The Singer

and the Scribe

European Ballad Traditions
and European Ballad Cultures
 75
Internationale Forschungen zur
Allgemeinen und
Vergleichenden Literaturwissenschaft

In Verbindung mit

Dietrich Briesemeister (Friedrich Schiller-Universität Jena) - Guillaume van

Gemert (Universiteit Nijmegen) - Joachim Knape (Universität Tübingen) - Klaus Ley
(Johannes Gutenberg-Universität Mainz) - John A. McCarthy (Vanderbilt University)
- Manfred Pfister (Freie Universität Berlin) - Sven H. Rossel (University of
Washington) - Azade Seyhan (Bryn Mawr College) - Horst Thomé (Universität Kiel)

herausgegeben von

Alberto Martino
(Universität Wien)

Redakteure:
Norbert Bachleitner & Alfred Noe

Anschrift der Redaktion:

Institut für Vergleichende Literaturwissenschaft, Berggasse 11/5, A-1090 Wien
 The Singer
and the Scribe
European Ballad Traditions
and European Ballad Cultures

Edited by
Philip E. Bennett
and
Richard Firth Green

Amsterdam - New York, NY 2004

Le papier sur lequel le présent ouvrage est imprimé remplit les prescriptions
de “ISO 9706:1994, Information et documentation - Papier pour documents -
Prescriptions pour la permanence”.

The paper on which this book is printed meets the requirements of “ ISO
9706:1994, Information and documentation - Paper for documents - Requirements
for permanence”.

ISBN: 90-420-1851-8
©Editions Rodopi B.V., Amsterdam - New York, NY 2004
Printed in The Netherlands
Acknowledgements

This volume of essays began as a colloquium of the Medieval and Early Renaissance
Studies Programme of the University of Edinburgh, held on Saturday 27th May, 2000, at
the Robertson Music Centre, St George’s School, Edinburgh. One paper, read on that day
but not included in the volume at the request of the speaker, was given by Dr Mary-Ann
Constantine of the Department of Welsh, University of Wales, Aberystwyth. The other
speakers at the colloquium are marked by an asterisk in the list of contibutors at the end of
the volume.
My thanks go first to my fellow organisers of the colloquium, Dr Roger Collins,
Department of History, and Dr Fran Colman, Department of English Language, University
of Edinburgh. I also wish to thank the Research Committee of the Faculty Group (Arts,
Divinity, Music) of the University of Edinburgh for financial support, and St George’s
School for donating the venue and providing generous hospitality. Finally my thanks are
due to Joanne Naysmith, Administrative Secretary in the Division of European Languages
and Cultures, University of Edinburgh, for her great patience and meticulous work in
preparing the colloquium and the volume that has proceeded from it, and to Thomas
Nicholas at the Ohio State University for overseeing the presentation of the final manuscript
copy.
Philip E. Bennett, February 2004
This page intentionally left blank
Contents

Philip E. Bennett and Introduction 5

Richard F. Green
Roderick Beaton Balladry in the Medieval Greek World 13
Ekaterina Love Story or Heroic Deed? (The Two Faces of 23
Rogatchevskaia Russian Balladry: Bylinas and Ballads)
Huw Lewis From Oral Adventure Story to Literary Tale of 37
Enchantment: the case of the Count Arnaldos ballad
Manuel da Costa Fontes A Morte do Rei D. Fernando and Floresvento: two 51
rare Portuguese epic ballads
Ad Putter Fier Margrietken: a medieval ballad and its history 69
William Layher Looking up at ‘Holger Dansk og Burmand’ (DgF 89
30)
Philip E. Bennett The Suppression of a Ballad Culture: the enigma of 105
medieval France
Richard Firth Green F.J. Child and Mikail Bakhtin 123
Charles Duffin Echoes of Authority: audience and formula in the 135
Scots ballad text.
Margaret Sleeman Estrea Aelion, Salonica Sephardic Tradition and the 153
Ballad of Imprisoned Virgil
Roger Wright Spanish Ballads in a Changing World 169
Thomas A. McKean The Stewarts of Fetterangus and Literate Oral 181
Tradition
List of Contributors 209

Index 211
This page intentionally left blank
 Introduction

Philip E. Bennett Richard Firth Green

The purpose of this volume is not to provide a comprehensive history of the European
ballad, nor to offer a complete comparative typology. It is rather to investigate the various
ways in which literate and non-literate, or para -literate, cultures have interacted in various
parts of Europe (and the New World as an extension of European culture by migration) to
shape the development of the ballad as a text for reading and a poem for singing from what
may be debatable origins in the late Middle Ages to the last years of the twentieth century.
Coverage, even within this limited framework, is not meant to be exhaustive. Celtic
balladry, whether Brythonic or Goedelic, is omitted, as is High German. The main thrusts
of the volume are towards the Iberian world, Hispanic and Portuguese, and the Anglo-Scots
traditions, with supporting studies of Dutch, French, Greek, Norse and Russian ballad
traditions.
A central problem for all who write about ballads is to arrive at a workable definition of
the object being studied, for without such a definition it is difficult to see how any kind of
external history can be undertaken at all. Nonetheless, many have refused to give abstract,
theoretical definitions. Typical of the ‘connoisseurship’ approach is that of Sir Arthur
Quiller-Couch in his introduction to The Oxford Book of Ballads where, before giving his
own selection of illustrative quotations from poems included in the anthology, he first
quotes W.P. Ker:

If the reader interrogate me concerning this Idea of the Ballad, as Mr Pecksniff demanded of Mrs Todgers her
Notion of a Wooden Leg, Professor Ker has my answer prepared: — ‘In spite of Socrates and his logic we
may venture to say, in answer to the question “What is a ballad?” — “A Ballad is The Milldams of Binnorie
and Sir Patrick Spens and The Douglas Tragedy and Lord Randal and Childe Maurice, and things of that
sort”’. 1

This notion, that the ballad cannot be defined, merely identified by an instinct born of
experience, is, of course, quite inadequate, and more recent scholars have tried to establish
the contours of the genre. Their efforts, however, which are frequently more prescriptive
than descriptive, are usually based on English and Scots traditions, or, where a broader
international view is taken, using the Anglo-Scots ballads as a yardstick. This has been
notably true of English-speaking scholars from either side of the Atlantic: Albert B.
Friedman’s excellent attempts to pin down the genre tended to err in that direction,2 and
even William J. Entwistle, still the only anglophone scholar to attempt a comprehensive
study of European balladry, began his chapter ‘What is a Ballad?’ by referring to Sir

1
The Oxford Book of Ballads, chosen and edited by Sir Arthur Quiller-Couch (Oxford: Clarendon Press, 1910;
repr. 1963), pp. xiii-xv.
2
Friedman, Albert B., The Ballad Revival, studies of the influence of popular on sophisticated poetry (Chicago;
London: University of Chicago Press, 1961), p. 6 et pass.; Princeton Encyclopedia of Poetry and Poetics, ed.
by Alex Preminger (Princeton: Princeton University Press, 2nd ed. 1974), pp. 62-64.
The Singer and the Scribe

Patrick Spens.3 Entwistle’s final working definition, however, is far broader than anything
that the study of Anglo-Scots conventions in isolation might mandate: ‘any short traditional
narrative poem sung, with or without accompaniment of dance, in assemblies of the
people’.4
Central to almost all definitions is this notion of ‘tradition’, with a consequent
opposition between ‘traditional’ (understood as ‘popular’ as a synonym of ‘folk’) and
‘literary’ ballads, the latter almost invariably being dismissed as aesthetically inferior to the
former as products of balladry. This view is particularly enshrined in Gerould’s otherwise
seminal study of what is significantly termed the ‘ballad of tradition’, in which qualitative
hierarchy is established between the products of a rural, northern, anonymous ‘folk’ and
those of southern, urban ‘rhymesters’, including the authors of broadside ballads, which
otherwise are seen as a crucial link in the transmission of the popular ballad.5 Yet even such
apparently fundamental sociological distinctions are not universally accepted as a basis for
aesthetic judgment. In the Hispanic tradition, for example (despite the habitual division of
the romancero into ‘ancient’ or ‘traditional’ and ‘artistic’ branches), they are simply
unknown.6 Indeed, though modern scholars, such as Roger Wright in his article in this
volume, may distinguish between folk romances and the products of literate poets from
Góngora to Lorca, the habit among Spaniards has been to treat the romancero as a unified
cultural good. This can clearly be seen in the nineteenth-century Tesoro de los romanceros,
in which, for example, anonymous romances dealing with the life of the Cid are freely
mixed with those on the same topic composed by Lope de Vega and Lorenzo de
Sepúlveda,7 the same mixture being apparent in all the other thematic sections into which
the volume is divided.
The question of nomenclature, and of making sure we are comparing like with like, is
one of the thorniest in ballad scholarship, and one which this volume highlights by its range
of studies, without seeking to resolve the issue. A central problem is caused by the adoption
in English in the eighteenth century of the word ‘ballad’ (which originally denoted a dance-
song) as the name for a particular type of closely focused, dramatised and impersonal
narrative poem, 8 which has caused much controversy over the boundaries separating the

3
Entwistle William J., European Balladry (Oxford: Clarendon Press, 1939), p. 16.
4
Entwistle, European Balladry, pp. 16-17.
5
Gerould, Gordon Hall, The Ballad of Tradition (Oxford: Clarendon Press, 1932), pp. 115, 241- 43.
6
Even ways of classifying the romances vary from authority to authority: Colin Smith links the romances viejos
and romances juglarescos as traditional types against ‘diverse kinds of romance aritificioso’, Spanish Ballads,
ed. by Colin Smith (Oxford: Pergamon Press, 1964), p. 20, while Dorothy Clotelle Clarke divorces the
romances juglarescos from the anonymous, traditional romancero viejo associating them with romances
eruditos and artisticos by individual poets, whether identifiable by individual name or not, Princeton
Encyclopedia of Poetry and Poetics, pp. 712-13.
7
Tesoro de los romanceros y cancioneros españoles, históricos, caballerescos, moriscos y otros, ed. by Don
Eugenio de Ochoa (Paris: Baudry, 1838), pp. 128-210.
8
The definition is first given by Gerould, The Ballad of Tradition, p. 192, and taken up by Friedman, Princeton
Encyclopedia of Poetry and Poetics, loc. cit.; for the dates of the use of ‘ballad’ to describe narrative poems
see Gerould, The Ballad of Tradition, pp. 235 and 250.

6
 Introduction

ballad from other types of traditional song on the one hand and from narrative romances
and epics on the other.
The most sensitive area is undoubtedly in defining the very permeable boundary
between ballad and folksong in general, although this boundary is seen as vital by David
Fowler in establishing his thesis that the Anglo-Scots ballad emerged only in the fifteenth
century by the conflation of metrical romances with folk lyrics (carols, riddling songs and
so forth).9 The porous nature of this border, if one takes a comparative international view, is
nowhere better illustrated than by the lack of a clear taxonymy in French, where the
chanson populaire (‘folksong’) is not subject to further classification. At various times the
‘ballad’ has been dubbed romance (borrowing the Spanish word), complainte and chanson
épico-narrative. This lack of clarity in isolating a type of poetry in the French tradition,
although the existence of the kind from at least the seventeenth century is regularly
admitted, has led to the paradoxical position, as Philip Bennett’s contribution seeks to
show, that the ballad is often held to be unknown in France at just that period, the thirteenth
to fifteenth centuries, when ballads, frequently exploiting French cultural materials, are
seen to be developing across much of the rest of western Europe.
An equally telling fluidity on the ballad-folksong border can be seen in the case of the
song type of the Baffled Knight, which tells the story of how a young lady or princess
preserves her virginity when meeting a knight in the country by assuring him she will marry
him if he escorts her home, then escaping from him and mocking him for his lack of sexual
enterprise. Fowler refuses to see this as a ballad, equating it with riddling songs or with
‘light romantic and comic pieces’ dealing with the battle of the sexes,10 yet essentially the
same scenario is presented by the Spanish Romance de la Infantina, with the exception that
in the Spanish version the knight pronounces satiric condemnation against himself for not
striking while the iron was hot.11 The Count Arnaldos Ballad, analysed in this volume by
Huw Lewis, offers another Iberian instance of the way the lyric impulse can be inextricably
woven into a ballad’s narrative structure, while even in the more austere Anglo-Scottish
tradition, as the association of the haunting lyric ‘O Waly, Waly,’ with the ballad of Jamie
Douglas shows, such interweaving is not entirely unknown. Tristram P. Coffin and Francis
Gummere (the latter famous for the now-discredited theory of the communal genesis of the
ballad) could hardly be further apart in their views of ballad origins, yet both would agree
on the primacy what Coffin calls ‘the emotional core’ of the ballad over its narrative
components.12

9
Fowler, David C., A Literary History of the Popular Ballad (Durham NC: Duke University Press, 1968). pp. 18-
19. Despite his statement in these pages that the popular ballad crystalised in Britain in the fifteenth century
Fowler’s insistance on using only manuscript evidence for the existence of ballads, his belief ‘that a given
ballad took the particular shape it has about the time it was written down, unless there is specific evidence to
the contrary’ (p. 5), and his rejection from the canon of all the earliest surviving poems normally considered to
be ballads effectively mean that the earliest ballads available for study according to him belong to the
sixteenth century.
10.
Fowler, A Literary History of the Popular Ballad, pp. 20; 9-30.
11
Romancero viejo, ed. by María de los Hitos Hurtado (Madrid: Edaf, 1997), pp. 165-67.
12
The Critics and the Ballad, ed. MacEdward Leach and Tristram P. Coffin (Carbondale: Southern Illinois
University Press, 1961), p. 246.

7
The Singer and the Scribe

Even on the other boundary, that on which the ballad approaches such traditional
narrative forms as the epic and the romance, Fowler sees what he calls a ‘rounded tune’ (p.
89) as being of utmost importance in defining ballads.13 Both Gerould, in studying the
English tradition, and Davenson, in analysing French chansons populaires, note the
importance of Gregorian chant as the foundation of ballad melodies,14 though due caution
should be applied before ascribing universal characteristics to ballad tunes. Roger Wright,
in this volume, indicates the wide range of musical styles adopted in different parts of the
Hispanic world for singing romances, some much more elaborate than others, and while
Ekaterina Rogatchevskaia points out that the nature of the supporting music is one feature
distinguishing the Russian ballada from the bylina, the latter, which, as Rogatchevskaia
argues in her contribution, should really be equated with epics, have frequently been
considered as belonging to the same tradition as western ballads.15
If the context of oral tradition, as a living if unconscious concept, and oral transmission,
as a mechanism of diffusion, have been vital to the ballad’s survival as a distinctive poetic
kind, such transmission has been practised throughout the whole period of the existence of
the ballad as a recognisable genre in a milieu which is essentially literate. Since ballads are
generally considered to have been sung and disseminated by an unlettered ‘folk’ this
paradox should be explored.
Romantic communalist theories of the origins not only of ballads but also of epics
posited, usually implicitly rather than explicitly, a state of society that would equate with
the Germanic tribes of the migration period or pre-Christian Celtic peoples not in contact
with Rome. The importance of Rousseau’s notions of the pre-societal Golden Age and of
Enlightenment theories of the Noble Savage should not be underestimated in the generation
of traditionalism as an explanation of ballad and epic origins, yet western Europe in the
fourteenth and fifteenth centuries, if we accept this as the likely period of the emergence of
the ballad in most of the areas covered by this book, is far removed from any society
isolated from the culture of the book. Whatever role is assigned to clerics, minstrels and
jongleurs / juglares in the composition and performance of ballads, both they and their

13
Fowler, A Literary History of the Popular Ballad, p. 80. This can be taken too far, however, as when he
excludes all the Robin Hood material from his view of the corpus of ballads because the poems refer to
themselves as ‘talkynges’. It should be noted that Middle English vocabulary is as imprecise in this sphere as
is Old and Middle French. The Middle English Dictionary lists ‘talkinge’ as a ‘tale’ while ‘song’ is glossed as
a ‘poem’: in neither case is the mode of delivery specified. Equally ‘singen’ has the meaning ‘to recite’ as well
as ‘to sing’, and ‘talken’ is associated with ‘dytees’ glossed as ‘songs’, with examples offering the Latin
equivalent ‘carmina’ (Middle English Dictionary, consulted at http://ets.udml.umich.edu/m/med/). Nor need
the comparative length of a ballad like The Gest of Robin Hood mean that it was invariably recited; The
Hunting of the Cheviot is also a substantial narrative, yet Sir Philip Sydney famously reports hearing it sung
by a ‘blinde Crowder’. For these reasons, it would be hazardous to conclude that Robin Hood poems were not
sung, or at least chanted, and to deny them the status of ballads.
14
Gerould, The Ballad of Tradition, pp. 216-23; Davenson, Henri (Henri-Irénée Marrou), Le Livre des chansons,
introduction à la chanson populaire française, Les Cahiers du Rhône (Neuchâtel: Editions de la Baconnière,
1944; repr n.p. [Paris]: Club des Libraires de France, 1958), pp. 65-69.
15
For example by Entwistle, European Balladry, pp. 354-80, though he does distinguish them from most western
European ballads.

8
 Introduction

audience, as well as the non-professional or ‘semi-professional’ singers equally active in the

tradition, cannot have failed to be aware of the whole range of texts, ecclesiastical, legal
and lay, which, while being diffused orally to most receivers, proclaim by their very
existence the centrality of the written word with its inherent if relative stability and
authority.16 On the one hand modern definitions of literacy, which often apply as much to
writing as to reading and include criteria such as the ability to sign one’s name, are not
applicable in a medieval context.17 On the other, in this case (as in so many involving
arguments over the oral dissemination of literature in the medieval and early modern
worlds) what is more important than the ability to decipher a greater or smaller amount of
written text is a mind set which is aware of the existence and importance of books.
This is to be sensed even in some of the very earliest ballads, which often seem to
depend upon more prestigious medieval narrative genres. For Fowler the dominant genre is
the romance, but this is another area where a broader European outlook will tend to
complicate assumptions based on a narrower Anglo-Scottish perspective. While Fowler is
not alone in assuming that early ballads represent a popularising of aristocratic medieval
romances,18 several of the essays in this collection serve to challenge this idea, and rather
than positing some kind of generic attraction between the two forms, we might do better to
think of the ballad as taking colouring from whatever prominent literary types happen to
coexist with it. In Russia, as Ekaterina Rogatchevskaia shows, national epics were
collected and written down comparatively late, so it is perhaps unsurprising to find Russian
ballads characterised not as popular romances but as ‘lower epic songs’. To make a similar
observations about the birthplace of Homer and Hesiod might seem absurd, but the epic
‘demotiko tragoudi’ described here by Roderick Beaton are similarly late, the product of
Greece’s heroic struggle for independence. Unlike Rusian ballads with their frequently
nameless heroes, however, Greek akritika may celebrate national heroes like Basil Digenes
Akrites. Similarly the great Scandinavian hero, Ogier the Dane, was being celebrated in
ballads from late fifteenth-century Sweden, whence, as William Layher shows, he found his
way into a fascinating church mural. If Swedish audiences could be entertained by stories of
Holger Dansk, and Dutch ones, as Ad Putter reminds us, by tales of the Germanic hero
Hildebrand, Portuguese audiences were apparently no less receptive to tales of the Spanish
Cid, and Manuel da Costa Fontes demonstrates how epic material from medieval Iberia has
been preserved down modern times in ballads from the Portuguese-speaking world. Even in

16
Gerould, The Ballad of Tradition, pp. 220–22; Friedman, Albert B., The Ballad Revival: studies in the influence
of popular on sophisticated poetry (University of Chicago Press, 1961) studies the interaction of literate and
non-literate or para-literate forms from the seventeenth to the twentieth century. In this volume Roger Wright,
Margaret Sleeman and Thomas McKean consider the interplay of the oral and the written in the repertoires of
twentieth-century singers, particularly in the case of those who sing predominently in the home or to a
restricted social circle and who do not perform to wider groups for any form of reward.
17
See Michael T. Clanchy, From Memory to written record, England 1066-1307, 2 nd ed. (Oxford: Blackwell,
1993), and Brian Stock, The implications of literacy : written language and models of interpretation in the
eleventh and twelfth centuries (Princeton, N.J. : Princeton University Press, 1983).
18
See Holger Nygard, ‘Popular Ballad and medieval Romance’, in Folklore International, ed. by D. K. Wilgus
(Hatboro, Pennsylvania: Folklore Associates, 1967), pp. 161-73; and Thomas J. Garbáty, ‘Rhyme, Romance,
Ballad, Burlesque, and the Confluence of Form’, in Fifteenth-Century Studies: Recent Essays, ed. by Robert F.
Yeager (Hamden, Conn.: Archon Books,1984), pp. 283-301.

9
The Singer and the Scribe

the Anglo-Scottish tradition, the influence of late medieval romance is far from
omnipresent. One would be hard pressed, for example, to categorise either the pro-Scottish
heroic ballad of the Battle of Otterburn, or its pro-English counterpart, the Hunting of the
Cheviot, as romances. No doubt the fact that these are early ballads from a remote and
lawless border region, helps explain their epic qualities, but if further evidence were needed
that no necessary generic connection exists between the ballad form and that of the literary
romance, we might cite the later dependence of the much-maligned early-modern broadside
on newspapers and chapbooks.
By the sixteenth century, when in the west all but the most remote communities had
their priest, lawyer and frequently schoolmaster, literary sensibilities had penetrated many
traditional communities. Such a community on the borders of Brittany is portrayed by Noël
du Fail in his Propos rustiques, a collection of ‘village tales’ representing local oral
tradition in the mid-sixteenth century. Notably the village has a former schoolmaster turned
wine maker who reads from his small library (the Shepherd’s Calendar, Aesop’s Fables,
the Romance of the Rose) on the same festive occasions as the villagers tell their tales.19 It
is also at this time that we find many ballads circulating, not for ‘court’ use but among what
would be regarded as traditional target audiences of singers, on broadside sheets in Britain
and in Spain on the equivalent pliegos sueltos.20 The parish of Myddle in Shropshire,
described by Richard Gough a century and a half after du Fail, gives an interesting insight
into the way such ballad communities might develop. Gough himself was an educated man,
able to quote Virgil and turn rustic epitaphs into Latin, but he was alert to the popular
culture of his own village. At one point he tells of a son of the parish who had gone to sea,
been captured by ‘the Turks of Tangiers’, and escaped only after a series of hair-raising
adventures. His fortunes temporarily restored, the man ‘came downe to Myddle, and was
there at what time they were singing ballads abroad in Markett townes of this adventure’.21
All of this has implications for our understanding of the ways in which oral traditions in
general and specifically ballad composition and singing as a sub-set of such traditions
operate in what should be termed para-literate communities. The pressures imposed by the
sense of the authority of the text, the sourcing of materials in the learned tradition and the
interplay between memorisation and formulaic re-composition of song materials cannot but
shape the transmission of the ballad.22 Equally significantly the acceptance that the oral
ballad tradition does not exist in vacuo only to be contaminated and killed by contact with
the literate has implications for the perception of the vitality of ballad production in the
modern world. The ballad ceases to be the ethnographic museum piece it is so often
regarded as being, and is re-instated as a living part of contemporary culture. In this volume
Margaret Sleeman’s study of the singing of Mrs Aelion from the London Sephardic
community, and Thomas McKean’s account of the multifold influences on the singing of

19
Du Fail, Noël, Propos rustiques, ed. by Gabriel-André Pérouse and Roger Dubuis, Textes Littéraires Français,
445 (Geneva: Droz, 1994), pp. 49–50.
20
Gerould, The Ballad of Tradition, ch. 9, ‘Ballads and Broadsides’, pp. 235–54; Smith, Spanish Ballads, pp. 18–
21.
21
Richard Gough, The History of Myddle, ed. David Hey (Harmondsworth: Penguin Books, 1981), p.115.
22
These phenomena have been studied by Finnegan, Ruth, Oral Poetry (Cambridge: Cambridge University Press,
1977), pp. 16-24; 246-62.

10
 Introduction

the Stewart family from Aberdeenshire reveal most comp ellingly the ways in which
tradition is kept alive, renewed and expanded even at the end of the twentieth and beginning
of the twenty-first century.
Although for increasingly obvious reasons ballads in all traditions are treated as literary
or at least as ethnographic artefacts, performance — singing or reciting —remains central to
the ballad as cultural phenomenon, and this perception informs a final group of essays in
this collection that seek to explore the inner history, as it were, of the ballad through a
reconstruction of the aesthetic demands made upon it by its audience. Charles Duffin gives
us the most theorised account of such an approach, using what he calls ‘the critical
authority of the non-literate audience’ as a yardstick against which to measure the formal
qualities of ballad narrative. On the other hand Richard Green’s attempt to gauge a thematic
pattern (in this case Bakhtin’s notion of the carnivalesque) with such a yardstick suggests
that the approach works better with form than content, ballad audiences proving too
disparate, both temporally and geographically, to be easily classified by any such reductive
typology. Ad Putter’s application of six of Axel Olrik’s laws of folk narrative to the
medieval Dutch ballad Fiere Margrietke offers a far better illustration of the way Duffin’s
oral aesthetics might play out in practice.
‘What allows the ballad to be told briefly,’ Putter writes, ‘without authorial explanation
regarding motivation, cause and effect, and so on, is its participation in a shared world of
understanding. The ballad is short because so much is understood’. In one way or another
all the studies in this book seek to reach across the bare and often enigmatic words of
recorded ballad texts to the shared experience of the audiences for which they were
composed and performed. It is here that a pan-European perspective can prove most
valuable, reminding us that just as there can never be one monolithic folk audience, so there
can be no typical ballad singer and no single reductive entity we may regard as an ur-ballad.
It is part of the perennial fascination of ballad scholarship that what is told and what is
understood must always vary from place to place and from era to era, yet such variety will
always occur within a set of recognis able parameters, shaped by clearly identifiable forces.
We hope that this volume has gone some way towards deepening the recognition of these
parameters and these forces.

11
This page intentionally left blank
 Balladry in the Medieval Greek World

Roderick Beaton

The term ‘ballad’ is a latecomer in the Greek language, imported in the second half of the
nineteenth century, and applied almost exclusively to foreign works and their conscious
literary imitation in Greek. But the nineteenth century was also the time of the discovery,
and dissemination for the first time in writing, of a thriving Greek oral tradition whose
roots, its first ardent promoters declared, went back hundreds and in some cases even
thousands of years.1
This is the tradition now known in Greek as demotiko tragoudi, a term equivalent to the
German Volkslied, the English ‘folksong’, and their cognates in other languages. Collected
in large numbers during the nineteenth century, these songs had become established by the
end of that century as the ‘unwritten history of the nation’, a unique — and uniquely Greek
guarantee of the continuity of language and culture through centuries of foreign oppression,
and also as a reference point for poets, journalists, politicians and almost every kind of
public utterance, comparable to the King James Bible or the plays of Shakespeare in
English.2 Most twentieth-century histories of Modern Greek literature begin with these folk
songs, assuming, rather than demonstrating, their historical primacy in the formation of a
modern Greek language, of collective moral and political values, and of a common
aesthetic. The story of the rediscovery and reception of Greek demotiko tragoudi between
about 1820 and the 1950s, runs closely parallel to the story of the reception of western
balladry which begins about half a century earlier.
Not all of the rich variety of song types encompassed by demotiko tragoudi can be
compared to western ballads. Greek folklorists at the turn of the last century established a
typology which despite its dubious historical premises, in some cases, is still in use today.
In the first place, about half of the collected corpus consists of lyrical songs, whose
prevalent themes are love, death and exile. These do not concern us here. Nor does the
sizeable category of songs devoted to the lives, and particularly to the violent deaths, of the
klefts, the brigands of the late eighteenth century who during the Greek war of
independence of the 1820s found themselves cast in the unlikely role of freedom-fighters
and as successors to the victors at Marathon and Salamis. Although some of their thematic
material and stock formulae may well be much older, these are narrative songs which
recount events historically placeable in the period 1750–1922 (when endemic brigandage
was finally, and officially, stamped out in Greece). Other songs again, such as the famous

1
This begins with the bilingual collection, in two volumes, with a long introduction, highly coloured with the
Romantic spirit of the time, by Fauriel, Claude, Chants populaires de la Grèce moderne (Paris: Firmin Didot,
1824-25).
2
See Beaton, Roderick, Folk Poetry of Modern Greece (Cambridge: Cambridge University Press, 1980), pp. 1-12;
for references to the major collections see pp. 213-20.
The Singer and the Scribe

account of the last mass in the cathedral of St Sophia before the fall of Constantinople to the
Ottomans on 29 May 1453, in its many variants, do take us back to the threshold of the
Middle Ages. But these songs are brief, and few in number. 3
The Greek equivalent of western balladry is to be found in two categories of song which
the folklorists have somewhat arbitrarily separated. These are known as akritika and
paraloyes.4 The thematic range and characteristic style of these two types of song provide
the best grounds for comparison with aspects of the medieval ballad tradition in the west.

Heroic and domestic ballads

The songs known conventionally as akritika have heroic themes. Their world is the male
domain outside the house. The hero is either male, or in an extreme case, a warrior-maiden,
an Amazon figure who takes on a male role. The protagonists in these songs fight. They do
little else. They fight alone, and always against enormous odds. If their enemies are
ordinary mortals, these come in thousands; if wild beasts, they are the fiercest: usually lions
and bears. Often these enemies are supernatural creatures, of either indeterminate shape or
with the ability to change shape at will: such as the drakos (a pleasant spoken young man
who can suddenly transform himself into St George’s fearsome adversary), or the deceitful
lamia (distant kin to the subject of Keats’s poem in ballad style).
The most memorable of all the hero’s adversaries, whom he encounters in a great many
variant versions from all over the Greek world, is Charos, the personification of death.
Charos in the Greek tradition is a grim old curmudgeon, in laments often ridiculed as a
bloody-minded functionary, hen-pecked by his wife or mother: the sort of petty bureaucrat
to whom all appeals are vain. In these heroic songs, when it is the hero’s turn to be
summoned by the black horseman, Charos, he refuses. Instead he pays his adversary the
compliment of treating him as an equal. The hero, whose attributes represent life at its most
vibrant, challenges his opposite, Death, to single-combat, unarmed, on the village threshing
floor. To emphasise the hardness of the struggle, this arena is not made of the usual beaten
earth, but of marble, bronze or iron.
This is the full text of a version transcribed in Crete and first published in 1876:
Eat and drink, sirs, and I shall tell you a tale.
I shall tell you a tale about a hero,
about a young man I saw out hunting on the plains,
he was hunting and coursing hares and after wild goats.
As he runs the young man grabs a hare, with a leap a wild goat,
the plumed partridge he leaves far behind.

3
Beaton, Folk Poetry, pp. 95-111.
4
Still the fullest definition and explanation for these terms is to be found in Kyriakidis, Stilpon, ‘Mnemeia tou
logou: asmata’, in idem, To demotiko tragoudi: synagoge meleton, ed. by A. Kyriakidou-Nestoros (Athens:
Ermis, 1978), pp. 61-65 (first published 1923), although many of the underlying assumptions have been
quietly dropped since the 1970s. For newer approaches to the subject see Loukatos, D., Eisagoge sten ellenike
laografia (Athens: Cultural Foundation of the National Bank [MIET], 1978); Sifakis, Grigoris, Gia mia
poietike tou ellenikou demotikou tragoudiou (Heraklion: University of Crete Press, 1988); Kapsomenos,
Eratosthenes, Demotiko tragoudimia diaforetike prosengese (Athens: Arsenides, 1990).

14
 Roderick Beaton

But Charos [Death] passed by and he was mad with jealousy.

‘Take off, young man, your clothes, take off your weapons,
bind your hands in the shape of the cross, for me to take your soul.’
‘I won’t take off my clothes, nor will I take off my weapons,
I won’t bind my hands in the shape of the cross, for you to take my soul.’
But you and me, as man to man, and both of us brave heroes,
come, let’s wrestle on the iron threshing floor,
so as not to shatter the mountains or lay waste the town.’

They went and wrestled on the iron threshing floor,

and nine times the youth threw Charos down
and on the ninth time Charos lost his temper.
He takes the young man by the hair, and pushes him to his knees.

‘Let go, Charos, of my hair and take me by the arms,

and then I’ll show you what pallikaria [brave young men] are made of.’
‘This is how I take them all: all the pallikaria ,
beautiful girls I take, and mature fighting men,
and infant children and their mothers too.5

So long as they fight on human terms, the hero keeps the upper hand: life is stronger than
death. Death wins his inevitable victory at the end, but only by cheating. Male heroism is
shown at its apogee in taking on the unequal struggle with death, and the strategy of these
songs makes the defeated hero the moral victor. These are perhaps the best known, and best
loved, of all Greek folk songs.
The other principal activity of the hero in the akritika is the abduction of a bride. Having
successfully abducted her, the hero has then to fight to prevent others carrying her off in
turn. There is a whole code of warfare around bride-snatching, and the corresponding
defensive activity, in these songs. Though a hero will perform extraordinary feats to protect
his home and his womenfolk, these acts are motivated not by a rosy picture of domestic
bliss, but by the imperative of preserving male honour intact.6
By contrast, the second group of narrative songs, conventionally termed in Greek
paraloyes, can be said to focus on domestic life, or at least on conflicts within, rather than
outside, the home. Here the primary role is often played by women. Some commentators
have emphasised the role of the supernatural in these songs, and their connection with
popular traditions and superstitions. In fact, the supernatural forms part of the familiar
world of almost all Greek folksongs. In the heroic songs too, horses and birds regularly talk;
the most memorable adversaries are mythical or legendary beings; the intervention of

5
Jeannaraki, A., Asmata kretika meta distichon kai paromion (Leipzig: 1876), p. 142; reprinted in Ioannou, G., Ta
demotika mas tragoudia (Athens: Tachydromos, 1966), p. 30 (the translation retains the tenses of the original).
There is no modern anthology of this group of songs, as there is for e.g. the paraloyes, but most of the
standard collections, from that of Politis, N.G, Eklogai apo ta tragoudia tou ellinikou laou (Athens: 1914),
onwards, include a section headed ‘akritika’.
6
See Mackridge, Peter, ‘None but the Brave Deserve the Fair: abduction, elopement, seduction and marriage in the
Escorial Digenes Akrites and modern Greek heroic songs’, in ‘Digenes Akrites’: new approaches to Byzantine
heroic poetry ed. by Roderick and David Ricks (Aldershot: Variorum, 1993), pp. 116-30.

15
The Singer and the Scribe

nature by supernatural means is taken for granted. A persuasive study by Margaret Alexiou
demonstrated that a group of the best known of these songs, featuring the supernatural, ring
the changes on a central conflict within the family group, namely that between blood
kinship and marriage kinship, between the centripetal demands of the nuclear family and
the centrifugal forces that threaten it through the need to find marriage alliances outside.7
The most celebrated of all the paraloyes — which already in the nineteenth century was
being compared to Burger’s Leonore and a range of Balkan oral variants — is called The
Song of the Dead Brother.8
A mother has nine sons and one daughter: there is some evidence that this was thought
of traditionally as the ideal family size and composition. A marriage suit for the daughter,
usually named Arete, comes from a faraway city. The brothers advise their mother against
acceptance, but the youngest, Kostantinos, is for the match. Kostantinos travels a lot; he
vows he will keep in touch and that, in extreme need, he will bring the daughter back to
visit her mother. So the daughter is married and leaves home. A plague then strikes, in
which the brothers, including Kostantinos, all die, and the mother is left alone ‘like stubble
on the plain’. She curses Kostantinos and reminds him of his vow. Such is the power of a
mother’s curse and of a solemn vow unfulfilled, that the dead Kostantinos rises from his
grave. Taking, in one version, a cloud for his horse, a star for his saddle, and the moon for
company, he rides off to bring his sister back. During the return journey, through a macabre
series of questions and answers with her brother, the unsuspecting Arete slowly discovers
the truth, that she is riding with a dead man. In keeping with the Greek tradition
Kostantinos is not an insubstantial ghost, that familiar figure from the northern European
ballad tradition and tales of the supernatural, but a physical corpse: a zombie or revenant,
for which the Greek term, from Serbian, is vrykolakas. In almost all versions the end is
tragic: Arete arrives home in time to die in the arms of her mother who dies at the same
moment, though in other versions she is transformed into a bird. As Alexiou reads this
song, an overemphasis on family ties at the expense of marriage ties leads to the destruction
of the entire family, since Arete, who would otherwise have survived the plague to prolong
the family line in her new home, is doomed by the embrace of her dead brother. 9
Another well-known song, whose theme runs along a rather different axis, is known in
Greek as the Bridge of Arta although among approximately three hundred collected
variants, several other locations for the bridge are mentioned. This is immediately
recognisable as a version of a tale-type, often called a ballad, found throughout southeast
Europe. 10

7
Alexiou, Margaret , ‘Sons, Wives and Mothers: reality and fantasy in some modern Greek ballads’, Journal of
Modern Greek Studies, 1 (1983), 73-111.
8
Politis, N.G., ‘To demotikon asma peri tou nekrou adelfou’, Deltion tes Istorikes kai Ethnologikes Etereias tes
Ellados, 6 (1885), 193-261; for a translated version see Alexiou, ‘Sons, Wives and Mothers’.
9
To demotiko tragoudi – paraloyes ed. by G. Ioannou (Athens: Ermis, 1970), pp. 31-43 (introduction and three
versions).
10
For worldwide variants see The Walled-Up Wife: A Casebook, ed. by A. Dundes (Madison: University of
Wisconsin Press, 1996); on the Greek versions: Megas, G. A. Die Ballade von der Arta -Brücke: eine
vergleichende Untersuchung (Thessaloniki: Institute for Balkan Studies, 1976).

16
 Roderick Beaton

In the Greek versions, a master-builder directs a large labour-force to build a bridge

across a river. The size of the labour-force suggests the high prestige of the project. But
what is built by day mysteriously falls down each night. The master-builder is in despair,
until the spirit of the river addresses him directly, and demands the sacrifice of the master-
builder’s wife, without which the bridge will never stand firm. The master-builder is in
despair, but finally consents. He tricks his wife into descending into the foundations of the
bridge, then orders his underlings to throw in stones and mortar after her. Some versions
end the story there, but in those which continue, the victim responds by placing a curse on
the bridge and all who cross it. But the master-builder reminds her that she has a brother
who may one day cross, and so she is persuaded, as she dies, to bless the work instead.11
In this, and many other songs, it is noticeable that the central role is played by a woman.
The Dead Brother is only one of many songs of this type in which the husband and father,
the traditional pillar of the patriarchal household, is not even mentioned. In The Bridge of
Arta, the man is shown as weak and vacillating; his prestige project is completed only
through the willing sacrifice of his wife. Here too, in the argument that it is her brother who
may one day cross the bridge, what the song ends by privileging is blood kinship over the
marriage relationship.
Stylistically, there is no difference between the two types of song, akritika and
paraloyes. Length is typically between about thirty and about seventy lines, with the longest
reaching a little over a hundred lines. Rhyme is unknown in Greek until the fourteenth
century, when it enters literary poetry in Crete through close cultural contact with Venice.
In most types of oral narrative poetry, rhyme has always been the exception.12 The
prevalent metrical form is the fifteen-syllable iambic verse, known to Byzantine literature
by the name of politikos, a term which probably means ‘ordinary’, ‘everyday’. In narrative
songs the fifteen-syllable verse is almost always end-stopped, has an obligatory caesura
after the eighth syllable, and a fixed stress-pattern at the end of each half-line: masculine at
the end of the first hemistich, feminine at the end of the second. This comes very close to
the English and Scottish ballad rhythm:
The King sits in Dunfermline town
Drinking the blood-red wine...,

with the addition of an obligatory, unstressed syllable at the end of the second line (or half-
line), which fits the characteristic stress-pattern of the language.13
Narrative style is generally abrupt, sometimes almost telegraphic, with a high proportion
of dialogue and frequent switching of tenses. These songs have generally been praised for
their immediacy and striking imagery. Events follow one another without causal links.
When sung (though we have no sound recordings from before 1930, and most date from
after the 1950s), the melody often demands repetition of all or part of a metrical line, and it

11
Paraloyes, ed. cit., pp. 44-8 (introduction and two versions); for discussion in English see Beaton, Folk Poetry,
pp. 120-4 (reprinted in The Walled-Up Wife, pp. 63-71).
12
Beaton, Roderick, The Medieval Greek Romance (London: Routledge, 2nd edition, 1996), pp. 199-206; idem,
Folk Poetry, pp. 148; 151-92.
13
Alexiou, Margaret, and David Holton, ‘The Origins and Development of Politikos Stichos: a select critical
bibliography’, Mantatoforos, 9 (1976), 1- 40; Beaton, Medieval Greek Romance, pp. 98-100.

17
The Singer and the Scribe

is common in some regions for a musical phrase to encompass a line together with part of
the line which follows, which may be broken off in mid-word, and then started again from
the beginning. The musicologist Samuel Baud-Bovy gave this technique the name strophe
cleftique, and it may be peculiar to the Greek tradition.14

Historical context
Both the akritika and paraloyes have been assumed, sometimes on very shaky grounds, to
be of great antiquity, and their supposed historical origins are even enshrined in the
(conventional) names by which they have been baptised by Greek folklorists. Akritika are
understood to be songs belonging to the epic cycle surrounding the medieval Greek hero
Digenes Akrites, whose exploits were sung on the eastern frontier of the Byzantine empire
before the twelfth century. Paraloyes, in the first half of the twentieth century, were
systematically compared to the plots of ancient Greek tragedy, from which they are
supposed to be directly descended through the popular mime performances of late antiquity.
The word paraloye, attested in only a single song text recorded in the late nineteenth
century, is said to corroborate this origin, being a shortened form of parakataloge, a term
which may have denoted some such theatrical performance in the fourth century AD.
Although studies of Greek oral tradition are no longer primarily directed to affirming
the antiquity of motifs, or to shoring up national claims to cultural continuity from a remote
past, these designations are still in use and these definitions are still repeated in standard
anthologies and textbooks.15
The alleged connection of the paraloyes with ancient Greek theatre does nothing to help
us understand either the songs or their historical background. Similarly, the term paraloye,
although now an accepted part of the language, has no real authority within the song
tradition itself, and certainly does not prove the continuity or derivation that its adoption
presupposes. This is not to say, of course, that the tradition from which these songs derive
may not be very old. But the specific connection with ancient tragedy is no more than a
distraction.
With the akritika we are on slightly safer ground. The exploits of Basil Digenes Akrites
were recorded in the same fifteen-syllable iambic metre as the heroic songs in a manuscript
from South Italy whose most recent editor dates it to around 1150. The same story is told in
no fewer than five further manuscripts of the fifteenth century. One of these, in a markedly
more popular style, it has been vigorously proposed in recent years, preserves a version
datable to the early twelfth century. This epic tale (or perhaps epic manqué) was
rediscovered in 1869 and first appeared in print in 1875, just at the time when folklorists
were active in recording and collating songs from the oral tradition. The names Digenes and
Akrites (or Akritas) also occur in the heroic songs, although not very often and never

14
Baud-Bovy, Samuel, ‘La strophe des distiques rimés dans la chanson grecque’, in Studia Memoriae Belae
Bartok Sacra (Budapest: 1956), pp. 365-83.
15
See note 4 above.

18
 Roderick Beaton

together. But this was sufficient reason for the whole group of heroic songs to be named
akritika, and once again the designation has stuck.16
It is not, however, only the names that link the heroic songs with the Byzantine epic.
The medieval poem, in all its versions, extols very much the same code of heroic behaviour
as we find in the songs: bride-snatching takes up more than half the story; single combat
against enormous armies, against a warrior-maiden, against wild beasts, and a drakos that
can change its shape, account for much of the rest. At the end, the hero, still young, faces
the one adversary he cannot overcome, death. Disappointingly, however, there is no trace,
in any of the medieval versions, of the single combat against Charos, the personification of
Death, which in the song tradition marks the high point of the ideal of male heroism.
In addition to this common ground with the written epic of the twelfth century, the
akritika preserve some historical memory of combat on a remote eastern frontier. The
Euphrates is a common landmark, warriors fight on horseback wielding sword and mace,
and many songs preserve elements of Byzantine military vocabulary which have elsewhere
dropped out of use. There is little doubt that some, at least, of these songs do preserve
memories of Byzantine warfare and, probably, of the male heroic code as it may have
existed between the eighth and the eleventh centuries, at a time when Byzantine power
extended throughout Anatolia and into Mesopotamia. Corroboration for this can be found in
the Byzantine epic of Digenes Akrites, which probably acquired written form in or about the
early twelfth century and is likely to derive from an earlier form of this same tradition.
There is also a much shorter heroic narrative song, preserved in two manuscripts of the
fifteenth century, the song of Armoures, which confirms the existence of verse narratives of
this type, at least by that period.17
It must be emphasised, however, that with the exception of its probable influence on the
written Digenes Akrites, and of the single ballad-type text of the fifteenth century, the
tradition of Greek oral narrative songs during the Middle Ages has left no direct evidence of
its existence. Manuscript evidence from the seventeenth century shows all the recognisable
features of the more recent tradition fully developed by then; and, of course, the wealth of
material transcribed from oral tradition, beginning in the early nineteenth century, is
without question the product of a long, but not otherwise specifiable, oral tradition.

Oral and literary traditions

We do, however, have evidence that songs similar to the ones we know from later periods
did exist in the Greek world from at least the twelfth century. Consideration of this
evidence raises the question central to the present volume: the relation between ‘the singer
and the scribe’.

16
The best way to approach this poem and the scholarly controversy that has raged since the 1870s is through the
edition of the two most important manuscripts by Elizabeth Jeffreys: see Digenis Akritis: the Grottaferrata
and Escorial Versions, ed. and trans by Elizabeth Jeffreys, Cambridge Medieval Classics, 7 (Cambridge:
Cambridge University Press, 1998). See also the essays collected in Digenes Akrites, ed. by Beaton and
Ricks).
17
Included in the edit ion of the E version of Digenes Akrites (in Greek only): Alexiou, Stylianos, Vasilios Digenes
Akrites kai ta asmata tou Armoure kai tou Yiou tou Andronikou (Athens: Ermis 1990).

19
The Singer and the Scribe

The Modern Greek vernacular makes its first appearance in Greek literature during the
twelfth century. Leaving aside the evidence of Digenes Akrites, whose earliest written form
is still much disputed, we find the vernacular used in a small number of court poems of the
mid-twelfth century.18 These bear affinities with goliardic poetry in the west, and are almost
certainly at least partly parodic in intent. But they do use the same verse-form as we find in
the later song tradition, notably with the mid-line caesura and the stylistic tendency to build
a line of out of two parallel or contrasting half-lines. One of the group of poems known as
Ptochoprodromika (Poems of Poor Prodromos) makes a parodic allusion to ‘Akrites’ as the
epitome of a hero; another, perhaps more tellingly, parodies the style of heroic single
combat with the mace, in a scene where the hen-pecked husband is shut out of the kitchen
and does battle with his wife using a broom-handle! Although the source here might be an
already-written version of the exploits of Digenes, the style and some of the vocabulary are
close enough to those of the later recorded akritika to make it virtually certain that oral
songs of this type, in this verse-form and with this subject-matter, were in circulation in
Constantinople in the mid-twelfth century.19
Later, during the fourteenth and fifteenth centuries, narrative poetry in the vernacular
becomes relatively well-developed in Greek. Many of these texts, in length between a few
hundred and twelve thousand lines, belong to the genre of the love-romance; others are
versified chronicles, more or less fictionalised history (the Tale of Belisarius), or allegorical
tales of the natural world. Once again, the standard verse-form is the fifteen-syllable line
with its fixed caesura; and from the fourteenth century onwards we increasingly find
recognisable phrases, images, and sometimes even whole lines, that will turn up again many
centuries later in songs transcribed from the oral tradition. Since there is, with very few
exceptions, no trace of the subject matter of these narratives in the later ora l tradition, the
likeliest explanation for this is that these stock elements and memorable images were taken
up by literary writers from the oral tradition, rather than that originally literary
compositions later bequeathed these — but only these — elements to the oral tradition. This
picture is complicated, however, by the fact that we find evidence, increasingly, for popular
dissemination of originally literary works from the fifteenth century onwards, and some
literary texts, such as the romance Erotokritos, written in Crete at the end of the sixteenth
century, subsequently enter the oral tradition and take on a life of their own, down to the
twentieth century. None of these, however, seems ever to have become fully assimilated
into the tradition of either the akritika or the paraloyes.20

18
Beaton, Medieval Greek Romance, pp. 13-15.
19
Hesseling, D.C., and H. Pernot, Poèmes prodromiques en grec vulgaire (Amsterdam: 1910), see Poem I, lines
155-97; Poem 3, lines 164-73; Eideneier, Hans, Ptochoprodromos, Neograeca Medii Aevi, 5 (Köln:
Romiosini, 1991) with German translations: see Poem I, lines 155-81; Poem IV, lines 189-544.
20
See Beaton, Medieval Greek Romance, pp. 164-206; idem, ‘Orality and the Reception of Late Byzantine
Vernacular Literature’, Byzantine and Modern Greek Studies, 14 (1990) 174-84; idem, ‘Les fortunes de
Digénis Akritis: de l’épopée médiévale au symbole du nationalisme grec’, in Formes modernes de la poésie
épique: nouvelles approches, ed. by Judith Labarthe-Postel (Colloque, Université de Tours, 25-27 April 2002:
forthcoming).

20
 Roderick Beaton

Conclusion
The evidence for balladry in the medieval Greek world is, then, indirect, but none the less
strong. The rich tradition of short oral narratives in verse, represented by collections made
since 1820 has almost certainly been in existence, and in some degree of coexistence with
written literature since at least the twelfth century. This is not to claim that individual song-
texts can be precisely dated. The oral tradition, though conservative, can be assumed never
to have been static. But it seems most probable that a tradition closely equivalent to the
ballad tradition in the West formed part of the cultural horizon of the later centuries of the
Byzantine Empire, and also of the Greek-speaking lands under western rule after the Fourth
Crusade of 1204.
Even if we cannot hope to recover the earlier medieval context in which these songs
took shape, they still offer a great deal both to admire as poetry and to set alongside the
better documented traditions of other parts of Europe.

21
This page intentionally left blank
 Love Story or Heroic Deed?

(The Two Faces of Russian Balladry: Bylinas and Ballads)

Ekaterina Rogatchevskaia

Aroused by F.J. Child’s publication of English and Scottish ballads public interest in and
scholars’ attention towards ballads have never faded.1 The first collection of Russian folk
songs was compiled in the eighteenth century and published only in 1804,2 although the
collecting of folk songs on a regular basis started in the middle of the nineteenth century.3
In the early anthologies Russian folk songs were not classified systematically and their
genres were not specified. Most of the published pieces were called songs, but later the term
bylina4 was also adopted, although this did not correspond to how singers would define the
songs they had been performing. Traditionally, all narrative songs were called stariny
(which derives from the word ‘staryi’ — ‘old’ and means something that happened a long
time ago) to distinguish them from purely lyrical and ritual songs. Since some folk pieces
did not seem to fall into the category of bylinas (an epic song with a strong heroic and
national message) the word ‘ballad’ started to appear in specialist literature on folklore. It
was P.V. Kireevskii, a nineteenth-century Russian scholar, who recognised the difference
between bylinas and ‘other songs’ and published them in a separate section in his volumes.5
A.I. Sobolevskii also distinguished between the two genres and coined the term ‘lower epic
songs’ applying it to those songs which are now commo nly acknowledged as ballads.6
While the word ‘ballad’ was internationally accepted as a term for ‘a number of different

1
See, e.g., these classical works on balladry: Gummere. F.B., The Popula r Ballad (London-Boston-New York:
Archibald Constable & Co. Ltd and Houghton, Mifflin & Co, 1907); Hart, W.M., Ballad and Epic. A Study in
the Development of the Narrative Art (Boston: Ginn & Co, 1907), Cohen, H.L., The Ballade (New York:
Columbia University Press, 1915); Hustvedt, S.B, Ballad Book and Ballad Men (Cambridge, Mass: Harvard
University Press, 1930); Entwistle, W.J., European Balladry (Oxford: Clarendon Press, 1939). For more
information see Edson, Richmond, W., Ballad Scholarship: an annotated bibliography (New York-London:
Garland, 1989).
2
References will be to the editions of the anthologies available in the UK: Danilov, Kirsha, Drevnie rossiiskie
stikhotvoreniia, sobrannye Kirshei Danilovym (Moscow, 1958).
3
Kireevskii, P.V., Sobranie narodnykh pesen Kireevskogo, Vol. 1, 2. (Leningrad: Nauka, 1983); Gilferding, A.F.,
Onezhskie byliny, zapisannye A.F.Gilferdingom letom 1871 goda (Moscow: Izdatel’stvo Akademii nauk
SSSR, 1949-1951); Rybnikov, P.N., Pesni, sobrannye P.N.Rybnikovym (Moscow: Tipografiia A.Semina,
1861-1862 (vol. 1-2), Petrozavodsk, 1864 (Vol. 3), and St Petersburg, 1867 (Vol. 4)).
4
The word ‘bylina’ derives from the Russian verb ‘byt’ (to be) and has the meaning of ‘something which has
really happened’.
5
Kireevskii, P.V., Sobra nie narodnykh pesen Kireevskogo.
6
Sobolevskii, A.I., Velikorusskie narodnye pesni (St Petersburg, 1895).
The Singer and the Scribe

forms in music and poetry, each of which has as much right as any to be so called’, 7 it also
made its way into the Russian academic tradition probably with the help of western
Slavonic scholars for whom the very existence of balladry in the western Slavonic traditions
was not as questionable as in Russian and other eastern Slavonic cultures. The term was
therefore introduced and a collection of songs under the title Russkaia ballada (Russian
Ballads) was published by V.N. Chernyshev and N.P. Andreev in 1936. 8 Neither the
definition of the Russian ballad, nor its main characteristics were established by the authors.
The question remained debatable for decades until the 1950s and early 1960s when D.M.
Balashov made an outstanding contribution to the study of this genre, publishing his
pioneering collection of ballads with a detailed introductory article in which he explained
his understanding of the definition of the genre.9 Although with hindsight he can be
criticised for inconsistency and errors, contemporary scholars would often refer to his
works, which are commonly recognised as classical. The principal question which had to
be solved by slavists, then, was that of the definition and characteristics of the genre,
because the term ‘ballad’ was acquired form foreign traditions to name a phenomenon
which had never previously been called a ballad in Russian or eastern Slavonic folk
practice.
Strangely enough reading the literature on western European balladry we find similar
concerns and debates focused on the issue of genre. Those scholars who deal with western
balladry usually operate with definitions close to that once formulated by T. Pettitt: a ballad
is ‘a narrative song, current in popular tradition, which tells its story in a particular,
specified way’. 10 Over the years research has identified other fundamental features of
traditional ballads, such as the nature of their narrative, including the objectivity of
narration, verbal repetitions and commonplaces or formulae. 11 At the same time F.G.
Andersen points out that ‘in the absence of a clear-cut definition, it has been customary
simply to define ballads by example, as songs similar in kind to those in Child’s volumes.’ 12
Having discussed and analysed some other prominent features and elements which can be
exploited to characterise ballads, Andersen comes to the conclusion that ‘ballads are both
relatively easy to spot, in terms of narrative technique, and extremely difficult to
characterise, in terms of varying ways in which this technique is employed to express
distinct cultural meanings in different circumstances.’13

7
Gerould, G.H., The Ballad of Tradition (Oxford: Clarendon Press, 1932), p. vii.
8
Russkaia ballada, ed. by V. I. Chernyshev, with an introduction by N.P. Andreev (Leningrad: Sovetskii pisatel’,
1936).
9
Balashov, D.M., Narodnye ballady (Moscow-Leningrad: Sovetskii pisatel’, 1963).
10
Pettitt, T, ‘Introduction. Ballads as Narrative’, in The Ballad as Narrative, ed. by F.G. Andersen, O. Holzapfel,
T. Pettitt (Odense University Press, 1982), p.1.
11
In recent years V. Propp’s talerole analysis was also applied to ballads. This attempt could also be seen in the
context of solving the problem of the definition of the genre. See, for example, Buchan, D., ‘Talerole Analysis
and Child’s Supernatural Ballads’, in The Ballad and Oral Literature, ed. by J. Harris (Harvard University
Press, 1991), pp. 60-77.
12
Andersen, F.G., ‘Technique, Text, and the Context: formulaic narrative mode and the question of genre’ in The
Ballad and Oral Literature, p.19, which also gives an overview of the literature on this particular problem.
13
Andersen, F.G., ‘Technique, Text, and the Context’, p. 38.

24
 Ekaterina Rogatchevskaia

From this it becomes clear that one of the key issues of definition is the so-called
‘cultural meaning’ or ‘cultural context’ that shapes ballads and influences their narrative
techniques. Since oral traditional cultures of any nation preserve more of the original
national features than written cultures, the use of similar terms applied to different pieces of
folklore can be problematic. 14 Therefore, it can be productive to examine what Slavonic
researchers and slavists understand by a ‘ballad’ and determine whether this approach has
any correlations to the approach adopted by the majority of scholars interested in western
European balladry.
It can be seen at first sight that the terms ‘ballad’ and ‘epic’ have different meanings
when applied to western European and Slavonic traditional cultures. Since the place of epic
poems (literature) is occupied in Europe by the Edda, the Nibelungenlied, Beowulf, the
Poema de Mio Cid or the Chanson de Roland, ballads (oral, traditional culture), although
connected with epic poems and sometimes dependant on or derived from them, 15 would not
claim the status of the national epic. 16 The situation in Slavonic cultures is different. In
those Slavonic countries where Orthodox Christianity was adopted as the official religion,17
literatures were almost entirely clerical and no lay impact was possible before the beginning
of the seventeenth century. This means that the national epic poems (originally pagan)
would be less likely to be incorporated into the written culture than in Catholic countries.
As a result, eastern and southern Slavs did not create fully developed epic poems in written
form (e.g. the Russian Slovo o polku Igoreve [The Lay of Igor’s Campaign] is more of an
epic-like poem or, as some would prefer to name it, vestiges of an epic tradition in literature
than a full-blown epic), and balladry (oral epic songs) acquired the features that in the
western tradition would be split up between literary epic poems and ballads.18

14
This idea was developed in his chapter on Slavonic popular ballads by Kravtsov, N.I., Problemy slavianskogo
folklora (Moscow: Nauka, 1972), pp. 167-99.
15
Hart, W.M, Ballads and Epic. A Study in the Development of the Narrative Art (Boston: Ginn & Company,
1907).
16
A. N. Robinson, a Russian scholar, who examined Russian mediaeval literature and culture in the context of
other mediaeval literatures and cultures, argued that lay literatures in western European vernacular languages
expanded so quickly that they ‘took over’ the epic genres of national folklore and eventually put an end to the
oral transmission of epic poems: Robinson, A.N. Literatura Drevnei Rusi v literaturnom protsesse
srednevekov’ia XI-XIII vekov (Moscow: Nauka, 1980), p. 169.
17
Western Slavs developed under circumstances which could be held comparable to conditions in other European
countries.
18
On Russian and Slavonic epic traditions see such fundamental works as Russian Epic Studies, ed. by R.
Jakobson and E.J. Simmons (Philadelphia: American Folklore Society, 1949); Putilov, B.N., Russkii i
iuzhnoslavianskii geroicheskii epos: Sravnitel’no-tipologicheskie issledovaniia (Moscow: Nauka, 1972);
Propp, V.Ia., Russkii geroicheskii epos. 1st ed. (Leningrad: Izdatel’stvo Leningradskogo Universiteta, 1955;
reprint: Moscow: Labirint, 1999); Zhirmunskii, V.M., Narodnyi geroicheskii epos: Sravnitel’no-istoricheskie
ocherki (Moscow-Leningrad: Gosudarstvennoe izdatel’stvo khudozhestvennoi literatury, 1962); Rybakov,
B.A., Drevniaia Rus’. Skazaniia. Byliny. Letopisi (Moscow: Izdatel’stvo Akademii nauk SSSR, 1963);
Meletinskii, E.M., Proiskhozhdenie geroicheskogo eposa: Rannie formy i arkhaicheskie pamiatniki (Moscow:
Izdatel’stvo vostochnoi literatury, 1963); Slonim, M., The Epic of Russian Literature, from its Origins through
Tolstoy (New York: Oxford University Press, 1969). Some more recent works worthy of note are: Arant, P.

25
The Singer and the Scribe

Since the terminology has been clarified to some extent19 it can be understood why
those researchers who deal with western balladry would use the term ‘ballads’ when
discussing Russian bylinas and other Slavonic folk songs similar to them. 20 Bylinas were
chosen to represent Russian balladry probably because other genres of Slavonic folk songs
were less known to scholars, who were not specialists of Slavonic folklore. While not trying
to bridge the gap in one short paper, I would like simply to point up certain issues which
have not received enough attention from specialists of either western or Slavonic traditional
cultures. In this article the term ‘balladry’ will be applied to all narrative folk songs existing
in both Slavonic and western European folklore, 21 although the distinction between the
bylina and the ballad as sub-genres within Slavonic balladry will be particularly stressed.22
As in various western oral traditions the principle quality of Slavonic balladry (in the
wide sense of this term which in the practice of slavists would be a synonym of ‘epic
genres’) is also its narrative nature. According to Iu.I. Smirnov, one of the most prominent
writers on Russian and Slavonic balladry, the various genres of southern, eastern and
western Slavonic traditions, some of which are unique to their own traditions, can be
summarised as follows.
In the tradition of the southern Slavs seven groups of ballad-like (epic in Smirnov’s
terminology) songs are presented:

M., Compositional Techniques of the Russian Oral Epic, the bylina (New York: Garland Publishing, 1990);
Hritsa, S., Ukrainskaia pesennaia epika (Moscow: Sovetskii kompozitor, 1990); Bailey, J., Study of Russian
Epic Verse by Trubetzkoy, Jakobson, and Taranovsky: reconsideration and continuation (Amsterdam: Har-
wood Academic Publishers, 1996).
19
The division between so -called western and Slavonic balladry is one of the theoretical assumptions which has
been made. To clarify this point, however, it should be recognized that balladry in Europe is not
homogeneous; as Bengt Jonsson indicates, ‘The Scandinavian medieval ballad is an unusually well-defined
genre (with a restricted number of stanzaic forms and easily recognisable style), consisting of some 830 types.
Borderline cases are strikingly few, and although there are clear subgenres or categories within the genre, I
particularly want to stress that we are dealing with what is fundamentally one single genre. In this respect the
Scandinavian ballad – like many of the English or Scottish Child ballads – differs from songs labelled as
ballads elsewhere, for example in Germany, where narrative songs appeared in many different forms and do
not constitute one special genre’, Jonsson, Bengt R. ‘Oral Literature, Written Literature: the ballad and Old
Norse genres’, in The Ballad and Oral literature, ed. by Harris, p.140.
20
Entwistle, W.J., European Balladry (Oxford: Clarendon Press, 1939); Magnus, L.A., The Heroic Ballads of
Russia (London: Kegan Paul, Trench, Trubner & Co.; New York: E. P. Dutton & Co., 1921). See also the
entry on the ballad in Encyclopaedia Britannica: ‘France, Denmark, Germany, Russia, Greece, and Spain, as
well as England and Scotland, possess impressive ballad collections […] In no two language areas, however,
are the formal characteristics of the ballad identical. For example, British and American ballads are invariably
rhymed and strophic (i.e. divided into stanzas); the Russian ballads known as byliny and almost all Balkan
ballads are unrhymed and unstrophic’, Encyclopaedia Britannica Online: http://search.eb.com .
21
However, this does not mean that we can agree with a view that all narrative songs should be considered epic;
Czernik, S., Polska epika Ludowa (Wroclaw — Kraków, 1958).
22
In the western academic tradition it is more common to speak about types of ballads (border ballad, historical
ballad, broadside ballad, etc.) rather than about genres. In the Slavonic academic tradition the term genre is
heavily exploited.

26
 Ekaterina Rogatchevskaia

songs with mythological plots; iunak songs (with a strong heroic element); ballads;
early historical songs; robbers’ songs; songs based on Christian plots and legends; late
historical songs.
Eastern Slavs have,
bylinas; ballads; early historical songs; robbers’ songs; Christian songs; late historical
songs.
In the Russian tradition all six groups are found whereas the Belorussian and Ukrainian
traditions would have their own variations (for example the Ukrainian duma cannot be
considered as a complete equivalent to the Russian bylina).
The tradition of the western Slavs would be represented by such genres as ballads, robbers’
songs, Christian songs, and late historical songs.23 It is also worth mentioning that eastern
and southern groups of Slavs would have more similarities in their epic traditions due to
similarities in geographical position and historical circumstances (for example, the
domination of the Golden Horde in Russia and Turkish rule in the Balkans), while western
Slavs were much more influenced by European balladry.24
The most relevant distinguishing features of Russian balladry can be summarised as
follows:
Russian balladry will share with other national traditions such characteristics as oral
transmission, narrative structure, set formulae, focus on a single situation and conflict,
minimisation of characteristics and descriptions, dramatic effect achieved through
purposeful actions and abruptness, certain tunes associated with particular plots.25
Russian balladry was virtually extinct by the time systematic research started (even then
it was located mainly in the far North of Russia) and since the 1950s and 60s no new
texts and presenters could be found.
Russian balladry existed exclusively in oral form — no such phenomenon as broadsides
was known in Russia.
A relatively weak mythological and supernatural element (unlike, for example, in the
Czech tradition). 26

23
Smirnov, Iu.I., Slavianskie epicheskie traditsii (Moscow: Nauka, 1974), pp. 90-91.
24
For example, among Czech scholars there is a tendency to study the Czech ballad in the entire European rather
than Slavonic context: see Nejedlá, J., Balada v promene doby (Praha: Ceskoslovenský spisovatel, 1989).
25
The importance of examining ballad tunes has been stressed in recent studies carried out by specialists in
traditional music, as for example, Mukharinskaia, L.S., Iakimenko, T.S. ‘Kizucheniiu muzykal’noi tipologii
narodnykh ballad’, in Pamiati K.Kvitki (Moscow: Sovetskii kompozitor, 1983), pp. 111-121; Marchenko Iu.I.,
Petrova, L.I., ‘Balladnye siuzhety v pesennoi kul’ture russko-belorussko-ukrainskogo pogranich’ia’, Russkii
folklor, 27 (1993), 205-255; 28 (1995), 290-348; 29 (1996), 110-191.
26
Harkins, W.E., ‘K sravneniiu tematiki i kompozitsionnoi struktury russkoi i cheshskoi narodnoi ballady’, in
American Contributions to the Sixth International Congress of Slavists, ed. by W.E. Harkins (The Hague-
Paris: Mouton, 1968), pp. 133-63.

27
The Singer and the Scribe

A relatively weak Christian element in balladry (Christian culture flourished in literature

and in some other literature-related genres of folklore, such as dukhovnyi stikh —
‘spiritual poem’).
The Russian bylina and ballad are distinguished from historical songs27 which either tell
of specific historical events and people (for example, songs about Ivan the Terrible) 28 or
depict typical historical situations (for example, such widespread songs as Avdot’ia
Riazanochka29 and Tatarskii polon (The Tatar imprisonment) which tell of Russian
women captured by the Tatars).30
Heroic plots and ‘family’ stories are usually split up between the bylina and the ballad,31
although this point is debatable.
As this paper is entitled ‘Love Story or Heroic Deed?’ this last point is of great interest to
us in the context of the development of Russian balladry, and its impact on Russian
literature and culture. In the case study below we shall examine the features which bylinas
and ballads share as forms of Russian balladry and their specific characteristics which allow

27
Some scholars (Smirnov, Iu.I., Slavianskie epicheskie traditsii; Azbelev, S.N., Istoricheski pesni, ballady
(Moscow: Sovremennik, 1986); Putilov, B.N., Istoricheskie pesni XIII-XVI vekov (Moscow-Leningrad:
Izdatel’stvo Akademii nauk SSSR, 1960) and Geroicheskii epos i deistvitel’nost (Leningrad: Nauka, 1988)
etc. would include at least some historical songs in balladry, whereas others would argue that the nature of
these songs is completely different, as, for example Kravtsov, N.I., Problemy slavianskogo folklora (Moscow:
Nauka, 1972). Having no intention to solve this complicated problem I tend to accept the first approach, as it
is more universal and accurate when applied to the songs created in the seventeenth century. Close links
between the historical song, the bylina and the povest’ (narration), one of the genres of medieval Russian
literature, were demonstrated in the analysis of the story of the heroic life and death of M.V. Skopin -Shuiskii,
one of the brightest politicians and generals at the Time of Trouble (the beginning of the seventeenth
century): Stief, C., Studies in the Russian Historical Song (København: Rosenkilde and Bagger, 1953);
Azbelev S.N., Istorizm bylin i spetsifika fol’klora (Leningrad: Nauka, 1982).
28
See Perrie, M., The Image of Ivan the Terrible in Russian Folklore (Cambridge: Cambridge University Press,
1987).
29
A.O. Amel’kin has discovered that this song had a precise historical foundation: see Amel’kin, A.O., ‘O vremeni
vozniknoveniia pesni ob Avdot’e Riazanochke’, Russkii folklor, 29 (1996), 80-85 which automatically settles
the question of whether this song should be considered a ballad or a historical song. It is interesting to
compare this with the findings in Rieuwerts, S., ‘The Historical Moorings of ‘The Gypsy Laddie’: Johnny Faa
and Lady Cassillis’, in The Ballad and Oral Literature, ed. by Harris , pp. 78-96, where the problem of a
historical foundation does not call in question the balladic nature of the song. At the same time the approach to
historical songs demonstrated in A. Amel’kin’s paper is not universally approved by all slavists: specialists on
the southern Slavonic popular tradition would include all historical songs in balladry. See, for example,
Czajka, H. Bulgarska i DFHGR ska historyczna SLHGE ludowa (Wroclaw-Warszawa-Kraków, 1968); Braun M.
Das Serbokraotische Heldenlied (Göttingen, 1961).
30
For more on this distinction see Kravtsov, N.I., Problemy slavianskogo folklora (Moscow: Nauka, 1972), pp.
184-85. The author also argues that, from the wide range of songs centred on the story of imprisonment found
in all Slavonic countries, these should not all be automatically considered historical songs – most of them
would possess characteristics of the ballad.
31
N.I. Kravtsov calls this feature one of the ‘specific characteristics’ of Russian balladry compared to western
European, and especially English and Scottish, balladry.

28
 Ekaterina Rogatchevskaia

us to treat them as different genres. As a result we hope to discover which characteristics of

the ballad provoked its transformation from a form of the traditional folksong into modern
forms of literature, folklore and even mass culture.32
As a rule, slavists would differentiate the bylina from the ballad 33 on the following
grounds:
the types of stories they tell (content),
the type of hero on whom the story is centred (content),
the narrative techniques they exploit (form),
the length of songs: bylinas are usually longer than ballads (form),
the music which accompanies the songs (form), 34
the time when these genres were established, when they were most popular and when
their popularity faded (historical context), 35
the forms of their disintegration: the bylina became too complicated for oral
transmission, and younger singers started to transmit the stories in prose, whereas
ballads were taken over by lyrics and lost most of their narrative nature (historical
context), 36

32
The impact of balladry on modern culture was analysed by J. Culík, although he concentrated on the process of
studying the ballad tradition rather than on examining texts of ballads: Culík, J., ‘Folk Song – The Collectors,
the Scholars and the Public. The development of English, Scottish and Czech Folk-Song Scholarship: a
comparison’, Scottish Slavonic Review, 21 (1993), 85-112.
33
I.Iu. Smirnov also argues that together with ‘classical’ ballads other ballad-like songs should be analysed.
Ballad-like songs in his interpretation are either songs that changed their function in traditional culture,
becoming, for example, part of a ritual (which Russian ballads never were), or songs with some literary
elements; see Smirnov, I.Iu., Vostochno-slavianskie ballady i blizkie im formy (Moscow: Nauka, 1988), pp. 6-
7. An overview of other approaches is given by Burlasová, S., Kompozícia balady ako kritérium žánru (Praha,
1983) and Burlasová, S., Katalóg slovenských naratívnych piesní (Bratislava: Veda, 1998).
34
For an example of thorough research based on the Belorussian ballad see Mozheiko, Z.Ia., Pesni belorusskoogo
poles’ia , Part 1 (Moscow: Sovetskii kompozitor, 1983); part 2 (Moscow: Sovetskii kompozitor, 1984).
35
Although most of the bylinas that have survived over the years and are presented now in anthologies and
collections were created no earlier than in the thirteenth century, many specialists would agree that they
possess certain features that prove their formation as a genre as early as the tenth or eleventh century. At the
same time V.Ia. Propp, the author of the first and still the only complete monograph on the Russian bylinas in
which almost all aspects of the poems are examined, suggests that the genesis of bylinas should be studied in
the context of mythology; see Propp, Russkii geroicheskii epos. Since the fifteenth century very few new
bylinas have been created; however, the tradition of oral transmission did not dry up completely until the end
of the nineteenth or the middle of the twentieth century. Ballads first appeared in the thirteenth to sixteenth
centuries in Russia while the active life of the genre ended in the eighteenth century when no new songs were
created.
36
Smirnov, I.Iu., Vostochno-slavianskie ballady i blizkie im formy (Moscow: Nauka, 1988), p. 7; Astakhova, A.M.,
V.V. Mitrofanova, and M.O. Skripil’, Byliny v zapisiakh i pereskazakh XVII-XVIII vekov (Moscow-
Leningrad: Nauka, 1960). For further analysis see Harkins, W.E., ‘Epicheskie i liricheskie elementy v

29
The Singer and the Scribe

the social status of the audience: the audience for the bylina in the Middle Ages
included the cultural elite, and only later did the bylina become common among peasant
communities, while the ballad was incorporated into the popular (mass) culture from the
very beginning.
As far as story types and character types are concerned it is now a commonplace that
bylinas tell stories of fighting for Rus’ (the medieval name for Russia), a Prince (a lord), or
for a wife (the oldest stories are based on the mythological motifs of obtaining a woman).
The stories are centred on the heroic deeds of the bogatyrs (sing. bogatyr’), such well-
known heroes as Il’ia Muromets, Dovrynia Nikitich, Alesha Popovich, Sviatogor, Mikhaila
Potyk and others (several dozens in total). The Russian ballad concentrates exclusively on
family and love stories and is focused on ordinary and often unnamed people, which is
unknown in the bylina. I. Horak considers there to be four groups of typical stories:
stories about lovers;
stories about husbands and wives;
stories about parents and children;
stories about sisters and brothers.37
I. Smirnov, however, proposes another classification based not solely on the characters
involved in the story but also on the situation these characters are in:
mother and son / daughter;
a stranger / enemy is looking for a girl;
a hero is looking for a girl.38
However, when we start analysing individual texts it becomes clear that a combination of
all possible distinctive features is not normal; moreover, borderline cases occur with great
frequency. Therefore from our point of view they deserve special attention. One of the most
popular stories presented in almost all Slavonic traditions (including those of western Slavs)
is a story about brothers who oppose their sister’s love for the young man she desires.39 In
the Russian tradition this type of story can be both a ballad and a bylina: the most
widespread titles are Alesha Popovich and Elena Petrovichna or Alesha Popovich and the
Sister of the brothers Petroviches.40 Within each genre there are several variants, offering a
variety of structural elements.41 The common element of the story in all its variants is the

slavianskoi ballade’, in American Contributions to the Ninth International Congress of Slavists, ed. by P.
Debreczeny, Vol. 2, Literature, Poetics, History (Slavia, 1983), pp. 189-200.
37
Horak, I., ‘Les ballades populaires slaves’, Revue des études slaves, 39 (1961), 33.
38
Smirnov, Vostochno-slavianskie ballady i blizkie im formy, p.12
39
Story N 46 in I. Smirnov’s list of narratives common in Slavonic balladry: see Smirnov, Slavianskie epicheskie
traditsii, pp.127-128.
40
The bylina is known in 23 variants collected in the northern provinces of Russia and in Siberia. In the central
provinces of Russia and in Belorus only ballads based on the same story have been collected.
41
Balashov, Narodnye ballady, pp. 75-81; Putilov, B.N., Byliny (Leningrad: Sovetskii pisatel, 1957), pp. 411-413;
Grigir’ev, A.D., Arkhangel’skie byliny i istoricheskie pesni, sobrannye v 1899-1901 gg., Vol.1 (Moscow,
1904), songs Nn 54, 72, 81, 85, 97, 100, 118, 125, 128, 173; Kireevskii, P.V., Pesni, sobrannye Kireevskim,

30
 Ekaterina Rogatchevskaia

brothers’ discontent with their sister’s lover (rarely – her husband) and the murder or
attempted murder of the sister, or of both the sister and her partner. The plot is an
international one and can be found in other ballad traditions. The fact that the popularity of
this story spread across Europe seems to prove its international character rather than support
W.J. Entwistle’s theory of cultural borrowings.42
The bylina Alesha and Elena Petrovichna was thoroughly studied by V.Ia. Propp;43
however, he treated ballads with condescension and was not interested in those variants of
this song which represented the ‘tastes of the Philistines’.44 The artificial exclusion of the
ballad from research could not be productive. In our opinion, in his study of this song V.Ia.
Propp exaggerated the pathos of the story. He probably did so partly owing to his desire to
challenge his predecessors (‘bourgeois scholars’ in his words) who had claimed that,
because it had been created relatively late, this bylina should have been presented as an
example of the decadence of the genre and the hero, Alesha Popovich. V.Ia. Propp tried to
show that Alesha and Elena Petrovichna was a late adaptation of an ancient story about
winning a bride through combat. He also argued that all the characteristics necessary to call
this song a bylina were in place in this particular case and that the ‘Philistine spirit’ (usual
for ballads from V.Ia. Propp’s point of view) had not spoiled the story, at least in those
versions which could be called bylinas. V.Ia. Propp’s conclusions that the bylina was
definitely the first to be created and was only later ‘spoiled’ and given the form of the
ballad is also questionable. Doubts about V.Ia. Propp’s speculations become stronger if we
consider the entire ‘checklist’ of the bylina and ballad features found in different versions
of the story.

The story and its elements

Bylina
At the feast (in Kiev and/or in Prince Vladimir’s presence – optional) the brothers are
boasting of their obedient sister who is kept shut at home and has not been seen by anyone.
Bogatyr’ (hero) Alesha replies that he has seen her (in an intimate situation – optional, but
intimate relations are implied). The brothers go home to murder their sister (the style of the
murder varies). Alesha rescues the girl and marries her (the brothers disapprove of this –
optional). For bylinas the ‘rescue operation’ is indispensable. In most of the versions
Alesha’s heroism and strength are not really impressive: in some of them the girl herself
encourages Alesha to act, in others Alesha succeeds by ruse. In the majority of the versions

Vol. 2 (Leningrad: Nauka, 1983), pp. 64-67; Astakhova, A.M. et al., Byliny Pechory i Zimnego Berega
(Moscow-Leningrad: Nauka, 1961), pp.357-62; 415-16.
42
‘The ballad of Dobrynja and Aljoša is certainly related to the German Moringer; and the ballad-novelette in
which he casts doubt on the chastity of the Petrovici’s sister […] reproduces in part the Imogen story, of
Italian origin; it is to be found also in the Greek ballad of Maurianos, so that there is no way of determining
how it entered Russia. Its foreign origin is, at least, assured’, Entwistle, European Balladry, p. 364.
43
Propp, Russkii geroicheskii epos, pp.418-27; 583-84.
44
Propp, loc. cit. This position was relatively common among Soviet scholars and even those who specialised in
the study of ballads had to stress constantly that they would distinguish between traditional, popular, ballads
and late, bourgeois, ones.

31
The Singer and the Scribe

Alesha rescues the girl when the brothers are away which means that in any case Alesha
does not fight for her.
Ballad
The story in ballad form can deploy various motifs for the start of the conflict: either the
lover (Alesha Popovich or others) publicly (usually at a feast) boasts of his relations with a
young girl, 45 or a young man shoots an arrow and breaks a girl’s window thereafter they
become lovers 46 and while the brothers interrogate him about his presence in their house he
tells them the truth. The ending of the story also displays several variants: 1) only the girl is
to be murdered by her brothers, the lover (Alesha Popovich only) saves her and they marry
in church; 2) both are murdered and two birch trees grow up in their memory;47 3) only the
young man is murdered by the brothers while the sister commits suicide and her brothers
regret what they have done. Other motifs can also be incorporated into this story in other
Slavonic traditions which are not discussed here.
In both bylinas and ballads this story is presented as a love story, not a heroic one. There
is no fight in the story and it appeals to listeners because of the love and death dichotomy.

The character / hero

Bylina:
The name of Alesha Popovich appears in the vast majority of variants, although this
character does not resemble the classical image of this bogatyr’ in other bylinas.
Ballad
The character might be called Alesha Popovich, or Ivan Dudorovich, or Fedor
Kolyshchatoi, or he might not be given a name at all. The loss of the name is probably a
signal that it was not really important for the idea of the story and was not connected to
Alesha Popovich as a hero.

Narrative techniques
The narration in both the bylina and the ballad is framed by built-in verbal repetitions which
develop the formulaic style shared by the two genres. Nevertheless, the two forms can be
distinguished by the intensity of formulaic usage. The comparison of the opening episodes
of the bylina and of the ballad (here only those versions of the ballad which begin with the
argument at the feast are taken into consideration) shows that the introductory episode takes
33 lines or more 48 in the bylina and 21 or 22 lines in the ballad. Although formulae are still

45
The same motif occurs, for example, in the ballad Molodets i korolevna – A Guy and a King’s daughter: see
Balashov, Narodnye ballady, pp. 83-86, or in the bylina Dunai i Nastas’ia -korolevichna – Dunai and
Nastas’ia, King’s daughter: see Putilov, Byliny, pp. 414-16.
46
Similar motifs in the bylina Dovrynia i Marinka, Putilov, Byliny, pp. 292-98.
47
A contamination of two well-known motifs can also be seen in this variant: the motif of transformation into a
tree and the motif of young, innocent lovers buried together.
48
In some versions there is a long introduction dealng with Alesha’s wanderings and whereabouts, and an
explanation of how he eventually decided to visit Vladimir’s feast to have a good drink: see Astakhova, et al.,
Byliny Pechory i Zimnego Berega, pp. 357-58.

32
 Ekaterina Rogatchevskaia

heavily used, the tendency to compress narration is clearly seen in the ballad. For example,
in the bylina the narrative style would force a singer to follow a certain verbal pattern:
U Vladimira-kniazia byl pochesten pir,
Da vse na piru napivalisia,
Da vse na chestnom naedalisia ,
Da vse na piru priraskhvastalis’:
Da inoi khvastat zolotoi kaznoi,
Kak inoi khvastat molodoi zhonoi,
Kak inoi khvastat konem ezzhalym-e,
49
Kak inoi khvastat bykom kormlenym-e.

[At Prince Vladimir’s feast everyone eats and drinks a lot, and having got drunk everyone starts to show off.
One brags about the gold he has, another brags about his young wife, another brags about his excellent horse,
another brags about his healthy ox.]

The two brothers who also attend this feast have nothing to boast of except their sister. In
their speech the verbal pattern elaborated above should be repeated in full: only then would
it be one hundred percent complete:
‘Uzh vy oi esi, dva bratelka!
Ne p’ete, ne edite, nichem ne khvastaete’. –
‘Uzh my chem zhe budem khvastati?
Ishsha netu u nas zolotoi kazny,
Ishsha netu u nas molodoi zhony,
Ishsha netu u nas byka kormlenogo,
Ishsha netu u nas konia ezzhalogo, –
Tol’ko est’ u nas edna sestritsa,
50
Ishsha ta zhe Elenushka Petrovna-svet.’

[Why don’t you, brothers, eat and drink, why don’t you show off? – What do we have to brag about? We
don’t have gold, we don’t have young wives, we don’t have a healthy ox, we don’t have a excellent horse. We
only have our sister, our sister – Elena Petrovna.]

The same technique is fundamental to the narrative structure of the Russian ballad, although
compared to the bylina the narrative economy of the ballad is stronger and affects even such
fundamental principles of narrative technique as repetition. At the same time there is one
variant of the bylina which is extremely long,51 although the length does not seem original.
The expansion of Vladimir’s role in the story, the extended introductory episode where
Alesha’s wanderings are described, the emergence of Ilia Muromets, another hero
(bogatyr’), who suggests his help in the rescue operation, though there is no need for it (no
battle is mentioned in the bylina) and some other details seem extraneous and make us think
that they were all added later to make the story resemble a bylina more closely.
In the Russian tradition the ballad exploited the same narrative techniques which had
been first established in the bylina, and this makes the question of origins more

49
Putilov, Byliny, p. 411.
50
Ibid. p. 411.
51
Astakhova, et al., Byliny Pechory i Zimnego Berega, pp. 357-62.

33
The Singer and the Scribe

complicated. Although narrative techniques were simplified to enable the ballad to achieve
more intense emotional effect, the similarities between the two forms of Russian balladry
are very deep. In our opinion the correlation between the two genres (in which the bylina’s
position is that of a first-born) does not necessarily have an effect on the history of every
single story. The analysis of the bylina and the ballad versions shows that the essence of the
story is that of a song of tragic love. It seems reasonable to argue that this international love
story was initially adopted by the Russian tradition in the form of a ballad, as it is a primary
objective of the ballad and not of the bylina to narrate stories about love and arouse
sympathy for the characters. The bylina, as I indicated above, has different objectives and
traditionally never concentrates on love stories. Therefore we can suggest that this story of a
young girl and her lover started its life as a ballad and later, in those provinces in the north
of Russia and Siberia where the bylina was still a living and productive genre in peasant
communities and where the tradition of singing bylinas was stronger than that of ballad
singing, the ballad was turned into a bylina.52 It was extended, associated with Alesha
Popovich (probably because he is the youngest and the most rascally hero in the Russian
bylinas) and a happy ending celebrating Alesha’s heroic deed was introduced.
Consequently, the formal features of the genre took over the story transforming it into a
bylina.
If this theory is correct it becomes clear how such a lengthy and unnecessarily detailed
variant as the one that I have discussed above could have been created. The priority of the
ballad form in this case also explains how motifs not strictly belonging to the main story
appeared in late versions of bylinas about Alesha Popovich and Elena Petrovichna (for
example, Alesha frankly describes his meetings with the girl and accuses the brothers’
wives — more new characters in the story — of adultery). 53 It seems like these additions
compensate for the lack of drama which the listeners were looking for in such stories. As a
result, the long and detailed bylina looks like a compromise, having the content of a ballad
and the form of a bylina.
As the Russian ballad at the very start was oriented towards popular rather than towards
elite culture, the listeners’ enthusiasm for empathising with the characters and for putting
themselves into the characters’ shoes was much greater than similar emotions that the
presentation of the bylina could have provoked. In my opinion, this is the key issue that
distinguishes the Russian ballad from the bylina. When the bylina gradually changed its
status in society and consequently its audience, the ballad came to influence the bylina, as
we can see in our case study, where one plot was transformed from ballad form to bylina
form. The main attribute of the ballad in Russian balladry is the possibility of imitating
suffering, fear, pain — everything, in fact, which people would normally like to avoid in
real life — fulfilling exactly the role that soap-operas do to modern popular culture. The
Russian ballad is one of the folklore genres which fills this gap in popular culture. In
contrast, the function of the bylina was to commemorate the glorious past and to establish
the idea of national identity. Having adopted this opinion, we can see that it becomes easier

52
A.M. Astakhova mentioned this as an assumption in her commentaries appended to the publication of this
bylina. However, she did not develop the idea: see Astakhova, A.M., Byliny Severa. Mezen’ i Pechera, Vol. 1
(Moscow-Leningrad: Izdatel’stvo Akademii nauk SSSR, 1938), pp. 34 & 553.
53
Astakhova, et al., Byliny Pechory i Zimnego Berega, pp. 357-62.

34
 Ekaterina Rogatchevskaia

to give reasons for the cruelty and violence that the ballads are full of; in the following
passge a girl threatens a boy who offended or harassed her:
Ia veliu brattsam podstrelit’ tebia,
Podstrelit’ tebia, potrebit’ dushu,
Ia iz kostochek terem vystroiu,
Ia iz rebryshek poly vysteliu,
Ia iz ruk, iz nog skam’iu sdelaiu,
Iz golovushki edbovu sol’iu,
Iz sustavchikov nal’iu stakanchikov,
Iz iasnykh ochei – chary vinnye,
Iz tvoei krovi navariu piva.

[I shall ask my brothers to shoot you, I shall ask my brothers to take away your soul. I shall build a house of
your bones, I shall make the floor in the house of your ribs, I shall make a bench of your arms and legs, I shall
make a bowl of your scalp, I shall make cups of your joints, I shall make wine glasses of your eyes, I shall
54
brew beer of your blood.]

The function of imitation can be assigned not only to the ballad but also to some folklore
genres in prose. However, this is indeed the function that gave the traditional Russian ballad
the power to transform itself into lyrics and revive itself in the form of the modern popular
‘cruel’ song. On the other hand, having lost its traditional area for transmission, the bylina
was transformed into short pieces in prose. The place of a historical account of the glorious
past was either taken over by historical songs or went out of favour with society. Therefore
the bylina lost not only its form but, what is probably more important, its function.
Although the links between the traditional popular ballad, the literary ballad which was
flourishing in the Romantic era and was initiated in Russian literature though translations of
British and German Romantic authors and the modern popular ballad labelled a ‘cruel
romance’ or a ‘city romance’, deserve special research, it is obvious that the function of the
ballad was demanded by people, and the traces of the traditional form of the Russian ballad
can be found in literature and modern popular culture.

54
Balashov, Narodnye ballady, pp. 142. The list of examples can be extended.

35
This page intentionally left blank
 From Oral Adventure Story to Literary Tale of Enchantment:

the case of the Count Arnaldos ballad

Huw Lewis

Whenever we study the oral culture of any language, few factors can make research more
difficult, or more interesting, than the need to untangle the complicated web of alterations,
confusions and intrusions from other sources that have contributed to the development of a
particular tale or, in this case, a ballad. Such changes may take place as the result of a
deliberate attempt to make a narrative more suited to a new locale, through inadvertent
confusion with other narratives or local lore, or, as is often the case in the Spanish ballad
tradition, through borrowings (ranging in length from a half-line to several lines) from other
ballads, as deemed appropriate by the singer or scribe. Thus it is not surprising to find
several variant forms of the most popular ballads (some of these coming, not from the
Peninsula itself, but from North Africa and even Latin America), and this is certainly the
case with the family of ballads that is to provide the focus of this paper, whose probable
evolution from an oral adventure story to a literary gem evoking magical powers and
contact with the supernatural otherworld provides an excellent case study of the factors that
can contribute to an oral narrative’s evolution. This study will attempt to show how such
change may have taken place, and to explain how an apparently innocuous tale became one
of the most lauded and studied ballads of the Spanish language, reaching beyond the
confines of its own time and place to influence creations as diverse as Longfellow’s ‘The
Sound of the Sea’, and Antonio Gala’s 1994 collection of essays, A quien conmigo va,
where the writer offers some of his reflections on life (the title, meaning ‘Whoever
accompanies me’, is a quotation from the ballad).
Of all Hispanic ballads, the Romance del Conde Arnaldos is one of the most beautiful
and haunting, but its ‘meaning’ has baffled a host of critics, who have interpreted it
variously as a simple adventure story,1 a Christian allegory,2 and an allegory of frustrated
love.3 The folkloric background to this ballad has also attracted some attention,4 and it has
been argued that El Conde Arnaldos may be related in theme and content to a group of
ballads from European tradition concerning the Wild Host. The extant versions of this

1
Menéndez Pidal, Ramón, Poesía popular y poesía tradicional en la literatura española (Oxford: Clarendon
Press, 1922), p. 14.
2
Hart, Thomas R.,‘“El Conde Arnaldos” and the Medieval Scriptural Tradition,’ Modern Language Notes, 72
(1957), 281-5.
3
Hauf A., and J.M. Aguirre, ‘El simbolismo mágico-erótico de “El infante Arnaldos”’, Romanische Forschungen,
81 (1969), 89-118.
4
Spitzer, Leo, ‘Notas sobre romances españoles’, Revista de filo logía española, 22 (1935), 153-74; ‘The
Folkloristic pre-Stage of the Spanish Romance “Conde Arnaldos”’, Hispanic Review, 23 (1955), 173-87.
The Singer and the Scribe

ballad are at least as numerous as the interpretations of the critics, for it is to be found in: a
sixteenth-century chapbook; the Cancionero de Londres (1471–1500), where it is attributed
to Juan Rodríguez del Padrón; the Cancionero de romances of 1550; the Cancionero de
romances of unknown date; various Sephardic sources. Whatever their own interpretation
of the ballad, however, all critics agree with Menéndez Pidal that the ballad (or its most
truncated form, at least, as found in the Cancionero de romances of unknown date) is a
masterpiece (‘una obra maestra’)5 of Spanish balladry. The following version is the one
which Pidal considered to be the jewel in the crown of the Spanish ballad tradition (this and
all other English translations are my own):
¡Quién hubiese tal ventura Who could have had such good fortune
sobre las aguas del mar, on the waters of the sea
como hubo el conde Arnaldos as did Count Arnaldos
la mañana de San Juan! the morning of St John’s day!
Con un falcón en la mano A falcon in his hand
la caza iba a cazar; he went out hunting,
vio venir una galera and saw a galley
que a tierra quiere llegar. approaching the land:
Las velas traía de seda, its sails were made of silk,
la ejarcia de un cendal; and its rigging of pure sendal;
marinero que la manda the sailor at its helm
diciendo viene un cantar came singing a song
que la mar facía en calma, which calmed the sea,
los vientos hace amainar; and made the winds drop,
los peces que andan nel hondo, and made the fish from the deep
arriba los hace andar; swim near the surface,
las aves que andan volando, and the birds flying above
nel mástel las faz posar. came to perch on the mast.
Allí fabló el conde Arnaldos, Then spoke Count Arnaldos,
bien oiréis lo que dirá: hear what he had to say:
Por Dios te ruego, marinero, -In God’s name, sailor
dígasme ora ese cantar. I ask you to teach me your song.
Respondióle el marinero, The sailor replied,
tal respuesta le fue a dar; this is what he said:
Yo no digo esta canción -I only teach my song
6 7
sino a quien conmigo va. to those who accompany me.

This narrative clearly raises a host of questions (who is the sailor, what is the strange power
of his song, what is the ‘good fortune’ that Arnaldos is about to enjoy?), and over the years
these obscurities have led many to believe that the ballad recounts an encounter with
supernatural forces.
In an attempt to counterbalance the Romantic interpretation of the ballad as an
expression of ‘primitive’ superstition and belief in the supernatural, however, Pidal pieced

5
Menéndez Pidal, Poesía popular, p. 10.
6
Romancero Antiguo, ed. by Juan Alcina Franch, 2 vols (Barcelona, 1971) II, 473-74 (hereafter RA).
7
The original of this version, as well as of most of the other texts discussed here, may be found in RA vol. 2.

38
 Huw Lewis

together what he described as a fuller version of the ballad based on the combination of the
sixteenth-century chapbook and twentieth-century Sephardic sources.8 The first half of the
following translation reproduces the chapbook, the second half the Sephardic addition:
¡Quién hubiese tal ventura Who could have had such good fortune
sobre las aguas de la mar, on the waters of the sea
como hubo el infante Arnaldos as did prince Arnaldos
la mañana de San Juan! the morning of St John’s day!
Andando a buscar la caza Heading out to hunt
para su halcón cebar to feed his falcon
vio venir una galera he saw a galley
que venía en alta mar; approaching on the high sea:
las áncoras tiene de oro, its anchors were made of gold,
las velas de un cendal; and its sails of pure sendal;
marinero que la guía the sailor at its helm
va diciendo este cantar: came singing a song
Galera, la mi galera, — My galley, my dear galley,
Dios te me guarde de mal, May God protect you from harm,
de los peligros del mundo, from the perils of the world,
de fortunas de la mar, and the fortunes of the sea,
de los golfos de León from the Gulf of Lions
y estrecho de Gibraltar, and the Straits of Gibraltar,
de las fustas de los moros from the Moorish vessels
que andaban a saltear. that are ready to pounce.
Allí habló el infante Arnaldos, Then prince Arnaldos spoke,
bien oiréis lo que dirá: hear what he had to say:
Por tu vida, el marinero, — I beg you, sailor,
vuelve y repite el cantar. come back and repeat your song.
Quien mi cantar quiere oír, — Whoever wishes to hear my song
en mi galera ha de entrar. must come on board my vessel.
Tiró la barca el navío The ship sent out its longboat,
y el infante fue a embarcar; and the prince went on board;
alzan velas, caen remos, they raised the sails and dropped the oars,
comienzan a navegar; and headed on their way;
con el ruido del agua the sound of the water
el sueño le venció ya. lulled him to sleep.
Pónenle los marineros Then the sailors
los hierros de cautivar; clapped him in irons;
a los golpes del martillo, the sound of the hammer
el infante fue a acordar. caused the prince to wake.
Por tu vida, el buen marino, — I beg you, good sailor,
no me quieras hacer mal: do not cause me harm:
hijo soy del rey de Francia, I am the son of the king of France
nieto del de Portugal; and grandson to the king of Portugal,
siete años había, siete, for seven years now, seven,

8
RA vol. 2, 4. 75-76.

39
The Singer and the Scribe

que fui perdido en la mar. I have been lost at sea.

Allí le habló el marinero: Then the sailor spoke:
Si tú me dices verdad, — If what you say is true
tú eres nuestro infante Arnaldos you are our prince Arnaldos
y a ti andamos a buscar. and it is you we have come to find.
Alzó velas el navio The ship raised its sails
Y, se can a su ciudad And returned to its home port
Torneos y más torneos, Tournament followed tournament
Que el conde pareció ya. For the Count was found at last.

In this version the details concerning the mesmerising effect of the song on the sea, fish,
and birds have clearly been omitted, to be replaced by the text of the song itself (this being
a simple invocation unto God to keep the vessel safe from the dangers of the sea), which is
also included in the highly corrupt version reproduced in the Cancionero de Londres.
It was on the basis of these two additions  the text of the song and the new, expanded,
ending  that Menéndez Pidal came to the conclusion that the dream-like, supernatural air
generated by the ballad had received undue attention, and that originally this was really
nothing more than a simple adventure story. Nevertheless, this conclusion is based on a
number of dubious assumptions, the first and foremost of which is the presumption that the
ending of the twentieth-century Sephardic version faithfully reflects what would have been
the ballad’s ‘original’ ending; in fact, this ending could be a much later addition, as is the
case with the other, highly unsatisfactory endings that have been associated with it, one
example being the version of the poem attributed to Juan Rodríguez del Padrón, where the
ending has blatantly been culled from the Ballad of Count Olinos, with little or no regard to
the logical development of the story, for, following the text of the sailor’s song, we are here
told the following:
Oídolo ha la princesa The princess heard him
en los palacios do está: from the palace where she lives:
— Si saliéredes mi madre, — If you came out, mother,
si saliéredes de mirar if you came out to look
y veredes cómo canta you would see how he sings,
la sirena de la mar. the siren of the sea.
— Que non era la sirena, — That is not the siren,
la sirena de la mar, the siren of the sea,
que non era sino Arnaldos, that is just Arnaldos,
Arnaldos era el infante, Arnaldos the prince
que por mí muere de amores, who languishes for love of me,
que le quería frustrar. which I do not return.
¡Quién le pudiese valer, If only he could be helped
9
que tal pena no pagase! so that he should not suffer so much pain!

Here, not only has the ballad been transformed into a tale of unrequited love, but the
borrowing has also led to the illogical twist that it is now Arnaldos who is apparently
singing, and not the sailor as we had been told in the first half of the text! Furthermore, we

9
RA vol. 2, 474-75.

40
 Huw Lewis

must also wonder about the reliability of the chapbook as a means of reconstructing the
‘orignial’ version of the ballad, for the inclusion of the text of the sailor’s song in the
sixteenth-century text could itself have been brought about by the intrusion of a source as
corrupt as the Cancionero de Londres, where we are also told, strangely, that Arnaldos is
hunting lizards (lagartos) when he encounters the ship. This reference itself indicates how
easily incongruencies can find their way into textual variants such as this, for the
identification of lizards as Arnaldos’s prey is itself probably the result of a scribe mis-
transcribing, or mis -copying ‘la garça’ (heron) which, according to the 1550 Cancionero de
romances, is his quarry. The failure to identify the strangeness of hunting lizards is also an
indication both of the lack of close attention often paid by scribes to the logic of their text,
and also the fact that, with the passage of time, this particular ballad becomes increasingly
associated with exotic and alien pursuits. If interference from a similarly erratic source has
indeed taken place in the chapbook, then this would completely undermine its value as
evidence of the ballad’s ‘original’ theme and content. Furthermore, Spitzer10 has astutely
drawn attention to a series of problems and incoherences in Menéndez Pidal’s
reconstruction:
1. We are told that Arnaldos has a ventura at sea, yet he is out hunting on dry land when
the poem opens. This is not explained convincingly by Pidal.
2. What is so special about the sailor's song, and why will he not repeat it? The text of
he song given by Pidal is hardly worth being so closely guarded.
3. Why does the sailor wish to lure Arnaldos on board his ship? If he is really out
searching for his long-lost lord, what advantage could he possibly hope to gain by
abducting a complete stranger?
4. Why does Arnaldos fall asleep when he boards the ship in the extended version of the
poem?
Finally, we must bear in mind that even if Menéndez Pidal’s reconstruction of the ballad is
accurate, we are still faced with the fact that at some stage in its development an unknown
(and for evermore unknowable) imagination saw fit to re-shape the text, giving it that
haunting quality which evokes contact with the supernatural. Even if its origins were
mundane (and this fact is by no means proven), the ballad was later transformed into
something far more enduring and significant, with more than a hint of the extraordinary.
There is, moreover, one other version of the ballad which we must take into account
before proceeding to a closer examination of the text itself. Bénichou,11 also collating
material from Sephardic sources, published yet another version of the Arnaldos ballad (a
token, perhaps, of the danger of relying too heavily on one modern variant as a means of
reconstructing the notional ‘original’ ballad), which is interesting in a number of respects.
First, the protagonist’s name is no longer Arnaldos but Fernando, although he is referred to
by the more archaic title, infante. It is the longest of all extant variants of the ballad,
containing the sailor’s song but not the reference to the joyous feasting at the end. It
provides us with some information about Arnaldos/Fernando’s estate: he is a successful

10
Spitzer, ‘The Folkloristic Pre-Stage’.
11
Bénichou, P., ‘Romances judeo-españoles de Marruecos’, Revista de Filología Hispánica, 6 (1944), 255-79
(at pp. 268-69).

41
The Singer and the Scribe

leader in battle and has captured seven cities together with Rome itself. Disappointingly,
however, this longer version does little to help clarify any of the more obscure aspects of
the shorter text. The reference to Fernando’s ‘rescue’ is indeed made all the more baffling
because it is difficult to imagine anyone in less urgent need of succour than this mighty
conqueror! We will see, however, that the incoherence of this version will, paradoxically,
help our attempts to unravel the complex web of confusions and omissions that has clouded
our understanding of the development of this family of ballads; an analysis of some of the
ballad’s central motifs will elucidate this point.

1. The good fortune at sea

The poem opens by alluding to Arnaldos’s adventure and good fortune at sea, the wording
of which is highly reminiscent of the Ballad of Floriseo:
¡Quién hubiese tal ventura Who could have been so fortunate
en haberse de casar when he came to be married
como hobo Floriseo as was Floriseo
12
cuando se fue a desposar ... the day he went to be wed!

In itself, this obvious repetition is not surprising, as it is common for such formulae to be
borrowed from one ballad to another. However, Menéndez Pidal himself suggests that
Floriseo could be the source of some of the material in Arnaldos,13 and the likelihood of
this is further reinforced by the fact that even specific details are echoed, to the extent that
the former also includes a description of a beautiful ship from which a melodious song
emanates, mesmerising fish and birds alike. Significantly, Floriseo and his companions are
not confronted by a sailor at this stage, but by a beautiful maiden (doncella hermosa), who
lulls the hero to sleep and takes him to her mistress, who, in turn, puts him under a spell,
forcing him to give her his love and to stay with her as her consort; in this way we are given
a plausible reason for the abduction of the protagonist: so that he may become a
supernatural maiden’s lover. The appearance of a damsel rather than of a sailor on the ship
is also far more in keeping with this motif as it is found elsewhere, especially in early Celtic
tales and in French romance, where women are far more common than men as
representatives of the supernatural otherworld. The whole description of Floriseo’s
kidnapping is, indeed, more logical and detailed than that of Arnaldos, and provides us with
an important key to the significance of some of the references in the latter, as will soon
become apparent.
Returning to the opening lines of Arnaldos, we remember that we are told of the
protagonist’s good fortune at sea. Folk belief frequently associates the supernatural with
water, and we are often told of supernatural beings who sail across water in search of some
sought-after mortal. In most of these traditions, the protagonists are approached by female
supernatural beings, and while male emissaries are not entirely unknown, they are far less
common; it is therefore possible that a female messenger has been replaced by a male sailor
in this ballad. Picking up on these associations between water, boats and the otherworld,
Débax and Martínez Mata in an analysis of this ballad have indeed suggested that it may
tell of an encounter with the supernatural, although their interpretation lays particular

12
RA vol. 2, 424-25 .
13
Menéndez Pidal, Poesía popular, pp. 16-17.

42
 Huw Lewis

emphasis on associations with death, pointing out that many traditions place the land of the
dead on the other side of a water barrier.14 Certainly, this interpretation has its merits, and is
particularly relevant when we broach the question of the ballad’s relationship to the theme
of the Wild Host. Interestingly, of all the Hispanic ballads allegedly related to this theme,
Arnaldos is the only one where the abductor and the abducted are of the same sex: in all the
others a male abductor makes off with a female victim. Thus, through the fusion of the two
possible sources already mentioned  Floriseo and the Wild Host  a male emissary may
have been substituted for a female one in the Arnaldos ballad, causing the tale either to
evolve in the following manner, or at least to evoke some of the following connotations in
the minds of those who eventually transformed it into the truncated account of an encounter
with the supernatural: (i) by analogy with Floriseo or similar narrative, we are presented
with a situation whereby a supernatural being (originally female) sails across the sea in a
ship of wondrous craftsmanship; (ii) this damsel sings a song that mesmerises birds and
fish, who congregate around the ship and so give added testimony to its marvellous origins;
(iii) the damsel approaches Arnaldos and attempts to lure him to the supernatural
otherworld, eventually being forced to cast a spell on him in the form of a song that lulls
him to sleep; (iv) in this way he is captured and abducted, the purpose of his abduction
being so that a supernatural maiden may win or force his love.
With the intervention of the theme of the Wild Host, coupled with increasing ignorance
of the significance of individual motifs and primitive beliefs, a number of these motifs
become confused. The female enchantress now becomes male (or, in Pidal’s reconstructed
text, a number of sailors), the more common sex of the abductor in Wild Host tradition;
Arnaldos is still lulled to sleep, but, in an attempt to rationalise the misunderstood motif of
the enchanting song we are told that it is the soporific effect of the lapping of the water that
brings about the stupor; and, surprisingly, Arnaldos must also be clapped in irons to prevent
him from escaping. This more physical and violent manifestation of the abduction motif
brings us closer to this theme as it is found in other traditions concerning the Wild Host,
where abduction is associated with that most extreme of all forms of violence, death.
Finally, since the abductor is now male, the original reason for the abduction (an attempt to
realise pas sionate desires) is now illogical and irrelevant, and is consequently omitted from
this altered version of the story; since the significance of the Wild Host as a portent of death
also seems to have been lost (explicitly, at least), the new ending owes more in its
conception to Floriseo’s rescue from the hands of Laciva, the sorceress who had put him
under her spell: when Arnaldos awakes from his enchanted sleep, it is as if he is returning
to ‘reality’ after seven years of enforced separation from his orig ins and conscious, mortal
existence. This reading also clarifies some of the incoherencies of Bénichou’s version of the
ballad: if Fernando has been living under a spell, then in spite of his martial success he is
still in need of being rescued from the clutches of his abductor. Thus Débax and Martínez
Mata’s interpretation faithfully reflects one of the most important elements which has
contributed to this ballad’s development, death certainly being a theme which seems to
underlie the whole of this text, although it would be a mistake to regard it as the main focus,
either at its conception or later.

14
Débax, Michelle, and Emilio Martínez Mata, ‘Lecturas del “Conde Arnaldos”’, reprinted in Francisco Rico and
Alan Deyermond eds, Historia y crítica de la literatura española 1/1. Edad Media: primer suplemento
(Barcelona: Crítica, 1991), pp. 224-9, at 227-8.

43
The Singer and the Scribe

The fact that the adventure takes place on St John’s Day (24 June) is also of no mean
importance. As Hauf and Aguirre indicate in their discussion, St John’s Day may be
associated in popular belief with misfortune, especially in the form of abduction, a theme
which is also intimately linked with that of the hunt. Thus, at the very beginning of this
ballad, we have a fusion of three separate contexts for abduction in popular lore: the arrival
of a seafaring enchanter/enchantress, St John’s Day, and the hunt.

2. The Hunt
The hunt is a central feature in this ballad. Once again, it parallels Floriseo, where the hero
and his retinue are out hunting when they encounter the marvellous ship and its
supernatural passenger. The motif of the hunt in the romancero and other ballad traditions
has been discussed and elucidated by Edith Rogers,15 who shows the different contexts in
which this motif may arise, and the different functions it may perform. Sometimes, hunting
is mentioned without special emphasis, merely as the habitual activity of a knight; it is a
means of getting him out into the countryside to a place where he might experience some
adventure. But the hunt can also be an important feature of the ballad, perhaps as a symbol
for the love-chase, or as a prelude to death, or else either as an activity which leads the
protagonist to the dwelling-place of supernatural or enchanted creatures, or, finally, as the
threshold or point of transition from this to an imaginary/marvellous world. It is in the latter
category that she places the Arnaldos ballad. Without question, the hunt functions on a
variety of levels in this poem. Symbolically, it may represent the love quest, as Hauf and
Aguirre have suggested. This is particularly true of the corrupt Cancionero de Londres
version of the ballad where the amorous dimension is brought to the forefront from the
outset:
¡Quién hubiese tal ventura Who could have had such good fortune
con sus amores folgar, enjoying the fruits of love
como el infante Arnaldos as did prince Arnaldos
16
la mañana de San Juan! the morning of St John’s day!

If this ballad is at all related in theme to a tale similar to that of Floriseo, then the allusions
to the love-chase are certainly not out of place here, for in the prose romance, Floriseo, we
learn that the sorceress Laciva falls in love with the protagonist and retains him at her court
as her protector and consort; ironically, therefore, Floriseo here turns out to be the hunted
rather than the hunter. Considering the development of all versions of the Arnaldos ballad,
it is reasonable to assume that we are being presented with a similar case of entrapment and
role-reversal. Such reversals are well-attested in folk tradition, where hunters of noble
descent frequently find that they are the unsuspecting prey of some supernatural adversary,
or the quarry sought by some benevolent protector. Indeed, chivalric romance from its very
conception is at pains to stress the twin roles of the king or nobleman as a ruler in society
but also as the servant of his God (or gods and goddesses in the case of the pagan tales
which are the ultimate source of some of these romances), and the necessity of his fulfilling

15
Rogers, Edith Randam, ‘The Hunt in the Romancero and Other traditional Ballads’, Hispanic Review, 42 (1974),
133-71; idem, ‘The Perilous Hunt: Symbols in Hispanic and European Balladry’, Studies in Romance
Languages, 22 (Lexington: University of Kentucky Press, 1980).
16
RA vol. 2, 474-75.

44
 Huw Lewis

his duty to those people under his protection. The hero of a romance therefore both leads
and is led, serves and is served, hunts and is hunted. It might even be said that the vital task
which all heroes of chivalric romance must accomplish before achieving their ultimate goal
of sovereignty is that of reconciling the essentially dual nature of their existence as
intermediaries between this world and divine will. Constantly trapped on a threshold
between two modes of existence, the mortal and the supernatural, it is inevitable that they
should continually find the traditional roles they assume being challenged and their own
identity (as leader, lover, hunter) being subverted and even overturned. It is fitting,
therefore, that in Arnaldos the hunter should himself be lured into a trap, one which itself
has possible associations with the love chase.
As well as possibly acting as symbol for the love chase, the hunt in the Arnaldos ballad
also acts as an introduction to the strange events that follow, a function which it performs
for a number of reasons. A hunt often takes place in a forest, an environment which is apt to
evoke images of enchantment and of the unknown, where strange beings may dwell, and
where it is easy for a man to be separated from his companions and lose his way. It is
therefore an ideal environment for the marvellous and for the unknown. The hunt also has
close links with the theme of the quest, and as such is an appropriate prelude to an
adventure which might furnish the hero with a store of experiences and insights of
inestimable value to him in future life.
But the significance of the hunt in the Arnaldos ballad transcends even this
interpretation. Again, comparison with the prose Floriseo provides us with important
additional information. We learn that the eponymous hero of this romance possesses a
magical sword which has the particular property of protecting its bearer from all forms of
enchantment and sorcery. However, because he is engaged in the peaceful and social
activity of hunting, Floriseo is without his sword, and so gives Queen Laciva precisely the
opportunity she has been waiting for in order to lure him into her clutches. Evidently, the
hunt performs a vital function here, for its conventions place the protagonist in grave
personal danger; without his protective sword, Floriseo is as vulnerable as any other mortal
to the spells and enchantments cast by his enemies, and it is really only through sheer good
fortune that his fate is not much worse, for it had been Laciva’s original intention to kill
him, until his beauty and charm won her heart. Certainly, Floriseo could be described as
having enjoyed good fortune in this respect, and his situation is somewhat comparable to
that of Fernando in the version of the Arnaldos ballad collected by Bénichou: Fernando is
himself a great warrior and leader in battle, and it is again the hunt that suggests his
vulnerability when he is confronted with the unexpected while unprotected by the normal
martial apparatus. Arnaldos, too, is seemingly devoid of any protection or accompaniment:
in each case the hunt places the protagonist at a disadvantage for he is moved outside the
usual security of the court or even his own powers, and at the mercy of the supernatural.
As has been mentioned, this ballad is but one of many which may be connected with the
Wild Host, a theme which, though of Germanic origin, was widely known throughout
Europe in the Middle Ages. Popular belief maintained that Woden, the Germanic god of
death, or a substitute (the Devil, Herod, or even King Arthur) at certain times of the year led
a horde of tormented souls over the world in a frenzied hunt for innocent, living victims.
Spitzer also relates ‘la chasse sauvage’ to the ‘danse macabre’, the belief that the spirits of
the dead become engaged in an eternal dance. All these traditions reflect a belief in the

45
The Singer and the Scribe

existence of malicious supernatural beings who, using the element of surprise, brute
strength, or enchantment, attempt to trap and abduct unwary mortals. It is to this belief that
Spitzer relates our ballad, adducing primarily the same theme as it is found in Rico Franco,
aragonés, where the protagonist carries off a beautiful maiden against her will.
Significantly, this ballad also opens with a hunt.
Of all Hispanic ballads of this type, Rico Franco, aragonés is undoubtedly the closest in
theme and content to the Germanic Halewijn, although it should be noted that the exact
details of the damsel’s abduction do not appear in this Spanish ballad, and no mention is
made of any form of charm or enchantment which might have caused her to fall into Rico
Franco’s clutches. Furthermore, it is significant that in this ballad, as in practically all the
others on the theme of the Wild Host, it is the abductor, and not the abducted, who is
depicted as being out hunting. In Arnaldos, on the other hand, the reverse is true, which
seems to indicate that more factors have interfered in the development of this ballad than
simple analogy with the Germanic Wild Host. It must, therefore, be emphasised once more
that the motif of the hunt performs a completely different function from that in the Wild
Host, and that, by transferring the role of hunter from the abductor to the abducted, the
emphasis has been shifted, the hunt being used, not as a symbol of death, but as a threshold
between this and the supernatural world, as Rogers has correctly indicated. However, it is
also interesting that in Rico Franco the damsel should have been abducted from a castle,
for, if we turn our attention once more to the version of Arnaldos collected by Bénichou, we
see that an oblique reference to the motif of the hunt is combined with abduction from a
castle, where Fernando (as the protagonist is called in this version) first hears his falcon
crying out with hunger, and then climbs the ramparts of his castle, from whence he sees the
strange ship approaching. This version, more than any of the others, highlights the extent of
the confusion in this part of the ballad: here the hunt motif is only fleetingly touched upon
by means of the reference to the falcon, and is developed no further. Indeed, the fact that
the falcon has not yet been fed indicates quite clearly that a hunt has not yet taken place,
although one is undoubtedly pending. The hunt is passed over because its significance has
been lost and its relevance to the ‘plot’ is highly questionable: Fernando is abducted by
sailors on board ship (with no obvious link with the Wild Host), and seems to be taken from
within the safety of his own castle (which allows no room for the hunt as a threshold to a
supernatural world). This version is a prime example of how the function and meaning of
such motifs may change radically over the centuries, their significance sometimes being lost
altogether with the passage of time. Nevertheless, remnants of these ideas continue to
survive in most versions of the ballad, and, as we begin to understand some of their (albeit
unconsious) connotations in folk belief, it becomes easier to understand how the
transformation of this ballad from adventure story to tale of enchantment might have taken
place.

3. The Boat
The description of the boat gives added weight to the notion that this ballad tells of a
supernatural encounter. Its workmanship is in keeping with descriptions of supernatural
boats elsewhere, which are frequently described as standing out from the ordinary, either in
workmanship, in their means of propulsion and navigation, or both. Certainly, a ship with
gold anchors, silk sails and rigging made of sendal stands out from the ordinary, and is apt
to conjure images of wondrous origins.

46
 Huw Lewis

It may also be relevant that two different words  navío (ship) and galera (galley) 
are used to refer to the ship in its various extant texts. The former, or some variant thereof
(nave; nao) has been in common use in Castilian since earliest times; certainly, nave was
used by c. 1140, and navío by 1275. 17 Galera, on the other hand, may only be traced to the
second quarter of the fifteenth century or, in the form galea, to the beginning of the
thirteenth, although it was known in Catalan by 1120. 18 If our sources are correct, in
medieval times these two words were used in very different contexts: according to
Corominas, galera was usually used of a merchant ship, while Martín Alonso indicates that
it was used predominantly of a ship powered by oars (and Menéndez Pidal’s twentieth-
century Sephardic text indeed mentions the use of oars).19 Navío seems to be a word used
for a fighting ship;20 certainly, it is a large ship with sails, capable of withstanding very
harsh weather.21 Furthermore, it is likely that galera  derived from the Greek galea and
probably passing into Castilian from French and Catalan 22  had more currency on the
Mediterranean seaboard than the Atlantic, where navío would have been more common. In
both French and Italian the word for a galley (galère and galea respectively) was
synonymous with the idea of a prison ship, and of course in 1605 Cervantes himself depicts
his most famous creation, Don Quixote, freeing prisoners bound for the Spanish galleys
from their chains. It may not be a total coincidence and irrelevance, therefore, that Arnaldos
should himself be clapped in irons by his abductors. The use of galera in certain versions of
this ballad (especially since this is the term used by the sailor in his song), may therefore be
a later development, which would provide a further indication that the text of the sailor’s
song did not form part of an earlier version of the ballad. Alternatively it may simply be a
product of Arnaldos’s assimilation into Mediterranean tradition, a possibility further
strengthened by the sailor’s reference in his song to Moors, whose threat would have been
felt most strongly on the Mediterranean seaboard. In either case, it seems likely that
references to the ship as a galley are comparatively late additions to the text, as is, in all
probablility, the sailor’s enchanting song. This brings us to a consideration of other matters
relating to this haunting melody.

4. The Song
Reminiscent of the song of the Sirens of Classical tradition in its ability to enchant and
captivate those who hear it, this song is, nevertheless, different in many respects: it is sung
by a single, male sailor rather than by a group of female enchantresses; and it is sung by
someone on board ship, captivating someone who is on dry land, rather than vice-versa. In
those versions of the ballad where the text of the song is included, it completely loses its
supernatural connotations and becomes a much more banal incantation beseeching God to
protect the ship against the dangers of the sea. Elsewhere, however, the very ambiguity
caused by its omission adds to its haunting attraction, and, even today, the reader can still

17
Corominas, J., and J.A. Pascual, Diccionario crítico etimológico castellano e hispanoamericano, 6 vols
(Madrid: Gredos, 1980), IV, 219.
18
Corominas and Pascual, Dicionario crítico, III, 34-35.
19
Alonso, Martín, Diccionario medieval español (Salamanca, 1986), s.v. ‘galera’.
20
Alonso, Martín, Diccionario medieval español, s.v. ‘navío’.
21
Covarrubias, Sebastián de, Tesoro de la lengua castellana o española (Barcelona, 1943), s.v. ‘navío’.
22
Corominas and Pascual, Diccionario crítico, III, 34-35.

47
The Singer and the Scribe

feel, like Arnaldos, the burning desire to learn the sailor’s song and to surrender to its
soothing melody. In analogous fashion, Floriseo, both in the prose romance and in its ballad
counterpart, is lulled to sleep by the enchanting song of a damsel who approaches in the
marvellous boat, and who ultimately derives her power from the enchantress, Queen
Laciva.
The song in Arnaldos is also closely parallelled in the Ballad of Count Olinos,
reinforcing theories concerning their common origins and connections:23
¡Conde Olinos, conde Olinos Count Olinos, Count Olinos
es niño y pasó la mar! was a child and crossed the sea!
Levantóse conde Olinos Count Olinos arose
mañanita de San Juan; the morning of St John’s day;
llevó su caballo al agua, he took his horse to the water
a las orillas del mar. on the shores of the sea.
Mientras el caballo bebe, While the horse drank
él se pusiera a cantar: he began to sing
— Bebe, bebe mi caballo; — Drink, drink my horse;
Dios te me libre de mal, may God keep you from harm,
de los vientos rigurosos from strong winds
y las arenas del mar. and the sands of the sea.
Bien lo oyó la reina mora The Moorish queen heard him
de las altas torres donde está: from the high tower where she dwells.
— Escuchad, mis hijas todas; — Listen, all my daughters;
las que dormís, recordad, those of you that are asleep, awake,
y oirédes a la sirena and you will hear the siren,
24
cómo canta por la mar. how he sings out at sea.

The wording here and in Arnaldos is so similar as to make it imp ossible not to conclude
that some interference has taken place (it should be noted that conclusion of the version of
Arnaldos attributed to Juan Rodríguez del Padrón in the Cancionero de Londres borrows
extensively from the Ballad of Count Olinos, proving that connections were certainly made
between these particular two texts). However, it is also possible that the Olinos ballad has
borrowed from Arnaldos: it makes no sense that Olinos should take his horse to the
seashore to drink, nor that he should ask for protection for his horse from winds and sand;
moreover, the Moorish queen clearly believes that the song is coming from somewhere out
at sea, not from Olinos. Is this, therefore, yet another confused version of the events in
Arnaldos, where the protagonist and his horse are on the seashore when he hears the sailor’s
song? In all events it is clear that both ballads depend heavily on the enchanting power of
the song as a means of capturing the attention of all concerned, and thus, in both cases, the
song is vital in the development of the ‘plot’. Without it, there can be no abduction; without
it, there can be no love interest; without it, there is no ballad.

23
Entwistle, W.J., ‘El Conde Olinos’, Revista de filología española, 35 (1951), 237-48.
24
RA vol. 2, 554.

48
 Huw Lewis
5. Arnaldos’s Sleep
In our ballad, the effect of the wind or of the waves could be regarded as an attempt to
explain the protagonist’s sleep rationally; but they could also be regarded as instruments of
some unseen supernatural power. The truncated version of the ballad tells us that the sailor
is capable of controlling animals and the elements alike; whether it is his song or the sound
of the wind and waves that induce Arnaldos’s sleep, there can be no doubt that the sailor is
the controlling force, and all doubts in this respect must be removed by comparison of this
ballad with the prose Floriseo and its ballad counterpart, where the soporific effect of the
damsel’s song is left beyond doubt.
Sleep, especially when it is induced by some sort of enchantment is, like the forest, the
hunt, or the seashore, a liminal milieu that acts as a boundary between the world of mortals
and the otherworld (be the latter a land of supernatural beings, or the afterlife). In the same
way that the hunt takes the protagonist into the ‘wilderness’, the no-man’s land between this
world and the next, and as water, too, is depicted as a half-way house between the world of
mortals and the beyond (traditional riddles often refer to the propitious properties of boats,
which also represent half-way states, for anyone in a boat is neither in water nor on dry
land), 25 so sleep is regarded as a middle stage between this world and the next,
consciousness and death. Thus it comes as no surprise that, especially in later, euhemerised
accounts, sleep is frequently the catalyst for an encounter with the supernatural otherworld,
pagan or Christian, either through a vision, or by direct, physical contact, as is the case here.
Arnaldos’s inexplicable loss of consciousness when he is in a most vulnerable position
therefore further intensifies the supernatural air of this poem, and the idea that it relates an
encounter with something otherworldly.
As we have seen, all the classic elements of an otherworld tale are patently present in El
romance del Conde Arnaldos: the hunt that first brings the protagonist into contact with the
supernatural; the boat that carries him away; the enchanted sleep that marks his transition
from one state to another, possibly from the immediately recognisable reality of the world
of mortals to the dream-like, ‘unreal’ surroundings of the otherworld; the strange animals
and beautiful music that so often act as a marker indicating the presence of the supernatural.
But these elements are arranged in a fashion quite unlike any that can be found elsewhere.
Certainly, the theme of the Wild Host is echoed in some of the tale’s structure, but it is by
no means the only element which has contributed to the evolution of this particular ballad.
The echoes of a traditional account of a hero’s journey to an otherworld abode are brought
to the fore by comparing this text with its parallel, Floriseo, but corruption and confusion
with the theme of abduction (embodied in the Wild Host) may have brought about several
important changes to the text, or to its interpretation in popular belief, not least of these
being the substitution of a male sailor for what might originally have been a female
messenger. With the passage of time even the Wild Host’s deadly associations were lost
from memory, causing yet more disjointedness and confusion, and giving rise to the
addition of new but highly unsatisfactory endings. As the ballad evolved, associations were
forged, modified and discarded, forever changing the significance of the ‘original’ text,

25
Sims-Williams, Patrick, ‘Riddling Treatment of the “Watchman Device” in Branwen and Togail Bruidne Da
Derga’, Studia Celtica, 12/13 (1977-8), 83-117.

49
The Singer and the Scribe

Clearly, a highly complex and confused web of beliefs and associations underlies this
ballad, ones which, in spite of our modern ignorance of much of their original connotations,
continue to cojnure-up images of the strange and the unearthly, and so ensure the
continuing status of this ballad as the one of the jewels of the Spanish ballad tradition. As
Antonio Gala asks in his 1994 volume, ‘What was the significance of that song to the
sailor?’ It may be impossible for us to know the full answer to this question, but for us the
mysterious song embodies a myriad of possibilities, brought into being by the evolution of
a song born in the oral tradition, and fossilised for future frustration and delight in a
truncated text that has fired the imaginations of audiences from the sixteenth century to the
present day.

50
 A Morte do Rei D. Fernando and Floresvento:

two rare Portuguese epic ballads *

Manuel da Costa Fontes

As the citations in the works of about fifty early authors demonstrate, ballads were very
popular in Portugal during the Middle Ages and the Renaissance,1 but, unlike what
happened in Spain, no one bothered to compile them. This confers special importance on
the modern Portuguese ballad tradition, for it provides the only adequate documentation of
what existed in the early period. Almeida Garrett began to investigate it in 1823, and
thousands of ballads have been collected since then.2 In Brazil, the first collection is a

*
The following abbreviations are used for works regularly cited in this study:
C1344 = Crónica Geral de Espanha de 1344, ed. by Luis Filipe Lindley Cintra, 4 vols (Lisbon: Academia
Portuguesa da História, 1951-90).
Cantos = Braga, Teófilo, Cantos Populares do Arquipélago Açoriano (Oporto: Livraria Nacional, 1869).
CVR = Crónica de Veinte Reyes (partly ed. in Menéndez Pidal, Ramón, Reliquias de la poesía épica española,
acompañadas de ‘epopeya y Romancero I’, ed. by Diego Catalán, 2nd ed. (Madrid: Gredos, 1980), pp. 240-
56).
Ferré = Ferré, Pere, Vanda Anastácio, José Joaquim Dias Marques, and Ana Maria Martins, Romances
Tradicionais ([Funchal]: Câmara Municipal, 1982).
PCG = Primera crónica general de España, ed. by Ramón Menéndez Pidal, with a study by Diego Catalán, 3rd
ed., 2 vols (Madrid: Gredos, 1977).
Primav. = Wolf, Fernando J., and Conrado Hofmann, Primavera y flor de romances, ed. by Marcelino Menéndez
Pelayo, in Antología de poetas líricos castellanos, 8, “Obras Completas,” 24 (Santander: Consejo Superior
de Investigaciones Científicas, 1945).
RPI = Fontes, Manuel da Costa, O Romanceiro Português e Brasileiro: Índice Temático e Bibliográfico (com uma
bibliografia pan-hispânica e resumos de cada romance em inglês) / Portuguese and Brazilian Balladry: A
Thematic and Bibliographic Index (with a Pan-Hispanic bibliography and English summaries for each text-
type). Selection and Commentary of the Musical Transcriptions by Israel J. Katz. Pan-European Correlation
by Samuel G. Armistead. (Madison, Wisconsin: The Hispanic Seminary of Medieval Studies, 1997).
VRP = Vasconcellos, José Leite de, Romanceiro Português, Acta Universitatis Conimbrigensis, 2 vols (Coimbra:
Universidade, 1958-1960).

1
Michaëlis de Vasconcelos, Carolina, Estudos sobre o romanceiro peninsular: romances velhos em Portugal
(Coimbra: Imprensa da Universidade, 2nd ed. 1934).
2
For a more detailed survey of the work undertaken since 1967, see RPI, 1, 29-32. See also Marques, José
Joaquim Dias, ‘Ballad Collecting and Ballad Research in Portugal’, in Ballad and Ballad Studies at the Turn
of the Century. Proceedings of the 30th International Ballad Conference (Bucharest: Deliana, 2001), pp. 142-
49; idem, La investigación sobre el romancero portugués (manuscript: 1967-2000). José Joaquim Dias
Marques read a draft of the present chapter, and I would like to thank him for his important observations.
The Singer and the Scribe

manuscript with eight versions compiled by Inácio Raposo in 1853, but, since it did not
reach print until much later, 3 it was Celso de Magalhães who published the first Brazilian
ballads in 1883. 4 So far, every state in the country has been investigated.5 The number of
ballads collected throughout the rest of the Portuguese-speaking world is negligible. There
are five text -types from Goa, probable fragments of a particularly popular poem, Bela
Infanta (RPI, I1),6 from Sri Lanka, two ballads in the creole dialect of Malacca, and a poor
version of Bela Infanta from the Cape Verde Islands (RPI, 1: 33-34), but, as far as I can
determine, nothing has been collected in the former colonies of Angola, Mozambique,
Guinea-Bissau, St. Thomas and Prince, Macau and Timor.
Since the ballad constitutes a pan-European phenomenon, except for a few text -types
created in Portugal, most of the early ballads in the country came through centrally located
Castile, which, besides exporting poems that had crossed the Pyrenees, contributed a good
number of ballads of its own creation. Thus, Luso-Brazilian balladry constitutes a branch
of the pan-Hispanic tradition as well. Within this context, Portugal represents a
conservative, lateral area closely related to the Northwestern block of the Iberian Peninsula,
which includes Galicia, Asturias and Leon, preserving some ballads that have disappeared
elsewhere, a few text -types that survive only in the rich tradition of the Sephardim, and a
good number of ballads that, although documented elsewhere, remain extremely rare. This
paper will focus on two such ballads, the epic A Morte do Rei D. Fernando, which
combines three text -types (RPI, A7 + A8 + A10) derived from the Castilian Cantar de la
muerte del rey don Fernando y cerco de Zamora (eleventh century), and Floresvento (RPI,
B10), a Carolingian poem derived from the French Floovent (twelfth century). In addition
to their epic origin, both poems reflect intermediate, condensed versions of the lengthy
songs on which they are based.7

I.
According to epic sources, King Ferdinand I divided his kingdom among his children as he
lay dying in 1065. Sancho inherited Castile, Alfonso received León, and Galicia went to

3
Ribeiro, Joaquim, and Wilson W. Rodrigues, ‘Romanceiro tradicional do Brasil (textos do século XIX)’, in 1.o
Congresso Brasileiro de Folclore. Anais, 3 vols (Rio de Janeiro: Ministério das Relações Exteriores, 1952-
53), 2, 23-112.
4
Magalhães, Celso de, A Poesia Popular Brasileira, ed. by Bráulio do Nascimento (Rio de Janeiro: Biblioteca
Nacional, 1973).
5
RPI, 1: 32-33; Nascimento 1972.
6
I refer to RPI’s classification next to ballad titles because, in addition to samples, that catalogue provides Luso -
Brazilian and pan-Hispanic bibliographies for each text-type, including pan-European correlations by Samuel
G. Armistead and musical selections by Israel J. Katz.
7
In the pages that follow, I synthesise findings presented in four previous papers: Fontes, Manuel da Costa, ‘A
Sephardic Vestige of the Ballad ‘Floresvento’, La Cronica, 10 (1982), 196-201; idem, ‘The Ballad of
‘Floresvento’ and its Epic Antecedents’, Kentucky Romance Quarterly, 32 (1985), 309-19; idem, ‘The Ballad
“A Morte do Rei D. Fernando” and the “Cantar de la muerte del rey don Fernando y cerco de Zamora”’,
Anuario Medieval, 8 (1996), 108-51; idem, ‘Uma Nova Versão do Romance “A Morte do Rei D. Fernando”’,
Estudos de Literatura Oral, 2 (1996), 115-23.

52
 Manuel da Costa Fontes

Garc ía. The King’s two daughters, Elvira and Urraca, were given the cities of Toro and
Zamora, respectively. Being the oldest, Sancho felt that he should have inherited the whole
realm. He eventually managed to take León, Galicia, and Toro from his siblings, and was
killed in 1072, as he laid siege to Urraca in Zamora. 8
These events formed the basis for a lost epic known to us through its prosification in the
Crónica Najerense (twelfth century),9 the PCG (thirteenth century),10 the CVR (thirteenth
century),11 the Portuguese C1344, and several ballads.12 Menéndez Pidal thought that there
were two epics on the subject, one on the partition and another on the tragic consequences,
but, as Charles Fraker argued, there was only one poem. 13
Three of the ballads derived from this poem are Doliente estaba, doliente (á-o; Primav.,
no. 35), which portrays Fernando in his deathbed, surrounded by three legitimate sons and a
fourth, illegitimate one; Morir os queredes, padre (á-a; Primav. 36), where princess Urraca
receives Zamora after complaining to her dying father that he has forgotten her in his will;
and Afuera, afuera, Rodrigo (á-o; Primav. 37), which is based on a meeting between the
Cid, Sancho’s second-in-command, and Urraca, during the siege of that city.
In 1548, Martín Nucio published the three ballads together on folios 157r-158v, placing
Afuera first, and thus out of sequence.14 In 1550, however, he printed them in correct order,
linked with transitions that allow them to be read as one poem, for, despite the omission of
Sancho’s death, they synthesise the events that occurred between 1065 and 1072 in a broad,
coherent manner. I stress the transitions in question below:
Doliente se siente el rey In pain was the King,
esse buen rey don Fernando good King Fernando,
2 los pies tiene hazia oriente his feet facing the East,
y la candela en la mano a candle in his hand.
a su cabecera tiene At the head of the bed
arçobispos y perlados archbishops and prelates,
4 a su man derecha tiene and at his right
a sus fijos todos cuatro all four of his sons;

8
Historically, Urraca and Elvira’s inheritance was lordship over all royal monasteries in Castile and Leon, and,
rather than having inherited Zamora, Urraca probably started a rebellion against Sancho in that city. For the
historical background, see Reilly, Bernard E., The Kingdom of León-Castilla under King Alfonso VI (1065-
1109) (Princeton NJ: Princeton University Press, 1988), pp. 14-67; Reig, Carola, ‘El cantar de Sancho II y
cerco de Zamora’, Revista de Filologia Española (Madrid: Consejo Superior de Investigaciones Científicas,
1947), Anejo 37, pp. 7-27.
9
Cronica najerense, ed. by Antonio Ubieto Arteta, Textos Medievales, 15 (Valencia: Anubar, 1966).
10
Primera Crónica general de España, ed. by Ramón Menéndez Pidal, with a study by Diego Catalán, 2 vols
(Madrid: Gredos, 3rd ed. 1977).
11
CVR, pp. 240-56.
12
Clavero, Dolores, Romances viejos de temas épicos nacionales. Relaciones con gestas y cronicas (Madrid:
Ediciones del Orto, 1994), pp. 191-313.
13
Fraker, Charles F., ‘The Beginning of the Cantar de Sancho’, La Corónica, 19 (1990-91), 5-21.
14
Nuncio, Martín, Cancionero de romances impreso en Amberes sin ano, ed. by Ramón Menéndez Pidal (Madrid:
Consejo Superior de Investigaciones Científicas, 1945).

53
The Singer and the Scribe

los tres eran de la reyna three were the Queen’s,

y el vno era bastardo a bastard was one.
6 esse que bastardo era He was the richest,
quedaua mejor librado and so would remain,
arçobispo es de Toledo Toledo’s Archbishop,
maestre de Santiago Grand-Master of Santiago,
8 abad era en Çaragoça Abbot in Saragossa
de las Españas primado. and Primate of Spain.
Hijo si yo no muriera ‘If I wasn’t dying, my son,
vos fuerades padre santo you’d be Pope,
10 mas con la renta que os queda but with all this money
vos bien podreys alcançarlo. you still have hope.’
Ellos estando en aquesto In the middle of this
entrara Vrraca Fernando Urraca Fernando came in ,
12 y buelta hazia su padre and turning toward her father,
desta manera ha hablado. these words she spoke.

Morir vos queredes padre ‘My father, you’re dying,

san Miguel vos aya el alma may your soul find it’s rest!
2 mandastes las vuestras tierras You have given your kingdom
a quien se vos antojara as you thought best,
a don Sancho a Castilla you’ve offered Sancho
Castilla la bien nombrada proud ancient Castilla,
4 a don Alonso a Leon León to Alfonso,
y a don Garcia a Bizcaya Biscay to García,
a mi porque soy muger but I am a woman,
dexays me deseredada and left with no home.
6 yrme yo por essas tierras I’ll wander the kingdom
como vna muger errada a woman alone
y este mi cuerpo daria and offer my body
a quien se me antojara to all men I see,
8 a los Moros por dineros to Moslems for money,
y a los Christianos de gracia to Christians for free,
de lo que ganar pudiere and the money I make
hare bien por la vuestra alma. for your soul it shall be.’
10 Alli preguntara el rey, Then the King asked:
Quien es essa que assi habla? ‘What woman so speaks?’
Respondiera el arçobispo The Archbishop replied:
Vuestra hija doña Vrraca. ‘Your daughter Urraca’
12 Calledes hija calledes ‘Silence, my daughter,
no digades tal palabra those words should be spurned,
que muger que tal dezia a woman who says them
merescia ser quemada deserves to be burned!
14 alla en Castilla la vieja There’s a place in Castilla
vn rincon se me olvidaua that now I recall,
Çamora auia por nombre the town of Zamora
Çamora la bien cercada with fortified walls,

54
 Manuel da Costa Fontes

16 de vna parte la cerca el Duero the Duero runs by it,

de otra peña tajada it’s on a steep hill,
del otro la moreria and the old Moslem quarter
vna cosa muy preciada is beautiful still;
18 quien vos la tomare hija who captures it from you,
la mi maldicion le cayga. my curse is on them!’
Todos dizen amen amen All except for Don Sancho
sino don Sancho que calla. they all say ‘Amen!’
20 El buen rey era muerto As soon as the King died
Çamora ya esta cercada Zamora was besieged,
de vn cabo la cerca el rey the King from one side,
del otro el Cid la cercaua the Cid from the other.
22 del cabo que el rey la cerc On the King’s side
a Çamora no se da nada they did not give an inch,
del cabo que el Cid la cerca but on the Cid’s side
Çamora ya se tomaua the city began to fall.
24 Assomose doña Vrraca Urraca appeared,
assomose a vna ventana she leaned out of a window,
de alla de vna torre mocha from a tower without a spire
estas palabras hablaua. she spoke these words.

A fuera a fuera Rodrigo ‘Get away, proud Rodrigo,

el soberuio Castellano Castilian son!
2 acordarse te deunia You ought to remember
de aquel tiempo ya passado the times that have gone,
quando fuiste cauallero when you were armed
en el altar de Santiago a knight before the altar of St. James;
4 quando el rey fue tu padrino my father was your sponsor,
tu Rodrigo el ahijado you, Rodrigo, his godson.
mi padre te dio las armas My father gave you the arms,
mi madre te dio el cauallo my mother gave you a horse,
6 yo te calce las espuelas and I put on your spurs
porque fuesses mas honrrado so as to honor you even more;
que pense casar contigo I thought we would marry
mas no lo quiso mi pecado but that couldn’t be,
8 cassaste con Ximena Gomez for you married Ximena Gómez,
hija del conde Loçano daughter of Count Lozano.
con ella vuiste dineros With her you had riches,
comigo vuieras estado when you should be with me.
10 bien casaste tu Rodrigo A good marriage, Rodrigo,
muy mejor fueras casado but it wasn’t the best,
pexaste [sic] hija de rey you married a vassal
por tomar de su vasallo. and spurned a Princess.’
12 Si os parece mi señora ‘If you wish it, Urraca,
bien podemos destigallo I’ll abandon my wife.’
mi anima penaria ‘My soul would suffer
si yo fuesse en discrepallo. if I were to disagree.’

55
The Singer and the Scribe

14 A fuera a fuera los mios ‘Be gone now, be gone,

los de a pie y de a cauallo my soldiers and knights!
pues de aquella torre mocha From that tower without a spire
vna vira me han tirado an arrow has pierced me tonight,
16 no traya el asta hierro the shaft isn’t metal
el coraçon me ha passado. but it’s wounded my heart,
Ya ningun remedio siento and for all of my life
15
sino biuir mas penado. I’ll be hurt by the dart.’

The first transition is found at the end of Doliente (vv. 11-12). As Menéndez Pidal
demonstrated,16 rather than being invented by Nucio, those verses are traditional, for they
correspond to the following passage in the CVR: ‘Ellos en esto estando, entro la ynfante
doña Vrraca con todas sus dueñas por el palaçio, metiendo bozes e faziendo el mayor llanto
del mundo, llamando e diziendo: ¡Padre señor! ¿que fiz yo porque ansy finco deseredada?’
(247.14-16). [When they were in the middle of this conversation, Princess Urraca came into
the palace with all of her ladies, shouting and crying very loudly and saying: ‘My Father
and Lord, what did I do in order to be left without an inheritance?’]
The second transition appears at the end of Morir (20 -25). Wolf and Hoffmann refused
to accept it as being traditional and placed it in a footnote (Primav. 36, n. 6), but it is
extremely difficult to attribute those verses to Nucio or to anyone familar with the
chronicles. According to the transition, the siege of Zamora begins as soon as Fernando
dies (20), but the chronicles describe in detail the events that occurred between the death of
the king in 1065 and the siege of the city in 1072. The PCG dedicates no less than 16
chapters (814-29) to those events. Contrary to what happens in the ballad (23), in the
chronicles the Cid does not manage to take any part of Zamora. On the contrary, in an
effort to prevent the siege, Sancho sends him as emissary to his sister Urraca, offering to
exchange Zamora for other holdings. Rather than addressing the Cid from a turret, as in the
ballad (24-25), the princess receives him in her palace (PCG, 2: 507a.4-5). Anyone familiar
with the chronicles would have known better than to depart from them in such a radical
manner.
This clear process of condensation and novelización occurs as the events on which a
poem is based are half-forgotten and altered through oral transmission. In the words of
Menéndez Pidal, 17 in 1550 Nucio ‘no inventó de su cosecha esos versos de unión, sino que
los halló, juntamente con dichas variantes, en la versión que le servía de norma para
corregir el texto del Cancionero sin año’ [did not invent those linking verses on his own, but
rather he found them, together with the aforesaid variants, in the version that he used in
order to correct the text of the previous edition]. In sum, Nucio printed the three ballads as a
trilogy in 1550 because they existed as one poem in the oral tradition.

15
The original text is taken from Cancionero de romances (Anvers, 1550), ed. by Antonio Rodríguez-Moñino
(Madrid: Castalia, 1967); all translations accompanying quotations in this chapter are mine unles otherwise
indicated.
16
Menéndez Pidal, Ramón, Romancero hispánico (hispano-portugués, americano y sefardí), 2 vols (Madrid:
Espasa-Calpe, 1953).
17
Menéndez Pidal, Ramón, Estudios sobre el Romancero, ed. by Diego Catalán, in Obras completas, vol. 11
(Madrid: Espasa-Calpe, 1973), p. 120 .

56
 Manuel da Costa Fontes

The ballad that follows, which I have entitled A Morte do Rei D. Fernando, also testifies
to the traditional character of this trilogy, for it is still sung in Portugal as one poem. In
Madeira and the Azores, it survives thanks to its contamination with Silvana (RPI, P1), a
ballad about a girl whose father, who is often a king, asks her to go to bed with him. Due to
space limitations, I will summarise this part of the ballad.18 When the king tells the girl that
he wants to spend a night with her, she reminds him of the pains of hell, and he replies that
the Pope would forgive their sin. Informed of what is happening, the queen goes to her
husband’s bed instead, and, thinking that he is with his daughter, he is greatly surprised to
discover that she is no longer a virgin. The queen reveals her identity and names the three
children that she has already had, one of whom is ‘D. Pedro de Castilha’ (28b). Hearing
this, the king curses his daughter.19 Although the epic trilogy that follows is free from
contaminations, the first three of the four verses that correspond to Doliente (34-36) were
changed from the original á-o to í-a, no doubt because of the influence of the preceding
Silvana, which also rhymes in í-a:
34 Nisto o rei adoeceu Then the King fell ill,
para a morte, que morria, he was bout to die,
mandá’ chumar os seus filhos and sent for his children
para fazer a partida. in order to do the partition.
36 Q’ando chegou à sua casa, When she arrived at the house
seu pai na cama o tinha, her father was in bed,
c’os pés coma um defunto, his feet placed like a dying person’s,
com a candeia na mão. a candle in his hand.
38 —Senhor pai que ‘tás à morte, ‘Father, you’re about to die,
Dês te tome parte n’alma. may God have your soul.
Deixastes os vossos bens You’ve left everything
a quem vos não era nada, to people who were nothing to you,
40 e sendo a vossa filha and I, who am your daughter,
me deixastes deserdada. find myself without an inheritance.’
A João deixo as casas, ‘I’m leaving the houses to João,
a Pedro terras lavradas. the ploughed lands to Pedro.’
42 Caminhou dali Silvana, Silvana walked away
como louca, desvairada. like a mad, bewildered person.
—Eu vou é por esse mundo ‘I’ll go through the world
como pobre, desgraçada, as a poor, disgraced woman,
44 vou co’a minha roca à cinta, ‘I’ll go through the world
já que espada não m’é dada. as a poor, disgraced woman,
C’a minha roca à cinta, My distaff at the waist,
mulher não tem outra arma. it’s a woman’s only weapon,
46 Nem de preto, nem de branco, and neither black nor white men
de ninguém serei guardada. will show me any respect.’
—Oh que vozes são estas ‘What words are those,

18
What follows is drawn from a synthetic version. For a full account of the versions and the verses on which it is
based: see Fontes, ‘The Ballad A morte do rey D. Fernando’, pp. 112-13.
19
In this portion of the ballad, the existing versions exhibit no less than seven contaminations with other text-
types: see Fontes, ‘The Ballad A morte do rey D. Fernando’, pp. 114-20.

57
The Singer and the Scribe

que eu oiço tão desmudadas? disturbed words I hear?

48 Que palavras são aquelas, What words are those,
mulher tão desaustinada? and such a crazy woman?’
—É vossa filha Silvana, ‘It’s your daughter Silvana,
que a deixais deserdada. because you’re leaving her disinherited.’
50 —Deus se não lembre de mim, ‘May God forget me
se tal filha me lembrava, if I remembered that daughter;
que eu deixava-lhe Sambóia, I’ll leave her Sambóia,
cercada, mui bem cercada. Sambóia, with its fortified walls.
52 Por uma banda corre ouro, Gold runs by it on one side,
por outra prata lavrada. wrought silver on the other.’
—Pa que quer’aquela terra, ‘What do I want that land for?
se D. João ma tomara? Don João will take it from me.’
54 —Se D. João ta tomara, ‘If he t akes it from you
pro maldição lha deixara. my curse will be upon him.’
Mas ‘inda o rei não era morto, The King was not yet dead,
já Sambóia era tomada, Sambóia was already taken,
56 com duzentos cavaleiros and completely surrounded
por toda a roda cercada. by two hundred knights.
—Tu não le lembras, Rodrigues, ‘Don’t you remember, Rodrigues,
daquele tempo passado, the times that have gone,
58 qu’o rei era tê padrinho when the King sponsored you,
e tu eras seu afilhado, you were his godson,
o meu pai te deu uma espada my father gave you a sword,
e minha mãe deu-te o cavalo, my mother gave you a horse,
60 e eu dei-te as esporas de ouro I gave you golden spurs
para te ver aumentado, so as to increase your honor,
e p’agora vires buscar and now you come to take away
todo o meu novo estado? all of my new property?’

The sixteen versions used to put together the preceding synthetic version of this ballad are
from the isolated archipelagoes of Madeira 20 and the Azores,21 two conservative lateral
areas in relation to the rest of Portugal. Estácio da Veiga published a synthetic version from
the Algarve as well, but I have not included it because it represents a different subtype.22

20
Ferré, nos. 1-2, pp. 246-48, 250, 254; Purcell, Joanne B., Novo Romanceiro Português das Ilhas Atlânticas, 1.,
ed. by Isabel Rodríguez-García and João A. das Pedras Saramago (Madrid: Seminario Menéndez Pidal-
Universidad Complutense, 1987), 2.1-4; Marques, José Joaquim Dias, ‘Imagens e Sons do Romanceiro
Português’, in El Romancero: Tradición y pervivencia a fines del siglo XX. Actas del IV Coloquio
Internacional del Romancero (Sevilla-Puerto de Santa María -Cádiz, 23-26 de Junio de 1987), ed. by Pedro
M. Piñero, Virtudes Atero, Enrique J. Rodríguez Baltanás and María Jesús Ruiz (Seville-Cádiz: Fundación
Machado-Universidad de Cádiz, 1989), pp. 381-412 (at pp. 388-389); Fontes, ‘Uma nova versão do romance
A Morte do Rei D. Fernando’, pp. 120-121.
21
Cantos 4; Purcell 1987, 2.5-6.
22
Veiga, S[ebastião] P[hilippes] M[artins] Estácio da, Romanceiro do Algarve (Lisbon: Joaquim Germano de
Sousa Neves, 1870), pp. 19-22. Veiga tampered considerably with his version, but Dias Marques, who
discovered the author’s original papers, demonstrated that he based himself on a traditional version: Marques,

58
 Manuel da Costa Fontes

The contamination with Silvana may have occurred because, in the broadest terms, that
ballad is about a father-daughter relationship, and, to a great extent, so is the early trilogy,
where Urraca plays a central role. This broad theme could have triggered the contamination
by itself. Although there is no hint of incest in the trilogy, there were rumours of such a
relationship between the impetuous Urraca and her brother Alfonso.23 These rumours could
also have played a role. Whatever the case may have been, the shared references to the
Pope, three children and a curse may have contributed to the contamination as well.
Whereas the king tells Silvana that the Pope can forgive their sin, in Doliente Fernando says
to his illegitimate son, an archbishop, that he would eventually make him Pope if he were to
live longer (9). When Silvana’s mother reveals her identity, saying that she has already had
three children and that one of them is ‘D. Pedro de Castilha’ (28b), this echoes Morir,
which, besides naming three children, stresses Sancho’s association with Castile (3).
Finally, the king’s curse in Silvana also echoes Fernando’s curse against anyone who dares
to take Zamora from Urraca in Morir (18).
The portion of the Portuguese ballad that corresponds to Doliente has been condensed to
only four verses (34-37). Since it is not possible to include here a detailed analysis of the
transformations that have taken place,24 I will stress only the details in which the
Portuguese ballad seems to follow the old Cantar better than Nucio’s version.25 As Joanne
Purcell indicated,26 the verse ‘mandá’ chumar os seus filhos / para fazer a partida’ [he sent
for his children in order to do the partition (35)] reflects the Cantar much better. According
to the epic, the king ‘mando llamar a sus fijos e partioles el reyno ante que muriese’ [sent

José Joaquim Dias, ‘Subsídios para o Estudo do Método Editorial de Estácio da Veiga no Romanceiro do
Algarve’ in Actas do 1.o Encontro Sobre Cultura Popular (Homenagem ao Prof. Doutor Manuel Viegas
Guerreiro), 25 a 27 de Setembro de 1997, ed. by Gabriela Funk (Ponta Delgada: Universidade dos Açores,
1999); for the original itself, see pp. 274-76. Contrary to the Insular renditions, this one opens with ‘A Morte
do Príncipe D. João’ (RPI, C5), which is about the death of Prince John, son of the Catholic Monarchs. For
additional studies of Veiga’s papers, see Marques, José Joaquim , ‘Os Manuscritos do Romanceiro do Algarve
de Estácio da Veiga Existentes no Museu Nacional de Arqueologia’, O Arqueólogo Português, 11-12 (1993-
94), 153-73; idem, ‘Contribuição para o Estudo do Romanceiro do Algarve de Estácio da Veiga à Luz de
Manuscritos Inéditos (Tese de Aptidão Pedagógica e Capacidade Científica, Unidade de Ciências Exactas e
Humanas, Universidade do Algarve, 1997). The latter includes a splendid study of this ballad (pp. 52-85).
See also his forthcoming ‘O Dom Rodrigo.’
23
Lévi-Provençal, E., and Ramón Menéndez Pidal, ‘Alfonso VI y su hermana la infanta Urraca’, Al Andalus, 13
(1948), 157-66.
24
Fontes, ‘The Ballad A Morte do Rei D. Fernando’, pp. 124-34.
25
The version from the Algarve refers to a will rather than to a partition because of the contaminatio n with A
Morte do Príncipe D. João: ‘Fazer quero testamento / desta pobre holanda minha’, Marques, ‘Subsídios para o
Estudo’, p. 275.
26
Purcell, Joanne B., ‘Recently collected ballad Fragments on the Death of Don Fermando I’, in Portuguese and
Brazilian Oral Traditions in Verse Form / As Tradições Orais Portuguesas e Brasilieras em Verso , ed. by
Joanne B. Purcell, Samuel G. Armistead, Eduardo M. Dias and Joanne E. March (Los Angeles: University of
Southern California: 1976), pp. 158-167(at p. 164). In this paper, the author summarises the most essential
findings presented in her doctoral dissertation, Purcell, Joanne B, ‘The Cantar de la muerte del rey don
Fernando in Modern oral Tradition: Its relationship to sixteenth-century romances and medieval chronicles’
(PhD dissertation, University of California Los Angeles, 1976).

59
The Singer and the Scribe

for his children and divided the kingdom among them before dying (CVR 242.1-2)], but, at
this point, Nucio’s version merely states that the king’s four sons are at his bedside (4), and
refers to the partition only in an implied manner, indicating that the illegitimate son did
better than his brothers (6).
The portion that corresponds to Morir (38-56) is much better preserved, and may reflect
the Cantar better than Nucio’s version in three instances. As Joanne Purcell pointed out,27
the verse in which the princess complains to her father that he has left her out of his will,
‘Deixastes os vossos bens / a quem vos não era nada’ [You’ve left everything to people
who were nothing to you (39)], seems closer to the chronistic ‘partistes vos los reynos e a
mi non me distes nada, e finco desanparada e lazrada’ [you partitioned the kingdom, gave
me nothing, and so I’m being left destitute and in misery (CVR 248.1-2)] than Nucio’s
corresponding verse, where Urraca says ‘mandastes las vuestras tierras / a quien se vos
antojara’ [you’ve given your kingdom as you thought best (2)].
In the second instance, the Portuguese poem specifies that the king realises that the
speaker is an extremely upset woman who shouts (47-48), whereas in its early congener he
merely asks who has just said such words (10). Here the modern poem may reflect the
Cid’s advice to Urraca in the prosified epic, where she asks him for his help in convincing
her father to leave her an inheritance. The Cid tells her to follow him shortly after he enters
her father’s chambers, ‘con vuestras dueñas faziendo muy grant llanto’ [crying very loudly
with your ladies (CVR 246.17)]. Urraca puts the Cid’s advice into practice, for she enters
with her ladies-in-waiting ‘metiendo vozes e faziendo el mayor llanto del mundo’ [shouting
and crying very loudly (CVR 247.15)]. This parallels the ‘vozes tão desmudadas’ [disturbed
words (47b)] that the king hears in the Portuguese ballad. Since these ‘vozes’ are very
specific and reflect the circumstances of Urraca’s stormy entrance much better than Nucio’s
version, the parallel is not likely to be a matter of pure coincidence. In all probability, at
this point the modern ballad preserves an epic detail absent from its early congener.
The third instance in which this part of the modern poem is closer to the Cantar occurs
in the answer to the king’s query. In Nucio’s version, the archbishop replaces the Cid and
replies to his father,28 revealing that the distraught woman is Urraca (11). In the modern
poem, an unidentified speaker adds that Silvana is being left without an inheritance (49),
thus paralleling the prosified Cantar, where the Cid replies: ‘Señor, es vuestra fija doña
Urraca que finca deseredada e pobre’ [Sir, it’s your daughter Urraca, who is being left poor,
without an inheritance (CVR 248.4-5)].
The Portuguese poem adds a new verse in which the princess protests to her father that
‘D. João’ will take the land that he is leaving her (53). It is only then that the king curses his
son if he should do so (54). D. João, of course, stands for Sancho. Since the princess’s
question makes good sense in view of the king’s answer, which in Nucio’s version appears
somewhat gratuitous, the modern verse could easily derive from another early version.

27
Purcell, ‘Recently collected ballad Fragments’, p. 164.
28
As Menéndez Pidal explained, Estudios sobre el romancero, pp. 119-20, the change probably came about
through oral transmission, being triggered by the union of Doliente and Morir, because Doliente mentioned
the archbishop, not the Cid.

60
 Manuel da Costa Fontes

The Portuguese ballad compresses the already condensed transition that reached print in
1550 (20-25) even more, reducing it to only one verse that merely states that ‘Sambóia’
(Zamora) was taken as soon as the king died (55). 29 The modern ballad also adds one verse,
according to which the besieging army consisted of two hundred knights (56). These
knights may parallel the fifteen knights who, in the prosified epic, accompany the Cid in his
embassy to Urraca: ‘et fuesse pora çamora con XV de sus caualleros’ [and he went to
Zamora with fifteen of his knights (PCG, 2: 506b.30-31)]; ‘E foisse logo pera a villa cõ XV
cavalleiros’ [And he went to the city right away, with fifteen knights (C1344, 3: 376.15)].
Rather than having introduced an innovation, the Portuguese poem probably retains here yet
another epic detail absent from the sixteenth-century version.
Like its early congener, the Portuguese poem then continues with Afuera, which is
based on the meeting between Urraca and the Cid, condensing it from seventeen to a mere
five verses (57-61). In the Insular renditions, the princess reminds the hero of their past
friendship, but the lovers’ quarrel that follows in Nucio is omitted.30 The version from the
Algarve, however, preserves this passage:
Casaste com Ximena Gomes, filha do conde Lousã,
com ela terás dinheiro, comigo foras honrado.
—Como isso é assim, eu ta mando já matar.
—Não permita Deus do céu, nem o seu sangue sagrado;
casamento que Deus ajunta, que por mim seja apartado.

[‘You married Ximena Gomes, daughter of Count Lousã,

you may have riches with her, but I would’ve brought you more honour.’
‘Since it’s that way, I’ll have her killed at once.’
‘May not God in Heaven or His holy blood permit
that a marriage that He blessed be undone on my account.’] 31

Contrary to Nucio’s version, where Urraca agrees to the Cid’s apparent offer to divorce
his wife —the cryptic ‘destigallo’ in v. 12a seems to be a misprint for ‘desligallo’32 — the
hero offers to kill his wife, but the princess refuses, saying that she does not want to be

29
This transition is also present in the version from the Algarve: ‘Noutro dia de manhã / Samora estava cercada’
[The morning after, Zamora was besieged], Marques, ‘Subsídios para o Estudo’, p. 275.
30
At this point, the PCG merely states that Urraca and the Cid were raised together (2: 507a.9-11), but, as
Armistead demonstrated, other chroniclers go out of their way in order to deny the existence of an affair
between the two, which suggests that oral versions of the Cantar brought up the scandal preserved in Afuera :
Armistead, Samuel G., ‘“The Enamored Doña Urraca” in Chronicles and Balladry’, Romance Philology, 11
(1957-58), 26-29; see now also idem, La tradición épica de las ‘Mocedades de Rodrigo’ (Salamanca:
Universidad, 2000), pp. 49-52,
31
Marques, ‘Subsídios para o Estudo’, p. 276. I have added the punctuation and changed the original’s
arrangement from seven-syllable lines into two hemistichs in order to save space.
32
This is how the word appears in another sixteenth-century version in Menéndez Pidal, Ramón, Pliegos poéticos
españoles en la Universidad de Praga, Serie Conmemorativa, 7, 2 vols (Madrid: Joyas Bibliográficas, 1960)
2, 121-24 (at p. 123).

61
The Singer and the Scribe

responsible for breaking up his marriage. As Dias Marques demonstrated (1997, 73), 33 the
verses ‘comigo foras honrado’ [but I would’ve brought you more honour] and ‘Não permita
Deus do céu,’ [May not Go d in heaven permit] which Nucio omits, appear in two other
early versions.34 This confirms that the modern ballad is independent from Nucio’s trilogy,
depending on another traditional version instead. Having omitted the affair, the Insular
versions close with the princess asking ‘Rodrigues’ how he dares to dispossess her of her
property in view of what she and her parents have done for him in the past (61). Since this
verse makes perfect sense, it could well derive from another early version.
In sum, Nucio’s trilogy, which represents a condensed form of the Cantar de la muerte
del rey don Fernando y cerco de Zamora, does not seem to have been reprinted elsewhere.
Rather than depending on Nucio, the Portuguese ballad reflects a different version, for,
besides including verses that he omits, it follows the Cantar more closely at times,
including additional epic details. Condensation, of course, is one of the main characteristics
of oral transmission. As we will see in the pages that follow, this transformation of a
lengthy epic into a much shorter poem is not unique.

II.
The extremely rare ballad of Floresvento (RPI, B10), which derives from another epic, the
French Floovent (twelfth century), including many elements from an Italian prose
derivative, Fioravante (fourteenth century), survives in two archaic Portuguese lateral
areas, Trás-os-Montes and the Azores.35 It may also have been sung in Brazil at one time,
for there is a version of Veneno de Moriana (RPI, N1), which opens with a three-verse
contamination from Floresvento.36 Since the Galician four-verse fragment collected in
Castiñeira (Ourense)37 depends upon the versions from Trás-os-Montes, and the

33
‘Contribuição para o Estudo do Romanceiro do Algarve de Estácio da Veiga à Luz de Manuscritos Inéditos’
(Tese de Aptidão Pedagógica e Capacidade Científica, Unidade de Ciências Exactas e Humanas, Universidad
do Algarve: 1997).
34
Timoneda, Juan, Rosa española , in his Rosas de romances (Valencia, 1573), ed. by Antonio Rodríguez-Moñino
and Daniel Devoto (Valencia: Castalia, 1963); Escobar, Juan de, Historia y Romancero del Cid (Lisboa,
1605), ed. by Antonio Rodríguez-Moñino, Introduction by Arthur L.-F. Askins (Madrid: Castalia, 1973).
Although Escobar’s Historia y romancero del Cid dates from 1605, there is no question that his version is
much older.
35
For a full bibliography, see RPI, B10; I will point out the main differences between the versions from these two
regions in my discussion.
36
Lopes, António, Presença do Romanceiro. Versões Maranhenses (Rio de Janeiro: Editora Civilização Brasileira,
1967), p. 231. It is more closely related to the Azorean renditions, for, unlike the Continental versions, it states
that the hero dishonoured seven maidens on Christmas Eve.
37
Valenciano, Ana. Os romances tradicionais de Galicia. Catálogo exemplificado dos seus temas, Romanceiro
Xeral de Galicia, 1 (Madrid-Santiago de Compostela: Publicacións do Centro de Investigacións Lingüisticas e
Literarias Ramón Piñeiro-Fundación Ramón Menéndez Pidal, 1998) No. 11.

62
 Manuel da Costa Fontes

contamination found in the Sephardic versions of El Conde Arnaldos is also incomplete, 38

the ballad is essentially Portuguese today.
In the French epic, Clovis, King of France, places his son Floovent in the care of
Senechaul, duke of Burgundy. As a prank, Floovent cuts off the old duke’s beard while he
sleeps. This was considered to be a serious crime at the time, 39 because all honest men
wore one (‘Adonc estoient tuit li prodome barbez,’ [All worthy men wore a beard in those
days] v. 63), 40 and thieves, being forced to shave them off, were ashamed of appearing in
public (vv. 69-70). Since the importance attached to a man’s beard was eventually lost —
Fioravante, where the duke’s name is changed to Salardo, the insult does not seem to be as
great41 — the ballad substituted the hero’s felony for a series of readily understood crimes.
He destroys churches, cas tles and cities, murders babies and priests, and dishonours several
royal maidens:
—Cruel vento, cruel vento, ah! roubador maioral!
2 Derrubaste três igrejas, todas três em Portugal.
Derrubaste três castelos, no reino de Portugal.
4 Derrubaste três cidades, todas três em Portugal.
Mataste três inocentes, todos três por baptizar.
6 Mataste três sacerdotes, revestidos no altar.
Desonraste três donzelas, todas três de sangue real.42

[“Cruel Vento, Cruel Vento, oh you biggest thief!

You demolished three churches, all three in Portugal.
You demolished three castles in the kingdom of Portugal.
You demolished three cities, all three in Portugal.
You killed three innocent children, all three still not baptised.
You killed three priests covered before the altar.
You dishonoured three maidens, all three of royal blood.”]

Besides increasing each crime from three to sevenfold, the Azorean versions state that they
are perpetrated on Christmas Eve (‘Caminhou Flores e Ventos / uma noite de Natal, //

38
Armistead, Samuel G., with the collaboration of Selma Margaretten, Paloma Montero and Ana Valenciano, El
romancero judeo-español en el Archivo Menéndez Pidal (Catálogo-índice de romances y canciones), 3 vols
(Madrid: Cátedra-Seminario Menéndez Pidal, 1978), H15; Fontes, ‘A Sephardic Vestige of the Ballad
Floresvento ’, pp. 196-201.
39
Menéndez Pidal, Ramón, Cantar de Mio Cid. Texto, gramática y vocabulario, 3 vols (3rd ed. Madrid: Espasa-
Calpe, 1954-56), 2, 498-99; Burt, John R., ‘Honor and the Cid’s Beard’, La Corónica, 9 (1981), 132-37.
40
This and all the references that follow are from La Chanson de Floovant , ed. by F.H. Bateson (Loughborough:
n.ed., 1938).
41
The boy cuts off the duke’s beard, but the text does not establish a difference between bearded, honest men and
beardless thieves (147). This and all the references that follow are to Andrea da Barberino’s version in I Reali
di Francia, ed. by Giuseppe Vandelli and Giovanni Gambarin (Bari: Laterza & Figli, 1947). Fioravante = pp.
143-244.
42
Sources: Vv. 1-2: Martins, P, and A. Firmino, Folclore do Concelho de Vinhais, 2 vols (Coimbra; Lisboa:
Impressa da Universidade; Imprensa Nacional, 1928-39), 1, 219; v. 3: VRP 36; v. 4: VRP 37; v. 5: VRP 35; v.
6: VRP 34; v. 7.

63
The Singer and the Scribe

desonrou sete donzelas, / todas de sangue real’ [Flores e Ventos left on a Christmas Eve, he
dishonoured seven maidens, all of royal blood] Cantos, 18), and add gambling to the crimes
committed by the protagonist (‘Jogou cem dobrões de ouro, / marcados e por marcar [He
bet a hundred gold doubloons, both marked and unmarked] ibid.). The fact that the Church
forbade sexual relations, even between husband and wife, during Lent and during religious
festivals (Armistead and Silverman 1977, 193-94), 43 increases the sin involved in the seven
deflowerings tremendously. Since the references to castles, a city (Rome) and gambling
already appear in the Sephardic contamination of Arnaldos, the substitution of the beard
episode for a series of crimes probably goes back to the years that preceded the Spanish
exodus, that is, to the end of the fifteenth century:
¡Quien tuviera tal fortuna sobre las aguas de la mar,
2 como el infante Fernando mañanita de San Juan,
que ganó siete castillos a vuelta de una cibdad!
4 Ganara cibdad de Roma, la flor de la quistiandad;
con los contentos del juego saliérase a passear.

[Upon the high seas to have such good fortune

as Prince Fernando on St. John’s day morning!
He won seven castles that surrounded a city,
he won the city of Rome, flower of Christendom,
44
5 and happy with his winnings he went out for a walk.]
(Bénichou 1968, 207)

As soon as he discovers what his son has done to Senechaul, the king, furious, determines
to punish him with death, but changes his mind thanks to the intercession of the queen,
banishing him for seven years instead: ‘Dame, ce dit le rois, je vos an doins le chief, /
Entreci et .VII. anz mar i metrai le pié’ [‘Lady, says the king, I grant you his head, but it
will go ill for him if he sets foot here these next seven years (vv. 139-40)]. In the Azorean
renditions the mother, besides intervening, is the one to suggest that the death penalty be
commutted to banishment (‘—Não mateis o nosso filho, / que bem custou a criar; //
mandai-o p’ra terras longes, / fora do céu natural’ [Don’t kill our son, it was hard to raise
him; send him far away, away from his homeland] Canto, 17). This is closer to Fioravante,
where the distressed mother suggests the idea of exile to the injured Salardo, promising that
her son would marry his daughter upon returning: ‘fategli dar bando del regno . . . e darògli
la vostra figliuola per moglie’ [have him banished from the kingdom . . . and I will give him
your daughter as wife (151)].
Floovent continues with the king’s proclamation forbidding his vassals from helping his
son with money, horses or supplies, threatening those who failed to do so with severe
punishment (vv. 147-53). As Menéndez Pidal indicated (1945, 56), 45 in the ballad the

43
Armistead, Samuel G., and Joseph H. Silverman, Romances judeo-españoles de Tánger recogidos por Zarita
Nahón, Musical transcriptions by Israel J. Katz (Madrid: Cátedra-Seminario Menéndez Pidal, 1977), pp. 193-
94.
44
Bénichou, Paul, Romancero judeo-español de Marruecos (Madrid: Castalia, 1968), 207 .
45
Menéndez Pidal, Ramón, ‘Poesia tradicional en el romancero hispano-portugués’, in Castilla: La tradición, el
idioma (Madrid: Espasa-Calpe, 1945), pp. 41-73 (at p. 56).

64
 Manuel da Costa Fontes

king’s provisions take the form of a curse, which, in Azorean variants, is at times proffered
by the mother herself:
Foge, foge, ó cruel vento, p’r’às bandas de além do mar.
Nas terras donde passares, nem água t’hão-de qu’rer dar;
as fontes donde beberem, logo se hão-de secar;
a mesa donde comeres, logo se há-de escachar,
e a cama donde dormires, em fogo s’há-de abrasar.

[‘Flee, Cruel Vento, flee overseas.

In the lands where you pass they won’t even give you water;
the fountains where you drink will dry up at once;
the tables where you eat will break up into pieces,
and the beds where you sleep will go up in fire.’]
(VRP, 35)
—Não o mateis, senhor rei, que é o nosso filho carnal;
desterrai-o para longe, longe do vosso reinado;
que não tenha pão nem vinho, nem comida o seu cavalo!

[‘Don’t kill him, my Lord and King, for he is our son;

banish him far away, far from your kingdom,
without bread and wine and food for his horse.’]
(Cantos, 18)

The privations endured by Floresvento show that he suffers the consequences, for everyone
refuses to sell him supplies (Cantos 18). This situation is not echoed in Floovent, where the
hero does not endure any hunger, but although his father does not forbid people from
selling him supplies in Fioravante, he becomes lost in a forest and goes without eating for
two days: ‘La terza mattina, non trovando abitazione, s’inginnochiò e raccomandossi a Dio,
perché la fame colla fatica molto lo noiava’ [On the third morning, unable to find any
houses, he kneeled and commended himself to God, for hunger and fatigue bothered him
greatly (155)]. Thus, once again, Fioravante coincides with the ballads in an important
detail missing from Floovent.
Although the protagonist’s amorous adventures are interwoven with his exploits as a
warrior in both Floovent and in Fioravante, they acquire much more emphasis in the latter,
where his mother, while preparing him for exile, ‘missegli una sopravesta verde, la quale
significava giovane innamorato’ [put a green overcoat on him, which meant “enamoured
young man (153)]. A disdained innkeeper’s daughter dies of grief when she realises that he
would rather have Drusolina (201), whom he eventually marries. The women in the castle
of Monfalcone, ‘quando lo viddono, tutte furono accese del suo amore’ [upon seeing him,
all fell in love with him (213)]. In Floovent, the hero is involved only with two women,
Florette and Maugalie (=Drusolina), marrying the latter. Consequently, the allusion to the
dishonoured maidens must derive from a version which followed the Italian form of the
poem more closely. Whereas in the epic the amorous episodes occur after the banishment
of the hero, in Floresvento they are enclosed, through displacement and a considerable
amount of condensation, within the initial series of crimes which have substituted for the
insult to the aged preceptor. It should also be noted that the protagonist’s prowess in the

65
The Singer and the Scribe

amorous encounters has been grossly exaggerated in the ballads, for in the epic he is
definitely consistent in his retreat before the enamoured maidens.
The epic deeds have also been displaced to the beginning of the ballads, where they are
echoed in the reference to the destroyed castles and cities which precedes the banishment.
Obviously, this has been accomplished through a process of condensation similar to the one
observed in the case of the exaggerated amorous adventures, and the hero’s role has been
equally distorted: rather than acting like the mindless wrecker and assassin portrayed in the
ballads, in Floovent and in Fioravante he is the opposite of the villain, always fighting on
the side of truth and justice.
In the Azores, some versions of Floresvento end with verses adapted from Claralinda
(RPI, M1), a well-known ballad that tells how an adulterous wife is caught by her husband
while trying to hide her lover in their bedroom, 46 but Floresvento adds a second lover:
—Quem eram esses dois homens que estavam na minha sala?
—Matai-me, homem, matai-me, que a morte tenho ganhado.
—Não te mato, D. Branca, mate Deus que te criou,
que isto tudo foram pragas que a minha mãe me rogou.

[‘Who were those two men in my bedroom?’

‘Kill me, my husband, kill me, for I deserve to die.’
‘I won’t kill you, Branca, only the God who created you should,
since this was because of the curses that my mother placed on me.’]
(Cantos, 19)

From an artistic point of view, this contamination could, perhaps, be explained as a

fitting retribution on the part of the oral tradition for the hero’s earlier transgressions, as
well as a consequence of variants of the curse placed on him by his mother, who had
wished ‘Que a mulher com quem casares / nunca te seja leal’ [May the woman you marry
never be loyal to you (Cantos, 19; cf. VRP 36)], adding in one version: ‘Os filhos que tu
tiveres, / nunca os há-de lograr’ [You will never enjoy the children you have (VRP 36)].
Fortunately, it is possible to establish at this point a precise parallel with an episode that
takes place near the end of Fioravante. Biancadora, the hero’s mother, had promised to
marry him off to Salardo’s daughter upon his return from exile, but he refused, choosing to
keep his word to the beautiful Drusolina, an infidel’s daughter who had helped him before.
When Drusolina gives birth to two male children, Biancadora, determined to obtain
revenge, locks a servant, Antonio, in her daughter-in-law’s chambers while she is asleep,
accusing her of adultery to her son. Enraged, Fioravante kills Antonio in Drusolina’s
bedroom, but doubts his wife’s adultery when his sword fails to cut through her flesh.
Biancadora insists on Drusolina’s guilt, swearing to put a curse on him if he fails to seek
revenge: ‘Ma io ti giuro que, se tu non ne farai vendetta, che io ti darò la mia maldizione’
[But I swear that, if you fail to take revenge, I will curse you (222)]. Drusolina is then
thrown into a furnace with her two children, but she prays to the Virgin for deliverance and
comes out unharmed. Nevertheless, Biancadora convinces Fioravante to exile her with the
children, claiming that she had escaped with help from the devil. The family is reunited
after several adventures.

46
Pérez Vidal, J., ‘Floresvento and La esposa infiel’, Douro Litoral, 4:9 (1952), 37-40 .

66
 Manuel da Costa Fontes

In Floresvento, then, the disloyalty of the hero’s wife and the suspect legitimacy of his
children have been displaced, becoming part of the curse proffered by his mother at the
beginning of the ballad, a curse which in turn represents, at least in part, a transformation of
the penalties attached to his banishment by the king. The contamination with Claralinda
was triggered by its thematic similarity with an episode found towards the end of
Fioravante: the protagonist catches two men in his wife’s bedroom (one in Claralinda),
but, unlike the adulterous Claralinda, Drusolina is innocent. Thus, Floresvento substituted
a portion of a well-known ballad for a similar episode already present in its epic
predecessor.47
In sum, Floresvento constitutes an abbreviated, condensed form of Floovent, whose
existence can be explained only through a lost Iberian prototype. Since Floresvento has
more in common with Fioravante, the Iberian poem was more closely related to the
narrative’s Italian form, although it was still different from both versions, representing,
therefore, a shorter, independent poem, which, in its turn, probably depended on an
intermediate version in Provençal. Like the ballads, the Iberian epic probably rhymed
mostly in á, and it was thanks to this condensed version that it managed to perpetuate itself
with such a wealth of detail.
Like Floovent, the Cantar de la muerte del rey don Fernando y cerco de Zamora also
existed in a highly condensed form. Besides confirming the traditional character of Nucio’s
trilogy, together with Floresvento, A Morte do Rei D. Fernando points to the transformation
of a few lengthy epics into shorter, condensed versions that could be sung in a more
reasonable amount of time. Thus, these two ballads constitute splendid, additional
examples of the manner in which the modern oral tradition often sheds new light upon the
past.

47
For an analogous contamination in another epic ballad, Celinos (RPI, B9), see Armistead, Samuel G. and
Joseph H. Silverman, ‘El romance de Celinos: Un testimonio del siglo XVI’, Nueva Revista de Filología
Hispánica, 25 (1976), 86-94 (at p. 94).

67
This page intentionally left blank
 Fier Margrietken: a medieval ballad and its history

Ad Putter

The ethnographer Ruth Underhill was once told by a Papago singer: ‘The song is so short
because we understand so much’.1 I am interested in what this aphorism might tell us about
the popular ballad. Ballads, too, are short and compressed, though scholars have usually
explained their terseness differently. Gordon Gerould, for instance, said of the ballads that
‘their compression, their centralization, with the impersonality that results from dramatic
treatment of a theme, and above all, the swiftly moving action, are precisely the qualities
that would arise, almost inevitably, from the practice of singing stories to brief tunes’.2 The
ballad, in other words, developed its distinctive qualities in a process of natural adaptation
to the circumstances of its production. The remark of the Papago singer adds to Gerould’s
evolutionary perspective an important insight: what allows the ballad to be told briefly,
without authorial explanation regarding motivation, cause and effect, and so on, is its
participation in a shared world of understanding. The ballad is short because so much is
understood.
This essay is concerned specifically with a late medieval Flemish ballad, Fier
Margrietken (‘Proud Maggie’), 3 and all that is understood by it. My aims are, in one sense,
limited: I should like to make the ballad accessible in an English translation; to discuss its
history — both the historical events that lie behind it and the processes of memorial
transmission that shaped these events into song; and, finally, to consider the qualities that
make it a ballad, and a beautiful one at that. I regret that I cannot do more to introduce
readers to the medieval ballads of the Low Countries. This whole area is terra incognita to
readers who have no Dutch, and ballad scholarship in Britain now all too often proceeds as
if it did not exist.4 But since the stories of scholars, too, must be longer in proportion as less
is understood, the urgent task of describing the field must be left to someone with more
space (and more expertise). However, since in oral tradition, which kept Fier Margrietken

1
Ramsey, Jarold W., ‘The Wife Who Goes Out Like a Man, Comes Back as a Hero: the art of two Oregon Indian
narratives’, PMLA, 92 (1977), 9-18 (p. 9); reprinted in Folk Groups and Folklore Genres: a reader, ed. by
Elliott Oring (Logan: Utah State University Press, 1989), pp. 209-23.
2
Gerould, Gordon Hall, The Ballad of Tradition (New York: Oxford University Press, 1957), p. 212.
3
I translate fier as ‘proud’ for want of a word that captures all the connotations of the word: ‘spirited, pretty’,
‘courageous, adventurous’, and, particularly important in the traditional ballad (see n. 68) ‘unyielding to male
suitors’.
4
Whence such unfortunate generalisations as David Buchan’s that ‘traditional balladry flourished in nonliterate,
homogeneous, agricultural society, dominated by semi- independent chieftains, that is situated in a remote,
hilly, or border region’: The Ballad and The Folk (London: Routledge and Kegan Paul, 1972; repr. East
Linton: Tuckwell Press, 1997), p. 47. In Holland and Belgium and large parts of Germany ballads flourished
in flat and urbanised land. A brief account of the Dutch and Flemish ballad is given by Entwistle, William,
European Balladry (Oxford: Clarendon Press, 1939). For more detailed surveys of the Dutch and the closely
related German ballad see respectively Kalff, G., Het Lied in de Middeleeuwen (Leiden: Brill, 1884) and
Kayser, Wolfgang, Geschichte der deutschen Ballade (Berlin: Junker and Dünnhaupt Verlag, 1936).
The Singer and the Scribe

alive until at least the nineteenth century, every theme and every formula ‘has around it an
aura of meaning which is put there by all the contexts in which it has occurred in the past’,5
the discussion of Fier Margrietken inevitably leads us into the larger tradition that gives the
words and ideas of each single ballad their full effect. The ballad of Fier Margrietken may
thus serve as a point of entry into the wider ballad context. As I hope to show, it may also
afford us a rare insight into the processes by which history is transformed when it is
remembered and retold in ballad form. For, unusually, even though both the ballad of Fier
Margrietken and the murder of which it tells go back to medieval times, the historical
events that lie behind the ballad are nevertheless remarkably well documented. Only rarely
do we have this kind of information about a ballad; indeed, some say that it does not exist
‘for any song within the received tradition of popular balladry’. 6 On this point the evidence
can speak for itself.

Margaret’s Murder and Cult

One evening — the date was later officially fixed at 2nd September 1225 — Amandus and
his wife were making preparations to join the Cistercian order at Villers-la -Ville (in
present-day Belgium). They had saved money towards their upkeep, and were to set out the
next day. Their preparations had come to the attention of a gang of criminals who entered
their guesthouse, the St George’s inn on the Muntstraat in Louvain, under the pretence of
being pilgrims in need of board and lodging. A servant girl, Margaret, was dispatched with
a pitcher to fetch wine for the ‘guests’, probably from the winepress in the ducal vineyard;
she returned to find the household murdered. Still clutching her pitcher she was taken first
to a house on the outskirts of Louvain (where she was probably raped) and then, when the
lady of the house became suspicious, to the river Dijle, where the robbers decided to
dispose of her. It seems that one of the criminals had some compunction, offering to marry
her; but Margaret had seen too much, and for an extra ten marks the man agreed to murder
her himself, cutting her throat and stabbing her in the chest, and then dumping her into the
river. However, the lady of the house, who had secretly followed them, witnessed the
murder, and the story spread like wildfire the next day. The hunt for the killers was in vain,
but fishermen later recovered her body and buried it on the riverbank. Some days later a
mysterious light shone around her grave; the body was dug up and transferred to the
collegiate church of St Peter in Louvain, where a wooden chapel was built adjoining the
church walls: ‘And even to this day various miracles are wrought, so great were her merits,
both in the place where she was killed, and also in the chapel to which her body was
translated’.7

5
Lord, Albert B., The Singer of Tales, 2nd ed. (Cambridge, MA: Harvard University Press, 2000), p. 148.
6
Andersen, Flemming G., and Thomas Pettitt, ‘The Murder of Maria Marten: the birth of a ballad?’, in Narrative
Folksong: new directions; essays in appreciation of W. Edson Richmond, ed. by Carol L. Edwards and
Kathleen E. B. Manley (Boulder, Colorado: Westview Press, 1985), pp. 134-78.
7
Caesarius of Heisterbach, The Dialogue on Miracles, trans. by H. von E. Scott and C. C. Swinton Bland, 2 vols
(London: Routledge, 1929), I, p. 447. There is a learned, if occasionally fanciful, reconstruction of the events
by Ollivier, Marie-Joseph,‘La Bienheureuse Marguerite de Louvain’, Revue Thomiste, 4 (1896), 592-618 and
van Even, Edward, La Bienheureuse Marguerite de Louvain, dite Marguerite la Fière: sa légende, son culte,
sa chapelle (Louvain: Peeters, 1896). I have drawn on Ollivier and Van Even, and on the entry on Margaret in

70
 Ad Putter

There is no reason to doubt that this horror story, or at least something very similar, did
indeed take place in Louvain. Caesarius of Heisterbach (ca 1180-1240), who is our earliest
authority for it, was writing close to the events,8 and had personal informants at the
Cistercian monastery of Villers which Amandus and his wife had hoped to join; the St
George’s inn was used by the monastery’s lay-brothers, one of whom Caesarius adduces as
an eye-witness to the diligence and humility with which Margaret served in the inn before
her murder. 9 Very quickly a local cult of Margaret established itself. Sites associated with
her became places of worship. Her room in the inn of St George was named after her. In
1308 the Beguine Margaret Utembruele built a chapel in the district where Margaret the
Blessed had been abducted and Margaret the Beguine had been born, a district known as the
‘Bruel’, then infamous on account of its prostitutes. The wooden chapel in which
Margaret’s bones rested was reconstructed ca 1435. It must have been a familiar landmark
to locals. In a municipal act dated 18th July 1453, the location of a house belonging to
‘Henricus de Halen’ is given as opposite the chapel ‘dicte Tfier Margrietken’; in 1479 the
builder Johan Oeghe was instructed to remove the roof tiles from the same house; again the
specified location is ‘opposite Fier Magrietken’.10 In the 1540s the wooden chapel was
replaced with a stone one, located inside the church walls, where Margaret’s shrine remains
to this day.
And of course there were miracles. Shortly after Margaret’s death, an old woman who
sold soup on the market had disdainfully refused to join in with the crowd that was
worshipping her at St Peter’s, dismissing the stories of her sanctity as ‘about as plausible as
the chance of the pan of soup I have taken off the fire coming to the boil’; but no sooner
had she uttered this slander than the soup came to the boil and, what is more, the pan never
emptied — until a member of the gang that abducted and killed Margaret had the nerve to
eat the soup. The villain and his confederates were immediately apprehended and sentenced
to death.
This miracle of the market-woman is told in the short Life of the Blessed Margaret of
Louvain, from an anthology of saints’ lives entitled Hagiologium Brabantinorum, compiled
ca 1480 by Johannes Gielemans at the monastery of Rouge-Cloître (near Brussels).11 Its
account of Margaret’s murder generally agrees with that of Caesarius, but more detail is
given. Thus Gielemans writes that the robbers had tried to rape Margaret, who had,
however, been ‘protected by the shield of godly piety from her treacherous enemies, who
attempted to take from her the precious treasure of her chastity.’ Her body, undefiled, had

Acta Sanctorum for information about her life and cult. A full and scholarly treatment of the legend (in Dutch)
is currently being prepared by Gilbert Huybens. I am very grateful for his comments on an earlier draft of this
essay.
8
The Dialogue is believed to have been completed in 1222; since Caesarius talks of the murder as having taken
place ‘a few years ago’ the traditional date of 1225 is almost certainly too late.
9
Dialogue, p. 446.
10
van Even, Marguerite, pp. 37-38.
11
An edition of the text is available in AS, 2 September, pp. 582-94. For a description of the Hagiologium
Brabantinorum see anon., Iohannis Gielemans (Brussels, 1895), pp. 11, 42-61; for further bibliography on the
Rouge-Cloître manuscripts used by the Bollandists see Dubois, Jacques, and Jean-Loup Lemaitre, Sources et
méthodes de l’hagiographie médiévale (Paris: Éditions du Cerf, 1993), p. 40.

71
The Singer and the Scribe

floated into Louvain against the current, carried by fish. And none other than the Duke of
Brabant (Henry I) and his wife had seen the mysterious light surrounding her corpse as they
stood on the castle walls; they had led the solemn procession of courtiers and clerics which
later accompanied the corpse to its resting place at St Peter’s. The Acta Sanctorum
reproduces a seventeenth-century engraving, apparently inspired by an original wall
painting in her chapel, 12 which shows Margaret afloat in the river — her mouth stuffed with
a white ball to stop her from screaming, the wine pitcher standing on the riverbank, while
from a castle window the duke and his wife look on. The wine pitcher remained, or turned
up, in the hands of Amandus’s family, the prominent Absolons and their heirs, who
entrusted it to the chapter so that it might be displayed on the great feast of Margaret, held
every seven years on 2nd September. It is now in the possession of the chapter of St Peter’s,
as is a white leather ball, which, like the wooden pitcher, had become an object of
veneration.

The Ballad and its Date

As stories are passed on they mutate. The classic book Remembering by the psychologist
Frederic Bartlett is devoted to the study of such mutation.13 In a series of controlled
experiments Bartlett set in circulation an oral story along a chain of subjects, from whose
memory he took down the story after every stage of transmission. Bartlett discovered that
the most powerful factor of narrative change is the conscious or unconscious influence of a
‘schema’, a deep structure of stereotypical patterns and norms which acts on remembered
experiences or narratives by re-framing them in its likeness. This process of
‘schematisation’ is discernible even in the official written record of Margaret’s murder,
which is clearly influenced by the genre of the saint’s life. It is interesting to see how the
hagiographical schema operated on the main obstacle to her sanctification, which was that
Margaret had probably been raped. A number of early chronicles state quite specifically
that she was raped,14 while the offer of marriage in Caesarius’s earliest account may imply

12
AS, p. 584; van Even, Marguerite, p. 21. The series of paintings that now adorn her chapel today, by P.J.
Verhaghen, date from 1760. I am grateful to Tom Shippey for taking the trouble to see the Chapel of the
Blessed Margaret in St Peter’s on my behalf.
13
Bartlett, F. C., Remembering (Cambridge: Cambridge University Press, 1932).
14
See e.g. Chronyk van Nederlant van Jaere 1027 tot 1525, in Chroniques de Brabant et de Flandre, ed. by
Charles Piot (Brussels: Académie Royale de Belgique, 1879): ‘Item, MCCXVI […] Ende doen was vermoerd
ende vercracht fiere Margriete te Leuven’, and Oud kronijkje van Leuven: ‘Int selve jaer (1225) wordden Tfir
Magrietken vercracht ende vermoort, te Loven, en dit deden die vleeschouwers’ (cited in van Even,
Marguerite, p. 19). verkrachten can mean to abduct in medieval Dutch, but I doubt that this is the intended
sense in these two instances. A ‘saintlier’ account is given in the Klein Kroniek, in Codex diplomaticus
Neerlandicus, ed. by M.C.A. Rethaan Macaré, 3 vols (Utrecht: Institute Historique d’Utrecht, 1853), III, p.
64, which states that Margaret, ‘nomine vulgariter Fier Margrietken, propter ejus castitatem interfecta est et in
Dyeliam fluvium projecta ab iis, qui eam stuprare et florem virginitatis privare studebant’ (‘popularly named
‘Proud Maggie’, was killed on account of her chastity and thrown in the river Dijle by those who attempted to
violate her and to steal the flower of her virginity’).

72
 Ad Putter

rape.15 But in clerical versions of Margaret’s martyrdom, rape could not be accommodated
since in the lives of women saints it is always narrowly avoided. Admittedly, St Lucy had
contemplated the prospect of rape with equanimity on the grounds that it would double her
reward in heaven — an argument that Margaret’s later apologists never fail to adduce16 —
but God had intervened in time to protect Lucy’s purity. Loss of virginity may be
threatened in medieval hagiography but never actually endured.17 Accordingly Gielemans
remembers Margaret dying with her virginity intact, miraculously protected by God; in later
versions she dies ‘defending her chastity’ or ‘refusing the solicitations of her murderers’, 18
and in the most creative rewriting she dies a virgin while travelling to Villers, where she
was to take the veil. 19
However, in the official records, access to earlier written versions keeps the influence of
the schema in check. In memorial transmission this constraint falls away, and the ballad of
Fier Margrietken dramatically illustrates what may happen to a story when a memorial
schema operates unchecked:
Het soude een fier Margrietelijn Proud Maggie was setting out,
Ghister avont spade Late yesterday evening,
Met haren canneken gaan om wijn; With her pitcher to fetch wine;
20
Si was daer toe verraden, ja verraden. She had been led on to this, yes led on.

Wat vantse in haren weghe staen? What did she find across her path?
Eenen ruyter stille. It was a silent knight.
‘Nu segt mi, fier Margrietelijn ‘Now tell me, proud Maggie,
doe nu mijnen wille, ja wille’. You will do my will, yes will.’

‘Uwen wille en doen ic niet. ‘I shall and will not do your will,
Mijn moerken soude mi schelden; My mum would tell me off;
Storte ic dan mijnen coelen wijn, If I should spill my wine so cool,
Alleyne soude ic hem ghelden, ja ghelden.’ I’d pay for it alone, yes alone.’

‘En sorghet niet voor den coelen wijn, ‘Never mind the wine so cool,
Mer sorghet voor u selven. But rather mind yourself.
Di waert is onser beyder vrient, The taverner is a friend to us both,
Hi sal ons noch wel borghen, ja borghe.’ He’ll give us credit, yes credit.’

15
It is to be remembered that this is a time when marriage between rapist and rape victim constituted a legally
recognised remedy for the victim. See Saunders, Corinne, Rape and Ravishment in the Literature of Medieval
England (Cambridge: D.S. Brewer, 2001), p. 56.
16
See the defence of Margaret, with a full list of authorities, in AS, p. 591.
17
Saunders, Rape and Ravishment, pp. 130-31.
18
AS, p. 585.
19
Henriquez, J. C., Menologium Cisterciense (Antwerp, 1630).
20
I can find no support for the loose translation offered by the editors of l.4 as ‘Dat werd haar ongeluk’ (‘that was
her downfall’); the sense of verraden that best fits the linguistic context (with adverbial daertoe ‘thereto’) is
the primary sense of verraden ‘to lead astray (by evil counsel)’: Middelnederlandsch woordenboek, ed. by E.
Verwijs and J. Verdam (The Hague, Martinus Nijhoff, 1927-52), s.v. verraden (1). The ballad tacitly assumes
our knowledge of the legend: the request for wine is part of the villains’ false pretence that they have come to
enjoy hospitality. In the words of the Papago singer: ‘The song is so short because we understand so much’.

73
The Singer and the Scribe

Hi namse in sinen witten armen He took her by her snow-white arms

Heymelick al stille; Secretly, without a sound;
Al in een duyster camerken In a darkened room
Daer schafte hi doe sinen wille, ja wille. He had his will with her, yes will.

Smorgens ontrent der middernacht The following day around midnight

Si ghinc haer kanneken soecken, She went looking for her pitcher,
Daer lach die moeyaart ende hi loech: While the scoundrel lay and laughed:
‘Het staet daer teynden mijn voeten, ja voeten.’ ‘It is standing at my feet, yes feet’

‘Mer dat daer teynden u voeten staet, ‘But for what stands there at your feet
Dat sal u noch lange berouwen. You will long be sorry.
Ic hebbe noch drie ghebroeders stout, I have three bold brothers,
Si sullen u dat hooft af houwen, ja houwen.’ They will cut off your head, yes head.’

‘Alle u ghebroeders stout ‘All your brothers bold and strong —

Die sette ick in mijn deeren. My heart goes out to them.
Ick sal alle dese somer lanck For all this summer long
Met Grietken houden mijn scheeren, ja scheeren. I’ll amuse myself with Maggie, yes Maggie.’

Ende hi nam eenen snee witten bal, And he took a snow-white ball,
Hi stackse al in haer kele, He thrust it in her throat,
Hi schootse tot eenderen veynsteren, He flung her from a window,
Hi schootse al in die Dijle, ja Dijle. He flung her in the Dijle, yes Dijle.

Teghen stroom quam si ghedreven uut Against the current she was washed up
Aen sint Jans cappelle. At the chapel of St John.
Dat sach so menich fijn edel man, So many a nobleman witnessed that,
21
So menich jonc gheselle, ja gheselle. So many a good fellow, yes fellow.

This ballad is no. 67 in the most important repertory of early Dutch ballads, Het Antwerps
Liedboek (henceforth AL), printed by Jan Roulans in 1544. The date of the only extant print
tells us little about the antiquity of Fier Margrietken, which, even in the form preserved by
AL, is older. Firstly, the 221 songs in AL appear to have been accumulated over a period of
years. The book is organised alphabetically, but the alphabetical sequence starts again at
song no. 172, and then again at no. 210. The likely explanation for this is that the 1544 print
was a conflation of at least one earlier edition.22 On the basis of historical songs about
topical events, the first cycle (containing songs 1-171) can be dated to ca 1535. The only
other internal information about the date of some of the ballads in AL is the rubric ‘an old

21
I have used the most recent edition by Vellekoop, K., and Wagenaar-Nolthenius, H., Het Antwerps Liedboek
(Amsterdam: Vereniging voor Nederlandse Muziekgeschiedenis, 1972), 2 vols I have made a few changes
with respect to punctuation and have regularised the refrain words ‘ja […]’, which the editors print only for
the second and final stanza. I have not attempted in my translation to render the rhyme and assonance, but
have tried to keep to the rhythm of the original.
22
I rely here on the excellent study by Koepp, Johannes, Untersuchungen über das Antwerpener Liedbuch vom
Jahre 1544 (Antwerp: De Sikkel, 1929). See also Joldersma, Hermina, ‘“Het Antwerps Liedboek”: A Critical
Edition’, 2 vols (Ph.D. dissertation, Princeton University, 1983), I, pp. xxxv-xxxviii.

74
 Ad Putter

song’ or ‘a new song’,23 which appears above some seventy songs in the collection. Fier
Margrietken has no such rubric.
In a few cases, versions of the ballads of AL are found in medieval manuscripts. The
Dutch ballad of Brandenborch (AL 81), for example, is first found in a fifteenth-century
manuscript; it goes back to a German original, which must be even earlier, though no
German versions are extant before the sixteenth century.24 The ballads of Heer Danielken
(AL 160) and Vanden ouden Hillebrant (AL 83) also derive from German ballads, versions
of which have survived in manuscripts from the fourteenth and fifteenth centuries.25 More
often our evidence for the antiquity of some of the ballads in AL is indirect. Popular ballad
tunes were commonly recycled for the use of religious songs, some of which are preserved
in books or manuscripts that predate AL. One of the best-loved Dutch ballads, Het daghet
inden Oosten (AL 73), introduced by Roulans as ‘Een oudt liedeken’, had already been
adapted into German by the fifteenth century,26 and its words and music inspired a number
of religious contrafacta, including a German lyric of 1421,27 and two Dutch ones that
survive in songbooks of 1539 and 1540.28 Possibly a Dutch contrafactum existed much
earlier, for a fifteenth-century Life of St Gertrude, a Beguine in Delft (died 1358), derives
her cognomen ‘Van Oosten’ from her habit of singing ‘daily for the love of Christ a song,
beginning Het daghet inden Oosten’.29

23
‘A new song’, it must be emphasised, can very well be old, for ‘new’ in this context may simply mean ‘adapted
to new circumstances, re-issued’. Thus in the following passage from the thirteenth-century Livre d’Artus —
‘Si comenca Grex un sonet novel quil ot apris en enfance’ (‘Then Grex began singing a new song which he
had learned in his childhood’) —novel evidently cannot mean ‘new’ in our sense (The Vulgate Version of the
Arthurian Romances, ed. by H. Oskar Sommer [Washinton: Carnegie Institution, 1913], vol. VII, p. 161). In
AL some songs designated as ‘Een nyeu liedeken’ (e.g. 29, 62) are actually adaptations of older German
originals.
24
Koepps, Untersuchungen, p. 118.
25
Balladen, ed. by John Meier, 2 vols (Leipzig: Reclam, 1935), I, p. 21; Koepps, Untersuchungen, p. 62, and van
Duyse, F., Het Oude Nederlandsche Lied, 3 vols (The Hague: Martinus Nijhoff, 1907), I, p. 40. In an attempt
to demonstrate the priority of the French ballad in the evolution of the medieval genre, Lajos Vargyas argues
that the songs Danielken and Hillebrant should not be regarded as ballads but as late heroic songs. In my view
the distinctions he wishes to draw between the heroic song and the ballad are too fine: if, as he himself argues,
the ballad usurped the place of the heroic epic in folk poetry, it should not surprise us that a number of early
Dutch and German ballads resemble heroic poetry. This does not mean they are ‘not authentic ballads’. See
Vargyas, Lajos, Researches into the Mediaeval History of Folk Ballad (Budapest: Akadémiai Kiadó, 1967),
pp. 261;276.
26
The earliest version survives in a manuscript datable to 1439. The scribe calls the ballad a ‘peasant song’. See,
Balladen, ed. by Meier, I, 201- 02.
27
Kalff, Lied, p. 155.
28
Vellekoop and Wagenaar-Holthenius, II, 49.
29
AS, 6 January, p. 349. The source is another hagiographical compilation by Gielemans, Johannes, Novale
Sanctorum (ca 1484), on which see Iohannis Gielemans, pp. 12, 61-80 and above, n. 10. van Duyse
(Nederlandsch Lied, I, 123) explained the association of St Gertrude with Het daghet in den Oosten as a later
rationalisation of Gertrude’s name; this is possible, though anyone attempting to rationalise ‘Van Oosten’
might have opted for something more obvious.

75
The Singer and the Scribe

Many Dutch and German ballads are thus incontrovertibly medieval. While we have no
clear evidence for the existence of any German and Dutch ballads before the thirteenth
century,30 thereafter the grounds for speculation become firmer. 31 In 1233 a pilgrim on her
way to Marburg heard a German song about the tragic separation of St Elisabeth of
Hungary and St Louis IV, Landgrave of Thuringia, who left her to go on crusade, where he
died in 1227.32 Some verses of the song have survived in a fourteenth-century German Life
of St Louis:
Ir ein daz ander umbeving They embraced each other
gar früntlich da mit armen; Most lovingly in their arms;
groz jamer durch ir herze ging, Great sorrow pierced her heart,
wen sold diz nicht erbarmen? Who would not feel pity for this?

The extract has all the appearance of a ballad stanza. One of the earliest Dutch historical
songs in ballad stanza (from a manuscript copied shortly after 1525) tells the story of St
Elisabeth and St Louis’s daughter, Sophia (1224 -1284), who married Duke Henry II of
Brabant, and who, according to the ballad (though not fact), died in childbirth, just as her
husband returned from war:
Al doen hij op die camer trat, And when he stepped into the room,
Hoe drufelijc dat sij op hem sach. How piteously she looked at him.
‘Sijt welcoem mijn prins mijn here ‘Welcome to you, my prince, my lord,
33
Nu sidt welcoem mijn alre liefste man.’ Welcome to you, my dearest man.’
The events of 1244 that lie behind the ballad are only dimly remembered in the ballad, but a
couple of centuries of memorial transmission may well account for this. A parallel can be
drawn with the German ballad Der edle Moriger, whose hero is a distant shadow of the
historical Minnesänger Heinrich von Morungen (early thirteenth century). In the ballad
Moriger leaves his lady for seven years and returns just in time to rescue her from a
marriage with his vice-regent. Could the ballad go back all the way to the thirteenth
century? It is a distinct possibility, for the dolorous parting of Moriger from his wife was
already the stuff of legend by the end of the thirteenth century, when an Austrian poet
expressed the pain of separation by recalling, just as the ballad does, how ‘der Morungaer’
had parted with sorrow from his love.34
Although the thirteenth century may have been the breeding ground for a handful of
early German and Dutch ballads, in the form we read them they tend to be the products of

30
For speculations (based on slender evidence) concerning the existence of the ballad in the early Middle Ages see
Metzner, Ernst Erich, Zur frühesten Geschichte der europäischen Balladendichtung: der Tanz in Kölbigk
(Frankfurt: Athenäum Verlag, 1972). A summary of Metzner’s theory in English can be found in his article
‘Lower Germany, England, Denmark and the Problem of Ballad Origins’, in The European Medieval Ballad,
ed. by Otto Holzapfel (Odense: Odense University Press, 1978), pp. 26-39.
31
Schneider, Hermann, ‘Ursprung und Alter der deutschen Volksballade’, in his Kleinere Schriften zur
germanischen Heldensage und Literatur, ed. by Hermann Schreider (Berlin: De Gruyter, 1962), pp. 96-106.
32
Kayser, Geschichte, p. 18.
33
Middelnederlansche Historieliederen, ed. by C. C. van de Graft (1904; repr. Arnhem: Gysbers en Van Loon,
1968), p. 47.
34
Kayser, Geschichte, pp. 20-21.

76
 Ad Putter

the fourteenth and fifteenth centuries. This pattern is consistent with the situation in other
European countries. There is, for instance, sporadic evidence from England and Denmark
that ballads existed by the thirteenth century: the Judas ballad is extant in a late thirteenth-
century manuscript;35 the opening of (what may be) a Danish ballad occurs in a manuscript
copied ca 1300, 36 but here, too, hard manuscript evidence becomes more substantial by the
end of the medieval period. Riddles Wisely Expounded (Child 1), St Stephen and Herod
(Child 22), Robyn and Gandeleyn (Child 115), and a number of Robin Hood ballads (Child
118-19, and perhaps 121) survive in fifteenth-century manuscripts. In Scandinavia, a
fragmentary stanza (the ‘Greenland stanza’) based on Hervarar saga was written on a map
of Greenland from 1427.37 Later (ca 1500) there are fragments of the Danish ballads The
Knight Transformed into a Hart and Marsk Stig.38
This brief overview of ballad chronology gives us a sense of what the bounds of
possibility for dating Fier Margrietken are. It is not impossible that the ballad developed
from a song composed in the immediate aftermath of Margaret’s murder, but, since her cult
continued unabated, the ballad may also have been prompted by later commemorations: by
the building of a chapel dedicated to her in ‘the Bruele’ in 1308; by the reconstruction of
her wooden chapel at St Peter’s in 1436; or by any of the seven-yearly feasts in her honour.
I think that this last possibility is more likely, since, like these ritual acts of remembrance,
the ballad is commemorative, and seeks to honour her relics by restoring them to the life-
story that gave them meaning: the pitcher of wine guarded by Margrietken with such
childlike innocence stands on the altar of the blessed Margaret every seven years; the white
leather ball, venerated by her devotees, is the ‘snow-white ball’ thrust in Margrietken’s
throat. And as Margaret’s worshippers would on her feast day, the ballad remembers her
martyrdom as if it happened today. Margrietken leaves home ‘late yesterday evening’, and
is killed the next day (‘Smorgens’). In true ballad fashion, the distant past is re-imagined as
the here-and-now; 39 in the editors’ words, ‘the old legend is presented as the latest news’.
But perhaps the story’s chronology implies, as its objects do, a far more precise relationship

35
The English and Scottish Popular Ballads, ed. by Francis James Child, 5 vols (Boston, Riverside Press, 1882-
98), no. 23. I see no reason for doubting, as some do, that the Judas poem is a ballad. An extreme position, that
there are no genuinely medieval English ballads, is taken up by Fowler, David C., A Literary History of the
Popular Ballad (Durham, North Carolina: Duke University Press, 1968), but Fowler ignores the wider
European picture. A cautious survey of the evidence for medieval England is given by Buchan, David, in
‘British Balladry: Medieval Chronology and Relations, in European Medieval Ballad’, ed. by Holzapfel, pp.
98-106. More recently Richard Firth Green has forcefully argued the case for the medieval English ballad in
‘The Ballad and the Middle Ages’, in The Long Fifteenth Century: Essays for Douglas Gray, ed. by Helen
Cooper and Sally Mapstone (Oxford: Clarendon Press, 1997), pp. 163-84.
36
Steenstrup, Johannes, The Medieval Popular Ballad, trans. by E.G. Cox (Seattle: University of Washington
Press, 1968), p. 254. In his recent study of the Scandinavian ballad, David Colbert argues that the Scanian
couplet should not be regarded as a ballad fragment, though he, too, finds indirect evidence of the existence of
ballads in the thirteenth century in the occurrence of ballad commonplaces in a number of Swedish romances,
the earliest being Herr Ivan (1303). See The Birth of the Ballad: the Scandinavian medieval genre
(Stockholm: Svenskt Visarkiv, 1989).
37
Colbert, Birth of the Ballad, pp. 56-61.
38
Steenstrup, Medieval Popular Ballad, p. 253.
39
For example, the opening of Lady Isabel and the Elf-Knight (Child 4) – ‘Fair Isabel sits in her bower sewing’.

77
The Singer and the Scribe

to Margaret’s cult. For if the evening that Margrietken sets out is our ‘yesterday’, then our
point of temporal reference is the day she dies. We appear to be dealing with a ballad that
was intended for Margaret’s feast day, and the ballad’s ‘yesterday’ may be a way, not just
of bringing the past back to life, but of remembering the liturgical significance of the day on
which the song was meant to be sung.
At any rate the ballad must be considerably older than the date of publication of AL. A
possible clue to its eventful textual history is the reference in the last stanza to St John’s
chapel: ‘Teghen stroom quam si ghedreven uut / Aen sint Jans cappell.’ Conceivably, the
reference is to one of the two St John’s chapels that existed in medieval Louvain.40 Or
perhaps, as the editors assume, ‘sint Jan’ is simply a corruption of ‘sint Pieter’, since ‘a St
John’s chapel is nowhere mentioned in the legend of St Margaret’. But there is more to be
said. Fier Margrietken survived in oral tradition well into the nineteenth century. In 1900
Albert Blyau and Marcellus Tasseel took down a version of the ballad (Machrietje) from
oral tradition in West-Vlaanderen:41
Des avonds in het klaar manesching, In the evening in the clear moonshine
Als Machrietje wierd uitgelaten, When Maggie was let out,
T was om te halen den rooden wijn, It was to bring some red wine home,
En dat op eenen avond late. And that was late one evening.

Machrietje onder haar weugetjes kwam, Maggie was on her way,

Den stout ruiter die kwam haar tegen: The bold squire came up to her:
‘Waar ga je, waar ga je, Marchrietje, men lief? ‘Whereto, whereto, Maggie, my love?
Zeg mij, en vanwaar zijn al jen weugen?’ And tell me why are you on the road?’

‘Vanwaar dat al mijn weugetjes zijn? ‘My reasons for being on the road?
En dat zal ik u zeere zeggen! I will tell you without doubt!
Sturt gij mijn kannetje met rooden wijn, If you spill the wine in my pitcher,
En dat zalt gie diere vergelden.’ You will dearly pay for it.’

Hij nam Machrietje bij haren hand He took Maggie by the hand,
En hij smeet ze in de riviere: And he threw her in the stream,
‘Ligt hier, ligt daar, Machrietje, men lief, ‘Lie here, lie there, Maggie my love,
En je ligt in de koele riviere!’ And you are lying in the river cold.’

Maar door den stroom van’t klaar watertjen But through the current of the waters clear,
Is Machrietje naar Gent gaan varen; Margaret floated to Ghent
De koopmans van Gent, se hen Machrietje verken, The merchants of Ghent, they knew her well,
En dat al aan haar roode wangen. By the redness of her cheeks.

Machrietje haddede een ringetje aan, Maggie wore a little ring,

En wat stoeg daarop geschreven? What was written thereon?
Alsdat den stouten ruiter, haar lief, That the bold knight, who was her love,
Hadde Marchrietje zoo kwalijk misdreven. Had so badly mistreated her.

40
Gilbert Huybens, personal communication.
41
Iepersch Oud-Liedboek: Teksten en Melodieën uit den Volksmond Opgeteekend, 2 vols (Ghent: J. Vuylsteke,
1900-1902).

78
 Ad Putter

Den dag passeerde en den avond kwam aan The day passed and the evening came,
Ze hebben den stouten ruiter gevangen; They have captured the squire so bold;
Ze hen hem in vier kaartieren gekapt, They have cut him in four parts,
En aan elken poort een deel g’hangen. And hung one part on each gate.

Al die dat niet gelooven en wilt, All those who are incredulous
En ze meugen al vrij gaan kijken: Are at liberty to go and see.
Toe Gent, al binnen Sint -Janskapel, In the chapel of St John in Ghent,
En daar ligt Machrietje te kijken. There Margaret lies to be seen.

Although the story has been resituated in Ghent, this ballad is recognisably a version of
Fier Margrietken, as is shown by various family resemblances. For instance, while in the
first stanza the words have changed, the rhyming sounds, always more tenacious in
memorial transmission,42 have been well preserved,43 as has the second couplet of the third
stanza. And, crucially, we have found another St John’s chapel, not in Louvain, but in
Ghent — and with it an explanation for the change of setting in Fier Margrietken. Though
originating in Louvain, the ballad may have spread to Ghent and attached itself to the city’s
most prominent landmark, the parish church of St John (now the St Baafs Cathedral). 44
Purists may regret the ‘error’, but ballads have always owed their longevity to their ability
to adapt to new places and so ensure their continuing relevance. A spectacular example of
the ballad’s adaptability is Earl Brand (Child 7), which travelled from Scotland to America,
where it lost its original place-names and acclimatised to its new surroundings. An
Appalachian singer who recited a version of Earl Brand told the ballad collector that the
story ‘happened way back yonder in Mutton Hollow. I was there myself’. 45 It would thus
appear that Machrietje descends from the original medieval ballad of Fier Margrietken in a
long line of memorial descent, though of course there is no absolute certainty. Perhaps AL
remained in circulation, in which case Fier Margrietken may have passed into oral
circulation much later. Unlikely: the book was banned in 1546, and the only surviving copy
was discovered some three centuries later in the Herzog August Library in Wolfenbüttel. 46
Perhaps Machrietje was inspired by a reading of Fier Margrietken in a modern edition of
AL. Unlikely: for if Machrietje goes back to the single nineteenth-century edition of AL (by
Hoffman von Fallersleben [1855]), it would be hard to explain the drastic transformation of
the original song. It may be simpler to see Machrietje as a witness to the long-lasting
vitality of Fier Margrietken in the oral tradition of Flanders.

42
See Rubin, David, Memory in Oral Traditions: the cognitive psychology of epics, ballads, and counting-out
rhymes (New Work: Oxford University Press, 1995), p. 83 and Buchan, Ballad and Folk, pp. 155-60.
43
‘Spade’ at l. 2 of Fier Margrietken has been modernised to ‘late’, while l. 4 of the original ballad, which would
be comprehensible only to those who know the legend of Margaret, has been rationalised
44
St John’s church was rededicated to Ghent’s patron saint, St Bavo, in 1539, and became the city’s cathedral in
1560. See Dictionnaire d’histoire et géographie ecclésiastiques, ed. by R. Aubert (Paris: Letouzey et Ané,
1912-), s.v. Gand. St John’s church may well have had in its possession a relic of Margaret of Louvain; but
unfortunately all medieval relics were destroyed in the religious troubles of the sixteenth century. I should
like to thank the present archivist of St Baaf’s, Ludo Collin, for his help on the early history of the cathedral.
45
Scarborough, Dorothy, A Song Catcher in the Southern Mountains (1937), cited by Hodgart, M.J.C., The
Ballads, 2 nd ed. (London: Hutchinson, 1962), p. 139.
46
See Joldersma, ‘Antwerps Liedboek’, for the history of AL.

79
The Singer and the Scribe
From History to Ballad
The earliest recorded version of Fier Margrietken already has behind it a history of
transmission and subsequent evolution. Its genuine popularity is further attested by two
religious lyrics that appropriated the ballad tune. One is extant in a songbook of 1565, also
printed in Antwerp.47 The other appears in Het Nieuw Amsterdams Liedboek of 1591; a
rubric specifies: ‘to the tune of Het zou een fier Margrietelijn / des avonds also spade’.48
We notice that the second line has changed: the durior lectio ‘Ghister avont spade’ in AL
has been replaced by the vague ‘des avonds also spade’ (‘Late in the evening’), 49
presumably because, in Amsterdam, the association of the ballad with the saint and her
feast day was no longer understood.
And only a long history of oral tradition can plausibly explain how the authentic story of
Margaret’s murder was reconfigured around the enduring schemata that guide our
memories of life and narrative. I want to argue that the two most powerful schemata that
shaped the style and substance of Fier Margrietken and brought about the drastic changes
to the original story of Margaret’s murder are the ballad genre and, more specifically, the
widespread ballad type of the wilful daughter.
Let me begin with the influence of the ballad genre. In his seminal essay ‘Epic Laws of
Folk Narrative’, Axel Olrik delineated the laws governing the world of folklore and
folksong (die Sagenwelt).50 While Olrik considered these ‘laws’ to be superorganic, that is,
‘requiring no reference to other orders of phenomena for an explanation of [their] origin,
development, and operation’,51 many of his laws seem to me to reflect the ingrained habits
and conditions of thought of non-literate societies. Nevertheless, Olrik’s laws continue to be
relevant, since they describe most accurately what can and cannot happen in ballads, and so
allow us to predict what must be added to or subtracted from life if it is to be remembered
in ballad form. I should like to list the most pertinent laws, followed by a brief discussion of
their implications for the ballad in general and Fier Margrietken in particular.
1. The Law of Opening and the Law of Closure
‘The Sage begins by moving from calm to excitement, and after the concluding event, in
which a principal character frequently has a catastrophe, the Sage ends by moving from
excitement to calm’. 52
Ballads respect the law of opening in a number of ways. They may often begin with
literal stasis — ‘Lady Margaret sat in her bower door’ (Child `104B), ‘In Oostenrijk daar
staat een huis’ (van Duyse, 18) — or with an introductory line: ‘It was a knight in Scotland

47
Vellekoop and Wagenaar-Nolthenius, Antwerps Liedboek, II, 46.
48
van Duyse, Nederlandsch Lied, I, 100. In his edition of Fier Margrietken Van Duyse actually prints these two
lines instead of those of AL. For the reasons I have indicated I consider it far more likely that AL is closer to
the archetype.
49
Compare t he even vaguer nineteenth-century version: ‘Op eenen avond late’.
50
In The Study of Folklore, ed. by Alan Dundes (Englewood Cliffs, N.J.: Prentice-Hall, 1965), pp. 131-41. The
essay was originally published as ‘Epische Gesetze der Volksdichtung’, Zeitschrift für Deutsches Altertum 51
(1909), 1-12.
51
The definition is taken from Alan Dundes’s introduction to Olrik’s essay (Study of Folklore, p. 129).
52
Olrik, ‘Epic Laws’, p. 132.

80
 Ad Putter

borne’ (Child 9A), ‘Wie wil hooren een goet nyeu liet?’ (van Duyse 22). But a more
economical solution is the use of an auxiliary verb that adds an ingressive aspect to the
action:
Daer soude haer een maget vermeyden (AL 22)
[A maiden was going out to amuse herself]
Daar zou ‘er een magetje vroeg opstaan (van Duyse 32C)
[A little maiden was to rise early]

In English the construction was + infinitive or the past progressive is the closest equivalent
to this use of zoude, which positions us towards the past as if it were ahead of us.53 As the
examples above show, Fier Margrietken begins in classic ballad style: ‘Het soude een fier
Margrietelijn’. We embark upon the past as if it were the future.
The law of closure, on the other hand, is typically observed by the ballad in a conclusion
that releases us from the tension of the story and moves us into a different temporal
modality. Olrik himself cites ballads as his examples: ‘Hundreds of folksongs end, not with
the death of the lovers, but with the interweaving of the branches of the two roses which
grow up out of their graves. In thousands of legends, one finds […] the punishment of the
villain appended to the principal action’. 54 The final stanza of the nineteenth-century
version of Fier Margrietken illustrates Olrik’s point, for it ends fittingly with the quartering
of the murderer. 55 The version in AL is subtler, and has a closer analogue in the Dutch
ballad Te Gherbeken binnen (AL 152), which tells the dramatic story of two lovers
murdered by a jealous husband, but which ends on a calm and lingering note in the final
couplet.
Hi namse daer beyde te gader, He took them both together,
Hi stackse in eenen sack, He put them in a sack,
Hi worpse al inder masen He threw them in the Maas,
Die mase die was nat. The river Maas was wet
Hie worpse al inder masen He threw them in the Maas.
Daer stont so menich here, so menich edel man, There stood so many a lord, so many a lord
Die dat met beweenden ooghen saghen an. Who saw this with tears in his eyes.

Fier Margrietken similarly concludes on a rallentando as the corpse drifts into the view of
the community who will tell her story:
Teghen stroom quam si ghedreven uut Against the current she washed up
Aen sint Jans cappelle. At the chapel of St John.
Dat sach so menich fijn edel man, So many a nobleman witnessed that,
So menich jonc gheselle, ja gheselle. So many a good fellow, yes fellow.

The complicated perambulations of Margaret’s real corpse — first fished up from the river,
then buried in a shallow grave, and finally led in solemn procession to a makeshift wooden
chapel at St Peter’s — have been telescoped into a single visualisable tableau, curiously

53
Middelnederlandsch woordenboek, s.v. sullen (3).
54
Olrik, ‘Epic Laws’, p. 132.
55
Gielemans too, could not let Margaret’s murder go unpunished, adding the capture and killing of the villains to
Caesarius’s account.

81
The Singer and the Scribe

reminiscent of the wall painting that was once in her chapel. 56 Perhaps the formula ‘dat sach
so menich edel man’ was originally meant to bring to mind the role of the Duke of Brabant
and his household, but the ‘menich jonc gheselle’ are the ‘good companions’ who sing and
listen to ballads.57 As psychological experiment has shown, stories have a better chance of
being remembered if they impinge upon the present. In the typical ballad closure, the story,
rounded off, enters the horizon of the present. The ballad thus etches itself into our
memories not only by virtue of its vivid story but also by making its entry into our world an
object of representation.
2. The Law of Two to a Scene
Only two individuals can interact in folk narrative, and these two individuals are further
subject to the Law of Contrast.
These laws need no further elaboration, for they correspond, after all, with our basic
categories of understanding (Self and Other, Good and Evil). In Fier Margrietken, the Law
of Two to a Scene accounts for the reduction of a band of criminals (who are not always of
one mind) to one single-mindedly evil villain. An important stylistic corollary of the Law of
Two to a Scene is the ballad’s preference for dialogue and questions and answers between
two speakers. Thus Margaret, who says not a word in Caesarius and Gielemans, has become
talkative in the ballad. Even if there are not two humans on the scene in a ballad a speaking
partner can be found for the protagonist, often in the shape of a bird or a tree. 58 The
narrative, too, is in dialogue with itself, asking questions and receiving replies: ‘Wat vantse
in haren weghe staen? / Eenen ruyter stille’. 59 These rhetorical questions are not simply
ballad formula but the hallmarks of oral communication, which, as Eric Havelock writes,
‘is “other-oriented” […] in the sense that the other is an audience, a “public” external to
the speaker, often symbolized in the vocative as a single person, but always palpably felt as
a listener who is a partner in the poetry’. 60
3. The Law of the Number Three
In the ballad, as in the popular proverb, ‘two is company, three is a crowd’, i.e. there are
two main characters per scene; any group of three is meant to represent a multitude. Thus
Margaret has three brothers: ‘Ic hebbe noch drie ghebroeders stout, / Si sullen u dat hooft af

56
See above page 72 and n.12.
57
Cf. the inscribed audience of Als al de eyckelen rijpen (Van Duyse 212): ‘Ghesellen, wilt dit onthouwen’
(‘Fellows, remember this’).
58
For examples of talking birds and trees in Dutch ballads see van Duyse 12 and 44 and Kalff, Lied, pp. 350-2.
For examples in the British ballad see Wimberly, Lowry Charles, Folklore in the English and Scottish Ballads,
2 nd ed. (New York: Dover, 1965). There is no need to attribute all such phenomena to superstitious beliefs in
the transmigration of souls to birds and plants, as Wimberly does.
59
Compare with this the German ballad Der verlorne Schuh (Meier 76), ‘Was fand es an dem Wege stan? Ein
Kneblein das was wolgetan’ (‘What did she find along the way? A boy who was most handsome’), and the
later Dutch ballad De Zavelboom, ‘Er zou een maagd om bloemetjes gaen / Om een wandeling te doene; /Wat
vond zy onder haer wege staen? / ’t Was een zavelboompje groene’ (‘A maiden was out to pick flowers / To
take a little walk; / What did she find on the way? / A little green juniper’).
60
Havelock, Eric, The Literate Revolution in Greece and its Cultural Consequences (Princeton: Princeton
University Press, 1982), p. 20.

82
 Ad Putter

houwen.’ In the ballad Des hadde een Swave een dochterlijn (AL 29) a maiden similarly
attempts to protect herself from her assailant by invoking her three brothers: ‘Noch hebbe ic
stoute broeders drie / Eenen rijcken vader tsoheyme’ (‘And I have three brothers bold and
strong / A rich father at home’). 61 The introduction of three brothers bring us to another
law.
4. The Law of the Family
As Max Lüthi has observed, in ballads protagonists tend to be located in the context of their
families; the law is perhaps most strikingly observed in the Judas ballad, which places the
betrayal of Christ by Judas in the background of Judas’s prior betrayal by his devious
sister.62 The history of Margaret has also been reshaped in accordance with this law.
Margaret, who lived away from home with her relatives, and whose parents are not even
mentioned in the early sources, has in the ballad acquired not only three brothers but also a
mother. And as mothers do in ballads,63 Margaret’s mother anxiously guards her daughter’s
sexual purity: ‘Uwen wille en doen ic niet / Mijn moerken soude mi schelden’.
5. The Law of Chronological Incompatibility
Events in folk narrative happen successively but never simultaneously; there exists only one
‘theatre of action’, as Vladimir Propp put it.64 Of course this law is not a peculiarity of the
Sagenwelt but of a fact of human experience. For although in life there are simultaneous
theatres of action, we are more aware of them than the ballad is. The recurrent ballad motif
of the protagonist looking behind him, or over his shoulder, to see, for example, his
relatives coming to the rescue is the ballad’s ingenious answer to the problem of how to
represent simultaneous events while safeguarding unity of action: we remain in one theatre
of action but from there see into a second.65 The law of chronological incompatibility
means that the ballad cannot tell the story as it really happened: Margaret had been sent to
buy wine, and, in the meantime, her foster family was murdered. Caesarius and Gielemans
can conjoin these two plot lines quite simply with the adverb interim, but in the ballad the
narration of simultaneous events is just not acceptable.
6. The Law of Action
‘Each attribute of a person must be expressed in actions — otherwise it is nothing.’
Again this law is not restricted to the Sagenwelt but belongs to oral literature in general.
To be memorable things must be visualisable and ‘so the preferred form of statement for
memorisation will be one which describes “from action”’. 66 The abstract propositions in
Margaret’s offical record — that ‘simplicity and an innocent life made her a martyr’, that
‘she served humbly and diligently’ — have no place in the ballad, which must make

61
van Duyse 288. The Dutch ballad was adapted from a German original, Der Schabentöchterlein (Meyer 75).
62
‘Der Familarismus der Volksballade’, in Lühti, Max, Volksliteratur und Hochliteratur (Bern: Francke Verlag,
1970), pp. 79-89.
63
See e.g. van Duyse, pp. 213 and 216.
64
Propp, V.I., Theory and History of Folklore (Manchester: Manchester University Press, 1984), trans. by Ariadne
Y. Martin and Richard P. Martin, ed. by Anatoly Liberman, Theory and History of Literature, 5), pp. 24-25.
65
For examples of this motif in the Dutch ballad see van Duyse, 17B-F.
66
Havelock, Literate Revolution, p. 137.

83
The Singer and the Scribe

concrete any abstract nouns (‘simplicity’, ‘innocence’) and unvisualisable abstractions

(‘virginity’). Few ballads do this better than Fier Margrietken, which turns Margaret’s
jealously guarded pitcher into an objective correlative of both her innocence and her
virginity. Like a child who has been sent out on an errand, and is overwhelmed by the
awesome responsibility of looking after her mother’s money, Margaret has blotted out all
thoughts apart from that of keeping her wine safe. So it falls to the sinister squire to tell her
that she has other things to worry about: ‘“En sorghet niet voor den coelen wijn, / Mer
sorghet voor u selven”’ Even after the rape she knows only that she must return with her
pitcher of wine:
Smorgens ontrent der middernacht
Si ghinc haer kanneken soecken,
Daer lach die moeyaart ede hi loech:
‘Het state daer teynden mijn voeten.’

Of course, the wine has been drunk; and the cruel breach of a child’s trust becomes a
metonym for her physical violation. The resources of tradition come to the poet’s aid.
‘Pouring wine’ or ‘tapping wine’ is a familiar euphemism in the ballad for sexual
intercourse,67 and by implication Margaret’s worry that her wine may be spilt signals in
advance (to the initiated, but not alas to Margaret) the threat to her virginity. Fier
Margrietken involves us in the ironic conflict between all that we know about sex and
sexual innuendo, and the little that Margaret knows — between a cynical world in which
‘spilling wine’ is slang for coitus and Margaret’s innocent world in which wine is only
wine, and yet the fear of spilling it more overwhelming than in the world that knows what
this loss really represents.
The few attributes that Fier Magrietken does contain actually support Olrik’s general
thesis. Thus when the ballad describes Margaret as ‘fier’, the adjective serves not primarily
as an insight into her character but as a cue to the actions that are to come. ‘Fier’ is the
traditional epithet for the girl who does not want a lover,68 but who inevitably acquires one
all the same. The official hagiographers vexed themselves needlessly about Margaret’s
epithet — how could this humble girl be called ‘proud’? 69 — for its function is not to
describe her character but her role in the story. In the case of the ‘ruyter stille’ who blocks
Margaret’s way, ‘stille’ is likewise not an attribute of the man but of his future actions: ‘Hi
namse in sinen witten armen / Heymelijc al stille’. Olrik’s dictum might thus be revised as
follows: attributes express actions, and actions express attributes.
From this brief overview of the ‘laws’ of the ballad and oral tradition, it will be evident
that the original history of Margaret has been reshaped in accordance with the ballad
schema. The most important schema behind Fier Margrietken, however, is a particular type
of ballad, well represented in the German and Dutch corpus, concerning a girl who, in
defiance of social convention, leaves home to fetch wine but fetches up with a man instead.

67
See AL 37 and 57.
68
Compare the ballad of Fiere Marienette (Kalff, Lied, p. 449) or ‘La Fiere’ in the romance Ipomedon by Hue de
Rotelande.
69
The ingenious answer worked out by the Bollandists is that Margaret was ‘proud’ only in the defence of her
chastity.

84
 Ad Putter

I should like to give two examples of this ballad type of the wilful daughter. I hope that the
similarities with Fier Margrietken will be obvious enough to require no further comment.
The first example, Het soude een meysken gaen om wijn, has an internal refrain
(italicised below), repeated in every stanza:
Tsou een meisken gaen om wijn, A girl was going out for wine
Hout u canneken vaste Hold your pitcher firmly
Savons in den maneschijn. In the evening in the moonshine
Bij nachte, by nachte At night-time, at night-time
Hout u canneken proper Dianeken, Hold your pitcher properly, Diana,
Hout u canneken vaste Hold your pitcher firmly

Wat vants in haeren weghen staen? What did she find on her way?
Een fijn gesel en dat was waer. A fine fellow, and that was true.

Den ruyter sprac dat meysken toe, To the girl the squire said,
Oft sij sijnen wille wou doen. Would she do his will.

Hoe weygerich dat dat meysken was, Reluctant though the girl might be,
Hy swanckse neder int groene gras. He laid her down on the meadow green

Doen hy syn willeken hadde gedaen, When his will had been fulfilled,
‘Schoon lief, ghij moecht wel thuyswarts gaen.’ ‘Fair love, you should be going home.’

Die ons dit liedeken eerstmael sanck He who first sang us this song,
70
Syn bellekens en gauen geen geklanck. His little bells made not a sound.
Johannes Koepp, who discovered this analogue to Fier Margrietken in the songbook
Pratum musicum (Antwerp, 1584), pointed out that a very similar refrain occurs in a Ger-
man ballad from the Ambraser Liedbuch (1586), also on the theme of the wilful daughter:
Es hett ein Schwab ein Töchterlein, A Swabian man had a daughter,
Halt die Kanne feste, Hold your pitcher firmly
Es wolt nit lenger ein Megdlein sein She wished to be a maid no more
Bei Nachte, fein sachte, At night-time, sweet and soft,
Halt die Kanna, schöne bas Anna Hold you pitcher, fairest Anna
71
Halt die Kanna feste. Hold your pitcher firmly.

My second example, apparently unknown to Koepp, is the widespread German ballad

Wein holen. A fragmentary first stanza of a ballad of this type is extant in a songbook ca
1535, 72 and further evidence of its early popularity is afforded by a number of fifteenth- and
sixteenth-century ad notam references to ballads beginning Es sout ein meiskin holen win or

70
The implication of the last stanza (conventional in the German and Dutch ballad) is that the singer of the song is
the ‘ruyter’ in the story, who absconds quietly on his horse. I cite the text from Leven en Werk van de
Antwerpse Luitcomponist Emmanuel Adriaenssen, ed. by Godelieve Spiessens, Verhandelingen van de
Koningklijke Academie voor Wetenschappen, Letteren en Schone Kunsten van Belgïe, Klasse der Schone
Kunsten, Jaargang 36 (Brussels: Paleis der Academïen, 1974), 2 vols, II, no. 57.
71
Cited by Koepp, Untersuchungen, p. 113.
72
The first full version is from a songbook of 1696. See Deutsche Voklslieder mit ihren Melodien, ed. by Rolf W.
Brednich, vol. VI (Freiburg: Verlag des Deutschen Volksliedarchivs, 1974), pp. 120-137.

85
The Singer and the Scribe

Het soude een meysken gaen om wyn.73 Although the earliest complete versions of this
ballad type are much later, we cannot therefore afford to ignore them, particularly since,
despite their lateness, they show marked similarities with Fier Margrietken. I cite below the
full text of an eighteenth-century version:
Es wollt’ ein schwarzbrauns Mägdelein A dark-brown maiden wished to go
Zum roten kühlen Wein, To get some cool red wine,
Zu Strasburg wohl über die Strassen, ja Strassen. On the road to Strassbourg, yes on the road.

Was begegnet ihr allda? Whom did she there meet?

Ein wunderschöner Knab’: A beautiful young man:
‘Feins Mägdlein, wollt ihr euch lassen, ja lassen?’ ‘Fine maid, do you want to play, yes play?’

‘Lasset ab, lasset ab, ‘Leave go, leave go,

Mein wunderschöner Knab! My beautiful young man!
Mein Mütterlein das tät mich schelten, ja schelten.’ My mother would scold me, yes scold me.’

‘Ei, warum verschütt’st du mir ‘Oh, why do you make me spill

Den roten, kühlen Wein The wine that is red and cool
Und bringst mich wohl um das Gelde, ja Gelde?’ And waste my money, yes money?’

‘Ja nicht um das Geld ‘It’s not for your money

Und den roten kühlen Wein, And the wine that is red and cool
Dafür darffst du nicht sorgen, ja sorgen: That you need to care, yes care:

‘Ist uns der Herr Wirt ‘The taverner he is to us

Der allerbeste Freund, The best of friends,
Der tut uns ein Kännlein borgen, ja borgen.’ He will credit us a pitcher, yes pitcher.’

‘Und wenn er uns ‘And if he will

nicht borgen will, Not lend to us,
So wollen wir ihn bezahlen, ja bezahlen,’ Then we will offer to pay him, yes pay him’

‘Mit lauter Silber, ‘With finest silver,

Mit lauter Gold, With finest gold,
Mit lauter dicken Talern, ja Talern.’ With fine thick coins, yes coins.’

Es hat das schwarzbraun Mägdelein The dark-brown little maid

Ihr Pantöffelein verloren, Has lost her little slipper
Sie kann es nicht mehr finden, ja finden. No longer can she find it, yes find it

Sie suchts wohl hin, sie suchts wohl her, She looks for it here, she looks for it there,
Und da sie es gefunden hat, And when at last she has found it,
Vor Freuden tät sie springen, ja springen For joy she starts to leap, yes leap.

Über zwei Berg und auch tiefe Tal Beyond two hills and valleys deep,
Da laüft ein schnelles Wasser, There runs a rapid stream,
Und wer sein Liebelein nicht länger haben will And he who no longer wants his love
74
Der kans ja lassen, ja lassen. Can leave her there, yes there.

73
Leven en Werk, ed. by Spiessens I, 231.
74
Meier 78B.

86
 Ad Putter
The family resemblance between these two ballads and Fier Margrietken suggests to me
that the history of Margaret was digested and remembered on the basis of the ready-made
schema provided by the ballad type of the accosted wine-fetching daughter. On the face of
it, the correspondences between this ballad type and the true history of Margaret’s murder
are slender and accidental — it so happened that Margaret had been sent out to fetch wine
from the winepress before being raped and murdered — but popular tradition had
developed for such experiences a morphology in the form of ballads about adventurous girls
who stray from their homes, and the ballad of Fier Margrietken naturally made good use of
it, as witness the revised story line: Margrietken meets with a ‘ruyter’, who takes her to a
tavern; he tells her she has no need to worry about her wine since they know the taverner,
who will supply more wine on credit, and so on. Of course the resemblance is imperfect.
Margrietken is out fetching wine because she has been misled (‘Si was daer toe verraden’),
not because she wants to; she does not end up looking for her lost ‘slipper’ or ‘shoe’, as
women who have lost (or are about to lose) their virginity commonly do in the ballad,75 but
for the pitcher of wine that had been entrusted to her. But despite these differences, Fier
Margrietken provides excellent support for the theory that remembered stories are subject
to ‘genre convergence’.76 In the course of memorial transmission the narrative gradually
changes to conform to the norms of the re levant genre (in this instance the ballad) or a
salient subgenre (the ballad type of the wine-fetching daughter). This explains why in the
nineteenth-century version of Fier Margrietken Margaret leaves home of her own accord,
and why the squire she meets has become ‘haar lief’ (‘her lover’). The ballad has lost its
individual characteristics and has converged so completely with the genre that Margaret of
Louvain is no longer recognisable in the ballad heroine.
The official hagiographers of Magaret must have been alarmed at the transformation of
the saint into the adventurous maiden of the ballad. The first reader response to Fier
Margrietken that I know of is that of a scandalised cleric, Johannes Molanus, who in his
Natales Sanctorum Belgii (Louvain, 1616-26), wrote that ‘there is extant a ditty about the
blessed Margaret […] in a book of profane songs, but the song is lascivious and deserves to
be burned’. In this intolerant criticism lies a kernel of truth: lasciviousness belongs firmly to
the type of ballad with which Fier Margrietken had been converging. But we are fortunate
that Fier Margrietken was spared the flames, for it is an exquisite ballad, much superior to
the nineteenth-century version and to the kinds of ballad that helped to shape it. Somehow
AL captured the song at just the right moment of its convergence. The ballad tradition of
wine-fetching maidens has influenced Fier Margrietken enough to energise its world with
libido and sexual innuendo, but Fier Margrietken has not advanced upon that tradition far
enough for the heroine to lose her childlike naivety. She wanders innocently with her
pitcher into a lewd scenario recognisable to anyone apart from herself. If the ballad schema

75
The loss of the shoe hints at the loss of virginity. See also the German ballad Der verlorne Schuh (Meier 76) and
the discussion in Deutsche Volkslieder, ed. by Brednich, p. 116.
76
Rubin, Memory, p. 280.

87
The Singer and the Scribe

had not modified the original story of Margaret’s murder this tension between an
unknowing heroine and a knowing world could not have reached its pitch of intensity, and
the tension would be lost if the influence of the schema had been any more extensive. Our
best ballads have been arrested at just the right stage of their development towards the
schema; I think Fier Margrietken is one of them.

88
 Looking up at ‘Holger Dansk og Burmand’ (DgF 30)

William Layher

One of the longstanding preoccupations in the Scandinavian ballad field concerns the
question of origins: how old are the oldest Scandinavian ballads, and where did they come
from? Although an oral ballad tradition survived well into the twentieth century in isolated
pockets of Norway, Sweden and the Faeroe Islands,1 the early days of this tradition are only
dimly understood. It is commonly agreed that the Nordic ballad was already well-defined
long before the first ballad manuscripts were assembled in Denmark in the mid to late
sixteenth century. How much earlier is uncertain; apart from a few suggestive references in
Swedish chronicles and other historical sources to public performances of narrative ‘songs’
that seem to resemble ballads,2 other attempts to date the origins of the first Scandinavian
ballads through comparative evidence gleaned from the fields of anthropology, history,
folklore or other literary sources have produced mixed results, with conclusions that range
from as early as the eleventh to as late as the middle of the sixteenth century. This dispute
may never be resolved, for the shortcomings of the manuscript sources for the period before
approximately 1550 are as legendary as they are insurmountable.3
Occasionally, however, evidence about the early days of the Nordic ballads comes to
light where it is least expected. A map of Greenland made by the Danish cartographer
Claudius Clavus in 1425 is perhaps the best known case of what we might call accidental
ballad transmission in the Scandinavian field, for this map uses words and snatches of verse
from a Nordic ballad as markers for the unnamed rivers, bays and inlets on the Greenland
coastline. Some have argued that Clavus composed the verses himself, while others read
them as a copy (or parody) of an older ballad, but regardless of provenence this map
represents the oldest verifiable textual record of a Nordic medieval ballad that we have
today.4 Another similar case — one that has not received the attention it deserves —

1
See most recently, for the Swedish tradition, Jansson, Sven-Bertil, Den levande balladen: medeltida ballad i
svensk tradition (Stockholm: Prisma, 1999).
2
Jonsson, Bengt, Svensk balladtradition (Stockholm: Svenskt visarkiv, 1967), p.19f. Jonsson defines the medieval
Scandinavian ballad as ‘a genre of orally transmitted song that is defined by its form (two-lined stanza with
one or two burdens, four-lined stanzas with one burden), its narrative content and its objective style, the latter
characterised not least by the frequent use of formulaic expressions and so -called commonplaces’ (p. 855).
3
For a critical discussion of the ‘medieval’ roots of the Nordic ballads see Layher, William, Queen Eufemia’s
Legacy: Middle Low German Literary Culture, Royal Patronage, and the First Old Swedish Epic (1301)
(Harvard University dissertation 1999), pp. 46-73, and idem, ‘Killing Erik Glipping, 1286: on the early years
of a Danish historical ballad,’ Lied und populäre Kultur/Song and Popular Culture: Jahrbuch des Deutschen
Volksliedarchivs, 45 (2000), pp. 13-34.
4
On the Clavus map see Hildeman, Karl-Ivar, Medeltid på vers. Litteraturhistoriska studier, Skrifter utgivna av
svenskt visarkiv, 1 (Stockholm: Almqvist & Wiksell, 1958), pp. 155-61 and Colbert, David W., The Birth of
The Singer and the Scribe

concerns a series of murals painted on the ceiling of a Swedish church in the fifteenth
century. High overhead in the eight triangular fields in the crown of vault IV of Floda
church, which lies some 80 km west of Stockholm in Södermanland province, is a
magnificent setting of eight warriors battling each other in single combat. Interestingly,
only one of these four pairs of combatants (David and Goliath) is drawn from Biblical
sources; the remaining six figures are not prophets, patriarchs, Biblical characters or figures
of sacred iconography but rather characters drawn from the vernacular literature of
medieval Scandinavia. Vault IV shows David killing Goliath, Sven Fötling beheading a
troll, Dietrich von Bern attacking Videke Velandsson, and Holger Dansk decapitating
Burmand with one stroke of his sword. Each of these figures is identified in a small text
scroll. Their names are given as dauit rex and golliat, swen fötling and trullet, diderik van
beran and wideke welandesson, but for the fourth pair, some extra information of note to
ballad scholars is given: the villain is named burmand while the hero is identified not by
name but by the words hollager dansk han van seger af burmand (‘Holger Dansk, he won
the victory over Burmand’) — a passage which is identical to the refrain of a popular
Danish and Swedish ballad about this famous battle. The oldest extant copy of this ballad is
found in a Danish manuscript from around 1580, but the text and image on the ceiling of
Floda church, which was decorated with murals around 1480 -1485 (see Figure 1), indicate
that a ballad on this topic was in circulation in eastern Sweden in the last quarter of the
fifteenth century, some hundred years before the oldest surviving copy of the ballad appears
in manuscript form.
The ballad pressed into service on the Claudius Clavus map derives, ultimately, from an
episode in Hervarar saga, an Icelandic ‘legendary saga’ of the thirteenth century, but it was
only after a twisted and elliptical route of oral transmission, redaction, truncation (and
perhaps parody?) that the ballad phrases were copied onto Clavus’s map. The ballad of
‘Holger Dansk og Burmand’ also goes back to a saga source — in this instance, a portion of
Karlamagnús saga — but the steps in its journey from saga prose to rhymed narrative song
have never been retraced in detail. 5 And yet, the surviving texts take us only so far. A
complete reconstruction of the early days of ‘Holger Danske og Burmand’ must take the
visual evidence on the ceiling of Floda church into account as well — and it is here, in the
murals, where the present investigation takes an interesting turn. Although the banderole
text hollanger dansk han van seger af burmand confirms that a ballad on this topic was
sung in Sweden in the 1480’s, the pictures of Holger Danske and, especially, of Burmund,
contain details that suggest that the artist may well have used another source besides the
fifteenth-century ballad when painting Holger Danske and Burmand on the ceiling of vault
IV.

Floda Church
The oldest walls of Floda church date to the twelfth century, but the nave and choir were
expanded in 1412-13 to roughly the dimensions they have today. Restoration efforts begun
after a disastrous fire in 1414 likely led to the first set of murals being painted in Floda

the Ballad: the Scandinavian medieval genre, Skrifter utgivna av Svenskt visarkiv, vol. 10 (Stockholm:
Svenskt visarkiv, 1989), pp. 56-61.
5
See, for example, the brief overview in Jonsson, Svensk balladtradition, pp. 26-27.

90
 William Layher

church around 1420. Traces of a Crucifixion scene are still visible in the church’s original
choir, i.e. in the easternmost vault in the northern aisle.6 A new vaulted roof was completed
around 1450, and the vaults over the northern aisle of the nave were decorated with murals
approximately thirty years later.
Three escutcheons in the northern aisle indicate that the murals were painted in the
1480s. Two of the family crests pay tribute to powerful local families with connections to
Floda church, while the third is a tribute to the local bishop’s family. The first escutcheon
shows a lily with the inscription Germunt Larenson wapn and the second depicts a boar’s
head with the text Hince jonsons oppå kopperberget. Germund Laurensson (d. 1459), a
member of the Lillie family, was one of the largest landowners in Floda parish. His
daughter Birgitta married a wealthy and influential local farmer named Mats Kagg shortly
after 1450. It is believed that Germund Laurensson donated much of the money towards the
construction of the vaulted roof in Floda church in the 1450s. The other family crest, that of
the boar’s head, refers to Hinse Jönsson of the Svinhuvud family. Jönsson, a wealthy
Stockholm businessman who was still alive as late as 1485, did not own a local estate, but
his daughter married Anders Laurensson (a brother or perhaps a son of Germund
Laurensson?), in the 1480s. The murals may well have been commissioned to
commemorate their wedding. A terminus post quem for the murals is indicated by the third
escutcheon, the family crest of Bishop Kort Rogge of Strängnäs (d. 1501), who was
elevated to that office in 1479. Despite the presence of his crest in Floda church it is
unlikely that Bishop Rogge commissioned the murals himself upon his installation; the
Rogge insignia — positioned underneath the enthroned Virgin Mary in the place of honour
in the first vault — is understood as a tribute to the seated bishop at the time of the painting,
not as an indication of artistic patronage.7
The murals in Floda church are attributed to Albertus Pictor, the most distinguished and
gifted church muralist in late-medieval Sweden, whose work graces some three dozen
churches in southeastern Sweden.8 His name was first attested in 1473, in a document
recording his marriage to the widow of master painter Johan Målare in Stockholm. Albertus
may have been a painter by trade — through his marriage to Anna, the widow of Johan
Målare, Albertus in all likelihood gained access into the city’s guild of painters — but he
earned a significant portion of his income as a textile craftsman in Stockholm, where he

6
The Birgittine iconography in this Crucifixion scene suggests a dating to the early fifteenth century. On these
murals see Bennett, Robert, Floda kyrka, Sveriges kyrkor, 205 (Stockholm: Almqvist & Wiksell Inter-
national, 1988), pp. 60-61.
7
Bennett, Floda kyrka, pp. 43-45. On the political and intellectual impact Bishop Rogge had on fifteenth-century
Swedish society, see Öberg, Jan, ‘Von Humanismus zum Traditionalismus. Die Einwirkung der politischen,
gesellschaftlichen und kirchlichen Verhältnisse auf das Kulturleben in Schweden am Beispiel von Kort Rogge
(um 1420-1501)’, in Ut granum sinapis: essays on neo-Latin literature in honour of Jozef Ijsewijn , ed. by
Gilbert Tournoy and Dirk Sacré, Supplementa Humanistica Lovaniensia (Leuven: Leuven University Press,
1997) 12, pp. 24-34.
8
Among the standard works on Albertus Pictor in Swedish are Lundberg, Erik, Albertus Pictor (Stockholm: P.A.
Norstedt, 1961); Cornell, Henrik, and Sigurd Wallin, Albertus Pictor. Sten Stures och Jacob Ulvssons målare
(Stockholm: Bonniers, 1972); see also, in English, the entry in Grove Dictionary of Art, ed. by Jane Turner,
34 vols (London, New York: Macmillan, 1994) I, 576.

91
The Singer and the Scribe

sewed and decorated religious vestments for use by the Church. Several city documents in
Stockholm refer to him as ‘Albert Pärlstickare’ (the pearl embroiderer) and attest that he
was still alive in 1509. His artistic talents were diverse: an elaborately carved and painted
altarpiece for the Nådendal convent in Finland has been recognised as one of his last works,
and he also earned a small sum as an organist and musician at the funeral of a fellow
guildsman in 1509.
Albertus Pictor’s strong connections to élite society in Stockholm strengthen the
hypothesis that Hinse Jönsson was indeed the patron of the murals in Floda Church. Of the
three dozen churches with murals attributed to Albertus Pictor, Floda church is one of the
most distant from the city of Stockholm. Although he painted in churches at a considerable
distance from Stockholm in Uppland province and Västmanland province, both to the north
of Lake Mälaren, Albertus received very few commissions to the south and west of the
Mälaren region and none more distant than Floda parish. Given Albertus Pictor’s attested
wealth and success in Stockholm, 9 it is very likely that Hinse Jönsson and Albertus Pictor
had become acquainted there, perhaps through business dealings or other artistic
commissions, and that these personal connections encouraged Albertus to accept the Floda
commission in the countryside west of the city. But before we turn our eyes to the visual
evidence linking the mural to the medieval ballad, let us discuss the ballad, its sources and
its monstrous figures.

‘Holger Dansk og Burmand’

The ballad survives in four Danish variants recorded in the late sixteenth and early
seventeenth century and in two Swedish versions that date from around 1630. 10 No records
of this ballad exist in the Norwegian, Icelandic or Faeroese tradition. The ballad is
summarised as follows:
TSB E 133 Holger Danske og Burmand — Princess releases prisoner Ogier le Danois to
fight giant for her.
The giant Burmand asks the king for his daughter Gloriant’s hand. Gloriant will not have
Burmand. She goes to her father’s pris on and asks one of the prisoners, Holger Danske, to
fight Burmand for her. She gives him a horse and a sword. Holger fights Burmand and kills
him. 11

9
Albertus owned two stone houses at Norreport in Stockholm, and paid taxes at a higher rate than any of his
colleagues; see Lundberg, Albertus Pictor, p. 11.
10
Danmarks gamle folkeviser, ed. by Svend Grundtvig, 12 vols (Copenhagen: div. publ., 1853 -1976) I, 384-97;
where the ballad is catalogued as DgF30. Svenska fornsanger; en samling af kampavisor, folk-visor, lekar och
dansar, samt barn-och vall-sanger, ed. by Ivar Arvidsson, 3 vols (Stockholm: Norstedt, 1834-42) I, 75-86.
The Swedish multiforms of this ballad are also taken up in the recent edition of Sveriges medeltida ballader;
see Sveriges medeltida ballader, ed. by Bengt R. Jonsson, Sven-Bertil Jansson and Margareta Jersild, 7 vols
(Stockholm: Svenskt visarkiv, 1983), 5.1, 104-112, where the ballad is catalogued as SMB 216. All ballad
translations are my own. Further references to these editions will be made parenthetically in the text.
11
Bengt Jonsson, Svale Solheim and Eva Danielson, The Types of the Scandinavian Medieval Ballad: a
descriptive catalogue, Skrifter utgivna av Svenskt visarkiv, 5 (Stockholm: Svenskt visarkiv, 1978), p. 255.

92
 William Layher

The earliest traces of the Holger Dansk figure are found in Old French literature: a
character named Oger de Denemarche [or: Danemarche] plays a minor role as one of
Charlemagne’s paladins in the early twelfth century epic La Chanson de Roland.12 But
already by the early thirteenth century, his heroic vita became the subject of independent
chansons de geste in the French tradition (e.g. Les Enfances Ogier [1195], La Chevalerie
Ogier de Danemarche [1200-1215]), and this new, expanded story left important traces in
other Old French works like the Pseudo-Turpin chronicle. 13 These younger texts form the
basis for Holger’s earliest reception in Scandinavia, which was accomplished through
Karlamagnús saga, a cycle of chansons de geste — all of them related to Charlemagne and
his deeds — that was translated into Old Norse in the middle of the thirteenth century.
Scholars recognise ten distinct narrative branches in Karlamagnús saga, whereby branch III
contains the story of Holger’s famous battle with Burmand.14
Although the ‘Holger Dansk og Burmand’ ballad ultimately goes back to scenes
narrated in Karlamagnús saga, it is highly unlikely that the Icelandic or West Norse saga
manuscripts served as the exclusive source for the ballad because new redactions of
Karlamagnús saga in Old Swedish and Old Danish began to appear in mainland
Scandinavia during the fifteenth century. One of them, the ODan Karl Magnus’ Krønike,
exists in a manuscript copy from around 1480,15 and in two printed versions, the first
edition in 1509 and a revised edition that appeared in 1534. 16 Another redaction of
Karlamagnús saga possibly translated from a lost Norwegian manuscript, the Old Swedish

12
On the historical development of the Holger Dansk figure see most recently Holzapfel, Otto ‘Holger Danske,’ in
Enzyklopädie des Märchens, ed. by Kurt Ranke and Rolf Wilhelm Brednich (Berlin, New York: Walter de
Gruyter, 1977-) 6, cols 1175-78; Togeby, Knud, Ogier le Danois dans les littératures européennes
(Copenhagen: Det danske Sprog -og Litteraturselskab, 1969) is indispensible reading.
13
See Togeby, Ogier le Danois dans les littératures européennes, pp. 38-109.
14
On the Old French sources of Branch III, entitled Oddgeirs tháttr danska in some manuscripts, see Skårup,
Povl, in ‘Contenu, sources, rédactions’, Karlamagnús saga. Branches I, III, VII et IX, ed. by Knud Togeby et
al., Ogier le Danois, 3 (Copenhagen: C.A. Reitzell, 1980), pp. 352-53, and Knud Togeby, ‘Filologiske Studier
over Karl Magnus’ Krønike (rev.)’ Studia Neophilologica, 37 (1965) 91-96. The manuscripts of Karlamagnús
saga are reviewed in Halvorsen, E.F., The Norse Version of the Chanson de Roland, Bibliotheca
Arnamagnæana, 19 (Copenhagen: Ejnar Munksgaard, 1959), pp. 32-37.and more recently in Loth, Agnete,
‘Les manuscrits norrois,’ in Karlamagnús saga, ed. by Togeby et al., pp 358-78. Unger’s 1860 edition of
Karlamagnús saga in normalised Old Icelandic is still the only complete edition of the saga; see
Karlamagnús saga ok kappa hans, ed. by C.R. Unger (Christiania: H.J. Jensen, 1860); a diplomatic edition of
branches 1, 3, 7, and 9 manuscripts A and B of Karlamagnús saga with a French translation in Karlamagnús
saga, ed. by Togeby et al.
15
Karl Magnus’ Krønike, ed. by Poul Lindegård Hjorth (Copenhagen: J.H. Schultz, 1960) and Hjorth, Poul
Lindegård, Filologiske studier over Karl Magnus’ Krønike (Copenhagen: J.H. Schultz, 1965). The oldest
manuscript version (ca 1480) is found in the ‘Børglum-Håndskrift’, Cod. Holm Vu 82, gl. nr. 12 b (Royal
Library, Stockholm). This Low Germany miscellany is described more fully in Borchling, Conrad,
‘Mittelniederdeutsche Handschriften in Skandinavien, Schleswig-Holstein, Mecklenburg und Vorpommern’,
in Nachrichten von der Königlichen Gesellschaft der Wissenschaften zu Göttingen, Philologisch-historische
Klasse, Beiheft (1900), pp. 109-13.
16
The different versions are evaluated in Karl Magnus’ Krønike, ed. by Hjorth, pp. xvi- xlviii. See also Hjorth,
Filologiske studier over Karl Magnus’ Krønike, pp. 97-157.

93
The Singer and the Scribe

Karl Magnus, dates to the first decades of the fifteenth century, in any case no later than
1420. 17 In contrast to Karl Magnus’ Krønike, however, the OSw text is an incomplete
treatment of its Icelandic source, as it contains only two of the later branches of the saga
neither of which concerns Holger Dansk. There is some evidence that a lengthier version of
the OSw Karl Magnus did exist at one time, however, because the language of the oldest
Karl Magnus’ Krønike manuscript of 1480 shows some significant Swedish influence
which can only have come from a lost Swedish codex of Karl Magnus that was more
complete than any of the surviving copies.18 Although the evidence for a lost version of the
OSw Karl is controversial, it leaves open the possibility that the battle between Holger
Dansk and Burmand was indeed in circulation — not merely in ballad form, but also in
Swedish prose — around the Lake Mälaren region in the latter part of the fifteenth century,
in the precise region of Sweden where Albertus Pictor did the majority of his church
painting.
That the ballad of Holger Dansk and Burmand was a popular subject in late-medieval
Denmark is confirmed by an editorial comment in another Holger Dansk book entitled
Kong Olger Danskes Krønike, which was published by Christiern Pedersen in 1534 (the
same year in which he printed the Karl Magnus Krønike).19 In his introduction to Kong
Olger Danskes Krønike, Pedersen makes reference to an old kempe vise or heroic ballad
about Holger Dansk, and he quotes the same ballad refrain that is found on the ceiling of
Floda church:
Den menige mand her i Riget viste icke før andet en at han hagde all
eniste verit en dansk kempe / som want seyer aff Burmand / som den
gamle kempe vise lyder / Olger dansk han want seyer aff Burmand.

[The common man here in this country knows nothing [about Holger Dansk] other than that he was a Danish hero
20
who won the victory over Burmand, as the old kempe vise says: Olger Dansk he won the victory over Burmand.]

17
Kornhall, David, Den fornsvenska sagan om Karl Magnus. Handskrifter och texthistoria , Lundastudier i nordisk
språkvetenskap, vol. 15 (Lund: GWK Gleerup, 1959), pp. 98-103. According to his reconstruction, the OSw
Karl Magnus was translated from a lost Norwegian manuscript of Karlamagnús saga (pp. 108-113).
18
On the disputed lost Swedish version of Karl Magnus, see Karl Magnus enligt Codex Verelianus och Fru Elins
Bok, ed. by David Kornhall, Samlingar utgivna av Svenskafornskriftsällskapet, 219, 63 (Lund: Carl Blom,
1957), pp. vii-xxviii; Kornhall, Den fornsvenska sagan om Karl Magnus, p. 298f; Hjorth, Filogiske Studier
over Karl Magnus Krønike, p. 161f and 309f, argues that there is no textual evidence proving beyond a doubt
the existence of a Swedish ‘missing link’ between the Icelandic saga and the Danish Karl Magnus’ Krønike ,
but admits that the Swedish inflections in the language of the oldest manuscript of the Danish text are best
explained if one assumes that a complete Swedish codex of Karl Magnus existed in the fifteenth century and
that this lost manuscript served as a source for the 1480 Danish text (p. 314); Kornhall, Den fornsvenska
sagan om Karl Magnus, p. 298f, disputes any direct link between the extant OSw and ODan texts, but agrees
that another version of Karl Magnus — one which was substantially different from the preserved mansucripts
— might have served as a source for the Danish translation of 1480.
19
Pedersen, Christiern, ‘Kong Olger Danskes Krønike’, in Christiern Pedersens Danske Skrifter, ed. by Carl
Joakim Brandt, vol. 5 (Copenhagen: Gyldendal, 1856), pp. 129-314. On the Old French source for Kong Olger
Danskes Krønike see Togeby, Ogier le Danois, pp. 227-229.
20
Pedersen, ‘Kong Olger Danskes Krønike’, p. 136 (my translation). Further citations to this edition will be given
parenthetically in the text.

94
 William Layher
Mindful of the ballad’s popularity in the Danish tradition, Pedersen is careful to point out in
his preface that while the foreign sources identify Holger’s opponent as ‘Brunamundus’
(p.164), he will call him ‘Burmand’ for good reason:
Denne samme kong Burnamand aff Egipten wor en mectig starcker
kempe / oc han kallis almindelige paa danske Burmand / thi vil oc ieg saa
kalle hannem her effter I denne Krønike / ath alle som høre eller lese
hanss naffn / ath de skulle vide hvad mand han wor. (p. 164)

[This king Burnamand of Egypt was a powerful champion and he is commonly known in Danish as Burmand,
and I will also call him that in this chronicle so that all who hear or read his name will know which man this
was.]

If the name Burmand was as popular in 1534 as Pedersen suggests, then we must consider
the source of its popularity. The Icelandic manuscripts of Karlamagnús saga calls him
‘Burnamend’ or ‘Burnament’, while Karl Magnus Krønike calls him ‘Bwrnemanth’ (1480)
or ‘Burnemand’ (1534). Only the Danish and Swedish ballad sources and Kong Olger
Danskes Krønike, whose debt to the ballad is made explicit in the preface, know him as
‘Burmand’ or ‘Burman’ (var. ‘Burrmand’, ‘Burmandt’, etc.). Pedersen makes it clear that
he is using the common form ‘Burmand’ in Kong Olger Danskes Krønike in order to avoid
confusing his audience in Denmark, for the latter name is the one which carries the weight
of tradition. The most likely source of this Danish familiarity with the disyllabic name
‘Burmand’ is the Nordic ballad which obviously enjoyed a commonality and universal
appeal that Pedersen’s new translation of unfamiliar French stories could not hope to equal.
This dispute over correct versus incorrect names also brings evidence of a stratified
reception at the turn of the sixteenth century to the surface: Pedersen suggests that non-elite
audiences who know the story of Holger Dansk solely through oral transmission (i.e.
ballads) will be receptive only to the name ‘Burmand’ while, on the other hand, elite
audiences of sufficient wealth and education to read the written sources — even though
they almost certainly knew the ballad as well — will recognise that ‘Burnement’ and
‘Burmand’ are the same figure, even if the orthography of the name is levelled in the 1534
edition so as to conform with its most popular (and populist) form. 21
Another point of disagreement among the sources concerns the manner of Burmand’s
death. All of the ballad variants agree that Burmand was killed at the end of his duel with
Holger Dansk, but his death is treated differently in the Icelandic, Danish and Swedish
sources. In Karlamagnús saga, Burmand is killed with a blow to the head and then — as is
common with giants, so as to prevent their reanimation after death — his head is amputated
from his torso. Burmand is killed with a blow to the chest, and then his head amputated in
Karl Magnus’ Krønike (p. 75) while Kong Olger Danske’s Krønike tells that Holger Danske
dealt Burmand a death blow from above that split his helmet down to the neck, causing the
victim to fall down dead (p. 170). In the Danish ballad variants, Burmand is simply

21
The fact that the name of Holger Dansk is likewise treated differently in the Icelandic, Old Danish and ballad
sources (Icl ‘Oddgeir’; ODan ‘Wdger, Olger’ [1480], ‘Olger’ [1534]; DgF 30 ‘Ollger Danske / Udgierd Dansk
/ Holgerd Dansk / Olger Dansk / Olger Danske’; Floda ‘Hollager Dansk’) should not detract from the main
point: that the form ‘Burmand’ attested in Floda church matches most closely with the surviving ballad texts,
and not with the surviving prose texts in any language.

95
The Singer and the Scribe

‘vanquished’ in general terms: thett var hynnd unge kong Burmand, faldt død for Holggers
fuod [ms A: the young king Burmand is struck, he falls down at Holger’s feet]; slagen bleff
goden Burmand Kamp, hand fall død til iord [mss BD: the valiant warrior Burmand was
slain, he fell dead to the ground]; kong Olger hand slaa then throll ihiell [ms C: king
Holger beat the troll to death] (I, 392-97). The Swedish variants, however, correspond in
close detail to Albertus’s artistic representation of the event on the ceiling of Floda church
— Burmand is killed with a single, decapitating blow to the neck: Så hugher han Burmans
hufve af, så blodet ran honom til dödha [ms A: thus he strikes Burmand’s head off, and he
bled to death], så högg han Burmans hufvud i tu, at blodet rann honom till döda [ms B: thus
he clove Burmand’s head in two pieces, that he bled to death]. I believe it is not a
coincidence that the Swedish ballads are most consistent with Albertus’s representation of
the scene — not because later ballad singers adapted their story so as to conform to the
visual evidence presented in Floda church, but because Albertus and the Swedish ballad
singers were following a lost common source that was unique to Sweden, one in which
Burmand, a monstrous figure with eyes like a cat, was killed with a single blow.

Burmand as monster?
By the time the murals were painted in vault IV, the figure of Burmand had undergone a
striking transformation in the written sources, from a haughty Egyptian king in the twelfth-
century Old French epics to something akin to a troll in the ODan translations of
Karlamagnús saga. Evidence of this progression can already be seen in two of the lead
manuscripts of Karlamagnús saga, known as A (AM 180c fol, ca 1350-1400) and B (AM
180d fol, seventeenth century). The older redaction A introduces Burmand as a king who
leads a company of twenty thousand brave warriors. No particular mention is made of his
appearance. But in the younger B manuscript, Burmand appears as a monstrous figure:
En Burnement hefir um allan aldr verit í bardögum, hann var mikill vexti ok illr kosti, svartr á hár ok hörund;
hann má engan mat eta nema hrán ok eigi vín drekka nema blóði væri blandat: hann hafði gul augu sem kettir,
ok Þó enn skygnri um nætr en daga. Þessi maðr var fullr galdra ok gerninga ok flærdar, ok mundi han tröll
kallar vera, ef hann kœmi norr hingat í heim.

[Burnament had been in many battles; he was large in size, and ill-shaped, black-haired and dark-skinned. He
ate no food except raw meat, and drank no wine unless it was mixed with blood. He had yellow eyes like a
cat, and could see better at night than during the day. This man was full of treacherous spells and tricks, and
22
he would be called a troll if he came northward into our land.]

22
Karlamagnus saga ok kappa hans, ed. by Unger, p. 108 (my translation). Ms B is a seventeenth-century paper
copy of a lost vellum manuscript that probably dates from the fourteenth century; see Halvorsen, The Norse
Version of the Chanson de Roland, pp. 32-35. We do not know if the details about Burmand’s monstrous
appearance in B were also present in its lost vellum source; perhaps they were added ex tempore by the
seventeenth--century copyist. Although clearly a redaction of Karlamagnús saga, the 1480 Børglum Karl
Magnus’ Krønike uses none of the extant saga manuscripts as its direct source. According to Hjorth,
Filologiske studier over Karl Magnus’ Krønike, p. 315, the ODan text was translated from a lost
Karlamagnús saga manuscript of the older A family — a significant claim, in that the A manuscripts do not
know Burmand as a troll — but the fact that the 1480 redaction does characterise Burmand as a monstrous
figure strongly suggests that its affiliation to the Aa manuscript family is suspect, and that another
Charlemagne/Holger Dansk/Burmand text in circulation in the fifteenth century influenced the

96
 William Layher

The ODan Karl Magnus’ Krønike follows the saga account in the younger Icelandic
manuscript very closely, omitting the references to Burmand’s powers of sorcery but
maintaining the description of his appetites, his strange eyes, and his troll-like appearance.
The 1480 redaction describes him thus:
Tha kam en k[onge] som bwrnamanth hether (…) hannum foldhae xxxM
men han otthae engen land tog bar han konngis naffn hans men skulle
altid sloss fenge the engen andre thaa sloues the inburdes alle hans dage
hadde han waretth I krygh ok orloff all hans madh skulle blandhes met
blodh han haffude øghen som en katth ok sogh baether om natthen en vm
daghen wore han her I landhet tha wore han lyger en troll.

[Then a king named Bwrnamanth arrived. He led a company of three thousand men. He did not own any land
[of his own] but he was known as a king. His men were always fighting. If they had no opponents they fought
amongst themselves. He spent all his days at war. All of his food must be mixed with blood. He had eyes like
23
a cat and saw better at night than during the day. If he were here in our land he would be called a troll. ]

The Danish ballads also know Burmand as a monstrous figure, their descriptions largely
following characterisations of him in Karlamagnús saga and the Krønike. Consider these
stanzas from variant B (Dg F, I, 393):
10. Manden er gram, oc hesten er grum, The man is gruesome, and his horse is fierce
det siger ieg eder for sant: I tell you this for true
ieg haffuer hørt, at dett er vist: I have heard, and swear it’s true
hand bider met ulffuens tand. he bites with wolfish teeth.

11. end blod med eeder blend. He will eat nothing else
end kiød aff christen mandt: than the meat of a Christian man
icke vill hand andet dricke, and wants to drink no other drink
Hand vill icke anet æde, than blood that is mixed with venom. (DgF, I, 393)

Variant A contains similar language describing Burmand’s Variant D also contains the
references to drinking blood and eating the meat of Christian men, although in this variant
Gloriant refers to Burmand explicitly as a ‘troll’:
13. Est du leffuende? Olger Danske, ‘Are you alive, Holger Danske
ieg siger dig paa min Sand: I tell you this for true
Her er en Trold, begerer mig, a troll [en Trold] is here, and wants to wed me
oc det er sorten Burmand. and he is called black Burmand.’

At the end of the D variant, Burmand’s fiendish nature is confirmed once again:
33. Olger reed til den skøne Iomfru: Olger rode to the lovely maiden.
“i tage nu eders Festemand: ‘Now you may join your betrothed.
Ieg haffuer dræbt met mit gode Suerd I have killed with my good sword here
den fule, forgifftige Aand.” the ugly, poisonous Thing (DgF, I, 395-97)

characterizations found in the Børglum manuscript; this discrepancy could be another shadow cast by the
‘missing link’ OSw Karl Magnus codex.
23
Karl Magnus’ Krønike, ed. by Hjorth, p. 60 (my translation).

97
The Singer and the Scribe
The Swedish ballads show Burmand as a grotesque figure whose appearance and appetites
have more than a shade of the comical. As Burmand sits on his horse and calls for the
maiden Gloria to come out to him, she replies,
17. Tu är inthe möcket vacker at se, ‘You are hardly pleasant to look at
tin kortel han är huiter, Your tunic it is white
tin näsa är tre alnar långh, Your nose is three spans long
tu est en skråpucke licker. And you look like a buffoon.’

In the next stanza, Burmand’s alleged ruthlessness is seriously undermined when he lets
forth an enormous fart as Holger approaches — a physical sign of weakness that betrays his
anxiety about facing Holger Danske in combat:
18. Burman holler för staden ut, Burmand waits outside the city
han lather sin rompa kröcka, And then his rump trumpets forth:
iagh fruchtar för holger danske, ‘I am afraid of Holger Dansk,
han skal och alla försöcka. He shall come and try his luck with me.’

In the light of the Swedish multiforms, it is notable that Albertus Pictor painted
Burmand not as a comical figure but as a monstrous one: note the shaggy fur, the cloven
feet, the beaked nose and the pointed chin, and the fact that Burmand wields a crudely-
fashioned club instead of a sword or another chivalric weapon: in contrast, the giant Goliath
wears plate armor and fashionable shoes, and brandishes a mace. Burmand’s eyes are also
exceptional in this mural. They are unusually well- defined, with arching eyebrows and
bright, glowing whites. By comparison, the eyes of the other figures in vault IV are
unremarkable. I would argue that the way Burmand’s eyes are depicted tells us a great deal
about the status of the Holger Dansk story in Sweden around 1485. We recall that Karl
Magnus Krønike says that Burmand ‘had yellow eyes like a cat, and could see better at
night than during the day’. Although none of the surviving ballads make any mention of
Burmand’s unusual eyes, Albertus obviously rendered the eyes with special care in his
mural. Since the entire study of medieval iconography is predicated on the assumption that
artists followed traditional themes (as they understood them) in their work, we must assume
for Burmand’s case in vault IV that Albertus Pictor knew some other version of the story
than that which is transmitted in the Danish and Swedish ballads — a version in which
Burmand’s cat-like eyes are one of his defining characteristics.
What, then, was Albertus Pictor’s source? It is highly unlikely that he ever came in
contact with the Icelandic manuscripts of Karlamagnús saga in the younger B redaction,
and even if he had, his comprehension of the Old Norse text would have been minimal. The
ODan texts are likewise ineligible as sources, because the younger incunabula prints of
Karl Magnus Krønike postdate the murals in Floda church while the oldest manuscript,
which was copied in Børglum in 1480, remained there, either in the private library of
Bishop Jakob Friis (d. 1486), whose escutcheon is found in the codex, or in the monastery
library in the city.24 There is no evidence that this manuscript or its contents had any
reception in Sweden. In fact, Albertus’ unique depiction of Burmand cannot be linked with
any of the surviving sources in the Nordic tradition. And yet the bright-eyed Burmand
remains. If we must identify a source for this, we might do well to consider some other

24
On the escutcheons in the Børglum miscellany, see Karl Magnus’ Krønike, ed. by Hjorth, pp. xx-xxiii.

98
 William Layher

factors which may have shaped Burmand’s reception in Sweden in the late fifteenth
century. Again our suspicions are drawn to the fabled “lost codex” of the OSw Karl
Magnus. It is assumed that this manuscript travelled from Sweden to Denmark at some
point before 1480, as it seems to have influenced the translation of the ODan Karl Magnus
Krønike. If this Karl Magnus was indeed a longer and more complete redaction of
Karlemagnús saga than the surviving copies, then it could be held responsible for more
than just the “extra details” about the monstrous Burmand that appear in the 1480 Børglum
manuscript of the Danish text; this OSw text — or some oral/traditional offshoot from it
that reached Albertus Pictor’s ears (another ballad? recitation from manuscript? local
legend?) — may also have influenced his artistic representation of the Burmand figure in
Floda Church.25

Gazing at the ceiling

Albertus drew much of the visual inspiration for his murals from scenes in the Biblia
Pauperum, a Dutch book of woodcuts printed in the 1460s, and from other blockbooks and
broadsides circulating in northern Germany and Sweden in the fifteenth century.26 This is
not to say that Albertus’s sources have been identified in every case; indeed, the battle
between Holger Dansk and Burmand appears to have no clear iconographic model, since
the portrayal in vault IV fails to match up cleanly with any of the known sources. Some
might argue that Albertus was working without a direct narrative source in this instance; it
is conceivable that Albertus modelled his shaggy Burmand upon the ubiquitous late-
medieval Wild Man figure, attested in southern Sweden in the Glimmigehus statue from the
late fifteenth century. This is highly unlikely, however, because the Wild Man was
commonly held to be a benevolent forest dweller, while Burmand’s malevolence is clearly
apparent in the mural and is also fundamental to the story itself.
If we want to analyse the Floda church murals in light of the Nordic ballads or other
written sources, we would do well to tread carefully. According to recent scholarship on the
dynamics of text and image in medieval vernacular culture, it is incorrect to regard images
as subordinate to the written word — as ‘illustrations’ of a particular text version, as it were
— since both image and text function as independent manifestations of ‘story’ in all of its
variety.27 For far too long, the Holger Dansk murals have been misunderstood and
mischaracterised as illustrations of the Nordic ballad texts we catalogue as Dg F 30 and
SMB 216. Albertus did not paint a ballad scene; what he did was represent a climactic
scene from the Holger Dansk story (as he knew it) in two adjacent frames in the ceiling of
vault IV. 28

25
The “lost codex” is not entirely theoretical. In the late sixteenth or early seventeenth century, the antiquarian
Johannes Bureus (d. 1652) noted the existence of what appears to be a Karl Magnus manuscript in Söderala
that is unknown to modern scholars and presumed lost; see Kornhall, Karl Magnus, p. ix.
26
Cornell and Wallin, Albertus Pictor, pp. 10-29.
27
See, for example, Rushing, James A., Images of Adventure. Yvain in the Visual Arts (Philadelphia: University of
Pennsylvania Press, 1995) and the bibliography listed there.
28
Rushing’s distinction between ‘representative scenes’ and ‘narrative scenes’ is helpful for the present discussion
of the visual evidence. Albertus’ murals are representative of the story of Holger Dansk because they do not

99
The Singer and the Scribe

One wonders, then, why these figures are painted there in the church. The eight warriors
in the crown vault take pride of place, but other Biblical and religious figures are depicted
in vault IV as well: Cain and Abel, St Olaf, St Antonius, the archangel Michael weighing
the soul of a dead man, Moses and the Israelites, John the Baptist, and even a rendering of
the Lord God.29 Occasional juxtapositions of secular and sacred iconography are not
unknown in medieval Scandinavian churches,30 but an intermixing on this scale and with
this degree of unity of theme is unusual. What possible message do the battles in vault IV
articulate? Anna Nílsen suggests that the images of the Nordic heroes serve a didactic
function, arguing that Albertus meant to accentuate the close relationship between the
sacred and the secular worlds by drawing on literary and folkloric motifs common in the
Nordic vernacular tradition. Thus in the same way that the Virgin Mary’s purity is
represented metaphorically by images of a unicorn hunt in some fifteen churches in
medieval Sweden, the three Nordic heroes in vault IV serve as correspondances
(‘motsvarigheter’) to the avenging Christ who defeats the Devil. 31 Old Testament
prefigurations of New Testament events were indeed commonplace in medieval
representation, and it is not surprising to see David and Goliath appear within that context.
But a similar type of ‘motsvarighet’ linking Christ with the other three Nordic heroes is
difficult to articulate, especially in light of the character Didrik av Bern, who in the Nordic
tradition was regarded as something of a dark figure and whose violent demise — being
dragged off to Hell on a black horse at the conclusion of Thiðreks saga — complicates any
direct correspondances between the brave heroes of vault IV and the miles Christi. It is
more likely that the unifying theme of the four battles in the crown of vault IV derives less
from Christological imagery than from the common topos of good triumphing over evil in
single combat, with the paradigmatic battle between David and Goliath serving as the
model followed by the other Nordic heroes and their battles. Moreover, each of the human
heroes is shown defeating a non-human antagonist, for David kills the giant Goliath,32 Sven

tell the entire story, nor is there a succession if images which could convey a sense of ongoing narrative. See
Rushing, Images of Adventure, p. 21.
29
These other murals in Floda church are discussed in detail in Bennett, Floda kyrka, pp. 60-79 (esp. pp. 75-6).
30
Consider, for example, twelfth-century baptismal fonts in Bohuslän and Jämtland that are carved with images of
king Gunnar trapped in the snakepit (a story told in the Old Norse poem Atlakviða and elsewhere), or the
thirteenth-century carvings in Norwegian stave churches that show Sigurd meeting Regin the Smith and
killing the dragon Fafnir. I do not mean to claim an equivalence between those cases and the heroic murals in
Floda church. In the former, the usual ornamentation that was common for baptismal fonts or church portals
was altered in a subtle and unobtrusive way so as to include a small image from the vernacular tradition, while
in the latter, the painted figures in vault IV stand at the crowning moment of a grand theological narrative, and
their inclusion adds a disruptive, heteroglossic element to the divine message.
31
Nílsen, Anna, Program och funktion i senmedeltida kalkmåleri (Stockholm: Kungl. vitterhets historie och
antikvitets akademien, 1986), pp. 444-45.
32
To the medieval mind, giants — even Biblical ones like Goliath — were not seen as equivalent to humans, as
‘men writ large’ so to speak; they were considered a different race of beings. Much has been written on the
origins and anthropology of giants in the Middle Ages; see, for example, Friedman, John Block, The
Monstrous Races in Medieval Art and Thought (Cambridge, MA: Harvard University Press, 1981) and
Williams, David, Deformed Discourse: the functions of the monster in mediaeval thought and literature
(Montreal; Kingston:McGill - Queen’s UP, 1996).

100
 William Layher

Fötling slays a troll and Holger Dansk dispatches the monstrous Burmand. Dietrich von
Bern fights against another human warrior, Vitege Velandsson, but I feel this encounter is a
special case.33 The explicit defeat of monstrosity, in my opinion, is a critical factor for this
grouping in vault IV, for if the artist wanted merely to show the triumph of good over evil
in a general sense, he could easily have found other famous confrontations from Biblical or
other religious sources to illustrate that point. Thus I would justify my special focus on
Burmand thus far by arguing that it is Goliath — not David — who serves as the leading
figures in vault IV, and that Albertus selected three other recognisable and familiar scenes
from Nordic folklore that articulated a similar message, viz. that brave but physically
weaker human heroes can indeed triumph over monstrous villains.
If today’s ballad scholars are at all familiar with the images in Floda church, this is
because the text in the banderole underneath the figure of Holger Dansk has, ever since its
‘discovery’ by Sven Grundtvig almost 150 years ago,34 consistently been interpreted as a
quotation from the ballad multiforms we know as DgF 30 and SMB 216 — and considering
the scarcity of hard evidence supporting the existence of medieval Scandinavian ballads in
the pre-manuscript era, this unique combination of text and image in vault IV of Floda
church is considered an unimpeachable source of the first order. I do not wish to dispute the
validity of the evidence in Floda church, but merely to suggest that the story Albertus knew
— as we reconstruct it from the images he painted —differed in significant ways from the
ballads which survive in the Danish and Swedish corpus. The hermeneutic circle that has
uncritically linked Floda church with DgF 30 and SMB 216 has heretofore never been
broken, let alone challenged; but this is a necessary step if we wish to inspire confidence
amongst our colleagues that our reconstructions of the early days of the Scandinavian
ballads are laid on solid foundations. Consider the following scenario: if the banderole
underneath Holger Dansk had indicated only the name ‘hollager dansk’ and not the phrase
hollager dansk han van seger af burmand, this entire discussion would have taken a
different track. In this hypothetical case, the pictures of Holger Dansk and Burmand in
Floda church, stripped of any overt connection to the extant ballad texts in the Danish and
Swedish tradition, would not be taken as proof of the existance of a medieval Scandinavian
ballad some hundred years before its oldest manuscript attestations, but rather as credible
evidence for the reception of Karlamagnús saga in Sweden in the late fifteenth century. It is
the text passage in the banderole — not the images themselves — that has steered the
reception of the Floda church murals as illustrations of the surviving ballad texts rather than

33
Their battle is different from the others, for no injury is inflicted. While the Floda murals show that Burmand’s
head is sliced off, Trullat’s head is knocked off and Goliath is struck in the head by a stone, Vitege only
retreats from Dietrich — his shield at his back as if he were fleeing — as the furious Dietrich spouts fire from
his mouth. Vitege is not explicitly referred to as a monstrous figure in Thiðreks saga or in its OSw analogue
Didrikskrönikan (transl. ca 1450 in Stockholm), but he is the grandson of a giant and was possessed of an
incredible strength and naive ruthlessness, and his actions show him to be a disreputable figure full of
treachery. There are no extant Nordic ballads about this encounter. I intend to discuss the fire-breathing
diderik van beran in another setting.
34
While Grundtvig was surely not the first antiquarian to take note of the curious murals in vault IV of Floda
church — Peringskiöld’s Kyrkor i Södermanland (published in 1686) had an engraving of them — Grundtvig
was the first to link them thematically to the ballad he classified as DgF 30. See, Danmarks gamle folkeviser,
ed. by Grundtvig, II, 654. and the discussion in Jonsson, Svensk balladtradition, p. 26 n. 58.

101
The Singer and the Scribe

as illustrations of the epic story as a whole. But regardless of how Holger Dansk and
Burmand are viewed in vault IV, the visual evidence in Floda church ultimately points in
the same direction: that Albertus Pictor knew a variant of the Holger Dansk story that is
otherwise unattested in Sweden in the 1480s. The vehicle which brought that story to
Albertus’ ears — perhaps a ballad, an episode from a prose epic, an oral recitation from
manuscript sources —is, however, unknown.
In closing let us recall Christiern Pedersen’s remark from 1534 about the popularity of
the old kempe vise. This likely held true some fifty years earlier as well, even in the
Swedish provinces where Albertus did his painting: that the common man knows little
about Holger Dansk ‘other than that he was a Danish hero who won the victory over
Burmand, as the old kempe vise says: Olger Dansk he won the victory over Burmand’. It is
indisputable that the Floda murals have a connection to the Nordic ballad tradition, for a
ballad with that same refrain was obviously in circulation in Sweden by the end of the
fifteenth century. But based on the above analysis of the iconography and history of the
Burmand figure, we should no longer assume that the surviving ballad multiforms — which
were copied some hundred years after Floda church was painted — are the same multiforms
that were ringing in Albertus Pictor’s ears in 1485. It was our preconceptions that led us to
believe that Albertus Pictor painted a scene from the ballad here — for looking up at the
ceiling of vault IV, that is all that we were prepared to see.

102
 William Layher

Fig.1

(Published by permission of the ATA, Office of National Antiquities, Stockholm)

103
This page intentionally left blank
 The Suppression of a Ballad Culture: the enigma of medieval France

Philip E. Bennett

Peter Haidu begins his book The Question of Violence, which deals with the relationship of
epic poetry (and by further implication epic-heroic activity) to medieval France as an
evolving state, with a consideration of the ways in which certain forms of modern, and
modernist, critical discourse do violence to and seek to suppress the medieval literary
artefact.1 What Haidu has postulated as a fate for medieval epic can equally be applied to
the ballad in medieval French culture. I wish to consider the problem from two angles:
initially looking at the treatment of ballads, as a manifestation of ‘folksong’, in recent books
published in France on the subject of ‘popular’ literature, then considering the ways in
which ‘literate culture’ in medieval France (which I will take to cover the twelfth to
fifteenth centuries), adapts, subverts and suppresses ballad culture. I will close by analysing
briefly one or two poems, which, in any cultural context other than that of France, would, I
contend, be considered ballads.
Part of the enigma of the position of France, which sat more or less at the centre of a
large number of cultural developments in the Middle Ages, consists in the paradox that,
while not apparently developing a ballad culture of its own, it furnished ample material
from epic and romance to neighbouring cultures, which they absorbed to nourish their own
ballad traditions. Two very obvious cases involve Iberia (which developed topics and
incidents from both Charlemagne and Arthurian stories into ballads) and Scandinavia
(where ballads are produced alongside prose saga adaptations of the same material). 2 Colin
Smith, in the very detailed introduction to his edition of a selection of Spanish ballads,
makes a couple of points about this transfer, and about the evolution of ballads in general,
which will be germane to the rest of my argument.3 Firstly (p. 5), he invokes the
conservatism of Spanish society and its culture, which permits the preservation of ‘popular’
medieval cultural and artistic forms long after they had been superseded in the rest of
Europe by ‘learned’ cultural artefacts, drawn from the Greco-Roman classical tradition.
Secondly (pp. 13-16), he points out that there is no manuscript evidence for the existence of
ballads, including the Charlemagne-based or the Arthurian-based ballads, before the
fourteenth century, from which he argues that the preservation of this material in ballad
form is a mark of changing tastes in educated court and bourgeois society, leading to the
emergence of these scattered shards of medieval culture in ‘popular’ traditions.4 Now there

1
Haidu, Peter, The Subject of Violence: the Song of Roland and the Birth of the State (Bloomington and
Indianapolis: Indiana University Press, 1993), pp. 1-16.
2
See, in this volume, contributions by Manuel da Costa Fontes and William Layher.
3
Spanish Ballads, ed. by Colin Smith (Oxford: Pergamon Press, 1964), pp. 5-25.
4
Smith’s point raises as ever the question of defining the ‘Middle Ages’ and the temporal borders by which they
merge into the ‘Renaissance’ or the ‘Early Modern’ period. For the purposes of this paper the Middle Ages in
The Singer and the Scribe

is a certain paradoxical incompatibility between these two positions, but taken together they
do highlight the principal problem in addressing the existence of ballads in medieval
France: tensions between popular and high culture, between the ‘people’ and the
‘aristocracy’. The fact that much the same material passes from France, through Plantagenet
England and possibly Scotland to Scandinavia, where it survives in ballads in Faroese to the
present day,5 emphasises this paradox even more, since, at the period of transmission of the
bulk of the material, southern European distinctions between the aristocracy and a wider
community of ‘warrior-farmers’, traders and sailors were either blurred or inoperative in the
North.6 One cannot therefore say, for Scandinavian examples, that the ballads which
survive alongside versions recorded in the Karlamagnús saga or the riddere sögur represent
a class-based dichotomy, even if a clerical-literate — oral-lay opposition is implied. In
assessments by French scholars of their own native traditions the two concepts (of class and
literacy) have become rather overlaid, and the whole question of the production of a ballad
literature so embroiled in the traditionalist — individualist debate on the origins and
composition of the epic that by a strict application of Cartesian logic they have been driven
to all but deny the existence of ballads in early French traditions.7 Indeed most critics go so
far as to deny the existence of popular tradition in France at almost any period. I shall return
to this point more fully in a moment, for now suffice it to say that the adjective most
regularly used to qualify ‘non-courtly’ medieval texts is popularisant, a word which is
difficult to translate with simple precision, but which may best be seen to imply ‘works
with a popular colouring’.

France will be held to include the fifteenth century, with the Renaissance / Early Modern beginning arbitrarily
in 1500.
5
Skårup, Povl, ‘La matière de France dans les pays du Nord’, in Charlemagne in the North , Proceedings of the
Twelfth International Conference of the Société Rencesvals, Edinburgh 4th -11th August 1991, ed. by Philip E.
Bennett, Anne Elizabeth Cobby and Graham A. Runnalls (Edinburgh: Société Rencesvals British Branch,
1993), pp. 5-20 (esp. pp. 5-6). The role of manuscripts prepared in England in the transmission of this material
to Scandinavia has been explored by André de Mandach, Naissance et développement de la chanson de geste
en Europe, IV Chanson d’Aspremont B-C la guerre contre Agolant, Publications Romanes et Françaises, 156
(Geneva: Droz, 1980), pp. 45 -52. For a more nuanced discussion of the routes of transmission of ‘French’
culture, suggesting variable routes including Germany, the Low Countries and England according to period
and material transmitted to Scandinavia see Lönnroth, Lars, ‘Charlemagne, Hrolf Kraki, Olaf Tryggvason:
parallels in the heroic tradition’, in Les Relations littéraires franco-scandinaves au moyen âge (Actes du
Colloque de Liège, avril 1972), Bibliothèque de la Faculté de Philosophie et Lettres de l’Université de Liège,
208 (Paris: Les Belles Lettres, 1975), pp. 29-52, and Foote, Peter, ‘Aachen, Lund, Hólar’, in Idem , pp. 53 -76.
6
Foote, Peter and David Wilson, The Viking Achievement (London: Sidgwick and Jackson, 1970), pp. 4; 79-80.
7
Spanish Ballads, ed. by Smith, pp. 9 and 12; for a concise but perspicuous account of the (neo-) traditionalist –
individualist debate see La Chanson de Roland, Frederick Whitehead’s Text with Notes and Introduction by
T.D. Hemming (Bristol: Bristol Classical Press, 1993), pp. xviii -xxv. A notable exception is Davenson, Henri
(Henri-Irénée Marrou), Le Livre des chansons, introduction à la chanson populaire française, Les Cahiers du
Rhône (Neuchâtel: Editions de la Baconnière, 1944; repr n.p. [Paris]: Club des Libraires de France, 1958), pp.
62-67, who indicates that themes, poetics, onomastics and musical structures of French folksongs, including
ballads, can all be traced back to the Middle Ages, although he agrees with most other scholars in seeing the
essential period of production of existing songs as being between 1500 and 1800 (pp. 60- 62).

106
 Philip E. Bennett

One other problem confronting those who would study French ballad traditions must be
mentioned before I proceed to the real heart of this paper. This is a problem of vocabulary,
because French has no word which unequivocally covers the concept represented by
English ‘ballad’ or Spanish ‘romance’. Since the native French word ‘ballade’ (which has
been borrowed into English) was specialised in the sense of a courtly lyric with refrain, the
French borrowed the Spanish word, pronouncing it in their own way as [romãs]. Even in the
earliest period of its use (it entered the language in 1719) it did not have the same clear
focus on a narrative poetic text showing Albert B. Friedman’s three essential characteristics
of ballads (focus on an incident, dramatic presentation of event and impersonality)8 as
either the English or Spanish equivalents. Although Marmontel, writing in the Encyclopédie
at the end of the eighteenth century, offers analyses of many medieval lyrico-narrative
poems, particularly chansons de toile, and stresses the importance of the heroic in the
subject matter of the early romance, for him the essence of the French romance is already
that of the slight and sentimental love lyric. 9 This attitude becomes so entrenched by the
late nineteenth century, despite the use of the word Romancero by Paulin Paris and Prosper
Tarbé or of Romances and its Germanised equivalent Romanzen by Karl Bartsch in the
titles of their anthologies of medieval French poetry,10 that any reference to narrative poetry
of what we would term a ‘ballad type’ is missing from discussions of the word. It has
become no mo re than a sentimentally moving song of a popular disposition. This absence or
dissolution of a clearly specialised lemma to cover the concept /ballad/ has re-inforced, and
given comfort to, the individualist tendency to deny the existence of the literary form
itself. 11
Thus in his recent book, Le Moyen Age et ses chansons (with its telling subtitle ‘ou un
passé en trompe-l’œil’), Michel Zink equates the romance with the whole broad range of
‘folksong’, which he is at pains to discredit as a notion, largely on the grounds of the
ambiguous use of Volk (‘people’) by writers of the Romantic-Traditionalist School to mean

8
Friedman, Albert B., ‘Ballad’, in Princeton Encyclopedia of Poetry and Poetics, ed. by Alex Preminger
(Princeton: Princeton University Press, enlarged ed. 1974), pp. 62- 64.
9
His attitude is best summed up in the article ‘Chanson’, republished in the Eléments de littérature:
Nous avons aussi des chansons plaintives sur des sujets attendrissants; celles-ci s’appellent romances; c’est
communément le récit de quelque aventure amoureuse; leur caractère est la naïveté; tout y doit être en
sentiment. [We have melancoly songs on themes which move to tears; they are called romances, which
habitually tell the tale of some amorous adventure; they are characterised by their simplicity and their
sentimentality.] Œuvres complètes de Marmontel, 18 vols (Paris: Verdière, 1818-19), Vol 12 (Eléments de
littérature, 1), p. 444.
10
Le Romancero françois: histoire de quelques anciens trouvères et choix de leurs chansons, ed. by Paulin Paris
(Paris: Techener, 1833); Romancero de Champagne, ed. by Prosper Tarbé (Reims: P. Dubois, 1863-1864;
repr. Geneva: Slatkine, 1980); Romances et pastourelles françaises des XII e et XIIIe siècles [Altfranzösische
Romanzen und Pastourellen der 12. und 13. Jahrhundert], ed. by Karl Bartsch (Leipzig: Vogel, 1870).
11
Davenson, Le Livre des chansons, uses the term complainte to refer to the sub-type of chanson populaire
otherwise known to French folklorists as ‘chansons épico-narratives’. His reason for this choice of label is that
he considers the French approach to ‘ballads’ as reducing the narrative element in favour of an introspective
meditation on the tragic dimensions of the narrative.

107
The Singer and the Scribe

both ethnic group and social class.12 His strictly individualist position leads him to argue
that the fragmentary evocations of ‘folk tradition’ which he finds in a number of literary
texts from the thirteenth to the seventeenth centuries are precisely exercises in false
nostalgia, whose purpose within the works in which he finds them is to define a lost Golden
Age of ‘popular’ culture, before the Fall, characterised by the author’s present within the
sophisticated culture of court or town. Zink is able to make a case for the inevitably
fragmentary nature of these evocations of what he paints as an illusory past on two counts.
In the first place he cites the use within a number of thirteenth-century romances of
quoted extracts from songs, which provide a thematic ‘paratext’ for the romance in
question. Among these are the much discussed fragments included in Guillaume de Dole by
Jean Renart of chansons de toile (narrative songs, usually on a love theme, with a female
protagonist and sometimes narrator-persona, both or either of whom may be portrayed as
weaving or spinning while executing, or participating in the action of the song in
question).13 Now this type of song does seem to offer one sort of camouflage for the ballad
in the French tradition, and is indeed held most regularly to be the original of the French
romance.14 It is under that designation that Léon Clédat includes in his Chrestomathie du
Moyen Age the song Bele Doette (the story of a young woman who takes the veil when the
knight variously described as her lover — ami — and her husband — sire — is killed in a
tournament). 15 Pierre Bec actually goes even further in his study of the early medieval lyric
in France and restricts the genres popularisants to women’s songs of this sort, which he
equates with the Spanish cantigas d’amigo.16

12
Zink, Michel, Le Moyen Age et ses chansons ou un passé en trompe-l’œil (Paris: Editions de Fallois, 1996), pp.
25-30, where both Herder in Germany and Gaston Paris in France are criticised for ‘confusing’ the genius of a
people with the spontaneous outpourings of the people, and pp. 113-17, where medievalists and folklorists are
accused of shifting between the ethnological and sociological usages of the word depending on whether they
are producing a diachronic or a synchronic study.
13
Zink, Le Moyen Age et ses chansons, pp. 148-60; the notion that the folklorist, like the anthropologist or
ethnologist, always arrives ‘just too late’ so that all that is available is the fragmentary wreckage of a myth, a
song or a ritual rescued fortuitously from oblivion, is dealt with on pp. 37- 44. He also embraces Davenson in
what he sees as a generally Romanticist view of the creative powers of the folk, pointing out that Davenson
too refers to the fragmentary and disjointed nature of surviving texts (p. 180). In fairness it should be pointed
out that at the place cited by Zink (Le Livre des chansons, p. 16) Davenson is evoking the psychology of the
reception of folksong in mid twentieth-century France. He makes his own critique of the Romantic-
Traditionalist school, in terms not dissimilar to those of Zink, on pp. 22-32 and gives a much more subtle
account of the interactions of popular and ‘élite’ poets in the production of folksong in France from the
Middle Ages to the early nineteenth century on pp. 42-47.
14
The chanson de toile or d’histoire has also been considered as the ‘source’ of the English and Scots ballad. See
Howard S. Jordan, ‘The Old French Chanson d’histoire as a Possible Origin of the English Popular Ballad’,
Revue de Littérature Comparée, 16 (1936), 367-78. See also below, n. 22.
15
Clédat, L., Chrestomathie du Moyen Age, ou morceaux choisis des auteurs français du moyen âge (Paris:
Garnier, 12th ed. 1932), pp. 328-30.
16
Bec, Pierre, Nouvelle anthologie de la lyrique occitane du moyen âge (n. p. [Avignon]: Aubanel, 1970), p. 70
with examples pp. 149-67; idem, La Lyrique française au moyen âge (XIIe – XIIIe siècles), contribution à une
typologie des genres poétiques médiévaux, 2 vols (Paris: Picard, 1977), I, 29-30 and 33-35. In the second work

108
 Philip E. Bennett

It is thus by a quirk of literary self-referentiality within a specific mode of thirteenth-

century narrative that Zink is able to make his case for popular culture (including ballad
culture) belonging to a time and place that are no more.17 Fragments of several chansons de
toile are in fact inserted many times in Guillaume de Dole, including the occasion on which
the mother of the heroine, in her rural castle far from the corruptions of the Imperial court,
introduces three quoted fragments with a reference to how women used to sing such songs
in days gone by.18 By a sleight of hand what starts as an aesthetic effect within a work of
fiction becomes for Zink, in this work, a statement of sociological reality in the extra-
textual world, although he avoids such unwarranted extrapolations in an earlier study.19
Moreover, in keeping with the violence I see as being visited on this type of text, Zink
portrays the singing of Bele Aiglentine en roial chamberine20 by a group of young knights
heading for a tournament as a salacious parody of the chanson de toile and the ‘lost’
conservative way of life it is held to represent. A contrast is established between the use of
fragments of song to evoke the ‘Never-Never Land’ of the women and their ‘folksongs’,
which operate reciprocally by mise en abyme to generate a world existing only in nostalgic
memory, and the singing in extenso of Bele Aiglentine to produce a comic deconstruction of
that world. There is, however, nothing in the narrative line of Bele Aiglentine (the story of a
young noblewoman whose pregnancy is discovered by her mother and who then gets to
marry the handsome knight with whom she is in love) to suggest more than a degree of
fantasising and wish-fulfilment on the part of the ‘bacheler de Normendie’ who sings it.21
Moreover, the contrast Zink establishes between the fragmentary singing of songs by
Lïenor and her mother and the singing of a whole song by the ‘bacheler de Normendie’ is,

Bec introduces the extra idea of ‘genres folklorisants’, which are a sub-set of the ‘genres popularisants’; the
‘sujet lyrique’ — we might say persona — of these poems is still a woman according to Bec. In no case does
this imply that the author (where one is known) will be female, indeed most of the examples given in the
Nouvelle anthologie are by identified male poets, who also compose courtly lyrics.
17
Zink, Le Moyen Age et ses chansons, pp. 158-60.
18
Jean Renart, Le Roman de la Rose ou de Guillaume de Dole, ed. by Rita Lejeune (Paris: Droz, 1936) ll. 1145-92.
All quotations are from this edition.
19
Zink, Michel, Belle, essai sur les chansons de toile suivi d’une édition et d’une traduction (Paris: Champion,
1978), pp. 3 -11. In this work Zink is clear that the effect of mise en abyme established by Jean Renart operates
in such a way that the only extra- textual world for the chansons de toile being performed is that of the
romance of Guillaume de Dole itself.
20
Guillaume de Dole, ll. 2235 -94.
21
Whether or not the fact that the bacheler is from Normandy in this romance written by an Easterner for a patron
from Eastern France is of any significance, the fact that he is a landless young knight presumably in quest of
an heiress undoubtedly is. The model of this type of knight in the late twelfth century is Guillaume le
Maréchal, who, after a career of some twenty years serving Henry II and his sons Henry Court Mantel and
Richard the Lionheart, was rewarded aged forty with the hand of the sixteen year-old heiress to t he county of
Pembroke. On his career see Duby, Georges, Guillaume le Maréchal ou le meilleur chevalier du monde (Paris:
Fayard, 1984). For an interpretation of the singing of Bele Aiglentine similar to mine, and which also notes the
similarity of structuree between this French chanson de toile and English ballads see Page, Christopher,
Voices and Instruments in the Middle Ages: instrumental practice and songs in France 1100-1300 (Berkeley:
University of California Press, 1986), pp. 35-37, where Page draws attention to the folkloric and proverbial
discourse of the Aiglentine text and its incremental narrative structure.

109
The Singer and the Scribe

unfortunately, misleading, since at the end of the romance another male character, the
nephew of the bishop of Liège, who is clearly part of the same ‘modern’ world as the young
knights attending the tournament at Saint-Trond, also sings a fragment of a chanson de
toile, ‘Or vienent Pasques, les beles, en avril’, whose heroine is la bele Aigline.22 In this
perspective we must also consider the interesting problem of the singing in the romance of
multiple versions of Main se leva bele Aaliz, including three which are sung on a picnic
where young men get to fondle the thighs of young women who offer them their chemises
as towels to dry their hands after washing before the meal. 23 Now this might be held simply
to re-inforce the contrast between the immoral modernity of the court and the respectable
conservatism of the provincial castle if the judgement of most recent critics is accepted, that
the Bele Aaliz texts presented here are not fragments of something longer but complete
rondets de carole (dance-songs with close generic links to what will, by the end of the
thirteenth century, become one of the main types of ‘fixed form’ courtly lyric: the
rondeau).24 However, Maurice Delbouille considered the rondet de carole at least as old-
fashioned as the chanson de toile in Jean Renart’s day,25 which nullifies the argument from
nostalgia. Nor has anyone actually responded to, let alone refuted Pierre Le Gentil’s
observation that to reduce the texts presented in Guillaume de Dole to regular rondets
requires considerable surgery.26 To take just the first example from Jean Renart’s romance:
Main se leva bele Aeliz
— Mignotement la voi venir —

22
Guillaume de Dole, ll. 5188-207. There is some dispute among specialists about the generic affinities of this
fragment, although the latest editor of the poem, Félix Lecoy, does identify it as a chanson de toile, Jean
Renart, Le Roman de la Rose ou de Guillaume de Dole, ed. by Félix Lecoy, Classiques Français du Moyen
Age, 91 (Paris: Champion, 1962), p. xxv. Lecoy, loc. cit., also explicitly identifies the chanson de toile and the
romance.
23
Guillaume de Dole, ll. 259-556.
24
Bec, La Lyrique française au moyen âge, I, 223-24; II, 150-55.; Guillaume de Dole, ed. by Lecoy, pp. xxvi-xxix.
The identification of the Bele Aaliz poems as dance-songs may not be totally irrelevant, however: Jordan, ‘The
Old French Chansons d’histoire’, p. 372, follows W.P. Ker in deriving ballads from dance-songs. This
connection is certainly alive in Faroese ballads, which, even when derived from Old French epic material, are
performed in the manner of the Old French caroles with a lead singer singing the stanza, the company singing
the refrain and dancing in line or in a round. In this manner, on the basis of a singing by Bárður Jákupsson
specially recorded for the event, the Faroese ballad ‘Runsivalstriðið’ was performed at the twelfth
international congress of the Société Rencesvals held in Edinburgh in 1991.
25
Delbouille, Maurice, ‘Sur les traces de “Bele Aëlis”’, in Mélanges de philologie romane dédiés à la mémoire de
Jean Boutière (1899-1967) ed. by Irénée Cluzel and François Pirot, 2 vols (Liège: Soledi, 1971), I, 199-218
(esp. p. 216).
26
Le Gentil, Pierre, ‘A propos de Guillaume de Dole’, in Mélanges de linguistique romane et de philologie
médiévale offerts à M. Maurice Delbouille, tome 2, Philologie médiévale, ed. by Jean Renson (Gembloux: J.
Duculot, 1964), pp. 381-97 (esp. pp. 390-92). See also van der Werf, Hendrik, ‘Jean Renart and Medieval
Song’, in Jean Renart and the Art of Romance: essays on Guillaume de Dole, ed. by Nancy Vine Durling
(Gainesville: University Press of Florida, 1997), pp. 168-69, where he too comments on the mis-match
between what we find in Jean Renart’s manuscript and the versions of the poems published by Friedrich
Gennrich, although with no reference to the earlier work by Le Gentil. None the less, van der Werf still
discusses all these songs from Guillaume de Dole as ‘rondeaux’.

110
 Philip E. Bennett

Biau se para, miex se vesti

Desoz le raim.
— Mignotement la voi venir,
Cele que j’aim! —

[Fair Aeliz rose early — I see her coming daintily towards me — she made herself beautiful, dressed herself
27
more finely under the branch. — I see her coming daintily towards me, the one I love!]

We can see that the poem, if complete, does not offer the habitual rhyme scheme of the
rondet or rondeau of ABabAaAB (where capital letters indicate the refrain) but a mono-
assonanced sequence of three lines with a self-rhyme in –ir provided by the repeated refrain
and a rhyme in –ain provided by a stanza line and a refrain line. While, on the one hand,
Maurice Delbouille noted this feature of the Bele Aaliz poems but offered no comment on
it,28 Pierre Le Gentil for his part observed that the underlying structure of these stanzas is
that of the strophe de romance (the French equivalent of a ballad stanza) and concluded that
the fragments quoted in Guillaume de Dole represent the disiecta membra of one or more
romances with Aaliz as heroine.29 Equally interesting is the fact that the variants offered of
Bele Aaliz suggest a lively and creative oral tradition capable of generating what amounts to
a singing competition among the picnickers: the reason each of these singings is
fragmentary is that, as the narrator indicates, people keep interrupting each other with their
own versions of the song.30
This feeling that Jean Renart is not consigning such traditions to an improbable and
unfashionable past arises also from the evident relationship between one of the songs sung
by the hero’s sister,
La bele Doe siet au vent;
Souz l’aubespin Doon atent.
Plaint et regrete tant forment
Por son ami qui si vient lent:
— «Diex! quel vassal a en Doon!,
Diex! quel vassal!, Diex! quel baron!

27
Guillaume de Dole, ll. 310-15; all translations appended to quotations in this article are mine unless otherwise
indicated. The lines in italics indicate the supposed refrain of the rondet, although they might be just a quoted
refrain from another song, unless it is the Bele Aaliz text that is quoted and the refrain original.
28
Delbouille, ‘Sur les traces de “Bele Aëlis”‘, pp. 211-12.
29
Le Gentil, ‘A propos de Guillaume de Dole’, pp. 393-95. A similar mono-assonanced stanza, though this time
with no indication of a refrain, is offered by the fragment ‘De Renaut de Mousson’ (Guillaume de Dole, ll.
2397- 403); Lecoy, Guillaume de Dole, p. xxvi, considers this to be a ‘chanson historique, une chanson d’éloge
en tout cas’, opining that this is an otherwise unknown genre in French poetry.
30
This also happens with the bishop of Liège’s nephew, whose song is cut short by another character coming in
with a version in French translation of two stanzas of ‘Can vei la lauzeta mover’ by Bernart de Ventadorn
(Guillaume de Dole, ll. 5208-27). For another interpretation of the scene in which multiple versions of ‘Bele
Aaliz’ are sung see van der Werf, ‘Jean Renart and Medieval Song’, p. 172 where the suggestion is made that
what we are witnessing is a form of ‘parlor game’ in which participants have to improvise a complete short
song on a refrain cité. This could bring us back to the idea of the pre-existing complete song, whose narrative
scheme would provide a basis for the improvisation of the participants in the game.

111
The Singer and the Scribe

Ja n’amerai se Doon non.»

[Beautiful Doe sits in the wind; under the hawthorn she waits for Doon. She complains and laments intensely
on account of her lover who is so slow to come to her: ‘God, what a fine fighter is Doon! God, what a fighter!
31
God, what a brave man! I’ll never love anyone but Doon.’]

and that published by Clédat:

Bele Doette as fenestres se siet,
Lit en un livre, mais au cuer ne l’en tient:
De son ami Doon li ressovient,
Qu’en autres terres est alez tornoier.
E or en ai dol.

[Beautiful Doette sits in a high chamber, she reads a book but can’t concentrate on it: sh e remembers her lover
32
Doon because he has gone to fight in tournaments in other lands. Oh, what grief is mine.]

In some ways the Guillaume de Dole version could indeed be held more archaising (the
evocation of Doon owes more to epic than to romance traditions; the allusion to the
hawthorn appears as a folklore reference) while Clédat’s text seems more modern (the
heroine is not doing needlework but reading a book seated in the airy solar of her castle).
However, the dramatic intensity of Lïenor’s singing has nothing to do with a vapid
evocation of the ‘good old days’: it is a reflection of the single-minded purposefulness of a
heroine who will take her reputation in her own hands when she goes to court to clear her
name of a scurrilous slander and win the man she loves. That Jean Renart has Lïenor sing
only a fragment is important, because the intertextual references set up by the suppressed
ending (which we must assume to follow the established theme as recorded in the Clédat
edition and have the bereft heroine retreat to a convent) call precisely on the audience’s
knowledge of a tradition to measure the conventional passivity of the heroine of the
romance against the very unconventional activity of the heroine of the romance.
The second type of poem which Zink is able to cite as a model of inherited folk tradition
as a shattered mirror is the kind of lyric poem, current again in the thirteenth century,
known in French as a chanson avec des refrains. This sort of poem may be of a purely
courtly nature (as in ‘Chançon ferai, que talenz m’en est pris’ by Thibaut de Champagne), 33
offer a mixed register, with a courtly theme wedded to a narrative structure borrowed from
the pastourelle (as in ‘Avant ier me chevauchoie’ possibly by Thibaut de Blaison)34 or
purely in the pastourelle or bergerie tradition (as in ‘El mois de mai par un matin’ by Jehan
Erart). 35 The defining characteristic of the type is the insertion of a different refrain at the
end of each stanza of the poem. The generally held view is that these refrains are borrowed

31
Guillaume de Dole, ll. 1203-16.
32
Clédat, Chrestomathie, p. 328.
33
Les Chansons de Thibaut de Champagne, roi de Navarre, ed. by A. Wallensköld, Société des Anciens Textes
Français, 69 (Paris: anc. Edouard Champion, 1925), pp. 76-82.
34
Les Poésies de Thibaut de Blaison, ed. by Terence H. Newcombe, Textes Littéraires Français, 253 (Geneva:
Droz, 1978), pp. 122-24.
35
Les Poésies du trouvère Jehan Erart, ed. by Terence Newcombe, Textes Littéraires Français, 192 (Geneva:
Droz; Paris: Minard, 1972), pp. 72-78.

112
 Philip E. Bennett

from pre-existing songs in a popular mode, although it is not inconceivable that on

occasions the poets invented refrains to adorn their poems.
This example from Jehan Erart gives an indication of how ‘ballad-like’ the type can be,
fulfilling all of Friedman’s criteria except that of the anonymity of its author. It also differs
from most ballads by the complexity of its verse form:
El mois de mai par un matin
S’est Marïon levee;
En un boschet lez un jardin
S’en est la bele entree. 4
Dui vallet, Guiot et Robin,
Qui lonc tens l’ont amee,
Pour li voër delez le bois
Alerent a celee. 8

Et Marïon qui s’esjoï

A Robin perceü,
Si dist ceste chançonete:
«Nus ne doit lez le bois aler 12
Sanz sa conpaignete.»

Robin et Guiot ont oï

Le son de la brunete.
Cil qui plus a le cuer joli 16
Fet melz la paëlete.
Guiot mult tres grant joie ot
Quant ot la chançonete;
Pour Marïon sailli en piez, 20
S’atempre sa musete.

Robin mult tres bien oï l’ot,

Au plus tost que il onques pot
A dit en sa frestele: 24
«Deus! quel amer,
Harou! quel jouer
Fet a la pastorele!»

Guiot a mult bien entendu 28

Ce que Robins frestele,
Si tres grant duel en a eü
A pou qu’il ne chancele.
Mes li cuers li est revenu 32
Pour l’amor de la bele;
Sa musele il a reposté
S’escorce sa cotele.

Un petitet ala avant 36

Delez Marïon maintenant,
Si li a dit tout en esmai:

113
The Singer and the Scribe

«Hé, Marionnete!
Tant amee t’ai!» 40

Marïon vit Guiot venir —

S’est autre part tornee,
Et quant Guiot la vit guenchir,
Si li dist sa pensee: 44
«Marïon mains fez a prisier
Que fame qui soit nee,
Quant pour Robinet ce bergier
Es si asseüree.» 48

Quant Marïon s’oï blasmer,

Li cuers li conmence a trenbler,
Si li a dit sanz nul deport:
«Sire vallet, vos avez tort 52
Qui esveilliez le chien qui dort.»
Quant Guiot vit que Marïon
Fesoit si male chiere,
Avant sacha son chaperon, 56
Si s’est tornez arriere.
Robin qui s’estoit enbuschiez
Souz une chasteigniere,
Pour Marïon sailli en piez 60
Si a fet chapiau d’ierre.

Marïon contre lui ala,

Et Robin deus foiz la besa,
Puis li a dit: «Suer Marïon, 64
Vous avez mon cuer
Et j’ai vostre amor en ma prison.»

[One morning in May Marion arose; the beauty went into a grove beside a garden. Two lads, Guiot and
Robin, who have loved her for a long time went secretly to the edge of the wood to see her. And with joy
Marion spotted Robin and sang this little song: ‘No one should go to the edge of the wood without his girl
friend.’
Robin and Guiot heard the tune the brown-haired girl was singing. The one with the merriest heart had the
best time. Guiot was overjoyed when he heard the song; on account of Marion he jumped to his feet and tuned
his bagpipe. Robin heard it well enough and as quickly as he could played on his pipe ‘God! what love,
Halloo! what sport the shepherdess has set up!’

Guiot clearly understood what Robin was piping, he was so struck with grief he nearly fell down, but heart
came back to him for love of the beauty; he put his bagpipe aside and hitched up his tunic. He advanced at
once a bit towards Marion and said to her with great emotion: ‘Hey, Marionnete, how much have I loved you!’

Marion saw Guiot coming and turned to one side, and when Guiot saw her swerve he said what was on his
mind: ‘Marion, you are the least worthy of any woman alive, when on account of Robinet, that shepherd, you
show such resolve.’ When Marion heard herself criticised her heart began to beat fast and she said without
any jesting: ‘Sir boy, you are wrong to wake a sleeping dog.’

114
 Philip E. Bennett

When Guiot saw how Marion was scowling he pulled his hood over his face and turned tail. Robin, who was
lying in wait under a clump of chestnut trees, leapt to his feet on account of Marion and made a chaplet of ivy.
Marion went up to him and Robin kissed her twice, then said: ‘Dear Marion, You have m y heart, and I have
your love as my prisoner.’]

Now Zink is right that this poem from the literate tradition is giving us glimpses only of
what purports to be popular, oral tradition, although nothing here indicates that a lost
culture is being evoked. What is perhaps most perplexing with the whole set of chansons
avec des refrains is that none of the refrains used, although they are re-used by more than
one poet, can actually be attached to a song outside the set. They are, in fact, refrains that
are known to be refrains, and nothing more.36 It is impossible to trace a verifiable tradition
of oral poetry, whether in ballad form or not, behind them. Even worse, the last pair of lines
used by Erart, and identified as a quoted refrain by the editor, do not have the air of being a
refrain at all: they could be the incipit, or any other pair of lines, taken not from a ‘folk’
song but from a courtly song.
However, if we look more closely at what Erart is doing in this poem, we see that the
whole tenor of it is one of parody of neo-epic activity, and neo-epic composition. The use
of formulaic language, including textual repetitions within the poem, the insistent use of the
simple temporal link quant [‘when’] to articulate an otherwise disjointed, paratactic
narrative line, the use of dialogue and quoted song to advance the plot all move us back into
the universe of the ballad in its heroic neo-epic manifestation, but in a burlesque register.
While this poem is not a ballad in the habitual sense of the term, we are fa r from the
evocation of the fragmented and lost tradition referred to by Zink. It actually presupposes a
current and active tradition against which the audience can measure Erart’s performance,
without which much of the humour would be lost.
The other bone of contention among French scholars, which has considerable
ramifications for the perception of the place of the ballad in medieval culture, concerns the
existence — or as most individualist critics would put it — the non-existence of those short
songs — cantilenæ — which traditionalist and neo-traditionalist critics see at the root of
epic poetry. The impression given in statements of theoretical position by individualists is
that Gaston Paris, who was the main advocate at the end of the nineteenth century of the
role of cantilenæ in generating the versions of epics surviving in manuscripts, was simply
inventing concepts ex nihilo.37 The argument broadly runs that, since no examples of such
songs exist in manuscript form today, they can never have existed. The witness of writers
such as William of Malmesbury, who refers to the cantilena Rollandi sung at the Battle of
Hastings, and Ordericus Vitalis who refers to cantilenæ sung vulgo (‘by the people’) on the
subject of Guillaume d’Orange, is either dismissed, or interpreted in a tendentious way,
which takes cantilena in its strictly Classical sense to mean ‘trivial song’ rather than ‘short

36
On the question of whether early refrains used in the Old French chanson avec des refrains genuinely derive
from otherwise lost traditional songs and on the propensity of poets to invent refrains to fit their own poems
see Zumthor, Paul, Toward a Medieval Poetics, trans by Philip E. Bennett (Minneapolis; Oxford: University
of Minnesota Press, 1992), pp. 195- 96.
37
A typical example is the dismissive statement by Timothy Hemming that ‘Gaston Paris […] lent a certain
intellectual reinforcement to this charming fantasy [that epic poems found their ultimate origins in laments and
eulogies sung immediately after the events celebrated]’, La Chanson de Roland, ed. cit., pp. xviii-xix.

115
The Singer and the Scribe

song’ and vulgo to mean ‘in the vernacular’, the two words being treated as forming a
pleonastic doublet implying that the songs referred to were of the same approximate length
and shape as those surviving in literary form, but were denigrated by clerical writers as not
belonging to Latin culture. Two points need to be made here: firstly, that there is no doubt
that for Gaston Paris and the (neo-) traditionalist school the songs referred to by medieval
writers as cantilenæ are short poems; secondly (and as a form of corroboration of that)
longer poems related to epic material are designated not cantilena but carmen in twelfth-
century sources, as we see from the Carmen de prodicione Guenonis, a Latin treatment of
the core material of La Chanson de Roland,38 and the Carmen de Hastingæ Prœlio, an
account of the Norman conquest of England drawing on both Classical and contemporary
vernacular epic modes of composition.39 Also, in terms of medieval witness, we should not
assume that when Wace portrays the jongleur-knight Taillefer singing about ‘Charlemagne,
Roland, Oliver and the vassals who died at Roncevaux’ before Duke William at Hastings40
he is singing anything approaching the Chanson de Roland as we have it, even in the
Oxford manuscript. It has, as a point of comparison, been calculated for Greek epic texts
and romances that at a standard rate of recitation it takes about three hours to ‘chant’ 1500
lines of text.41 This is clearly incompatible with the scenario implied by Wace, and stated
explicitly by William of Malmesbury that
Tunc cantilena Rollandi inchoata, ut martium uiri exemplum pugnaturos accenderet, inclamatoque Dei auxilio,
prelium consertum, bellatumque acriter, neutris in multam diei horam cedentibus.

[Then they struck up the song of Roland (cantilena Rollandi), to fire them as they went into battle with the
example of a heroic warrior, and calling on God’s help (shouting the warcry of the Normans), came to grips
42
and fought furiously, neither side giving way till a late hour.]

The brevity (or incomplete nature) of the song implied in cantilena, coupled with the
shouting of the Norman war cry (“Dieus aïe!”) at the mo ment when battle is to be joined led
Du Cange to identify the cantilena itself as a ‘clamor militaris’, 43 and definitely orients us
towards the sort of ‘epic fragment’ which, if self-contained, becomes virtually
indistinguishable from a ballad. However, that such ‘proto-epics’ should be considered also
to be ‘proto-ballads’ is itself a matter of controversy. Gordon Hall Gerould and Albert B.
Friedman (who also quotes the opinions of both Gerould and W.P. Ker) refuse to make the

38
Carmen de prodicione Guenonis in Les textes de la Chanson de Roland ed. by Raoul Mortier, 10 vols (Paris: La
Geste Francor, vol. 3, 1941), pp. 105-17.
39
The Carmen de Hastingae proelio of Guy, Bishop of Amiens ed. by Catherine Morton and Hope Muntz (Oxford:
Clarendon Press, 1972).
40
Le Roman de Rou de Wace, ed. by Anthony J. Holden, Société des Anciens Textes Français, 93, 3 vols
(Paris:Picard, 1970-73), ll. 8013-18.
41
Personal communication by Prof. Elizabeth Jeffreys based on her work on Byzantine practice.
42
William of Malmesbury, Gesta regum Anglorum: The History of the English kings, ed. and trans. by R.A.B.
Mynors; completed by R.M. Thomson and M. Winterbottom, Oxford Medieval Texts, 2 Vols (Oxford:
Clarendon Press, 1998-99), p. 455, with my glosses in parentheses.
43
Carolus Du Fresne, Dominus Du Cange, Glossarium ad Scriptores Mediæ et Infimæ Latinitatis, 3 Vols (Paris: L.
Billaine, 1678), I, col 765.

116
 Philip E. Bennett

connection,44 but they were both concerned with the English tradition, which implies a
certain poetics of the ballad, not necessarily found in other parts of Europe. Certainly two of
the earliest Old French epics, Gormont et Isembart and La Chanson de Guillaume have
undeniable refrains and the oldest of all, La Chanson de Roland in the Oxford manuscript
has an enigmatic notation ‘AOI’ at several points which can be interpreted as a refrain.45
We do in fact have two witnesses to such products in France in the thirteenth century.
One is the parodic, and frankly scatological Audigier,46 which recounts in the form of a
short epic a burlesque battle between grotesque peasants, not unlike the battle in Torelore,
in Aucassin et Nicolette,47 where the opposing forces fight with soft cheese and rotten
apples. The other is the presumably satirical and topical Bataille d’Annezin. This poem,
which consists of fifty lines on a single assonance relates the story of a battle between
Christians and Saracens ‘in the marshes at Annezin’ (near Lille).48 The battle is aborted,
however, when the two sides are reconciled by the appearance of a cup of wine. The extent
to which either of these poems is in the ‘popular’ or ‘aristocratic’ tradition is open to
debate, a debate which may in any case be irrelevant, since from one point of view all
vernacular traditions may be regarded as ‘popular’ in contrast to the ‘learned’ traditions of
clerical Latin. Audigier is linked to the popular tradition because a line of it is sung by a
coarse peasant (who is immediately silenced by his more refined fellow-shepherds and
shepherdesses) in Adam de la Halle’s play Le Jeu de Robin et Marion, a dramatised
pastourelle-cum-bergerie, presented at the court of Louis d’Anjou in Naples, probably at

44
Gerould, Gordon Hall, The Ballad of Tradition (Oxford: Clarendon Press, 1932), pp. 102-03; 194-96; Friedman,
Albert B., The Ballad Revival, studies in the influence of popular on sophisticated poetry (Chicago: London:
University of Chicago Press, 1961), pp. 19-21.
45
Gormont et Isembart, fragment de chanson de geste du XII e siècle, ed. by Alphonse Bayot, Classiques Français
du Moyen Age, 14 (Paris: Champion, 3rd ed. 1931), which has a variable intercalated quatrain as a refrain at
certain moments of the narrative; e.g. ‘Quant il ot mort le bon vassal / ariere enchalce le cheval; / puis mist
avant sun estendart: l’em la li baille un tuenart’. [When he had killed the good vassal / he drove his horse back
behind his lines; then he raised his standard: he was handed a large shield.], ll. 5-9, 37-40 (with minor
variation), 61-64, 83-86, 134-37 which punctuates a series victories by the pagan leader Gormont over French
warriors; La Chanson de Guillaume, ed. and trans. by Philip E. Bennett, Critical Guides to French Texts,
121.ii (London: Grant & Cutler, 2000), which has a series of refrains referring to days of the week throughout
the text; e.g. ‘…Crie “Munjoie!” – l’enseigne Ferebrace. / Lunsdi al vespre: / Cil le choisirent en la dolente
presse’. [He shouts “Munjoie!” – the warcry of Ferebrace. / Monday at vespertide: / They caught sight of him
in the grievous throng.], ll. 447- 49; La Chanson de Roland, ed. cit. The notation ‘AOI’ ocurs for the first time
at the end of l. 9 and then frequently though irregularly throughout the rest of the poem. Since nothing like this
occurs in any other Old French epic in a corpus of some 120 poems, it is reasonable to assume that this feature
is a survival of an early stage of epic production, implying an approach which favoured the lyric, perhaps with
ritual dance elements, over the narrative which dominates later epic production in France.
46
Conlon, D.C., ‘La Chanson d’Audigier, a scatalogical parody of the chanson de geste edited from ms BN f. fr.
19152’, Nottingham Medieval Studies, 33 (1989), 21-55.
47
Aucassin et Nicolette, chantefable du XIIIe siècle, ed. by Mario Roques, Classiques Français du Moyen Age, 41
(Paris: Champion, 2nd ed. 1962); the Torelore episode occupies sections XXVIII-XXXIII (pp. 29-33).
48
Badel, Pierre-Yves, ‘La Bataille d’Annezin: une parodie de chanson de geste’, in ‘Plaist vos oïr bone cançon
vallant?’ mélanges offerts à François Suard , ed. by Dominique Boutet et al., Collection Travaux et
Recherches (Villeneuve d’Ascq: Université Charles - de - Gaulle – Lille 3, 1999), pp. 35 - 44.

117
The Singer and the Scribe
49
Christmas 1283. The joke is multi-layered, and the habitual critical dichotomy of popular
and courtly traditions negated, because of the rejection of coarseness by polite shepherds
and the self-evident fact that Adam’s princely and aristocratic court audience must have
recognised and reacted to the quoted fragment. The Bataille d’Annezin is harder to place,
and any deductions must be made from its appearance in a sole manuscript. This too
operates at a variety of levels, since character-types from the chansons de geste are placed
in ‘homely’ local surroundings, and reduced in status to being bourgeois militia. So far we
have found in both these texts elements already identified in the analysis of the pastourelle
by Erart: a degree of humour that can only have worked by reference to a serious litera ry
model, and a systematic debasement of content which is both necessary to an appreciation
of another level of humour and designates the model as unworthy of consideration by a
cultivated audience. The last level of meaning in La Bataille d’Annezin, though, is a more
serious one than we have found in any poem so far, since it involves an allegorical reading,
making of the whole a political satire on the military mismanagement of King John. It is
this which leads me to the last category of poem I would like to consider as candidates for
the lost, or suppressed, ballad culture of the medieval French-speaking world: certain types
of Anglo-Norman political song.
The varied collection of poems published by Isabel Aspin in the Anglo-Norman Text
Society series in 1953 has never attracted the attention of scholars which it deserves.50
While it would be wrong to claim that all the poems in the short anthology are ‘ballads’ in
the sense we would want to give to the term (it includes a set of macaronic proverbs, satires
on the state of the church and the realm, attacks on excessive taxation — again in Latin and
French — reflections on morality and the Three Estates) some of the poems Aspin
published fit clearly within a tradition that we see to be that of the ballad across Europe at
the end of the Middle Ages. I would like to single out just two for brief consideration.
The first is a short narrative (96 lines) in rhyming octosyllabic couplets presenting the
colourful career of a self-styled double agent (Thomas Turberville) who, in 1295, offered
first to guarantee the conquest of England to Philip IV of France, then to arrange the
reconquest of Edward I’s lost territories in France. Its overt form is that of the romance or
lai.51 However, the opening,
Seignurs e dames escutez,
De un fort tretur orrez
Ke aveit purveu une tresun;
Thomas Turbelvile ot a non. (ll. 1-4)

49
Adam de la Halle, Le Jeu de Robin et Marion, ed. by Kenneth Varty (London: Harrap, 1960); Audigier is quoted
at l. 729.
50
Anglo-Norman Political Songs, ed. by I.S.T. Aspin, Anglo -Norman Texts, 11 (Oxford: Anglo -Norman Text
Society; Blackwell, 1953).
51
While the use of octosyllabic rhyming couplets associates the poem with narrative literature in French rather
than with sung forms it should be noted that the Castilian romance uses a very similar form of octosyllabic
lines assonating abcbdb etc., see Spanish Ballads, ed. by Smith, pp. 25-30. Aspin’s identification of this poem
as a ‘song’ rather than as a ‘tract’ is thus not necessarily incongruous.

118
 Philip E. Bennett
[Lords and ladies, listen, you shall hear about a bold traitor who had planned to commit treason; his name was
Thomas Turberville],
is borrowed from epic-oral style. So, most probably, is the persistent calling of the King of
France ‘Charles’, since the first mention of the name ‘Charlys’ (unadorned by a title)
evokes a formulaic reference to Saint Denis, which again suggests an oral-epic origin for
the substituted name. At the end of the poem the evocation of Eustace le Moine (an early
thirteenth-century pirate who commanded the French invasion fleet of 1216 and became the
subject of a roman d’aventure) alongside ‘Duc Lowys’ (identified as Louis VIII, although
he never bore such a title, being known simply as ‘le Prince Louis’) maintains the blend of
popular-historical and literary reminiscence. Objective history is in fact the victim of this
tendentious retelling of heroic English resistance to the invasion of forces despatched by a
bloodthirsty and unreasoning tyrant:
Ore purra Charles pur ver
Aprés li longement garder,
Einz k’il venge pur la treison
Demander de li garison.
Sire Edeward pur la grant navye
De France ne dona une aylle;
De vaillante gent fist la mer
De tut part mut ben garder.
De Engleter sunt failliz
Ly Franceys e sunt honiz […]
Ore sunt tuz, jeo quide, neëz
Ou en lur teris retornez
E penduz pur lur servise,
Ke Engleter n’aveyent prise;
E ceo Charles lour premist
Si nul de eus revenist. (ll. 63-84)

[Now Charles can truly watch for him a long time before he comes to ask the reward of his treason. Lord
Edward didn’t give a clove of garlic for the great French fleet; he had the sea watched on all sides by valiant
men. The French have failed to take England and are put to shame […] They are, I think, now all drowned or
returned home and hanged for their service, since they hadn’t taken England; that’s what Charles promised
52
them if any returned home.]

This is the type of the cantilena that Gaston Paris proposed as the source of the epic
tradition. Whether they were sung, chanted in the manner of liturgical responses or simply
recited53 is not as important as the evident fact that this text, although conserved in a

52
Aspin, pp. 51-53.
53
Stevens, John, Words and Music in the Middle Ages, song, narrative, dance and drama, 1050-1350, Cambridge
Studies in Music (Cambridge: Cambridge University Press, 1986); pp. 199-267 deal with ‘narrative melody’,
which, for French, Stevens limits to early saints lives, chansons de geste and related laisse and stanza-based
forms. However, he does note (pp. 213-14) that German translations of French romances using continuous
sequences of couplets offer the possibility of chant as an alterntive to recitation, which also raises the question
of how the French originals were ‘produced’.

119
The Singer and the Scribe

manuscript, is not meant to be pored over: it is for immediate oral communication.

Nonetheless, it might be objected that the explicit authorial comment of l. 79, and the
implied comment of ll. 83-84 prevent this poem from being a ballad, since it fails the test of
impersonality.54
The same sense of oral immediacy is found in the quite different Trailbaston, a first-
person narrative of a man forced to flee and live as an outlaw to escape what are presented
as the extortions practised by commissioners of Trailbaston, whose own original function
was to prosecute acts of extortion or criminal violence, a remit rapidly extended to cover
conspiracies to pervert the course of justice and similar abuses of power in local officials.55
Composed in mono-rhymed quatrains with lines of varying lengths, the poem opens with a
mixed appeal to courtly lyric and epic modes, proceeds with a brief allusion to pastoral
mode, and finishes with a reference to the composition and first diffusion of the poem:
Talent me prent de rymer e de geste fere
D’une purveaunce qe purveu est en la terre.
Mieux valsit uncore que la chose fust a fere;
Si Dieu ne prenge garde, je quy qe sourdra guere.

**********
Pur ce me tendroi antre bois, suz le jolyf umbray;
La n’y a fauceté ne nulle male lay,
En le bois de Belregard, ou vole le jay
E chaunte russinole touz jours santz delay.

**********
Cest rym fust fet al bois, desouz un lorer,
La chaunte merle, russinole e eyre l’esprever;
Escrit estoit en parchemin pur mout remenbrer
E gitté en haut chemin qe um le dust trover.

[I have a desire to compose a poem and a noble story about provisions made in the land. It would be better if
the thing were still to be done; unless God takes care, there will be war [...] So I shall take to the woods, under
the merry shade; there exist no falseness or evil law, in the forest of Belregard, where the jay flies and the
nightingale sings all day without a break [...] This poem was composed in the forest, under a laurel, where the
blackbird and nightingale sing and the sparrowhawk flies; it was written on parchment so that it would be well
56
remembered, and thrown on the highway where people would find it.]

54
Despite the regularly asserted impersonality of the ballad, many of the most celebrated in the English tradition
do present authorial points of view and reflections. A typical example, similar to the closing lines of the
extract from ‘Turbeville’ quoted above, is Alison Gross: ‘O Allison Gross, that lives in yon towr, / The ugliest
witch i the north country’ (Child 35, 1-2); these lines apparently inspired Sir Arthur Quiller Couch to add the
refrain: ‘Alison Gros, she must be/ The ugliest witch in the north country!’ The Oxford Book of Ballads, ed. by
A. Quiller Couch (Oxford: Clarendon Press, 1910 [repr. 1963]), pp. 52-52.
55
For the legal and historical background to the Commissions of Trailbaston in the years 1297-1305 see ‘Early
Trailbaston Proceedings from the Lincoln Roll of 1305’, ed. by Alan Harding, in Medieval Legal Records
edited in memory of C.A.F. Meekings, ed. by R.F. Hunisett and J.B. Post (London: Her Majesty’s Stationery
Office, 1978), pp. 142-68.
56
Aspin, pp. 69-73.

120
 Philip E. Bennett
While the bulk of the subject matter (generalised protest against the application of the
Trailbaston commissions represented as personal experience) does not fall within the
normal scope of balladry, the appeal in the frame narrative to the Robin Hood tradition of
outlaw ballads is unmistakable, 57 and bears witness to the popularity of the type even at this
early date. Equally interesting is the mixed appeal to lyric and epic in in the first line quoted
above, which, like the last two lines, implies a fusion of written and oral media. If this poem
is not exactly a ballad in the form in which we have it, the launching of it into the oral
tradition, so vividly represented by the throwing of the parchment on which it is inscribed
onto the highway, indicates that its author envisaged its transformation into something akin
to a ballad in the mouths of the wandering singers tacitly charged with disseminating it.
From these varied observations a few tentative conclusions may be drawn. The first of
these is that songs in the oral tradition, similar to ballads, most probably did exist in the
French-speaking world in the Middle Ages. On the Continent, however, they are constantly
subverted by courtly poets who offer parodic versions, which none the less through their
own deforming mirror attest to the existence of the models on which they depend for their
effect. Even so, the best one can say of their existence is that it is precarious and
‘underground’, with no formal recognition in the manuscripts which ‘canonise’ the
literature of the day. In addition it has to be said that the range of material offered by such
poems in Continental France is very narrow. Occasionally an epic or chivalric model may
be glimpsed as the necessary hypertext or intertext allowing interpretation of a parody, and
we have but one tantalising glimpse of a song on the theme of aristocratic largesse, but in
general love and the pastoral are already the staple fare of romances in the thirteenth
century as they will be in the eighteenth. On the other hand in Anglo-Norman England
poems are produced and recorded in manuscripts, which give some glimpses of what a
medieval French ballad tradition might have been, including elements of the heroic.
It is perhaps no coincidence that the nearest approach to a ballad literature we have in
the French-speaking world in the Middle Ages comes from England, which was on the
point of producing a ballad culture in the increasingly dominant English language. Yet one
cannot easily invoke either a supposed conservatism or a popular (non-aristocratic) bias to
explain the phenomenon. The enigma of France is not that it failed of all the countries of
Western Europe to produce a ballad culture in the Middle Ages, but that it went to such
lengths to suppress that culture. It is even more of an enigma that in the highest seats of
learning in France there was still, at the end of the twentieth century, a desire to deny the
reality of that tradition in the name of promoting a purely literate ‘aristocratic’ culture.

57
Although the surviving ‘Robin Hood’ poems are all much later than this, it seems most probable that the legends
developped at the very end of the thirteenth or in the early fourteenth century. At all events not only were the
legends fully formed but ‘rymes of Robyn hood’ were circulating by the 1370s. On the rather tangled early
chronology of Robin Hood see Dobson, R.B. and J. Taylor, Rymes of Robyn Hood, an introduction to the
English outlaw (London: Heinemann, 1976), pp. 1-36. The ‘Trailbaston’ poem would, therefore, be
contemporary with the development of the legends, if not with the earliest poems.

121
This page intentionally left blank
 F.J. Child and Mikail Bakhtin

Richard Firth Green

One of the most striking features of the Child Ballads is their generally sombre tone.1
Though this view has been questioned,2 it remains true that except for a handful of ballads
printed, seemingly almost as an afterthought, in Volume Five, 3 what humour there is, is
largely incidental. This does not, of course, mean that it is inconsequential. The brio with
which the heroine of Tam Lin dismisses a busybody who has remarked upon her thickening
waist could hardly be bettered:
‘Haud* your tongue, ye auld* fac’d knight, hold / old
Some ill death may ye die!
Father my bairn* on whom I will, child
I’ll father nane * on thee.’ (Child 39A:12) none
Neither could the indignation with which Thomas Rymer learns that he has been granted ‘a
tongue that can never lie’:
‘My tongue is mine ain* ,’ True Thomas said, own
‘A gudely* gift ye wad gie* to me! goodly / would give
I neither dought* to buy nor sell, I’d have no skill
At fair or tryst * where I may be.’ meeting

‘I dought neither speak to prince or peer,

Nor ask of grace from fair ladye:’
‘Now hold thy peace,’ the lady said,
‘For as I say, so must it be.’ (Child 37C:18-19).

If such wry moments provide one of the great pleasures of reading the Child Ballads,
another comes from the ironic situations we encounter in ballads like Robin Hood and the
Potter (Child 121), or Queen Eleanor’s Confession (Child 156). But even as we concede
the importance of such features, we must recognise that there is an almost complete lack of
the broader, more racy kind of humour we might expect from these products of folk culture.

1
The English and Scottish Popular Ballads, ed. by F.J. Child, 5 vols (Boston: Houghton Mifflin, 1882-1889).
Ballads from this collection will be cited by number.
2
See Roberts, Warren E., ‘Comic Elements in the English Traditional Ballads’, Journal of the International Folk
Music Council, 3 (1951), p. 76-81.
3
They include Our Goodman (274), Get up and Bar the Door (275), The Friar in the Well (276), The Farmer’s
Curst Wife (278) and The Keach in the Creel (281). Others in this group, such as The Wife Wrapt in Wether’s
S kin (277), are evidently intended as comic, though their humour has not worn well. It is worth noting that,
though more than a third of Child’s Ballads have not survived down to modern times, all of these (with the
partial exception of The Wife Wrapt in Wether’s Skin) are still widely sung today; see Traditional Tunes of the
Child Ballads, ed. by B.H. Bronson, 4 vols (Princeton: Princeton University Press, 1959-1972), 4: 95-277.
The Singer and the Scribe

That we should have any such exp ectation is due in part to the enormous influence of
Mikhail Bakhtin’s study, Rabelais and his World.4 For Bakhtin, as is well known, the
Middle Ages were marked by a dramatic division between the official culture of the
ecclesiastical and feudal authorities and the popular-festive culture of the marketplace. In
his pages, the cursing, brawling, guzzling, copulating folk flaunt the disorderly imagery of
the grotesque body in the face of their political masters, asserting the positive, dynamic,
regenerative triumph of festive laughter over the ascetic seriousness of official ideology. It
is this carnival spirit, he claims, that is reflected in the work of François Rabelais. If
Rabelais himself was the great conduit through which the carnivalesque flowed down to
later ages, however, Bakhtin leaves us in no doubt that its ultimate source is to be sought in
the ageless rituals of the folk themselves.
In light of Bakhtin’s rich and compelling model, we might naturally ask why the ballad
folk, at least as Child represents them, should have been almost totally excluded from this
realm of festive laughter. True, women sometimes turn the tables on men,5 but elsewhere in
the ballads challenges to official ideology — as, for example, in Lamkin (Child 93) —
produce an effect that is anything but carnivalesque.6 What Bakhtin calls ‘the popular-
festive system of images’ (p. 176) has left little impression on the ballads. Feasts, even
marriage feasts, are rarely festive:
And aye* she served the lang* tables, ever / long
With white bread and with brown;
And aye she turned her round about
Sae* fast the tears fell down. (Fair Annie, Child 62A:18) so
Bloody brawls are fought in deadly earnest, with little sense of regenerative vitality:
It’s five he’s wounded, an* five he slew, and
I * the bonny braes* o’ Yarrow; on / banks
There came a squire out o’ the bush;
An pierced his body thorough. (The Braes o’ Yarrow, Child 214B:9)

Unless recuperated by an advantageous marriage, 7 sexual irregularity is almost always

punished:
‘O wae* be to you, my fause* brethren, woe / false
And an ill death mat * ye die! may
Ye mith* slain Clerk Saunders in open field, might (should) have
And no* in bed with me.’ (Clerk Saunders, Child 69F:23) not

Furthermore, Child’s most conspicuous example of a grotesque body — ‘Her teeth was
a’ like teather stakes [stake posts], / Her nose like club or mell [hammer]’ (King Henry,

4
Bakhtin, Mikhail, Rabelais and his World, trans. by Helene Iswolsky (Cambridge, Mass.: MIT Press, 1968).
5
E.g. Lady Isabel and the Elf Knight (Child 4D), The Broomfield Hill (Child 43), or The Baffled Knight (Child
112).
6
One important exception is The Twa Magicians (Child 44).
7
E.g. The Knight and the Shepherd’s Daughter (Child 110), Lizzie Lindsay (Child 226), or The Beggar-Laddie
(Child 280).

124
 Richard Firth Green

Child 32:6) — serves merely to reinforce the prestige of the king who is challenged by this
monstrous creature.
It is this paradox of Child’s uncarnivalesque folk ballad, then, that I propose to explore
in this essay. Apart from matters of national temperament, which seem to me comparatively
unimportant in this context,8 there are, I suggest, three principal ways in which we might try
to account for such a paradox:
1) we might maintain that broadly comic ballads exist, or existed, but that Child
deliberately excluded them from his collection;
2) we might argue that the folk humour characterised by Bakhtin was peculiar to the late
Middle Ages and the early Renaissance and that by the time that most of the Child
Ballads were being collected it was no longer in vogue;
3) we might claim that there is something in the ballad form itself that is inimical to the
carnivalesque.
I shall consider each of these propositions in turn.
In 1972 the singer and musicologist A.L. Lloyd (1908 -1982) introduced a performance
of The Widow of Westmorland’s Daughter with a story, quite possibly apocryphal, of the
song’s scornful rejection by F.J. Child from the canon of traditional balladry.9 The Widow
of Westmorland’s Daughter is a charming tale of a girl who, to stop her mother’s nagging,
returns to her seducer to get her virginity back:
So he kissed her and undressed her,
And he laid her on the bed,
And he set her head where her feet was before,
And so give back her maidenhead.

The folk motif of the restored maidenhead is a very old one,10 and though, like so many
ballads of seduction, this one, too, ends in a recuperative marriage, its air of carnivalesque
inversion is unmistakable; Lloyd’s intuition that it must have been this that put it beyond
the pale for Child is at least plausible. Unlike Sir Arthur Quiller-Couch, the editor of an
Edwardian anthology of ballads, who believed that it was better ‘to omit a stanza from
Glasgerion […] or to modify a stanza in Young Hunting, than to withhold these beautiful

8
Though the great scholar of European balladry, William J. Entwistle, ‘Ballads and Tunes which Travel’, Folklore,
50 (1939), 333, would assign France and the British Isles to different ‘ballad areas’, his criteria hardly apply at
the level of abstraction with which Bakhtin is concerned. Entwistle himself does not discuss humorous
ballads, but it is worth pointing out that the tragic Spanish ballad Blanca niña, which he traces to a thirteenth-
century French fabliau (pp. 335-37), lives on in the English-speaking world in its original comic guise as Our
Goodman (Child 274).
9
Classic A.L. Lloyd, CD (Workington: Fellside Recordings, 1994) track 24.
10
A recipe from 1485 includes such ingredients as seven miles of moonlight, the song of seven Welshmen, the
left foot of an eel, and the creaking of a cartwheel; Reliquiae Antiquae, ed. by Thomas Wright and J.O.
Halliwell, 2 vols (London: Pickering, 1841-1843), 1: 250-51. The story-type appears in Poggio's Facetiae, no. 157.

125
The Singer and the Scribe

things altogether from boy or maid’,11 Child was too good a scholar to bowderlise his texts.
Nevertheless, even if his notorious characterization of popular broadsides as ‘veritable
dunghills’ (in a private letter) had not survived, his genuine ‘revulsion from bawdy
material,’ 12 could reasonably have been deduced, not only from his actual selections, but
also his presumed rejections: there are several songs that must certainly have been known to
him whose exclusion from The English and Scottish Popular Ballads seems to confirm
Lloyd’s belief in his prudery.
Perhaps the most striking of such omissions is John Barleycorn.13 This ballad, which
can be traced back to the early seventeenth century, was adapted by Robert Burns, and
versions of it were printed by at least three nineteenth-century collectors known to Child:
Robert Jamieson (1806), James Henry Dixon (1846), and Robert Bell (1857). It is old, it is
in ballad metre, and it tells a story, so that Child’s reasons for excluding it must primarily
have been based on its subject matter. A drinking song about the cultivation of barley
(buried in the winter, springing up with the new year, cut down and threshed in harvest
time, and finally reborn again through the brewer’s art), John Barleycorn conforms closely
to Bakhtin’s characterisation of a ‘traditional system of images: uncrowning, travesty,
thrashing’ (p. 198), all elements, as he says, ‘steeped in “merry time”, time which kills and
gives birth, which allows nothing old to be perpetuated and never ceases to generate the
new and the youthful’ (p. 211). 14
Almost as striking an omission from The English and Scottish Popular Ballads is the
song of The Derby Ram (Kennedy 304), which was also printed by nineteenth-century
collectors known to Child: George R. Kinloch (1827) and Llewellyn Jewitt (1867). No
doubt Child’s prejudice against it derived in part from his suspicion of animal ballads, a
suspicion which led to a similar exclusion of such ancient ballads as The Frog and the
Mouse (Kennedy 294) and Old Daddy Fox (Kennedy 301). No doubt he also felt, rightly,
that its narrative line was slight (a feature, however, insufficient to disqualify other ballads
like Captain Wedderburn’s Courtship [Child 46] in his eyes), but the traces of a death-and-
rebirth motif are nonetheless quite easily detected (particularly in some of the more bawdy
versions),15 and its link with the kind of ‘grotesque realism’ that Bakhtin finds at the heart
of the carnivalesque is still more obvious:

11
The Oxford Book of Ballads (Oxford: Clarendon Press, 1910), p. xi. The stanza omitted from Glasgerion (Child
67) is “He did not take that lady gay / To boulster nor to bedd, / But downe vpon her chamber-flore / Full
soone he hath her layd” (A.11). His modification of Young Hunting (Child 68) is indicated by the bracketed
text: “[She has kiss’d him by] the candle light / And the charcoal burning red”; Quiller-Couch’s version is by
and large a conflation of Child A and J, but no original of these two lines appears in any of Child’s versions.
12
Palmer, Roy, ‘“Veritable Dunghills”: Professor Child and the Broadside’, Folk Music Journal, 17 (1996), p.157.
13
Kennedy, Peter, Folksongs of Britain and Ireland (London: Oak Publications, 1984), no.276. Ballads and songs
from this collection will be cited by number; for ease of reading, I have punctuated quotations from Ken-
nedy’s text, without further comment.
14
Cf. My Father Had an Acre of Land (Kennedy 300), a ballad first printed in James Orchard Halliwell’s Nursery
Rhymes of England (1842), another collection known to Child.
15
I.e., ‘It took all the women in Yorkshire for to wheel away his stones / And all the women in Dorset to cover up
his bones’; James Reeves, The Everlasting Circle (London: Heinemann, 1960), p.92.

126
 Richard Firth Green

And when this ram was kill-ed sir, there was a terrible flood,
Took four and twenty butcher boys to wash away the blood. (Kennedy 304:9)

Though songs of such grotesque hyperbole are not common in the surviving English ballad
tradition,16 their roots go back a long way,17 and it is difficult to see why Child should have
decided that riddling ballads were worthy of inclusion, whereas liars’ songs were not.18
One of the most fruitful sources for Child’s collection was Bishop Percy’s Folio
Manuscript (ca 1650); 19 indeed, he dedicated his English and Scottish Popular Ballads to
Frederick J. Furnivall on the grounds that ‘without the Percy MS no one would pretend to
make a collection of the English Ballads, and but for you that manuscript would still, I
think, be beyond reach of man’. 20 Fourteen years earlier Furnivall (and his co-editor, John
W. Hales) had published an edition of the manuscript that had omitted a number of its
‘loose and humorous songs’, but Furnivall, who was certainly no prude, had taken care to
have these published privately in a separate edition.21 In a letter to James Russell Lowell,
Child warns his friend that if this latter volume should ever come his way, ‘I advise you to
put it up the chimney (where it will be in its element) or into the fire — where the authors
no doubt are!’ (quoted by Rieuwerts, p.13); hardly surprisingly, then, it makes no
appearance in Child’s voluminous bibliography. At least one of the poems in Furnivall’s
second collection, however, is an obvious contender for inclusion in an anthology of
popular ballads (see Palmer, p. 157). The Sea Crabb (pp. 99-100) is in ballad metre and
tells a fabliauesque tale of a pregnant wife who feels a sudden craving for crabmeat. Her
long-suffering husband returns home late at night with a live crab he has bought for her;
faute de mieux he deposits it in the chamber pot, with farcical consequences which hardly
need elaborating. This ancient ballad was still being sung in the 1950’s (Kennedy 196).
While Child might be prepared to accept the mildly salacious ballad of The Boy and the
Mantle (Child 29) from the Percy Folio Manuscript (no doubt because of its Arthurian
setting and its magical machinery), such a ‘loose and humorous’ piece as The Sea Crabb
clearly formed no part of his notion of the traditional ballad.
A second ballad in the Percy Folio Manuscript is yet more interesting in that Child and
Furnivall both print it, but under completely different guises. Furnivall includes the ballad
he calls Lillumwham among his ‘loose and humorous songs’ (pp. 96-98), because he sees it
as being based (like The Widow of Westmorland’s Daughter) on the motif of the restored
maidenhead. In Child’s English and Scottish Popular Ballads, on the other hand, it is titled
The Maid and the Palmer (Child 21), and is introduced as a version of the story of Jesus
and the Samaritan Woman ‘blended with medieval traditions concerning Mary Magdalen’.

16
The Herring Song (Kennedy 296) came to light only after Child’s collection was complete.
17
Cf. ‘The Felon Sew: a mock-heroic poem from the fifteenth-century’, ed. by G.H. Cowling, in Essays and
Studies by Members of the English Association, OS 8 (1922), pp. 79-89.
18
Cf. Kittredge, G.L., ‘Notes on a Lying Song’, Journal of American Folklore, 39 (1926), 195-99.
19
See Fowler, David C., A Literary History of the Popular Ballad (Durham, N.C.: Duke University Press, 1968),
pp. 132-82.
20
On Child’s involvement with the Percy Folio Manuscript see Rieuwerts, Sigrid, ‘“The Genuine Ballads of the
People”: F.J. Child and the Ballad Cause’, Journal of Folklore Research, 31 (1994), pp. 11-12.
21
Bishop Percy’s Folio Manuscript: loose and humorous songs, ed. by Frederick J. Furnivall (London: Printed by
and for the Editor, 1868).

127
The Singer and the Scribe

In Furnivall’s defence, we might notice that Lillumwham has a refrain that seems con-
spicuously ill-suited to so solemn a subject; indeed, its verbal exuberance (derived in
this case perhaps from the ploughboy or carter’s calls) reminds one of Bakhtin’s stress on
the role of licentious language in popular-festive laughter:
The maid shee went to the well to washe,
Lillumwham, Lillumwham!
The mayd shee went to the well to washe,
Whatt then? what then?
The maid shee went to the well to washe,
Dew ffell of her lilly white fleshe;
Grandam boy, Grandam boy, heye!
Leg a derry, Leg a merry, mett, mer, whoope, whir!
driuaunce, larumben, Grandam boy, heye! (1-9)

It is no part of my brief here to try to adjudicate between Furnivall and Child on this matter;
merely to point out how deaf Child shows himself to be to such carnivalesque echoes in one
of the balldas he prints.
Two other poems in Furnivall’s collection, Fryar and Boye (pp. 9-28) and Panche (pp.
61-67) must have been quite easy for Child to reject for they are not written in any
recognised ballad metre, but they do raise the tantalising question of whether there may
once have existed carnivalesque ballads now lost to us. Fryar and Boye (better known as
Jack and his Stepdame) is a particularly fascinating example. 22 Popular from the fifteenth to
the eighteenth century, it tells of a boy who forces his stepmother’s lover, a friar, to caper
among thorny briar bushes by playing a pipe that compels all who hear it to dance. This
pipe had come to him as one of three wishes granted for an act of kindness, and another of
these wishes enables him to afflict his stepmother with flatulence whenever she reproves
him. Stepmother and friar haul him off to the ecclesiastical court where he sets not only his
accusers but all the court officials cavorting uncontrollably until he is finally pardoned.
Where this tale exhibits the carnivalesque mockery of official culture, another aspect of
Bakhtin’s popular-festive comedy, the grotesque body, is conspicuous in Panche. Panche is
a man of Gargantuan appetite:
soe great a stomacke had hee,
his wiffe did him provide
ten meales a day, his hungar to lay,
yet he was not satisfyed. (5-8)

Invited to his sister-in-law’s wedding, he creates havoc, partly by the enormous quantities
of food he consumes at the banquet, and partly by his wandering about in the middle of the
night in search of further snacks. Though Fryar and Boye and Panche look nothing like
traditional ballads as they stand, they do serve to show that early modern England was
hardly less receptive to carnivalesque humour than Rabelais’s France. Moreover, since

22
For a recent edition see Ten Fifteenth -Century Comic Poems, ed. by Melissa Furrow (New York: Garland, 1985),
pp. 67-153.

128
 Richard Firth Green

similar non-ballad narratives in the Percy Folio turn up in ballad guise later on,23 there
seems no reason in principle for denying that such tales might have furnished the materials
for popular balladry. An obvious way of accounting for their apparent failure to do so, is to
see this as primarily a matter of changing tastes.
This point will serve to introduce the second of my three possible explanations for the
paucity of carnivalesque humour in the Child Ballads: that such humour was restricted to a
specific time period and was no longer thriving during the era of the great eighteenth- and
nineteenth-century ballad collectors. Though in some ways counter-intuitive (we like to
think of folk culture as immensely conservative, not subject to whims of fashion like the
culture of the elite), such a position receives a measure of support from Bakhtin himself.
Much of his chapter on the history of laughter (pp. 59-144) is concerned with the way
changing critical fashions have obscured for later readers the true sources of Rabelais’s
comedy, but there are times when he seems to imply that these fashions themselves reflect
deeper-seated changes in cultural values. One of Bakhtin’s central tenets, after all, is that
the peculiar social conditions of the Renaissance allowed a cross-fertilisation between
official and non-official culture of a kind that was to become impossible at a later period:
“Later, in times of absolute monarchy and the formation of a new official order, folk
humour descended to the lower level of the genre hierarchy. There it settled and broke away
from its popular roots, becoming, petty, narrow, and degenerate” (p. 72). Were it not for
such breaking away from its popular roots, we could imagine that the traditional ballad
might have represented one of the lower levels of ‘the genre hierarchy’ for Bakhtin, but
whatever he means precisely by this claim, it seems quite clear that he felt that folk culture
itself became degenerate after the Renaissance. Of the ‘romantic grotesque’ of the late
eighteenth century, for example, he writes that ‘the direct influence of folk spectacles and
carnival forms, which were still alive though degenerate, […] was apparently not
considerable’ (p. 37).
David C. Fowler has argued strongly against the older view that many traditional
ballads represent oral survivals from the Middle Ages, claiming rather that most are no
older than the period in which they were first collected. The classic era of ballad collecting,
the age of Thomas Percy (b. 1729), David Herd, Joseph Ritson, Walter Scott, William
Motherwell, and Robert Jamieson (d. 1844), coincides with the period that Bakhtin claims
witnessed a degeneration of folk culture, so if both Fowler and Bakhtin are correct we have
no problem accounting for the absence of true popular-festive forms in Child’s collection.
The difficulty is, of course, that Bakhtin’s claims for the degeneracy of folk culture are
somewhat difficult to substantiate.
While it is true that recent historical studies of popular-festive forms in England, most
notably those of David Underdown, David Cressy, and Ronald Hutton,24 have tended to
focus on the period before the Civil War, it could hardly be said that this work has borne

23
Cf. John the Reeve (ed. by Furrow, pp. 177-234) and King Edward the Fourth and a Tanner of Tamworth (Child
273).
24
Underdown, David, Revel, Riot, and Rebellion: popular politics and culture in England, 1603-1660 (Oxford:
Oxford University Press, 1985); Cressy, David, Bonfires and Bells: national memory and the Protestant
calendar in Elizabethan and Stuart England (London: Weidenfeld and Nicholson, 1989); Hutton, Ronald, The
Rise and Fall of Merry England: the ritual year 1400-1700 (Oxford: Oxford University Press, 1994).

129
The Singer and the Scribe

out Bakhtin’s conclusions. In particular, the relationship between official and unofficial
cultures now appears far more complex than Bakhtin was prepared to recognise; and, lest
we should suppose that this situation was peculiar to England, Peter Burke is there to
remind us that it obtained on the Continent as well. 25 While most historians would
recognise that something changed in the relationship between popular and elite after ca
1650, not all would characterise that change in terms of popular degeneration; Burke
himself is quite happy to use the ballad as evidence for the nature of traditional forms (pp.
118-30). Perhaps the historian whose work poses the greatest threat to Bakhtin’s model,
however, is E.P. Thompson, for his Customs in Common, in particular, demonstrates the
continuing vitality of popular-festive forms well into the nineteenth century.26
For the literary scholar, the most obvious place to turn for evidence of such survival is
to the Mummers’ Play. Mummers’ Plays are archetypically festive comedy: their language,
imagery, and action are frequently grotesque, they are deeply concerned with themes of
death and rebirth, and they belong indisputably to the folk. Textually, the Mummers’ Plays
have never had their F.J. Child, in part because none of the three great collectors, Thomas
Fairman Ordish, Reginald Tiddy, or Alex Helm, lived to complete his projected edition, and
in part because the texts themselves are far less manageable: they show both a wider range
of local variation and a far greater uniformity of subject matter than the popular ballad —
all Mummers’ Plays, indeed, can be reduced to three basic types: the Sword Dance, the
Plough or Wooing Play, and the Hero-Combat Play. Nonetheless, when E.K. Chambers
wrote his classic study, The English Folk Play, in 1934, he knew of the existence of some
157 texts,27 some still in performance in his own day. The Oxford scholar Reginald Tiddy
alone had been able to collect twenty-eight texts from live performance in the years
immediately before the First World War in which he lost his life, 28 and in 1981, at the time
Alex Helm’s The English Mummers’ Play was posthumously published, new texts were
still being discovered.29 Even the frankly bawdy Wooing Plays, which clearly offended
Victorian sensibilities,30 have survived down to recent times. Why, then, does the
carnivalesque live on in this one form of folk performance, and not apparently in the other?
Is it perhaps, as I hypothesised at the beginning of this paper, that there is something in the
very form of the ballad that makes it inhospitable to festive comedy?
In many ways, Peter Kennedy’s Folksongs of Britain and Ireland is the diametrical
opposite of Child’s English and Scottish Popular Ballads. Inclusive, where Child had been
selective, it makes no attempt to segregate narrative ballads from other kinds of folksong,
nor does it show any signs of the narrow prescriptivism that led Child to reject, say,
broadside material or nursery rhymes. Above all, where Child’s work had been largely
dependant on manuscript and printed materials collected in libraries, Kennedy assembled

25
Burke, Peter, Popular Culture in Early Modern Europe (Aldershot: Scholar Press, rev. ed. 1994).
26
Thompson, E.P., Customs in Common: studies in traditional popular culture (New York: The New Press, 1993).
27
Chambers, E.K., The English Folk Play (Oxford: Clarendon Press, 1933), pp. 236-44.
28
Tiddy, R.J.E., The Mummers’ Play (Oxford: Clarendon Press, 1923); the book actually contains thirty-three
texts, but four are from the nineteenth-century and one was contributed by Tiddy’s editor, Rupert Thompson.
29
Helm, Alex, The English Mummer’s Play, ed. by N. Peacock and E.C. Cawte (Woodbridge: D.S. Brewer for the
Folklore Society, 1981), p. 66.
30
Helm, Alex, The English Mummer’s Play p. 19.

130
 Richard Firth Green

his collection entirely from the living tradition, from field recordings, made mostly by
himself in the 1950s. A glance at Kennedy’s chapter headings (for example, ‘Songs of
Seduction’, ‘Songs of Country Life’, ‘Songs of Good Company’, ‘Songs of Diversion’)
suggests that here, if anywhere, we should expect to find evidence of the survival of
Bakhtin’s popular-festive forms, yet this expectation is only partially met. What we find, by
and large, is that the festive songs are non-narrative, and that the narrative songs, even
where they deal with earthy subject matter, convey only a limited sense of the
carnivalesque.
Though the old theory, associated with Francis B. Gummere, of the communalist origins
of the ballad is generally discredited today, it is quite clear that certain folksongs do owe
their origin and development to collective merry -making. The Welsh Was You Ever See?
(Kennedy 309) is an obvious example of a song that can be carried by a roomful of people,
individuals chipping in with improvised verses as inspiration takes them. 31 It is equally
clear that such songs will totally lack the kind of narrative focus that might qualify them as
ballads; when they are not entirely random, what structure they possess tends to be that of
the simple list. This is, for example, quite evident in drinking songs: The Barley Mow
(Kennedy 265) lists measurements, implements, and tradespeople associated with the
brewing of beer, Billy Johnson’s Ball (Kennedy 266) lists guests, Here’s to the Grog
(Kennedy 274) lists garments, Tam Broon (Kennedy 283) lists playing cards, and
Twankydillo (Kennedy 286) merely gives a random list of people (the blacksmith, the
gentleman, the pretty girl, the Queen) whose healths the company is invited to drink to. One
variant of this pattern, When Jones his Ale was New (Kennedy 287), is reminiscent of
presentation formulae in the Mummers’ Plays, ‘Here comes I, Old Father Christmas’
(Chambers, p. 20), but even this is still a long way from providing it with a narrative
structure. While it remains possible to reshape a song like Was You Ever See? (309) to
make it tell a story, or at least comment on a local event (Kennedy, p. 682), at that point the
individual performer enters on the scene and something of its original communal anarchy is
lost.
There is no denying a strong carnivalesque element in songs like these, but as soon as
we turn to the narrative pieces in Kennedy’s collection this element becomes noticeably
more attenuated. Songs of seduction and copulation are a staple of the folk repertoire, of
course, but two common elements tend to weaken the effect of full-bloodied popular-festive
comedy in them. One is the use of a first-person narrator (the antithesis of the celebrated
impersonal narration of the traditional ballad), and the other is the use of comic
euphemisms. The following incident from The Overgate (Kennedy 187), for example,
would undoubtedly feel more carnivalesque if told in the third person:
She’d four hot pies and porter
She swallid* them baith* galore. swallowed / both
She ate and drank as much hersel’
As an elephant or a score.

31
Cf. Hench, Atcheson L., ‘Communal Composition of Ballads in the A.E.F.’, Journal of American Folklore, 34
(1921), 386-89; Hatch remarks of verses improvised to the refrain ‘Hinkie Dinkie Parlez-vous’ by soldiers in
the First World War, ‘unfortunately obscenity and filth prevent the printing of any but the harmless ones’
(387).

131
The Singer and the Scribe

O, then we baith gat up the stair

To hae * a contented night, have
When an a’ful* knock cam to the door dreadful
At the breakin’ o the light.

O, it was a big fat bobby,

He got me by the top o’ the hair,
And he gave me the whirligig
Right doon* to the foot o’ the stair. (Stanzas 4-6) down

One has only to compare this with one of Child’s few genuinely comic ballads, The Keach
in the Creel (281), to appreciate the difference in tone. By contrast, The Jolly Tinker
(Kennedy 177) is told in the third person, but its potential for broad comedy is greatly
weakened by its air of knowing jocularity:
The tinker he came down the lane
And on the door did knock:
‘O have you got some pots and pans,
With rusty holes to block?’
And I’ll be bound she had.

She brought him through the kitchen,

She brought him through the hall.
The cook cried: ‘It’s the devil,
He is going to block us all!’
And I’ll be bound he could . (Stanzas 4-5)

Interestingly, Vivian de Sola Pinto records an example of this ballad, ‘current in the Forces’
in 1957 (when Britain still had conscription), which is considerably less euphemistic and
puts a correspondingly greater emphasis on grotesque realism. 32 Despite the limitations of
both first-person narration and euphemism, some of Kennedy’s songs (The Greasy Cook
[129], for example) display carnivalesque humour of a fairly high order, but few would
qualify for inclusion in an anthology of traditional ballads, narrowly defined.
As a matter of fact, Bakhtin himself provides us with a way of accounting for this
phenomenon. One of the most fundamental aspects of the carnivalesque for Bakhtin is its
roots in communal consciousness: ‘medieval laughter is not a subjective, individual, and
biological consciousness of the uninterrupted flow of time. It is the social consciousness of
all the people’ (p. 92). Thus, even death can be the target for festive laughter since in the
experience of the community it represents the promise of regeneration and new life; only at
the level of individual self-reflexivity is the prospect of annihilation a source of terror.
Subjectivity, then, is the enemy of festive laughter; whatever it cannot translate into ‘cold
humor, irony, sarcasm’ (p. 38), it is likely to banish as mere vulgarism. For these reasons
the medieval comic theatre (a theatre without footlights, as he is at pains to stress [p. 257])
was for Bakhtin the literary medium closest in spirit to the world of the carnival. From this
perspective it is entirely understandable that the Mumming Play should best represent the

32
The Common Muse: an anthology of popular British ballad poetry, 15th -20th Century, ed. by Vivian de Sola
Pinto and A.E. Rodway (1957; Harmondsworth: Penguin Books, 1965), pp. 596-97.

132
 Richard Firth Green

carnivalesque in our own times, and that the ballad with its performer, its individualised
narrator, should preserve only echoes of an older communal ethos. As we have seen, these
echoes are made all the fainter where the narrator draws attention to himself, representing
his story as a personal history and relating it with a knowing wink. Indeed, we might go still
further and suggest that a strong narrative line (and spare, taut narration is after all a quality
much admired in the traditional ballad) is by its very nature unlikely to prosper in the
sprawling, cyclical, perpetually inchoate domain which festive laughter naturally inhabits.
Perhaps, then, it is no accident that, except for a few anomalous survivals like John
Barleycorn or The Derby Ram, the traditional ballad should have proved so relatively
inhospitable to the carnivalesque.
F.J. Child was not temperamentally attracted to the kind of humour I have been
discussing in this paper, yet even had the job of editing The English and Scottish Popular
Ballads fallen to someone like F.J. Furnivall it seems doubtful that the overall shape of the
collection would have been very different. No doubt this is partially due to changing social
conditions which made later ballad performers, confronted with popular-festive forms,
more likely to mimic the self-conscious drollery of men like Tom D’Urfey (a drollery
reflected in the arch first-person narratives still common today), but its ultimate causes may
lie deeper yet. They are to be sought perhaps in a clash between two distinct modes of folk
culture: between individual performance and communal participation, between ironic
restraint and grotesque overindulgence, between, finally, an awareness of the painful
brevity of linear time and a celebration of the ageless rhythms of death and rebirth.

133
This page intentionally left blank
 Echoes of Authority:

audience and formula in the Scots ballad text.

Charles Duffin

When it comes to drawing literary interpretations from orally rooted, popular ballads it
seems to be as a matter of course that we concentrate our attention on the singers and
scribes who, on the face of things, share joint responsibility for producing the actual texts
that we work from. These texts, viewed as editorially mediated products of oral creativity,
provide a fertile source of material, amenable to the various kinds of analyses that mark the
literary process in its quest for critical authority. A common barrier to all of those
approaches arises when we are forced to confront the presence of an audience, the ghost in
the ballad text. Critical response to this audience tends to vary in proportion to the degree of
literariness demanded by the particular strategy employed, so, at one extreme, that presence
might be completely exorcised from the text in the interest of the reading, while at the other
it is invoked contextually as the creative genie of the text, reducing in the process the
individual singer’s role to that of automatic mouthpiece of the tradition. The common
ground between these extremes is the lip service that most of us pay by acknowledging that
audience while skirting the implications of its presence in the text.
Of course, it is the implied nature of that presence which makes it difficult to assign the
audience a role in the critical equation that is in some useful way measurable against that of
singers and scribes. This difficulty is compounded by the need to model that audience
according to a set of cultural circumstances which for a number of reasons, not least of
which is literacy itself, cannot be reconstructed other than theoretically. We can seek
general assistance in the vast amount of field work carried out among currently vital oral
traditions, but cultural anomalies notwithstanding, varying degrees of literary
interdependence and the sublime interference of a new wave of technology-based orality
can all introduce potentially distorting elements when re-applied to a specific, historically
bound tradition like the Scots ballads. We can go some way to tracing the shape of our
notional audience by this means but it seems to me that its essence resides ultimately in or
around the text.
So, what are we looking for? Initially, we are looking for textual evidence which
suggests how the creative and critical dynamic between singer and audience might express
itself, but how do we discriminate between those aspects of the text that we assign to
individual creativity and those which express a more engaged commonality? Choral
components aside, there is very little on the surface of the text that would allow us to make
that kind of neat distinction. Unless we can find a way of tracking those elements of
performance which are dedicated to that relationship, it is difficult to propose a role for the
audience in the critical process. It is here, I believe, that singers and scribes come to our
assistance. We may not be able to reconstruct an authentic oral context for the ballad text,
The Singer and the Scribe

but because of the enthusiasm of early ballad collectors and the continuing vitality of the
ballad in a post-literate culture we can apply ourselves to determining what happens to what
we might describe as the negative space that the audience occupies in the oral text as it
collapses in on itself under pressure from the literary process. Once we have taken account
of the effects of literary intrusion on that space we are able to concentrate our attention on
what remains and how that relates to the audience’s role in the critical process.
An approach like this depends to a great extent on the Homeric scholar, Milman Parry’s
theory of oral-formulaic composition, developed by his pupil Albert Lord and applied to the
Yugoslavian epic tradition from where it has been drafted in and, with varying degrees of
success, applied as a critical apparatus to the Scots ballad tradition, most notably by David
Buchan.1 This has not been without controversy. Parry, himself warned of the dangers of
the partial application of his work to other oral cultures, but despite that warning the
attitude to audience in the oral-formulaic approach to ballad tradition has been one of
general conformity to Cecil Sharp’s liberal extrapolation of Parry’s thinking, ‘the individual
invents; the community selects’ — this, without providing any real explanation, or
evidence, of how that dynamic can be witnessed operating within the text. One notable
exception to that conformity is Flemming Andersen, whose sceptical view of oral-formulaic
method has, despite an occasional tendency to throw the baby out with the bathwater, given
up a great deal of useful material relating to the singer’s utilisation of formulaic material to
arouse narrative expectation in the audience.
One of the main reasons, in my view, for glossing over the audience’s role is a general
tendency to import only those parts of Parry’s thinking which apply to singers, so avoiding
a confrontation with the intangibilities of audience. Parry himself, may, to some extent,
share in the blame for this. His decision quietly to sideline a significant element of his own
theory, following the publication of his original thesis, may have gone some way to
promote an imbalance in the broader application of his original definition of what
constitutes a formula in oral poetry — ‘a group of words which is regularly employed under
the same metrical conditions to express a given essential idea’.2
For Parry, however, the oral style to which the poet and the audience are committed
jointly by tradition is not simply dependent on the same fixed repeated groups of words. In
such a case there would be no way in which change could be accommodated and this would
result in a sense of cultural stasis leading, ultimately, to a moribund tradition.3 Creativity in

1
Buchan, David, The Ballad and the Folk (London: Routledge and Kegan Paul, 1972).
2
Parry, Milman, The Making of Homeric Verse, ed. by Adam Parry (Oxford: Clarendon Press, 1971), p. 272.
3
Parry, The Making of Homeric Verse, p. 334. Change in an essentially conservative oral culture is, Parry claims,
slow but sure and comes through change in spoken language and the association between peoples with the
same language but a different dialect. Different dialects effect change but words and forms are not thought of
as words of a certain locality, they serve simply to carry the style above the commonplace everyday speech.
As the spoken language changes so does traditional poetic diction, but there is no requirement to give up
existing formulae which provide an archaic element. The poet may also introduce artificial elements by
analogy with the real created by the need for a formulaic unit of a certain length which can only be got by this
means. Adaptation also gives rise to the artificial whereby part of a word may be changed, part may be
archaic. Constraint on the technique of versemaking ensures longevity of formula even though the language

136
 Charles Duffin

oral poetry is, for Parry, brought about through analogy which creates new formulary
expressions on existing particular word-models and on the sound-pattern of existing
formulaic expressions. This view of formula allows for flexibility in the ideas of oral
poetry, which cannot always be inferred from the central premise of his theory, but it is a
view which he considered as important enough to share italicisation in the same paragraph
as his original definition. He continued: ‘the formulas in any poetry are due, so far as their
ideas go, to the theme, their rhythm is fixed by the verse-form, but their art is that of the
poets who made them and the poets who kept them’. 4
The controlling medium of this flexibility is the tradition itself geared to an aesthetic
imperative.
The oral poet expresses only ideas for which he has a fixed means of expression. He is by no means the
servant of his diction: he can put his phrases together in an endless number of ways; but they still set bounds
and forbid him the search of a style which would be altogether his own. For the style that he uses is not his at
5
all: it is the creation of a long line of poets or even of an entire people.

This notion aroused a great deal of antagonism from his contemporaries and led ultimately
to Parry being accused of being the Darwin of Homeric scholarship.6 The innate Darwinian
implications of Parry’s theory were deliberately downplayed by himself to deflect that kind
of criticism and this may have led to a critical oversight by those who went on to apply his
ideas to the ballad tradition. Given the underlying dependence of the recent critical history
of Scots ballad texts on Parry’s oral-formulaic theory it is appropriate to review the
application of Parry’s ideas to the Scots ballad tradition not only in the light of his own
Darwinian construct but also in that of contemporary neo-Darwinist scrutiny of cultural
development as complementary to scientific evolutionary process.
Neo-Darwinism applied to the cultural aspects of human behaviour invites controversy,
particularly when its proponents are moved to suggest that language mutates in the manner
of genes. Whether this kind of explanation is viewed as analogy masquerading as science or
science working through analogy is not yet clear, but the underlying notion of language as a
generational artifact of a biological process invites us to look again at Parry’s notion of a
formulaic discourse where formulae are replicated, mutate into variables and those which
are selected by the tradition accrete layers of associative meaning which ensures cultural
vitality. Viewed from a neo-Darwinist perspective, Parry’s formulaic discourse illustrates
potentially a working memetic complex, where a meme is defined in simple terms as a ‘unit
of cultural inheritance’ — the psychical parallel of a gene.7

has become archaic. Events may be new but they are told in the traditional way, so that the new becomes
possible by fullest use of the old.
4
Parry, The Making of Homeric Verse, p. 272.
5
Parry, The Making of Homeric Verse, p. 270.
6
Wade-Gerry, Henry Theodore, The Poet of the Illiad (Cambridge: Cambridge University Press, 1952), p. 38.
7
Dawkins, Richard, Unweaving the Rainbow (London: Penguin Press, 1998), p. 301. Dawkins, drawing on the
work of memetic theorists, Daniel Dennett and Susan Blackmore, extends the definition of the ‘unit of cultural
inheritance’ thus: ‘A meme is, by analogy, anything that replicates itself from brain to brain, via any available
means of copying’. Recognising the dangers inherent in the breadth of that definition, he raises his own

137
The Singer and the Scribe

When Parry talks about a poet adopting a style which belongs to an entire people he
makes it clear that it is this formulaic discourse, which allows the poem to be made in a
particular way, that is being appropriated by the performer. That formulaic discourse has its
roots, as Walter Ong notes, in the day-to-day thought processes of a pre-literate culture
where the entire noetic economy, or mental storehouse, is dominated by formulae and poets
simply maximise those processes in order to produce artistic products.8 That creativity is
governed by the formulaic diction itself. Parry is not denying the poet’s individuality, he is
simply proposing that individuality is to be expressed within the context of a formulaic
diction that is somehow fixed by common agreement. In order to sustain a dynamic and
vital culture there must be within the formulaic discourse some mechanism to allow new or
noteworthy information to enter the tradition. It follows from his remarks that if a
formulaic diction belongs to an entire people and the poet is governed by that diction then it
is that people, as a whole, who must stand guarantor for the way it is utilised and it is the
audience, in conjunction with singers, who authorise those new elements of style which can
be accommodated within the formulaic discourse and become a useful fixture in the poetic
economy. Parry’s notion of a formulaic discourse suggests a process of negotiation between
generations of poets and audiences by which means those innovations come to be absorbed
within the complex generational matrix which establishes and sustains a tradition through
time.
For many of those critics who have applied oral-formulaic theory to the ballad tradition,
the mechanism of that process of negotiation resides in the narrative itself. David Buchan
relies heavily on Lord’s recurrent dictum: ‘the tale’s the thing’. This is worth noting
because Lord, working with a living Yugoslavian tradition, took particular pains to avoid
any Darwinist implication in developing his tutor’s theory.9 When pressed to explain how
the oral narrative coalesces in performance, Lord substitutes for Parry’s Darwinism, the
notion of a tension of essences, something analogous to Adam Smith’s invisible hand of

questions about the limited usefulness of the ‘scientific poetry’ that draws parallels between genes and memes.
His unease seems justified given the scientific priority attached to generating and testing against evidence.
Within such a broad definition as ‘the unit of cultural inheritance’ the scientific premise is rendered
meaningless. If, however, it is possible to generate and test within a culture-bound system of recognisable
units of cultural inheritance, in this case Parry’s conception of an oral-formulaic diction geared to narrative
production, then the outcomes should be measurable to a degree that is at least coherent in terms that
Dawkins, as a scientist and a Darwinian, would recognise.
8
Walter Ong, Rhetoric, Romance and Technology (Ithaca; London: 1971), p. 286.
9
Buchan, The Ballad and the Folk, p. 53. Extrapolating from Lord’s notion Buchan attempts to subsume its
vagueness within the recognisable architectonic features of the Scots ballad form: ‘A conceptual pattern
called by Lord “the tension of essences”, whereby certain narrative elements automatically cohere, would
suggest that there are other hidden patterning forces, as yet undissected working within oral tradition. Just as
the aural patterns reflect the non-literate person’s highly developed sense of sound, so these architechtonic
patterns reflect how his mode of apprehension is spatial as well as simply linear and sequential’.

138
 Charles Duffin

God in economics, and, it has to be said in neo-Darwinian terms, an unquestionable

skyhook: that is to say, it is an argument suspended from a mystery. 10
Filtered through Lord and Buchan, Parry’s ideas limit the critical role of the traditional
audience to their demand for a consistent narrative without drawing on Parry’s Darwinian
construction of how that is sustained. Perhaps paradoxically, it is left to an oral-formulaic
sceptic like Flemming Andersen to broaden the implications of how the modes of narration
function as a possible critical apparatus in ballad texts. Commenting on the associative, or
what he describes as ‘supra-narrative’ function of ballad formulae, he remarks: ‘Appearing
in the same kinds of contexts in a large number of ballads the formulas have become
imbued with specific overtones, which have bearing on the immediate narrative contexts in
which they are employed’.11
Here, it appears to me that Andersen consciously, or otherwise, takes up Parry’s notion
of the formula as a kind of poetic meta-language which generates the narrative orientation
of the ballad. The formula is deployed in such a way as to ensure audience and singer
occupy, simultaneously, the same psychological space in the course of the narrative. This
assumes that individual singers are sufficiently skilled in their art so as to manipulate that
space in the performance, but he further claims: ‘Formulas may also serve as signals of
what is to come. Different formula families point to different ways of resolving the
conflicts, and consequently produce narrative variation by presaging different kinds of story
endings’.12
Here he is advancing a more powerful case for the critical role of the audience in which
they interpret through the formulaic disposition of the ballad what the likely narrative
outcome might be. If the singer is not sufficiently skilled, then, by implication, this meta-
language becomes dysfunctional, with audience and performer both occupying a different
psychological space as they engage with the immediate narrative. The singer who generates
narrative expectations by utilising a particular type of formula has, quite literally, lost the
audience if those expectations are not subsequently met. This is as far, I think, as Andersen
can go in asserting the critical role of audience. He rejects Parry’s original definit ion of
what constitutes a formula (‘a group of words which is regularly employed under the same
metrical conditions to express a given essential idea’) on the grounds of what he describes
as the multiform nature of ballad formulae, moulding his own adaption of Chomsky’s ‘deep
structure’ linguistic theories into a generative model of formulaic construction, in
preference to Parry’s substitutionary model of formulaic creation by analogy. Nevertheless,

10
Darwinians, such as Dawkins, Dennett and Blackmore have a tendency to apply the ‘skyhook’ argument to any
non-scientific approach to explaining cultural history and activity. It appears to me that Parry’s theory of an
oral-formulaic discourse applied to a tradition provides an ideal opportunity to test their theory against the
evidence generated by the tradition that resides in the orally based text. If that process confirms their ideas
about how language and culture ‘evolve’, we are some way to a coherent evolutionary theory of culture. If,
however, that fails to sustain the memetic theory, they could find themselves swinging from the biggest
skyhook of all, the meme itself.
11
Andersen, Flemming G., Commonplace and Creativity (Odense: Odense University Press, 1985), p. 25.
12
Andersen, Commonplace and Creativity, p. 25

139
The Singer and the Scribe

I believe that Andersen highlights here the Darwinist imp lications of Parry’s extended
definition of the formulaic discourse.
It is worth noting W.J. McCarthy’s claim that there is no obvious obstacle which disbars
Parry’s ‘substitutionary’ models of formula (creation by analogy) from operating as, or in
tandem with, Andersen’s linguistically derived ‘generative’ deep-structure models.13
Daniel Dennett, a contemporary neo-Darwinist, is another who sees no real obstacle in
accommodating Chomsky’s transformational grammar within an evolutionary model aside
from the not insignificant fact of Chomsky’s personal reluctance to condone it.14
Andersen’s insistence on exactly the same kind of metrical sound and sense precision
for a mature epic tradition and for a ballad tradition which operates under a completely
different set of cultural conditions, including an imminent confrontation with mass literacy,
is asking too much of Parry. The point surely is to examine how formulae emerge from the
fusion of those metrical conditions themes and ideas which support the particular tradition
and broadly speaking, it seems to me, the single-line ballad-formulae work in a way that is
coherent in respect to Parry’s notion of what constitutes a formulaic discourses where they
are not disrupted by literary encroachment. Andersen, however, prefers to group formulae
together in families, varying in length from one line to several stanzas which, he claims,
excite the same narrative expectations under different metrical circumstances.
He recognises those formulae that appear in linear form but argues that it is their verbal
aspect and not their metrical shape that expresses the significant narrative idea. To illustrate
his contention he examines several examples of the ‘sewing’ formula which is common to
the ballad tradition, of which is typical:
Fair Margaret sat in her bonny bower, St.2
Sewing her silken seam.
15
Tam Lin (Child: 39D)

The narrative function (essential idea) here is, according to Andersen, the conventional
depiction of a woman of some standing which overlays a supra-narrative expression of

13
McCarthy, W.B., The Ballad Matrix - Personality, Milieu and the Oral Tradition (Bloomington: Indiana
University Press, 1990), p. 11. McCarthy also notes that Andersen’s ‘rejection of the term oral-formulaic
seems more political than conceptual, since he admits a recreative role for the singer, and describes balladry
as an oral art built around the formula, albeit a more restricted kind of formula than Lord posits’. The
recognition of Lord’s influence on Andersen, over Parry, once again suggests that Parry has only really been
considered as a secondary source for oral-formulaic theory which may account for Andersen’s consistently
taking for granted the mechanical basis of Parry’s work as emphasised by Lord rather than in its full aesthetic
compass as outlined by Adam Parry.
14
Dennett, Daniel C., Darwin’s Dangerous Idea (New York: Simon and Schuster, 1991), p. 398. Dennett’s
contention that Chomsky is simply avoiding the implications of his own theory is given some licence by
Chomsky’s own notion that if, as he maintains, language is an innate function, cognitive structures should be
studied in the same way as organic structures. See also, Pinker, Steven, The Language Instinct (London:
Penguin Press, 1995), pp. 8-10.
15
The English and Scottish Popular Ballads, ed. by F.J. Child, 5 vols (Boston: Houghton Mifflin, 1882-1889). As
this is generally accepted to be a standard reference work, all quoted ballad texts are identified where
appropriate according to Child’s classification. Each stanza quoted is identified by ‘St. number’.

140
 Charles Duffin

longing for a lover. This latter aspect emerges from the tension between the act of sewing
which occupies the hands and the freedom of thought which is the corollary of that
mechanical activity. This subconscious longing is borne out by narrative events which
occur in its wake such as a sexual encounter. Andersen also argues that the ‘same overtones
are evoked in other ballads’ by formulaic phrases which denote other similar pre-
occupations as with:
Fair Margaret sat in her bower window, St .3
A combing of her hair.
Fair Margaret and Sweet William (74A)

These two distinct formulae, he argues, are members of the same formula family, are
interchangeable and carry the same narrative function and associative potential. His
evidence for this comes from a comparison of the opening stanzas of two other Child
versions (41A & 41B: Hind Etin) in which, he claims, this interchangeability is
exemplified. While it might be possible to regard this ‘family’ relationship in terms of
Parry’s notion of creation of new formulae by analogy, it is not possible to accept
Andersen’s idea, based on his own explanation of formulaic construction, that they are
invested with exactly the same associative function as each other. The act of hair-combing,
particularly when that hair is a significant colour (yellow in 41B), in the ballad tradition
generates a more profoundly sexual connotation than the sewing formula, particularly
where it appears with the ‘kirtle’ or ‘skirt’ formula in close attendance, as it does in both of
Andersen’s examples where the combing formula is utilised. Andersen himself notes the
propriety of the action of sewing in a surface-narrative sense and it is also significant that in
comparing the two versions of Hind Etin, the narrative outcome is significantly different,
where those different formulae have been used with the more overtly sexual narrative (41B)
producing a negative outcome and the more sensual than sexual ‘sewing’ formula (41A)
producing a positive transformation of narrative circumstances.
The point here is to illustrate that formulaic diction can be seen to operate not casually,
or by accident, during performance. Andersen assumes that these so called ‘inter-
changeable’ formulae are deployed by artists of equal skill and that they have access to a
common but variable formulaic vocabulary. He does not consider that one or the other may
have simply been less skilled in the deployment of formulaic diction and as a consequence
produce a significantly altered rendition.
In order to assess those qualities, these single-line formulaic phrases would have to be
examined in the context of their relationship with attendant formulae and the subsequent
narrative action. If, for example, a traditional audience was led to expect an overtly sexual
encounter because of the presence of a ‘combing’ formula in conjunction with a ‘skirt’
formula and what arose was a more subtle variation such as the advent of a lover’s ghost
(which Andersen elsewhere relates to the ‘sewing’ formula) then it might be fair to assume
that the performer had, in a traditional sense, launched the wrong associative signal and as a
consequence could be regarded as less skilled than the one who properly evoked the
supernatural resonance of the ghost through the ‘sewing’ formula.
When he comes to regard the longer narrative units, which he describes as stanzaic
formulae, Andersen cites the narrative condition of separation between principal characters

141
The Singer and the Scribe

that necessitates the call for a ‘bonny boy’ to act as go-between. The ‘Bonny Boy’ formula,
according to Andersen can take the form of a single line, the ubiquitous ‘Where will I get a
bonny boy?’ or a whole stanza:
Gin I had but een o my father’s merry men St.3
As aft times I’ve had mony,
That wad run on to the gates o Aboyne
Wi a letter to my rantin laddie
The Rantin Laddie (240)

This is an unusual choice because it is a linguistic variant, chosen perhaps to illustrate

Andersen’s point about multiformity. In other versions of this same ballad printed by
Child, however, there are alternative versions of the ‘bonny boy’ formula which conform
more readily to the most common utilisation of this formula, of request and echoed
response, which Andersen also quotes from a variety of ballad sources to illustrate the
linguistic variation in the formula including
‘Where will I get a bonny boy, St.10
Will win gold to his fee,
And will run into Chiel Wyet’s,
With this letter from me?’

‘O here I am, the boy, says one, St.11

Will win gold to my fee
And carry away any letter
To Chiel Wyet for thee’
Lord Ingram and Chiel Wyet (66A)

He then quotes a trinary stanzaic pattern which may involve a non-formulaic interpolation
by the singer which according to Andersen forestalls the ‘acceptance’ response
‘O where will I get a bonny boy, St.7
To rin my errand soon,
That will run into fair England,
And haste him back again?’

O up it starts a bonny boy, St.8

Gold yallow was his hair,
I wish his mither meickle joy,
His bonny love meikle mair.

‘O here I am a bonny boy, St.9

Will rin your errand soon
I will gang into fair England,
And come right soon again.’
Johnny Scot (99A)

At this level of construction, Andersen claims that the formula has almost become an entire
ballad scene, but it appears to me that all of the examples above can be read as a single-line
‘bonny boy’ formula, sometimes contextualised by a request, or an answer and response

142
 Charles Duffin

orientation, and qualified in narrative terms by attendant formulae such as ‘will win gold to
his/my fee’, as well as by subsequent narrative information that relates the single-line
formula to the overall narrative and conceptual architectonic of the ballad.
Andersen recognises that these formulae are, ‘traditional expressions, moulded over
many generations, and that their linguistic flexibility is not completely unrestricted’. I
would suggest that it is perhaps more restricted than he would have us believe.16
Even if we were to consider the trinary stanzaic form above a complete formulaic unit
(even though, as Andersen notes, the middle stanza is not in itself formulaic in his own
terms) we must also find reasons to explain its variation in economy. Interpolations of a
subjective and self-conscious kind by the narrator, typified by the central stanza, are
strongly indicative, not of an oral mindset but of a nascent literary sensibility. That same
interpolation also includes another formulaic phrase —‘Gold yallow was his hair’ — which,
as noted above, may have sexual connotations for a traditional audience whereas here it
could rather be a sign of literary sentimentality, alien to the ballad narrative. Taken in
conjunction with the ‘messenger’ formula this may be considered as both narratively and
conceptually inappropriate. It may or may not appear to be formulaic, it may be wrapped in
tradition, but it may also be the result of a flawed performance, or a literary interpolation in
a transitional work addressed to the emerging priorities of a literary audience.
Evolution, in an oral culture, although it allows for variability, does not, from Parry’s
perspective, allow for the kind of expansiveness apparent in Andersen’s notion of
multiformity. In a culture that is obliged to conserve its noetic stock in memory alone,
economy is the key to retaining information. Where Parry’s studies produced strong
evidence to support the notion that Homeric works are tightly metrically bound, this in itself
might only be an indication that the particular tradition upon which those works were based
had reached a degree of associative conciseness that made such close metrical economy
possible. That is to say that the formulaic diction, operating on the principal of utility,
exercised on the listener and artist together the maximum connotative potential in the
tightest possible narrative conditions — the metrical circumstances prescribed by Parry.
The relative stability of some linear formulae in the ballad tradition suggests that there
are formulae which have been refined to a degree of sound and sense which approaches
those circumstances and it is this that points us to the essential critical role of the audience.
What is it that stops variations from developing into highly individualised original songs?
The singer’s individuality is the key to variation in the oral economy but any attempt to
express that individuality outwith the confines of formulaic discourse results in material
which is incomprehensible to the culture as a whole. Innovation is vital but the meaning
which attaches to that innovation cannot be left to the singer alone. In order for an
innovation to become part of the tradition it must accrete narrative and associative meaning
that is universally comprehensible to the tradition. When that innovation attains a level of
common currency that allows it to release that accreted meaning into the narrative under
optimum metrical conditions and it functions, in co-operation with other established
formulae, in order to produce a satisfactory narrative outcome it has, in Parry’s terms,

16
Andersen, Commonplace and Creativity, p. 59

143
The Singer and the Scribe

evolved to the point where it is an essential element in formulaic discourse. That can only
be achieved where the singer’s drive to individuality is attenuated over time into
commonality by meeting the narrative expectation of the audience. Through that process the
conceptual basis of the original message is preserved and its resonance enhanced to
accommodate change through time. The establishment of a working formula is in itself the
hallmark of critical authority. Every time a coherent formula appears in a ballad narrative it
echoes that critical authority; consequently when it is rendered incoherent by narrative
intrusion and the effects of literacy, the non-literate audience suffers a loss of critical
significance.
If we had the luxury of being able to trace the development of a single ballad through
several generations of a particular oral process we might indeed be able to identify that
evolutionary process at work in the ballad tradition. The history of ballads as texts,
however, is such that it would be difficult to trace given that, by the time they were thought
worthy of collecting, they had for some time been interacting with literary forces through
broadsheets and manuscripts and were also being subjected to editorial pressures by early
collectors. Thanks to those same collectors we do, however, have access to an enormous
selection of ballad-texts and versions of ballad-texts from a variety of sources which, when
we take account of the equally varied editorial practices applied to them, might help us
identify any breakdown in the formulaic diction.
Walter Scott’s protracted thesis of decay in his Minstrelsy of the Scottish Border
illustrates a paradox that lies behind the explosion of ballad-collecting: that literary elites
were fascinated by the ballads as cultural products while their literary sensibility, driven by
the quest for originality of expression, was repulsed by the formulaic burden which attached
to them.
The collections of rhymes, accumulated by the earliest of the craft, appear to have been considered as forming
a joint stock for the common use of the profession; and not mere rhymes only, but verses and stanzas, have
been used as common property, so as to give the appearance of sameness and crudity to the whole series of
17
popular poetry..

Scott has a recurring knack of recognising the essence of oral poetry and simultaneously
missing the point of his own observations. What he is condemning here is the very stuff of
oral tradition, the formulae and architectonic repetitions which may ensure not only the
delivery of the narrative message but also, in the manner of their disposition, the gloss of
currency that attaches to that message at a particular point in the tradition, which seeks an
evaluatory response from the traditional audience. Scott, however, is simply reflecting
literary intolerance to formulae which are regarded in terms of redundancy. With the
complicity of his educated readers, Scott quite literally edited the critical authority of the
non-literate audience out of cultural contention. In editing and glossing out the
commonplaces and so-called clichés of formulaic discourse that offended the romantic
literary mind, he stifled the means by which the traditional audience could understand and
evaluate the current state of the tradition. These frozen and often adulterated texts were
collectively transformed into a literary dream of national identity mouthed by virtual bards

17
Walter Scott, Minstrelsy of the Scottish Border, rev. and ed. by T.F. Henderson, 4 vols (Edinburgh: Oliver and
Boyd, 1932), 1, 8.

144
 Charles Duffin

for the benefit of their own cultural aspirations. The complicity of Scott and his ‘educated
reader’ meant that ballad-texts were effectively re-configured into the software of a literary,
romantic discourse that assumed critical authority over the products of tradition.
This is not to say that we have no ballad-texts that are genuinely oral in their conception
or that all texts are by definition over-mediated in the process of transformation. Collectors
like William Motherwell show a remarkable prescience in regard to formulaic diction when
discussing the language and mechanisms of ballad-discourse and even that bloodthirsty
vegetarian Joseph Ritson demanded a respect for the oral process. It has to be said, though,
that his powers of discrimination when applied to that purpose often fell short of his
ambitions. These and other exceptions aside, many texts are profoundly marked by the
effects of literacy at the point of performance and transcription and to recognise this
provides a counter-weight when it comes to gauging what may, or may not, constitute
elements of a working oral-formulaic discourse in any given ballad-text.
One such opportunity exists in a comparative study of those texts delivered to us at
different times by Mrs Brown, who stands, according to Child among others, as the doy-
enne of Scots ballad singers.
Although there is no evidence to support the notion that Mrs Brown ever had a public
audience for her work until the demand from literary collectors created it, she had learned
traditional songs and techniques from people who may have had such a public role and it
seems fair to assume that her learning process was to some extent attuned to the critical
priorities of oral performance. Yet the only critical audience she ever had was a literary
audience and it is to the priorities of that literary audience, as she understood them, that her
work is ultimately addressed. She was familiar with the social and sexual politics of the
literary world as distinct from the ballad world and on both occasions when her texts were
being transcribed in 1773 and in 1800 she delivered censored versions of ballads with a
sexual content about which David Buchan admits it is fairly safe to assume deliberate acts
of omission on Mrs Brown’s part.18
If that is a safe assumption, then we must also assume a critical role for the literary
audience of which Mrs Brown was not only aware but which she was willing to
accommodate in her compositional strategy. Critical regard was not paid, directly or
indirectly, to a traditional non-literate audience, but to her literate transcribers, a fourteen-
year-old boy in 1773, and a Church of Scotland minister in 1800. More importantly, it was
the aesthetic priorities of the audience beyond the transcription that were exciting Mrs
Brown’s interventions — a wider, literary audience of ballad-collectors and a public
increasingly directed by their rules of aesthetic propriety which was essentially literary. It
was not simply that audience’s sense of sexual impropriety which was protected by Mrs
Brown but also their literary sensibility. Buchan recognises this up to a point:

18
Buchan, The Ballad and the Folk, p. 138. This self-censorship must hold some implications for the notion of
spontaneous re-creation. If it was so in Mrs Brown’s case, at what level of consciousness does censorship
take place? Buchan does not address that question. Considering that these ‘performances’ were to be
transcribed for a literate audience the level of intervention must have been attuned to the expectations which
Mrs Brown shared in common with that audience.

145
The Singer and the Scribe

We are faced with a paradox: the woman who preserved the finest representatives of the
old oral tradition, the tradition of the non-literate rural folk was herself an educated woman,
daughter of a professor and wife of a minister. This paradox is, however, apparent rather
than real, because it is possible, at a certain point in the tradition, for a person to be both
literate and an oral composer. It is only when a person ceases to be re-creative along
traditional lines and accepts the literate concept of fixed text that he or she can no longer be
classed as oral.19
However, a comparison of the texts delivered by Mrs Brown in 1773 and reconfigured
in 1800 illustrates that dynamic precisely. I would argue that that the evidence strongly
suggests that Mrs Brown’s two renditions of The Lass of Roch Royal (76D-76E) are
profoundly altered by the more modern version’s relationship to the text printed in Herd’s
Scottish Songs in 1776, reprinted by Child as (76B) The Bonny Lass of Lochroyan.
Comparing Mrs Brown’s two versions of The Lass of Roch Royal (Child: 76D, 76E)
will, Buchan says, support Bertrand Bronson’s claim that, what Mrs Brown was trying for
in the version of 1800 was not to recover her own text of 1783, but to recover, or re-create,
the ballad itself: the essential, ideal “Lass of Roch Royal”.20
This ‘ideal’ is, for Buchan, underpinned by Parry’s notion of an oral-formulaic creation
but if we do examine these texts we find evidence of an intrusive literary ambience which
undermines the proposition that the work is being re-created in an exclusively oral mode:
‘O wha will shoe my fu fair foot? St.1 ‘O wha will shoe my fu fair foot? St.1
An wha will glove my han? And wha will glove my hand?
An wha will lace my middle gimp And wha will lace my middle jimp,
Wi the new made London ban? Wi the new made London band?

‘Or wha will kemb my yallow hair, St.2 ‘And wha will kaim my yellow hair, St.2
Wi the new made silver kemb? Wi the new made silver kaim?
Or wha’ll be father to my young bairn, And wha will father my young son,
Till Love Gregor come hame? Till Love Gregor come hame?

Her father shoed her fu fair foot St.3 ‘Your father will shoe your fu fairfoot St.3
Her mother glovd her han Your mother will glove your hand;
Her sister lac’d her middle gimp Your sister will lace your middle jimp
Wi the new made London ban. Wi the new made London band.

Her brother kembd her yallow hair St.4 ‘Your brother will kaim your yellow hair St.4
Wi the new made silver kemb Wi the new made silver kaim;
But the king o heaven maun father her bairn And the king of heaven will father your bairn
Till Love Gregor come haim. Til, Love Gregor come hame.’
(76D) (1783) (76E) (1800)
When we examine the opening stanzas of Mrs Brown’s two versions, the conscious
anglicisation of the 1800 (76E) version obliges us to consider whether we are dealing with a
process of traditional idealisation or of literary improvement. Orthography alone might lead
us to speculate that the translation itself might have been an ‘improved’ version of a

19
Buchan, The Ballad and the Folk, p. 64
20
Bronson, Bertrand H., ‘Mrs Brown and the Ballad’, California Folklore Quarterly, 4 (1945), 129-40.

146
 Charles Duffin

traditional performance and this view would be supported not only by the substitution of the
vernacular, as in ‘son’ for ‘bairn’ in the later text, but also by the shift to a future tense. The
tonal ramifications of this, however, suggest a more trenchant strategic shift by the
performer. Comparison of the third lines of stanza four in each these versions reveals the
narrative consequences of those substitutions. In the original version (76D) the essential
narrative idea is conveyed in an impersonal and non-subjective manner which is consistent
with Child’s own idea of a traditional oral style. In the more modern version (76E) the
narrative idea is expressed by a narrator whose voice is critically affected by the
substitution of the standard English ‘will’ for the vernacular ‘maun’ (must) which
‘improves’ the poetic language and maintains the consistency in tense but fundamentally
alters and personalises the tone. This is no longer fatalistic, but positive. The strategic aim
appears to be an attempt to enhance the ‘traditional’ gra mmar in order to infuse the ballad
with a literary polish, but the narrative consequences are to distort fatally the traditional
sense. This undermines Buchan’s notion that this second version is being re-created
according to a traditional poetic grammar as described by Parry and suggests either the
intrusion of an active literary imagination, or that the second text is being recreated from a
different model.
In 1776 Herd’s Scottish Songs contained a version of the ballad, ‘The Bonny Lass of
Lochroyan’, reprinted by Child as 76B:
‘O wha will shoe thy bonny feet? St.1
Or wha will glove thy hand?
Or wha will lace thy middle jimp,
With lang, lang London whang?

‘And wha will kame thy bonny head, St.2

With a tabean brirben kame?
And wha will be my bairn’s father
Till Love Gregory come hame?’

Thy father’ll shoe his bonny feet, St.3

Thy mither’ll glove his hand;
Thy brither will lace his middle jimp,
With a lang, lang London whang.

‘Mysel will kame his bonny head, St.4

With tabean brirben kame;
And the Lord will be the the bairn’s father,
Till Love Gregory come hame.’

The tonal confluence between (76B) and (76E) is marked from the outset, suggesting a
closer narrative relationship at this point between these two works than between either of
these and Mrs Brown’s earlier version. The orthographic similarities are noteworthy, as is
the shared use of the future tense. The essential link between these two texts, however, is
the personalised narrator who in this (76B) version intrudes upon the action. Here, as in Mrs
Brown’s (76E) version, the narrative voice explicitly addresses the ballad-actor in contrast
to the impersonal narration of the early text (76D) where the action is described in a
traditional way. This is not evidence of a passive traditional voice but of an active, literary

147
The Singer and the Scribe

strategy. If Mrs Brown is looking to re-create an ‘ideal’ version, it is not the ‘ideals’ of
tradition that are uppermost in her mind.
This is reinforced by Mrs Brown’s alternative treatment of the traditional triangular
relationship between Annie, her lover and his mother who represents the threat to that
relationship. The entire nature of the threat and consequently of the relationship alters
between the two texts.
Long stood she at her true-love’s door St.9 ‘O open the door Love Gregor’, she says St.9
And lang tirld at the pin ‘O open and let me in’
At length up gat his f’ase mither ‘For the wind blaws through my yellow hair
Say’s wha that would be in? And the rain draps o’er my chin.’
(76D)
‘Awa, awa’ ye ill woman St.10
You’r na come here for good.
You’r but some witch, or vile warlock,
Or mer-maid of the flood.’
(76E)

In the earlier text the audience is made aware from the outset of the mother’s deception of
Annie. The ‘fa’se mither’ is a heavily charged formula which alerts the traditional audience
to the imminent deception. The relationship between Mrs Brown’s (76E) version and the
Herd version (76B) is reinforced here by the use of a stanza which appears in almost
identical form at the same stage of narrative development in both of these but not in the
original (76D) text.
Now open, open , Love Gregory, St.17
Open and let me in!
For the rain rains on my gude cleading,
And the dew stands on my chin.
(76B)

In the modern (76E) text the deception is not revealed until the exchange between Annie
and the mother is over and her lover wakes and recounts his dream, to which the mother
responds:
‘O there was a woman stood at the door, St.24 ‘Gin it be for Annie of Rough Royal St.19
Wi a bairn until her arms That ye make a’ this din,
But I would na lat her within the bowr She stood a’ last night at this door,
For fear she’d done you harm. But I trow she wan no in.
(76D)
‘O wae betide ye, ill woman, St.20
An ill dead may ye die!
That ye wouldno open the door to her,
Nor yet woud waken me.’
(76E)

In the original (76D) version the deception is still intact but understood by the traditional
audience through the use of the ‘fa’se mither’ formula which signifies to that audience that
the mother has deliberately misled both Annie and her lover despite her certain knowledge
of Annie’s identity, established by a test of tokens. This realisation is not shared by the son
in the original version where he remains ignorant of the deliberate nature of the deception.

148
 Charles Duffin

In the later version the deception, which is narratively unambiguous, is sprung by the
mother, increasing the emotional impact. This is consistent with the Herd text, which also
fails to utilise the ‘fa’se mother’ formula. Once again, a common stanza appears at a critical
stage of the narrative development of those two versions which does not appear in Mrs
Brown’s early text.
Awa, awa, ye wicket woman, St.22
And an ill dead may ye die!
Ye might have ither letten her in,
Or else have wakened me.
(76B)

The connotative power of a single traditional formula in (76D) is sacrificed to the literary
shock tactic of dramatic revelation. The mystery is not understood, but revealed. The fact
that Mrs Brown needed to explain suggests that by 1800 she was unsure of the ability of her
audience to appreciate to the connotative reverberations of the ‘fa’se mither’ formula. The
surprise element of the alternative strategy also suggests a literary sensibility at work in the
composition.
According to the Parry view of oral composition, an aesthetic contract between
traditional singer and audience is founded on a common understanding of the tale to be told.
There is no room for narrative surprises and novel elements are only introduced and
accepted by mutual agreement that the tradition is ready to absorb them. 21 The narrative is
fundamentally altered by the demands of the new approach and this culminates in distortion
in the final stanzas. The message compressed within the ‘fa’se mother’ formula is made
literal in (76E) where the protagonist launches a curse on his mother for deceiving him. He
is explicitly aware of the deception. In the (76D) version, however, he is left to mourn the
loss of his lover while it is the traditional audience alone who are aware of the root of that
tragedy. This awareness arises from an educated understanding of the connotative force and
strategic disposition of the formulaic phrase within the narrative.
O he has mourned o’er Fair Anny St.32 ‘O wae betide my cruel mother, St.26
Til the sun was gaing down, And ill dead may she die!
Then wi a sigh his heart it brast, For she turnd my true love frae my door
An his soul to heaven has flown. When she came sae far to me.’
(76D) (76E)

The religious and sentimental expressions in the final stanza of the original are themselves
evidence of an intrusive literary sensibility in the early text, but it is the narrative alteration
that concerns us here. Even in the Herd text the son dies. The consequences of that

21
Vansina, Jan, Oral Traditions as History (Madison, Wis.: University of Wisconsin Press, 1985), p. 35. Vansina’s
experience of contemporary oral performance suggests that whatever the quality of the performance, ‘the tale
must be well known to the oral audience: they must already know the tale so that they can enjoy the rendering
of various episodes, appreciate the innovations and anticipate the thrills to come’. This supports the idea that
variations, such as the one perpetrated here by Mrs Brown, could in no sense have been intended for an
exclusively oral audience. The ‘innovation’, at narrative level, effectively confounds the sense of anticipation
which the connotative ambience of the ‘fa’se mither’ formula evokes.

149
The Singer and the Scribe

relationship when comparing the modern version with the earlier text suggest a significant
disruption of the relationship between the formulaic disposition and the traditional narrative
to the point where the more modern text carries a romantic literary ambience which sets it
apart from the simple formulaic and narrative efficiency of the earlier text.
The literary infestation of the text results in a fundamental alteration in relationships
between the ballad-actors to the point where it is difficult to imagine how that version might
make sense to a pre-literate audience attuned to the narrative expectations raised by any
formulaic discourse. This is borne out ultimately in the alteration of the narrative outcome
where, instead of dying as he did in the earlier version (and, it is worth noting, in Herd’s
version), the lover, who has been deceived by his false mother, is instead condemned to life
as a tragic figure straight from the pages of a literary romance.
Mrs Brown’s re-configuration of this ballad demonstrates something fundamental to our
understanding of Parry’s formulaic discourse which is that something significant happens to
a ballad when a singer fails to recognise the critical authority of the traditional non-literate
audience. The narrative itself is transformed and the key to that transformation lies in the
aesthetic priorities of her perceived audience.
Buchan’s failure to recognise the full connotative power of the individual formulaic unit
and his attempt to incorporate Parry’s definition in an unmodified form has provided fuel
for those critics, like Flemming Andersen, who claim that oral-formulaic studies are
inappropriate to the ballad tradition.
The need to establish the usefulness and limitations of oral-formulaic theory is driven by
the recognition that it provides, perhaps uniquely, a critical framework which allows a
literary scholar the opportunity to appreciate and address the Scots ballad tradition in terms
of a distinctive, non-literary, oral aesthetic. Any other literary approach entertains distortion
rooted in the procedures and prejudices of the time when literary culture assumed
ownership of traditional products, for psychological, political and commercial purposes —
a literary process.
Ballad formulae where they appear, echo the critical authority of the oral audience. The
more frequently they appear to be working coherently within the narrative the more I think
we can assume that the text has been measured by that critical authority. Where they are
displaced or rendered incoherent by the literary interpolations of editors and the
compromised aesthetics of singers they may provide us with comparative evidence of the
kind of space that critical authority occupies in variant text s of the same ballad that are
relatively free from that kind of aesthetic disruption. In terms of understanding what is left,
it seems to me that Parry’s theory, in its broadest application, provides the most
comprehensive means of doing so which both respects fully the critical role of audience and
does not rely on skyhooks.
This does not mean that I am making a case for cultural Darwinism, at this stage I am
more concerned with reviewing critical history in order to make a case for the critical
authority of the oral audience, but I do think that the undeniably snug conformity to the
Darwinian paradigm of Parry’s theory raises interesting possibilities. If Parry’s ideas allow
us a sense of the creative and critical dynamic that operates in a text that is judged to be
relatively free from literary encroachment, we have in those texts an ideal opportunity to
attempt what scientists are always demanding of the humanities — to generate and test

150
 Charles Duffin

against available evidence. In the formulae we have, in neo-Darwinian terms, a measurable

memetic unit and in the formulaic discourse we have, potentially, a dynamic assembly of
memetic units engaged with each other in an associative drive toward narrative production.
If an examination of this process supported the idea of science working through analogy
then, deprived of our skyhooks, we may have to address the implications of a scientific
basis for oral aesthetics; if, however, it reveals analogy masquerading as science then we
may find in our critique of Parry a valuable source of evidence to set against the claims of
cultural Darwinists. It may be singers and scribes who generate the evidence that we find in
the Scots ballad text, but, if Parry’s ideas are to be usefully applied to that tradition, it is
ultimately against the critical role of the audience that such evidence must be tested.

151
This page intentionally left blank
 Estrea Aelion, Salonica Sephardic Tradition

and the Ballad of Imprisoned Virgil

Margaret Sleeman

In this paper I give a succinct portrait of Estrea Aelion, a remarkable singer originally from
Salica. I describe briefly my work with her in London, and indicate the interest of her ballad
repertoire. After this, I pick out for detailed discussion one of the ballads she sang,
Virgilios. The ballads of the Sephardic Jews have long been valued for their conservatism,
and rightly so. The tradition includes ballads which are now little known in the Iberian
Peninsula, or are no longer sung there at all, and Sephardic texts, together with those of
other isolated groups, can provide invaluable information about earlier stages of a ballad.
There is, though, another aspect, that of the accommodation of what was originally
Hispanic material to Sephardic culture, for the singers have made the ballads their own. I
aim to emphasise this aspect in my commentary. Before approaching my main subject,
however, it will be useful to fill in the background.
The Jews formed a small but significant group within the population of the Iberian
peninsula during the Middle Ages. Towards the end of that period, however, their situation
changed dramatically. In 1391, a series of pogroms in the towns of Castile and Aragon
caused panic, and many Jews embraced Christianity (though it has to be said that individual
conversions of these ‘New Christians’ or conversos varied in sincerity — the necessary
outward observance of Christianity often being a cloak for the continued and secret practice
of Judaism). Conversion did, however, open up new career possibilities in the Church and
in civil administration, and many conversos rose to positions of power in these areas. Their
new power was envied, and a combination of envy and suspicion about their religious
orthodoxy seem to have been key factors in the establishment of the Spanish Inquisition in
1478. And fears that the continued presence of considerable numbers of unbaptised Jews
might tempt the conversos to revert to Judaism led, eventually, to the expulsion of the Jews.
In 1492, Ferdinand and Isabella effectively decreed that all remaining Jews in Castile and
Aragon choose between baptism and expulsion. There was a further wave of conversions,
but many held out, and had to leave. The great majority fled to Portugal where they had a
brief respite, but in 1497, all Jews in Portugal were forced to leave or be baptised. For those
who left the Peninsula, favoured destinations were neighbouring Provence, Italy, and North
Africa. Expelled Jews also made their way to the Ottoman territories of the eastern
Mediterranean. It was in the towns of North Africa and the Ottoman Empire that they were
able to find greatest long-term security; they formed communities, were able to practise
their faith, and kept their language and important elements of popular culture.
Of the secret or crypto-Jews who had remained in the Peninsula, many managed to
leave over the following years and centuries, and either joined the Jewish communities
established earlier, or founded new communities in Northern Europe — in France,
The Singer and the Scribe

Germany, Holland, and England — where they were able to practice Judaism openly as
conditions allowed.
The settled communities of the Sephardim (descendants of Jews who lived in Spain and
Portugal before the Expulsion) have now largely disappeared. The Eastern communities
were affected by the Balkan Wars of 1912–13, the First World War, and in addition, in the
case of the Sephardic community in Salonica, the great fire of 1917, which destroyed large
parts of the city. Many Jews left, the United States and South America being popular
destinations. A group of almost 700 Eastern Sephardim from Salonica and Turkey came to
London to join an old established Sephardic community dating from the seventeenth
century (and by now thoroughly anglicised). Events over the following decades continued
to erode, and in some cases extinguish the Eastern Jewish communities. In the new Turkey,
Mustafa Kemal Atatürk’s modernisation projects made the country less attractive for Jews.
Nazi atrocities virtually wiped out the huge Jewish community in Salonica, others in
Greece, and those in Yugoslavia. The North African communities have now also dispersed.
From the middle of the twentieth century many Moroccan Jews have chosen to move to
Israel, the United States, Canada, France and elsewhere, continuing a desire to emigrate
which had begun earlier. 1
Much important ballad fieldwork was carried out in the old communities. The work of
the Spaniard Manuel Manrique de Lara was outstanding, conducted in all the major Eastern
centres in 1911, and in the Moroccan ones between 1915 and 1916. His meticulous
handwritten texts are housed in the Archivo Menéndez Pidal in Madrid, as are his
transcriptions of the tunes of some of these ballads. Inevitably, from the middle of the
twentieth century the focus has been on the new centres. Here the towering figures are the
American scholars Samuel G. Armistead, Joseph H. Silverman, and Israel J. Katz, who
have collected in the U.S.A., Morocco and Israel. I.J. Katz works on the music of this
collection.
Among the Sephardic communities which had not been investigated was the British one,
and during the 1980s I carried out fieldwork there. By far the most knowledgeable
informant I found was Estrea Aelion (born Matalón), originally from Salonica, who had
settled in London with her family many years before. Details of Mrs Aelion’s life are from
my recorded conversations with her, correspondence with her daughter and granddaughters,
and from a lengthy tape-recording which she made for her family at the time of her
hundredth birthday, a copy of which was very kindly given to me by her daughter. Estrea
Aelion was born in 1884 in Salonica (then part of the Ottoman Empire), to a family which
was apparently quite comfortably off. Her father Jacob Samuel Matalón was a jeweller who
specialised in dealing in pearls, and her mother Rahel Errera was the daughter of the owner
of a substantial shop which sold clothing and household goods.2 Estrea was the third of five

1
The break-up of the old communities and migration to new, usually non Spanish -speaking countries has dealt a
death blow to Judeo-Spanish. For a detailed account of the current situation, see Harris, Tracy K., Death of a
Language: The History of Judeo-Spanish (Newark: University of Delaware Press, 1994). The speech of
Moroccan Jews, however, was by the time of their resettlement in other parts of the world heavily influenced
by Andalusian Spanish, the result of the Spanish presence in Morocco.
2
The Jews lived together in certain districts. In 1883, the year before Estrea’s birth, the very substantial Jewish
population of Salonica was estimated at 50,000, while Turks numbered 25,000 and Greeks 20,000. See

154
 Margaret Sleeman

children, and the only girl. When she was fourteen, her mother decreed that she should
leave school and prepare for eventual married life. She went to classes in dressmaking. She
later said she had enjoyed sewing all her life, but also insisted that she had never stopped
reading: newspapers, and devotional books given to her by her father.
In 1905, Estrea married Saltiel Jehuda Aelion, the head of one of the departments in
grandfather Errera’s shop. In 1907 and 1910 their son León and daughter Irma were born.
After the First World War, it was decided that there was little future in Salonica, and the
family moved first to Istanbul, then to Paris in 1923 after Mustafa Kemal Atatürk came to
power in Turkey. Finally they moved to London where they had business interests. The
children, now grown up, moved first, Irma with her husband Maurice Frances. Mr and Mrs
Aelion arrived in 1936. Their business was an import and export business, a family concern.
The Aelion family were members of the Holland Park synagogue, opened in 1928 by the
newly-arrived Sephardim from Salonica and Turkey.
After her husband’s death, Mrs Aelion lived with her daughter Mrs Irma Frances in St
John’s Wood, in a very attractive and elegantly furnished flat overlooking Regent’s Park. It
was there that I first met Estrea in January of 1982. She was then 97. I visited her
intermittently until 1988, the year when she died, aged 104. Despite her age, Mrs Aelion
had a remarkable memory. She was a strong character, with a lively sense of humour, and
she very much enjoyed discussing earlier times in Salonica and singing her ballads and
other songs, which were clearly very important to her.
Estrea Aelion told me that she had learned her ballads from her mother, but she also
heard them from others in her family. When I asked her how she learned them, she simply
said (and I translate) ‘Everybody sang in my family, everybody sang, and I was small, and I
listened, and learned them. Then Irma heard them from me.’ The family’s living
arrangements (which were in line with local custom) must have reinforced what were
already close ties: Estrea’s family lived in the same building as her father’s brother, and her
maternal grandparents Errera had two apartments in their large house on the outskirts for
two sons and their families. As a girl, Estrea often stayed with one of these younger
families for a month at a time, since she was a great favourite of her aunt.
Ballads were almost entirely women’s songs, and were bound up with a way of life.
Judeo-Spanish was the vehicle, and that was the language of the home. Women in Salonica
were occupied with children and the house, while men concerned themselves with the
synagogue and their jobs or business interests. Ballads were used to lighten household
tasks, and were sung to the children as lullabies, or at bedtime if they were older. They
would be sung at parties and gatherings, such as the day, some time before a girl was
married, when her friends came to her house to help wash the wool for the mattresses
outside in the courtyard. Estrea described how the girl’s father would purchase a quantity of
sheep’s wool. A sunny day would be chosen, and the girls would sing all day long —
ballads and songs suited to the occasion. (She cited a song which contained the lines: Me
dišeron qu’eras novia, / y te vine a ayudar — They say you’ll be married, /and I’ve come to

Nehama, Joseph, Histoire des Israélites de Salonique, 7 vols (Salonica: Librairie Molho, 1935 [vol. I]; Paris
and Salonica: Librairie Durlacher and Librairie Molho, 1935–36 [vols II–IV]; Salonica: Fédération Séphardite
Mondiale, 1959 [vol. V]; Salonica: Communauté Israélite de Thessalonique, 1978 [vols VI–VII]), VII, 772.

155
The Singer and the Scribe

help you) Weddings and deaths were times when ballads reserved for the occasion were
sung. At weddings and other festive occasions semi -professional singers and tambourine
players (cantaderas and tañederas) would come to the house to sing ballads and other
songs. Until quite recent times, professional mourners (women) came to the house of the
deceased.
Mrs Aelion’s repertoire as collected by me comprised thirty-five ballads, plus songs
associated with dates in the Jewish calendar, numerous lyrics, nursery songs and sayings,
also topical songs which recorded some recent notable event. I worked with a checklist
(based on Attias’s Salonica collection), 3 but some ballads were sung unsolicited. All were
recorded on tape, with their tunes where Mrs Aelion remembered them, and with her
comments. The ballad repertoire is typical of others collected from that centre; there are
Carolingian ballads, two biblical ballads, and a large group of the so-called ‘novelesque’
ballads. Mrs Aelion also knew some which were ultimately of Greek origin, for a number of
Greek ballads had been adapted as Judeo-Spanish ballads, the product of bilingualism
among some singers. A few of Mrs Aelion’s ballads, El sueño de Doña Alda (Lady Alda’s
Dream), El paso del Mar Rojo (The Crossing of the Red Sea), El veneno de Moriana
(Moriana’s Poison), are considered very rare, in that few texts have been collected.4
Estrea Aelion’s texts are notable for their completeness, very unusual today when most
people are able to remember only a few of the commonest ballads, and these only in
fragmentary form. Mrs Aelion’s ballads are comparable in quality with those collected in
Salonica by Manuel Manrique de Lara in 1911 from people who were after, all her,
contemporaries. None of the repertoires collected by him is, however, as extensive. Her
ballads, with their tunes, are an invaluable document of Salonican tradition at the beginning
of the twentieth century.
There is little doubt that the fact that Estrea was able to use Judeo-Spanish on a daily
basis was a factor in her retention of her repertoire; that was the language she spoke with
her daughter Irma. What was surprising, however, was that Irma did not know her mother’s
ballads (despite her mother’s declaration, noted above, that she had sung them for her as a

3
Attias, Moshe, Romancero sefardí: romanzas y cantares populares en judeo-español, 2nd edn (Jerusalem:
Hebrew University, Ben-Zewi Institute, 1961); in future cited as: Attias.
4
A full account of Estrea Aelion’s repertoire is included in my monograph (in final stages of preparation), Stories
in Song: Judeo-Spanish Ballads Collected in Britain ; musical transcriptions by Judith R. Cohen. A selection
of Mrs Aelion’s ballads were played at the Colloquium: El robo de Elena (The Abduction of Helen of Troy),
La vuelta del marido (The Husband’s Return ), and Virgilios (Virgil).
The ballads sung by Estrea Aelion which are adaptations of Greek ballads are El pozo airón (from the
ballad known as The Haunted Well), La moza y el Huerco (from Charon and the Girl); La vuelta del marido is
also remarkably similar to a Greek ballad. For texts, see Argenti, Philip P. and Rose, H.J., The Folk-Lore of
Chios, 2 vols (Cambridge: Cambridge University Press, 1949), II, 718–19; 722–23; 778–81. The Greek
contribution to the eastern Sephardic repertoire is investigated in Armistead, Samuel G., and Silverman,
Joseph H., En torno al romancero sefardí (Hispanismo y balcanismo de la tradición judeo-española) (Madrid:
SMP, 1982), pp. 151-93.
Mrs Aelion knew only some words and phrases in Greek, but one of Rina Benmayor’s informants knew the
language quite well, and sang for her The Ballad of Susan, in Greek: see Benmayor, Rina, ‘A Greek
“Tragoúdi” in the Repertoire of a Judeo-Spanish Ballad Singer’, Hispanic Review, 46 (1978), 475–79.

156
 Margaret Sleeman

child). 5 Irma expressed amazement on hearing them. She did remark, however, that she had
sometimes heard her mother quietly humming to herself.
Several months after I started to work with her, Mrs Aelion sent some material to the
Proyekto Folklor started by Moshe Shaul of the Sephardic branch of the Israeli
Broadcasting Service, Kol Israel. She sent him two tapes, containing fourteen ballads and
some other songs. I have copies of the tapes, and it is interesting to compare them with
versions I collected myself. Some of the later versions are fuller, as Mrs Aelion recalled
additional lines, but my recordings are, I believe, ‘warmer’ than those recorded ‘for the
record’, since they were sung for an audience.
In a short essay such as this it is impossible to describe adequately the many ways in
which the ballads have been adapted by Sephardic singers from Eastern communities, for
the process pervades every level. Some instances drawn from Estrea Aelion’s repertoire
can, however, give an idea of the processes involved. On the verbal level, her texts (like
those of other singers of course) include words borrowed from Turkish and the Balkan
languages, words current in the Judeo-Spanish they spoke. Hebrew words also appear.
There are also local touches. In La mala suegra (The Evil Mother-in-Law), the false older
woman promises to look after her son carefully during her daughter-in-law’s absence by
cooking fine meals. In Peninsular versions, fine wine and fine bread are on offer, but
Eastern Sephardic versions promise, naturally enough, the favourite dishes of that group:
Mrs Aelion’s version has pešcado con agra *e, fish with a sharp sauce, others have pigeons
or savoury meatballs. There is also the question of the treatment of the frequent Christian
references in Hispanic ballads, studied most recently by Samuel Armistead and Joseph
Silverman in an article ‘El substrato cristiano del romancero sefardí’.6 They note
differences in the way these are dealt with which include retention (rare in the East), term
retained, but with a different sense, term replaced by one acceptable to the Jews: bautizar
(baptise), for example, is replaced by nombrar (name) in Eastern versions of Las hermanas
reina y cautiva (The Two Sisters: Queen and Captive).
The treatment of the key term ‘misa’ (Mass) in Eastern texts is interesting. In Hispanic
ballads, things happen when the protagonist is going to church to hear Mass, or is attending
Mass. The word indeed is used in Eastern Sephardic ballads — in the forms misa, milsa,
mesa — but in the examples I have encountered, with different senses. As discussed below
with reference to Virgilios, there is a Sarajevo text in which the following suggestion is
made: ‘digamos presto la misa’, suggesting that singers thought that the ‘misa’ was some
sort of prayer or service which could be rattled off. In some ballads, the context makes it
clear that the sense of misa is not that of ‘Mass’, nor even ‘prayer’, but of the building
where it is said, ‘church’. Eastern versions of La bella en misa (The Beauty in Church)
confirm this. Mrs Aelion’s version, which is typical, opens:
Tres damas van a la mesa para hazerż la orasión.
Three ladies go to the mesa to pray.

5
Irma Frances did, however, know nursery rhymes and topical songs, learned from her mother. It is probably
relevant that the family left Salonica when she was quite young.
6
In Armistead and Silverman, En torno al romancero sefardí, pp. 127–48.

157
The Singer and the Scribe
Here, three ladies go to la mesa to say their prayers, and a few lines further on, la mesa is lit
up by the exceptional beauty of the protagonist; the sense of mesa is clearly ‘church’. While
the priest who is dazzled by the girl’s beauty becomes, in these Salonica texts, the papazico,
*
the Orthodox priest familiar to the singers. Mesa has the same sense in Landarico. Mrs
Aelion’s text of Landarico opens:
E rey que mucho madruga para la mesa se iva.
Topó la mesa serrada, par’ ande la reina andava.

[And the King who rises very early sets off for the mesa
He found the mesa closed, he went to meet the Queen.]

Here, la mesa, which is found to be closed, is clearly a building, a church or chapel. Other
Eastern versions of Landarico have similar formulations. The word is used in this same
sense in La choza del desesperado (The Hopeless Lover’s Hermitage) where the wretchedly
unhappy man says he will build himself a misa, torre, or (Mrs Aelion) una ermita. I suggest
that in Mrs Aelion’s version of Virgilios given below, misa has the sense ‘church’.
Most interesting of all are cases where the inherited Hispanic material has been
completely recast in line with Sephardic sentiment and values. Such a case is the ballad
known as La tormenta calmada (The Quieted Storm), which derives, as Armistead and
Silverman have discovered, from a sixteenth-century Spanish erudite ballad which survives
in oral tradition only among the Eastern Sephardim. 7 Central to the Spanish ballad is a
shepherd’s prayer to ‘the Heavens’ that the storm cease, and his delight when it does. As
Armistead and Silverman observe, the Sephardic ballad transforms the original, which has
‘a minimum of re ligious overtones’ into a ballad which ‘has come to celebrate in the most
fervent terms God’s miraculous response to the prayers of a true believer’ (p. 140). The
relevant verses, in the version sung by Estrea Aelion, are as follows:
Bendicho el Siñor del mundo, que tantas maravías mos haz e ?.
Mo las haga agora y siempre y de prisa y que non de tadre.
El quita naves del golfo, a la parida cuando pare.

[Blessed the Lord of the world, who performs so many miracles for us
He perpetually performs them, urgently not tardily
He saves ships from the abyss in all hazards.]

There is an element of such a transformation in Salonican versions of Virgilios.

The Ballad of Imprisoned Virgil: ‘Virgilios’

A courtier (named Virgilios or Vergilios in early texts) makes love to a maiden at court,
generally the king’s niece, an action injurious to the king himself. The king imprisons him
and forgets all about him. One day he remembers him (or is reminded). He goes to see him
in prison and questions him about his situation. The prisoner’s exemplary answers so
impress the king that he frees him. In many versions, marriage to the lady follows.

7
Armistead, Samuel G., and Joseph H. Silverman, The Judeo-Spanish Ballad Chapbooks of Yacob Abraham Yoná
(Berkeley, Los Angeles, London: University of California Press, 1971), pp. 134–44.

158
 Margaret Sleeman

The above summary gives the story told in the majority of versions. It is a story which
may be linked to the medieval legends of Virgil (the Roman poet) as a lover who gets
himself into scrapes.
Versions of Virgilios were printed in pliegos sueltos in the early sixteenth century. One
version of the ballad was included in the Cancionero de romances ‘sin año’ of ca. 1548,
and also in later collections. Another, rather different version, was printed in a pliego
suelto, formerly in the collection of the Duque de T’Serclaes, now lost, of which a
handwritten copy is housed in the Archivo Menéndez Pidal; this version is titled Romance
nuevamente trobado del grande poeta Vergilius por muy gentil estilo [Newly found ballad
about the great poet Virgil in very noble style].
The ballad has also been collected from the modern oral tradition. Judging from the
numbers of texts collected, Virgilios seems to have been well known among the Moroccan
Sephardim and, in the Eastern Mediterranean, among Jews from Sarajevo and Salonica. It
appears to have been somewhat less well known in the other Eastern centres.8 In the Iberian
Peninsula, by contrast, the ballad is little known today. A mere handful of versions have
been collected from the Spanish provinces of Palencia, Zamora, Orense, and over the
border in Portugal, almost all of them at the beginning of the twentieth century. A recent,
intensive collecting campaign by Maximiano Trapero in the small island of El Hierro, part
of the Canary Islands archipelago, has, however, revealed that the ballad survives there with
some vigour. 9
Trapero has followed up his important discovery with a short monograph dedicated to
the ballad.10 A particularly valuable feature of this book is an appendix which contains a
selection of published texts as well as a number of previously unedited texts from the
Archivo Menéndez Pidal, the latter from the Peninsular and the Sephardic traditions.11 The

8
Armistead and Silverman, The Judeo-Spanish Ballad, pp. 134- 44.
9
His texts, recorded in 1982, and the very first of Virgilios to be collected in the Canaries, are published in his
Romancero de la Isla del Hierro (Madrid: SMP and Cabildo insular del Hierro, 1985).
10
El romance de ‘Virgilios’ en la tradición canaria e hispánica (Las Palmas: El Museo Canario, 1992).
11
For examples from Morocco see: Alvar, Manuel, ‘Los romances de La bella [en] misa y de Virgilios en
Marruecos’, Archivum (Oviedo), 4 (1954), 264–76, at pp. 266–70; Armistead, Samuel G. and Joseph H.
Silverman, Romances judeo-españoles de Tánger recogidos por Zarita Nahón (Madrid: CSMP, 1977), nos
14A–C; Bénichou, Paul, Romancero judeo-español de Marruecos (Madrid: Castalia, 1968), p. 99; Larrea
Palacín, Arcadio de, Romances de Tetuán, 2 vols (Madrid: CSIC, 1952), nos 51–53; Martínez Ruiz, J., ‘Poesía
sefardí de carácter tradicional (Alcazarquivir)’, Archivum (Oviedo), 13 (1963), 79–215, no. 49; Weich-
Shahak, Susana, with the collaboration of Paloma Díaz -Mas, Romancero sefardí de Marruecos: Antología de
tradición oral (Madrid: Alpuerto, 1997), no. 18 a–d. For texts from the Eastern tradition, see Armistead,
Samuel G., and Joseph H. Silverman, with the collaboration of Biljana Šljivi- Šimši, Judeo-Spanish Ballads
from Bosnia (Philadelphia: University of Pennsylvania Press, 1971), A1, B1, C3. This collection is referred to
in future as Bosnia ; Armistead, Samuel G., and Joseph H. Silverman, Judeo-Spanish Ballads from New York
Collected by Maír José Benardete (Berkeley, Los Angeles, London: University of California Press, 1981), no.
18. Cited in future as Benardete; Attias, 7; Danon, Abraham, ‘Recueil de romances judéo-espagnoles chantées
en Turquie’, Revue des Études Juives, 32 (1896), 102–23, 263–75, at p. 268; González-Llubera, Ignacio,
‘Three Jewish Spanish Ballads in MS. British Museum Add. 26967’, Medium Aevum, 7:1 (1938), 15–28, at pp.
21–23; Hemsi, Alberto, Cancionero sefardí, edited by Edwin Seroussi in collaboration with Paloma Díaz-

159
The Singer and the Scribe

selection of Eastern Sephardic texts is, however, restricted to those from Sarajevo, Belgrade
and Jerusalem, and Trapero’s analysis of the Eastern tradition follows suit. The quite
substantial number from Salonica is ignored.
In this paper I shall attempt to rectify the omission, all the more important since Mrs
Aelion’s version and two unedited texts in the Archive exemplify what amounts to a
different story, and one which reflects specifically Jewish preoccupations. These Salonican
versions have to be considered in the context of the tradition of the ballad as a whole, and I
start with the version printed in the Cancionero de romances ‘sin año’. It opens with the
imprisonment. Virgil is imprisoned for ‘violating’ (‘forzar’) a young lady, Isabel, in the
king’s palace. The king forgets all about him, until one day at Mass, he suddenly
remembers him. He checks with his men, and is reminded that Virgil is in jail. He says he
will go to see Virgil immediately after lunch, but the queen (for a reason which is not
explained), says that she will not sit down to eat without him. They go to the prison, and the
king asks Virgil what he is doing. He replies that he is combing his hair and his beard,
which started to grow in prison and will go white there. The king tells him that he has three
years still to serve, to which Virgil replies that he would gladly stay for the rest of his life,
should the king command it. As a result of his patience, Virgil is released and pardoned,
and later rewarded with marriage to Isabel. An important feature of this version is Virgil’s
great popularity among those at court (see verses 7, 11, 22– 23 below), most particularly
with Isabel herself, who is described as thrilled at his release. Thus, Virgil’s action, rather
than being one characterised only by violence, as suggested by the use of forzar (v. 3), must
have been one in which love was present, i.e. ‘sleep with’, ‘make love to and deflower’
(Isabel was, after all, a virgin, a doncella). The king’s concern for his prisoner also supports
this milder view. As it happens, another ballad (again contained in the Cancionero de
romances ‘sin año’), which opens ‘Medianoche era por filo …’, has interesting similarities
with Virgilios. Count Claros makes love to the Princess Claraniña (whom he had long
adored) after an encounter in which compliments and highly suggestive comments are
exchanged. Count Claros is arrested and condemned to death, the girl’s father referring to
his crime as ‘forzar’. Count Claros is eventually saved by the devoted Claraniña, who
makes a persuasive case to her father the king. There is no way in which this passionate and
loving relationship could be described as a violation12 and the king’s use of forzar is all the
more odd since the huntsman who informed him described the couple as kissing and
embracing as they made love. Should forzar be construed as ‘deflower’? In both Conde
Claros and Virgilios, the use of forzar is problematic, and it is not surprising that in
Virgilios, singers eventually substituted another verb.

Mas, José Manuel Pedrosa and Elena Romero (Jerusalem: The Jewish Music Research Centre, The Hebrew
University of Jerusalem, 1995), no. 17. Referred to as Hemsi; Molho, Michael, Literatura sefardita de Oriente
(Madrid and Barcelona: CSIC, 1960), p. 73. Bosnia A1 and González-Llubera’s text are of eighteenth-century
date. Unedited texts from the Archivo Menéndez Pidal are listed in the following work: Armistead, Samuel
G., with the collaboration of Selma Margaretten, Paloma Montero, Ana Valenciano, and with musical
transcriptions edited by Israel J. Katz, El romancero judeo-español en el Archivo Menéndez Pidal (Catálogo-
índice de romances y canciones), 3 vols (Madrid: CSMP, 1978). Cited in future as CMP. I am very grateful to
Professor Diego Catalán for allowing me to consult these texts, and the copy of the T’Serclaes text.
12
This ballad is edited and provided with extensive notes in Roger Wright, Spanish Ballads, second revised
impression (Warminster: Aris & Phillips, 1988), pp. 12–21.

160
 Margaret Sleeman

I quote the text of Virgilios,13 followed by a literal English translation (this and all
subsequent translations are my own).
Mandó el rey prender Virgilios The king ordered his men to seize Virgil
y a buen recaudo poner and to imprison him securely
2 por una traición que hizo 2 for a crime he had committed
en los palacios del rey, in the king’s palace,
porque forzó una doncella ‘violating’ a maiden
llamada doña Isabel. called doña Isabel.
4 Siete años lo tuvo preso 4 He held him prisoner for seven years,
sin que se acordase d’él forgetting him entirely.
y un domingo estando en misa One Sunday, while at Mass,
mientes se le vino d’él. Virgil came to mind.
6 — Mis caballeros, Virgilios, 6 — my Knights,
¿qué se había hecho d’él?— what has become of Virgil?
Allí habló un caballero Then a knight spoke up,
que a Virgilios quiere bien: one who loved Virgil well.
8 — Preso lo tiene tu alteza 8 — Your Majesty holds him prisoner,
y en tus cárceles lo tien. he is in your jail.
— ¡ Vía comer, mis caballeros; — Hurry, let us eat, knights;
caballeros, vía comer! knights, let us hurry and eat.
10 Después que hayamos comido 10 After we have eaten,
a Virgilios vamos ver. — we will go and see Virgil.
Allí hablara la reina: Then the queen spoke up:
— Yo no comeré sin él.— — I won’t eat without him.
12 A las cárceles se van 12 They went to the prison
adonde Virgilios es. where Virgil was being kept.
— ¿Qué hacéis aquí, Virgilios? — What are you doing, Virgil?
Virgilios, ¿aquí qué hacéis? Virgil, what are you doing?
14 — Señor, peino mis cabellos 14 — My lord, I am combing my hair
y las mis barbas también; and my beard as well.
aquí me fueron nacidas, My beard began to grow here
aquí me han de encanecer and here it will turn white,
16 que hoy se cumplen siete años 16 for it is seven years today
que me mandaste prender. since you imprisoned me.
— Calles, calles, tú, Virgilios, — Be quiet, be quiet, Virgil,
que tres faltan para diez. you have three years to go before ten are up.
18 — Señor, si manda tu alteza, 18 — My lord, if Your Majesty orders it,
toda mi vida estaré. I will stay here for the rest of my life.
— Virgilios, por tu paciencia — Virgil, because of your patience
conmigo irás a comer. you will come and dine with me.
20 — Rotos tengo mis vestidos, 20 — My clothes are in shreds,
no estoy para parecer. I’m not fit to appear.
— Que te los daré, Virgilios, — I shall give you clothes, I shall give orders

13
From Cancionero de romances ‘sin año’, edited by Paloma Díaz-Mas, in Romancero (Barcelona: Crítica, 1994),
pp. 346-48.

161
The Singer and the Scribe
yo dártelos mandaré. – that you are provided with clothes.
22 Plugo a los caballeros 22 The knights were glad,
y a las doncellas también; the maidens as well,
mucho más plugo a una dueña but the lady who was best pleased
llamada doña Isabel. was doña Isabel.
24 Ya llaman al arzobispo, 24 The archbishop was summoned,
ya la desposan con él; and she was married to Virgil.
tomárala por la mano He took her by the hand,
y llévasela a un vergel. and led her into a garden.

It is very interesting to compare the other sixteenth-century text (T’Serclaes) with this
one. Essentially the story told is the same as in the Cancionero de romances text. Virgil’s
crime is again described as forzar, and his punishment is again severe—a ten-year period of
imprisonment. Again, the king suddenly remembers about his prisoner, rushes to see him in
prison, is impressed by (in this version) Virgil’s humility in stating his unconditional
willingness to serve the remainder of his sentence, and again, the king lets him go early, and
also permits him to marry the woman he knows he loves. There are, however, some
significant differences. After v. 3 of the Cancionero de romances text, T’Serclaes has an
additional verse which gives detail of the lady’s parentage; she is the daughter of an
archbishop and the niece of the king: “hija es de un Arçobispo / sobrina era del buen rey.”
Second, the verse in which the queen speaks up is lacking in T’Serclaes. Finally, the
T’Serclaes text describes the prisoner’s suffering in much more harrowing terms than the
other text. In T’Serclaes, the king asks his prisoner what his life in the dungeon (mazmorra )
is like, and Virgil replies that there is no need to ask, for the results of the imprisonment can
be seen in his face: when he entered prison, his beard was beginning to show (i.e. he was a
youth), but now, for his sins, it was already going white—a formulation found too in textts
in the modern oral tradition and one which, as Paloma Díaz -Mas noted, Romancero, p. 347,
is a little different from the formulation in the text she edited, in which Virgil declares that
his beard will go white in prison (v. 15) due to the length of his incarceration. The
concluding verses of the two versions also vary. In T’Serclaes, Virgil is brought forth from
prison weighed down with handcuffs, leg irons, and chains, a pitiful sight, we are told (vv.
47– 52):
Ya lo sacan a Vergilios / de la cárcel segun veis
con esposas en las manos / y unos grillos a los pies
una cadena en su cuello / que dolor es si lo veis.

The final verses of T’Serclaes (vv. 55–56) explicitly praise the magnanimity of kings.
Like the early texts, those in the modern oral tradition have assonance in –é. Sephardic
texts from Morocco and Sarajevo and Salonica contain a verse which describes the lady’s
relationship to the king, as is found in the T’Serclaes version, and texts from Sarajevo and
Salonica have the verse in which the queen speaks up for the prisoner, a verse found in the
Cancionero de romances text and not in the other early one. There is one feature, however,
which is found only in the modern tradition: the woman dressed in mourning whose
presence serves to remind the king about his prisoner. This feature, shared by Sephardic,

162
 Margaret Sleeman

Peninsular, and Canarian versions, suggests that there was, at an early stage, a slightly
different version of the ballad which contained this detail. 14
The Moroccan version given below, from Tetuán tradition, is typical of Moroccan texts,
though the usual forzar becomes in this text, euphemistically, faltar (see v. 3). In Moroccan
tradition it is the sight of Isabel, dressed in mourning, which prompts the king to ask about
Vergico, as he is known here. Again, the king rushes to see the prisoner. In Morocco, as the
ensuing dialogue shows, it is Vergico’s devotion to the lady which is to the fore, for when
the king explicitly asks him how much he loves her, Vergico makes his feelings clear: he
declares that he would stay in prison double the time to be with his lady. For his steadfast
devotion to his lady, the prisoner is released and married to her. This particular Moroccan
version of the story, in which the love and devotion of both parties is so evident, again
suggests that the crime was that of impetuous passion rather than a violent assault on an
unwilling victim.
Preso llevan a Vergico, Vergico has been taken prisoner,
el rey le mandó prender, the king ordered that he be arrested,
2 por una traición que ha hecho for a crime he committed
en los palacios del rey, in the king’s palace,
de faltar a una donzella, for ‘wronging’ a maiden
la cual se llama Isabel; named Isabel,
4 hija era del obispo, the bishop’s daughter,
sobrina del señor rey. the king’s niece.
Un día indo el rey a missa, One day, as the king was going to Mass,
encontrara una mujer he met a woman
6 toda vestida de luto, completely dressed in mourning,
ella y sus damas también. she and her ladies too.
Preguntó el rey a su alcalde The king asked an official
que quién era essa mujer: who that woman was.
8 — Vuestra sobrina, mi señor, — Your niece, my lord,
vuestra sobrina Isabel. your niece Isabel.
— ¿Por quién va vestida en luto, — For whom is she wearing mourning.
ella y sus damas también? she and her ladies too?
10 Por Vergico, mi señor rey, — For Vergico, my lord the king,
que en vuestras prisiones es. he is in your prison.
— Pronto, pronto, mis criados, — Quickly, quickly, servants,
poní mesas a comer; prepare the tables for the meal;
12 mientras las mesas se aprontan, while the tables are being set
a Vergico iré yo a ver. I shall go and see Vergico.
— En buena hora estés, Vergico. — Greetings, Vergico.
— Bien vengáis, mi señor rey. — I greet you, my lord the king.
14 — ¿O qué años o qué tiempos — How many years, how much time
que en las mis prisiones es? have you been in my prison?

14
This feature was noted by Menéndez Pidal, Ramón, Romancero hispánico (hispano-portugués, americano y
sefardi, 2 vols (Madrid: Espasa-Calpe, 1953), I, 348. He argues that this version of the ballad, circulating in
oral tradition in the fifteenth century, but not printed in a pliego suelto, must have been the most widespread.

163
The Singer and the Scribe
— Siete años ha, mi señor rey, — Seven years now, my lord,
siete años y más un mes. seven years and a month.
16 Cuando yo entrí en ellas, When I entered prison,
no empeçaba a embarbecer, I did not have a beard,
y ahora por mis pecados and now, for my sins,
ya me empecí a encanecer. I am going white.
18 — ¿Qué darías tú, Vergico, — What would you give, Vergico,
por hablar con Isabel? to be able to speak to Isabel?
— La vida de las prisiones — I would be prepared
yo la doblaré otra vez. to double my time in prison.
20 — Pronto, pronto, mis criados, — Quickly, quickly, servants,
sacay Vergico a comer. free Vergico from prison and let him eat with us.
Y a otro día a la mañana, Next morning,
Vergico con Isabel. Vergico and Isabel were married.
(Bénichou, Romancero , p. 99)

Of the available Eastern texts, I turn first to the Sarajevo group (studied by Trapero, El
romance de ‘Virgilios’, pp. 37–43). In this group the verb used to describe the man’s crime
is not forzar but amar (make love to). The woman, again the king’s niece, is called Zadé,
while the man’s name (as is general in Eastern texts, as well as those from the Iberian
Peninsula and El Hierro) is prefixed by the courtesy title Don (i.e. Don Virgile, Bosnia A1,
an eighteenth-century text), the title sometimes being incorporated in Eastern texts giving
rise to forms such as Doverdjeli (Bosnia C3), Doverchile (CMP F8.3), an indication that the
sense of the title was not understood.
The last-mentioned version, CMP F8.3, from a manuscript obtained by Manuel
Manrique de Lara in Sarajevo in 1911, and edited in Trapero, El romance de ‘Virgilios’, p.
96, opens as follows:
Traición a Doverchile por el palacio del rey,
por amar una doncella que se llamaba Zadé,
ni más alta ni más baja, sobrina era del rey.
Tanto era el mal que hacía, fue en oídos del rey (vv. 1–4)

[Doverchile has committed a crime in the palace of the king

by making love to a maiden called Zadé
neither higher nor lower born, she was the king’s niece.
The king soon heard about the wrong he had done.]

The king duly imprisons the delinquent. A striking feature of this group of texts is the stress
given both to the severity of the punishment and to the neglect of the prisoner.15 The
eighteenth-century text, Bosnia A1, includes the verse: “Pasa tyenpo i vyene tyenpo i
ninguno se akodra dél” (v. 6). Roughly translated as ‘days come and go, and nobody
remembers him’, variants of this verse with its forlorn message are found quite widely in
Eastern texts. In others of this same group, the terms of the sentence are made plain at the
outset, and they are harsh. In the Sarajevo text cited above, CMP F8.3, the king’s orders are

15
Trapero also points out (El romance de ‘Virgilios’, pp. 38–39), that here the king is more closely involved in the
imprisonment, since he keeps the keys himself.

164
 Margaret Sleeman

as follows: “–diez años esté en la cárcel, ninguno non lo vaya á ver” (v. 7). The order not to
visit is, however, ignored by Doverchile’s sorrowing mother who goes to see him every day
and brings him food:
Su madre la desdichada cada día lo va á ver.
Debajo del brazo izquierdo un pan le lleva á comer (vv. 8–9)

That Doverchile is also forgotten is clear from the fact that, again, the king is reminded of
his existence by the woman in black, in these versions his mother. The king urges that
‘Mass’ be hurried,16 that the meal be eaten, in order to visit the prisoner quickly.
After mutual greetings, Doverchile tells the king that he submits to the full sentence,
seven years of which he has served, and he describes how his beard is already going white.
The king is impressed by this show of loyalty, and orders that Doverchile be freed, bathed
and united with Zadé. Marriage is the usual conclusion in this group, but there is one
version, CMP F8.6 (see Trapero, El romance de ‘Virgilios’, pp. 93–94) in which, after his
release, the prisoner drops dead; the victim of his ill-treatment, perhaps, in the estimate of
some singers.
Counting both published texts, the unedited versions in the Madrid Archive, and Mrs
Aelion’s new text, I have seen thirteen texts from Salonica. Whereas the Sarajevo versions
are fairly uniform, the same can not be said of the Salonican ones. The versions from this
second centre can, I believe, be placed in three groups. The first, comprising Attias’s text,
no. 7, CMP F8.21, and the one written down by the fourteen-year-old Elisa de Bottón in a
MS collection acquired by Manuel Manrique de Lara (CMP F8.19), resemble the Sarajevo
texts quite closely, albeit with features which distinguish them. The most important is that
these, and the other Salonica versions I have seen do not end with marriage.
The second grouping (which I call transitional) consists of a number of versions,
including Benardete 18, CMP F8.16, 17, 18, which appear at first glance to be somewhat
disordered, but which reveal the workings of a curious logic. Finally, I link together three
very reduced versions, Mrs Aelion’s, and F8.20 and 22 from the Archivo Menéndez Pidal,
in which I find the tensions of the transitional group resolved in a way which radically
recasts the story. Further, over and above these divisions, the Salonican texts show a very
interesting specific adaptation to the Sephardic milieu: the addition of verses which contain
an invocation to God, and which also constitute a commentary on events.
Of the first group, both Attias and the shorter Bottón text give a full version of the
ballad story. They open similarly; I quote from Attias:
Traición de Don Argilis / por el palacio del rey,
por amar una doncella, / por amarla y bien quierer (vv. 1–4)

When he discovers the misdemeanour with his niece (here the milder amar y bien quierer)
the king has the young man arrested and put in isolation in a deep dungeon:
El buen rey cuando lo supo, / en cárcel lo fué a meter,
en unas prisiones fondas, / que dinguno lo va a ver (vv. 9–12)

16
Trapero, El romance de ‘Virgilios’, pp. 40–41, discusses the curious ‘Digamos presto la misa’, literally, ‘let’s
say Mass fast’, which suggests that singers construed it as some sort of prayer which they could rush through.

165
The Singer and the Scribe
The Bottón text also suggests harsh treatment; the man is thrown into prison and chains.
Attias’s text has verses lacking in the other, verses which explicitly mention and then
comment on Don Argilis’s abandonment:
Tiempos van y tiempos vienen / dinguno s’acodra d’él.
(El Dió s’acodre de todo el mundo / y de mosotros también;
la hora que lo llamamos, / que mos aresponda en bien) vv. 17–22

The first two verses have already been encountered in a Sarajevo version, the third and
fourth, which are distinctive of a number of Salonica texts, represent a pious wish that God
not forget his people, a significant addition which expresses Sephardic religious sentiment.
The final verses elaborate on the previous two, and in my corpus are found only in Attias.
As in the Sarajevo texts, in both Attias and Bottón the king is reminded about the
prisoner by the presence of his mother, dressed in black. In Attias, the king questions Don
Argilis, and is told that his hair is going white, that he has been there for fifteen years, but
would stay until sixteen were up (dieciséis satisfies the assonance requirement) if the king
wished. In Attias, the king orders Don Argilis to be released, dressed richly, wear the king’s
crown, and that people should acknowledge his (the king’s) magnanimity. The shorter
Bottón text has a very similar prison dialogue, but ends abruptly with the prisoner’s offer to
stay up to sixteen years in prison.
The texts which I have called transitional have some unusual features. In most of these
texts, the woman loved in the palace is the king’s niece, in one she is just a señora. Again,
as in Attias, the verbs used are often amar y bien quierer, which express true love and
affection. In only one version in this group, F8.17, is the courtier explicitly sent to prison
for loving the woman. In the remainder, we discover later that the man is indeed in prison,
for the king is reminded of him by the presence of a woman in black — in all the versions
in this subgroup she is his wife. A number of puzzles present themselves. Is the wife the
same woman the courtier had initially dearly loved (and subsequently married)? Why then
was the courtier put in prison? Was the imprisonment some unexplained whim of the king?
Except for F8.17, where the prisoner is released, all end on a gloomy note with the prisoner
lamenting his fate. The suggestion of the arbitrary nature of the imprisonment is
strengthened by the final group, which includes Mrs Aelion’s version, to which I now turn.
Estrea Aelion’s sung version of Virgilios, which she told me she learned from her
mother, was recorded at her home in London on 7th February, 1982. It is as follows:
Un día, el rey estando’n la mesa, One day, when the king was in church,
vido venir una muzer he saw a woman approaching,
2 entera vestida de preto completely dressed in black
de la cavesa fin a los pies. from her head right down to her feet.
Demandó el rey a su gente The king asked his people
y quién er’ aquea muzer. who that woman was.
4 — Muzer era de Undergile — It is the wife of Undergile
qu’en la cársel lo tenéš. whom you hold in your prison.
Estava metido en cársel, He was in prison,
y en las priziones del rey. in the king’s jail.
6 Pasan días y vienen noches Days and nights come and go,
y el rey no se acodra d’él. and the king never remembers him.

166
 Margaret Sleeman

Se acodre el Dio de mozż otros, May God remember us,

y que mos quiera mucho bien. and may things go well for us.

I also give the short handwritten text CMP F8.22, which is housed in the Archivo
Menéndez Pidal. Written on a blank sheet of paper in a bold, sloping hand, it is initialled at
the end; the initials, which are not clear, seem to be J.G.C. Armistead tentatively assigns
this text to Salonica tradition, and suggests that it may have been obtained by Manuel
Manrique de Lara when he collected in the town in 1911. CMP F8.22 is as follows:
Un día estando en la milsa / vide [sic] venir una mujer,
2 vestida entera de preto / de la cavessa asta los pies.
Demandó el rei a su gente / ken era akea mujer.
4 – Mujer era de un gran vergiles / ke aki en preso lo tenech.
Días van i días vienen / y el rei non se acodra de él.
17
6 S’acodre el Dio de todo el mundo / i de nosotros también.

The two versions are very similar but for the awkward first verse of F8.22. A third Archive
text, F8.20, again a handwritten text (but with no indication of the transcriber), attributed to
Salonica tradition by Armistead, is also very like them, but lacks the verse containing the
expression of hope that God will remember his people. In view of their similarity to the
Aelion text there seems little doubt that the two Archive texts are indeed from Salonica.
None of these versions contains the usual reference to a crime, and the explicit
punishment of imprisonment for it. All begin with the king’s sighting of the woman in
black (again identified as the prisoner’s wife). Since I elicited Mrs Aelion’s version by
quoting the opening of Attias’s version (which does contain the usual opening) and by
giving a plot-summary, I am sure that what she sang was, for her, the complete ballad; she
normally indicated if she had forgotten part of a ballad. This version of the ballad,
exemplified by these three texts, tells, I believe, a radically different story. An individual,
Undergile (Aelion), 18 un gran vergiles (F8.22) un vergile (F8.20) is in prison, we know not
why, nor with what justification. His situation is lamented by his devoted wife, dressed
entirely in black (Aelion and F8.22), completely covered in mourning clothes (F8.20):
“Todo cuvierta de luito [sic] de la cavesa asta los pies” (v. 2). The king’s question about
this woman in black elicits the reply that she is the wife of the man he holds in prison.
Having been given this information, the king then consigns him to oblivion: days come and
go, and the king does not remember him. Whereas in most versions of the ballad the king’s
forgetfulness had, in narrative terms, brought about the testing of the prisoner’s stoicism,
loyalty, or love, the display of which brought about his release, in this little group of
Salonican versions, the king’s forgetfulness, which follows the reference to the man’s
situation rather than preceding it, has an entirely different function, for it amounts to total
abandonment.

17
CMP gives acodra , but the subjunctive form acodre is clear in the original transcription.
18
Establishing word boundaries can sometimes be a problem when transcribing a ballad. Since I wanted to know
whether Mrs Aelion analysed ‘Undergile’ as one word, or as an indefinite article + noun, I asked her to write it
down. The first syllable of the man’s name recalls the title Don, and assimilation accounts for the change in
the following consonant. Clearly we are now far from the Virgil of the early versions.

167
The Singer and the Scribe

In this version of the story, I would argue, the roles are reversed; the prisoner and his
faithful wife are victims of the king’s utter neglect, which now surely becomes the crime.
The final verse expresses the singers’ fervent hope that God never abandon them:
Se acodre el Dio de mozżotros, y que mos quiera mucho bien (Aelion, v. 7)
S’acodre el dio de todo el mundo / i de nosotros también (F8.22, vv. 11–12)

The underlying concept is surely that of God’s remembrance of his covenant with his
people, and the presence of the verse at the end of the ballad focuses on the Sephardic Jews’
trust in God. Further, I believe that the irresponsibility and cruelty of an earthly ruler, given
such prominence here, reflected a real concern; Mrs Aelion did not make the point when
she sang her version, but other diaspora Jews have commented to me that their group has
acute political antennae.19 In this case, it could be argued that these versions illustrate the
opposite of the common process of novelización, where a ballad on a historical theme
develops, in oral tradition, the personal, often amorous aspect, and that here, an amorous
ballad has become one with a sombre political message.

19
There is in fact evidence that, at a time of weak central control in the Ottoman Empire, Salonica Jews did suffer
at the hands of corrupt local officials. Nehama, Histoire des Israélites, VI, 37-38, describes the activities of a
late seventeenth-century governor who made false accusations against Jews, and notes too that corrupt
officials could imprison people who displeased them.

168
 Spanish Ballads in a Changing World

Roger Wright

At the Edinburgh colloquium ‘The Singer and the Scribe’, the purpose of this paper was to
demonstrate that the Spanish ballad is and has been a performance genre, better appreciated
when heard than when being read off a page. It may seem somewhat curious, then, to find a
version printed in this volume: readers need to bear in mind that several of the texts here
printed were presented to the colloquium on tape in their performed state, with the text and
English translations circulated to the audience on paper.
For an initial example of the Spanish ballad genre as it exists today, we can consider a
ballad recorded on 24th October 1977 in Moshav Mata, Israel. The singer is seventy-one
year old Rahel Gabai, born in Tetuán in 1906, who immigrated into Israel from Morocco in
1956. Her family is descended from one of the Spanish-speaking Sephardic Jewish
communities that were expelled from Spain in and after 1492, and eventually settled all
round the shores of the Mediterranean. The tale, Gerineldos, is a well-known one of ancient
origin, and the most widely attested of all the Spanish ballad stories in the modern world;
several hundred versions have been collected by field workers over the last century. The
text of this version was printed by Alexander et al. (1994:125-26, with the addition of two
extraneous lines) as their reference no. Y 2121(1); 1 the translation is mine, and intended to
be merely literal:
— Girineldo, Girineldo, ‘Girineldo, Girineldo,
mi caballero polido, My fine knight,
¿quién te me diera esta noche Would you give yourself to me tonight,
tres horas a mi servicio? Three hours at my service?’
— Como soy vuestro criado, ‘Since I’m your servant, [5]
Señora, burláis conmigo. My Lady, you are joking with me.
— Yo no burlo, Girineldo ‘I’m not joking, Girineldo,
Que de veras te lo digo. I’m saying this to you seriously.’
— Y ¿a qué hora podrá, Señora, ‘And at what time will you be able,
y a qué hora lo prometido? My Lady, to fulfil this promise?’ [10]
— A eso de la media noche ‘At about midnight,
cuando el rey ya está dormido. When the King is already asleep.’
Y a esa hora son las doce It is that time, twelve midnight,
cuando canta el gallo prim o, when the first cock crows;
medianoche ya es pasada, And now it is after midnight, [15]
Girineldo no ha venido. and Girineldo has not come.

1
Alexander, Tamar, Isaac Benabu, Yaacov Ghelman, Ora Schwarzwald, and Susana Weich -Shahak, ‘Towards a
typology of the Judeo-Spanish folksong: Gerineldo and the romance model’, Yuval, Studies of the Jewish
Music Research Centre, 6 (1994), 68-163
The Singer and the Scribe

— Malhaya tú, Girineldo, ‘Damn you, Girineldo,

¿quién amor puso contigo Who is loving you?
Medianoche ya es pasada Now it is after midnight,
y tú no has venido. And you have not come.’ [20]
Ellos en estas palabras But at these words
y a la puerta dio un suspiro. There came a sigh at the door.
— ¿Quién es ése o cual es ése ‘Who is that? Who is that
que a mi puerta fue atrevido? Who dared to come to my door?’
— Girineldo soy, Señora, ‘I am Girineldo, My Lady, [25]
que vengo a lo prometido. I’m coming for what was promised.’
Pusole escalera de oro, She let down a stairway of gold,
por ella se había subido; He has climbed up it;
se agarraron de la mano They took each other’s hand
y hacia la alcoba se han ido; And went to the bedroom; [30]
se pusieron a jugar They began to play
como esposa con marido. Like husband and wife.
A eso de la medianoche At about midnight,
el rey pidió su vestido: The King asked for his clothes:
— Que lo traiga Gerineldo ‘Girineldo should bring it to me, [35]
como mozo más antiguo. as my oldest page.’
Unos dicen: - No está aquí - Some say ‘He isn’t here’,
otros dicen: - Ya se ha ido - Others say ‘He’s gone’;
Y el rey que estaba en escuchas The King heard them,
a su encuentro le ha salido And went out to look for him. [40]
Y encontró a los dos durmiendo He found the two of them asleep
como esposa con marido; Like husband and wife.
— O si mato a la princesa, ‘Should I kill the Princess?
tengo mi reino perdido; I’ll be ruining my kingdom;
o si mato a Girineldo, Should I kill Girineldo? [45]
lo he criado entre chico. I’ve raised him since a boy;
Más vale que mire y calle Better to watch and keep quiet
Y no diga nada a ninguno. And to say nothing to anyone.’
Sacó alfaje de su cinto He took his scimitar from his belt
y entremedias lo ha ponido; And put it between them. [50]
con el frío del alfaje With the cold of the scimitar,
la reina se ha conmovido. The Princess has stirred;
— Girineldo, Girineldo, ‘Girineldo, Girineldo,
ya estamos los dos perdidos; We’re both of us doomed!
que el alfaje de mi padre For my father’s scimitar [55]
nos lo han puesto por testigo. They’ve put here as a witness.’
— ¿Por dónde me iré yo ahora ‘Where can I go now,
para no ser consentido? So as not to be found out?’
— Vete por esos jardines ‘Go out through those gardens
cortando (floras) rosas y lirios. Picking roses and lilies.’ [60]
Y el rey que estaba en escuchas But the King was listening out
a su encuentro le ha salido; And has come out to meet him.

170
 Roger Wright

— ¿Dónde vas tú, Girineldo, ‘Where are you going, Girineldo,

tan triste y tan amarillo? So sad and so pale?’
— He venido de un jardín, ‘I have come from a garden, [65]
de cortar rosas y lirios. From picking roses and lilies.’
— No lo niegues, Girineldo, ‘Don’t deny it, Girineldo,
que con mi hija has dormido. You’ve slept with my daughter.’
— No lo niego yo, buen rey, ‘I cannot deny it, My Lord,
Que yo la culpa he tenido. And the blame is all mine.’ [70]
— No has tenido tú la culpa, ‘No, the blame is not yours,
que mi hija la ha tenido. The blame was my daughter’s;
Y el sábado por la noche And on Saturday night,
seréis esposa y marido … You’ll be husband and wife ’…

At this point, in the middle of the musical phrase, the singer goes straight off into a different
ballad (La boda estorbada). She has a hoarse voice, but the words are clear; the music is
patterned into four-line octosyllabic verses, with a break after each pair of lines, but the
whole ballad runs on continuously without any break between these four-line musical
groups. She accentuates the rhythm of the music over the natural stresses of the words, in
the usual Hispanic manner, and sings slowly enough for an appreciative audience to have
time to react appropriately at various points. The English translation lacks the rhythm and
assonance of the original, and is strictly utilitarian. The text was transcribed from the tape-
recording; she was herself singing, perhaps partly improvising, from memory, not from a
written text.
This may well be the genuine oral tradition. The tale and much of the wording is
directly descended from late medieval Spain, and in this case it is possible that the music
has too. This version of Gerineldos was chosen for the present purpose largely because the
performance may approximate those of the medieval tradition, although of course many
singers would have been rather more sprightly than this one. It is also possible, however,
that none of the music that is heard with the Spanish ballads nowadays is similar to that of
the Middle Ages.
The tradition of singing ballads like this is old and deep-rooted in every part of the
Hispanic world. Specialists in social anthropology and in Hispanic literatures keep finding
them still, even now, in every Hispanic area and every Hispanic language and dialect; that
is, in the Iberian Peninsula (Portugal, Galicia, León, Castile, Navarre, Aragón, Catalonia,
Valencia, Andalucía, the Balearics), in the Canary Islands, Madeira and the Azores, and in
every country of Latin America that has been investigated; as well as in the communities of
the Sephardic Jews who spread throughout the Mediterranean, now mostly to be found in
Israel or The United States. Almost all the stories that are heard in these songs nowadays
are ancient, although new themes and plots are developed from time to time, particularly in
Mexico, as we shall see later. And it is still in essence an oral genre rather than a written
one. That comment is not intended to imply that the ballads are never written down, but that
they are still usually sung from memory rather than following a written script. This means
that a ballad can exist for centuries without anyone ever recording it in writing, and for that
reason their presence in former times is almost invisible to us now. Conversely, the
invention of magnetic tape means that they have recently been tape-recorded in vast
numbers, and although the specialists often adopt a pessimistic tone and say that the genre

171
The Singer and the Scribe

is falling out of fashion and likely to disappear, especially as television pre-empts the
anthropological niche that the genre used to fill, it is also true that the documentation of
these oral performances is much greater now than ever before. Most years in the 1970s and
1980s, for example, teams of researchers went out from Madrid into the remote valleys of
Northern Spain and elsewhere to record more versions on their tapes, so there is now a large
archive of such recordings in Madrid, waiting for someone to come and listen to them.
There is much less evidence for the living oral genre as it existed before the invention of the
tape-recorder, of course, but even so it is generally assumed that this was more of a living
genre in the past than it is now. So we might be justified in believing that performances
such as this are our best evidence of the nature of the medieval performance genre.
But we have got some old texts. At times in the past, it has fortunately been fashionable
for editors to collect these songs from the mouths of singers in order to print the words. In
Spain, the first time that this became a fashionable enterprise was in the mid-sixteenth
century, and there are huge collections from that period, known as Romanceros.2 But all
such editions involved some adjustment to suit normal written conventions, and for that
reason it has been argued that since the oral genre has probably not changed greatly over
the years a genuine oral ballad from the present day is a better witness to the nature of its
fifteenth- or sixteenth-century oral precursor than are the versions found in the huge
Romanceros of that time. The words can vary, though, and to some extent they have to
change if the ballad is to survive at all, so this claim needs to be treated with care.
Before going any further, I should explain what a Spanish ballad, known in Spanish as a
romance, actually is. It is defined in a formal, metrical, way and no other. It can be short, it
can be long. The shortest found in the old collections is of twelve lines, although,
obviously, singers are often singing odd one- or two-line snatches while doing the washing-
up, without performing the whole song. The longest of the ballads in the old collections has
over a thousand lines, but that is quite exceptional; the average length is of fifty to sixty
lines. Nearly all the lines have eight syllables. This syllabic regularity is not consciously
worked out arithmetically by the singers, of course; the point is just that in this way the
words fit the simple music that accompanies them. At least, the music is usually simple in
Spain, but, as we shall see, the Eastern Sephardim have sometimes taken over Eastern
music of greater sophistication. These octosyllabic lines have alternate assonance; that is,
the end of every alternate line shows vowel rhyme from the final stressed syllable onwards,
regardless of the consonants; thus in this version of Gerineldos the assonance pattern is -í-o
(polido, servicio, conmigo, etc., all the way through to marido). That is all that is involved
in the definition of a Spanish ballad. If a verse has eight-syllabled lines with alternate
assonance, it counts as a romance.
The themes of the ballads that are most popular nowadays are those of the most basic
human emotions, such as love, sex, grief, pride, honour, duty, fear and poetic justice.
Emotional reactions tend to take priority over the narration of the events that cause them, as
the listening audience are invited to feel the same emotions as the protagonists. Reactions to
the death of a loved one, for example, are presented more often, and more graphically, than
the actual death. Most of the stories concern, at least partly, the adventures of young

2
For relevant bibliography see Wright, Roger, Spanish Ballads with English Verse Translations (Warminster:
Amis and Phillips, 1987), pp. xiii-xvi.

172
 Roger Wright

women. The women in the ballads are almost all more warm, more human, more intelligent,
more caring, more enterprising and generally more interesting than the men. Half the
listeners may well assume that that is only to be expected, because after all women are
indeed more warm, caring, intelligent, etc, than men; but this feature may also be both the
cause and the effect of the fact that nowadays, at least, it tends to be women rather than men
who sing them. Perhaps it has always been a predominantly female genre. It is tempting to
agree with the feminist viewpoint that in this way many of the old ballads were implicitly
subversive of traditional male authority in Spanish society (see the excellent article to this
effect by Teresa Catarella). 3 But most of the emotions felt by protagonists within the
ballads can be experienced by anyone in the audience, of whatever gender. In the old
tradition this dependence on the powerful appeal of universal human emotions was
probably just as great as it is now, even though many of the first printed ballads were
ultimately based on striking historical events of the medieval Iberian Peninsula. As time
goes by, naturally, people gradually forget historical details, and many of these stories lose
their roots and turn into more general tales of love and adventure; in these cases a specialist
in literary history can sometimes see where the plot has originated from, but the singers and
their audience need know nothing at all about the original historical context for a
performance to be a success within its own context, engaging the sympathetic attention of
the listeners.
Only a few ballads are to be found in the enormous manuscript song-collections of
fifteenth-century Spain, or even in the gigantic printed Cancionero General of 1511,
because at that time, in the early Renaissance period, the ballads were thought to be of
unacceptably low quality, being rustic rather than literary. Describing these collections as
enormous is not an exaggeration; there are over 700 fifteenth-century Spanish poets whose
names we know, some of them prolific, plus a large number of anonymous poems. These
writers ignored the ballad genre almost entirely. But once the invention of printing became
more widely accessible, ballad texts were often printed on single sheets of paper, or in
leaflets containing some four to six ballad texts, known as pliegos sueltos. We know now
that there were many thousands of these sheets. Very few indeed of these survive, but
sometimes the only early text that we have of a ballad that is now famous exists in just one
surviving sheet: this happened to the famous ballad of Melisenda, for example. Then in
Antwerp, the cultural capital of the Spanish Netherlands, the printer Martín Nucio gathered
together many of these separate sheets and published them in two large volumes, one in
(probably) 1548 and one in 1550. Over the next few years several other collectors followed
his example in Spain. The texts included in these volumes are generally treated by modern
investigators as if they were more or less authentic, that is, recorded as they were sung; but
even so, it is often possible to suspect that the collector or the editor or the printer has
smartened the texts up to some extent. Unfortunately, this period presents only a short-lived
window of opportunity for the researcher; towards the end of the sixteenth-century, at the
height of the Spanish Golden Age, the ballad form became a fashionable one for educated
poets to comp ose their own works in and some were set to new music by court musicians
such as Narváez and Milán. In this way it became a written and even a learned genre also.
Thus there is a vast collection of poems in romance form, in nine volumes, printed at the

3
Catarella, Teresa, ‘Feminine historicising in the romancero novelesco’, Bulletin of Hispanic Studies, 67 (1990),
331-43.

173
The Singer and the Scribe

start of the seventeenth-century, which is largely useless for the modern folklore specialist
interested in the history of genuine oral ballads. From then on many of the best Spanish
poets and playwrights, including Góngora, Lope de Vega, Meléndez Valdés, the Duque de
Rivas, Antonio Machado, and Federico García Lorca in his Romancero Gitano (Gypsy
Ballads) and in some of his plays, have composed in the ballad form. Those are not the
ballads being considered here today. Even so, it is worth pointing out that the genuine oral
tradition is more important within mainstream literary culture in Spain than ballads are in
most other countries, and are indeed usually included on University literature courses.
In the nineteenth century, three centuries after Martín Nucio, specialists suddenly
discovered once again, apparently to their surprise, that out there in the countryside there
was still a very live tradition of singing the folk ballads, and the literary anthropologists
who went out collecting then found that many of the same themes and stories were being
sung as were present in the great sixteenth-century collections. They also came across
several ballads which seemed to be ancient, but which had not been recorded on paper in
the sixteenth century (so far as we know). This lack of sixteenth-century documentation can
be explained; it looks as if those early collectors had a particular interest in the ballads
based on stirring events in national history, tending to ignore the more romantic tales of the
adventures of lovers, princesses, witches and so on, that were of greater interest to the
collectors of the nineteenth century, being children of the Romantic age.
The earlier history of the oral genre, however, is a more controversial topic. There has
been considerable scholarly discussion of when ballads first began to be sung in Spain. All
is informed guesswork and conjecture, but it seems likely to me, at least, that this is a very
old genre indeed. Mine is not a new view, since many of the nineteenth-century Spanish
scholars thought the same, but they also thought that the attested written ballad texts were
older than the apparently oral thirteenth-century Spanish epic tradition; which cannot be
true, since we know now that no attested ballad text comes from before the later fifteenth
century. The commonly accepted dating of the supposedly earliest written text to 1421 has
been plausibly reassigned to the 1470s,4 which probably leaves the earliest known text as
one dating from about 1450. But on the other hand we also know now that the genre can
exist in perfect health without being written down at all, because that is what happened in
the eighteenth century, for example; so it is not at all irrational to propose that the genre
existed in the early Middle Ages, perhaps even in Visigothic Spain before the Moslem
invasion of the peninsula in 711, and conceivably even in Roman Spain, since we know that
Latin popular verse included octosyllabic rhythmic verse forms. This is not to suggest that
exactly the texts that we hear now, or read in the Golden Age collections, were being
performed in precisely the same way in the year 1000. But the essence of a folk ballad on a
universal human theme can remain more or less the same for centuries, even if details of the
plot, the length, the wording and the grammar change from performance to performance in
the ordinary way. And whatever the chronology of its first appearance, we can be sure that
the genre and several of the tales that are still heard nowadays were also available in the
later medieval tradition, as well as possibly some of the music, and it is easy to understand
the excited way that the specialist researchers in Madrid feel they are performing a kind of
national emotional archaeology, rediscovering essential and basic aspects of their human

4
Aubrun, Charles V., ‘Le romance gentil dona gentil dona, une énigme littéraire’, Iberoromania, 18 (1983), 1-8.

174
 Roger Wright

heritage. I have written two articles arguing for the existence of ballads in twelfth- and
thirteenth-century Spain, which have failed to provoke any disagreement at all, on the
chronological point, at least.5
The taste for composing ballads on contemporary history flourished in the fifteenth
century, particularly in the frontier area between the Moslem kingdom of Granada, on the
south coast, and the Christian area of Andalucía to the north and west of it. Several ballads
about the events of those conflicts were subsequently recorded in written form. Since the
original version of this paper was given in Edinburgh, it is both appropriate and a pleasure
to be able to declare that the most acute analysis of the social context of these frontier
ballads is probably that of Angus Mackay.6 It seems likely that many of these ballads were
composed by actual participants in the events recounted, and they are remarkable in many
ways, not least because several are noticeably pro-Moslem. On the whole these areas were
at peace rather more than they were at war, particularly between 1350 and 1460, and the
people on each side of the nominal frontier tended to get on reasonably well together. War
was promoted by rulers who lived further away, in Castile to the north, or in the Moslem
capital city of Granada itself, and unwelcome to those on the frontier.
Our next textual example is one of these: the ballad of AbnÁmar, as attested in the
Zaragoza collection of 1550-51. This derives from events of the 28th June 1431, when King
John II of Castile led an army south to Granada and camped overlooking the city from the
north. From here there is a spectacular view of the city still, with the Sierra Nevada in the
background, snowy even in late June. The last part of the ballad recounts the ensuing
skirmish outside the city walls, which in real life ended with the King’s Moslem guide, the
exiled dissident AbnÁmar, being installed briefly as a puppet ruler. But as time went by, the
first part of the ballad was the only section to survive in the collective memory; in this the
King and his guide discuss the view of Granada, and the King is so taken by its beauty that
he proposes marriage to the city, whose answer comes back on the wind, replying in the
negative. Most of the original historical circumstance thus vanished from view. Other
Golden Age versions correspond to only the first thirty lines of the 1550-51 version; these
run as follows, with my own translation7 intended in this case to approximate the effect of
the rhythm and the assonance, sometimes to the detriment of literal semantic equivalence:
— Abenámar, Abenámar, ‘Abnámar, Abnámar,
moro de la morería, Moslem and friend -
¿qué castillos son aquéllos? what are those towers
altos son, y reluzían. so shining and splendid?’
— El Alhambra era, Señor, ‘That’s the Alhambra,[5]
y la otra la Mezquita, that’s the Mezquita,
Los otros los Alixares that’s Alixares,
labrados a maravilla; with beautiful features;
el moro que los labró the Moor who designed it

5
The articles have been reprinted in Wright, Roger, Early Ibero-Romance (Newark: Juan de la Cuesta, 1995)
where they constitute chapters 19 and 20.
6
Mackay, Angus, ‘The Ballad and the Frontier in Late Medieval Spain’, Bulletin of Hispanic Studies, 15 (1976),
15-33.
7
Wright, Spanish Ballads, p. 106.

175
The Singer and the Scribe

cien doblas ganava al día; earned hundreds of doblas; [10]

la otra era Granada, the rest is Granada,
Granada la noblescida, Granada the noblest,
de los muchos cavalleros with many fine horsemen
y de gran ballestería. and many fine bowmen.’
Allí hablara el Rey Don Juan, King John then cries out [15]
bien oyréys lo que diría: these words of good omen:
— Granada, si tú quisiesses ‘If you’re willing, Granada,
contigo me casaría; please be my bride;
darte he en arras y dote I can offer you Córdoba,
a Córdoba y a Sevilla Sevilla besides, [20]
y a Xerez de la Frontera and also the town
que cabe sí la tenía. of Jerez can be yours -
Granada, si más quisiesses, if you want it, Granada,
mucho más yo te daría.’ I’ll give you much more.’
Allí hablara Granada, But the breeze from Granada [25]
al buen rey le respondía: this answer has carried:
— Casada so , el rey Don Juan, ‘I’m married, King John,
casada so, que no viuda, not widowed, but married;
el moro que a mí me tiene my husband the Moor
bien defenderme quería.’ will fiercely defend me.’ [30]

Lines 31-52 then recount the ensuing battle, and how the Castilians were paid off with great
riches to go away.
The following is a version of the ballad from modern Morocco, where nostalgia for the
glory of Granada persisted more than the historical details of the 1430s. The first and last
parts of this version correspond quite closely to the first ten lines of the 1551 version. This
is taken from a commercial record (Cantos Judeo-Españoles de Marruecos, Saga, 1984)
made at a time when there was a fashion for making such records, and the singer (Clarita
Benaim) and guitarist (Jesús García) are professional musicians. The music is arranged in
four-line verses with a guitar passage between each verse, whereas probably the early
ballads were continuous, like the sefardí performance of Gerineldos above; but this is still
recognizably the same genre, performed with the same slow deliberation that encourages
audience identification with the characters and situation. The development of the words in
this case, ignoring all but the start of the original ballad, is an instructive example of how
the oral tradition changes in a changing world, managing thereby to survive where a version
corresponding to the written text from the past would have little modern resonance at all.
(my literal translation):
— Abenámar, Abenámar, ‘Abnámar, Abnámar,
moro de la morería, Moor of the Moors,
el día que tú naciste The day you were born
grandes señales había. Great were the signs.
Estaba la mar en calma, The sea was calm,
la luna estaba crecida The moon was full.
Moro que en tal signo nace A Moor born under such a sign
no debe decir mentira. Should tell no lie.’

176
 Roger Wright

— Y no os la diré, Señor, ‘And I will not tell you one, Sir,

aunque me cueste la vida, Even if it costs my life,
que de chico y muchacho For as a young child and boy
mis padres me lo decían: My parents told me this:
que mentiras no dijese, That I should tell no lies,
que era grande villanía. For that was an evil thing.
Pregunta pues, el buen rey, So ask me, Good King,
que la verdad te diría. For I would tell you true.’
— ¿Qué palacios son aquéllos? ‘What are those palaces?
Altos son y relucían; They are tall, and they shine’
— La Alhambra era, Señor, ‘That’s the Alhambra, My Lord,
palacio de gran valía. A palace of great price.
El moro que lo labraba The Moor that designed it
cien duros ganaba al día, earned a hundred duros a day;
y el día que no los labra any day he did not work on it,
otros tantos se perdía. he lost the same amount.’

The Alhambra is the only one of the original buildings to remain in the modern version,
presumably because it was the only one still standing in the twentieth century; the coins
earned by the architect are no longer the now obsolete doblas but the modern duros (five-
peseta pieces); the military guide seems to have become a soothsayer, and the original
military context has disappeared; but some of the phrases are identical, and there is no
doubt that this has descended directly from the tale as told almost five hundred years
before, in 1550.
Two rather startling examples from contrasting Hispanic environments can help further
to illustrate this theme, of the way that the genre has been adapting to a changing world.
The first, Mas ariva i mas ariva, is a recording made in Israel of a genuine old Spanish tale,
concerning some poor fishermen who manage to fish out a rich duke, and the words are
undoubtedly in ballad form. But the singer of this ballad is an immigrant from Turkey,
Berta Aguado, performing unaccompanied; the words are Eastern Judeo-Spanish, and the
music is essentially Turkish. Below is the text as printed in the booklet that accompanies
the tape (Kantes Djudeo-Espanyoles, edited by Moshe Saul, no.5; published in Israel by
Sefarad, the Society for the Conservation and Diffusion of Judeo-Spanish Culture). The
spelling is the most distinctively Judeo-Spanish feature of the language, in fact, for the
recommended spelling in Israel is quite deliberately different from that of the same words
in Spain, such that in the event the oral version might be more intelligible to a Spaniard
than is this written text:
Mas ariva i mas ariva Further up and further up
en la sivda de Silivria In the city of ‘Silivria’,
Ay avia peshkadores There there were poor fishermen
peshkando sus proverias Trying their luck.
Vieron vinir tres en kavayo They saw three men come on horseback
asiendo gran polveria Raising much dust;
Vinieron serka del rio They came close to the river;
a la mar lo echarían They threw something in the sea.
Echo ganchos i gancheras One fisherman lowered hooks and lines,

177
The Singer and the Scribe

por ver lo ke le salia To see what came out for him;

Le salio un duke de oro A golden duke came out for him,
ijo del rey paresia He seemed like a son of the king.
Si se lo trayian bivo If they took him to the king alive
ombres grandes los haria he would make them great men;
Si se lo trayian muerto If they took him to the kin g dead
sus presentes les daria He would give them presents.
Kamiza de Holanda yeva He is wearing a fine linen shirt,
kaveson de sirma i perla A collar of silver thread and pearl;
Aniyo yeva en el dedo He has a ring on his finger;
sien proves rikos se asia He’d make a hundred poor men rich.

The combination of extremely slow Turkish music, including several extended melismas on
individual syllables, with a recognisable Spanish tale, sounds quite amazing. What the
singer and her listeners probably have no idea of is the original context of the events
recounted, which in fact refers to the disposal and subsequent recovery of the body of
Cesare Borgia’s murdered brother Giovanni, son of Pope Alexander VI, in the Tiber in
Rome in the year 1497; 8 hearing that men had been seen dumping the body, the Pope
ordered the river to be dragged and the richly dressed victim was then discovered. The
ballad is always presented from the viewpoint of those who dragged the river, and usually
printed elsewhere under the title of La muerte del duque de Gandía. The fact that historical
accuracy was no concern of the singers once the original circumstances had faded from
view is attested in striking form by the name of the place where the tale is set; in the
modern Sephardic tradition, it never seems to be set in Rome (the ballad seems not to
survive still in Spain itself, although it is found in Golden Age pliegos sueltos and
collections). Silivria (line 2) seems to be a garbled form reminiscent of Sevilla; in four
originally Turkish versions recorded in the United States in the 1970s, the placename is
once ‘Vizirolú’ (Benmayor no. 2a), with a musical transcription by Judith Mauleón), once
Marseilles (Benmayor no. 2c), and twice Messina.9 Singers, notoriously, can be equally
cavalier about names of people. In many of the Sephardic communities, where not only
were most of these songs never written down but there did not even exist the Roman
alphabet to do that in if they wished, those tales that were able to survive four centuries and
more did so by losing the context, acquiring a more generally intriguing feel and adapting
to the local musical styles, and the result can send shivers down the spine.
The last example here comes from the other end of the Hispanic world. In Mexico, the
tradition of inventing new songs on rousing political and military events of the present, such
as has hardly existed in Spain since the fall of Granada in 1492, is still alive, and has been
particularly strong in the twentieth century. Although in form they are obviously the
modern descendants of the old Spanish ballads, they are usually called something else,
corridos; and the music tends to have a tinge of mariachi, since these too have
accommodated to the local musical tastes. One type of newly-invented corridos at the

8
Benmayor, Rina, Romances judeo-español de oriente (Madrid: Gredos, 1979), pp. 31- 41: Further references are
to Benmayor plus an item number.
9
Benardete, Maír José, Judeo-Spanish Ballads from New York (Berkeley: University of California Press, 1981), p.
23.

178
 Roger Wright

moment is that of the so-called narco-corridos, about the adventures of Mexican drug-
traffickers, particularly those that operate over the border into the United States. What
follows here is one of several which were originally composed during the Mexican Civil
War of 1910-18. The inventors of corridos are just as likely to create rhythmic prose tales,
but when they burst into song it tends to be in the old octosyllabic ballad form, a process
fascinatingly described by Cardozo Freeman.10 The one here printed is called the Corrido
de Juan Carrasco, based on events of 1913 which need a brief explanation.
Francisco Madero (1873-1913) was a rich cotton farmer who challenged the Mexican
dictator Porfirio Díaz for the presidency in 1910. He was elected in 1911 and assassinated
in 1913. A wide movement to avenge this murder started, particularly in the north, under
Carranza and Pancho Villa, against the federales of the next president, Huerta. The towns
mentioned in this ballad, Quelite, Mazatlán and Acaponeta, are in Sinaloa province, just
inland of the west coast, opposite the southern end of Baja California, some 450 miles
north-west of Mexico City. The local leader of this movement and hero of this corrido, Juan
Carrasco, was an uneducated cowboy, expert horseman and notorious drinker, and his
colleague Ángel Flores was a docker, but their local rebellion, though disorganised, was
successful11 and the two became popular heroes; Pancho Villa, in his memoirs, even
referred to him as ‘General’ Juan Carrasco.12 In another corrido he is referred to as ‘Juan
sin Miedo’ (‘Juan the Fearless’).13 These were local corridos of the time, remarkably
similar in some respects to those inspired by minor but stirring military events on the
fifteenth-century Granada-Andalucía frontier. All the historical references are assumed to
be understood by the listener. The version of the ballad here presented is that performed
three decades later on a 78 rpm record by a singer called Juan Meza, recorded in about
1942. In form this is the same as a medieval ballad, except for the fact that it is structured
into four-line stanzas with mariachi music between each one, and the assonance pattern
regularly changes with each verse, often in the event using full rhyme; also Juan Meza, like
many Latin American singers, has the irritating habit of letting the rhythm of the music
(././../.) override the word stresses, such that in the second line, for example, the clitic la and
the second syllable of muerte receive linguistically undue prominence (‘por lá muerté de
Madéro’), but otherwise the corrido and its performance are most attractive:
Carrasco quedó sentido Carrasco was upset
por la muerte de Madero; By the death of Madero,
por eso se levantó And so he rebelled
con la gente del potrero. With the people of the plains.
Su hijito le decía His young son said to him
— Padre mío, no te metas; ‘My father, don’t attack;
¡ahí vienen los federales Here come the government troops
por el río de Acaponeta! Up river from Acaponeta!’

10
Cardozo Freeman, Inez, ‘Creativity in the Folk Process: the birth of a Mexican corrido’, in El Romancero hoy:
nuevas fronteras, ed. by Antonio Sánchez Romeralo et al. (Madrid: Cátedra Seminario Menéndez Pidal,
1979), pp. 205-14.
11
Knight, Alan, The Mexican Revolution, Vol.II (Cambridge: Cambridge University Press, 1986), p. 33.
12
Guzmán, Martín Luis, Memoirs of Pancho Villa (Austin: University of Texas Press, 1965), p. 275.
13
Knight, The Mexican Revolution, II, 222.

179
The Singer and the Scribe

Juan Carrasco ambicionaba But Juan Carrasco wanted

la libertad de su pueblo, The freedom of his people,
Y le gritaba a su gente: And shouted out to his followers:
— Si no lo cumplo, ¡me cuelgo! ‘If I don’t succeed, I’ll hang!’
Decían que no traía parque, They said he had no ammunition,
y que él traía malas armas; And that he had bad guns;
en el pueblo de El Quelite In the town of Quelite
les avanzaron las cargas; They attacked him;
morían los federales The government troops died
por toditas las labores, Despite their efforts,
de miedo que les tenían For the fear that they felt
a Carrasco y a Ángel Flores. Of Carrasco and Ángel Flores.
Como que quiere hacer aire, As sure as the wind’s going to blow,
como que quiere llover, As sure as it’s going to rain,
el que no quiera a Carrasco Anyone who won’t follow Carrasco,
algo le va a suceder. Something bad will happen to him.
Vuela, vuela, palomita, Fly, fly, little dove,
descansa en aquel peñasco; And rest upon that rock;
aquí se acaba el corrido This is the end of the corrido
del valiente Juan Carrasco. Of brave Juan Carrasco.

This ballad is at a very early stage of decontextualisation, for thirty years after the events
not much can plausibly be inserted that contradicts known facts; but even here the process
has begun, and if this song should still survive in the oral tradition of the twenty-first
century, untouched by the record, Carrasco could become as separated from his roots in
reality as AbnÁmar.

Thus the four recently recorded versions included in this lecture demonstrate how the Spa-
nish ballad has indeed been changing in the modern world; Juan Carrasco is a mere 30
years down the road from its subject matter, the Zaragoza AbnÁmar 120, the Turkish Mas
ariva 500, the modern Moroccan AbnÁmar 550, and in the case of the last two literary research
has been needed to explain the circumstances which the ballad itself cannot supply. Thus
it may be the case that Gerineldos also has some kind of now irrecoverable historical origin
origin (it has been suggested, for example, that the name, at least, might come from Charle-
magne’s secretary Egginhard). But it may also be true that from listening and watching
modern performances we can get a better idea of what the older oral genre was like than
we could if we were strictly limited to the necessarily partial evidence of the sixteenth-
century written texts, for so much of the impact of a performance is in the hands of the per-
former rather than in the text14 that we can never be sure in any case that a singer might not
have explained the historical circumstances of the tale in advance for example.15

14
Wright, Roger, ‘Point of View in the Ballad Performer’, Hispanic Research Journal, 1 (2000), 97-104.
15
I would like to express my particular gratitude to Isaac Benabu for his help with Gerineldos, Christy MacHale
for his help with AbnÁmar, Margaret Sleeman for her help with Mas ariva i mas ariva, and both Kenny
Murray and José Saval for their help with the Corrido de Juan Carrasco.

180
 The Stewarts of Fetterangus and Literate Oral Tradition

Thomas A. McKean

The Stewarts of Duke Street, Fetterangus, are one of Scotland’s outstanding singing
families,1 celebrated for their remarkable fund of traditional lore: stories, riddles, music,
songs of many kinds and particularly classic ballads.2 This essay examines how the skills of
reading and writing work in the context of their largely oral culture. 3 As will become clear,
I do not see oral tradition as existing in opposition to written tradition, but rather each is
dependent on the other for propagation and perpetuation. In this light, we can see oral
tradition as the process of passing on (social) culture through shared practices and lore,
what David Atkinson calls, fundamentally, ‘a cognitive function of individuals and groups
of people situated in time and space’.4
There have always been at least two types of oral tradition: the transcendent and the
transient,5 embodying, respectively, the long-term sweep of cultural knowledge – passing
on and preserving material over centuries – and the microcosmic tradition at the individual
and family level – one person’s memory of a song, one person’s passing on of that song to a
daughter, son, or contemporary. We might think of these as vertical and horizontal
traditions, respectively. Vertical tradition has certainly altered in nature with the advent of
literacy, for songs and stories have travelled with ease back and forth between oral and
written culture for centuries now. 6 It is immediately clear, then, that to study oral tradition
today, one must examine its longstanding relationships with writing and print. Although I

1
So called to differentiate the various related families of Stewarts from this area and the North-East of Scotland in
general. They are second cousins to the “Stewarts of Blair”: Alec, Belle and their daughters Sheila and Kathie.
2
I use the term roughly as defined in The English and Scottish Popular Ballads ed. by Francis Child (Boston:
Houghton and Mifflin, 1882–98, reprint New York: Dover, 1965, reprint Northfield, MN: Loomis House,
2002–), but without adhering exclusively to Child’s closed list of 305 types.
3
See David Atkinson’s thorough study, The English Traditional Ballad: Theory, Method, and Practice. Ashgate
Popular and Folk Music Series (Aldershot and Burlington, VT: Ashgate, 2002), p. 20–26 and 245–246, for
discussions of the use of song books and broadsides by English singers and by song editors.
4
Atkinson, p. 27, but note a differentiation drawn between the material of tradition and the culture which uses it.
These are not necessarily the same thing (p. 248).
5
The terms ‘transcendent’ and ‘transient’ are borrowed from Atkinson, p. 248, who sees the transcendent element
as ‘located in an individual, conscious, volitional, affective engagement in a relationship with people, but also
with cultural forms, across time’ (p. 28).
6
Atkinson, pp. 18–19, discusses the impact of broadside print on song tradition. David Buchan also addresses this
question in The Ballad and the Folk (London: Routledge, 1972, reprinted, Phantassie, East Lothian: Tuckwell,
1997), chapter 18. For a discussion of the interrelationship between print and oral narrative traditions in
Scotland and Ireland, see Bruford, Alan, Gaelic Folk-Tales and Mediaeval Romances (Dublin: The Folklore
Society of Ireland, 1969), particularly part 3.
The Singer and the Scribe

agree with William Motherwell that, ‘it is worthy of remark how excellently well tradition
serves as a substitute for more efficient and less mutable channels of communicating the
things of past ages’,7 I am not concerned with the antiquity of the traditions discussed here,
but the horizontal ‘tradition as personal relationship’,8 which is fundamentally unchanged in
form and in function. It has, of course, taken on the influence of print as an added reference
point, but it remains an interpersonal, social and, above all, verbal art.
The change which has taken place is one of specifics, rather than of essential nature. I
do not subscribe to the school of thought that holds that ‘oral tradition’ becomes mere
‘verbal tradition’ over time and oral re-creation mere memorisation in a literate society,9 but
instead see as key the ways an individual regards her source, oral or printed. A good oral
performer and tradition-bearer can exist within literate and non-literate societies. The issue
is not the singer’s environment, or the immediate source, print or oral, but whether aural or
visual memory is being used. Such a distinction exists independently of the existence of
writing. Indeed a single singer can employ either system, or a combination of the two: one
song can be remembered by actual or mental reference to writing or print, or perhaps the
place in which it was learned, another as an aural pattern, recalling the voice of the singer
from whom the song was learned, or the sound of a current or previous performing
environment.10 Traditional singers, in my experience, have always referred to an ‘original
source’, or authority, whether that source be an oral or a written one. So, in the pre-literate
past, when a singer wished to refer to the ‘correct’ version of a song, that is the original
source, she would consult the singer from whom she learned it. Today, such an original
source may be a person, a page, or a recording. What matters is that the singer feels there is
an authority from whom the legitimacy of their version or performance extends. That
feeling is the same as it was many hundred years ago, and is entirely independent of
literacy. If tradition is process, as I believe it is, rather than content, the mechanics of
tradition are essentially the same today as they were in pre-literate times.
This brief study is really just one way to explore an intense emotional bond between
Elizabeth Stewart and her aunt Lucy that is symbolised in their relationship of song, and its
manifestation in the intersections of memory, orality and literacy.
The Stewarts of Fetterangus occupy an interesting space in the transition between
domestic, private tradition and the public, commercialised tradition we take for granted
today. This is analogous to the earlier transition from an oral to a literate society explored
by David Buchan in The Ballad and the Folk . The steps from non-literacy to literacy and
from private to public tradition create new repertoires and new environments for traditional
performance, along with new techniques of learning and propagation, without necessarily
destroy pre-existing means, which undoubtedly continue to flourish where individual

7
Motherwell, William, Minstrelsy Ancient and Modern: with an historical introduction and notes (Glasgow:
Wylie, 1827), p. 3.
8
Barry McDonald, quoted in Atkinson, p. 28.
9
Buchan, p. 2.
10
It is accepted that many storytellers and singers visualise the action of a tale or ballad as they perform; see
Macdonald, Donald A., ‘A Visual Memory’, Scottish Studies, 22 (1979), 1–26 and Turriff, Jane, Singin is Ma
Life (Kingskettle: Springthyme SPRCD 1038, 1996), p. 11. For more on the textuality of oral verse and the
untextuality of printed verse, see English Traditional Ballad, p. 12.

182
 Thomas A. McKean

abilities and suitable social environments exist. Both these transitions actually take place on
a myriad of individual and family levels, in many places, and at many times, rather than to
the whole of society at a single moment in history. The Stewarts, then, are just one strand in
a complex, changing cultural fabric.
Fieldworkers collecting from oral and manuscript sources in the North-East of Scotland
have created an almost unbroken, centuries-old trail of written and recorded collections of
songs, yielding a fresh snapshot of tradition with each generation. This extensive body of
material has held the attention of many scholars, but often the human dimension is all too
easily lost. Song is only a part of the Stewart family’s life and legacy; when Elizabeth
Stewart remembers her aunt or her mother through song and music, she triggers an infinite
variety of memories, memories of their humour, their tragedies, their wit, their humanity,
just as anyone might recall a whole ethos from a single family photograph or an evocative
scent. To pigeonhole anyone simply as a singer or ‘tradition-bearer’, then, is to miss the
defining characteristic of traditional music itself: it is inseparably intertwined in the fabric
of life. Examining the long-standing, multi-layered interactions between oral and literate
traditions, and between the two main characters in this discussion, would be virtually
impossible without reference to the culture from which the singers’ culture, songs, and their
transmission and performance arise.
The key figure in this discussion is Lucy Stewart of Fetterangus, Aberdeenshire (1901–
1982), an outstanding, though lesser-known, figure in the consciousness of the modern folk
revival (fig. 1).11 A Traveller born in Skene Street, Aberdeen, Lucy made her living as a
dealer of new and used goods, but she was also a wonderful performer of ballads, songs,
Märchen, legends and riddles. Of the ballad, she is an acknowledged master, one of
Scotland’s greatest artists, and yet, since the LP of Child ballads produced by Kenneth S.
Goldstein in 1964, 12 and the odd track on Alan Lomax’s compilation albums of British
traditional songs,13 she has hardly appeared on record. Unlike Jeannie Robertson, Lizzie
Higgins, the Stewarts of Blair, Flora Macneill and other traditional singers who became part
of the fabric of the professional and semi-professional folk revival, she never passed on her
material to a paying public. Indeed, Lucy never performed in public at all, preferring to
keep her domestic traditions where they belonged: ‘*No, niver, no’, according to her niece
Elizabeth, ‘she wis too shy. [She sang] in e hoose an fan she wis oot in the cairtie wi me an
Jane’.14 Stewart was, however, recorded on many occasions, firstly by the late Arthur Argo,
then Hamish Henderson with Lomax and, most exhaustively, by Goldstein, the American
folklorist who lived in the nearby village of Strichen in 1959–1960. ‘*He jist aboot bade wi
us’, says Elizabeth, who witnessed many of Kenny’s persistent recording sessions. Several

11
She is, curiously, only given a passing mention in Munro, Ailie, The Folk Music Revival in Scotland (London:
Kahn & Averill, 1984), p. 117, reissued as The Democratic Muse: Folk Music Revival in Scotland (Aberdeen:
Scottish Cultural Press, 1996), p. 70.
12
Lucy Stewart, Traditional Singer from Aberdeenshire, Scotland, Child ballads, vol. 1, produced by Kenneth S.
Goldstein (East Lothian: Greentrax CTRAX 031, 1989, originally Folkways/Rounder, 1964).
13
These recordings made between 1949 and 1969 are now reissued as Folk Songs of England, Ireland, Scotland &
Wales: Classic Ballads of Britain and Ireland, vols 1 and 2 (Cambridge: Rounder 1775 and 1776, 2000).
14
Interview, 2000. Quotes preceded by an asterisk are taken down as close to verbatim as possible.

183
The Singer and the Scribe

other families he recorded have said similar things, so he must have been constantly at
work.
This process of periodic collecting itself, extending from 1951 to 1982, surely had its
effect on Lucy Stewart and the repertoire she chose to sing, and yet she still never
performed outside the home. Indeed, she could scarcely be persuaded to sing for visitors
and hers remained a very private tradition to the end of her days, for sharing with family
and trusted friends. Ironically, it was this reticence that allowed the public performing
careers of her sister Jean and niece Elizabeth to flourish through passing on songs and by
looking after first Jean’s and later Elizabeth’s children.
Lucy’s repertoire, like that of most traditional singers in Scotland, is quite broad and
includes centuries-old traditional ballads and children’s songs handed down through the
generations, to Country and Western and music-hall material. Older material was learned
principally from her mother, Elizabeth Townsley, nicknamed Aul Betty (fig. 2) and various
uncles who also sang; her father was more of a musician than a singer.
Later influences with which she may have come in contact included 78 rpm recordings
and the radio, along with print media, such as broadsides and song books. The privileging
of non-literate sources by singers and collectors is, of course, a well known phenomenon15
and according to Elizabeth, Lucy did not learn any of her material from these latter sources:
‘*I dinna think so. She cuidna read affa well’, as she looked after the younger memb ers of
the family and never got as much schooling as her siblings. Lucy’s was largely an oral
world and her style, learned in the traditional manner, is that of the oral singer. That is not
to say, however, that her entire tradition is an oral one, for while we can say with some
confidence that her personal tradition and modus operandi are oral, her repertoire reveals
the ‘the constant, dynamic, mutual reinforcement of ballad-singing and ballads in print’ that
we have come to expect in British folk tradition.16
In the latter half of her life, Lucy had, and continues to have twenty years after her
passing, an influence on the Scottish folk scene through performances by her niece
Elizabeth, through the use of recordings by the staff and students of the Royal Scottish
Academy of Music and Drama degree programme in traditional music, and through revival
singers such as Ray Fisher, Alison McMorland and Arthur Argo, all of whom spent time
learning songs and singing-style from this shy, unmarried woman with a wicked sense of
humour and great sense of the ridiculous, only to be seen if one had acquired her complete
trust.
Among the most frequently heard of her songs at just about any Scottish festival are The
Plooman Laddies and I Am a Miller Tae Ma Trade, often sung with little or no awareness
on the part of the singer or the audience as to the recent source of the song.17 Apart from the

15
See, for example, an anecdote about singer Harry Cox, in Atkinson, p. 21.
16
Atkinson, p. 25.
17
The Plooman Laddies can be heard on Elizabeth Stewart, ’Atween You an Me (Woodhead, Aberdeenshire,
Scotland: Hightop Imagery, 1993). Lucy’s singing of the Miller is not available commercially, but a version
learned from her is on The Fisher Family (London: Topic TOP 12T137, 1966). It has also been recorded by
Hamish Imlach and numerous other revival singers.

184
 Thomas A. McKean

one album and the few tracks mentioned before, her material has remained almost entirely
in archives and libraries.
For the purposes of this chapter, therefore, I will consider Lucy representative of a
‘pure’, unadulterated, non-commercial style and repertoire of North-East Scottish
traditional singing. These adjectives are, of course, indefensible, but by their use I contend
that her style and repertoire have been affected mainly by feedback from family and friends
(including, of course, ballad-hungry fieldworkers in the 1950s and 60s), and the usual
traditional influences of people, broadcast media, and print, rather than by folk festival and
commercial audiences. This is not a value judgement, simply a statement of fact that allows
us to look at Lucy’s repertoire and style in a slightly different shade of light than that which
shines on a more exposed performer. And here I reiterate my contention that Lucy remains
a traditional, oral singer even though living in a society driven by literacy. For her, these
were simply other influences, not qualitatively different ones. They did not transform the
nature of her tradition. In a nutshell, defining Lucy’s world in this way allows me to draw a
contrast with that of her niece. But first a few words about Lucy’s sister, Jean.
Jean Stewart, 1911–1962, enjoyed a remarkable career as a pianist, accordionist,
broadcaster and band leader, for which she is still well-remembered today (fig. 3). When
her own family was young, this was made possible by Lucy looking after her children while
she was out playing for BBC Radio Aberdeen and at village-hall dances throughout the
area. She held diplomas from the Royal Academy of Music in London and her father was a
fiddler and piper who taught her to read music at an early age. She was a bit of a polymath
who not only performed, but wrote, composed, sang and acted. Even today, anyone over the
age of fifty recalls her music with great affection (see fig. 4).
Jean’s eldest daughter, Elizabeth, born in 1939, is a mixture of the private and the public
embodied in her aunt and in her mother (see fig. 5). Naturally, the whole family has private
traditions outsiders never see, so my outsider observations are here combined with
Elizabeth’s from within. All three of these women have strong public and private strands to
their lives, but as far as the traditional arts are concerned, Lucy’s was a very private,
domestic tradition, as opposed to Jean’s and Elizabeth’s more public, commercial one. The
latter two took their traditions out of the domestic setting in a way that Lucy never did,
considerably adapting it in the process, which is, after all, part of what perpetuates a living
tradition.
In the 1960s and 70s, Elizabeth played Scottish dance music for Foxtrots, Quick-steps,
Tangos, Two-steps, Waltzes and the more indigenous Eightsome Reel and Gay Gordons.
She also adapted some of Lucy’s ballads, ‘up-tempo’ for the dance floor, and played some
of the Rock and Roll, Jazz, and Blues material of the time. These adaptations and
excursions in style may be little different from a performance situation in anyone’s life,
where different contexts demand differing material. What is clear, however, is that
Elizabeth chose, or in a sense was compelled, to place herself in those circumstances and
yet found a way, within the particular constraints of that context, to transpose material
organically from one environment into another.18

18
For a fuller examination of these adaptations, see McKean, Thomas ‘“You Make Me Dizzy Miss Lizzie”:
Elizabeth Stewart’s Up-tempo Traditional Ballads’, Northern Scotland, 18 (1999), 103–15.

185
The Singer and the Scribe

How is a good singer made? Well, the first question is whether the idea of a good or bad
singer is relevant in the traditional context. Virtually anyone can acquire songs, sing or
recite them and still be a valuable contributor, regardless of their (by definition, subjectively
judged) artistic talents. To get a clear, almost statistical, picture of a song tradition, in
relevant comparative scale to the rest of our culture, we really ought to record anyone and
everyone and find out just exactly what percentage of people know one to five Child
ballads, what percentage know six to twenty, and so on. Collectors often concentrate on the
manuscript that survived, or the outstanding singer, not the rubbish that did not survive in
oral tradition, or the truly ordinary singer, though there are honourable exceptions to this
tendency. I leave this sort of statistical survey to others.
So what criteria can we use to assess a singer’s quality? There are two basic avenues
open to us: (1) skill at singing and (2) various assessments of repertoire. I do not propose to
discuss skill and technique in any great detail here; that has been addressed elsewhere at
greater length, though not yet definitively.19 That leaves us with repertoire. Kenneth S.
Goldstein maintained that this was the criterion privileged by singers themselves, and their
acquaintances, when judging a performer’s skill, particularly size: number of songs, number
of verses. The more songs the better; the more verses the better.20 I must acknowledge that I
have heard singers describe their versions as better simply because they are bigger, ‘I sang
the full fifty-six verse version’, but again, this is not the criterion I wish to dwell upon here.
Instead, I wish to concentrate on a few particular songs sung by Lucy and Elizabeth
Stewart and look at how and why they are remembered and passed on. All singers are
collectors. They share much with the professional fieldworker: an interest in songs, an
interest in the ‘best’ versions (judged by various criteria), and an interest in the contextual
links of the material, albeit in a more holistic, personal way than that of an outsider
academic fieldworker.
Much is said these days about how the young are not taking an interest in tradition and
traditional music, but surely this might have been said in any era: not every young person a
hundred years ago would have taken an interest in traditional song, or everyone approached
by a fieldworker would have been a singer with a sizeable and interesting repertoire. No,
exceptional tradition-bearers are a self-selecting subcategory of society. We all, to varying
degrees, bear tradition — we carry information that defines and creates our culture — but in
this case, I mean those people who carry on traditional song, or other structured lore,
through a generation or two. So, they are self-selecting, but are also then selected by others
to receive traditional material. This is an obvious pattern, usually overlooked, as tradition-
bearers are often considered the remains of a society in which, it is vaguely implied,

19
See, for example, McKean, Thomas, ‘Gordon Easton’s “The Aul Beggarman”’, in Ballads into Books, ed. by
Tom Cheesman and Sigrid Rieuwerts (Bern: Peter Lang, 1997), pp. 237–248 and Idem ‘Equating Traditional
Singers’ Terms with Melodic Adaptation’, Acta Ethnographica Hungarica, 47, 1-2 (2002), 91–97. I am also
currently working on a book on the traditional ballad-singer’s art. See also, Munro, Democratic Muse, pp. 91,
95–96, 216, and Porter, James, ‘Jeannie Robertson’s “My Son David”: a conceptual performance model’,
Journal of American Folklore, 89 (1976), 7–26.
20
Goldstein, Kenneth. S., ‘Notes Toward a European-American Folk Aesthetic: lessons I have learned from
singers and storytellers I have known’, Journal of American Folklore, 104, no. 412 (Spring 1991),164–178,
particularly p. 168 ff.

186
 Thomas A. McKean

everyone knew hundreds of songs.21 Undoubtedly, in former pre-literate times people did
know more oral poetry as a general rule, but there would always have been those who knew
more than others, those who made a particular study of it, from the highly trained Gaelic
bards and court poets, to the village ‘character’ who knew songs, ditties, literary and
broadside ballads (such as Peter Buchan’s James Rankin).22
Extraordinary singers like Lucy Stewart and Jeannie Robertson, then, do not simply
occur. Rather they are the product of their environments, coupled with innate intellect,
interest, and talent. In fact, exceptional singers I have known display an almost obsessive
and professional interest in song above all things, even to the point, in the modern era,
where sometimes family and social responsibilities are somewhat neglected in pursuit of
their art. This is only a problem when venues for the performance of this material are
outside the home, which, formerly, would have been the usual environment for such
material. In the twentieth century, as traditional music became more commercial, the
demand for village-hall and theatre performances has escalated, removing the material and
its practitioners from the home environment. What was only recently a social and sociable
art, which helped to build and stabilise family units, has largely become the opposite.
Paradoxically, this shift helped to develop Elizabeth Stewart’s own passion for her
aunt’s material, for while her mother Jean was out playing with her band, Lucy was looking
after Elizabeth and her sisters. Consequently, she was exposed to the songs and stories in a
natural domestic context far more than she might otherwise have been; a clear cut example
of context — in this case family social function — affecting how tradition is passed on and
develops. In this case it was the song tradition, but with her mother a professional musician,
she was also drawn into that life. Even today, it is these two threads that define the cultural
side of Elizabeth’s life. The third thread, drawn from her Traveller heritage, is dealing in
antiques and used goods, her current means of making a living.
Following the tragically early death of her mother, in Elizabeth’s own years of young
motherhood, Lucy cared for Elizabeth’s children when she was out playing with her own
dance band. Consequently, they have been exposed to tradition a generation older than their
own mother’s. They are familiar with the songs, and sometimes discuss them, though they
do not perform in public. This is partly due to their youth, but also because their mother is
the senior singer. As Elizabeth deferred to Lucy, who the older children knew well, so
Jeannette, Elizabeth, and Michael do to Elizabeth. Their time will come to inherit the

21
C[arl] W[ilhelm] v[on] Sydow noted that only a minority takes part in perpetuating tradition Selected Papers on
Folklore, Published on the Occasion of His 70th Birthday. A selection of papers written from 1932–45
(Copenhagen: Rosenkilde and Bagger, 1948; reprinted Arno Press, 1977), pp. 12 and 48.
22
For more on Gaelic bards and court poets, see Thomson, Derick, Introduction to Gaelic Poetry (London:
Gollancz, 1974), particularly chapters 1 and 2. For more background on Peter Buchan and his sources, see
Buchan, pp. 205–222, and Rieuwerts, Sigrid, ‘The Case Against Peter Buchan’, in The Flowering Thorn, ed.
by Thomas A. McKean (Logan: Utah State University Press, 2003).

187
The Singer and the Scribe

mantle of actively carrying on the tradition.23 In the meantime, their spiritual bond with
their aunt Lucy is just as intense and absolute.
To return to the idea of singer-collectors, there are really two levels on which they are
active. The first is the most natural, in which a singer gathers her repertoire together over
time from of a range of sources: other singers, print/written media and, latterly, audio
recordings and radio. They learn songs that fit their criteria: a good song (many verses, an
interesting story, tune, or resonant family connection) or a useful song (for entertaining
children or some such practical function). Prominent examples of this type of collector in
the twentieth century include Belle Stewart (Blairgowrie, Perthshire), Jeannie Robertson
(Aberdeen), and Calum and Annie Johnston (Barra). 24 Older examples might include Agnes
Lyle of Kilbarchan, one of William Motherwell’s most prolific sources, and Anna Gordon,
the ‘Mrs Brown of Falkland’ whose texts feature prominently in Child’s collection.25 All of
these people were literate to one degree or another and there is a good chance that they
wrote some of their own songs down, not necessarily in a rigorous way, but at least in the
informal way that Elizabeth Stewart does, and that I shall address below.
In the second group of collector-singers, more wholly on the collecting side of the coin,
we find people like Margaret Gillespie, supplier of 466 songs (484 song versions) to the
collector James B. Duncan, and the curious case of Bell Robertson, who sent 398 songs
(425 song versions) to Gavin Greig, not one of them with a tune.26 Here we have an insider
collector who is not a singer at all, and who brings us out the other end of the continuum
between non-literate singer and non-singing literate and into the realms of non-participant
observer-fieldwork.

23
As such, they are passive tradition-bearers in the sense proposed by Goldstein, Kenneth, ‘On the Application of
the Concepts of Active and Inactive Traditions to the Study of Repertory’, Journal of American Folklore, 84
(1971), pp. 62–7.
24
For Belle Stewart, see Queen Among the Heather (Cockenzie: Greentrax CDTRAX 9055); The Stewarts of Blair
(London: Topic TOP 12T138, 1965[?]);The Sang’s the Thing, ed. by Sheila Douglas (Edinburgh: Polygon,
1992), pp. 138–143; The King o the Black Art and Other Folk Tales, ed. by Sheila Douglas (Aberdeen:
Aberdeen University Press, 1987); Scottish Traditional Tales, ed. by Alan Bruford and Donald A. Macdonald
(Edinburgh: Polygon, 1994), pp. 333 - 336. For Robertson, see Porter, James, and Herschel Gower, Jeannie
Robertson: emergent singer, transformative voice, Publications of the American Folklore Society, new series
(Knoxville: University of Tennessee Press, 1995); Jeannie Robertson, The Great Scots Ballad Singer
(Kingskettle: Springthyme Records SPRC 1025, 1988, first issued 1959); Jeannie Robertson, Jeannie
Robertson: the queen among the heather, The Alan Lomax Collection Portrait Series (Cambridge: Rounder
1720, 1998); and Scottish Traditional Tales, pp. 55-64, 240-244. For the Johnstons, see Scottish Tradition 13,
Calum and Annie Johnston: songs, stories and piping from Barra (Cockenzie: Greentrax CTRAX 9013,
1995, originally issued in 1980) and Scottish Traditional Tales, pp. 41-43, 64-69, 301-303, 339-341.
25
Agnes Lyle is the subject of a detailed study, McCarthy, W.B. The Ballad Matrix (Bloomington: Indiana
University Press, 1990), as is Anna Gordon, see Buchan, David, The Ballad and the Folk, esp. pp. 62-173.
26
My thanks to Emily Lyle of the School of Scottish Studies, University of Edinburgh, for these figures. The two
women’s immense contributions are published in Greig, Gavin, and Duncan, James B., The Greig - Duncan
Folk-Song Collection, general editors Patrick Shuldham -Shaw and Emily B. Lyle (Aberdeen: University of
Aberdeen Press: Edinburgh: Mercat Press, 1981–2002). Volume eight contains brief studies of Gillespie’s and
Robertson’s material.

188
 Thomas A. McKean

I am particularly interested in a middle type of singer-collector, one who actually writes

down or records songs that interest them, or that they feel worthy of note. Such a man was
Willie Mathieson (born ca. 1880), a farm servant in King Edward and Ellon, Aberdeenshire.
Scotland, and the North-East in particular, has an extensive history of manuscript song
collection,27 so Mathieson is no anomaly, but an assiduous example, collecting for himself
and posterity, but with no thought of publication or commercial gain. Mathieson
consciously collected songs, learned them and wrote them in three large notebooks, the
ultimate in academic fieldwork terms. His interest lasted many decades and so devoted to
songs was he that when he lost some notes in a fire, he immediately rewrote them from
memory. Not only, then, did he collect the songs from others for his own repertoire, but he
collected them from himself in order to replace his lost manuscripts. He had known Gavin
Greig and was also one of the American collector James Madison Carpenter's contributors
during his 1928–1935 tours of the British Isles.28 Two decades later, he performed the same
service for Alan Lomax and Hamish Henderson on their 1951 Scottish collecting tour. A
few years ago, the cycle began again as his grand-daughter contacted me having heard
something about ‘a singer, Willie Mathieson, from Ellon’, on the radio. She frequently
attended Traditional Music and Song Association festivals, and had a liking for North-East
song, but before hearing about him on the radio, was not aware that her own grandfather
had sung. I told her that photostats of the notebooks were preserved at the School of
Scottish Studies, along with the tune of each song on tapes made by Hamish Henderson.
She has now consulted these notebooks, and even learned a few of the songs – a good
example of the value of fieldwork, and perhaps all the more so in that the collector was an
insider with an acute awareness of the value of the material.29
Another way that some insiders have collected material is on tape. An astonishing
number of people in the North-East had reel-to-reel tape recorders in the 1950s and 60s, and
had the awareness to record their traditional songs. Naturally, a lot of the material they
recorded was the current music of the day – Country and Western and pop – but a
surprising amount of older material, from ballads to music hall, is also preserved. Jane
Turriff, also of Fetterangus, is one such collector who recorded herself, her sisters and her
mother in the 1950s and 60s, singing all kinds of material on tapes she still prizes today.
The Stewarts of Fetterangus were and are literate: Lucy could certainly read, though her
writing was not as fluent. Elizabeth and her late mother were both highly literate textually
and musically, and could write in Scots as well. This latter skill is almost entirely self-
taught, sometimes by the example of books of Burns’ poetry, local newspaper columns, or

27
Many well known eighteenth- and nineteenth-century song collectors drew, directly and indirectly, on North-
East material. For a comprehensive résumé of this material, see Buchan, pp. 278–280. The Aberdeenshire
collecting tradition is largely fieldwork based, making F. J. Child’s work something of an anomaly, though
some of his texts were, admittedly, only a generation or two away from oral transmission, e.g. some
broadsides.
28
See Bishop, Julia, ‘“Dr Carpenter from the Harvard College in America”: an introduction to James Madison
Carpenter and his collection’, Folk Music Journal, 7:4 (1998), 402–420 (the rest of the issue is devoted to
aspects of the Carpenter collection).
29
For more information on Mathieson, and several of his songs, see Tocher, 43 (Edinburgh: School of Scottish
Studies), pp. 22–39; see also Henderson, Hamish, Alias MacAlias (Edinburgh: Polygon, 1992), pp. 25 and 32.

189
The Singer and the Scribe

song books featuring words like awa, wey, aa or a’, windin’, etc. Until the 1930s, and the
coming of affordable and ‘portable’ recording machinery, the ability to write was essential
for any singer-collector. Scotland has seen widespread literacy for many years and certainly
since the 1872 Education Act nearly every Scot would have acquired some skill in reading
and writing English to a greater or lesser degree, depending on necessity, ability, and actual
school attendance. I specify English because the reading and writing of Scotland’s other
languages and dialects — chiefly Gaelic and Scots — was a much more hit and miss affair,
dependent on individual parochial or village provision. In the case of Scots, the teaching of
writing and reading still lags far behind, where it exists at all. It is worth remembering that
the two skills are quite different and often develop to different degrees.30
I would contend that, at least in Elizabeth Stewart’s case, the confidence to write texts in
a language she was not formally taught to write comes from her unshakeable belief in the
tradition itself. Since the songs are a central pillar of her individual, family and community
identity, they have an authority to define what is correct. The hand-written text, therefore,
reinforces the cultural veracity of the songs. This is in contrast to the assumption made by
early collectors that an oral version is but a broken down version of a printed, read
‘correct’, original and that a print version is seen by singers as authoritative. The former is,
of course, a general notion with which most Westerners live through our reliance on the
perceived authority of newspapers and print in general.
Where then, does an interested singer-collector in the North-East of Scotland look for
material? Apart from oral sources, there are old collections, broadsides, song-books, music-
hall performances, radio and, latterly, audio recordings. All of these are seen by traditional
singers as authorities to one degree or another. Some will privilege a source singer, others a
book or a recording; sometimes a singer will privilege a singer as the source of one song, a
text for another. What matters is that there is an external source to which to appeal for an
authoritative version.
The North-East of Scotland is well served with song publications, from those of the
Christies of Monquhitter to that of John Ord 31 and, of course, the Greig -Duncan collection.
Material was therefore readily available to those of sufficient means, or, in the early part of
the twentieth century, to readers of the Buchan Observer, which carried Gavin Greig’s
regular song column.32 Not unexpectedly, then, Elizabeth Stewart has a fine collection of
ballad and song-books. In later years, Elizabeth has sometimes used her books to look at
other versions of the family’s songs, to check historical references and to confirm the
song’s importance by noting its appearance in a published work. On the surface, all three of

30
Atkinson discusses the question in a little more depth and offers annotated references, in English Traditional
Ballad, p. 17–18 and notes.
31
Christie, William (1817–1885) and William Christie, Traditional Ballad Airs: Arranged and Harmonized for the
Pianoforte and Harmonium, from Copies Procured in the Counties of Aberdeen, Banff, and Moray
(Edinburgh: Edmonston & Douglas, 1876–81). See also, Alburger, Mary Anne, ‘The Christie Family of
Monquhitter: preservation and “refinement” of traditional music and song’, Northern Scotland, 18 (1999),
117–33 and Ord, John, Bothy Songs & Ballads of Aberdeen, Banff & Moray, Angus and the Mearns (Paisley:
Alexander Gardiner, 1930; reissued John Donald, 1990).
32
Collected in: Grieg, Gavin, Folk-Song of the North -East (Peterhead: Scrogie, 1914; reprinted Hatboro, PA:
Folklore Associates, 1963).

190
 Thomas A. McKean

these uses would seem to privilege the text in a familiar way, but when I ask Elizabeth
about it, she is unequivocal: Lucy’s version of the song is superior in every way, as is the
contextual narrative that accompanies it. The privileging of songs passed down within the
family is a common enough theme,33 as the personal and contextual associations clearly
derive from the unwritten social traditions surrounding the songs. If one accepts David
Buchan’s proposal regarding Anna Gordon that ‘it is only when a person ceases to be re-
creative along traditional lines and accepts the literate concept of a fixed text that he or she
can no longer be classed as oral’ (p. 64), we can say that Elizabeth’s tradition is ‘oral’. As
we shall see, she has not accepted the ‘fixed text’. Whatever the case, I do not believe that a
simple dichotomy — between re-creative/oral and fixed-text/verbal — can be applied to
anyone, whether they be Anna Gordon in the eighteenth century, or Lucy and Elizabeth
Stewart in the twentieth.
The fact that a song that Lucy knew, and taught to Elizabeth, can be found in the books
perhaps lends authority to Lucy, to the family, and to the oral tradition as a whole. Though
this process confers status on the family repertoire, it is not the texts themselves that lend
authority, but merely the fact that such a thing exists at all. Primacy is given to the oral and,
in Elizabeth’s case, to Lucy’s version, of the song: ‘*I never look at them. I like tae hae
them, bit they’re niver lookit at’. Because of the centrality of this process, we may therefore
say that the workings of Elizabeth’s immediate family tradition are still oral in nature, and
fundamental to family cultural identity, even though many songs show strong relationships
with parallel print culture. Print material has probably not been an important source of the
Stewart repertoire for the last century or so. The uses of literacy discussed here, therefore,
relate to the family’s transient tradition of the last generation or two.
Within Elizabeth’s and Lucy’s oral-song relationship, let us examine a few verses of
two of their classic ballads, noting levels of correspondence. According to Child, The Jolly
Beggar (Child 279) was first printed in 1751, in The Charmer, though he feels the Scottish
version to be far superior, and later in Herd’s 1769 collection. There is also a similar
narrative to be found in Pepys’ collection c. 1670-75. Only one of Child’s texts is similar to
the Stewarts’ version as opposed to about a third of those in Bronson’s Traditional Tunes of
the Child Ballads.34 Most versions in the Greig -Duncan collection do not share the famous
chorus, used by Byron in ‘So we’ll go no more a-roving’, preferring instead, ‘Wi his fal an
his dal an his dandy o/ Wi his teerin owre an eerin oure an andy o’, as couplet refrains.35
The two women’s singing styles are radically different, a measure of the role of
individual creativity within tradition. Elizabeth is very aware of the stylistic differences,
describing Lucy’s as ‘plain’ and her timing as leisurely and stately.36 This highlights that it

33
See Atkinson, p. 29, for a brief résumé of the tendency.
34
For full references, see Child, and Bronson, Bertrand H., The Traditional Tunes of the Child Ballads, with Their
Texts, According to the Extant Records of Great Britain and America (Princeton: Princeton University Press,
1959–1972).
35
For North-East examples similar to the Stewarts’, see Mrs Gillespie’s contributions, noted January 1909 in
Greig-Duncan, vol. 2, p. 299 (versions E and G). ‘The tune given in the books’, Greig writes, ‘is less
objectionable than usual’, Gavin Greig, article 30.
36
Questions of style being virtually impossible to deal with on the printed page, I leave readers themselves to
listen to the two singers on the two recordings cited above.

191
The Singer and the Scribe

is the process — the fact that she has learned the songs, values them, feels them, and
understands them on a deep level — that perpetuates the tradition. Elizabeth does not need
to copy the songs slavishly, nor Lucy’s singing of them, to carry on her aunt’s legacy.
The Jolly Beggar (Child 279)
Lucy Stewart (1959) Elizabeth Stewart (1994)
There was a jolly beggar, There was eence a jolly beggar,
And a-beggin he wis boun, And a-beggin he wis boun,
An he’s taen up his quarters, An he’s taen up his quarters,
In some landward town. In some langward toon.

An we’ll gang nae mair a-rovin An we’ll gang nae mair a-rovin
Sae late intae the nicht Sae late untae the nicht
We’ll gang nae mair a-rovin, An we’ll gang nae mair a-rovin,
Lat the mune shine e’er sae bricht. Let the meen shine e’er sae bricht.

He widnae lie in barns, He widna lie in barns,

Or he widnae lie in byres, Nor he widna lie in byres,
In ahint the ha door, Bit in ahint the ha door,
Or before the fire. Or jist afore the fire.

The beggar’s bed wis made at een, Oh the beggar’s bed wis made at een,
Wi guid clean strae an hey, Wi gweed clean strae an hey,
An just ahint the ha door, An just ahin the ha door,
An there the beggar lay. There the beggar lay.

Elizabeth herself comments on her tendency to introduce verses with a more than
occasional ‘fill-in’ word, such as ‘O’ or ‘An’. This perhaps derives from her more
emotional, extrovert, dynamic style, which virtually demands attention. Lucy’s singing, in
contrast, invites the listener into the song, the ballad-world, in their own time, so to speak.
The singers’ words correspond almost exactly, the differences partly due, no doubt, to
natural dissimilarities in any two renditions. There are a few variations in dialect, Lucy’s
‘mune’ versus Elizabeth’s ‘meen’ in line 8, for example, the latter a late nineteenth-century
form. A curious feature, also found in other songs of Lucy’s is the frequent use of English
words, e.g. ‘town’ (l. 4), where Elizabeth sings — and one would expect — ‘toun’
(pronounced ‘toon’). Similarly in verse 2 of Binnorie, Lucy sings the English ‘take’ where
her niece sings the more natural ‘tak’. Hamish Henderson has written of ‘the flexible
formulaic language of the older Scottish folk song — which grazes ballad-English along the
whole of its length, and yet is clearly identifiable as a distinct folk-literary lingo’. 37 Much of
the existence of such a dialect can surely be ascribed to the semi -migratory nature of
traditional ballads from other areas within the ‘single great ballad zone’ of England, Scots-
speaking Scotland and English-speaking Ireland, as Henderson goes on to say,38 but I think
one must also recognise the growing influence of spoken English as a lingua franca in the
era of the great collections of Scottish song — the late eighteenth-century to the present

37
Henderson, Hamish, ‘The Ballad, the Folk and the Oral Tradition’, in The People’s Past, ed. by Edward J.
Cowan (Edinburgh: Polygon, 1980, reissued 1990), p. 82.
38
Henderson, ‘The Ballad, the Folk’, p. 82.

192
 Thomas A. McKean

day. Oral versions collected in the age of sound recordings are often more Anglicised than
the broader Scots of older written sources. There is an undoubted tendency towards
Anglicisation-in-performance for the benefit of outsider collectors, an effect operating in
Lucy’s case and, I believe, in similar examples involving other singers. Elizabeth Stewart
agrees, saying that Lucy modified her language considerably for visitors like Goldstein, and
that she herself served as interpreter with other contributors on some occasions. To my
mind, the is sue is largely one of cultural self-assurance. As English made inroads into the
North-East consciousness, confidence in local dialect and culture wavered, yielding a more
Anglicised register in song. In recent years, especially for someone as proud of her family’s
culture as Elizabeth Stewart, renewed cultural pride allows singers to extend their natural
speaking dialect into the more formal setting of a classic ballad. Jock Duncan’s singing is a
classic example of this tendency; the dialect in his ballad versions is scarcely different from
that of his bothy songs (cornkisters) which originated in the late eighteenth to mid
nineteenth centuries. Composed by fee’d farm servants, hired on a six-monthly basis, bothy
songs discuss farmwork and farm life, enumerating the characteristics, good and bad, of the
farmers and the other workers. They are almost invariably sung, even for outsider
collectors, in the broadest of dialect.
The Two Sisters (Binnorie, Child 10)
Most versions of The Two Sisters begin, ‘There were two sisters lived in a bower’, ‘There
were two sisters lived in a ha’, or ‘There were three ladies playin at the ba’. Not one of
Bronson’s, Child’s, or Greig’s and Duncan’s ninety-seven, twenty-seven, and twenty-two
versions, 39 respectively, feature the Stewarts’ opening line ‘O there were twa sisters lived
in this place’, or Lucy’s refrain for that matter. At verse three, their version begins to match
with many collected in North America, and in Britain by Gavin Greig, among others. Lucy
was three years old when Greig began collecting seriously, just seven miles down the road,
but he was apparently too busy with all the immediately local material to move further
afield.
‘The Swan Swims Sae Bonnie o’
Lucy Stewart Elizabeth Stewart
Oh there were twa sist ers lived in this place, There were twa sisters lived in this place
Hey o ma nannie o, Hey oh, binnorie, oh.
Een was fair an the ither was deen, Een o them wis fair an the ither wis deen [dark]
An the swan swims sae bonnie o. An the swan swims sae bonnie o

Again the correspondences are close, though there are expected, natural variations, and the
first refrain line is, surprisingly in a conservative tradition, completely different. The two
versions are not quite as close as Elizabeth feels they are, but this highlights the fact that the
essential factors of traditionality are cultural fidelity, and the tribute paid in the
remembering and singing of the song, rather than slavish word-for-word reproduction.
Textually, then, Lucy’s and Elizabeth’s versions of songs reveal a high degree of exacti-
tude, with only minor variations. The differences, as might be expected within a single
family and between two emotionally close people, are of the same general type as minor
differences between oral and printed texts taken from oral sources, though, as in the case of

39
Greig-Duncan, vol. 2, song 213, pp. 76–92.

193
The Singer and the Scribe

broadsides, these may be more extreme if the song has been through several generations of
print. Of course, a great many such texts are originally from oral sources, with various
amounts of doctoring, as opposed to songs which have their ultimate origin in a literary
tradition.40
Perhaps the most common printed source for songs and ballads in Scotland is the
broadside; there were numerous publishers, primarily in Dundee, Glasgow and Edinburgh
and their material undoubtedly reached up into the North-East.41 One of several songs in
Lucy’s repertoire undoubtedly from the broadside tradition is the story of murderous Dr
Pritchard. There are two songs in Bodleian Library Broadside Collection, being,
The lament of Mr Taylor for his wife and daughter, who were cruelly poisened [sic] by the daughters husband,
Dr Pritchard, who is now lying under sentence of death, in the prison of Glasgow to be executed on the 28th
42
July, 1865.

The case was sensational and there was a spate of articles, broadsides, poems and sermons
about the murders and the execution, Glasgow’s last public hanging.43 I have yet to trace a
specific source for the Stewarts’ version, which is quite different from both Bodleian
examples. The date of 1865 means, I think, that Lucy Stewart is unlikely to have learned it
directly from a broadside. Perhaps a grandparent did, but that is really immaterial. By
Lucy’s time, the song was probably in oral tradition, having become detached from its print
source. In any case, the fact that it has such ultimate origins is not really germane to a
primarily contextual study of its oral life, centred in a period a hundred years later. The
song is particularly important to Elizabeth; its content moves her and she learned it from her
beloved aunt.
This song neatly demonstrates the interaction between print sources and the oral
tradition: a song can exist in parallel worlds simultaneously, in this case crossing only at its
point and time of origin, or shortly thereafter. Indeed there is some case to be made for
there now being two totally separate, individual songs, one which lives in an oral world, the
other in the two dimensional, unchanging world of print. Or to put it another way, without
privileging either one, a living, evolving one and a fixed, static version. These two worlds

40
For a look at the instability of broadside printed texts, which show many of the same supposedly oral unique
types of variation and localisation as oral versions, see Atkinson, p. 25.
41
For look at an Aberdeen broadside seller, see Brown, Mary Ellen, ‘An Excursus into Historical Documents: the
life and songs of Charles Leslie’, in The Ballad Today: history, performance, revival, ed. by Georgina Boyes.
th
Proceedings of the 13 International Folk Ballad Conference, Société Internationale d’Ethnologie et de
Folklore, Kommission für Volksdichtung (Doncaster and Addiscombe, England: January Books, 1985), pp.
72–79. Sellers and printers are the focus of Hindley, Charles, The Life and Times of James Catnach, Late of
Seven Dials, Ballad Monger (London: Reeves & Turner, 1878; reprint Welwyn Garden City: Seven Dials
Press, 1970) and Shephard, Leslie, John Pitts: ballad printer of Seven Dials, London, 1765-1844 (London:
Private Libraries Association, 1969). For a general introduction to broadsides, see Shephard, Leslie, The
Broadside Ballad: a study in origins and meaning (London: Herbert Jenkins, 1962).
42
Oxford, Bodleian Library, 2806 c.14 (52); the second song is Bodleian 2806 c.11 (60).
43
See Roughead, William, Trial of Dr Pritchard , Notable Scottish Trials (London: Sweet & Maxwell, 1906), p.
304. The book contains full transcripts of the trial, along with pictures of Pritchard and the other players in the
drama. Sermons on the doctor’s sins can be found on the Internet even today.

194
 Thomas A. McKean

should be considered ‘not as mutually exclusive cultures’, but to derive from, ‘mutually
supportive mental habits’. 44
More recent print sources of songs in the North-East of Scotland include the published
work of Willie Kemp, George S. Morris and other music-hall practitioners of the bothy
ballad tradition. By the end of the nineteenth century, and into the twentieth, the short-lived
tradition was taken over by music-hall composers whose songs are self-conscious, often
humorous, parlour parodies of the rough farm songs. As the lifestyle itself passed away, so
the nostalgic, golden sunset appeared in its song tradition. Kemp, Morris and others were
well-known throughout the North-East and their work appeared on 78s and in book form
quite early on.45 Indeed some of Jean Stewart’s professional work was accompanying
Willie Kemp, John Mearns, George Elrick, Ludovic Kennedy and other artists in their BBC
radio work in Aberdeen. Many Aberdeenshire performers sing some of their songs, which
are now generally accepted as traditional; many have entered the oral tradition, either by
this means or by radio and recordings, to the point where it is now virtually impossible for
anyone to maintain convincingly that they are, or are not, traditional songs. It is clear, then,
that the Stewarts would have had exposure to a wide range of material, also in the past
because of a Travellers’ itinerant lifestyle. In practice, however, they tended to stick to their
own tradition for the big ballads and lyric songs that had been in the family for some time.
One curious song Lucy’s and Jean’s immediate forebears may have learned directly
from the music-hall tradition is The Russian Jew. This is an unusual title for rural Aber-
deenshire, to say the least, and a very misleading one. It derives from the macaronic 1890s
‘Says I, “Ciamar a tha sibh an diugh?”’, which pokes fun at the Highlander by parodying
his English, as do a number of songs in the Greig -Duncan collection (one verse of this song
appears in volume 8). The original song ends each verse with the title line, meaning ‘Says I,
“How are you today?”’ in Gaelic. In The Russian Jew, this line has been transformed into
‘Says I, come a Russian Jew’, or ‘Says, Here comes a Russian Jew’, which sounds very
similar. The first field recording of this was made in the late 1950s by Kenneth Goldstein,
who told me in 1995 about hearing it for the first time from Elizabeth, her sister Jane and
their aunt Lucy: ‘*They had probably never even seen a Jew before, and here am I a
Russian Jew from New York recording this song from Aberdeenshire Travellers!’ 46
The most recent type of source material is audio recordings (such as 78s), radio and
television. While not literate in the strict sense, they are often given a similar sort of
authority by those in an oral tradition, much as we might give The Beatles’ original
rendition of a song authority over a cover version of the same song. I call these a kind of
literate source as they are reproducible and can serve as a permanent template, which is the
salient element of a printed text. Again, I draw the reader’s attention to the main function of
such sources, whether they be singers, books, or recordings: they serve as an external

44
Atkinson, p. 18, drawing on the work of Engler, Balz, ‘Textualization’, in Literary Pragmatics, ed. by Roger D.
Sell (London: Routledge, 1991), p. 185.
45
See, for example, Kerr’s ‘Cornkisters’ (Bothy ballads) as Sung and Recorded by Willie Kemp (Glasgow: James
S. Kerr, [1950]), or Kerr’s Cornkisters: Bothy Ballads (Glasgow: James S. Kerr, 1951).
46
Personal communication.

195
The Singer and the Scribe

authority for the singer. It is worth remembering, too, that once someone has learned a song
it is affected by the same rules that govern ordinary oral transmission and variation.47
Literacy also makes its presence felt in the song making tradition of the North-East.
Elizabeth Stewart has made a number of songs, including a traditional-style ballad, based
on historical events of the fourteenth century.48 Her mother was famous for her satirical
songs about fellow villagers in Fetterangus, including one that lead to a threatened arrest,
and her aunt has composed several verses, written on the backs of Christmas and birthday
cards and kept in a canister atop her piano. Writing plays a prominent role here, with the
song often being written down as soon as it is complete in the mind of the composer.
One singer, Bill McKinnon of Peterhead, also recorded by Kenneth Goldstein in 1959,
started making songs only in 1993. Unusually for a composer within the tradition, he
always writes them down as he makes them, rather than waiting until they are complete
and, in fact, he is often unable to sing them without the paper in front of him. He has told
me that he makes the verses while walking, motion being a common feature of the process
of traditional song composition, so at some stage they do exist only orally. I think the paper
copy is therefore largely necessary for the performance of the song, especially in front of
others.49
I have already mentioned Willie Mathieson, who wrote all his songs in three manuscript
notebooks, but in the course of my years in the North-East of Scotland, I have come across
numerous singers who write their traditional songs down, some of them the most capable
singers I know: Bill McKinnon, Jane Turriff, and Bill’s sister Jean Mathew, to name a few.
Elizabeth Stewart is one of these, though her mode of operating is rather different. While
McKinnon, Turriff and Mathew all write and wrote notebooks (or in Jane’s case Christmas
cards) full of songs, Elizabeth is constantly renewing her notebooks, starting a fresh one,
usually coincidentally with a public concert or competition, and occasionally to pass the
time. She does have one main book containing transcriptions of Lucy’s songs, with
occasional melodic outlines in sol-fa notation, but it would be misleading to say that she
keeps a single notebook. She also has two versions of some songs. Unusually, Lucy
sometimes had two very different versions, text and melody, of the same song. Like most
traditional singers, however, Elizabeth only ever sings one version.

47
See McKean, ‘Gordon Easton’s “The Aul Beggarman”’ about a singer drawing on a recorded source before
making a song his own. For the classic discussion of adaptation, variation and localisation in traditional song,
see Coffin, Tristram P., The British Traditional Ballad in North America, revised edition with a supplement by
Roger de V. Renwick, Bibiliographical and Special Series published through the cooperation of the American
Folklore Society (Austin: University of Texas Press, 1977), introduction, pp. 1-19.
48
For an examination of this song and its traditional themes, structure and creation, see McKean, Thomas A., ‘The
Making of Child 306’, in Bridging the Cultural Divide: Our Common Ballad Heritage/ Kulturelle Brücken:
Gemeinsame Balladtradition, ed. by Sigrid Rieuwerts and Helga Stein, 28. Internationale Balladenkonferenz
der SIEF – Kommission für Volksdichtung (Hildesheim; Zürich; New York: Olms, 2000), pp. 273–290.
49
For an interview with Bill McKinnon, see The Broken Fiddle 4 (Macduff: Aberdeenshire Council, 1995), pp.
56–61; for recordings of two of his songs, see North-East Tradition 1: New Recordings from the North -East
Folklore Archive (Kingskettle: Sprin gthyme SPRC 1040, 1994). For a description of the role of movement in
traditional song composition, see McKean, Thomas A., Hebridean Songmaker (Edinburgh: Polygon, 1997),
pp. 114-123.

196
 Thomas A. McKean

The simplest function of writing songs down is as an aide memoire, not that Elizabeth
ever refers to the words in the book. Rather, it is the process of writing that revitalises the
words in her mind, and restores the flow, reminiscent, perhaps, of listening to Lucy singing,
somehow externalising the source of the song again. I saw an example being made recently
when, just prior to the Strichen Festival singing competitions, Elizabeth sat in the hall
writing out the song she was planning to sing, reinforcing it in her mind. Two Pretty Boys
(The Twa Brothers Child 50), Elizabeth Stewart. (Figure 6)
When Elizabeth sings the song, the last line is repeated, line 4 ends with ‘boy’, line 7
ends with ‘wood’, and verse three runs,
So they went down to the merry green wood,
Tae try a wrestling for,
Big brother John took oot his little penknife,
And stabbed William to the ground.

She does not write out every word, but used her own abbreviations and shortened lines,
fully aware of their meanings. Sometimes, she says, she just writes down the first two
words, e.g. ‘An they’, and that is enough to trigger her memory and assure her that she
remembers the line. This shows that the process of writing is the most important function of
such a written text.
The most important of Elizabeth’s notebooks is one made in the late 70s when Lucy was
in her declining years. Elizabeth realised that it was an important time and took down
dozens of songs from Lucy’s dictation. The volume holds particular significance for her:
‘*It’s had Lucy’s hands on it’. (Figure 7)
Writing down the songs that meant so much to both of them brought them closer and
would have reassured Lucy that her legacy would be preserved, not that she necessarily
considered the matter consciously. Elizabeth honours her aunt’s memory with the sort of
monument that print texts are usually thought to be. As proof of this, I offer the fact that
Elizabeth is keen to have a book in print about her Lucy, Jean and herself.50 Though she
comes from a peerless oral culture, few of her immediate fa mily have taken up the reins; a
printed legacy is therefore the only monument that matters to the wider literate-minded
public. Such a document serves to confirm the legitimacy of the song, blunting the two-
edged sword of orality, by giving it parity with versions found in the Greig-Duncan
collection’s eight weighty volumes and others already mentioned, although with the proviso
that it is the text’s mere existence that is important, not the version itself.
For Elizabeth, writing songs down also gives them a context for performance when the
domestic setting in which they were passed on to her no longer exists. In the silent
performance, hearing the song in her head as she writes, she is a conduit in the dialogue
between the written song and the oral version. The page offers feedback in a way that an
audience or, in the past, a community might have. It is an inaudible performance, but given
that she does not refer to the written text again, we may consider the act of writing a
performance — like singing — in itself.

50
Elizabeth is currently working with Alison McMorland on such a book, which will contain family history, songs,
music and photographs relating to the Stewarts of Fetterangus.

197
The Singer and the Scribe

But what of the apparent differences between a text and a performance? Folk song is
said to be a very conservative tradition, for example in the retention of certain words, or at
least homonymic phrases, even if the singer does not understand them (e.g. ‘he tirled at the
pin’ in American versions of Scottish ballads, or the macaronic phrase, ‘Ciamar a tha sibh
an diugh?’), but despite this overall conservatism, there is a strong tendency for traditional
singers to individualise, to internalise a song. Having, or learning the text of a song does not
mean that it will sound the same each time, or that the text will remain the same after the
process of internalisation has taken place. An example of this difference can be found in
Jock Duncan’s singing of The Battle o Harlaw (Child 163), the text of which he says he
learned from John Ord’s Bothy Songs of the North-East of Scotland,51 with the last verse
remembered from his uncle Charlie’s singing: ‘Well, I got a version oot o a book, bit I got
the style o singin’t fae ma uncle, Charlie Duncan. At’s e wey he likit tae, tae dee’t’.52 He
learned the song, or at least stimulated the re-learning, by looking at the book, but he adapts
the text to reconstruct his uncles’ version and in the process also makes it his own. In re -
building his uncle’s version from memory, around forty percent of Ord’s text gets
recomposed as he adapts the song, sometimes radically, consciously or subconsciously, to
fit his own idiom. I suspect that his version differs from his uncle’s as much as it does from
Ord’s, which he has obviously used as a structural key. Elizabeth Stewart clings to Lucy’s
individuality, as Jock Duncan clings to Charlie’s. As we have seen, however, they both
recast songs, using their own experience, making them their own. This is as good a
definition of traditional singing as one is likely to create.
On balance, I think we may safely say that literacy can have much to do with the quality
of a text, but very little to do with the quality of a song, where song is taken to mean the
totality of performance, or that which creates a compelling version of a song. It is obvious
that one need not be non-literate to possess an excellent oral version of a song. Nor is
literacy connected with one’s ability to render said song in a moving way. They are
separate, but related skills, much like reading and writing; the ability to do one well does
not ensure fluency in the other. Literacy and orality are therefore not diametrically opposed,
as is usually assumed, and often asserted in scholarly work on the oral tradition, but can
operate in parallel, with little or no detrimental effect to each other. This is no surprise, for
as the oft-quoted Alan Lomax wrote in 1951,
The Scots have the liveliest folk tradition of the British Isles, and paradoxically, it is also the most bookish….
Everywhere in Scotland I collected songs of written or literary origin from country singers; at the same time, I
53
constantly encountered learned Scotsmen who knew traditional versions of the great folk-songs.

In fact, all the finest ‘traditional’ singers one can think of today — Jock Duncan, Elizabeth
Stewart, Gordon Easton, Norman Kennedy, Sheila Stewart — and the finest ‘revival’
singers — Ray Fisher and Alison McMorland, for example — are all highly literate in both
reading and writing terms. You would be hard pressed to say that they are not as good as

51
Ord, Bothy Songs, pp. 473–75.
52
Tape c1994.63.B in the North-East Folklore Archive, Mintlaw, Aberdeenshire. Summaries and transcriptions of
my 1993–1995 audio recordings of North-East tradition — ‘The Banff and Buchan Collection’ — can be
found at http://www.nefa.net/.
53
World Library of Folk and Primitive Music, compiled and ed. by Alan Lomax, Vol. 3 Scotland, The Alan
Lomax Collection, The Historic Series (Cambridge: Rounder 1743, 1998), booklet, p. 6.

198
 Thomas A. McKean

non-literate singers of the past, except through privileging the latter for the sake of it, or for
the sake of some hypothetical, purely oral mental process.54 The coming of literacy, then, is
not a death knell to the oral tradition. It is the arrival of another tool, used by singers like
Elizabeth Stewart to reinforce deeply held convictions about their oral material.
What role did literacy play in Lucy Stewart’s life of song? A small one, as far as I can
tell. Her ability to read was useful in daily life, as was a talent for figures in her work as a
dealer and shopkeeper, writing probably less so. She was undoubtedly fiercely intelligent,
retaining family genealogies, birth, marriage and death dates and detailed historical
information, such that she was an acknowledged authority in the village. On one occasion,
she told a Glasgow University history professor about one of Robert the Bruce’s
illegitimate sons. He was sure she was wrong and told her so, but after returning to
Glasgow, and researching the question, he telephoned Lucy to admit his mistake. 55
Naturally, such a story has the useful function of reinforcing faith in the oral tradition
through future generations. I think we can safely say that Lucy’s knowledge of tradition
existed almost entirely independently of her literacy.
What then of Jean’s and Elizabeth’s literacy? Jean was very well educated and would
have needed rather different skills from Lucy’s to pursue her broadcasting and teaching
work. For her professional life, textual and musical literacy were essential, though she
certainly never put on airs about this, nor turned her back on her tradition.
Jean’s abilities are reflected in Elizabeth, who was also a prize-winning pupil at school.
In addition to using these skills as her mother did, Elizabeth harnesses them in the service
of her tradition, whether performing or teaching: she uses them to reinforce her confidence;
she uses them to pay tribute to her aunt Lucy; she uses them to pay tribute to her wider
family and their traditions; she uses them to find out more about the songs she sings,
heavily privileging the oral knowledge passed down to her. An individual’s literacy, then,
does not have a great effect on the traditions of the Duke Street Stewarts of Fetterangus.
That is not to say that general community literacy does not; there is the obvious effect on
repertoire of broadside ballads for example. The fact that a song entered the repertoire from
print, however, has little or no bearing on how it is handled by oral tradition. Even those
songs which emerged from print culture have often undergone a thorough dose of the folk
process and come out the other end as oral ballads. Initially, therefore, individual examples
of literacy have an effect, but over time these individuals’ actions are polished by a
functional oral tradition, one that gathers influences of every kind, including those of
printed and written song-texts. The Stewarts of Fetterangus show us that this same process

54
Jock Duncan, Ye Shine Whar Ye Stan (Kingskettle: Springthyme SPRCD 1039, 1996) and Tae the Green Woods
Gaen (Ellon: Sleepytown SLPYCD010); Gordon Easton, In Sicht o Mormond (Aberdeen: Grampian Society
for the Blind, 1993) and several tracks on The Bothy Songs and Ballads of North-East Scotland, 3 vols (Ellon:
Sleepytown SLPYCD001, 006 and 011); Norman Kennedy, Songs and Stories of the Old People and More
Songs… (Sioux Falls, SD: Golden Fleece Publications, 1990); Sheila Stewart, From the Heart of the Tradition
(London: Topic TSCD515, 2001); Ray Fisher, Traditional Songs Of Scotland (Wotton-Under-Edge: Saydisc
SDL 391); Alison McMorland, The Belt wi Colours Three (London: Tangent TGS 125), Cloudberry Day:
Scots Songs and Ballads (Kilmarnock: Living Tradition LTCD1003) and (with Geordie McIntyre), Rowan in
the Rock (Kilmarnock: Living Tradition LTCD3002).
55
Elizabeth Stewart, personal communication.

199
The Singer and the Scribe

of synthesis, now with even more influences, is still happening. Reasons for using writing
and reading, and other media skills, can vary widely, but they have all become tools in the
repertoire of an educated singer-collector.
The importance of writing to Elizabeth Stewart has more to do with the symbolic power
of song than the power of the skill itself, or of texts. As we have seen, it reinforces family
history and continuity; it symbolises links with the past. The written texts and the large
number of tapes of Lucy and her mother serve as a legacy to the next generation. It is here
that literacy may come into its own, for if the culture of performance, transferred directly
from person to person, is lost, then all that will be left are the documents of tradition.
Without the cultural performance contexts of the past, or some successor, such a the folk
club, those documents are little more than fossilised bones.
Elizabeth Stewart’s song texts are a literate link between her own and Lucy’s oral
worlds. Now that Lucy herself is gone, these manuscripts help to make her traditions
concrete once more. While they are certainly no substitute for Lucy herself, they do serve as
a representation of her world that is external to Elizabeth and that, I think, in some way
provides a crucial form of comfort for that which she has lost.
Jean Stewart died in 1962, Lucy in 1982. Their gravestone in Fetterangus cemetery
stands as a memorial to their life and legacy of traditional music and song. And curiously
enough, there is an echo of literacy in the elegy on the stone. (Figure 8)

200
 Thomas A. McKean
List of Illustrations
Fig. 1. Louisa ‘Lucy’ Stewart, Fetterangus, Aberdeenshire, Scotland, ca. 1959 (photo
Kenneth S. Goldstein, courtesy Rochelle Goldstein).
Fig. 2. Elizabeth Townsley, ‘Aul Betty’, mother of Lucy and Jean Stewart, mid 1930s
(postcard, copied by Kenneth S. Goldstein, courtesy Elizabeth Stewart).
Fig. 3. Jean Stewart with her accordion class, New Deer Public Hall, early 1950s
(courtesy Gordon Easton).
Fig. 4. The People’s Journal, March 26, 1955.
Fig. 5. Elizabeth Stewart performing in America, 1995 (photo Thomas McKean).
Fig. 6. ‘Two Pretty Boys’ (courtesy Elizabeth Stewart).
Fig. 7. The title page from Elizabeth Stewart’s notebook of her aunt’s songs (courtesy
Elizabeth Stewart).
Fig. 8. Detail from the headstone of Lucy, Jean and Robert Stewart (photo Thomas
McKean).

201
The Singer and the Scribe

Fig 1.

202
 Thomas A. McKean

Fig.2

Fig.3

203
The Singer and the Scribe

Fig. 4

204
 Thomas A. McKean

Fig. 5

205
The Singer and the Scribe

Fig. 6

206
 Thomas A. McKean

Fig. 7

Fig. 8

207
This page intentionally left blank
List of Contributors

In the following list those marked with an asterisk (*) spoke at the colloquium of 27th
May, 2000.

*Roderick Beaton, Professor of Byzantine Studies, King’s College, London

*Philip E. Bennett, Reader in French, University of Edinburgh
Manuel da Costa Fontes, Professor of Spanish and Portuguese, Kent State University,
Ohio
*Charles Duffin, professional musician, Edinburgh
Richard Firth Green, Professor of English, Ohio State University
*Thomas A. McKean, Research Fellow, Elphinstone Institute, University of Aberdeen
William Layher, Assistant Professor of Germanic Languages and Literatures,
Washington University in St Louis, Missouri
Huw Lewis, Senior Lecturer in Spanish, University of Edinburgh
Ad Putter, Reader in English, University of Bristol
Ekaterina Rogatchevskaia, British Library, London
*Margaret Sleeman, Honorary Research Fellow, Department of Hispanic Studies,
University of Aberdeen
*Roger Wright, Professor of Spanish, University of Liverpool

209
This page intentionally left blank
Index Aspin, Isabel S.T 118 -19n52, 120n56
Astakhova, A.M 29n36, 31n41,
32n48, 33n51, 34nn52 -3
Aberdeen 183, 185, 188, 195 Atatürk, Mustafa Kemal 154
AbnÁmar 175, 180, 183n15 Atero, Virtudes 58n20
Acaponeta, Mexico 179 Atkinson, David 181-2, 184nn15 - 6,
Acta Sanctorum 72, 73nn16&18 190 n30, 19133, 194n40, 195n44
Adam de la Halle, Jeu de Robin et Atlakviða 100n30
Marion 117 Attias, Moshe 156, 159n11, 165-7
Aelion, Mrs Estrea 10, 153-7, 165- Aubert, R 79n44
168n18 Aubrun, Charles V 174n4
Aelion, Saltiel Jehuda 155 Aucassin et Nicolette 117
Aesop, Fables 10 Audigier 117
Aguado, Berta 177 Azbelev, F.N. 28n27
Aguirre, J.M 37n3, 44 Azores, the 57-8, 62, 66, 171
akritika 9, 14, 17-20
Albertus Pictor 91-2, 94, 98 -100, 102 Badel, Pierre-Yves 117n48
Alburger, Mary Anne 190n31 Bailey, J 26n18
Alesha Popovich 30 Bakhtin, Mikail 11, 124 -33
Alexander, Tamar 169 Balashov, D.M 24, 30n41, 32n45,
Alexander VI, pope 180 35n54
Alexiou, Margaret 16-17n13 Ballad and song titles
Alexiou, Stylianos 19n17 Danish and Swedish
Algarve, the 58,59n25, 61n29 Holger Dansk og Burmand 89 -
Alhambra, the 177 102
Alonso, Martín 47n19 Knight Transformed into a
Alvar, Manuel 159n11 Hart, the 77
Ambraser Liedbuch 85 Marsk Stig 77
Amel’kin, A.O 28n29 Runsivalstriðið 110n24
Andersen, Flemming G 24, 70n6, Dutch
136, 139-43, 150 Brandenborch 77
Andrea da Barberino 63n41 Daghet inden Oosten, het 75
Andreev, N.P 24 Hadde een Swave een
Antwerp 80, 173 Dochterlijn, des 82
Antwerps Liedboek 74 -5, 78 -81, 82- Fier Margrietken 77-87
84, 87 Heer Danielken 75
Arant, P. M 25n18 Machreitje 78-9
Archivo Menéndez Pidal 154, Soude een meysken gaen om
160n11, 165-7 wijn 87
Argenti, Philip P 156n4 Te Gherbeken binnen 81
Argo, Arthur 183 -4 Vanden ouden Hillebrant 75
Armistead, Samuel G 61n30, 63n38, Zavelboom, de 82n59
64n43, 67n47, 154, 156n4, 157-8, Finnish
159nn8&11 Nådendal 92
Arteta, Antonio Ubieto 53n9 French
Arvidsson, Ivar 92n10 Avant ier me chevauchoie 112
Askins, Arthur L. -F 62n34
The Singer and the Scribe

Bataille d’Annezin 118 Battle of Harlaw, the 198

Bele Aaliz 110-11 Battle of Otterburn, the 10
Bele Aiglentine 109 Beggar-Laddie 124n7
Bele Doette 108, 111-2 Billy Johnson’s Ball 131
Chançon ferai, qu talenz m’en Bonny Lass of Lochroyan, the
est pris 112 147-9
De Renaut de Mousson 111n29 Boy and the Mantle, the 127
Floovent 52, 62-5 Braes o’ Yarrow, the 124
El mois de mai 112-5 Broomfield Hill 124n5
Thomas Turberville 118 Captain Wedderburn’s
Trailbaston 120-1 Courtship 126
German Childe Maurice 5
Edle Moriger, der 31n42, 76 Clerk Saunders 124
Halewijn 46 Derby Ram, the 126, 133
Verlorne Schuh, der 82n59 Douglas Tragedy, the 5
Wein holen 85-6 Dr Pritchard 194
Greek Earl Brand 79
Ballad of Susan, the 156n4 Fair Margaret and Sweet
Bridge of Arta, the 16 William 140
Charon and the Girl 156n4 Fair Flower of Northumberland
Haunted Well, the 156n4 [Child 9] 81
Maurianos 31n42 Farmer’s Curst Wife 123n3
Song of Armoures, the 19 Friar in the Well 123n3
Song of the Dead Brother, the Frog and the Mouse 126
16 Gest of Robin Hood, the 8n13
Italian Get up and Bar the Door 123n3
Fioravante 63-7 Glasgerion 125
Latin Greasy Cook, the 132
Carmen de Hastingæ Prœlio Gypsy Laddie, the 28n29
116 Here’s to the Grog 131
Carmen de prodicione Herring Song 127n16
Guenonis 116 Hind Etin 141
Russian Hinkie Dinkie Parlez-vous
Alesha and Elena Petrovichna 131n31
30-4 Hunting of the Cheviot, the 10
Avdot’ia Riazanochka 28 I Am a Miller Tae Ma Trade
Dobrynja and AljošAša 31n42 184
Molodets i korolevna 32n45 Jamie Douglas 7
Slovo o polku Igoreve 25 John Barleycorn 126, 133
Tatarskii polon 28 Johnny Scot 142
Scots and English Jolly Beggar, the 191-2
Alison Gross 120n54 Jolly Tinker, the 132
Aul Beggarman, the 186n19, Judas 77, 83
196n47 Keach in the Creel 123n3, 132
Baffled Knight, the 7, 124n5 King Edward the Fourth and a
Barley Mow 131 Tanner of Tamworth 129n23

212
 Index

King Henry 124 Wife Wrapt in Wether’s Skin

Knight and the Shepherd’s 123n3
Daughter 124n7 Young Hunting 125
Lady Isabel and the Elf-Knight Spanish and Portuguese
77n39, 124n5 A Morte do Rei D. Fernando
Lamkin 124 52- 4, 56-7, 63
Lillumwham 127-8 AbnÁmar 175-6, 180
Lizzie Lindsay 124n7 Afuera, afuera, Rodrigo 53, 61
Lord Ingram and Chiel Wyet Bela Infanta 52
142 Bella en misa, la 157
Lord Randal 5 Benardete 165
Maid and the Palmer 127 Blanca niña 125n8
Murder of Maria Marten, the Boda estorbada, la 171
70n6 Cantar de la muerte del rey don
My Father Had an Acre of Land Fernando y cerco de Zamora
126n14 52- 6, 58, 61-2, 67
My Son David 186n19 Celinos 67n47
O Waly, Waly 7 Choza del desesperado, la 158
Old Daddy Fox 126 Claralinda 66 -7
Our Goodman 123n3, 125n8 Conde Claros 160
Overgate, the 131 Conde Olinos, el 48
Plooman Laddies, the 184 Corrido de Juan Carrasco 180-
Prince Heathen [Child 104] 81 183n15
Queen Eleanor’s Confession Doliente estaba, doliente 53,
123 56 -7, 59-60n28
Rantin Laddie, the 142 Doverdjeli / Doverchile 164
Riddles Wisely Expounded 77 Floresvento 62-67
Robin Hood and the Potter 123 Floriseo 42-5
Robyn and Gandeleyn 77 Gerineldos 169, 171-2, 176, 180
Russian Jew, the 195 Landarico 158
Sea Crabb, the 127 Mala suegra, la 157
Sir Patrick Spens 5-6 Mas ariva i mas ariva 177-79,
St Stephen and Herod 77 180n15
Tam Broon 131 Melisenda 173
Tam Lin 123, 140 Morir os queredes, padre 53,
Thomas Rymer 123 60n28
Twa Brothers, the 197 Moza y el Huerco, la 156n4
Twankydillo 131 Muerte del duque de Gandía, la
Two Sisters, the (Binnorie) 5, 180
192-3 Paso del Mar Rojo, el 156
Was You Ever See 131 Pozo airón, el 156n4
When Jones his Ale was New Rico Franco, aragonés 46
131 Robo de Elena, el 156n4
Widow of Westmorland’s Romance del Conde Arnaldos,
Daughter 125, 127 el 7, 37-50, 63-4
Romanceiro do Algarve 58

213
The Singer and the Scribe

Silvana 57-60 Brednich, Rolf Wilhelm 85n72,

Sueño de Doña Alda, el 156 87n75, 93n12
Tormenta calmada, la 158 Brittany 10
Veneno de Moriana, el 62, 156 Bronson, Bertrand H 123n3, 146,
Virgilios 153, 159-60, 165 191, 193
Vuelta del marido, la 156n4 Brown, Mary Ellen 194n41
ballada (Russian) 8, 24 Brown, Anna Gordon, of Falkland
Baltanás Rodríguez, Enrique J 58n20 145 -50, 188, 191
Bartlett, Frederic C 72 Bruford, Alan 181n6, 188n24
Bartsch, Karl 107n10 Buchan, David 24n11, 69n4, 77n35,
Basil Digenes Akrites 9, 18 79n42, 136, 138-9, 145 -7, 150,
Bateson, F.H 63n40 174-5, 181n6, 182, 188n25,
Baud-Bovy, Samuel 18 189n27, 191, 194-5
Bayot, Alphonse 117n45 Buchan, Peter 187
Bárður, Jákupsson 110n24 Bureus, Johannes (d. 1652) 99n25
Beatles, the 195 Burke, Peter 130
Beaton, Roderick 11, 13n2, 17nn11- Burlasová, S 29n33
12, 18n13, 20n18, 21n20 Burns, Robert 126, 189
Bec, Pierre 108, 110n24 Burt, John R 63n39
Belgrade 160 bylina (Russian) 8, 29-35
Bell, Robert 126 Byron, lord George Gordon 191
Benabu, Isaac 169n1, 180n15
Benaim, Clarita 176 Caesarius of Heisterbach 70 -1,
Benmayor, Rina 156n4, 180 81n55, 82
Bennett, Philip E 106n5, 115n36, Canary Islands 159, 171
117n45 Cancionero de Londres (1471-1500)
Bennett, Robert 91nn6-7, 100n29 38, 40-1, 44, 48
Beowulf 25 Cancionero de romances (ca. 1548)
Bernart de Ventadorn 111n30 38, 159-60, 162
Bénichou, Paul 41, 43, 45- 6, 159n11 Cancionero de romances (1550)
Biblia Pauperum 99 38-41
Bishop, Julia 189n28 Cancionero General (1511) 173
Blackmore, Susan 137n7, 138n10 Cantigas d’amigo 108
Bland, C. C. Swinton 70n7 cantilenae 115 -6
Blyau, Albert 78 Cantilena Rollandi 116
Bodleian Library 194 Cape Verde Islands 52
Bohuslän 100n30 Cardozo Freeman, Inez 179
Borchling, Conrad 93n15 Carpenter, James Madison 189
Bottón, Elisa de 165-6 Carranza, Venustiano 179
Boutet, Dominique 117n48 Carrasco, Juan 179-80
Boyes, Georgina 194n41 Catalán, Diego 53n10, 160n11
Børglum 96, 98n24 Catarella, Teresa 173
Braga, Teófilo, Cantos Populares 51, Cawte, E.C 130n29
58n 21, 64-6 Caxton, William 125n10
Brandt, Carl Joakim 94n19 Cervantes, Miguel de, Don Quixote
Braun M 28n29 47

214
 Index

Chambers, E.K 130-1 Cox, E.G 77n36

chansons avec des refrain 112-15 Cox, Harry 184n15
Chanson de Guillaume 117n45 Cressy David 129
Chanson de Roland 25, 93, 106n7, Crete 14, 17, 20
116n37, 116-17 Crónica de Veinte Reyes 53n11, 56,
chansons populaire 8, 107 n8,11 60
chansons de toile 107-110 Crónica Geral de Espanha de 1344
Charlemagne 180 53, 61
Cheesman, Tom 186n19 Crónica Najerense 53
Chernyshev, V.N 24 Culík, J 29n32
Chevalerie Ogier de Danemarche 93 Czajka, H 28n29
Child, Francis James 76n35, 127- Czernik, S 26n21
142n15, 145 -7, 181n2, 183, 186,
191, 193 Danielson, Eva 92n11
Chomsky, Noam 140 Danilov, Kirsha 23n2
Chronyk van Nederlant van Jaere Danon, Abraham 159n11
1027 tot 1525 72n14 Darwin, Charles 137
Christie, William (1817–1885) 190 Davenson, Henri 8, 106n7, 107n11,
Cid, Cantar de Mio 6, 25, 63n39 108n13
Cid, the 55-6, 60-1 Dawkins, Richard 137n7, 138n10
Cintra, Luis Filipe Lindley 53, 61 Débax, Michell 42-4
Civil War, English 129 de Mandach, André 106n5
Clanchy, Michael T 9n17 Delbouille, Maurice 110
Clarke, Dorothy Clotelle 6n6 Delft 75
Clavero, Dolores 53n12 demotiko tragoudi 9, 13-16
Clavus, Claudius 90 Dennett, Daniel 137n7, 138n10, 140
Clédat, Léon 108, 112 Devoto, Daniel 62n34
Clovis, king of France 63 Deyermond, Alan 43n14
Cluzel, Irénée 110n25 Débax, Michelle 43
Cobby, Anne Elizabeth 106n5 Didrikskrönikan 101n33
Codex diplomaticus Neerlandicus Digenes Akrites 18 -20
72n14 Dijle, River 70, 72n14, 74
Coffin, Tristram P 7, 196n47 Dixon, James Henry 127
Cohen, H.L 23n1 Díaz, Porfirio, Mexican president 179
Cohen, Judith R 156n4 Díaz-Mas, Paloma 159n11, 161
Colbert, David W 77nn36 -7, 89n4 Dobson, R.B 121n57
Collin, Ludo 79n44 Douglas, Sheila 188n24
Conlon, D.C 117n46 Dovrynia Nikitich 30
Constantinople 14 Du Cange, Charles Du Fresne 116
Cooper, Helen 77n35 du Fail, Noël, Propos rustiques 10
Cornell, Henrik 91n8, 99n26 Dubois, Jacques 71n11
Corominas, J 47n17 Duby, Georges 109n21
corridos 179 Duffin, Charles 11
Covarrubias, Sebastián de 47n21 dukhovnyi stikh 28
Cowan, Edward J 192n37 duma (Ukrainian) 27
Cowling, G.H 127n17 Duncan, Charlie 198-99

215
The Singer and the Scribe

Duncan, James B 188n26 ——— Portuguese and Brazilian

Duncan, Jock 193, 195n54, 198, Balladry 51, 52nn5-6, 59n22,
199n54 62n32, 67n47
Dundee 194 Foote, Peter 106nn5-6
Dundes, Alan 16n10, 80nn50 -1 Fourth Crusade (1204) 21
Duque de Rivas, the 174 Fowler, David C 7-9, 77n35, 127n19,
Duque de T’Serclaes, the 159, 162 129
D’Urfey, Tom 133 Fraker, Charles 53
Durling, Nancy Vine 110n26 Frances, Irma 155-7n5
Frances, Maurice 155
Easton, Gordon 186n19, 196n47, Franch, Juan Alcina 38n6
198, 199n54 Friedman, Albert B 5, 6n8, 9n16,
Edda 25 107, 113, 116
Edinburgh 169, 175, 194 Friedman, John B 101n32
Edson, Richmond W 23n1 Friis, bishop Jakob (d. 1486) 98
Edward I, king of England 118 Fryar and Boye 128
Edwards, Carol L 70n6 Funk Gabriela 59n22
Egginhard 180 Furnivall, Frederick J 127-8, 133
Eideneier, Hans 20n19 Furrow, Melissa 128n22
El Hierro (Canary Islands) 159
Ellon, Aberdeenshire 189 Gabai, Rahel 169
Elrick, George 195 Gala, Antonio 37, 50
Enfances Ogier 93 Gambarin, Giovanni 63n41
Entwistle, William J 5, 6, 8n15, 23n1, Garbáty, Thomas J 9n18
26n20, 31, 48n23, 69n4, 125n8, García, Jesús 176
Erotokritos 20 García Lorca, Federico Romancero
Escobar, Juan de 62n34 Gitano 174
Euphrates 19 Garrett, Almeida 51
Eustace le Moine 118 Gennrich, Friedrich 110n26
Gerould, Gordon Hall 6, 8, 9n16,
Faeroe Islands 89 10n20, 24n7, 69, 116
Fauriel, Claude 13n1 Ghelman, Yaacov 169n1
Felon Sew, the 127n17 Ghent 79
Ferdinand I, king of Castile 52 Gielemans, Johannes 71-3, 75n29,
Ferdinand V, king of Castile 153 82-3
Ferré, Pere, Romances Tradicionais Gilferding, A.F 23n3
58n20 Gillespie, Margaret 188, 191n35
Fetterangus, Aberdeenshire 181-3, Glasgow 194, 195
189, 196-7, 195-200 Goa 52
Finnegan, Ruth 10n22 Goldstein, Kenneth S 183, 186,
Firmino, A 63n42 188n23, 193, 195-200
Fisher, Ray 184, 198, 199n54 González-Llubera, Ignacio 159n11
Floda Church 95-102 Gormont et Isembart 117n45
Floriseo 42-9 Gough, Richard 10
Fontes, Manuel da Costa 9, 52n7, Gower, Herschel 188n24
57n18, 58n20, 59n24, 105n2 Góngora 174

216
 Index

Granada 175-79 Hildeman, Karl-Ivar 89n4

Green, Richard Firth 11, 77n35 Hindley, Charles 194n41
Greig, Gavin 188 -90, 193, 195, 197 Hjorth, Poul Lindegård 93nn15– 6,
Greig-Duncan Folk -Song Collection 94n18, 97n23, 98n24
190 -1, 193, 195, 193 Hodgart, M.J.C 79n45
Grigir’ev, A.D 30n41 Hoffman, Conrado 56
Grundtvig, Svend 92n10, 101 Holden, Anthony J. 116n40
Guillaume le Maréchal 109n21 Holland Park Synagogue 155
Gummere, Francis B 7, 23n1, 131 Holton, David 17n13
Guzmán, Martín Luis 179n12 Holzapfel, Otto 76n30, 93n12
Homer 9, 136-7, 143
Hagiologium Brabantinorum 71 Horak, I 30
Haidu, Peter 105 Hritsa, S 26n18
Hales, John W 127 Hue de Rotelande, Ipomedon 84n68
Halliwell, James Orchard 125n10, Huerta, Victoriano, Mexican
126n14 president 179
Halvorsen, E.F 93n14, 96n22 Hunisett, R.F 120n55
Harding, Alan 120n55 Hurtado de los Hitos, María 7n11
Harkins, W.E 27n26, 29n36 Hustvedt, S.B 23n1
Harris, J 24n11, 26n29 Hutton, Ronald 129
Harris, Tracy K 154n1 Huybens, Gilbert 71n7, 78n40
Hart, Thomas R 37n2
Hart, W.M 23n1, 25n15 Iakimenko, T.S 27n25
Hastings, Battle of 115-16 Il’ia Muromets 30
Hauf, A 37n3, 44 Imlach, Hamish 184n17
Havelock, Eric 82, 83n66 Ioannou, G 15n5
Heinrich von Morungen 76 Ipomedon 84
Helm, Alex 130 Isabella I, queen of Castille 153
Hemming, Timothy D 106n7, 115n37 Iswolsky, Helene 124n4
Hemsi, Alberto 159n11 Ivan the Terrible 28
Hench, Atcheson L 131n31
Henderson, Hamish 183, 189, 192 Jack and his Stepdame 128
Henderson, T.F 144n17 Jakobson, R 25n18
Henricus de Halen 71 Jamieson, Robert 126, 129
Henriquez, J. C 73n19 Jansson, Sven-Bertil 89n1, 92n10
Henry I, duke of Brabant 72 Jämtland 100n30
Henry II duke of Brabant 76 Jean Erart, El mois de mai 112-13,
Henry II, king of England 109n21 115, 118
Herd, David 129, 146-50, 191 Jean Renart, Guillaume de Dole 109-
Herder, Johann Gottfried 108n12 111n30
Herod 45 Jeannaraki, A 15n5
Herr Ivan 77n36 Jeffreys, Elizabeth 19n16, 116n41
Hervarar saga 77, 90 Jersild, Margareta 92n10
Hesiod 9 Jerusalem 160
Hesseling, D.C 20n19 Jewitt, Llewellyn 126
Higgins, Lizzie 183 John II, king of Castile 175

217
The Singer and the Scribe

John the Reeve 129n23 Layher, William 9, 89n3, 105n2

Johnston, Annie 188 Le Gentil, Pierre 110 -11n29
Johnston, Calum 188 Lecoy, Félix 110n22, 111n29
Joldersma, Hermina 74n22, 79n46 Lemaitre, Jean-Loup 71n11
jongleurs / juglares 8 Leslie, Charles 194n41
Jonsson, Bengt R 26n19, 89n2, 90n5, Lewis, Huw 7
92nn10-11, 101n34 Lévi-Provençal, E 59n23
Jordan, Howard S 108n14, 110n24 Liberman, Anatoly 83n64
Jönsson, Hinse 91-2 Liège 110, 111n30
Life of the Blessed Margaret of
Kagg, Mats 91 Louvain 71
Kalff, G 69n4, 75n27, 82n58, 84n68 Livre d’Artus 74n23
Kapsomenos, Eratosthenes 14n4 Šljivi -Šimši, Biljana 159n11
Karl Magnus (?OSw) 94, 99 Lloyd, A.L 125
Karl Magnus’ Krønike 93-9 Lomax, Alan 183, 188n24, 189
Karlamagnús saga 95-8, 101, 106 London 153
Katz, Israel J 64n43, 159, 160n11 Longfellow, Henry W 37
——— El romancero judeo-español Lope de Vega 6, 174
168 Lopes, António 62n36
Kayser, Wolfgang 69n4, 76nn32&34 Lord, Albert B 70n5, 136, 138-9n9,
Keats, John 14 140n13
Kemp, Willie 195 Lorenzo de Sepúlveda 6
Kennedy, Ludovic 195 Loth, Agnete 93n14
Kennedy, Norman 198, 199n54 Louis d’Anjou 117
Kennedy, Peter 126n13, 130-2 Louis VIII, king of France 119
Ker, W.P 5, 110n24, 116 Loukatos, D 14n4
Kerr, James S 195n45 Louvain 70, 78
Kiev 31 Lowell, James Russell 127
King Arthur 45 Lönnroth, Lars 106n5
King Edward, Aberdeenshire 189 Lundberg, Erik 91n8
Kinloch, George R 126 Lüthi, Max 83
Kireevskii, P.V 23, 30n41 Lyle, Agnes 188
Kittredge, G.L 127n18 Lyle, Emily B 188n26
Knight, Alan 179nn11&13
Koepp, Johannes 74n22, 75nn24–5, Macdonald, Donald A 182n10,
85 188n24
Kong Olger Danskes Krønike 94-5 Machado, Antonio 174
Kornhall, David 94nn17-8, 99n25 MacHale, Christy 180n15
Kravtsov, N.I 25n14, 28nn27&30–1 Mackay, Angus 175
Kyriakidis, Stilpon 14n4 Mackridge, Peter 15n6
Kyriakidou-Nestoros, A 14n4 Macneill, Flora 183
Madeira 57-8, 171
Labarthe-Postel, Judith 20n20 Madero, Francisco 179
Larrea Palacín, Arcadio de 159n11 Madrid 154, 165, 172, 174
Laurensson, Anders 91 Magalhães, Celso de 52
Laurensson, Germund (d. 1459) 91 Magnus, L.A 26n20

218
 Index

Malacca 52 Menéndez Pelayo, Marcelino 51

Manley, Kathleen E.B 70n6 Menéndez Pidal, Ramón 37n1, 38,
Manrique de Lara, Manuel 154, 156, 40-3, 47, 51, 53, 56, 59n23,
164-5, 167 60n28, 61n32, 62n37, 63nn38-9,
Manuel Pedrosa, José 160n11 64nn43&45, 156, 161, 165n14,
Mapstone, Sally 77n35 167, 169, 181n10
Marathon, battle of 13 Mesopotamia 19
Marburg 76 Messina 180
Marchenko Iu.I 27n25 Metzner, Ernst Erich 75n30
Margaretten, Selma 160n11 Meza, Juan 179
Marmontel, Jean François 107 Michaëlis de Vasconcelos, Carolina
Marques, José Joaquim Dias 51n2, 51n1
58-59, 61nn29&31 Mikhaila Potyk 30
Marseilles 180 Milán (musician) 173
Martin, Ariadne Y 83n64 Mintlaw, Aberdeenshire 198n52
Martin, Richard P 83n64 Mitrofanova, V.V 29n36
Martins, Ana Maria 51 Molanus, Johannes 87
Martins, P 63n42 Molho, Michael 155n11
Martínez Mata, Emilio 42-4 Montero, Paloma 155n11
Martínez Ruiz, J 159n11 Morris, George S 195
Mary Magdalen 127 Morocco 154, 159, 162-3, 169, 176,
Matalón, Jacob Samuel 154 180
Matalón, Rahel Errera 154 Mortier, Raoul 116n38
Mathew, Jean 196 Morton, Catherine 116n39
Mathieson, Willie 189, 196 Moshav Mata, Israel 169
Mauleón, Judith 180 Motherwell, William 129, 145, 182,
Mazatlán, Mexico 179 188
Mälaren, Lake 92 Mozheiko, Z.Ia 29n34
Målare, Anna 91 Mukharinskaia, L.S 27n25
Målare, Johan 91 Mummers’ Plays 130 -1
McCarthy, W.B 188n25 Munro, Ailie 183n11, 186n19
McCarthy, W.J 140 Muntz, Hope 116n39
McDonald, Barry 182n8 Murray, Kenny 180n15
McIntyre, Geordie199n54 Myddle, Shropshire 10
McKean, Thomas A 9n16, 10, Mynors, R.A.B 116n42
185n18, 186n19, 187n22,
196nn47&49 Naples 117
McKinnon, Bill 196 Narváez (musician) 173
McMorland, Alison 197n50, 198, Nascimento, Bráulio do 52n4
199n54 Natales Sanctorum Belgii 87
Mearns John 195 Nehama, Joseph 154n2, 168n19
Meekings, C.A.F 120n55 Nejedlá, J 27n24
Megas, G. A 16n10 New York 195
Meier, John 75n26, 82n59, 86n74, Newcombe, Terence H. 112n34 -5
87n75 Nieuw Amsterdams Liedboek 79
Meletinskii, E.M 25n18 Nílsen, Anna 100

219
The Singer and the Scribe

Nucio, Martín 56, 59-62, 67, 173-4 Politis, N.G 15n5, 16n8
Nygard, Holger 9n18 Porter, James 186n19, 188n24
Post, J.B 120n55
Ochoa, Don Eugenio de 6n7 Pratum musicum 85
Oeghe, Johan 71 Preminger, Alex 107n8
Olivos 49 Primera crónica general de España
Ollivier, Marie-Joseph 70n7 51, 53, 56, 61
Olrik, Axel 11, 80-1, 84 Propp Vladimir Ia 24n11, 25n18,
Ong, Walter 138 29n35, 31, 83
Ord, John 190, 198 Ptochoprodromika 20
Ordericus Vitalis 115 Purcell, Joanne B 58n20, 59-60
Ordish, Thomas Fairman 130 Putilov, B.N 25n18, 30n41, 32n45
Orense 159 Putter, A.D 9, 11
Oring, Elliott 69n1
Öberg, Jan 91n7 Quelite, Mexico 179
Page, Christopher 109n21 Quiller-Couch, Sir Arthur 5, 125
Palencia 159
Palmer, Roy 126n12 Rabelais, François 124, 128-9
Panche 128 Ramsey, Jarold W 69n1
Pancho Villa 179 Ranke, Kurt 93n12
Papago 69, 73n20 Rankin, James 187
paraloyes 14-18, 20 Raposo, Inácio 52
Paris, Gaston 108n12, 115-6, 119 Reeves, James 126n15
Paris, Paulin 107n10 Reig, Carola 53n8
Parry, Adam 136n2 Reilly, Bernard E 53n8
Parry, Milman 136-41, 143, 146-7, Renson, Jean 110n26
149-51 Renwick, Roger de V 195n47
Pascual, J.A 47n17 Rethaan Macaré, M.C.A. 72n14
Peacock, N 130n29 Ribeiro, Joaquim 52n3
Pedersen, Christiern 94, 102 Rico, Francisco 43n14
Pepys, Samuel 191 Rieuwerts, Sigrid 28n29, 127n20,
Percy Folio Manuscript 127-8 187n22, 196n48
Percy, Thomas 127 Ritson, Joseph 129, 145
Peringskiöld, Johan 101n34 Robert the Bruce 199
Pernot, H 20n19 Roberts, Warren E 123n2
Perrie, M 30n28 Robertson, Bell 188
Peterhead 196 Robertson, Jeannie 183, 186n19, 187-
Petrova, L.I 27n25 8
Pettitt, Thomas 24, 70n6 Robin Hood 8n13, 77, 116n38, 121,
Pérez Vidal, J 66n46 123
Philip IV, king of France 118 Robinson, A. N 25n16
Pinker, Steven 140n14 Rodrigues, Wilson W 52n3
Piñero, Pedro M 58n20 Rodríguez del Padrón, Juan 38, 40,
Piot, Charles 72n14 48
Pirot, François 110n25 Rodríguez-García, Isabel 58n20
pliegos sueltos 10, 159, 173, 178

220
 Index

Rodríguez-Moñino, Antonio 56n15, Salonica 153-6, 158-60, 162-6,

62n34 168n19
Rodway, A.E 132n32 Sarajevo 157, 159, 160, 162, 164 -6
Rogatchevskaia, Ekaterina 8-9 Saramago, João A. das Pedras 58n20
Rogers, Edith Randam 44 Saul, Moshe 177
Rogge, Kort, bishop of Strängnäs (d. Saunders, Corinne 72n17
1501) 91 Saval, José 180n15
Romance de la Infantina 7 Sánchez Romeralo, Antonio 179n10
Romance of the Rose, the 10 Scarborough, Dorothy 79n45
Romanceros 172 Schneider, Hermann 75n31
Rome 41, 178 Schwarzwald, Ora 169n1
Romero, Elena 159n11 Scott, H. von E 70n7
Roques, Mario 117n47 Scott, Walter 129, 145
Rose, H.J 156n4 Sell, Roger D 195n44
Rouge-Cloître, monastery of 71 Sephardim 10, 38-41, 52, 63-4, 153-
Roughead, William 194n43 68, 169 -70, 178
Rousseau, Jean Jacques 8 Seroussi, Edwin 159n11
Rubin, David 79n42, 87n76 Sharp, Cecil 136
Ruiz, María Jesús 58n20 Shaul, Moshe 156
Runnalls, Graham A 106n5 Shephard, Leslie 194n41
Rushing, James A 99nn27-8 Shepherd’s Calendar 10
Rybakov, B.A 25n18 Shippey, Tom 72n12
Rybnikov, P.N 23n3 Shuldham-Shaw, Patrick 188n26
Sifakis, Grigoris, 14n4
Sacré, Dirk 91n7 Silverman, Joseph H 64n43, 67n47,
St Antonius 100 156n4, 159nn8&11
St Baafs Cathedral (Ghent) 79 Simmons, E.J 25n18
St Denis 118 Sims -Williams, Patrick 49n25
St Elisabeth of Hungary 76 Skårup, Povl 93n14, 106n5
St George 14 Skopin-Shuiskii, M.V. 28n27
St George’s Inn (Louvain) 70-1 Skripil’, M.O 29n36
St Gertrude 75n29 Sleeman, Margaret 9n16, 10, 180n15
St John’s Chapel (Louvain) 77 Slonim, M 25n18
St John’s Church (Ghent) 78 Smirnov, I.Iu 26-30n39,
St John’s Day 44, 48 Smith, Adam 138
St John’s Wood 155 Smith, Colin 6n6, 105, 106n7,
St Louis 76 118n51
St Louis IV, Landgrave of Thuringia Sobolevskii, A.I 23
76 Solheim, Svale 92n11
St Lucy 73 Sola Pinto, Vivian de 132
St Olaf 100 Sommer, H. Oskar 75n23
St Peter’s Church (Louvain) 70-2, 77, Söderala, Sweden 99n25
81 Spanish Inquisition 153
St Sophia Cathedral 14 Spiessens, Godelieve 85n70, 86n73
Saint-Trond 110 Spitzer, Leo 37n4, 41, 45
Salamis, battle of 13 Sri Lanka 52

221
The Singer and the Scribe

stariny 23 Turner, Jane 91n8

Steenstrup, Johannes 77nn36&38 Turriff, Jane 182n10, 189, 196
Stein, Helga 196n48
Stevens, John 119n53 Underdown, David 129
Stewart, Alec 181n1 Underhill, Ruth 69
Stewart, Belle 181n1, 188 Unger, C.R 93n14, 96n22
Stewart, Elizabeth (1939– ) 183 -200 Utembruele, Margaret 71
Stewart, Elizabeth, jr 187
Stewart family, the (of Fetterangus) Valdés, Meléndez 174
181-188, 199-200 Valenciano, Ana 62n37, 160n11
Stewart family, the (of Blair) 181n1, van de Graft, C. C 76n33
183 van der Werf, Hendrik 110n26,
Stewart, Jane 195 111n30
Stewart, Jean (1911–1962) 185, 195, van Duyse, F 75nn25&29, 80,
200 82nn57-8, 82n57&58,
Stewart, Jeannette 187 83nn61,63&65
Stewart, Kathie 181n1 van Even, Edward 70n7, 72n14
Stewart, Lucy (1901– 1982) 183-200 Vanda, Anastácio 51
Stewart, Michael 187 Vandelli, Giuseppe 63n41
Stewart, Sheila 181n1, 198, 199n54 Vansina, Jan 149n21
Stief, C. 28n27 Vargyas, Lajos 75n25
Stock, Brian 9n17 Varty, Kenneth 118n49
Stockholm 90-92, 101n33 Vasconcellos, José Leite de,
Sviatogor 30 Romanceiro Português 51, 63n42,
Sydney, Sir Philip 8n13 65-6
Veiga, Sebastião Philippes Martins
Tale of Belisarius, the 20 Estácio da 58n22
Tarbé, Prosper 107n10 Vellekoop, K 74n21, 75n28, 80n47
Tasseel, Marcellus 78 Venice 17
Taylor, J 121n57 Verdam, J 73n20
Tesoro de los romanceros 6 Verhaghen, P.J 72n12
Tetuán, Morocco 163 Verwijs, E 73n20
Thibaut de Blaison 112 Villers, Cistercian Monastery of 70 -1,
Thibaut de Champagne 112 73
Thiðreks saga 101n33 Villers- la -Ville, Belgium 70
Thompson, E.P 130 Virgil 10, 169
Thompson, Rupert 130n28 Volkslied 13
Thomson, Derick 187n22 Von Fallersleben, Hoffman 79
Tiber, River 178 von Sydow, Carl Wilhelm 187n21
Tiddy, Reginald 130
Timoneda, Juan 62n34 Wace, Roman de Rou 116n40
Togeby, Knud 93nn12&14, 94n19 Wade-Gerry, Henry Theodore 137n6
Toro 53 Wagenaar-Nolthenius, H 74n21,
Tournoy, Gilbert 91n7 75n28, 80n47
Townsley, Elizabeth (1939– ) 184 Wallensköld, A 112n33
Trapero, Maximiano 159 -60, 164-5 Wallin, Sigurd 99n26, 91n8

222
 Index

Weich-Shahak, Susana 159n11,

169n1
West-Vlaanderen 78
Whitehead, Frederick 106n7
Wild Host 37, 43, 45-6, 49
Wild Man 99
William of Malmesbury 115-16
Williams, David 100n32
Wilson, David 106n6
Wimberly, Lowry Charles 82n58
Winterbottom, Michael 116n42
Woden 45
Wolf, Fernando, and Conrado
Hofmann, Primavera 56-57
Wolfenbüttel 79
Wright, Roger 6-8, 9n16, 160n12,
172n2, 175nn5&7, 180n14
Wright, Thomas 125n10

Zamora, Spain 53-6, 59, 61, 159

Zaragoza, Spain 175, 180
Zhirmunskii, V.M 25n18
Zink, Michel 107-9, 112, 115
Zumthor, Paul 115n36

223

Download more eBooks here: http://avaxhm.com/blogs/ChrisRedfield