s

Previous Conference Publications ..EDITING

. AND EDITORS:
A RETROSPECT
1965 Editing Sixteenth-Century Texts, ed. R. J. Schoeck (1966)
1966 Editing Nineteenth-Century Texts, ed. John M. Robson (1967)
1967 Editing Eighteenth -Century Texts, ed. D. I. B. Smith (1968)
1968 Editor, Author, and Publisher, ed. WilliamJ. Howard (I 969)
Papers given at the twenty-first annual
1969 Editing Twentieth Cen tury Texts, ed. Francess G. Halpenny Conference on Editorial Problems
(1972) University of Toronto
1970 Editing Seventeenth-Century Prose, ed. D . I. B. Smith (1972) 1-2 November 1985
1971 Editing TextsoJthe Romantic Period, ed. John D . Baird (1972)
1972 Editing Canadian Texts, ed. Francess G. Halpenny (1975)
1973 Editing Eighteenth-CenturyNovels, ed. G. E. Bentley, Jr. (1975)
1974 Editing BritishandAmerican Literature, 1880-1920, ed. Eric
EDITED BY
W. Domville(1976)
RICHARD LANDON
1975 Editing Renaissance Dramatic Texts: English, Italian, and
Spanish, ed. Anne Lancashire (1976)
1976 Editing Medieval Texts: English. French, and Latin Written in
England, _d . A. G. Rigg (1977)
1977 Editing Nineteenth-Century Fiction, ed. Jane Millgate (I 978)
1978 Editing Correspondence, ed. J.A. Dainard (1979)
1979 Editing Illustrated Books, _d. William Blissett (I 980 )
1980 Editing Poetry from Spenser to Dryden, ed. A. H . de Quehen
(198 1)
1981 Editing Texts in the History ofScience and Medicine, ed.
Trevor H. Levere (1982)
1982 Editing Polymaths: Erasrnus to Russel/, ed . H . J. Jackson (1983)
1983 Editing Early English Drama: Special Problems and New ' L. .) .
Directions, ed. A. F. Johnston (1987) AMS PRESS, INC. U r /,)
~ lc
( '! ..
New York J
,.-v .- ...,
. , ,
1984 Editing, Publishing and Computer Technology, ed. Sharon Butler
and William P. Stoneman (1988) I

I .N .
) ,.0,\- "
:

I ... /
t' ..

" ,
.J
 'Epistulae Venerunt Parum Dulces':
The Place of Codicology
III the Editing of Medieval Latin Texts

Leonard E. Boyle a.p.

The title of this twenty-first annual Conference on Editorial

Problems is "Editing and Editors : A Retrospect", and I am
glad that such a broad title has been chosen. Too often the
general heading for a conference such as this is "Textual
Criticism" or "Problems of Textual Criticism" and the like.
Maas's well-known manual, to which everyone turns, is enti-
tled Textual Criticism; so is Martin West's recent attempt to
replace Maas. I There has been a spate of conferences of a
similar bent over the past twenty or thirty years, notably, for
legal circles, the two-volume La critica del testa which em-
bodies the proceedings of an international congress of the

11 P. Maas, Textkritik (Leipzig, 1937', 1950' , 1957', 1960'), English trans-

lation by B. Flower, Textual Criticism (Oxford. 1958), from 3rd edition;
M. West, Textual Criticism and Editorial Technique Applicable to Greek
and Latin Texts (Stuttgart, 1973) .

30 I MEDIEVAL LATIN TEXTS BOYLE / 31

Societa italiana di Storia del Diritto, and ranges far and wide B.ecause it was in print, it was sacrosanct, and no thought was
in the field! gIven to the fact that the punctuation was the editor's not
To me, the problem is that a title such as "Textual Criti- Bacon's, and was plainly at variance with the syntax of the
cism" presumes that one already has a text to criticize. Martin passage. 3
West in his recent volume even goes so far as to counsel his I think it therefore not unfair to say that the greatest threat
readers that when one is preparing an edition one should first to an editor's independence and to an unprejudiced presenta-
provide oneself with an interleaved copy of an existing edition tion of a textual tradition is the presence of an existing edition.
of the text in question, and then proceed from library to Yet this threat is rarely seen for what it is, the modern equi-
library, noting on the interleaves the chief variants that one valent of what is decried as "contamination" in a medieval
comes across in the various manuscripts in those libraries. setting. Hence a recent and distinguished editor unblushingly
This sort of approach is, I am sure, quite all right in its states that his base text for his edition "was constructed from
way, but it has its dangers. Taking the trouble to consult five manuscripts which were collated in their entirety against
other manuscripts of a text, if it be discerned in a medieval the most recent edition." Another editor, an apostle--and a
scribe or scribe-scholar, is dubbed "contamination." When very good one--<Jf computerized editions, dismisses as
practised by a modern scribe-scholar (and the modern editor inad~quate .a previous edition of a work he is editing, yet, in
is no more than that) it is termed "scholarship." Yetcontami- a twmkle, mforms us that the first stage of his own edition
nation, if one accepts the term at its pejorative worst, is not will consist of recording the earlier edition on magnetic tape,
confined to the medieval scene. If contamination means, as it there to eliminate its errors. In both these cases, not to speak
seems to mean, an unwarrented influence of one textual trad- of West's interleaving, the danger is that one will wind up
ition on another, then the most pernicious form of all is the with an edition of an edition. It hardly can be called a new
printed text. A printed text has a stark, imperative quality, edition based on the manuscript tradition, unless one were to
and it is hard not to be influenced if not overawed by it. allow that existing editions are, to all intents and purposes,
Generations of Roger Bacon scholars, for example, have
fought over the date of his birth and the stages of his early
career, the chief cause of contention being an autobiographical 31 Fratris Rogerii Bacon Compendium Stud;i Theologiae, ed. H. Rashdall
statement in one of his works that survives in one manuscript. (British Society for Franciscan Studies Ill , Aberdeen, 1911) 34: "Etiam
logicalia fuerunt tarde recepta et lecta [Oxonie]. Nam Beatus Edmundus
Yet, in fact, it is not the autobiographical statement that proves
Cantuariensis archiepiscopus primus legjt Oxonie librum elenchorum tem-
to be the problem, but the manner in which the editor,
p~ribus meis : et vidi .magistrum Hu gonem, qui primo legit librum post-
Rashdall, punctuated it when editing it for the first time.
enorum, et [hbrum] eJUS conspexi ... As the text is printed, Bacon seems to
say, and has been accepted by scholars as saying, that Edmund of Abingdon
taught at Oxford "in my day ," whereas "temporibus meis" really goes with
21 La mt:ca del testa , 2 vols. (Florence, 1971); see also L. E. Boyie, the following "et ... et" clause: .. . Nam Beatus Edmundus Camuariensis
Medieval Latin Palaeography: A Biblwgraphical Introduction (Toronto, archiepiscopus primus legit Oxonie librum elenchorumj temporibus meis
1984), nn. 1954-2042, for various works on the approaches to textual et vidi magistrum Hugonem, qui prima legit librum posteriorum, et librum
criticism. eius conspexi. "
32 / MEDlEVAL LATIN TEXTS
BOYLE / 33
part of the manuscnpt , tra d'!twn,
, and should be treated as
such" , procedure :0 follow if ,o~e is ever to arrive at a text upon
The real trouble with all of this is that It puts the cart which to bnng textual cntlClsm to bear. For it is an inescapable
before the horse, It presumes that one already has to hand a fact that the only way in which we know of the text we are
ready-made text, all nicely parcelled, to :"ork on, whereas editing-even when it has had many editions-is through the
the primary problem with which most editors are faced, no codices that carry it, And unless these carriers are examined
matter what editions exist, is the fact that they have to d~al as t,h,oroughly ~s is possible, one is never going to be in a
with a text that is in no way set, but varies from manuscnpt pOSltl?n to subject that text as it is carried by the codices to
anythmg approaching a critical analysis,
to manuscript, ," ,
I am not suggesting that Textual CnucI,sm IS a bad ?r Most manuals of textual criticism make little or no distinc-
unnecessary thing, Far from it, It has a, d:fmlte and essential tion between the criticism of the tradition or transmission of
role to play in the editing of texts, b,u,t ~t IS a role that c,annet texts and textual criticism, And because textual criticism is
be played out until there is first a cntlClsm of the tradmon of gener,ally the goal of ,these manuals, the tradition is only
the text, exammed from the pomt-of-view of the text with little or
Unfortunately, criticism of the tradition of texts (I do not nothing on the codices as such, Yet withou: these codices
say of the textual tradition) doe~ not have as lar~e an appeal there would not be a text to edit, Editing, in a word, has to
as textual criticism, Everyone wishes to be an editor and not start from the codices, has to respect the codices at all points,
simply a reporter; and to be in a position t~ n~te .s:gely m and may only depart from them when the editor is Sure that
one's apparatus, "conieci", or, rapturously, scnp~l. ,These they have been squeezed dry of every possible piece of evi-
dence, whether textual or physical.
heady moments when one makes one's own contnbutlon to
the textual tradition are indeed to be savoured, but they,n:ay In a piece I read recently, which I shall not identify further,
only be indulged in if o~e is q~ite sure just what the tradmon the o.ne eX,tan t, c~dex of a text is described as a "slovenly
of text is to which one IS addmg, , , c~py , wh~ch m ltself IS a rather ungracious remark, since
An assessment of the tradition is a prereqUisite of textu~1 Without thiS umque slovenly copy the editor would hardly
criticism and it is time-consuming, demanding and~ or so It have been in a pOsition to edit the text and make--or unmake--
seems to some, degrading, At times it seems to be miles away his name, But, the author goes on, "grammatical errors are
from the text because it has to concentrate not on the text but common in the text, and he promises to correct them, Need-
on the textual setting, on the codices, that is, that carry the less to say, this is a hazardous procedure at best, Our modern
text rather than on the text itself, Yet this is the only logical scribe-scholars are all too eager to re-write and all too often
fall into, traps as horrendous as those for which, again all too
temeranously, they would crucify medieval scribe-scholars
or even simple scribes, Above all else, some modern scribe~
41 The two cases above, and other similar cases cited below, are not inven-
,
Hons on m y part . But I do not give the name of the author or the source
.
in these cases, lest in any way I should seem to impugn the profeSSIonal
sch~lar~ (and, I repeat, an editor is no more or less than that)
integrity of authors or editors who have a different approach to the very ~re mclme~, perhaps under the influence of Housman, to pay
personal one expressed here. llt,tl: mentIOn to two fundamental principles that should guide
cntlclsm, and especially emendation: no emendation should
34 I MEDIEVAL LATIN TEXTS
BOYLE / 35
be attempted until what is in the MS (or MSS) has been viewed itself, proved to say exactly the opposite of what the editor
from every possible angle; and second, if an emendation has had concluded. What had happened was that, because of the
to be made, then the less tampering the better with the text tight binding of the codex, the microfilm on which our editor
as it is transmitted in the codices. had relied had been able to record only the outer part of the
In the present case, there is a nice example of a failure to gloss-a part which, oddly, made perfect sense on its OWn
observe these principles, when the phrase "ut lectoris meus and, although it held an unusual opinion, was never questioned
videat prae oculis ea quae dicimus", which clearly is awry, is by the editor.
emended to "ut lector meus videat prae oculis ea quae But even where the codices are examined at first hand ,
dicimus." Since a confusion of Un" and "un is rampant in some scholars are content to give what are no more than
medieval MSS (and in most of our own handwriting), then perfunctory descriptions. Obvious things such as foliation
the first thing to do before rushing to emendations such as and gatherings are noted, but few scholars, if any, bother to
"lector meus" for "Iectoris meus" is to see whether "Iectoris note second or third folio incipits, the size of the frame of
mens" makes any sense--which of course it does ("ut lectoris writing, the presence of such details as "fillers" and other
mens videat prae oculis ea quae dicimus", i.e. in the diagram sc~ibal conceits. In the light of the work of Gilbert Ouy, lOon
immediately following), where "lector meus", apart from tam- GIhssen and others on, for example, the make-up of quires
pering with what is in front of one, betrays an acute insensitiv- and the layout of pages, old-style descriptions have had their
ity to what is going on in the codex. . day. 5 ~ven if one does not take to the term "Codicology",
Since it is what is in the codex that matters, the fIrst step, the plam fact remains that a rigorous physical description of
then, on the way to an edition is the tiresome one of catalogu- each codex--call it Codicology or Handschriftenkunde or
ing or describing each of the codices as carefully as possible. L' Archeologie du livre or what you will-a rigorous physical
It is not sufficient to note the size, the date, the script of each description is now de rigeur. I may add, however, that "phys-
codex. The whole physical setting of the text has to be ac- ical setting" is probably a more useful term than "physical
counted for, from the binding to the flyleaves and pastedowns. description." "Physical description" of itself only covers the
From this point of view it is surprising to find that there are ~ake-up as such of a codex, whereas "physical setting"
editors who do not bother with the physical make-up of a mcludes both the codex and the text as it is found from page
codex. Too often they rely on descriptions in catalogues and to page, with marginal or interlinear notes, rubrics , decora-
are happy to work from microfilms or photographs of the tions, gibbets, fillers, doodles and the like. JUSt how a text is
part of the codex in which their text occurs and without ever physically transmitted may be just as important for the overall
seeing the whole codex itself. This is quite perilous. One tradition of the text as the quality of the text itself in a given
edition which came my way recently, but luckily just before codex.
publication, made much in its introduction of the fact that
the medieval author, whose work survives in just one codex,
5/ See for example L. Gilissen, Pro/egomene, a la codic%gie (Ghem, 1977);
had taken a firm stand against the mainstream of allegorical G. Ouy, uPour une archivi stique des manu5crits medievaux," Bulletin des
interpretation of the scriptures. The editor based himself on bibliotheques de France 3 (1958) 897-923; and nn. 1570-1594. 1616- 1632,
a gloss of the author, which, when examined in the codex etc. in BoyIe, Medieval Latin Palaeography.
36 / MEDlEVAL LATIN TEXTS
BOYLE / 37
This codicological setting is, as I see it, the first of three cancellation, annotation, gap, erasure, correction, inversion,
stages through which one has to pass when preparing an edi- misspelling, filler, grammatical inanity, homoeoteleuton and
tion, the second of which is the establishing of the text from all.
the extant codices, the third the editing of that established text. . .What the Recensionist editor winds up with at this point
The second stage, establishing the tradition of the text IS Simply a scrupulously transcribed but utterly unedited text
precisely as it is found in the codices, has, I need hardly of the chosen first witness. Since this text opens the way
remark, received far more attention from scholars than the to a Recensio of the codices as a whole, it may be called a
first stage, that of the physical setting-and this in spite of Recension-text. With its help, the other codices may now be
the fact that it is something less than prudent to attempt to recorded on the recension sheets with a certain expedition,
establish a text from its codices if one does not know the for only the variations, whether textual or physical, from
codices intimately. codex to codex, will need to be set down or transcribed.
There are two main theories as to how One extracts or When all the witnesses have been set out in relation to
establishes a text from a plethora of codices. The first, which one another On the recension sheets, it should not be too
for obvious reasons one may call "Optimist", argues, in the difficult to see at once what is common to all the codices (a
wake of Bedier (1928), that the sensible thing to do, when line here, a paragraph there) and what is not. The difficulty,
confronted with a number of codices of a work, is to single of course, is defined by what is not common. All the same,
out for editing that codex which is on the whole the most the position is not hopeless. For although one is now faced
satisfactory, and then to follow this "codex optimus" through with passages or phrases which are not common to all the
thick and thin, except, of course, where it is obviously defec- witnesses, it is possible, the Recensionist claims, to find out
tive or unintelligible. The second, to which one may assign which witnesses depend on or are close to one another. And
the label "Recensionist", argues, claiming Karl Lachmann as thus the Recensionist achieves, by cataloguing and plotting,
its progenitor, that one should make a complete "Recensio" "common variations" in the codices at the points where they
or review of a given text as it is found in each of its codices, differ.
and this in order to find out, first, what is transmitted in The usual term found in manuals and articles is "common
common by the extant witnesses, then what is not held in errors", but "common variations" is more appropriate, since
6
common. it eliminates the subjectivity present in the term "errors" and
The Recensionist procedure is tedious if followed faith- embraces both textual variations and those of a codicological
fully (and even if one cuts corners). First of all, one of the or physical kind. 7 What constitutes a "common variation" is
codices (it does not matter which, but usually the oldest) is not whether it is right or wrong (for, with respect to the text,
transcribed faithfully, completely 'slavishly, and with every we have no means at this point of judging this in most cases),
but the simple physical fact that two or more witnesses have
61 On the twO main systems see L. E. BoyIe, "'Optimist and Recensionist: ~ome featureJn common that a third does not have. Anything,
Common Errors or Common Variations ?, in Latin Script and Letters A.D. III fact, that IS not shared by all the extant witnesses can be a
400-900: Fest5Chrift presented to Ludwig Bieler, ed. J. J. O'Meara and B.
Naumann (Leiden, 1976) 264-274. 71 Maas, ed. cit. n. 1, pp. 2-9, provides a classic statement of "errors",
38 / MEDIEVAL LATIN TEXTS BOYLE / 39

variation, from a press mark to a doodle, from a garbled word sense that no one has ever been on this road before. The
to a cancelled passage, from a change of ink or a change of problems now are the editor's problems and his alone; and it
hand to word-separation, from glosses to alternative readings. is here that his mettle as editor and critic is tested to the full.
Hence the importance of the "physical setting" which I noted It is of little use at this point to turn to theoretical approaches
when speaking of the first stage. or manuals of criticism. All they will provide are instances of
It is a Recensionist belief that by observing what variations how other editors have proceeded when confronted with their
are common to what codices where the codices differ and own problems. But these were their problems. They are not
what are not, one can trace retrogressively the path of these the problems of this present or of any other editor, though
common variations and arrive at the sourCe or sources of the they may provide some useful parallels. The editor's problems
different sets of common variations. If the Recensionist are his and his alone, just as his text is his and his alone. It
wishes, he may plot or make a diagram of these variations. is no uSe wailing, "What am I to do now?" What to do is up
He may even call it a Stemma codicum, but only if it is based to him and to no one else. Here the editor has to rely on his
on common variations between the codices and not on com- own intelligence, his own imagination, his own wide reading,
mon errors between the text as carried by the codices (in his own scholarship-and the help of his friends (not, of
which case it is not aStemma codicum but aStemma textuum). course, to provide him with an answer, but rather to criticize
From the relationships between the codices which he has or to consolidate his own solutions from their varying areas
detected through the phenomenon of "common variations" , of expertise). Thus a recent edition of a medieval philosophical
the Recensionist is now enabled to See a way through the text of no great significance could have benefitted from some
codices where they are textually at variance with one another. friendly criticism before being exposed in public to gleeful
He can see noW which manuscript or group of manuscripts purveyors of textual inanities. For at one point it proffers two ,:
syllogisms which, as printed, might give medieval scholasti- ,-il,f
is, where they differ, more likely to carry the text as it com-
I)
monly circulated (I do not say the "true" text). And he is cism a bad name: "Omnis canis creditur ; celeste sydus est
noW in a position to establish from the codices as a whole the canis; ergo celeste sydus creditur. Iterum: Quidquid creditur, ,-I~
"
text that they on-the-whole carry. The text that he establishes habet pedes; celeste sydus, ut dictum est, creditur; ergo habet "I

from the codices is thus the text that is common to all, plus pedes." A friend at least could have suggested without undue ii"
what he decides with the help of common variations to be the strain that there is such a thing as a dog-star, and, since dogs :~
have legs and legs are for running, that "currit" might be more !:i,-
likely vulgar text where the codices disagree.
~;
There remains the third stage, that of textual criticism. logically and palaeographically believable than "creditur. "
The established text is simply the text transmitted by the Helpful friends apart, being on one's own does not mean
codices and reported as faithfully as possible by the editor. that an editor now has a free hand to be creative: to emend,
It is a wholly unedited text, and it may be inconsistent in to conjecture or, indeed, to invent. He must always remember
some spots and delirious in others. What it needs now is an that he is dealing with the results of a tradition of text, and
editorial eye. that, if he must keep within the bounds of reason, he must
Here in the third stage the editor really comes into his also not venture beyond the confines of the codices from
own. He is in fact on his very own; and he is alone in the which he established the text now before him. At any point
40 I MEDlEVAL LATIN TEXTS BOYLE I 41
of uncertainty all the codices may and should be called upon, text on the basis of common variations and the Stemma. Nor,
even those which the editor knows from common variations indeed, is the common text itself sacrosanct simply because
and the resulting Stemma codium to be direct copies of earlier it is common. It may make no sense at all On occasion. Even
codices. when it does, one has to be on the alert.
Of course, since most Stemmata codicum are really Stem- In a recent edition of the letters of Seneca, for example,
mata textuum, it follows almost inevitably that when it is the editor prints the sentence, "Quae sint quae antiquos mov-
evident from a stemma that one text is a direct copy of another, erint vel quae sint quae anti qui moverint, dicam", noting that
or what is called a "codex descriptus, then that direct copy the phrase "vel quae sint quae antiqui moverint" is found only
is eliminated from textual consideration as a codex descriptus. in a Bamberg MS of the 9th century, the earliest extant, though
But here again, this is to take the text for the codex. For what "quae antiqui moverint" is found as an interlinear gloss in a
is really "descriptus" is the text not the codex. The codex Vatican MS of the mid-12th century. 8
must retain its place in the tradition of the text, and may be The Vatican codex in question, MS Pal. lat. 869, which
called upon when the work of editing begins. A so-called is MS C in Reynolds' edition and in his Medieval Tradition
codex descriptus may well have been in the possession of some of Seneca's Letters (Oxford, 1965), will occupy the remainder
able scholar who had access to manuscripts or testimonia that of this paper. For in fact it proves to be more interesting than
are no longer extant from which he made notes or took variant
readings. On a careful codicological examination they may
one would have guessed from the somewhat bleak description
of it in Reynolds and others. When examined codicologically,
","
prove to carry discerning conjectures or readings which cannot it shows signs on the one hand of having been copied from ,:"
be overlooked. an Insular model (on occasion the obsolete Insular symbol "i~
No matter how logical or consistent or revealing aStemma for "eius" is reproduced, as though the scribe was not quite
codicum may turn out to be, it is the codices as such that sure how to expand it); on the other hand, the codex clearly
matter in the long run, not the Stemma codicum. For all its was collated, shortly after it was written, with a manuscript
seeming inexorable contours, the Stemma codicum is not a of the Bamberg tradition, possibly by its very first owner.
vice from which there is no escape. At best it is no more than This second conclusion is based on the presence in the
a prop: an instrument that enables one to shore up some of Vatican codex of a series of interlinear notes, most of which
the text temporarily, where the witnesses are at variance with are introduced by the "vel" siglum. Reynolds records a few
one another, and hence have weakened the common bond of of these in his apparatus, but seems to look upon them as
text that holds them all together as the work of one author. simple glosses. This is not at all the case. These "vel" clauses
Where the established text is not common to all the codices, are in fact alternative readings, and their distinction from sim-
an editor is not stuck with it simply because it is dictated by ple glosses is clear from the fact that the "vel" siglum is gen-
common variations and the Stemma codicum. When this estab- erally followed not by a full word but by the part of the word
lished text, made up of common and uncommon texts, is seen underneath for which a variant has been found in another
as a whole in this third stage of editing, one or other of the
alternative readings may well prove on critical examination to 81 L. D. Reynolds. ed. L. Annaei Senecae ad Lucilium Epistulae Morales
be more fitting than that selected ad hoc for the established (Oxford, 1965), n. 473 (letter 113).
42 / MEDIEVAL LATIN TEXTS BOYLE / 43

manuscript. Thus over the word "obiurgamur" in letter 93 manuscript, especially when one notes that this clause (though
(fol.33') there is a neat interlinear note "vel mus" above "mur", without "vel") occurs interlinearly above the opening clause
meaning that another MS reads "obiurgamus." "Quae sint quae antiquos moverint" in the Vatican manu-
This "vel" phenomenon, indicating an alternative reading, script. Given that the sentence as a whole in the Bamberg MS
seems to have been a common feature of 12th-century schol- is a little odd, though seemingly all editors accept it without
arship, though I have yet to document it fully. It is nicely to question, and given the fact that "vel" clauses often denote
be seen in the earliest MS of the letters of Peter the Venerable, an alternative reading, it is at least arguable that the original
attesting to the existence of a further early manuscript for sentence may have been a simple "Quae sint quae antiquos
which in fact the recent editor of the corpus of Peter's letters moverint, dicam", and that "vel quae sint quae antiqui
made no allowance in his discussion of the early history of moverint" was originally an alternative reading over the initial
these letters. 9 phrase in some manuscript or other from which the Bamberg
In the present case of Seneca's letters, these "vel" clauses scribe, in a fashion not untypical of scribes, copied both the
are as much part of the tradition of the letters in MS Pal. lat. initial phrase and the alternative as one sentence: "Quae sint
869 as the text itself, and they enable us to add a second lost quae antiquos moverint vel quae sint quae antiqui moverint,
codex to the tradition of the letters, to go with the presumed dicam."
Insular model of the Vatican codex. Of course one may dismiss My final example of the place of codicology in the editing
these "vel" clauses as examples of medieval meddling or "con- of medieval texts is that from which the tide of my paper I'
I'
tamination", but they are part of the codicological tradition comes. Again the text is a letter of Seneca, and again the I

of the text of Seneca's letters and, in fact, provide splendid manuscript is Pal. lat. 869. Writing in his usual moralistic way i
examples of what I have termed "codicological variations" to Lucilius, Seneca chides him on feeling miserable and out
between manuscripts. To ignore them, or simply pass them of sorts, saying "Vesicae te dolor inquietavit, epistulae vener-
over as glosses, is to misrepresent the tradition of the letters unt parum dulces, detrimenta continua-propius accedam, de
in this codex and so miss a rich part of the history of the text capite timuisti. ,,10
as transmitted. This is the received text, except that where all MSS read
An awareness of the significance of "vel" clauses of this "epistulae vero erunt parum dulces", most editors read "ven- "~:
kind may even be helpful when editing an established text. It erunt" with vonJan forthe difficult "vera erunt." The meaning ..
~:

could, for example, lead one to question the accepted reading, is something like this: A pain in your bladder bothered you,
quoted above, of "Quae sint quae antiquos moverint vel quae letters came from you that were hardly pleasant, everything ..
sint quae antiqui moverint, dicam", from the Bamberg went wrong-let me put it bluntly, you began to fear for your
life (or, as translated in the Loeb Classics) "It was disease of
the bladder that made you apprehensive; downcast letters came
9/ The Letters of Peter the Venerable, ed. G. Constable, 2 vols.
(Cambridge. Mass., 1967). The presence of the other early manuscript is
acknowledged in Peter the Venerable. Selected Letters, ed. J. Martin and
G. ConStable (Toronto, 1974). 101 Ep. 96, ed. Reynolds (n. 8 above), n, 401.
44 / MEDIEVAL LATIN TEXTS BOYLE / 45

from you; you were continually getting worse; I will touch If I am correct in taking the text as emended in Pal. lat.
the truth more closely and say that you feared for your life. "" 869 seriously, then what may have happened was that at some
Now in the midst of all these physical ailments, the refer- point or other before the ninth century, when the Bamberg
ence to far-from-sweet letters may seem a little out of place. MS was written, the "epulae" was contracted to "eple", with
But it does not bother modern editors, though one or two a hook through the "I", and because this is very nearly the
cite an alternative to "epistulae" preferred by Erasmus and abbreviation for "epistulae", it would have been quite easy
others after him. 12 Where Erasmus got it I do not know as for the scribe of the Bamberg MS or its exemplar to have
yet, but the reading "epulae" is borne out by an erasure and expanded "eple", meaning "epulae", to the more common
correction in MS Pal. lat. 869, that 12th-century codex which "epistulae. "
was so assiduously collated against another manuscript or It is at least odd that the 12th-century owner of the Vatican
manuscripts by its 12th-century owner, as I presume. In Pal. manuscript should have been moved to change "epistulae" to
lat. 869 (fol 44V) part of "epistulae" has been rubbed out to epulae", if the presence of "epistulae" was as unexceptional
make way for "epulae", and, for good measure, "epulae" is as modern editions suggest it to be. Was "epulae" a conjecture?
written clearly in the margin . Granting, but not at all accept- Hardly, when one remembers that Pal. lat. 869 was so labori-
ing, vonJan's "venerunt" fo r '(vero erunt", the text now reads: ously collated with another or other manuscripts of Seneca's
to
"Vesicae te dolor inquietavit, epulae venerunt parum dulces, Letters. But even if "epulae" be a conjecture, should it there- ,,.
detrimenta continua-propius accedam, de capite timuisti." fore be denied consideration or excluded from the apparatus?
Seneca, then, according to the corrected text in Pal. lat. If a conjecture or emendation by a modern scholar such as
869, is simply listing the complaints of Lucilius, and, for van Jan is deemed worthy of attention, why should not a
effect, goes from bad to worse in, so to speak, ascending medieval conjecture be granted a hearing? At least it is part ,:
B
order: "A pain in your bladder bothered you; eating became
less of a pleasure; everying went wrong-to put it bluntly,
of the codicological tradition of the text and as such merits
the attention of any editor worth his salt.
,.
"
E
you began to feel that you were going off your head. " These two examples from Seneca's Letters of a very ten- "
Given his ailing bladder, it is hardly unexpected to find tative application of codicology in editing text may not seem "
that Lucilius had gone off his food. This fact alone inclines very momentous. Perhaps I may seem to be making too much
me to "epulae" rather than "epistulae." To insist, with Bam- out of nothing, especially in the eyes of these who see codicol-
berg and other MSS, on "epistulae" is to disrupt what is clearly ogy as simply an examination of the physical make-up of a
one disaster after another. Hence Seneca goes on, "A long codex: size, stitching, gatherings, and the like.
life includes all these troubles. Did you not know, when you But to me codicology in its full colours is an examination
prayed for long life, that this was what you were praying for?" of a codex precisely as it is a carrier of a text. If, as is the
general tendency, one simply extracts a text from a codex,
11 1 Seneca. Ad Lucilium Epistulae morales, ed. R. M. Gummere, III (Cam- then the text is bereft of its setting, and the codex is ignored.
bridge, Mass., 1971 ), 107. If, on the contrary, one concentrates on the physical make-up
121 Cl. Seneque. Lettres a Luciiius, ed . F. Prehac, V (Paris, 1964), 114 of a codex, then the danger is that the codex will be treated
(apparatus). in isolation from the text, as though it were any codex and
46 / MEDIEVAL LATIN TEXTS

not just this specific codex carrying this specific text. Codicol-
ogy, then, must include the text-not of course, the text as
text, but as physically carried by the codex, whether this be History of Editing the
the size of the columns, the spirals of initials, the annotations,
or indeed, the smudges of readers.
Greek New Testament'
So codicology is not at all "ueberlieferungsgeschichte",
which, as everyone knows, Housman characterized as "a Bruce M. M etzger
longer and more noble name than fudge. "IJ In other words,
codicology is a history of the fortunes not of a text as text,
but of a text as it is carried by codices. It is a simple and
necessary recognition of the fact that texts have survived be-
cause of codices, and that each codex in turn carries a text in
its own unique fashion.

,.
c
,.
"E
I~.

1. EDITIONS PRIOR TO PRINTING ,;:::

Historians are unable to say exactly when and by whom the "
I:
first edition of the Greek New Testament was made. It is ..!~
;.
13/ M. Annasi Lucani. Belli civilis librj decem, ed. A. E. Housrnan (Oxford, altogether probable, however, that toward the close of the
1926), p. xiii. This remark occurs in a characteristic introduction , in which first century of the Christian era someone set about to collect
he excoriates previous editors and notably C. M. Francken. of whose editton copies of as many epistles of the apostle Paul as could be
in 1896-1897 he writes: "Hardly a page of it can be read without anger found. According to a theory popularized earlier this century
and disgust. Francken was a born blunderer, marked cross from the womb by Edgar J. Goodspeed, this unknown individual was a great
and perversej and he had not the shrewdness or modesty to suspect that admirer of Paul, who having read the account of the apostle's
others saw clearer than he did , not the prudence and decency to acquaint missionary journeys in the Acts of the Apostles, thereupon
himself with what he might have learnt from those whom he preferred to
decided to search out his surviving epistles and to issue an
contradicL" (p. xxxiv), It is therefore a shock to find that Housman's own
edition of the corpus Paulinum at Ephesus in the last decade
edition was not based on a first-hand acquaintance with the MSS of Lucan :
KMy reports of the manuscripts are selected from the apparatus criticus of
Mr. Hosius' third edition, except in a few places.". where I chance to have . / A version of this paper has been published in the Proceedings of the
independent information." (p. xxxv). American Philosophical Society, 130 (1986),148-158.


Leonard Eugene Boyle

*Creggan, Ballybofey, County Donegal, Irish Free State, 13 November 1923
Rome, 25 October 1999
Analecta Hibernica 38 (2004)