Greek and Roman Musical Studies 2017.1 by Greek and Roman Musical Studies

Greek and Roman Musical Studies 5 (2017) 1-17
brill.com/grms
Is the Idea of ‘Musical Emotion’ Present in

Classical Antiquity?
Andreas Kramarz, Ph.D.

Legion of Christ College of Humanities, 475 Oak Avenue, Cheshire,
CT, 06410, USA
[email protected]
Abstract
This article investigates to what degree the concept of ‘musical emotion,’ a term
coined by contemporary psychology, can be traced in antiquity. Hence, it is neces-
sary to begin by clearly defining ‘music’ and ‘emotion,’ in both ancient and modern
understandings. The distinctions between ‘musically induced emotions’ and ‘musical
emotions’ strictly speaking, and between the ‘referentialist’ and ‘absolutist’ (or ‘cogni-
tivist’) school in music psychology structure the question. While most ancient theo-
rists believe that the impact of music on the passions (παθήματα) is of pedagogical or
therapeutical relevance as it is able to create ethos in the human soul through mimēsis,
others, similar to the cognitivists, limit its effect to (aesthetic) pleasure. Emotions
unique to music are not explicitly discussed by the ancient theorists, although an indi-
rect acknowledgment possibly exists in form of metaphorical descriptions of musical
experiences, a certain notion of specifically musical pleasure, and the idea of music’s
magical power.
Keywords
music – emotions – ethos – mimesis – psychology – magic
© koninklijke brill nv, leiden, ��7 | doi 10.1163/22129758-12341286

Downloaded from Brill.com09/02/2020 01:17:12AM
via University of Cambridge
2 Kramarz
Introduction1
In the third book of the Republic (3.401d), Plato has Socrates say to Glaucon
that ‘upbringing2 in music3 is most decisive (or powerful), given that both
rhythm and melody sink most deeply into the inner part of the soul and most
strongly take hold of it.’4 This statement comes at a place in the dialogue where
Plato is laying out the educational principles for the guardians of his ideal
State. While physical exercise shapes their bodies, μουσική is responsible for
shaping their souls.5 Especially during young age, when their souls are most
malleable, they should be imbued with stimuli that form the character that
is appropriate and necessary for their leadership in the State. Also Aristotle
(Pol. 8.1340a22-3) observes that we ‘change our soul when we hear such things,’
referring again to rhythm and melody.6 But what is it that happens to the soul
when music touches it? In the Laws (7.812c), Plato explains that the soul under-
goes passions (παθήματα) in or through rhythm and melody; with this he seems
to mean what today we would call emotions.
The relationship between music and emotions, as obvious as it seems to
be, is actually quite difficult to understand; it already engrossed the authors
of antiquity, and it still keeps contemporary psychology and philosophy of
music engaged in studious research aimed at bringing light into the mysteri-
ous power that music exercises on the human psyche or soul, as the ancients
called it. But before entering into the question whether the concept of a
‘musical emotion,’ a term coined by contemporary music psychology, can be
traced in ancient classical music theory, we should clarify our terminology, for
the terms ‘music’ and ‘emotion’ have not been used univocally in antiquity and
in our own day.
1 This article is a revised version of a paper presented on January 7th, 2016, at the 147th Annual
Meeting of the Society for Classical Studies in San Francisco, CA.
2 Literally: ‘nourishment’ (τροφή).
3 Shorey 1930 translates with “music.” While Emlyn-Jones/Preddy 2013 translate with ‘arts,’ the
following specification ‘rhythm/harmony’ shows that Plato here thinks of music in a more
specific sense. Barker 1984, 135, simply transliterates ‘mousikē.’
4 κυριωτάτη ἐν μουσικῇ τροφή, ὅτι μάλιστα καταδύεται εἰς τὸ ἐντὸς τῆς ψυχῆς ὅ τε ῥυθμὸς καὶ
ἁρμονία, καὶ ἐρρωμενέστατα ἅπτεται αὐτῆς.
5 R. 403c-e; cf. also Cri 50d; Ti 88c; Arist. Pol. 8.2.3 1337b24; 8.4.5 1339a23-5.
6 μεταβάλλομεν γὰρ τὴν ψυχὴν ἀκροώμενοι τοιούτων.

Is the Idea of ‘ Musical Emotion ’ Present in Classical Antiquity ? 3
Definitions
To begin with ‘music’: the problem we have today in defining music centers
largely on questions like how we can distinguish music either from noise
(for instance, by the particular structure of the musical sound wave) or from
spoken language (by the emphasis on pitch in music and timbre in speech).7
For purpose of exploring ‘musical emotions,’ I propose considering music to be
organized vocal and/or instrumental sound (i.e. defined by a set of rules or pat-
terns such as scales, systems, etc., to differentiate it from speech), as produced
and perceived by human beings.8 In order to explore the relationship between
music and emotions, it is necessary to isolate it from lyrics (text). Of course, in
song and most other vocal forms, both text and melody (and other elements)
are intimately united, but only if we try to abstract from any extra-musical
‘message’ can we examine how music on its own influences the emotions.
It is important to notice that separating music from text and movements
like dance for the sake of analysis is not foreign to the Greeks. It is true that
‘μουσική’ often comprises more or different things than what the definition just
given of ‘music’ describes. In Plato alone it can mean the following: philoso-
phy (Phd. 61a);9 the art of properly playing the cithara, singing, and dancing
(Alc 1.108c-d);10 the part of education that corresponds to the soul (as opposed
to the body) and includes words (meaning stories not sung: λόγοι, μύθοι)
(R. 2.376e)11 and song (ᾠδή or μέλος). Song, for its part, is again divided up into
7 See e.g. Kivy 2007; Patel 2008, 86.

8 For a more detailed discussion of the definition of music, see Kramarz 2016, 12-8. The
study of common dictionaries and encyclopedias shows that the primary meaning of
‘music’ is applies to human beings while attributing music to animals or other realities is
taking ‘music’ in a wider or analogous sense.
9 Philosophy is called ‘the greatest music,’ while writing poetry is the popular (δημώδη)
kind; see also the supposed etymology in Cra. 406a. In Ti. 88c (in conjunction with 47b-e),
μουσική seems to be mean ‘harmony,’ as a part of philosophy which, as a whole, keeps the
soul in proper motion (πλάττοντα τὰς τῆς ψυχῆς ἀνταποδοτέον κινήσεις, μουσικῇ καὶ πάσῃ
φιλοσοφίᾳ προσχρώμενον); LSJ translates the term here with ‘generally art or letters.’
10 ἡ τέχνη, ἧς τὸ κιθαρίζειν καὶ τὸ ᾁδειν καὶ τὸ ἐμβαίνειν ὀρθῶς; cf. also Leg. 2.655a: ἐν γὰρ
μουσικῇ καὶ σχήματα μὲν καὶ μέλη ἔνεστιν, περὶ ῥυθμὸν καὶ ἁρμονίαν οὔσης τῆς μουσικῆς: ‘for
in music there are both the dance figures and the melodies/songs, being music about
rhythm and harmonia.’ Originally, any of the arts corresponding to the Muses could be
covered by the term μουσική.
11 That Plato first meant to discuss poetry not sung emerges from R. 3.398d; on the other
hand, already at 397b-c, ἁρμονίαι and rhythm are mentioned, maybe just referring to
spoken pitch and rhythm.
Greek and Roman Musical Studies 5 (2017) 1-17 Downloaded from Brill.com09/02/2020 01:17:12AM
4 Kramarz
text (λόγος), melody (ἁρμονία),12 and rhythm (ῥυθμός) (R. 3.398c-d). But this
latter distinction and the individual discussion of harmoniai and rhythms
that follows in the Republic (and the scrutiny of appropriate text ‘message’
in what precedes) show clearly that Plato here considers the effect of these
musical elements by themselves.
The same can be said about Aristotle who, in his eighth book of the Politics,
discusses the educational value of music in a manner similar to Plato. He
defines music as comprising melody- (or song-) making and rhythm13 and
seems to distinguish μουσική with melody (μετὰ μελῳδίας) from ‘bare’ (ψιλή)
music, which is usually translated with ‘instrumental.’14 All of Aristotle’s fol-
lowing treatment is to be understood to be about music as defined above,
regardless of lyrics. Expositions of later authors follow the same course.15
12 Barker in GMW 1.130 n. 18 stresses that an “organised scheme of pitches” is meant, not
simply ‘melody.’ Plato’s following discussion of harmoniai in the sense of modes seems
to confirm this understanding. I still translate with ‘melody’ as a term covering a general
musical parameter as opposed to rhythm, instrumentation, or harmony. Unfortunately,
since many of the musical terms in ancient Greek are not used consistently with the
same meaning, one needs to ponder carefully the meaning in each case.
13 μουσικὴν ὁρῶμεν διὰ μελοποιίας καὶ ῥυθμῶν οὖσαν (8.7.2 1341b23-4). He does not explicitly
include dance here, which is mentioned in Po. 1447a26-28.
14 τὴν δὲ μουσικὴν πάντες εἶναί φαμεν τῶν ἡδίστων, καὶ ψιλὴν οὖσαν καὶ μετὰ μελῳδίας (8.5.1
1339b21). Rackham 1932 translates: ‘whether instrumental or instrumental and vocal
music together.’ Barker GMW 1.174 n. 8 ad loc. supports ‘instrumental;’ this is also the defi-
nition given by LSJ ad loc., but LSJ offers for ‘ψιλός’ also the meaning ‘mere poetry, without
music’ (e.g. Pl. Smp. 215c), wherefore someone might argue in favor of translating ‘bare
of melody = text without song.’ However, the whole context suggests that Aristotle talks
about ‘music’ strictly speaking and not about the effect a song may have because of its
textual meaning (see the reference to the harmoniai later on in 8.5.8 1340a41 or 8-5.10
1340b18). The separation of various musical elements even in artistic practice, each one
eliciting its own ἤθη καὶ πάθη καὶ πράξεις: ‘ēthē, passions/emotions, and actions,’ with the
same effect (οὖσαι τοιαῦται τὴν δύναμιν), appears also in Poet. 1447a25-30; cf. also Pl. Plt.
268b with a similar idea (οὐκ ἄλλος κρείττων παραμυθεῖσθαι καὶ κηλῶν πραΰνειν, μετά τε
ὀργάνων καὶ ψιλῷ τῷ στόματι τὴν τῆς αὑτοῦ ποίμνης ἄριστα μεταχειριζόμενος μουσικήν: ‘no
other encourages and by charming soothes better, with instruments and with the mouth
only, administering the music best for his flock’). Similar reasons and the strong argu-
ments presented by Ford 2004, 320-29, support the much-discussed Susemihl emenda-
tion for Po. 1340a14, suggesting the consideration of rhythms and tunes without words.
15 Aristoxenus’ remark that the ethos of a musical piece should come from considering the
whole and not (only) from its parameters might actually be a certain reaction to such dis-
secting approaches; see in Ps.-Plu. De Musica 32.1142d-33.1143d and Rocconi 2012, 76-81.

Next, we need to clarify the concept of ‘emotion.’ Based on current

psychology,16 I suggest the following definition of emotions as transient neu-
rophysiological states that arise in response to a conscious or sub-conscious,
real or imaginary stimulus (‘object’) provided by the exterior or interior senses
or by the mind only.17 The term ‘emotion’ is to be distinguished from ‘moods’
(affective states, less intense but lasting longer than emotions and without
a clear ‘object’), ‘feelings’ (the subjective experience of emotions or moods),
and ‘affections’ (a generic label containing all of the above).18 Typical basic
emotions are joy, sadness, anger, and fear. Many models exist to classify
the emotions and to subdivide them further.19
In antiquity, the term most equivalent to ‘emotion’ would be πάθος or
πάθημα. Plato distinguishes three parts of the soul: the ‘rational’ (λογιστικόν
or φιλόσοφον), the ‘passionate’ (θυμοειδές), and the ‘desiring’ (ἐπιθυμητικόν).
We might associate emotions with the passionate and desiring parts since
they are both described as opposed to reason.20 In his Nicomachean Ethics,
Aristotle uses a simplified division of the soul into a rational and an irrational
part.21 The sensations of pleasure (ἡδονή) and pain (λύπη)—on which emo-
tions are based—reside primarily in the sensitive faculty (De an. 2.2.413b23-25)
but are also interrelated with the intellect (id. 3.7.431a1-3.8.432a14). Aristotle
offers several lists of emotions in the Eudemian Ethics and in De Anima and
undertakes a more systematic treatment in his Rhetoric.22 Most of the later
16 The anthropological continuity of the human species over the past millennia does not
suggest a significant difference between antiquity and today regarding the natural pro-
cesses that underlie what we today call ‘emotion.’ The concept itself is more difficult to
compare across times and cultures; see Konstan 2006 and below.
17 The OHME 10 offers this definition: “a quite brief but intense affective reaction that usually
involves a number of sub-components—subjective feeling, physiological arousal, expres-
sion, action tendency, and regulation—that are more or less ‘synchronized’. Emotions
focus on specific ‘objects’ and last minutes to a few hours.” The “focus on specific objects”
is a point of debate, as we shall see.
18 Definitions paraphrased from OHME 10. See also Arist. EN 2.5.1-6 1105b19-1106a13 who indi-
vidualizes the states of the soul according to emotions (πάθη), faculties/powers (δυνάμεις),
and habits (ἕξεις, to which correspond virtue and vice).
19 See, for instance, Hevner’s ‘Mood Wheel,’ which applies the same to moods and emotions
and is used in music therapy: Wigram et al. 2002, 57-61.
20 R. 10.603a-605b; see also Lg. 9.863b: here θυμός is a πάθος or an element of the soul, and is
opposed to ἠδονή, thus possibly preparing the Aristotelian distinction of the irascible and
appetitive passions within the irrational part of the soul.
21 E N 1.13.9-19 1102a27-1103a3.
22 What appears in the table in EE 2.3.4 1220b38-1221a12, even if called πάθη, are really ἤθη
of vices, as Aristotle analyzes the relationship between emotionally provoked negative
6 Kramarz
authors in antiquity follow these conceptions about the soul and the passions
with minor adjustments.23
However, the attempt to gain a clear concept of how the ancient Greeks (and
Romans) defined and experienced emotions is met with complications. David
Konstan, reviewing much of the current scholarship on the issue, has called
attention to the fact that the general understanding of what emotions are
and their precise taxonomy depend on socio-cultural factors and, therefore,
is subject to change.24 On the other hand, he concedes that “there are broad
similarities between the ancient pathê (. . .) and modern emotions” (2016, 260)
as a general human experience. While it is not possible here to demonstrate
this in depth, I hope that the definition given above is open and clear enough
to identify cross-culturally a specific kind of response to music.25 At the same
time, we shall need to keep in mind that the varying nuances within the
states and the virtue as the proper mean. A smaller list of actual emotions appears earlier
at 2.2.4 1220b12-14: θυμὸν φόβον αἰδῶ ἐπιθυμίαν, ὅλως οἷς ἕπεται ὡς ἐπὶ τὸ πολὺ ἡ αἰσθητικὴ
ἡδονὴ ἢ λύπη καθ᾽ αὑτά: ‘anger, fear, shame, desire, generally those [emotions] that in
themselves are mostly accompanied by sensory pleasure or pain.’ It is interesting that
Aristotle introduces here pleasure and pain as a common factor in all emotions (in a way
already in Pl. Lg. 2.653a-b), something which modern psychology calls “valence” (positive
or negative: attractiveness or aversiveness of the object of the emotion; for the application
to music e.g. OHME 608-15). Other lists of passions/emotions in Aristotle can be found in
De an. 1.1 403a7-8 (ὀργίζεσθαι, θαρρεῖν, ἐπιθυμεῖν, ὅλως αἰσθάνεσθαι / anger, courage, desire,
all of sensation) and 1.1 403a17-18 (θυμός, πραότης, φόβος, ἔλεος, θάρσος, ἔτι χαρὰ καὶ τὸ φιλεῖν
τε καὶ μισεῖν / anger, gentleness, fear, pity, courage, joy, love, hating)— it is interesting
that here he also inquires to what degree the body is involved. For the Rhetoric, see book 2,
chapters 1-11, followed by a discussion of ēthē in 12-7.
23 Some Stoics (such as Chrysippus) emphasize more a cognitive side of the emotions or
passions as ‘evaluative judgments’ and reject the notion of a tripartite soul, as laid out in
Nussbaum 1993, 97-149.
24 Konstan 2006. By comparing catalogs of ‘basic emotions’ in multiple authors and analyz-
ing the examples given especially in Aristotle’s Rhetoric, Konstan holds the various socio-
logical milieus responsible for the differences in the emotional vocabulary. For instance,
the pre-Hellenistic Athenian society “understood emotions as responses (. . .) to actions,
or situations resulting from actions, that entail consequences for one’s own or others’
relative social standing” (id., 40) as opposed to an individual ‘inner state’ (id., 31), which
would be the prevalent aspect in defining emotion during Hellenistic or even modern
times.
25 Konstan (2006, 39) holds that “the feelings inspired by music (. . .) do not count as pathê
for Aristotle.” This may be true for the Rhetoric but not for the Politics where the word
‘πάθος’ is used in the context of music (e.g. 8.5.5 1340a13 or 8.7.5 1342a5), and I would argue
that the concept is present throughout the section.

emotional spectrum may generate incongruences between both perception

and description of the emotional impact of music in antiquity and today.
We are now prepared to investigate the concept of ‘musical emotions.’ This
term can be understood in two ways: first, we could call any emotion ‘musical’
whenever it is triggered by a musical experience (e.g. by listening to, actively
producing, or imagining or dreaming of music).26 This occurs, for instance,
when people say that they feel happy or uplifted when listening to a tune
that they experience as rousing. The expression ‘musically induced emotion’
may serve to differentiate this general type from the other usage of ‘musical
emotions’ properly speaking, which refers to emotions that are thought to
be unique to music, i.e. which occur only if one is exposed to music. There is
much debate about whether ‘musical emotions’ in this sense actually exist,
but we need to hold off that question for a moment. Notice that ‘musical emo-
tion’ does not refer to the cases where music is used to express emotion, but
to provoke emotion. For in the first case, the emotion preexists independently
from music and is thought to be translated somehow into a musical ‘language;’
we are here concerned with the emotions that follow a musical experience as
their direct and primary cause.27
Ancient Use of Musical Emotions
Musically induced emotions are an everyday experience both now and then,
and are reflected throughout in literary writings; a well-known example is
Achilles delighting (τέρπειν) while he accompanies his own singing with the
phorminx (Hom. Il. 9.186-9).28 For the ancient theorist, they are relevant, first,
as mentioned above, within the context of educating the young, specifically of
character formation, and, second, they are considered for acquiring or restor-
ing a proper emotional state, a procedure similar to what today we call music
therapy.
26 This is the definition employed in the OHME, see p. 10.

27 Statistically, the expression of existing emotions is far more frequently reported in
accounts of music in ancient literature; see chapter 2 in Kramarz 2016.
28 Admittedly, the literary sources often do not distinguish whether an effect is brought
about mainly through the music itself or through the text or the necessary combination
of both.
8 Kramarz
Ethos
In order to describe the kind of emotions to be induced, the theorists employed
another related concept that they called ēthos (ἦθος). Ethos, in the meaning of
‘individual patterns of being or behavior’ (which are not necessarily moral),29
usually refers to individual people (such as gentleness, irascibility, courageous-
ness, relaxedness, etc.), but can then also be applied to groups of people, ani-
mals, and other realities—such as music. Most ancients believed that music
forms character by means of habituation, through frequent repetition of the
emotional response to music that possesses a specific ethos.30 This leads to
the question of how does music acquire ethos, and how can it influence human
ethos? Without being able to give an answer with the necessary detail,31 here is
a brief account of what the ancients thought about it.
Mimēsis
Music as an art form is identified as a kind of mimēsis, a term that can range
from ‘imitation,’ to ‘representation,’ or ‘expression.’32 Plato is the first known
author to bring a pre-existing notion (and practice) of mimēsis to the level of
explicit theoretical reflection.33 To give an example from the context of music:
ethical qualities of manliness (or courage) and self-control are to be instilled
through music that ‘imitates’ or expresses the utterings of someone who shows
manliness (or courage) in warfare and endures in failure (R. 399a-b).34 It is hard
for us to imagine how a melody could possibly resemble such utterings in any
29 “Ethos” is close, but not identical, to what we call ‘character,’ in part because it does not
need to mean a stable disposition but can be a simple ‘characteristic’ of something, and
also because it would describe only some trait and not all of the character of a person. Some
descriptions of ethos are used for virtues or vices, but ethos would coincide with them
only when describing human habits that incline positively or negatively towards action.
Hence the question whether music disposes to virtue, something which Philodemus
vigorously denied.
30 See Arist. Pol 8.5.6 1340a23-29; cf. Aristid. Quint. 2.6 63.31-64.9.
31 This is done in Kramarz 2016 where the existing scholarship on the issue is reviewed and
updated; see also Barker 2005.
32 The most thorough study of the term and its history, with particular emphasis on
Plato and Aristotle, is Halliwell 2002 (who deals with the pre-Platonic development
on pp. 17-22).
33 Halliwell calls him the “ ‘founding father’ of mimeticism” (2002, 24; see also p. 37) who,
at the same time, employs the concept within a large range of contexts and treats it in a
quite complex, by no means uniform way.
34 I am summarizing; Plato’s description of the situation is even more elaborate. In Barker’s
translation (GMW 1.131), we read: ‘the sounds and cadences of a man who is brave in deeds
of war and in acting under pressure of any kind, and who, if he is faced with wounds or

precise way, but the point is that Plato supposes a parallelism between musi-
cal structure and ethic-revealing sound or behavior in real life, and that this
parallelism transmits the musically expressed ethos to the listener. Aristotle
speaks in similar terms when he affirms that the ‘similarities’ of certain music
make souls ‘enthusiastic,’35 or show anger, mildness, manliness or courage or
self-control and their opposites, or in general, they make people ‘feel along’
(συμπαθεῖς), for ‘songs themselves contain the mimēmata of ēthē.’36
The fact that our souls are susceptible to music at all is explained by Plato
(and in a similar vein by many later authors) in the Timaeus: the cosmic and
human soul are governed by the same principles, and music, being based on
mathematically describable proportions, forms the link between the two.
Thus, to music corresponds the role of “tuning” the soul according to cosmic
harmony.37 Aristides Quintilianus develops the perhaps most complex theory
of how various musical parameters together create specific ethos in the dif-
ferent parts of the human soul. For one part, he endeavors to assign ethical
qualities to all musical parameters, even to the individual notes, intervals,
modes, melodic styles, rhythms, etc., and finally to the musical piece as a
whole. Secondly, he establishes a male-female polarity, on which these quali-
ties are based, and this polarity can also be found in the structure of both the
universe and the soul. For Aristides Quintilianus, mimēsis is not a simple musi-
cal mimicking of extra-musical forms, but an intrinsic equivalency between
music, the soul, and the cosmos, through which they communicate in ethos.
This nexus provides the paideutic or therapeutic musician with the ability to
arrange musical ethos in accordance with the ethos of text, and both accord-
ing to the ethical needs of each individual. Aristides’ system is perplexing and
fascinating at the same time, and we cannot appreciate it here in detail,38 but
in its multiple explicit and implicit references to earlier authors of antiquity
it constitutes the culmination of a tradition according to which the interior
makeup of music creates or influences pathos and ethos in the human being.
As we see, there is plenty of evidence that ancient music theorists subscribed
death or falls into any other catastrophe, confronts his fate in all these situations with
self-discipline and steadfastness.’
35 ταῦτα γὰρ ὁμολογουμένως ποιεῖ τὰς ψυχὰς ἐνθουσιαστικάς, ὁ δ᾽ ἐνθουσιασμὸς τοῦ περὶ τὴν
ψυχὴν ἤθους πάθος ἐστίν (Pol. 8.5.5 1340a10-12; the following references are from lines 20-21
and 13). It is significant that Aristotle calls ‘enthusiasm’ here a passion (or emotion) of
ethos, in other words, an emotion that leads to a specific ethos.
36 ἐν δὲ τοῖς μέλεσιν αὐτοῖς ἔστι μιμήματα τῶν ἠθῶν (ibid., lines 39-40).
37 See Barker 2005, 33-47; 120-8; Pelosi 2010, 68-113.
38 For a thorough analysis of this author and his theory, see Kramarz 2016, 315-48.
10 Kramarz
to the general concept of musically induced emotions. We may add that in any
case the addition of text or other extra-musical elements are certainly capable
of specifying the ethos further and of increasing the degree of the emotional
impact in a musical piece.39
The Debate about the Relationship between Music and Emotion
However, there are two known ancient authors (the Epicurean Philodemus of
Gadara and the Skeptic Sextus Empiricus, along with a brief section (1.13) from
the so-called Hibeh Papyrus from the fourth century BC) who deny music any
usefulness for education or therapy. Both Philodemus and Sextus Empiricus
refuse to acknowledge any specific emotional impact of music that would go
beyond the aesthetic experience of pleasure (ἡδονή).40 Their position stands
as a minority against a mainstream of authors who take the ethical impact of
music for granted.41 It is interesting to notice how these two camps align with
two contemporary schools of thought which Leonard Meyer has called ‘refer-
entialists’ and ‘absolutists’: the former claim that music in some way refers “to
the extra-musical world of concepts, actions, emotional states, and character,”
39 One could suspect a contradiction in Plato when he attributes to melody and rhythm
clear ethical characteristics (R. 401d) but elsewhere (Lg. 669e) complains about the
difficulty to identify the ethos of music without text. However, the context of the Laws
passage is about confused and corrupted music which, for that reason, lacks clear ethos.
What remains true is that identifying exact musical ethos is the task of the musical expert
(see R. 398d-e; 400a-c; also Ps.-Arist. Pr. 19.27 919b36-37), but it does find its ultimate effec-
tiveness in combination with text.
40 A closer analysis of their reasoning reveals that, for the most part, they limit themselves to
simply stating that the various supposed effects of music do not exist or should be attrib-
uted to the lyrics (text) instead of the music. Their only argument of true substance is the
empirical evidence that some who are exposed to music of a specific ethos do not show
signs of having assimilated that particular ethos (Hibeh 1.13, lines 13-22; S. E. M. 6.15; Phld.
Mus. 3.11 Kemke = 65 Delattre); for a more detailed analysis, see Kramarz 2016, 275-98.
41 Bonds 2014, 17, draws the lines in a different way, between Orpheus and Pythagoras,
by stating that they “embody two fundamentally different perspectives on music that
together circumscribe the foundation of Western attitudes toward the art. As a musi-
cian, Orpheus demonstrated music’s effect; as a philosopher, Pythagoras explained its
essence.” As intriguing as this might at first sound, when Bonds later (p. 22) asserts that
“Pythagoras’s explanation of music’s essence was at the same time an explanation for its
effect, as realized through the skill of Orpheus,” this relationship of “mutual reinforce-
ment,” his initial distinction does not seem to be as fruitful as the one presented here,
which focuses on what kind(s) of effect music can have at all.

while the latter see the meaning of the experience of music only “within the
context of the work itself, in the perception of the relationships set forth within
the musical work of art.”42 Peter Kivy, one of the most pronounced promoters
of musical ‘cognitivism’ (which is widely equivalent to ‘absolutism’), holds that
music, on its own, cannot arouse ordinary emotions because these emotions
depend on a corresponding ‘object’ (or cause) which music does not provide.43
It is certainly true that nobody has a reason to feel literally ‘sad’ just because of
listening to a Prelude by Chopin in a minor key, since there are no disappointed
expectations or negative consequences for one’s life. The ancient critics of the
common view of musically induced emotions did not employ this argument
directly. Instead, Philodemus, for instance, claims that it is a fallacy to attribute
the change of affective states to music instead of the real cause, which is the
text,44 or that music can display ethos (and, hence, instill virtue) just as little as
cookery,45 a typical reductio ad absurdum. Philodemus and Sextus do not offer
much proof for their objections, but their reservations seem, at least in part,46
42 Meyer 1956, 1-3; the author notes that this distinction is not the same as “formalist” and
“expressionist,” because among the latter, there are “absolute expressionists” who believe
that an emotional response to music functions without reference to extra-musical reali-
ties, and “referential expressionists” who “would assert that emotional expression is
dependent upon an understanding of the referential content of music” (id., p. 3). Reviews
and assessments of the most important theories offer Budd 1985 and Davies 2010.
43 E.g. Kivy 1990, 171: “Being moved by music and the descriptions we give of music in emo-
tive terms—sad, hopeful, happy, angry, and the like—are independent phenomena,
related only in the sense that I might be aroused to ecstasy by the beauty (say) of a par-
ticularly anguished passage in a musical work;” and 194-5: “Music alone is about nothing
at all, and the inference from its sadness or joy, tranquility or turbulence, to its ‘aboutness’
a false one. (. . .) The expressive properties of music alone are purely musical properties,
understandable in purely musical terms.”
44 E.g. Phld. Mus. 3.64-66, 57, 67 Kemke = 95.29-97.45 Delattre. For Philodemus, something
irrational such as music cannot have an effect on a rational moral disposition (virtue)
(3.39, 31 Kemke = 83 Delattre). Sextus Empiricus argues that it is philosophy, not music,
which restrains the soul’s passion: M. 6-7 and 17 (paragraph numbering according to the
edition by Greaves 1986). At the most, music offers a distraction but no real change of
the mind: M. 16.
45 E.g. Phld. Mus. 4.3 Kemke = 117.28-35 Delattre. Sextus holds that ethical states such as
love or intemperance (in the case of Achilles in Hom. Il. 9.186-189) attract certain music
instead of the music creating these states (M. 19)—actually an implicit (and most prob-
ably unintended) argument in favor of musical mimēsis: how else would certain states of
mind attract specific kinds of music?
46 Other reasons include the specifically Epicurean and skeptical approach to music and
science in general.
12 Kramarz
to stem from a general sense that the connection between ordinary emotions
and music is not satisfactorily accounted for.
This is not the place to resolve the intriguing debate between these schools,47
but one problem that I see in the cognitivist position and which I would like
to comment on here is that it does not consider the possibility of ‘musical
emotions’ properly speaking.
Specifically Musical Emotions
Most of us would probably agree that the emotional experience of a Chopin

Prelude is indeed not quite sadness, and we might call it “sad” simply for lack
of a better description.48 Kivy does admit the possibility of an emotional
response (such as awe) to the beauty of music in its structure or understood
expressiveness, but according to him, any other emotion is owed to factors
extrinsic to music. I venture to disagree and propose that it is possible, based
on contemporary research on the various mechanisms that induce emotion
through music,49 to explore phenomena such as ‘emotional contagion’ for a
model of emotions that originally derive from ordinary life experiences (such
as ‘tension’ and ‘relaxation’); these experiences reverberate emotionally in
musical patterns and thus become ‘musical emotions,’ without requiring
‘objects’ which define the ordinary emotions. Thus understood, musical emo-
tions may be genetically related to, but not identical with, ordinary emotions.
47 Psychological studies have proven “quite conclusively that music does evoke emotions,
and that the strong version of cognitivism is thus untenable” (Juslin, P.N. / Sloboda, J.A. in
the OHME 83).
48 This lack of vocabulary would explain why also the ancients characterized music in most
instances with terms taken from ordinary emotions, which in itself does not disprove
that they were sensitive as well to realize a certain incommensurability of those terms.
Martha Nussbaum (2001, 249-294) builds her argument in favor of musical expressivity
on emotional content that is not necessarily linguistically describable, wherefore music
might ‘describe’ emotions even more precisely than language. However, Nussbaum does
not consider the concept of specifically musical emotions.
49 See e.g. Juslin in the OHME 619-23. Some of these mechanisms are: rhythmic entrainment,
evaluative conditioning (frequently pairing music with other emotional stimuli), visual
imagery (visualizations while exposed to music), episodic memory (emotionally charged
events of the past linked to music), musical expectancy, and cognitive appraisal (whether
music fulfils or not an expected function).

In addition, the enjoyment felt in the aesthetical experience of music in its

harmonic beauty could be identified as a source for musical emotion.50
Our final question now is whether anywhere in ancient Greek or Roman
descriptions or theoretical reflections on music the idea of specifically musical
emotions was entertained. As far as the theorists are concerned, their primar-
ily paideutic interest in modelling human ethos does not motivate them to
reflect about emotions independently from the ability of music to constitute
a mimēsis of extra-musical ethical realities. The ‘absolutists’ Philodemus and
Sextus Empiricus could have explored the non-ethical effects of music, had
they not limited themselves to ridiculing and deconstructing the usefulness
of music as assumed by their adversaries. But there are still other traces of an
intuition that the effect of music can transcend its ethical dimension.51
In literary sources we can observe that the Greek language is particularly
rich in the use of characterizations for music that are often metaphorical
and borrowed from other senses.52 From a host of possible examples, I shall
illustrate this here with just a few. The Greek language is capable of providing
a great variety of compound terms with the root for ‘honey’ (μελ-) or sweet-
ness/pleasantness in general (ἡδυ-) which pervade in a special way Pindar’s
odes53 but are also very frequent in Latin authors (suavis, dulcis, mell-).54 Other
examples could be ἐρατός (‘lovely’),55 λεπταλέος (‘fine, delicate’),56 or, on
50 For a more detailed exposition of this model, see Kramarz 2016, 417-39.
51 This is not to say that musical emotions strictly speaking are non-ethical; but insofar
as they break up the identification of the musical experience with ordinary emotions,
their ethos would not align anymore so easily with the characterizations used for musi-
cal paideia which need to be labeled according to the standards of human ethos (or
character).
52 See Kaimio 1977, Rocconi 2003, and the synopsis of terms in Kramarz 2016, 65-136.
53 E.g. O. 7.7-8: the poet sends ‘nectar’ as gift of the Muses, ‘sweet fruit of the heart/mind’
(γλυκὺν καρπὸν φρενός); then 11-12: ἄλλοτε δ’ ἄλλον ἐποπτεύει Χάρις ζωθάλμιος ἁδυμελεῖ θαμὰ
μὲν φόρμιγγι παμφώνοισί τ’ ἐν ἔντεσιν αὐλῶν: ‘Charis, giving the bloom of life, looks upon
the one or the other man, often with sweet-sounding phorminx or with the all-voiced
instruments of the auloi.’
54 E.g. Apul Met. 5.15.8-10: Nec tamen scelestarum feminarum nequitia vel illa mellita
cantus dulcedine mollita conquievit: ‘And yet the vileness of the wicked woman did not
rest softened even by that honey-like sweetness of the song.’ Latin, however, displays a
much more limited use of compounds.
55 E.g. h. Hom. 3.515 or 4.153, 423, 455 (all referring to the cithara).
56 Hom. Il. 18.569-71: τοῖσιν δ᾽ ἐν μέσσοισι πάϊς φόρμιγγι λιγείῃ ἱμερόεν κιθάριζε, λίνον δ᾽ ὑπὸ
καλὸν ἄειδε λεπταλέῃ φωνῇ: ‘in their midst a boy played with the clear phorminx lovely,
he sang the beautiful Linos tune with a fine voice.’ This quote combines a number of
characterizations, underlined.
14 Kramarz
the unpleasant end, ἐχθρός (‘hated, hateful’),57 or in Latin, compounds from

horr-.58 Since specifically musical emotions elude precise descriptions, trans-
ferring words from other experiences can be taken as an attempt to approxi-
mate the various effects of music.
Another concept within which an idea for something specifically musical
could be hidden is that it creates pleasure or enjoyment (ἡδονή). Again, liter-
ary descriptions abound.59 Plato is not oblivious to this fact,60 and Aristotle
acknowledges music as one of the most pleasant experiences.61 On the other
hand, to my knowledge, neither Plato nor Aristotle or any of the later authors
reflect beyond the point of intensity about the kind of pleasure which music
produces in comparison to other perceptions,62 and so we are left for the most
57 A. Th. 870, describing the victory song (παιάν) of Hades, along with the shrieking hymn of
the Erinys (τὸν δυσκέλαδόν θ᾽ ὕμνον Ἐρινύος).
58 E.g. Ov. Fast. 4.190: ‘horrendo lotos adunca sono / the bent pipe with terrible sound.’
59 One for many: Thgn. 1.531-3· Aἰεί μοι φίλον ἦτορ ἰαίνεται, ὁππότ’ ἀκούσω αὐλῶν φθεγγομένων
ἱμερόεσσαν ὄπα. χαίρω δ’ εὖ πίνων καὶ ὑπ’ αὐλητῆρος ἀκούων, χαίρω δ’ εὔφθογγον χερσὶ λύρην
ὀχέων. ‘It was always dear to me to melt the heart when I hear the charming voice of
sounding auloi. I rejoice over drinking well and listening from the aulos player; I rejoice
holding a well-sounding lyre in the hands.’
60 Despite all that is usually considered to be severe and morality-driven censorship in
Plato’s education system, he actually does not intend to suppress pleasure in music; he
only rejects it as a criterion for properly judging music; see e.g. Lg. 653d-654a; 655c-d;
802a-dc.
61 Pol. 8.5.1 1339b20-21; cf. 8.5.10 1340b17: ἡ δὲ μουσικὴ φύσει τῶν ἡδυσμένων ἐστίν: ‘music is by
nature one of the sweetest/most enjoyable things.’ Aristotle’s common sense approach
speaks in these lines (Pol. 8.5.4 1340a4-6): ἔχει γὰρ ἡ μουσική τιν᾽ ἡδονὴν φυσικήν, διὸ πάσαις
ἡλικίαις καὶ πᾶσιν ἤθεσιν ἡ χρῆσις αὐτῆς ἐστι προσφιλής: ‘music dispenses pleasure of a
natural kind, so that the use of it is beloved by all ages and characters’ (translation Barker
GMW 1.174).
62 In Ti. 80b, Plato distinguishes ἡδονή as felt in the perception of the polar interplay of
musical movements by people who do not think (τοῖς ἄφροσιν) from εὐφροσύνη as felt by
those who do (τοῖς ἔμφροσι). This (intellectual) ‘delight’ is owed to the mimēsis of divine
harmonies in mortal motions. Whether both reactions (which differ not on the side of
music but on the side of the perceiving subject) are specific to music or would be the
same in the case of other (non-musical) forms of mimēsis remains open; in the first case,
we would have a hint to two different musical emotions. I am grateful to an anonymous
reader for reminding me of that reference for this context.
Aristotle does mention that senses other than hearing do not transmit ethos or do so
only in a symbolic way (as σημεῖα) but not by themselves as do melodies/songs (Pol. 8.5.6-
10 1340a23-1340b19). In addition to forming character, music is ideal for education, since
by its enjoyment it sweetens up the pedagogical procedure (ibid., at the end lines 15-17);
see about the same in Arist. Quint. 2.3 55.4-56.5.

part with the general notion that music is simply “enjoyable,” i.e. it predomi-
nately produces emotions of a positive, agreeable kind.63
One last hint at an emotional force of music that otherwise does not
exist could be seen in the descriptions of its magical power which at times
is unescapable—with the Sirens as its notorious personification.64 Here are
meant not the instances of miraculous musical power over plants, animals,
and even rocks as attributed especially to Orpheus,65 but the enchanting or
mesmerizing effect that music is often reported to elicit, usually described with
verbs such as θέλγειν,66 ἐπαοιδεῖν,67 κηλεῖν,68 or, in Latin, especially mulcēre.69
But also here, musical magic, by its very definition, needs to remain mysterious
and without explanation. Furthermore, it is hardly ever clear whether musical
magic is imagined to function on the level of emotions or through some other
coercive mechanism.70 It seems reasonable to surmise, though, that if music
can be thought to work at times on the emotions in a way that the ancients
interpret as magical, it is quite likely that with this they try to describe an effect
that is different from ordinary emotional experiences.
63 Examples for not enjoyable music are much less frequent in ancient literature than
the positive ones; they are mostly due to a bad job on the part of the musician (e.g. Hor.
Ars 347-350) or because the musical context already carries a negative emotional charge,
such as lamentations, danger, etc.
64 The original reference is Hom. Od. 12.39-54, 183-198, but the topos has been used by many
later authors to describe the luring character that music can have, e.g. D.Chr. 32.47; Arist.
Quint. 1.19 90.27-30.
65 A compilation of most of the many stories about Orpheus in previous authors provides
Ovid in Met. 10.1-105 and 11.1-66.
66 E.g. Pi. P. 1.12: κῆλα δὲ καὶ δαιμόνων θέλγει φρένας: ‘the shafts [of the phorminx] enchant the
minds of the gods.’
67 E.g. Pl. Lg. 812c, speaking of the παθήματα of the soul affected by good or bad mimēsis of
songs, wherefore the souls of children should be ‘charmed’ (ἐπᾴδῃ ταῖς τῶν νέων ψυχαῖς) by
good ones.
68 E.g. E. Alc. 359, wishing to be able to charm with hymns like Orpheus (ὕμνοισι κηλήσαντά).
69 E.g. Ov. Met. 10.303: mea si vestras mulcebunt carmina mentes: ‘if my songs will charm
your minds.’ Another vivid expression be cited from Macr. 2.1.5: puella ex industria supra
naturam mollior canora dulcedine et saltationis lubrico exerceret inlecebris philosophantes:
‘the girl—made up to be more voluptuous than nature intended—would charm them
with her sweet singing and lewd dancing while they philosophized’ (transl. Kaster 2011).
70 I say “hardly ever” because the Plato reference from above (n. 59) does establish a connec-
tion between the emotions and “enchantment.”
16 Kramarz
Conclusion
Our conclusion, then, is that ancient authors did explore abundantly the phe-
nomenon of musically induced emotions, especially for the sake of their ethi-
cal implications. Emotions unique to music, on the other hand, did not make
it into their theoretical reflection and can at most be conjectured from meta-
phorical literary descriptions, from an aesthetical experience that appears
to distinguish music from other art forms, and from the possibility that the
ancients take recourse to the idea of magic because the power of music tran-
scends the ordinary perception of emotion and evades other explanations.
Bibliography
Barker, A. 1984/1989. Greek Musical Writings. 2 vols. (Cambridge/New York) (cited as

GMW).
Barker, A. 2005. Psicomusicologia nella Grecia antica. Tr. Angelo Meriani (Napoli).
Bonds, M.E. 2014. Absolute Music: the History of an Idea (Oxford/New York).
Budd, M. 1985. Music and the Emotions: the Philosophical Theories (London/Boston).
Davies, S. 2010. Emotions Expressed and Aroused by Music. Philosophical Perspectives, in:
Juslin, P.N., Sloboda, J.A. (eds.) Handbook of Music and Emotion: Theory, Research,
and Applications (Oxford/New York), 15-43.
Ford, A. 2004. The Power of Music in Aristotle’s Politics, in: Murray, P., Wilson, P. (eds.)
Music and the Muses: the Culture of ‘Mousikē’ in the Classical Athenian City (Oxford/
New York), 309-336.
Halliwell, S. 2002. The Aesthetics of Mimesis: Ancient Texts and Modern Problems
(Princeton).
Juslin, P.N., Sloboda, J.A. (eds.) Handbook of Music and Emotion: Theory, Research, and
Applications (Oxford/New York) (cited as OHME).
Kaimio, M. 1977. Characterization of Sound in Early Greek Literature (Helsinki).
Kivy, P. 1989. Sound Sentiment: an Essay on the Musical Emotions, Including the Complete
text of The Corded Shell (Philadelphia).
Kivy, P. 1990. Music Alone: Philosophical Reflections on the Purely Musical Experience
(Ithaca).
Kivy, P. 2007. Music, Language, and Cognition: and Other Essays in the Aesthetics of
Music (Oxford/New York).
Konstan, D. 2006. The Emotions of the Ancient Greeks: Studies in Aristotle and Classical
Literature (Toronto/Buffalo/London).
Kramarz, A. 2016. The Power and Value of Music: Its Effect and Ethos in Classical Authors
and Contemporary Music Theory (New York).

Meyer, L.B. 1956. Emotion and Meaning in Music (Chicago).

Nussbaum, M.C. 1993. Poetry and the Passions: Two Stoic Views, in: Brunschwig, J./
Nussbaum, M.C. (eds.) Passions and Perceptions: Studies in Hellenistic Philosophy of
Mind (Cambridge).
Nussbaum, M.C. 2001. Upheavals of Thought: The Intelligence of Emotions (Cambridge/
New York).
Patel, A.D. 2008. Music, Language, and the Brain (Oxford/New York).
Pelosi, F. 2010. Plato on Music, Soul and Body. (Cambridge/New York).
Rocconi, E. 2003. Le parole delle Muse: la formazione del lessico tecnico musicale nella
Grecia antica (Rome).
Rocconi, E. 2012. Aristoxenus and Musical Ethos, in: Huffman, C.A. (ed.) Aristoxenus of
Tarentum: Discussion (New Brunswick), 65-90.
Wigram, T., Pedersen, I.N., Bonde, L.O. (eds.). 2002. A Comprehensive Guide to Music
Therapy: Theory, Clinical Practice, Research, and Training. (London/Philadelphia).
brill.com/grms
The Symphony of Temperance in Republic 4

Musical Imagery and Practical Models
Dr Tosca Lynch
Jesus College, University of Oxford, Turl Street, Oxford, OX1 3DW
[email protected]
Abstract
This paper calls into question a long-lasting but ill-founded tenet of Platonic scholar-
ship, namely that Plato was not interested in, or aware of, the technical implications of
the musical concepts he employed in the dialogues. Conversely, I will show how Plato
exploited the technical and practical features of the concept of symphōnía dià pasôn,
and of choral singing more generally, to highlight the unique role played by temper-
ance (sōphrosýnē) in the ideal city. More precisely I contend that Plato’s musical images,
far from being decorative or purely metaphoric devices, enrich our understanding of
this ethical notion precisely by means of their technical and performative implications,
which were very familiar to the original readers of the Republic. Hence musical theory
and practice, in addition to being central elements of the cultural context in which
Plato’s reflections must be interpreted, represent also a repertoire of concepts that sig-
nificantly informed his philosophical theories.
Keywords
Plato – musical imagery – temperance – symphōnía – choral practice – octave –

harmonics – Aristotelian problems
The reader of the Republic is presented on many occasions with thought-

provoking musical characterisations of the most important ethical values of
the ideal city: temperance (σωφροσύνη) and justice (δικαιοσύνη), as well as
the souls of the individuals who embody these ethical ideals, are repeatedly
depicted as being ‘harmonious’, resembling a symphonic blend of sounds

The Symphony of Temperance in Republic 4 19
(συμφωνία) or even a proper ‘tuning system’ (ἁρμονία).1 Socrates introduces

these curious depictions in a very matter-of-fact way, as if describing ethical
notions in these terms were only natural. However, why should one think of
ethical excellence in musical terms at all? And how do these musical char-
acterisations reflect Plato’s understanding of the social and psychological
function of these ethical skills? In order to answer these questions, I will first
examine the musical terminology employed in Republic 4 to depict temper-
ance and its wide-ranging effects; secondly, I will compare Plato’s usage with
Classical accounts of the nature of specific intervals and musical practices,
so as to reveal how Plato exploited well-known technical portrayals of these
musical concepts for his own philosophical purposes.
1 See especially Pl. R. 4.430e1-2, 4.431e7-8, 4.432a8, 4.443d6-e3. Cf. also 3.402b-c, where the hall-
mark of a true mousikós is said to be precisely his ability to recognise ethical excellence (aretḗ)
in all its forms and appearances. Whenever possible, I will avoid translating the Greek word
aretḗ as ‘virtue’, because this translation runs the risk of evoking a Christianising conception of
virtue as a stable disposition of the soul, a gift which is bestowed by God’s grace and does not
necessarily require an active response on the part of the virtuous individual (see e.g. Aquinas,
Sum. Th. i-ii.55.4). Conversely the nature of the Greek notion of aretḗ, which has its roots in
the heroic world, is essentially performative and agonistic: it is a kind of ethical excellence
that is revealed by, and indivisible from, a person’s actions and, therefore, entails a constant
tension and active effort on the part of the agent (cf. Arist. EN 2.1103a14-1105a12 and 1.1098b30-
1099a30). A similar problem arises in translating the Greek word sōphrosýnē, since most
modern renderings (e.g. self-control, moderation, self-restraint and so on) evoke an exclu-
sively censorious idea of this ethical skill as the ability to curb excessively intense or unruly
emotions (e.g. sexual desire or the overly exuberant behaviour of young children). However,
Plato’s conceptualisation of this kind of ethical excellence centres on the well balanced
attitude (closer to the original meaning of σώφρων/σώφρον, ‘of sound mind’) that results from
exercising both this censorious power and the opposite skill to establish a truly harmonious
psychological state: in order to be perfectly sṓphrōn, a person must be able to soothe exces-
sively intense feelings as well as enhance the emotional tone of overly austere spirits (see e.g.
Lg. 2.665d-666d; cf. also Plu. Virt. Mor. 443c-d and 445a, where sōphrosýnē is defined as a
mean between anaisthēsía and akolasía). For this reason, I will translate sōphrosýnē with the
rather old-fashioned word ‘temperance’, which emphasises the ‘intermediate’ character of a
soul whose psychological disposition is neither too ‘tense’ nor too ‘slack’ (cf. R. 3.410c-412a).
Furthermore, the musical overtones evoked by this word make it particularly fitting to reflect
Plato’s aim to employ music and its temperaments as a tool to modulate, and not suppress,
human emotions.
20 Lynch
1 The Choir of Temperance: A Symphony of Natural Differences
The question of what role music should play in the ideal city is introduced
for the first time in Book 2 of the Republic, immediately after a discussion of the
extraordinary psychological nature of the future Guardians of the state: they
will have to be ‘aggressive and spirited’ against external enemies and, at the
same time, ‘gentle and amenable’ toward their fellow citizens. But how to pro-
duce such a paradoxical character, ‘gentle and stout-hearted’ at once (πρᾶον καὶ
μεγαλόθυμον ἦθος, R. 2.375c6-7)? Socrates turns instinctively to the traditional
recipe that prescribed music and poetry for the care of the soul and physical
exercise for the body. By the end of Book 3, however, Socrates and Glaucon
find out that both music and gymnastics must be practiced primarily for the
good of the soul, since they affect the two leading faculties of the psychḗ—
the spirited (thymoeidḗs) and the rational (logistikón) part—by tending and
relaxing them as if they were lyre strings.2 So only a person who is capable of
harmonising these psychological components with each other can be regarded
as a perfectly accomplished mousikós, much more than someone who is sim-
ply able to tune the strings of an instrument. Still, as Socrates revealed in his
earlier examination of some technical aspects of music, this all-important psy-
chological harmony can be gained only through practical mastery of specific
modes and rhythms, which have been accurately selected on the basis of the
models of ethical excellence that must be promoted in the ideal city.3
In Book 4, these questions are approached from a completely different
perspective. Having completed the foundation of the political and cultural
institutions of the new state, Socrates and his interlocutors are now in a posi-
tion to observe the model they created in its entirety and clarify the functions
and features proper to individual ethical skills by observing the dynamics of
the social interactions that take place between the ideal citizens. Since the
2 ἐπιτεινομένω καὶ ἀνιεμένω, R. 3.412a1; cf. 4.441e8-442a2. For a similar connection between lyre
harmoníai and ethical composure, see Pl. Prt. 326a-b, where lyre-masters are said to ‘take care
of the children’s temperance’ (οἵ τ᾽ αὖ κιθαρισταί [. . .] σωφροσύνης τε ἐπιμελοῦνται) by teach-
ing them to perform ‘compositions of other good poets, the song-makers, intoning them to
the accompaniment of the lyre’ (εἰς τὰ κιθαρίσματα ἐντείνοντες). By means of these practical
performances, the teachers will force ‘rhythms and harmoníai to settle in the souls of the
children’ (καὶ τοὺς ῥυθμούς τε καὶ τὰς ἁρμονίας ἀναγκάζουσιν οἰκειοῦσθαι ταῖς ψυχαῖς τῶν παίδων).
For pre-Platonic examples of this commonplace, see e.g. Ar. Nu. 960-968, Eq. 984-996, V. 959
and 989. On the origin and technical meaning of the verbs ἐπιτείνω/ἐντείνω and ἀνίημι in
musical contexts see Rocconi 2003, 13-21.
3 Cf. R. 3.398c-400c, with Lynch 2016 on the relationship between the two harmoníai selected
for the education of the future Guardians and the technical features of lyre tunings.

beginning of this discussion, great emphasis is put on the importance of pre-

serving and reinforcing the unitary nature of the new constitution, but the
kind of whole represented by the ideal city and its inhabitants is very peculiar
and complex to define.4 Far from being a homogeneous and undifferentiated
entity, Plato’s ideal of a truly unified city (μία πόλις, R. 4.423a-d) rather resem-
bles the perfect combination of the pieces of a jigsaw puzzle: this model repre-
sents a unitary, stable and ‘harmonic’ organisation of multiple elements but, at
the same time, does not at all entail that all the pieces must be exactly identical
to each other—quite the opposite, in fact.5 In keeping with this approach, the
overall excellence of the ideal constitution is said to arise from the presence of
different traits in different people, provided that these individual talents are
incorporated into a sharing system which allows them to flourish and become
the best in their own area and, at the same time, allows the citizens to col-
laborate harmoniously by exchanging their services.6 So, for instance, wisdom
(σοφία) will not be present in each and every citizen but only in very few indi-
viduals; nonetheless, the city as a whole will be wise thanks to the contribu-
tion provided by these exceptionally gifted souls. Similarly, courage (ἀνδρεία)
is described as a kind of ethical excellence which belongs to a particular group
of citizens, including soldiers and other members of the military forces; still,
the brave actions undertaken by these select individuals will make the whole
city courageous.7
The approach changes drastically as soon as Socrates turns to temperance,
which significantly is also the first virtue to be described in musical terms.
Differently from wisdom and courage, the notion of temperance is introduced
in a slightly oblique fashion by means of a musical image, which does not
4 On these thorny issues and their implications for the relationship between happiness and
virtue, see e.g. Vlastos 1977, Kamtekar 2001, Schofield 2006, 30-5 and 212-27, and Prauscello
2014, esp. 8-9 and 21-34.
5 This model of unity that arises from the combination of markedly different elements is dis-
cussed also in many other dialogues and applies both to ethical and political matters: see
e.g. Pl. Prt. 329d-330c, 349b-d and 359a-b, where the ‘parts’ of ethical excellence are likened
to different parts of a face that play different functions and not identical parts of gold; and
Plt. 311a-c, where the ideal constitution is likened to a well-spun fabric that weaves together
people with different ethical and intellectual qualities, who are bound to each other by like-
mindedness (homónoia) and friendship (philía). Both of these notions play a central role also
in the musical definition of temperance offered in the Republic: cf. R. 4.432a7 (quoted below)
and 4.442c10-d2, with notes 8, 9 and 33.
6 R. 4.423a-424a. For the identification of justice with the principle of specialization, cf.
4.433a-434d.
7 R. 4.428e7-429c3.
22 Lynch
provide a precise definition of sōphrosýnē but vaguely likens it to ‘some sort of

symphōnía and harmonía’ (4.430e3-4). Prompted by his interlocutor, Socrates
explains that temperance may be correctly regarded as a type of harmony
because it is capable of producing an orderly structure (κόσμος) in the soul by
harmonising the different, and potentially conflicting, pleasures and desires
that stem from its individual faculties. And it is exactly the presence of this
tension between multiple needs and identities that makes sōphrosýnē so cru-
cial and philosophically interesting: differently from the self-contained and
individual nature of wisdom and courage, temperance is fundamentally rela-
tional because its very essence consists in creating a harmonious system out of
naturally contrasting elements. In other words, temperance does not attempt
to dissolve or neutralise the tensions that arise from the presence of different
individual qualities but works with them, organising them into a new system
in which each element performs its own specific function, just like different
notes in a scale.8 Socrates’ second and more elaborate definition of sōphrosýnē
clarifies even better how temperance can be properly conceived as a form of
musical harmony, moving from the inner dimension of the soul to society as
a whole:
Ὁρᾷς οὖν, ἦν δ᾽ ἐγώ, ὅτι ἐπιεικῶς ἐμαντευόμεθα ἄρτι ὡς ἁρμονίᾳ τινὶ ἡ

σωφροσύνη ὡμοίωται;
Τί δή;
Ὅτι οὐχ, ὥσπερ ἡ ἀνδρεία καὶ ἡ σοφία ἐν μέρει τινὶ ἑκατέρα ἐνοῦσα ἡ μὲν
σοφήν, ἡ δὲ ἀνδρείαν τὴν πόλιν παρείχετο, οὐχ οὕτω ποιεῖ αὕτη, ἀλλὰ δι᾽ ὅλης
ἀτεχνῶς τέταται διὰ πασῶν παρεχομένη συνᾴδοντας τούς τε ἀσθενεστάτους
ταὐτὸν καὶ τοὺς ἰσχυροτάτους καὶ τοὺς μέσους, εἰ μὲν βούλει, φρονήσει, εἰ δὲ
βούλει, ἰσχύϊ, εἰ δέ, καὶ πλήθει ἢ χρήμασιν ἢ ἄλλῳ ὁτῳοῦν τῶν τοιούτων. ὥστε
ὀρθότατ᾽ ἂν φαῖμεν ταύτην τὴν ὁμόνοιαν σωφροσύνην εἶναι, χείρονός τε καὶ
8 The specific kind of concordance which characterises temperance corresponds to the sym-
phonic relationship of similarity, but not identity, established between the higher and the
lower notes of an octave: cf. section 2 below. As I will show in detail elsewhere, it is not a
coincidence that temperance and justice are the only two virtues to be characterised in musi-
cal terms in the Republic. Justice is associated with a more complex kind of harmonía, which
builds upon the symphonic agreement produced by temperance and is said to resemble
the harmony of the three boundaries of a lyre tuning: cf. R. 4.443c9-444a2, with Lynch 2013,
97-103. On the similarity between justice and temperance, see also R. 1.351d, where dikaiosýnē
is said to produce ‘like-mindedness and friendship’ (ὁμόνοιαν καὶ φιλίαν), two crucial ele-
ments that in Book 4 are associated with temperance (R. 4.432a7 and 4.442c10-d2; cf. notes 5,
8 and 33).

ἀμείνονος κατὰ φύσιν συμφωνίαν ὁπότερον δεῖ ἄρχειν καὶ ἐν πόλει καὶ ἐν ἑνὶ
ἑκάστῳ.
Pl. R. 4.431e7-432b1
Do you see, then—I said—that we prophesied aptly before, saying that

temperance resembles some kind of harmonía?
And why?
Because unlike courage and wisdom which, each being present in a
specific part, respectively make the whole city wise and courageous, tem-
perance does not make the city temperate in the same way. Rather, it is
stretched right through the whole city and makes the weakest, the stron-
gest and those in the middle—with regard to intelligence, if you wish, or
physical strength, number, money or any such thing—sing the same song
together in octaves. Therefore, we would be absolutely right in saying
that temperance is this kind of like-mindedness, a natural symphony
of the better and the worse about which part should rule, both in the city
and in each individual.
Here the essential trait of temperance is identified in its inclusiveness: differ-

ently from wisdom and courage, temperance will be present in some degree in
all the citizens of the ideal city and does not belong exclusively to a single social
group. On the contrary, this diffused virtue is spread throughout the city and
acts as a unifying bond between the citizens, weaving so to speak the political
fabric of a truly unified community on the basis of a shared agreement on who
should exercise political power.9 Explicitly building upon his earlier portrayal
of temperance as a kind of symphony and harmony,10 Socrates now depicts
this inclusive notion in greater detail by means of a famous musical image:
9 The central role played by this kind of homónoia in Plato’s definition of temperance fur-
ther illuminates his choice to depict it as a harmonious choir: in fact, the very practice of
choral singing was a living and familiar example of the importance to accept and follow
the lead of the chorus-leader (χορηγός). Ignoring this principle brought about disastrous
consequences that Plato and his readers must have witnessed or even experienced in
person while performing in a chorus. For this reason, the chorus and its strictly musi-
cal implications represented a shared point of reference that Plato could employ to give
shape to his innovative philosophical theories, making them more immediately under-
standable to his audience. For the image of a political fabric comprising individuals with
different talents and inclinations, cf. Pl. Plt. 311a-c and note 5 above.
10 Cf. 4.430e3-4: ὥς γε ἐντεῦθεν ἰδεῖν, συμφωνίᾳ τινὶ καὶ ἁρμονίᾳ προσέοικεν [sc. ἡ σωφροσύνη]
μᾶλλον ἢ τὰ πρότερον. This musical characterisation of temperance appears again at
4.442c9-d2: σώφρονα οὐ τῇ φιλίᾳ καὶ συμφωνίᾳ τῇ αὐτῶν τούτων, ὅταν τό τε ἄρχον καὶ τὼ
24 Lynch
the well-ordered and temperate city is presented as a harmonious choir in

which all the citizens sing the same melody together, regardless of their natu-
ral or social differences.
Even on a superficial reading, this image represents very effectively the
workings of a shared network of relationships that enables the citizens to
coexist peacefully and collaborate with each other. But many other metaphors,
such as that of a beehive for example,11 could have illustrated this concept just
as well, so why did Plato choose this specific musical image instead?
We can answer this question by taking a close look at the musical expres-
sions featured in this passage and their wider implications. To begin with,
we should note that Socrates’ musical depictions of temperance are based
on the concept of symphōnía, which by definition admits the presence of
melodic lines on different pitches, and not on the notion of homophōnía,
which by contrast would have entailed that each and every member of the
choir sung exactly the same notes in unison. At first sight this distinction
may seem inconsequential, but in my view it undermines significantly one of
the most terrifying claims made about Plato’s ethical theories, namely that his
ideal citizens should relinquish their individual traits and motivations in order
to embrace an impersonal ideal of virtue that eradicates all personal features.12
ἀρχομένω τὸ λογιστικὸν ὁμοδοξῶσι δεῖν ἄρχειν καὶ μὴ στασιάζωσιν αὐτῷ; cf. also R. 4.443d-e,
Lg. 2.653a5-c3, 2.659c10-660a8 and 3.689a5-b2.
11 See e.g. Hes. Op. 305, Th. 593-5, Semon. fr. 7.90-2 West, E. Hipp. 77-80. As far as Plato is
concerned, see e.g. R. 7.520c, Plt. 301d-e and Phd. 82b, where bees are mentioned together
with wasps and ants as reincarnations of ‘temperate’ individuals. Differently from the
Republic, however, the concept of temperance outlined in the Phaedo is much simpler and
more ascetic, in keeping with the stark dualism that opposes body and soul in this dia-
logue. This monolithic conception of the soul, which does not admit degrees or complex
relationships, and its complete identification with reason undermines the possibility of
defining it as a kind of harmony, which is by nature a sýntheton prâgma (Phd. 92a-95a; for
a detailed examination of the arguments deployed in this passage, see Pelosi 2010, 181-3).
Furthermore, as Socrates points out at Phd. 93b-94b, if the soul is essentially identified
with harmony in itself (as opposed to something that can be harmonised), then different
virtues and vices cannot be regarded as types of harmony or disharmony, a vital aspect of
Plato’s conception of these notions. The tripartite psychological model discussed in the
Republic overcomes both this simplistic view of the soul and its unqualified opposition to
the body, thereby opening the way for a truly harmonic conceptualisation of the psychḗ
and its activities. See also the discussion offered in the Timaeus, where the World Soul is
not identified with harmony per se but is presented as a separate entity that ‘partakes in
reason and harmony’ (λογισμοῦ δὲ μετέχουσα καὶ ἁρμονίας ψυχή, Tim. 36e6-37a1).
12 See e.g. Annas 1981, esp. 267-70 (‘Their motivation is thus very abstract . . . they make
an impersonal response to an impersonal demand’) and especially 333-4 (‘No value is

But if this were indeed the case the concept of homophōnía, which was well
attested in Plato’s time,13 would have served his purpose much better than
the flexible and inclusive notion of symphōnía. Secondly, Plato specifies that the
kind of symphony he has in mind is a harmonía dià pasôn, an expression that is
particularly effective because it conveys two crucial points at once: on an ethi-
cal level, the expression διὰ πασῶν (literally ‘through all [the strings]’) conveys
the idea that this perfect harmony will include all the members of the city; on
a musical level, it gives some precise indications about the nature of the choral
attached to any individual life apart from the fulfilment of a social role for the common
good. Stress on the wholly impersonal viewpoint is responsible for much of what is inhu-
man about Plato’s political proposals. People are seen from a perspective which avowedly
ignores everything which makes for individual and personal commitment . . . Plato goes
on to put forward the impersonal viewpoint. His doing so is unfortunate’). Annas contrasts
this impersonal ethical ideal with a political complementarity between different social
roles but claims, erroneously in my view, that this complementarity takes no account of
individual preferences or inclinations: ‘One thing Plato does not mean is that individual
differences between people are important and that society benefits when these are devel-
oped and encouraged . . . differences of talent are seen solely as means towards the greater
good of the whole . . . Plato never argues for his assumption that even when need is not in
question people may still reasonably be expected to live lives that are determined by their
ability to contribute to the common good’, 74-6. Later on, Annas recognises that the souls
described in the Myth of Er are ‘characterized by more than impersonal love of wisdom’
(347) but frames this as a flaw in Plato’s argument rather than an element that casts some
doubts on her earlier claims. Of course, this is not to say that abstract ethical ideals do not
play any role in the psychological and ethical development of Plato’s ideal citizens, but
I believe it is crucial to remember that they represent normative ideals and not practical
models to be embodied by actual individuals: in other words, these abstract ideals should
be regarded as a ‘model in the sky’ (ἐν οὐρανῷ παράδειγμα, 9.592b2), which is meant to
guide individuals in their own personal growth and not supplant it entirely.
13 See e.g. A. Ag. 158 (τοῖς δ᾽ ὁμόφωνον αἵλινον αἵλινον εἰπέ) and Ps-Arist. Pr. 19.13, 19.16 and
19.39, which highlight precisely the similarity but not identity between homophōnía and
the special kind of symphōnía represented by the octave. Cf. also Arist. Pol. 2.1263b34-5,
where Socrates is unfairly accused to advocate in favour of a state that represents a com-
plete and undifferentiated unit, ‘just as if one turned a symphōnía into a homophōnía’
(ὥσπερ κἂν εἴ τις τὴν συμφωνίαν ποιήσειεν ὁμοφωνίαν, Pol. 2.1263b34-5); however, as we have
seen above, this is precisely what Socrates avoids doing both from a philosophical and
strictly musical point of view. This shows clearly how Aristotle’s rhetorical and, at times,
tendentious readings of Plato’s texts should not be taken at face value as reliable and
‘objective’ accounts. Cf. Annas 1981, 188 (‘Politics II, 1-6 . . . often [is] surprisingly crass
and literal-minded, much below Aristotle’s best’), Vegetti 2000 and 2002; on Aristotle’s
biased account of Socrates’ selection of different musical modes in the Republic, see
Lynch 2016.
26 Lynch
performance Plato was thinking of, since the expression διὰ πασῶν, in a musi-
cal context, means ‘in octaves’.14
2 A Symphony of Opposites: Choral Singing, the Octave and the

‘sound of correspondence’ (ὁ τῆς ἀντιφώνου φθόγγος)
Plato’s musical image, however, entails much more than the two points I have
just mentioned. As is revealed by the collection of texts known as Pseudo-
Aristotelian Problems, a technical source on music that is both chronologi-
cally and culturally close to Plato, the notion of singing in octaves entails some
crucial aesthetic and performative implications, which shed some light on the
theoretical overtones that this musical image would have evoked in the minds
of well-educated Athenians, such as Plato’s original readers. For example,
one of the Problems asks:
Διὰ τί ἡ διὰ πασῶν συμφωνία ᾄδεται μόνη; μαγαδίζουσι γὰρ ταύτην, ἄλλην δὲ
οὐδεμίαν. ἢ ὅτι μόνη ἐξ ἀντιφώνων ἐστὶ χορδῶν, ἐν δὲ ταῖς15 ἀντιφώνοις καὶ
τὴν ἑτέραν ἐὰν ᾄδῃ, τὸ αὐτὸ ποιεῖ; ἡ γὰρ μία τρόπον τινὰ τὰς ἀμφοτέρων ἔχει
φωνάς, ὥστε καὶ μιᾶς ᾀδομένης ἐν ταύτῃ τῇ συμφωνίᾳ ᾄδεται ἡ συμφωνία·
καὶ ἄμφω ᾄδοντες, ἢ τῆς μὲν ᾀδομένης τῆς δὲ αὐλουμένης, ὥσπερ μίαν ἄμφω
ᾄδουσιν. διὸ μόνη μελῳδεῖται, ὅτι μιᾶς ἔχει χορδῆς φωνὴν τὰ ἀντίφωνα.
Ps.-Arist. Pr. 19.18
Why is the octave the only concord that is used in singing? For people
magadise16 in this concord, but in no other. Is it because this is the only
concord that arises from opposite but corresponding notes (antiphṓnōn)
14 E.g. Ps.-Arist. Pr. 19.32, 19.35; Ps.-Plut. Mus. 1138e-1139c; Procl. In R. 1.212.24-213.1; Plu. Quaest.
Conv. 657b, Plat. Quaest. 1008d, De An. Proc. 1018e-1021c, 1028e-1029b.
15 ταῖς P, τοῖς C, X, Y. Perhaps the feminine is preferable in this context: cf. Pr. 19.19, 919a9-
10. However, it is worth noting that the expression ἐν τῇ/ταῖς is never employed in the
Problems to refer to strings or notes: the feminine always refers to consonant intervals
(συμφωνία or ἁρμονία: cf. 19.13, ἐν τῇ διὰ πασῶν, 19.17, ἐν ταύτῃ τῇ συμφωνίᾳ, 19.25, ἐν ταῖς
ἁρμονίαις, 19.38, ἐν τῇ συμφωνίᾳ ὁ λόγος, 19.39, ἐν τῇ διὰ πασῶν συμφωνίᾳ . . . οἱ ἐν τῇ συμφωνίᾳ
φθόγγοι . . . ἐν τῇ διὰ πασῶν μαγαδίζουσιν). By contrast, individual notes or sounds are
referred to in the masculine or neuter (e.g. 19.13, ἐν ἀμφοῖν, 19.14, τὰ γὰρ ἐν τοῖς ὀξέσιν, 19.34,
ἐν ὅλοις ὅροις, 19.39, ἐν τοῖς φθόγγοις περιόδων, 19.49, Διὰ τί τῶν τὴν συμφωνίαν ποιούντων
φθόγγων ἐν τῷ βαρυτέρῳ τὸ μαλακώτερον;).
16 In this passage, the verb magadízein can be roughly translated as ‘singing in parallel
octaves’; more generally, it indicates a musical practice which consisted in doubling,

and if one sings just one of two corresponding notes, one produces the
same result? For the one note contains in some way the voices (phonás)
of both, so that when even one note is sung in this concord, the [whole]
concord is produced; and also when they sing both, or when one note is
sung while the other is played on the aulós, they both sing as one. This
is why only this concord is used in melody, because corresponding melo-
dies at the octave17 have the voice of one note.
First of all this passage attests that, in Classical times, only the interval of an
octave was employed in sustained choral polyphony, to the exclusion of the
fifth and the fourth as well as all dissonant intervals18—a piece of informa-
tion that shows clearly how Plato’s musical image was not a figment of his
philosophical imagination but is based upon precise features of contempo-
rary choral practices.19 Secondly, the distinctive aesthetic characterisation
of the nature of this interval, which is expressed by the enigmatic adjective
antíphōnos,20 parallels exactly the unique relationship of similarity, but not
identity, that Plato associates with temperance: in fact when two ‘corresponding’
or perhaps also echoing, a given melody at the octave. Cf. Ath. Deip. 14.634c-637a (with
Barker 1984, 293-8), Ps.-Arist. Pr. 19.39 and 19.14 with Barker 1988, West 1997.
17 This is one way to translate the neuter plural tà antíphōna [sc. mélē]. Alternatively, it
could be rendered as ‘corresponding intervals at the octave’ (tà antíphōna diastḗmata).
18 In other words, the octave was the only interval at which Greek choruses performed
parallel melodic lines. This restriction applies only to choral performances of whole
melodies sung at two different pitches simultaneously and does not extend to the rela-
tionship between vocal melodies and their instrumental accompaniment, which regu-
larly featured both consonant and dissonant intervals: see e.g. Ps.-Plu. Mus. 1137b-c and
Barker 1995.
19 See Pr. 19.39 below, where tò antíphōnon is said to arise from the combined voices of
young children and men, which correspond to the highest and lowest strings of an octave,
and Aristox. fr. 99 Wehrli (τὴν μάγαδιν ὀνομάσαντα ψαλμὸν ἀντίφθογγον, διὰ τὸ διὰ δύο γενῶν
ἅμα καὶ διὰ πασῶν ἔχειν τὴν συνῳδίαν ἀνδρῶν τε καὶ παίδων). Cf. also Ph. Jud. De Agr. 80,
Vit. Cont. 88, Vit. Mos. 2.256.
20 This adjective indicates a distinctive aesthetic quality which combines opposition and
correspondence and becomes especially noticeable when listening to melodies performed
in parallel octaves: on this complex phenomenon and its treatment in the Aristotelian
Problems, see Barker 2015a. Cf. also Plato Leg. 7.812d-e, with Barker 1984, 194-5: here the
word antíphōnon seems to have a more general meaning, which may be loosely translated
as ‘notes in contrast with each other’. The exact musical nature of this contrast is harder
to identify: it could refer to some kind of antiphonal style employed in a song’s instru-
mental accompaniment or perhaps to the use of arpeggiated chords produced by striking
a row of strings with a plectrum, as opposed to the simultaneous sounds produced by
28 Lynch
notes are played at the octave, whether by two voices or by a voice and an
instrument, they sound as if they were one and the same but, at the same time,
do not lose their individual features entirely.21
The author of this Problem, however, goes further and argues that the natu-
ral correspondence between the two notes of an octave is so strong that they
automatically respond to one another even if only one of them is actually per-
formed, since each note ‘somehow (τρόπον τινά) contains the voices of both’.
Other passages of the Problems help us make sense of this seemingly absurd
observation and indicate that it refers to the phenomenon of sympathetic
vibration, spelling out exactly how either of the two notes could be said to
encompass the sounds of both. On the one hand, the lower note of an octave
was said to ‘contain’ the higher,22 a view that most likely took shape in the con-
text of lyre playing: in fact, the harmonic overtone an octave above the basic
pitch of a string can be easily produced by touching it lightly at the halfway
point while plucking or striking it with the plectrum, a gesture that reveals
the high note ‘hidden’ in the lower string. Conversely, the higher note of the
octave was regarded as being able to ‘arouse’ the lower,23 a notion that once
again must have originated in lyre practice: as we are told in Problem 19.24,
for example, if one plucks the highest string of a lyre and then damps it, only
the lowest string seems to respond by echoing its sound an octave lower
(ἀντηχεῖν), a natural kinship that reflects their being sýmphōnos.24 But this can-
not simply mean that the notes produced by these two strings are concordant
with each other: otherwise, the same phenomenon should arise also with
plucking them (cf. Pl. Ly. 209b, esp. 5-7: οὐ διακωλύουσί σε οὔτε [. . .] ἐπιτεῖναί τε καὶ ἀνεῖναι
ἣν ἂν βούλῃ τῶν χορδῶν, καὶ ψῆλαι καὶ κρούειν τῷ πλήκτρῳ).
21 Cf. Ps.-Arist. Pr. 19.38, 921a2-6: συμφωνίᾳ δὲ χαίρομεν, ὅτι κρᾶσίς ἐστι λόγον ἐχόντων ἐναντίων
πρὸς ἄλληλα . . . τὸ δὲ κεκραμένον τοῦ ἀκράτου πᾶν ἥδιον, ἄλλως τε κἂν αἰσθητὸν ὂν ἀμφοῖν τοῖν
ἄκροιν ἐξ ἴσου τὴν δύναμιν ἔχοι ἐν τῇ συμφωνίᾳ ὁ λόγος. (‘But we enjoy concord because it
is a mixture of opposites standing in proportion to each other . . . Anything that is mixed
is more pleasant than what is unmixed, especially if it can be perceived by the senses
and the proportion in the concord has the dýnamis of both extremes in equal balance’;
in other words, the dýnamis of the two notes of the octave is equal (ἐξ ἴσου), not identi-
cal, while the lógos in which they stand is double (ἐκ διπλασίου): cf. Arist. ap. Ps. Plut.
De Mus. 1139c4-6, Iambl. In Nic. 121.1-6, Porph. In Harm. 103.2-4). Cf. Pr. 19.39 below, esp.
921a17-31, 19.13, 19.19 and 19.35, with Pelosi 2009. See also Theophr. ap. Porph. In Harm.
63.28-30 Düring (Ἀλλ᾽ ἐπεί ἐστί τι σύμφωνον ἰσότητα δηλοῦν ἀμφοῖν τοῖν φθόγγοιν, ἰσότης ἐστὶ
τῶν δυνάμεων διαφέρουσα τῇ ἰδιότητι ἑκατέρᾳ), with Barker 2015b, 224-5, Raffa 2016, 432-3
and note 30 below.
22 E.g. Pr. 19.7, 19.8, 19.12 and 19.23.
23 Cf. Pr. 19.13, 19.24 and 19.42.
24 Pr. 19.24: ἢ ὅτι συμφυὴς μάλιστα γίνεται τῷ φθόγγῳ ὁ ἀπὸ ταύτης ἦχος διὰ τὸ σύμφωνος εἶναι.

strings tuned in fifths and fourths, which of course is not the case. In con-
trast, in this passage the word sýmphōnos seems to identify a special property
that belongs to octaves alone,25 an issue that is pursued in greater detail in
Problem 19.17:
Διὰ τί <διὰ> πέντε οὐκ ᾄδουσιν ἀντίφωνα; ἢ ὅτι οὐχ ἡ αὐτὴ ἡ σύμφωνος <ἐν
ταύτῃ> τῇ συμφωνίᾳ,26 ὥσπερ ἐν τῷ διὰ πασῶν; ἐκείνῃ27 γὰρ <ἡ βαρεῖα>28 ἐν
τῷ βαρεῖ ἀνάλογον, ὡς ἡ ὀξεῖα ἐν τῷ ὀξεῖ· ὥσπερ οὖν ἡ αὐτή ἐστιν ἅμα καὶ ἄλλη.
αἱ δὲ ἐν τῷ διὰ πέντε καὶ διὰ τεττάρων οὐκ ἔχουσιν οὕτως, ὥστε οὐκ ἐμφαίνεται
ὁ τῆς ἀντιφώνου φθόγγος· οὐ γάρ ἐστιν ὁ αὐτός.
Ps.-Arist. Pr. 19.17
Why do people not sing corresponding melodies at the fifth? Is it because

the sýmphōnos is not the same in this concord (symphōnía) as it is in the
interval of an octave? For in that sýmphōnos the low note in the low
range29 is analogous to the high note in the high: so it [sc. the sýmphōnos]
is, as it were, simultaneously the same and different. By contrast, the con-
cords in fourths and fifths are not like this and, consequently, the sound
of the correspondence does not arise from them: for it is not the same.
Once again, the practice of singing antíphōna in octaves is related to a distinc-

tive kind of concordant sound (sýmphōnos) which, unlike that produced by
other concords, arises from the combination of two notes that are different
but somehow ‘analogous’ to each other. Elsewhere we are told that this anal-
ogy reflects the equality (ἰσότης), and not the identity, of the melodic functions
performed by these two notes30 and it is precisely this equality that gives rise
to the distinctive aesthetic effect proper to this interval:
25 This idiosyncratic usage may be echoed in Porph. In Harm. 108.19 Düring, where the
octave alone is labelled as sýmphōnon, as opposed to the lesser symphōníai of the fifth
and the fourth. Cf. note 30 below.
26 τῇ συμφωνίᾳ ΜSS, <ἐν ταύτῃ> τῇ συμφωνίᾳ Barker 2015a, 142 n. 42; cf. Pr. 19.18: ἐν ταύτῃ τῇ
συμφωνίᾳ ᾄδεται. Alternatively, if we retain the MSS text, this sentence could be translated
as follows: ‘Is it because the sýmphōnos is not the same as the symphōnía, as in the interval
of an octave?’ (cf. note 30 below).
27 ἐκείνῃ Barker 2015a, ἐκείνη ΜSS. I take the feminine dative to refer to the sýmphōnos
(sc. τῇ συμφώνῳ), since this feminine form does not agree with the previous reference to
the interval of an octave (ἐν τῷ διὰ πασῶν).
28 <ἡ βαρεῖα> Gevaert-Vollgraff 1903.
29 Or in the low melody (mélos).
30 Pr. 19.14: τὰ γὰρ ἐν τοῖς ὀξέσιν ὄντα οὐχ ὁμόφωνα, ἀλλ᾽ ἀνάλογον ἀλλήλοις διὰ πασῶν. ἢ ὅτι
ὥσπερ ὁ αὐτὸς εἶναι δοκεῖ φθόγγος, διὰ τὸ ἀνάλογον; <τὸ γὰρ ἀνάλογον> ἰσότης ἐπὶ φθόγγων,
30 Lynch
Διὰ τί ἥδιόν ἐστι τὸ σύμφωνον τοῦ ὁμοφώνου; ἢ <ὅτι>31 καὶ τὸ μὲν ἀντίφωνον
σύμφωνόν ἐστι διὰ πασῶν; ἐκ παίδων γὰρ νέων καὶ ἀνδρῶν γίνεται τὸ ἀντίφωνον,
οἳ διεστᾶσι τοῖς τόνοις ὡς νήτη πρὸς ὑπάτην. συμφωνία δὲ πᾶσα ἡδίων ἁπλοῦ
φθόγγου (δι᾽ ἃ δέ, εἴρηται), καὶ τούτων ἡ διὰ πασῶν ἡδίστη· τὸ ὁμόφωνον δὲ
ἁπλοῦν ἔχει φθόγγον. μαγαδίζουσι δὲ ἐν τῇ διὰ πασῶν συμφωνίᾳ, ὅτι καθάπερ
ἐν τοῖς μέτροις οἱ πόδες ἔχουσι πρὸς αὑτοὺς λόγον ἴσον πρὸς ἴσον ἢ δύο πρὸς ἓν
ἢ καί τινα ἄλλον, οὕτω καὶ οἱ ἐν τῇ συμφωνίᾳ φθόγγοι λόγον ἔχουσι κινήσεως
πρὸς αὑτούς. τῶν μὲν οὖν ἄλλων συμφωνιῶν ἀτελεῖς αἱ θατέρου καταστροφαί
τὸ δὲ ἴσον τοῦ ἑνός (‘For the notes in the higher pitches are not in unison [with the lower]
but are analogous to one another at the octave. Or is it because the sound seems to be
the same, due to the analogy? For, in the case of sounds, analogy is equality and the equal
is characteristic of the one’). Porphyry’s explanation of the origin of the term antiphōna
confirms that the equality of the two opposite notes of an octave reflects the equality and
not the complete identity of their melodic function (δύναμις): ‘For this is what it is for two
notes not to differ in function from just one: it is what happens when the function arises
from two notes as if it resulted from one. This is why the notes are also called antíphōnoi,
just as someone equal to a god is called antítheos and the Amazons are called antiánei-
rai, being equal in power to men even though they are women’ (τοῦτο γάρ ἐστι τὸ δύο
ἀδιαφορεῖν ἑνὸς κατὰ δύναμιν, ὅταν ἐκ δυεῖν ἀποδέδωται δύναμις ὥσπερ ἀπὸ τοῦ ἑνός. διὸ καὶ
ἀντίφωνοι οἱ φθόγγοι λέγονται ὡς ἀντίθεος ὁ ἰσόθεος καὶ ἀντιάνειραι αἱ ἀμάζονες αἱ τῇ δυνάμει
ἀνδράσιν ἰσούμεναι καίτοι οὖσαι γυναῖκες, Porph. In Harm. 104.8-12 Düring; cf. Barker 2015a,
247, Barker 2015b 316-9 and Raffa 2016, 518-9). Later on in this treatise, Porphyry describes
the methodology followed by ‘some Pythagoreans’ in order to determine which of the
symphōníai is most concordant (symphṓnous mâllon; for the details of this mathematical
procedure, cf. Barker 1989, 34-5) and the octave is singled out as the only interval whose
‘remainder’ amounts to one, making it the only concord whose ‘unequal’ remainder is
identical to the ‘equal’ units subtracted from the two numbers that define each interval
(Porph. In Harm. 108.7-16 Düring). Significantly, in the same passage the octave alone is
identified as sýmphōnon precisely on the grounds that its ‘unequal difference’ amounts to
one, while the fifth and the fourth are said to ‘follow’ it as if they represented ‘defective’
versions of the perfect concordance of the octave: ‘They say that those whose unequals
are smaller are more concordant than the others. [The?] sýmphōnon is the concord of an
octave, since its unequals are 1; after it comes the fifth, since its unequals are 3; and last
is the fourth, since its unequals are 5’ (Porph. In Harm. 108.18-21 Düring: ἐφ᾽ ὧν δ᾽ ἄν φασι
τὰ ἀνόμοια ἐλάσσονα ᾖ, ἐκεῖνα τῶν ἄλλων εἰσὶ συμφωνότερα. σύμφωνον μέν ἐστιν ἡ διὰ πασῶν,
ὅτι ταύτης τὰ ἀνόμοια ἕν· μεθ᾽ ἣν ἡ διὰ πέντε, ὅτι ταύτης τὰ ἀνόμοια τρία· τελευταία δ᾽ ἡ διὰ
τεσσάρων, ὅτι ταύτης τὰ ἀνόμοια πέντε). Does this reflect some aspects of early Pythagorean
terminology, namely a soon-lost distinction between the perfectly equal and concordant
sýmphōnos of the octave harmonía and the lesser concordance of the other symphōníai?
Cf. Pr. 19.24 (διὰ τὸ σύμφωνος εἶναι), Porph. In Harm. 113.27-114.21 Düring and Thrasyllus ap.
Theo Sm. Math. Plat. 48.16-49.2.
31 <ὅτι> Jan 1895.

εἰσιν, εἰς ἥμισυ τελευτῶσαι· διὸ τῇ δυνάμει οὐκ ἴσαι εἰσίν. οὖσαι δὲ ἄνισοι,
διαφορὰ τῇ αἰσθήσει, καθάπερ ἐν τοῖς χοροῖς, ἐν τῷ καταλύειν μεῖζον ἄλλων
φθεγγομένοις ἐστίν. ἔτι δὲ ὑπάτῃ συμβαίνει τὴν αὐτὴν τελευτὴν τῶν ἐν τοῖς
φθόγγοις περιόδων ἔχειν. ἡ γὰρ δευτέρα τῆς νεάτης πληγὴ τοῦ ἀέρος ὑπάτη
ἐστίν. τελευτώσαις δ᾽ εἰς ταὐτόν, οὐ ταὐτὸν ποιούσαις, ἓν καὶ κοινὸν τὸ ἔργον
συμβαίνει γίνεσθαι [. . .]
Ps-Arist. Pr. 19.39
Why is it that the sýmphōnon is more pleasant than unison? Is it because

the correspondence of opposites (antíphōnon) is also the sýmphonon that
goes through all lyre strings (i.e. the octave)? For the antíphōnon arises
from the voices of young children and men, who are separated in pitch as
the highest (nḗtē) and the lowest (hypátē) string of a lyre. Concordance
(symphōnía) as a whole is more pleasant than a simple note (we have
already explained why), and the octave is the most pleasant of concords,
whereas unison has a simple sound. And people magadise in the concord
of an octave because, just as in metres the feet exhibit the ratio of equal
to equal or two to one or some other, so also the notes in a concord have
a ratio of movement to each other. But in the case of the other concords,
the endings of one note or the other are incomplete, finishing at the half-
way point; for this reason, they are not equal in function (dýnamis)32 and,
being unequal, there is a difference in how they are perceived by the
senses, as there is in choruses when, at the end of a song, some sing
longer than others. Conversely, it happens that hypátē has the same
ending of the periodic movements in its notes: for the second blow on
the air made by nḗtē is hypátē. They end at the same time but do not
accomplish the exact same thing, and so partake in an action that is one
and common [. . .].
32 On the theory of sound that informs this passage, see Barker 2015a, 234-8. In brief, the
movement made by the two notes of an octave is not exactly the same since the higher
note produces more impacts on the air than the lower; however, they partake in one and
the same activity (érgon) because the second impact of the higher note corresponds to
the first of the lower. By contrast, the second impact of the higher note of a fifth or a
fourth falls in-between the impacts of the lower note (see Figure 2 in Barker 2015a, 235).
Given that their movements do not coincide at this point in time, the two notes ‘are not
equal in their function’. On the functional equality of the two notes of an octave, see
Porph. In Harm. 104.5-12 Düring, with note 30 above and Barker 2015a, 247-8 (at 234-5,
Barker argues for a different interpretation of the word dýnamis in this Aristotelian pas-
sage; I hope to justify my reading elsewhere in due course).
32 Lynch
Far from obliterating the intrinsic features of each component, the perfect
concordance of opposites embodied by the octave arises from an orderly
arrangement of naturally contrasting movements: just like the ideal symphony
of Plato’s choir, this harmonious correspondence originates in the simulta-
neous and synchronised performance of distinct notes sung by naturally dif-
ferent individuals. These notes do not accomplish exactly the same physical
movement or melodic function but contribute as equals to a joint perfor-
mance. Like singers of a well-trained chorus, the individual notes of an octave
partake ‘in an action that is one and common’ and bring it to completion by
achieving the same end. This unmitigated blend of distinct but, at the same
time, equal notes creates a unique type of harmony: the sound of correspon-
dence (ὁ τῆς ἀντιφώνου φθόγγος), which appears to be simultaneously one and
many, alike and different, and strikes the senses as the most pleasant kind of
concordance.
3 Conclusions
The evidence preserved by the technical sources we have just examined pro-
vides us with some crucial tools to appreciate how Plato’s musical image
reflects the peculiar nature of sōphrosýnē much more precisely and meaning-
fully than is generally acknowledged. Plato’s choir of temperate citizens is not
only a metaphorical depiction of the collaborative and peaceful relationships
established among virtuous individuals with different natures. On the con-
trary, his precise reference to singing in octaves reflects some ordinary features
of choral performances and, at the same time, allows him to exploit for his own
philosophical purposes the complex interplay of unity, equality, identity and
difference that musical theorists associated with the unique aesthetic effect of
this interval.
Through his elaborate and moving musical representation of temperance,
Plato evoked a network of cultural notions he shared with his readers and
played on them to elucidate his innovative construal of the traditional notion
of temperance, as well as its social and political implications. While each
note of an octave has its own distinctive qualities, just as individual citizens
do, they still sound as one or rather ‘have the voice of one note’, as one of the
Aristotelian Problems says well (Pr. 19.18). In the same way, in Plato’s view, a
group of individuals can give rise to a truly unitary political system without
renouncing their individual differences, skills and talents—indeed emphasis-
ing their vital importance for the good of the community as a whole. By ‘singing
the same song in octaves’, Plato’s temperate citizens will become ‘simultane-

ously the same and different’, creating a symphonic agreement of natural dif-
ferences in the social system of the ideal city.33
Bibliography
Annas, J. 1981. An Introduction to Plato’s Republic (Oxford).

Barker, A. 1984. Greek Musical Writings 1—The Musician and his Art (Cambridge).
Barker, A. 1988. Che cos’era la ‘mágadis’?, in: Gentili. B. and Pretagostini, R. (eds.), La
Musica in Grecia (Rome/Bari), 96-107.
Barker, A. 1989. Greek and Roman Musical Writings 2—Harmonic and Acoustic Theory
(Cambridge).
Barker, A. 1995. Heterophonia and poikilia: accompaniments to Greek melody, in: Gentili,
B. and Perusino, F. (eds.) Mousike: Metrica Ritmica e Musica Greca in memoria di
Giovanni Comotti (Pisa/Rome), 41-60.
Barker, A. 2015a. Musical Pitch and the Enigmatic Octave in Problemata 19, in: Mayhew, R.
(ed.), The Aristotelian Problemata physica. Philosophical and Scientific Investigations
(Leiden), 226-54.
Barker, A. 2015b. Porphyry’s Commentary on Ptolemy’s Harmonics. A Greek Text and
Annotated Translation (Cambridge).
Gevaert, F. A. and Vollgraff, J. C. 1903. Les Problèmes Musicaux d’Aristote. Texte grec avec
traduction française, notes philologiques, commentaire musical et appendice (Ghent).
Jan, C. von. 1895. Musici Scriptores Graeci (Leipzig).
Kamtekar, R. 2001. Social justice and happiness in the Republic: Plato’s two principles,
HPTh 22.2, 189-220.
Lynch, T. 2013. Training the Soul in Excellence—Musical Theory and Practice in Plato’s
Dialogues, Between Ethics and Aesthetics. PhD Thesis, University of St Andrews.
Lynch, T. 2016. Why are only the Dorian and Phrygian harmoniai accepted in Plato’s
Kallipolis? Lyre vs. Aulos, in: Bravi, L., Lomiento, L., Meriani, A. and Pace, G. (eds.),
Tra lyra e aulos. Tradizioni musicali e generi poetici (Pisa-Bari), 267-84.
33 ὁμόνοιαν . . . κατὰ φύσιν συμφωνίαν, R. 4.432a6-b1, with note 9 above. Cf. Plu. De Amic.
Mult. 96e-f: probably referring to this Platonic image, Plutarch says that ‘in the harmony
produced in harps and lyres, the σύμφωνον is produced by means of opposite notes
(δι᾽ ἀντιφώνων), a likeness being somehow engendered between the higher and the lower
notes’. However, in his subsequent observations Plutarch distances himself from Plato’s
nuanced and flexible approach and adopts a much simpler model, according to which
‘in the symphony and harmony of friendship there must be no element unlike, uneven,
or unequal (οὐδὲν ἀνόμοιον οὐδ᾽ ἀνώμαλον οὐδ᾽ ἄνισον εἶναι δεῖ μέρος), but all must be alike
to engender agreement in words, counsels, opinions, and feelings, as if a single soul were
divided among many bodies’.
34 Lynch
Pelosi, F. 2009. Suoni simultanei: prassi esecutiva, ēthos e psicologia nei Problemata pseu-
doaristotelici, in: Martinelli, M. (ed.), La Musa dimenticata. Aspetti dell’esperienza
musicale greca in età ellenistica. (Pisa), 205-24.
Pelosi, F. 2010. Plato on Music, Soul and Body (Cambridge).
Prauscello, L. 2014. Performing citizenship in Plato’s Laws (Cambridge).
Raffa, M. 2016. Claudio Tolomeo—Armonica con il Commentario di Porfirio (Milan).
Rocconi, E. 2003. Le parole delle Muse. La formazione del lessico tecnico musicale nella
Grecia antica (Rome).
Schofield, M. 2006. Plato—Political Philosophy (Oxford).
Vegetti, M. 2000. La critica aristotelica alla Repubblica nel secondo libro della Politica,
il Timeo e le Leggi, in: Vegetti. M. (ed.), Platone, La Repubblica—Libro 5 (Naples),
439-52.
Vegetti, M. 2002. La critica aristotelica alla Repubblica di Platone nel secondo libro della
Politica in: Migliori, M. (ed.), Gigantomachia: convergenze e divergenze tra Platone e
Aristotele (Brescia), 179-90.
Vlastos, G. 1977. The theory of social justice in the polis in Plato’s Republic, in: North, H.
(ed.), Interpretations of Plato: A Swarthmore Symposium (Leiden), 1-40; reprinted in
Vlastos, G. 1995. Studies in Greek Philosophy (Princeton), II.69-103.
Wehrli, F. 1967. Die Schule des Aristoteles—Aristoxenos, vol. 2 (Basel and Stuttgart).
West, M.L. 1997. When is a harp a panpipe?, CQ 47.1, 48-55.

brill.com/grms
Aristotle and Musicologists on Three Functions

of Music
A Note on Pol. 8, 1341b40-1
Pierre Destrée
Université Catholique de Louvain, Belgium
[email protected]
Abstract
In Pol. 8, we find two rather different threefold divisions of the aims, or usages, of music.
At the very beginning of his analysis, Aristotle first lists (1339a11-26): amusement and
relaxation; moral education; leisure . Strikingly enough though, when it comes up again
at the end of the treatise on musical education, this threefold division has undergone
a few remarkable changes. Now, the division comes up between moral education,
emotional purgation/purification, and “thirdly”, Aristotle says, “leisure, rest and relax-
ation of one’s tensions (τρίτον δὲ πρὸς διαγωγὴν πρὸς ἄνεσίν τε καὶ πρὸς τὴν τῆς συντονίας
ἀνάπαυσιν)” (1341b36-41). The main difficulty that this new enumeration creates is nota-
ble: how to explain that now the third aim of music seems to consist in the ensemble of
leisure, repose and relaxation, while leisure and relaxation were first introduced as two
distinct aims? I argue that πρὸς διαγωγὴν should be best considered a gloss.
Keywords
Aristotle – musicologists – music – leisure
In Pol. 8, we find two rather different threefold divisions of the aims, or usages,
of music. At the very beginning of his analysis, Aristotle first lists (1339a11-26):
amusement (παιδιά) and relaxation (ἀνάπαυσις), the former being for the sake
of the latter (ἡ δὲ παιδιὰ χάριν ἀναπαύσεώς ἐστιν, 1337b38-9); virtue (ἀρετή), or
as he will say, παιδεία, ie moral education; leisure (διαγωγή). This is apparently
a firm and definitive division which he repeats a little further along: music
aims εἰς παιδείαν ἢ παιδιὰν ἢ διαγωγήν (1339b13-14), each of these aims being a
genuine one, depending on the circumstances and persons involved. Strikingly

Downloaded from Brill.com01/17/2019 05:30:01PM by [email protected]
via Zoe Rath
36 Destrée
enough though, when it comes up again at the end of the section on musical
education, this threefold division has undergone a few remarkable changes.
Now, the division comes up between moral education, emotional purgation/
purification (παιδείας ἕνεκεν καὶ καθάρσεως), and “thirdly”, we read, “leisure, rest
and relaxation of one’s tensions (τρίτον δὲ πρὸς διαγωγὴν πρὸς ἄνεσίν τε καὶ πρὸς
τὴν τῆς συντονίας ἀνάπαυσιν)” (1341b36-41). The main difficulty that this new
enumeration creates is quite notable: how to explain that now the third aim of
music seems to consist in the ensemble of leisure, repose and relaxation, while
leisure and relaxation were first introduced as two distinct aims?
Curiously enough, most recent English translations do not seem to have
noticed the problem, and translate the transmitted text literally and without
further ado (which is also the text that the three last editors, Ross, Dreizehnter
and Aubonnet, have printed)—the most recent one, by Reeve, reads “and
third, for leisured pursuit, for rest, and for the relaxation of one’s tensions”.1
But as R. Kraut acknowledges in his important Clarendon commentary, “It is
likely that the manuscripts do not here convey what Aristotle originally wrote”
(Kraut 1997: 209).2 For, not only does that second division not match the first
one, but also, and more alarmingly, it flatly contradicts it. In his first division of
the aims of music, relaxation and amusement are like sleep and wine: they are
supposed to give us some rest after hard work or heavy stress, or, as Aristotle
says citing Euripides, “put an end to our worries”; consequently, they have
nothing to do with “serious matters (οὐδὲ τῶν σπουδαίων)” (1339a16-19). Quite
to the contrary, music for leisure is closely associated with intelligence (πρὸς
φρόνησιν, 1339a25-26), being an “intellectual pastime” as some have translated
it, which (whatever that may exactly amount to) makes it a rather “serious”
and highly valuable usage of music (taking σπουδαῖος to mean both “serious”
in opposition to amusing, and valuable in opposition to worthless, or at least
less worthy). That distinction was already announced in the strongest manner
at the very beginning of Aristotle’s defense of his plan for a perfect city. There
1 Reeve 1998: 240. Other translations have a similar rendering (eg. Lord 1984, Simpson 1997).
Many previous English translations proposed the same sort of non-committal reading
(eg. Barker 1946: “a third is benefit of cultivation, with which may be linked that of recre-
ation and relaxation from strain”). This is also the case in other languages—again, I guess,
because of the authority of the three main editions in use. See eg. Pellegrin 1990: 542: “. . . et
en troisième lieu, elle [la musique] sert à mener une vie de loisir et à se reposer de ses efforts”.
2 While he points to this, Kraut himself proposes no solution and translates the transmitted
text. This is also what Susemihl & Hicks did in their 1894 edition, resolutely writing that it
is not “possible to reconcile the three advantages attendant on the use of music here with
the three ends of musical education enumerated” earlier (608, n. 40), and yet they print the
received text.
Greek andDownloaded
Romanfrom Musical Studies 5 (2017) 35-42
Brill.com01/17/2019 05:30:01PM by [email protected]
via Zoe Rath
Aristotle And Musicologists On Three Functions Of Music 37
(1337b33-38a6), he opposed activities for the sake of the amusement that suit
workers who indeed need relaxation after hard work (ὁ γὰρ πονῶν δεῖται τῆς
ἀναπαύσεως) to activities for leisure (using synonymously the verb σχολάζειν,
the phrase ἐν τῇ σχολῇ διαγωγή, and the word διαγωγή), that suit “free men”,
that is, citizens who do not need to work. “Being able to enjoy noble leisure
is the core principle of everything (σχολάζειν δύνασθαι καλῶς. αὕτη γὰρ ἀρχὴ
πάντων μία)” (1337b31-32), Aristotle says emphatically: music is one of these
“things” that constitute the truly happy life of those free citizens of his per-
fect city. In its most valuable usage, music must not be something “useful”
properly speaking (such as a means toward recovery from hard work) but an
end in itself, which can only be enjoyed in a leisured life and by “free men”
(1338a13-32). Linking music for amusement and relaxation, and music for
leisure would amount to downplaying, even denying, everything Aristotle has
hitherto said to defend and promote his views on the true, perfect happiness
the citizens of his best possible city should be supposed to enjoy. Thus, what
may appear to be a small philological or exegetical quibble actually points to
a crucial issue, indeed the very core, of Aristotle’s proposal when it comes
to describing the perfect happiness that only a well-conceived musical educa-
tion of young people makes possible.
Editors have proved too cautious in editing the transmitted, yet highly
problematic, text and many translators have followed them blindly. However,
several emendations have been proposed, especially by late 19th and early
20th century English and German editors and commentators. First of all, and
before considering any exegetical issue, the Greek of the transmitted reading
sounds philologically odd as we would normally have expected something
between πρὸς διαγωγήν and πρὸς ἄνεσίν τε καὶ πρὸς τὴν τῆς συντονίας ἀνάπαυσιν.
Some scholars have tried to save the transmitted text by taking πρὸς ἄνεσίν τε
καὶ πρὸς τὴν τῆς συντονίας ἀνάπαυσιν to be the explanation of πρὸς διαγωγήν,
as Rackam does in his Loeb translation: “and thirdly it serves for amusement,
serving to relax our tension and to give rest from it”.3 But as Newman had
already remarked in his 1902 annotated edition, if such a reading follows com-
mon Greek where διαγωγή often meant “pastime” or “amusement”,4 it would
ignore the specifically Aristotelian usage of that term throughout this book of
the Politics, and would in fact contradict what we have read in Aristotle’s first
3 This was also Bernays’ reading: “drittens zur Ergötzung, um sich zu erholen und abzus-
pannen” (1880: 7). Alternatively, Saunders (1981: 473) considers relaxation as the means to
διαγωγή: “and (iii) to promote civilized pursuits, by way of relaxation and relief after tension”.
4 Aristotle too knows this common usage: see NE 1127b33-34: Οὔσης δὲ καὶ ἀναπαύσεως ἐν τῷ
βίῳ, καὶ ἐν ταύτῃ διαγωγῆς μετὰ παιδιᾶς.

via Zoe Rath
38 Destrée
enumeration, where he firmly distinguished amusement and leisure. In the

app. crit. of his Teubner edition, Immish contrariwise suggests adding μὲν οὐ
after πρὸς διαγωγήν, amounting to reading: “and thirdly, it is not for leisure, but
for repose and relaxation”. But if this suggestion underlines the difference
between these two kinds of aims, it would be very strange to have Aristotle here
denying what he previously stated, namely that besides amusement and moral
education, there is also a third kind of aim, leisure. Newman himself, following
a suggestion made by Susemihl, suggested adding an ἤ before πρὸς ἄνεσιν.5 This
would be a very light emendation that would allow us to maintain the distinc-
tion between leisure and relaxation. As Schütrumpf (who also endorses this
suggestion) rightly notes, this would mean that these two aims must form a sort
of generic unity opposed to the unity formed by moral education and katharsis
(2005: 651). But that would in turn contradict the way Aristotle has presented
this division so far: in fact, if there is a unity between two of these aims, it
should rather be between relaxation and purgation, or purification, which are
both described as a sort of medicine (compare 1339b15-7 about relaxation: τῆς
γὰρ διὰ τῶν πόνων λύπης ἰατρεία τίς ἐστιν; and 1342a8-11 about katharsis: ἐκ τῶν δ’
ἱερῶν μελῶν ὁρῶμεν τούτους [. . .] καθισταμένους ὥσπερ ἰατρείας τυχόντας)—one
for working people who need to recover from hard work, and the other for emo-
tional people who need to restore their emotional balance. As for relaxation
and leisure, it is true that Aristotle at one point recognizes that both share the
fact of being very pleasant (1339b15-19). But being pleasant is a general feature
of music, and music for moral education must be pleasant too if one wants it
have a real effect on the moral dispositions of youth (1340a14-18). Thus plea-
sure cannot be taken as a criterion that would link amusement and leisure
in any particular way. In fact, in his first division, Aristotle does present lei-
sure as yet another, quite distinctive aim to be added to amusement and moral
education (καὶ γὰρ τοῦτο [ie leisure] τρίτον θετέον τῶν εἰρημένων, 1339a25-26)—
and indeed, it is Aristotle’s main argument, which he vigorously states (in fact
against Plato), that only leisure provides the framework for our appreciating
music for itself, and not as a means toward a further end, such as recovery from
hard work in the case of relaxation, and improvement of one’s dispositions in
the case of moral education (see especially 1338a11-13; and 1339b25-27). Finally,
following Zeller (1921: 771), one might be tempted to read a τέταρτον, “fourthly”,
5 Interestingly enough, this is a suggestion Susemihl made in the app. crit. of his 1872 Teubner
edition, which he then inserted into the Greek text in his 1879 Greek-German edition with
Engelmann. (Moerbeke’s translation: tertio autem ad deductionem ad remissionem que et ad
distensionis requiem, might tempt one to infer that he read either ἤ or καί before πρὸς ἄνεσίν;
in fact though, que et is Moerbeke’s usual way of translating τε καί).
Greek andDownloaded
via Zoe Rath
before πρὸς ἄνεσίν, which would avoid the critiques made against the previous
suggestions. But that reading, reasonable as it may at first sight appear, would
be very much at odds with both the threefold division of the aims of music
made previously, and more immediately with the threefold division between
songs (μέλη), or tunes (ἁρμονίαι), which is at stake in this passage, ie between
τὰ μὲν ἠθικά, τὰ δὲ πρακτικά and τὰ δ’ ἐνθουσιαστικά (1341b34).6
Before I come back to how I suggest we should read 1341b40-1, let me face
a related problem that this last solution brings to the forefront, namely why
Aristotle now introduces katharsis as a further aim of music, one that was not
mentioned, or alluded to, at all in his first division. Actually, here Aristotle does
not present this threefold division between tunes or songs as his own: this is
a division that has been proposed by “some people who are experts” in music
(τινες τῶν ἐν φιλοσοφίᾳ, 1341b33), that is, we may suppose, people seriously
working on music theory, whom we would nowadays call “musicologists” (such
as Aristotle’s own disciple Aristoxenus). In other words, Aristotle is not restat-
ing his own division of the aims of music, and therefore is not adding a further
aim to those he described earlier; he is reporting a division those experts have
proposed, and katharsis was one of the three aims of music they have enumer-
ated. To be sure, Aristotle does say that he endorses their division. That, we
may suppose, is because it at least partly implements the division of the aims
of music he has been proposing and defending all along. Indeed, that is the
reason why he goes on to review these kinds of songs, or tunes, one by one:
“frenetic songs” (ἐνθουσιαστικά) are suited to emotionally unbalanced persons
(and perhaps to anyone who may be affected by violent emotions from time
to time too); “reinvigorating songs” (πρακτικά) for working people; and finally,
“right-minded songs” (ἠθικά), that is Dorian songs (and not Phrygian ones as
well, as Socrates wrongly asserted in the Republic), are what we need for the
moral education of youth. Those songs implement Aristotle’s own division of
the aims of music, with reinvigorating songs for relaxation from hard work,
and right-minded songs for the moral education of youth. (As for katharsis, we
6 For the sake of completeness, I should also mention the proposal Hicks & Susemihl offer in
an excursus (1894: 638-39): they suggest reading ταύτης δ´ ἢ πρὸς διαγωγὴν ἢ πρὸς ἄνεσίν τε
καὶ πρὸς τὴν τῆς συντονίας ἀνάπαυσιν, taking ταύτης to refer to καθάρσεως at b38, which would
mean that katharsis works as the means to both leisure and relaxation. Besides its requiring
a rather heavy change in the text, this is a very implausible solution: how would katharsis
work as a means to leisure since Aristotle has so strongly insisted that learning to play the
aulos (which is of course the instrument for ecstatic music) cannot help us in developping
our intelligence (πρὸς τὴν διάνοιαν οὐθέν ἐστιν ἡ παιδεία τῆς αὐλήσεως, 1341b6-7) while music
for leisure is “for the sake of intelligence” (πρὸς φρόνησιν, 1339a25-26)?

via Zoe Rath
40 Destrée
may surmise that Aristotle considered that aim to be part of relaxation, both
being essentially restorative like medical remedies).
If we take this seriously into account, one firm conclusion should offer itself,
from which one philological proposal on the strange formulation of 1341b40-1
must follow. As we can see from the text itself, it is the case that those music
experts don’t mention anything like a special tune or song that would corre-
spond to music for leisure; as I have just said, the three kinds of songs, τὰ μὲν
ἠθικὰ τὰ δὲ πρακτικὰ τὰ δ’ ἐνθουσιαστικά, are those best suited to moral educa-
tion, relaxation and purgation/purification respectively. Thus, the most obvi-
ous conclusion that is to be drawn is that those music experts quite simply
never thought, let alone explicitly spoke, of leisure as a special, distinct aim of
music. Hence, the philological proposal that logically follows (and which as a
matter of fact was proposed long ago by James Welldon in his 1883 translation
of the Politics but, as far as I am aware, was never followed by anyone else): πρὸς
διαγωγήν should be best considered a gloss.7 Read without those two words,
τρίτον δὲ πρὸς ἄνεσίν τε καὶ πρὸς τὴν τῆς συντονίας ἀνάπαυσιν is philologically
flawless and makes perfect sense. Ἄνεσίς and ἀνάπαυσις come as the third aim
of music after παιδεία and κάθαρσις, which in the rest of the chapter Aristotle
links to the respective tunes musicologists have distinguished. Reversing the
order of his announcement, he reviews first the ἐνθουσιαστικά (1342a4-15), then
the πρακτικά (1342a15-28),8 and finally the ἠθικά (1342a28-b17).
Of course, when one proposes deleting a reading, especially when it is the
only reading that we find in all our MSS, a plausible explanation must be given
as to why someone would have invented it. The reason why some ancient
7 Welldon translates the phrase under discussion: “and thirdly for the relaxation or recreation
of the tense condition of the soul”, barely indicating in a footnote: “Omitting πρὸς διαγωγήν”
(1883: 245).
8 Reading with Sauppe (followed also by Ross, but not by Dreizehnter and Aubonnet) πρακτικά
at 1342a15 instead of MSS καθαρτικά. Following Schütrumpf (see his very helpful note ad loc),
I take 1342a16-28 to be the explanation of how those “reinvigorating songs” contribute to
relaxation for workers. Generally, translators who keep the awkward—and unparalleled in
Greek literature—MSS καθαρτικά consider those lines as constituting another paragraph, but
they then lose the continuity, and in fact the very meaning, of Aristotle’s argumentation in
this whole discussion of songs, or tunes. For a detailed defence of the MSS καθαρτικά, see
esp. Lord 1982: 132-4, who argues, quite unconvincingly to my mind, that these καθαρτικά are
meant to refer to the frenetic songs that normal people do enjoy without harm, and with-
out undergoing the curative katharsis mentioned earlier. It is much more natural to refer
that “harmless pleasure” (χαρὰν ἀβλαβῆ, 1342a16) to the reinvigorating songs (see 1339b25-27:
“harmless pleasures are suitable not only because they promote the end of life, but because
they promote relaxation too”), which Plato would have quite evidently considered harmful.
Greek andDownloaded
via Zoe Rath
reader felt himself entitled to add those words may be plausibly reconstructed
as follows. Reading the sentence without the words πρὸς διαγωγήν could have
easily left one wondering why διαγωγή, which has so far been so prominently
defended, is conspicuously absent in this analysis. Thus, such a reader might
have felt the need to write those words down, most probably in the margin of
his MS—perhaps with a question mark, indicating his perplexity as to why
διαγωγή was missing. It should then come as no surprise that a later copyist
using that MS did not hesitate to insert it into the text: as διαγωγή commonly
meant amusement, or entertainment, and since Aristotle himself presented
διαγωγή as the third section in his first division of the aims of music, adding
the words πρὸς διαγωγήν to the text may easily have been seen as a happy solu-
tion that would make the two divisions seem congruent with one another.
(As we have seen, quite a few respectable modern scholars have tried to justify
this reading, too).
The emendation of the text I propose after Welldon is not only required
by the context, where Aristotle discusses moral education, katharsis, and
relaxation, but not at all leisure. It also has an important philosophical dimen-
sion. It prevents us readers from confusing διαγωγή with entertainment or
amusement,9 and highlights the importance and originality of Aristotle’s own
understanding of what διαγωγή amounts to in his grand picture of perfect
happiness: when it comes to the value of music and its importance in human
life, one should consider music as an end in itself, not simply a means toward
something else, and therefore as part and parcel of the end of human life,
happiness.10 No one else among his predecessors, be they philosophers (such
as Plato who defended music only for moral improvement) or music experts,
ever proposed such a usage for music. Adding the words πρὸς διαγωγήν into that
sentence obscures the most original and interesting idea Aristotle defends in
his own version of musical education. Admittedly, even emended, our text still
leaves us with the pending question of what type of song, or tune, might corre-
spond to that usage of music. This is a question that one may suppose Aristotle
must have asked himself. Emending the text in the way I have defended should
9 This is a confusion that even Andrew Barker demonstrates in his otherwise excellent
translation of this passage in his Greek Musical Writings: he there translates our text,
“and thirdly at amusement for the sake of relaxation and relief from tension”, while he
had translated παιδιά (notably at 1339b15) by the same word, ‘amusement’ (Barker 1984:
180 and 174 respectively).
10 On the originality of Aristotle’s approach on music for leisure, see especially Nightingale
2004: 240-252.

via Zoe Rath
42 Destrée
force us interpreters to reconstruct what his answer might, or should, have

been; but this is not a task I shall attempt to undertake here.11
Bibliography
Aubonnet, J. 1989. Aristote: Politique. Livre VIII. Paris: Les Belles Lettres.
Barker, A. ed 1984. Greek Musical Writings 1. The Musician and his Art. Cambridge:
Cambridge University Press.
Barker, E. 1946. The Politics of Aristotle, Oxford: Clarendon.
Bernays J. 1880, Zwei Abhandlungen über die Aristotelische Theorie des Drama, Berlin:
Hertz.
Destrée, P. forthcoming. “Aristotle on Music for Leisure”, in: A. D’Angour & T. Phillips
(eds), Music, Texts, and Culture in Ancient Greece, Oxford University Press.
Dreizehnter, A. 1970. Aristoteles’ Politik. Munich: W. Fink.
Immisch, O. 1909, Aristotelis Politica. Leipzig: Teubner.
Kraut, R. 1998. Aristotle. Politics Books VII and VIII. Oxford: Clarendon.
Lord, C. 1982. Education and Culture in the Political Thought of Aristotle, Ithaca/ London:
Cornell UP.
Lord, C. 1984. Aristotle, The Politics. Chicago: University of Chicago Press.
Newman, W.L. 1887-1902. The Politics of Aristotle. 4 vols. Oxford: Oxford University
Press.
Nightingale, A. 2004. Spectacles of Truth in Classical Greek Philosophy. Cambridge:
Cambridge University Press.
Pellegrin, P. 1990. Aristote. Les Politiques. Paris: GF-Flammarion.
Reeve, C. D. C. 1998. Aristotle. Politics. Indianapolis: Hackett.
Ross, W. D. 1957. Aristotelis Politica. Oxford: Clarendon.
Saunders T. 1981, Aristotle. The Politics. London: Penguin.
Schütrumpf, E. 2005. Aristoteles Politik. Buch VII/VIII. Berlin: Akademie Verlag.
Simpson, P. 1997. The Politics of Aristotle. Chapel Hill: University of North Carolina
Press.
Susemihl, F. & Hicks, R. D. 1894. The Politics of Aristotle: A Revised Text, With Introduction,
Analysis and Commentary. London: MacMillan.
Welldon, J.E.C. 1883. The Politics of Aristotle. London: MacMillan.
11 I have offered such an attempt in Destrée forthcoming. I am very grateful to Armand
D’Angour and the GRMS anonymous referee for their remarks and insightful suggestions.
Greek andDownloaded
via Zoe Rath
brill.com/grms
Re-Thinking Lupercalia
From Corporeality to Corporation
Z. Alonso Fernández
Departmento de Filología Clásica, Universidad Autónoma de Madrid
[email protected]
Abstract
This paper is an addendum to the article “Choreography of Lupercalia: Corporeality in

Roman Public Religion”, published in the latest issue of GRMS. In my previous essay,
I explored the methodological possibilities of the notion ‘choreography’, a concept
that has been critically re-elaborated by dance scholars in the past two decades, and
applied it to the wandering of the Luperci in order to understand the performative role
of their mobility and physicality as traits shared with other dances within the realm of
Roman public religion. The aim of the current article is to refine the approaches pro-
posed for Lupercalia by examining aspects of training, performance, and reception that
are intrinsic to this choreographic practice, and to observe these elements in light of
the Roman idea of sodalitas (‘corporation’). This approach will allow us to determine
how dancing—and, more exactly, corporeality—works in the construction of Roman
identities.
Keywords
Roman dance – ludus – choreography – Lupercalia – Roman identity
Foreword
Year after year, around the Ides of February, the area of the ancient oppidum at
the Palatine was covered by masculine bodies commemorating in their move-
ment the pre-foundational origins of the Roman civilization. The nudity of the
runners, the ludic atmosphere of the races, and, of course, the hectic interaction
with passers-by reinforced the chaotic ambiance of one of the most ‘irreverent’

44 Alonso Fernández
festivals of the Roman calendar: the Lupercalia.1 The apparent improvisation

of the Luperci, however, was carefully orchestrated—or, more specifically,
choreographed—by the agency of the participants, while the annual rep-
etition of the ceremony counterbalanced any lack of control. This aspect of
recurrence—the cyclical rhythmicity of Lupercalia—has, in fact, a curious
effect that reverberates on the level of the modern scholarship, for most stud-
ies of the ritual tend to revisit, one way or another, the manifold singularities
of such a peculiar occasion.2 A 2008 article by John North and Neil McLynn,
“Postscript to the Lupercalia: from Caesar to Andromachus”,3 perhaps best
exemplifies this tendency to return, as it specifically implies the authors’ need
to keep working on the Lupercalia or, at least, the appropriateness of a precise
afterthought.
In my previous contribution to this journal, I offered a few preliminary con-
siderations on Roman religious dance—or, at least, as I warned, a branch of
Roman religious dance—and relied on the Lupercalia as a means to obtain
the necessary tools for my task. Pondering the lack of an explicit musicality
that usually drives any research on ancient dance, I turned to the wandering
of the Luperci in search for other qualities that characterize a choreographic
practice, such as agency, visuality, spatial motion, and corporeality.4 A ritual
which is not orchestic by nature and which cannot be considered ‘dance’, the
Lupercalia nevertheless fits well within the modern definition of ‘choreogra-
phy’ as it has been re-articulated by dance scholars in recent years, that is to
say, “the conscious designing of bodily movement through space and time”
(Kwan 2013, 4) or “purposeful stagings of structured, embodied movements
that aim to communicate an idea or create meaning for an actual, conceptual,
or purposefully absent audience for aesthetic and social reasons” (Morris and
Giersdorf 2016, 7).5 On the one hand, this new lens allowed me to extrapolate
the choreographic elements from the ritual and to analyse their properties
in regard to other Roman ceremonies. On the other, I was able to see how
this kind of theorising permits us to observe the process of choreography as
1 On Lupercalia cf. Var. L. 6.34, D.H. 1.32, Ov. Fast. 2.267-382, Liv. 1.5.1-4, V. Max. 2.2.9, Plu. Rom.
21.4-10, Caes. 61.2, Ant. 12.1, Quaes. Rom. 68, and Tert. Spect. 5. See also Michels 1953, Piccaluga
1962 and 1965, Brelich 1976, Bremmer 1987, Wiseman 1995a and 1995b, Scheid and Granino
Cecere 1995, Ziolkowski 1998-9, Tortorella 2000, Rüpke 2007 and 2008, Valli 2007, North 2008,
McLynn 2008, North and McLynn 2008, and Ferriès 2009.
2 Cf., for instance, the works by Piccaluga, Wiseman and Rüpke quoted in n. 1.
3 North and McLynn 2008, an addendum to North 2008 and McLynn 2008.
4 See Alonso Fernández 2016a, with bibliography.
5 Cf. above all Foster 2009 and 2010, and Lepecki 2007.

Re-thinking Lupercalia 45
intersecting with issues of gender, class, and ethnicity, thus becoming a cul-
tural means for agency in the creation of social identities.
The foundation of these ideas came, in fact, from an analysis of the Latin
vocabulary of dance and body movement; in particular, of those expressions
which are commonly used in the context of a religious practice and which are
able “to describe movements within the realm of the mind, the body, or both at
once” (Alonso 2016a, 318), such as the verb moueo, but also tripudio, ex(s)ulto,
ludo, and their derivatives. Such vocabulary revealed that most of the ritual
manifestations that are conceptualized with these terms respond to a series of
common parameters in substance and form. It was therefore necessary to find
a suitable concept comprising all the particularities implied by a highly het-
erogeneous lexicon. The conception of these rites as collective choreographies
and social mechanisms permitted, then, to re-elaborate the notion of corpo-
reality as a ritual aspect that pertains univocally to all these processes and to
detect the common traits that the Lupercalia shares with other practices that
have been traditionally characterized as dance.
Many questions have arisen from this proposal and the peculiarities of such
a complex rite require a more thorough examination. So now I return to the
Lupercalia. Following the kind suggestion of the editors of this journal,6 this
second article on the choreographies of the Luperci attempts to elucidate the
content of my previous ideas and to advocate more fully the value and neces-
sity of a different discourse on Roman religious dance. It is not the purpose
of this paper to re-investigate the religious aspects of the rite, but rather to
embrace the rhythmic recurrence of the Lupercalia and explore the cultural
embodiment of such a ludic atmosphere.
1 Roman Choreographies
In the last decades of the twentieth century, researchers in contemporary the-

ory started to incorporate the body as an area of inquiry. Heirs of the develop-
ments achieved in multiple disciplines within cultural studies,7 dance scholars
6 I thank the generosity of Professors Barker and Moore as well as the comments of the
anonymous referees of my previous article, whose recommendations have been crucial for
the elaboration of this paper.
7 That is, feminist approaches to the problematisation of gender, the literary focus on non-
written texts, theoretical debates about body politics, anthropological studies with an inter-
est in non-documented human behaviours, and other conceptions of race and ethnicity after
postcolonialism.
in particular began to approach physicality as a site of meaning-making, thus

understanding the reality of embodiment and using the body as a critical
term.8 Above all, these specialists considered bodily reality “as a tangible and
substantial category of cultural experience” (Foster et al. 1996, x) and coined
the term ‘corporealities’, insisting on the idea that bodies are not merely vehi-
cles for the expression of something else. At the same time, they argued that
“bodies develop choreographies of signs through which they discourse” (ibid.),
and deployed this notion (‘choreography’) as a thinking tool, a metaphorical
frame that generated the analysis, not only of dance practices, but of any other
action or event imbued in ‘meaning-filled physicality’.9
According to these approaches, dancing is “a cultural practice that cultivates
disciplined and creative bodies, a representational practice that explores rigor-
ously strategies for developing bodily signification, and an endeavour through
which cultural change is both registered and accomplished” (Foster et al. 1996,
xii). It is therefore an excellent lens through which to study a ritual like the
Lupercalia. In the case of ancient Rome, there is a significant number of situ-
ations that we could examine in light of choreography: choreographies are,
for instance, the walks and promenades of the elite through the streets of the
city, the schematic patterns of the army, the training exercises at the Campus
Martius, the funerals, triumphs, and processions of the Roman ludi, the gladi-
atorial performances, the chariot-races, and many other religious festivals that
may or may not incorporate the element of dance.10 In all these occasions
there is ‘inscription’ in motion,11 so the ideas of embodiment and physicality
allow us to explore, among other things, the bodies’ role in the production of a
8 Especially in North America, after the creation of a PhD program at the University of
California, Riverside. For an explanatory genealogy of dance studies and the beginning
of this theoretical focus on dance, cf. Giersdorf 2009.
9 For the descriptive and theoretical possibilities of this notion cf., above all, Foster (ed.)
1995, Franko 1995, Foster (ed.) 1996, Martin 1998, and Foster 2009 and 2010.
10 Important works that study these forms of ‘choreography’ from other perspectives are
those by Dupont 1993, Flower 1996, Bodel 1999, Sumi 2002, Corbeill 2004, Beard 2007,
Fowler 2007, Favro and Johanson 2010, O’Sullivan 2011, and Östenberg, Malmberg and
Bjørnbye (eds.) 2015, among others.
11 According to the terminology of Connerton (1989), dance is one of the ‘incorporating
practices’ through which societies transmit their memories, as opposed to the ‘inscribing
practices’ of social reproduction. Habinek (2005, 158-60), however, recalls how these two
categories relate to one another and observes the phenomenon of Roman song (i.e. song,
music, and dance) in regard to both.

narrative of Rome, in the construction of collective identities, or in the articu-

lation of an unconscious memory.
Now, if the focus of this work is to investigate more broadly Roman religious
dance, why should we look at the ritual of the Lupercalia, which is, at first sight,
the least formal and most uncoordinated movement of all the Roman choreog-
raphies listed above? How can we explore the aspects that characterize Roman
dancing if all we have is the incoherent improvisation of a disjointed corpora-
tion that runs a-rhythmically through a non-patterned space? Unsurprisingly,
the answer to these questions is linked to the theories exposed along these
lines, for the Lupercalia becomes an invaluable framework for observing
“aspects of training, technique, rehearsal, performance, and reception” that
are intrinsic to any choreographic practice (Morris and Giersdorf 2016, 7). As
we shall see, the lack of cohesion of the Lupercalia, the nudity, the wanderings,
and the improvisation are precisely those elements which will elucidate why a
‘non-danced’ practice turns out to be useful for studying dance. It is because of
the moving corporeality of this choreography that we will be able to see how
the Luperci (like other sodales, such as the Salii or the Arval Brothers) inform
political and social relations in the late republican and early imperial Rome.
2 The ludus of Lupercalia
The remains of a Campana relief from the Augustan period contain what
Stefano Tortorella (2000, 251) has described as the only example of an actual
representation of the running at the Lupercalia. A small fragmentary piece
found in the area of the House of Livia, on the Palatine Hill, this terracotta was
originally interpreted as a gymnastic scene, for it featured a group of young
boys exercising together.12 Since no other pieces related to the Lupercalia festi-
val were found around the object, it was difficult to decide, at first, the nature
of the represented scene, but the relative proximity to the Lupercal cave as well
as the characteristics of the naked runners implied strong connections with
the ritual in question:
12 Von Rohden and Winnefeld 1911, 152. Although, as Tomei (1999, 438) recalls, already Pietro
Rosa described it as ‘le corse Lupercalia’ in a letter from 1869. Cf. Veyne 1960, Tortorella
2000, 251 and Romano 2005, 91.
figure 1 Campana relief. Augustan period. Currently at Museo Nazionale

Romano (inv. 4359).
In this image, the standing naked figure of a man is surrounded by three other
masculine bodies that run scattered in various directions while holding stripes
of cloths in their hands.13 The agility of their wandering helps to create an over-
all sense of movement that sharply contrasts with the stillness of the character
situated in the foreground of the scene. The nudity of this man also allows us
to appreciate the differences between his more mature and thick body and the
13 Image retrieved from Rohden and Winnefeld 1911 (tav. XLVIII). Heidelberg University
Library. Digital Images (http://digi.ub.uniheidelberg.de/diglit/stradonitz1911/0093).

slim and vigorous traits of the others. He might represent a leader or magister,14
in which case his status is further marked by the whip that he holds in his left
hand and his peculiar hairstyle, which has been described as a wig or a mask.15
From this visual information we can discern some of the most distinctive
characteristics of the Lupercalia that I highlighted in my previous essay.16 First
of all, the image represents the only portrayal of a group of Luperci who are
utterly naked, a feature that is normally avoided in visual arts17 but which con-
stitutes a common trait of the literary descriptions of the festival, particularly
in republican and early imperial sources.18 The nudity of the Luperci relates
originally to the wildness of a primeval state of the ritual, which goes back
to the story of Romulus and Remus and the pastoral worship of Faunus, so it
has been connected to the leuitas of a carnival-rite.19 Considered as a sort of
disguise, nakedness situates the runners in a remote reality, distant from the
urban environment and characterised by the attributes of an animal land. In
the words of Cicero, the bodies of the Luperci conform ‘a quite savage brother-
hood, downright rustic and uncouth . . . founded long before civilisation and
law’ ( fera quaedam sodalitas et plane pastoricia atque agrestis . . . ante est insti-
tuta quam humanitas atque leges, Cael. 26), that is to say, at the dawn of the
Roman history.
This natural landscape represents, at the same time, a channel through
which the runners embrace an ‘original’ form of manliness or, to follow Ramsay
Burt (2007, 19-22), an image of masculinity “conceived of as instinctive and
14 For magistri Lupercorum, cf. Rüpke (2008), who mentions three of them: Mark Antony
(Dio Cass. 45.30.2, Nic. Dam. Vit. Aug. 21), A. Castricius Myriotalenti (CIL 14.2105 =
ILS 2676), and Clesipus Geganius (ILLRP 696 = CIL 12.1004 = CIL 10.6488 = ILS 1924, Plin.
Nat. 34.11).
15 To support the idea of a mask, Tortorella (2000, 251) quotes Tertullian (Spect. 5) and
Lactantius (Inst. 1.21.45): the former compared the Luperci to ludii and the latter claimed
that they used to run nudi, uncti, coronati, aut personati aut luto obliti. This interpreta-
tion, however, can be problematic because, as we know from McLynn (2008, 170), in Late
Antiquity theatrical professionals had replaced the original runners. Since we cannot see
the heads of the runners in our image, it is impossible to guess whether they had a similar
coiffure. In any case, none of the extant images of Luperci represent such an odd feature.
16 Cf. Alonso Fernández 2016a, 319-25.
17 Normally, the Luperci are presented as wrapped in a kilt or long cloth that covers their
lower body and legs. Cf. Veyne 1960, Tortorella 2000, and Romano 2005, 91.
18 According to Varro (L. 6.34), Dionysus of Halicarnassus (1.80), Ovid (Fast. 2.284 ff.), Livy
(1.5.1-4), Valerius Maximus (2.2.9), and Plutarch (Rom. 21), the Luperci run naked or wear-
ing small goatish loincloths. Cf. Wiseman 1995b, 11-12 and North and McLynn 2008, 178.
19 Cf. Alonso Fernández 2016a, 322, with bibliography.
innate”. Half-animals, the young Luperci impose their manly endurance to the
viewers, a ‘hyper-masculine’ display that naturalises fierceness and resolution
in order to show that they are neither females nor effeminate men. Moreover,
the heterogeneity of their similarly naked bodies creates the sense of a united
sodalitas, a virile bond which lies at the basis of the Roman notion of manhood
and separates them from other groups in the society. These sodales, however,
are not yet Roman uiri, but adolescents aiming at conforming to the body of
the elite,20 so their naked wanderings enact the liminal period in which they
are still deprived of their toga uirilis.21 In fact, they commemorate the ath-
letic training of the mythical twin brothers as a form of exercise—what Ovid
(Fast. 2.365-69) describes as lusus—that prepares them for their future lives.22
As Thomas Habinek (2005, 114) recalls, “it is adolescence or young adulthood,
the period between physical and social maturity, that the Romans most com-
monly associate with play”. Thus, in the case of Lupercalia, the soon-to-be virile
bodies incarnate the properties of a ‘ludic’ atmosphere in the form of a par-
ticular training, their play providing a transitional experience that constitutes
mockery, transgression, preparation, and rehearsal.
Habinek (2005, 116) argues that “the reality generated by ludus is a real-
ity experienced and carried in bodies”. In the terracotta relief, the ludic cor-
poreality of the runners is represented by the explicit nudity of their young
figures, but it is articulated, above all, through the indistinct trajectory that
their motion describes. This irregular pattern defined by the written sources
as discurrere (‘to run to and fro’, ‘to run about’),23 is what momentarily breaks
the impression of cohesion created by the multiplicity of naked individuals, so
the original homogeneity of participants gives way to a more independent per-
formance, which seems to be almost improvised. At first sight, the Luperci’s
freedom of choice suggests that each of the runners follows his own instinctive
direction without considering the course of the others. Gradually, a collision of
20 The Luperci are normally described as iuuentus (Ov. Fast. 2.365, Liv. 1.5.2, V. Max. 2.2.9, Tert.
Spect. 5). However, as North and McLynn (2008, 177) note, after the legislation imposed
by Augustus (Suet. Aug. 31.4), the run was not supposed to include imberbes (‘beardless
young men’). It is, therefore, not a coincidence that most the imperial records provide
information about Luperci who are in their early twenties (Rüpke 2008, 8-9).
21 Although the evidence suggests that many of the runners were a little older, it is interest-
ing to reconsider the idea of Brelich (1976, 121 n. 35), according to which the Lupercalia
takes place only one month before the Liberalia of March 17th.
22 Plutarch (Caes. 61.2) claims that the Luperci run naked ‘for sport and laughter’ (ἐπὶ παιδιᾷ
καὶ γέλωτι). See also Plu. Ant. 12.1.
23 On the unspecific direction of their wandering and the use of this and other verbs, cf.
Alonso Fernández 2016a, 323 and, in particular, n. 55.

separate actions creates a wave of expansive movement that distorts the per-
ception of a disoriented audience. In agreement with the literary sources, the
image of the relief implies that the wandering of the Luperci aims at reaching a
vast amount of territory, no matter the specificity of the path, yet always enact-
ing a simultaneous performance that fills up the streets of the urbs.24 This way,
their running about becomes a hectic choreography that is at once collec-
tive and individual, an asymmetrical manoeuvre through which the runners
achieve a sense of cohesion as well as their status of privileged actors separate
from the crowds.
To a certain extent, we could borrow Barbara Kowalzig’s notion of ‘arrhyth-
mia’ as a necessary constituent of a collective rhythmicity in the construction
of what she calls ‘the bodily social’, this time applied to a specific group within
the society, such as the young members of the male elite.25 For Kowalzig (2013,
188), “rhythm is itself a form of variety. It is the succession of the quasi-identical.
It is variety within a regularity”, so that “in having a necessary element of
arrhythmia, rhythm is precisely not about sameness but about variety trough
time” (Kowalzig 2013, 190). In the wandering of the Lupercalia, ‘arrhythmia’
characterizes an extremely uneven performance, particularly as time relates
to each of the runners’ spatial mobility—that is to say, their dis-currere—.
Nonetheless, the moving figures of the Campana relief invite us to see a sort of
rhythmicity within their irregular running. These dynamics indirectly reverse
and complement Kowalzig’s ideas, according to which the different bodies
end up stepping back into the anonymity of the communal movement and
embodying collective time.26 It is precisely this ‘rhythmic arrhythmia’ that rep-
resents the materialization of Lupercalia’s ludic occasion while constituting
the driving force of the sodalitas in the construction of social identities.
But the discurrere of the Luperci is not exclusively a form of embodied
‘arrhythmia’. Turning now to another kind of performance of the iuuentus, we
may see the similarities between the wanderings of Lupercalia and the cre-
ative process that gave birth to the ludi scaenici. According to Livy (7.2.1-7) and
24 On the notion urbem celebrare (‘to crowd the city’) and the bodily evocation of the city’s
pre-foundational past, cf. Alonso Fernández 2016a, 325.
25 Cf. Kowalzig 2013, 187-190.
26 On group cohesiveness and individual innovation as mechanisms for the Luperci’s
social relation, cf. Alonso Fernández 2016a, 323-24. Perhaps in these cases, the figure of a
magister like the one in the Campana relief helps solidify the bonds between the various
individuals by re-orchestrating their collective move.
Valerius Maximus (2.4.4),27 an outbreak of pestilence in 364 BCE prompted the

Roman authorities to import a group of dancers from Etruria (the so-called
ludiones) to appease the angry gods, while the young Romans performed their
own improvisation of movements. Whereas Livy focuses on the actions of the
iuuentus as an imitation of the Etruscan performance (imitari deinde eos iuuen-
tus, Liv. 7.2.5) or rather, a parody of their dances in gestures and words (. . . simul
inconditis inter se iocularia fundentes uersibus, coepere; nec absoni a uoce motus
errant, ‘at the same time, they hurled insults at each other with rude verses;
and their movements were not incongruous with their voice’, 7.2.5-6), Valerius’
version is the reverse: first the jesting youth (iuuentus . . . iocabunda) added
gesticulations to the compositions addressed to the gods—‘with a rude and
disarranged movement of the body’ (rudi atque inconposito motu corporum,
2.4.4)—, and because of the success of their improvisations, the authorities
called real artists from Etruria to perform a more refined form of dance. No
matter the succession of events, these stories aid our understanding of the role
of the Luperci’s ‘ludic corporeality’ as a result of the significative function of
the iuuentus’s disarticulated mobility, particularly in connection to the aspects
of coarseness, mockery, and improvisation that the authors highlight.
Andrew Feldherr (1998, 181) points out that, in Livy’s account of the ludi scae-
nici, “what had been a ludus, a word that recalls both the public festivals that
provided the occasion for the drama and the jesting of the indigenous Roman
youths, became an ars, a craft of profession”. Similarly, in the Lupercalia, the
ludus of the celebration, the mockery that commemorates mythical training,
becomes rehearsal for real life. In the words of Thomas Habinek, “play does
not exist except in dynamic interrelationship with reality” (2005, 111), so the
irreverent mobility of these groups of young men turns out to be an essen-
tial medium by which they acquire their future status.28 In both these events
commemorating Rome’s history and traditions, a leading group of male aristo-
crats carries out a bodily performance that defies the standards of a normative
behaviour. Their physical involvement in the process—and, more than that,
the subversive choreographies that their bodies configure—reverse certain
forms of communal or religious communication that are regarded as accept-
able, to the extent that they—the iuuenes—become special agents within
27 Although Varro is thought to be the common source for these two authors, Oakley (1998,
77) proposes that Valerius’ account derives at least in part from a source other than Livy.
On these questions, cf. Moore 2012, 2, with bibliography.
28 Following Morel (1969), Habinek (2005, 120-21) interprets the iuuentus as an organized
company of free adolescent males, a cadre of young men of military age, like the Luperci.
Cf. above n. 20.

the social body: they are allowed to perform in this way because of who they
are, and they are allowed to be who they are because of how they perform.
Furthermore, the destabilizing actions of these youths discourage other groups
from taking their place in challenging the norms, so with their collective per-
formance they prevent any attempt to social change.
By means of their chaotic corporeality, the Luperci participate in a choreo-
graphic process that contributes to the structuring of their power and generates
specific forms of social knowledge. This kind of knowledge is based, above all,
on the kinaesthetic properties of an audience that relies on their own body’s
ability to feel motion, recognising the meanings and values of the performance
through perceptual and cognitive mechanisms.29 Scholars in dance studies and
other related fields have recently coined the observers’ capacity to perceive
the performers’ physicality and to identify somatically with that motion as
‘kinaesthetic empathy’, a form of motor reaction which draws the audience
“into a powerful sense of participation and identification” (Olsen 2017).30 A
process which is closely related to the phenomenon of ‘emotional contagion’,31
kinaesthetic empathy focuses on the embodied impact of what the observers
witness as well as on their channels for interpersonal understanding, but it
is essentially conditioned by a set of social and cultural assumptions. Thus,
in the case of Lupercalia, the choreography of runners is inevitably informing
and encouraging the empathetic responses of bystanders, to the extent that
they acknowledge their performance as a shared expression of common
ideals, even when there is a clear-cut distinction between them and the exclu-
sive group of Luperci.
Such a response to the performance is not just a product of the transitory
identification of the audience with the ludic motion they perceive in the exact
moment of the running. Rather, the observers recognise such orchestration
of Roman identities by means of a kinetic familiarity with the recurring per-
formance that takes place year after year, in the month of February.32 Thus,
the reiterated discurrere of the Luperci constitutes one of the activities that
29 As highlighted by Noland (2009, 36), Marcel Mauss (1934) was the first to claim that the
motor body makes meaning, thus challenging his contemporaries’ emphasis on the cul-
turally constructed body.
30 Cf. Sklar 2008, Noland 2009, Foster 2010, and Reynolds 2013. Advances in neuroscience
and cultural anthropology are crucial for their studies.
31 Reynolds (2013, 213-14) claims that the difference between ‘empathy’ and ‘contagion’ is
one of degree, rather than kind.
32 As Valerius recalls, ‘the memory of that cheerfulness is renewed with the annual return of
the festival’ (cuius hilaritatis memoria annuo circuitu feriarum repetitur, 2.2.9).
Paul Connerton (1989) defines as ‘incorporating practices’ in the transmission

of the social memory of the citizen body and, in particular, of the memory of
the elite. Because of the need to create and maintain their special image, the
Roman aristocracy becomes a choreographer of the festival, demonstrating its
agency in the annual continuance of the rite.33 The runners, at the same time,
transmit their bodily practice from one generation to the next, embracing and
performing traits that will characterize, repeatedly and indirectly, the ideal
notion of the Roman uir.34
We cannot forget, in any case, that the Luperci articulate their agency in
their explicit interaction with passers-by. As the ancient sources recall, they
strike them with their lashes in order to purify the crowds and to stimulate
fertility among young married women.35 Visual representations of the ritual
insist on this aspect as one of the most distinctive dynamics of Lupercalia.
Thus, in the funeral stele of Claudius Liberalis (2nd c. CE),36 the young runner
seems almost to be coming out from the gravestone as he faces the front of the
scene with the whip in his right hand and his feet ready to run.37 More striking
are the mosaic floor from Thysdrus, in Tunisia (ca. 3rd c. CE), and the panel of
Aelia Afanacia’s sarcophagus (late 3rd c. CE), both of which show moments
of deliberate flagellation of women during the course of the rite.38 Now, whereas
the woman in the mosaic “looks back over her left shoulder at where her dress
is raised to bare her body for the blow” (Wiseman 1995a, 16), the female of
the sarcophagus is exposing herself to the whipping in a conscious and open
way: she throws herself into the multitude of Luperci and stares directly at the
33 Interestingly, as McLynn (2008, 170) notes, the Roman elite was implicated in the mainte-
nance of the Lupercalia even when the runners were not members of the aristocracy, but
professional actors hired specifically to perform the rite.
34 On the experiences and impressions that come through such physical practice as a learn-
ing process (ludus), cf. Collins (2014), who examines, in the case of the Attic chorus, “the
ways in which words and movements are apprehended by the senses, minds, and bodies
of a chorus-in-training”.
35 V. Max. (2.9.9): obuios . . . petiuerunt; OGR 22.1: occursantes quosque; Plu. (Rom. 21.5):
τὸν ἐμποδὼν. Cf. Wiseman 1995b, 84.
36 On this monument cf. Veyne 1960, 104-105, Tortorella 2000, 249, Wiseman 1995a, 16, and
North and McLynn 2008, 178.
37 The agency of Liberalis can be inferred as well from the inscription in the base (CIL 6.3512),
which reads sodalis desiderantissimus, perhaps indicating that he had been nominated
as a member of the Luperci but had not yet made his first run. Cf. North and McLynn
2008, 178.
38 Cf. Tortorella 2000, 252-53. For a more complex and contextualized interpretation of these
scenes and their dramatic qualities, cf. North and McLynn 2008, 179-80.

person who is holding her arms. North and McLynn note the active partici-
pation of these two women (2008, 179), suggesting that “the Lupercus shares
the spotlight with the victim” and stressing the increasing amount of drama
in what seems to be already more a form of public spectacle than a religious
rite. Despite the transformations of the ritual in the late Roman empire,39 how-
ever, I argue that through these third-century representations of the Lupercalia
it is still possible to perceive a form of kinetic involvement, according to
which at least the female members of the audience enact their own bodily
responses to the Luperci’s performance. In these and other instances that the
written sources recollect,40 the subtle forms of kinaesthetic identification
with the ludus have given way to the full involvement of the audience’s bodies
in the game.
3 Corporeality and Corporations
The choreographic study proposed along these lines has served to relativize
many of the assumptions that we had about the Lupercalia. By looking at aspects
of training, rehearsal, technique, and reception, we have been able to encap-
sulate the ambivalent nature of the Luperci’s ludic corporeality, the orchestra-
tion of an improvised performance, and the impact of a ‘disjointed’ mobility
in the dynamics of collective/individual, self/other, participant/observer. In
the introductory section of this essay, I stressed the necessity of examining the
physicality of a performance in order to understand the role of dance and body
movement in ancient Roman religion, proposing the Lupercalia as a case study.
My argument at that point was that even the least articulated of all the Roman
choreographies would allow us to observe how the moving corporeality of the
participants informs political and social relations in the late republican and
early imperial Rome. But if that is the case, what bridges the anarchic corpore-
ality of the Luperci with the notions that characterize the Roman uir?
In his study on the iconography of the Lupercalia, Stefano Tortorella (2000:
248-50) includes a series of representations that sharply differ from what the
written sources recount about the ritual and its participants. Coming from dif-
ferent periods, styles, and media, they include a statue of a Lupercus from the
first century CE, the monument of Claudius Liberalis, and a funerary relief of
39 Cf. McLynn 2008.

40 Plu. (Caes. 61.2), for example, claims that they present (παρέχουσιν) their hands to be
struck, like children at school. Juvenal (2.142), for his part, also suggests that they ‘put out’
(praebere) their own hands to be struck by the Lupercus.
an equus publicus, both from the second century CE.41 The most surprising fea-
ture of all these representations is that they do not portray naked Luperci, but
young boys wearing a long cloth under their waists and holding real whips. In
view of this evidence, T.P. Wiseman (1995a, 16) wonders “what has happened
to the hilaritas and lasciuia of the republican ritual” and suggests that during
the early imperial period the ritual could have been deliberately neutralized,
especially after the legislation imposed by Augustus.42 To North and McLynn
(2008, 178-80), this view contradicts the stories transmitted by the majority of
early imperial authors, who insist particularly on the nakedness of the runners.
They therefore propose that each of these representations shows a “Lupercus,
not poised and ready for the Lupercalia run, but rather wearing a dress uni-
form, marking his membership with reference to his ‘nakedness’ and to the
ritual gear, but not attempting to represent the reality that was, on this view,
only to be seen on 15 February itself”.
In addition to the element of clothing, the Luperci of these monuments
display a quite particular corporeal attitude, with solid and delineated tor-
sos, their sight facing straight at one direction, and a strong air of heaviness
which, even in the case of Claudius Liberalis,43 transmit a sense of what Don
Fowler (2007, 5) described as ‘immobile constantia’. Compared to the dynamic
group of naked runners from the Campana relief, the bodies of these single,
static Luperci radiate ideas of grauitas, auctoritas, and self-control, all markers
of a form of superiority that was central to the Roman conceptions of bodily
instance.44 Drawing on Bourdieu’s terminology of habitus (1977, 72)—or,
rather, of hexis as its corporeal dimension—, we could say that these young
males have embodied a set of practices, linked to a system of social mean-
ings and values, so they show a permanent physical disposition that com-
plies with the normative behaviour of the elite’s representatives, that is to
say, the male aristocracy. In this respect, it is necessary to clarify that man-
liness, in ancient Rome, is not exclusively a category of gender, as Connolly
(2007, 92) claims, but a complex notion that encompasses aspects of class,
status, and citizenship. Even if the virility of these draped Luperci no longer
constitutes an obvious display imposed on the viewers (i.e. they are not naked
41 For the statue of Lupercus (Fondi, Museo Civico), cf. Faccenna 1954 and Wrede 1983,
189. For the monument of Claudius Liberalis (Tibur, Musei Vaticani. Galleria Lapidaria)
and the funerary relief (Benevento, Museo del Sannio), cf. Veyne 1960, 104-105. Cf. also
Romano 2005, 91.
42 Cf. above, n. 20.
43 For the active pose of this Lupercus, cf. above p. 54.
44 For these bodily characteristics in relation to walking and standing, cf. Corbeill 2004,
118-22, Fowler 2007, 4-7, and O’Sullivan 2011, 17-20.

any more), the qualities that we deduce from their bodies are still those which
characterize the Roman uir.45 Above all, the masculinity of these bodies estab-
lishes clear-cut distinctions with regard to other groups in the society, like
foreigners, women, and slaves. The Romans’ ‘performance’ of uirilitas—to put
it in Judith Butler’s terms (1993)—serves to accentuate such social distinc-
tions in order to make them appear as ‘essential’. Moreover, as Connolly (2007,
95-96) emphasizes, the male aristocracy also retains gender as an important
category in critical discourse, particularly because it “helps recast the competi-
tive nature of intra-elite relations as a contest for virtue”. And it is here where
we get to the idea of ‘corporation’.
In my previous essay, I insisted on the similarities between the Salii and
the Luperci and tried to demonstrate how these two religious sodalitates func-
tioned together in ritual and cultural spheres.46 Examining qualities of specta-
cle and performance highlighted by other scholars concerning these festivals47
and, in particular, the ideological implications of these ceremonies with
regard to the language of empowerment of the Roman uir, I then relied on
the thesis of Thomas Habinek (2005, 36-38), according to which “ideal manli-
ness is located in communal performance”. Clearly, the dance of the Salii—but
also the Arval choreographies or the equestrian parade of the lusus Troiae—
constitutes a training mechanism (ludus) carried out by the Roman aristocracy
in order to incorporate, generation after generation, a set of features and values
(Bourdieu’s habitus) that help to construct the identity of the uir Romanus:48
the ternary rhythm of their characteristic tripudium, for instance, the strength
of their jumps against the ground, and the methodical alternation of agile
movements in call and response are all examples of mensura, constantia,
and grauitas, that is, the markers of a ‘virile’ identity.49 Similarly, the kind of
movements that they carry out fosters a form of social relation that combines
group cohesiveness with individual performance and promotes the aristo-
cratic balance between parity and competition.
45 For the study of masculinity in ancient Rome, cf. also Gunderson 2000.
46 Cf. Alonso Fernández 2016a, 326-27. On the Salian sodalitas, cf. Wissowa 1902, Cirilli 1913,
Lambrechts 1946, Bloch 1958, Dum.zil 1966, Rüpke 1990, Bremmer 1993, Torelli 1990 and
1997, Habinek 2005, 8-33, Glinister 2011, Sarullo 2014, and Granino Cecere 2014.
47 Cf. Piccaluga 1965, 147-57.
48 Cf. the words of Quintilian (Inst. 1.11.18-19) about this process of embodied apprentice-
ship, a thought that somehow anticipates the modern theories of Mauss (1934), Bourdieu
(1977), and Connerton (1989). Cf. Corbeill 2004, 2-3.
49 For a more thorough analysis of these elements with regard to the Salian dance, cf. Alonso
Fernández 2016b, with bibliography. I leave the question of the Arval Brethen and the
lusus Troiae for future research on these matters.
From the formal point of view, the figures of the draped Luperci, embody
notions that coincide with all these markers of the uirilitas, thus contrasting
with the leuitas of the rite. Isolated from the other sodales, these static run-
ners are not explicitly involved in a ludic discurrere, so their portrayals cor-
respond to a more permanent disposition of a uir Romanus, and not to the
exact moment of their running about. Yet, in order for them to consolidate
this permanent disposition I argue that the irreverent element of Lupercalia
is equally as necessary as the rest of the above-mentioned practices that help
construct the body of the male elite. As Connolly (2007, 90) claims, “authentic-
ity and naturalness are made synonymous with the manly performance”; these
go back to the times of Romulus and Remus, and have played a fundamental
part since the transition from the Romans’ original agrarian life. In the case of
Lupercalia, it is therefore essential that the sodales cultivate their particular
channels for agency, their spontaneity, and their naturalness. This is why to
Virgil (A. 8.665) the Luperci need to appear as ideologically linked to the Salii
in the shield of Aeneas, for their ‘irreverent’ bodies are nothing but the other
side of the coin, articulating an even more accentuated form of ludus that
fulfils the same ritual and cultural aims.
Moreover, in both these occasions it is perhaps the fact that the participants
enact collective exercises that makes the individual performance of manhood
merge with the ‘bodily social’ or, in other words, that turns the performance of
gender into a performance of the civic ideals of Rome. We shall recall, at this
point, that Appius Claudius used to dance inter collegas (Macr. Sat. 3.14.14), and
that Seneca (dial. 9.17.5) hailed Scipio Africanus and “the men of old” (antiqui
illi uiri) because they “used to dance in a manly style” (uirilem in modum tripu-
diare). These references to communal performances remind us of the Greek
chorus, as I noticed in my previous essay,50 yet in the case of the Roman col-
lective choreographies, they seem to be almost exclusively a practice for men
of the elite, where notions of ‘Romanness’ and foundation cross-pollinate with
the element of manhood:51 no matter the characteristic form of the manly
50 Cf. Alonso Fernández 2016a, 313.

51 Apart from the choral performance of children of both sexes celebrated for the ludi saecu-
lares in 17 BCE (CIL 6.32323 = ILS 5050), there is only one historical mention of a female
chorus in Rome, which performed in 207 BCE (Liv. 27.37.14). As for other references to
female choruses in Roman literature, such as those in honour to Diana at Nemi (Prop.
2.28.60; Prop. 2.32.9-10; Hor. Carm. 2.12.18-20), they could be the product of the poets’
imagination. For agrarian choral dances, cf. Verg. G. 1.344-350 and, for the choral perfor-
mance of the Roman tibicines, cf. Var. L. 6.3. For an overview of these questions, cf. Curtis
(forthcoming).

performance, we can trace the creation of communal identities out of a sum

of individual bodies and see how these communal identities construct and
are construed by ideas of their essential manliness. It is in all these occasions
when ‘corporealities’ become ‘corporations’ and when the singular bodies of
men become the referent for the Roman patriarchal social body.
Acknowledgements
This work has been possible thanks to a research grant from the RCC at Harvard
University. I want to express my gratitude to the Department of the Classics at
Harvard for their hospitality and, in particular, to Naomi Weiss for her read-
ings. I also thank Professors Barker and Moore, as well as the anonymous refer-
ees, for their suggestions and feedback.
Bibliography
Alonso Fernández, Z. 2016a. Choreography of Lupercalia: Corporeality in Roman Public

Religion, GRMS 4.1, 311-332.
Alonso Fernández, Z. 2016b. Redantruare: Cuerpo y cinestesia en la ceremonia saliar,
Ilu 21, 9-30.
Beard, M. 2007. The Roman Triumph (Cambridge/London).
Bloch, R. 1958. Sur les danses armées des Saliens, Annales (ESC) 13, 706-15.
Bodel, J. 1999. Death on Display: Looking at Roman Funerals, in: Bergmann, B.,
Kondoleon, C. (eds.) The Art of Ancient Spectacle (New Heaven/London), 259-81.
Bourdieu, P. 1977. Outline of a Theory of Practice. Trad. Richard Nice (Cambridge/
New York).
Bremmer, J.N. 1987. Romulus, Remus and the Foundation of Rome, in: Bremmer, J.N.,
Horsfall, N.M. (eds.) Roman Myth and Mythography (London), 25-48.
Bremmer, J.N. 1993. Three Roman Aetiological Myths, in: Graf, F. (ed.) Mythos in mythen-
loser Gesellschaft: das Paradigma Roms (Stuttgart), 158-74.
Brelich, A. 1976. Tre variazioni romane sul tema delle origini (Roma).
Burt, R. 2007. The Male Dancer. Bodies, Spectacle, Sexualities. Second Edition (London/
New York).
Butler, J. 1993. Bodies that Matter: on the Discursive Limits of “Sex” (New York/London).
Cirilli, R. 1913. Les Prêtres-Danseurs de Rome : Étude sur la Corporation Sacerdotale des
Saliens (Paris).
Collins, J.H. 2014. Dancing the Virtues, Becoming Virtuous: Procedural Memory and
Ethical Presence, Ramus 42.1-2, 183-206.
Connerton, P. 1989. How Societies Remember (Cambridge).

Connolly, J. 2007. Virile Tongues: Rhetoric and Masculinity, in: Dominik, W., Hall, J. (eds.)
A Companion to Roman Rhetoric (Malden), 83-97.
Corbeill, A. 2004. Nature Embodied. Gesture in Ancient Rome (Princeton/Oxford).
Curtis, L. Imagining the Chorus in Augustan Poetry (Oxford), forthcoming.
Dumézil, G. 1966. La religion romaine archaïque, avec un appendice sur la religion des
Étrusques (Paris).
Dupont, F. 1993. Ludions, lydioi: les danseurs de la pompa circensis. Exégèse et discours
sur l’origine des jeux à Rome, École française de Rome (ed.) Spectacles sportifs et
scéniques dans le monde étrusco-italique (Roma), 189-210.
Edwards, C. 1993. The Politics of Immorality in Ancient Rome (Cambridge).
Favro, D., Johanson, C. 2010. Death in Motion: Funeral Processions in the Roman Forum,
JSAH 69:1, 12-37.
Faccenna, D. 1954. Statua di Vittimario scoperta a Fondi, ArchCl 6, 23-42.
Feldherr, A. 1998. Spectacle and Society in Livy’s History (Berkeley).
Ferriès, M.C. 2009. Luperci et Lupercalia de César à Auguste, Latomus 68, 373-92.
Flower, H.I. 1996. Ancestor Masks and Aristocratic Power in Roman Culture (Oxford).
Foster, S.L. (ed.) 1995. Choreographing History (Indianapolis).
Foster, S.L. (ed.) 1996. Corporealities. Dancing, Knowledge, Culture, and Power (London).
Foster, S.L. 2009. Choreographies and Choreographers, in: Foster, S.L. (ed.) Worlding
Dance. Studies in International Performance (London), 98-118.
Foster, S.L. 2010. Choreographing Empathy: Kinesthesia in Performance (London/
New York).
Fowler, D. 2007. Laocoon’s Point of View: Walking the Roman Way, in: Heyworth, S.J.
et al. (eds.) Classical Constructions: Papers in Memory of Don Fowler, Classicist and
Epicurean (Oxford), 1-17.
Franko, M. 1995. Dancing Modernism / Performing Politics (Bloomington).
Giersdorf, J.R. 2009. Dance Studies in the International Academy: Genealogy of a
Disciplinary Formation, Dance Research Journal 41.1, 23-44.
Foster, S.L. 2013. The Body of the People: East German Dance since 1945 (Madison).
Glinister, F. 2011. ‘Bring on the Dancing Girls’: Some Thoughts on the Salian Priesthood,
in: Richardson, J.H., Santangelo, F. (eds.) Priests and State in the Roman World
(Stuttgart), 107-36.
Granino Cecere, M.G. 2014. I Salii: tra epigrafia e topografia, in: Urso, G. (ed.) Sacerdos.
Figure del sacro nella società romana (Firenze), 105-28.
Gunderson, E. 2000. Staging Masculinity: The Rhetoric of Performance in the Roman
World (Ann Arbor).
Habinek, T. 2005. The World of Roman Song: from Ritualized Speech to Social Order
(Baltimore).

Kowalzig, B. 2013. Broken Rhythms in Plato’s Laws. Materialising Social Time in the
Chorus, in: Peponi, A.E. (ed.) Performance and Culture in Plato’s Laws (Cambridge),
171-211.
Kwan, S. 2013. Kinesthetic City: Dance and Movement in Chinese Urban Spaces (Oxford/
New York).
Lambrechts, P. 1946. Mars et les Saliens, Latomus 5, 111-19.
Lepecki, A. 2007. Choreography as Apparatus of Capture, The Drama Review 51.2, 119-23.
Mauss, M. 1934. Les techniques du corps, Journal de Psychologie 32, 271-93.
McLynn, N. 2008. Crying Wolf: The Pope and the Lupercalia, JRS 98, 161-75.
Michels, A.K. 1953. The Topography and Interpretation of the Lupercalia. TAPhA 84,
35-59.
Moore, T. 2012. Music in Roman Comedy (Cambridge).
Morel, J.P. 1969. La iuventus et les origines du théâtre romain, REL 47, 208-52.
Morris, G., Giersdorf, J.R. (eds.) 2016. Choreographies of 21st Century Wars (Oxford).
Noland, C. 2009. Agency and Embodiment. Performing Gestures/Producing Cultures
(Cambridge).
North, J.A. 2008. Caesar at the Lupercalia, JRS 98, 144-60.
North, J.A., McLynn, N. 2008. Postscript to the Lupercalia: From Caesar to Andromachus,
JRS 98, 176-81.
Oakley, S.P. 1998. A Commentary on Livy Books VI-X. Vol. 2. Books VII-VIII (Oxford).
Olsen, S. 2017. Kinesthetic choreia: Empathy, Memory, and Dance in Ancient Greece, CP
112.2, forthcoming.
Östenberg, I., Malmberg, S., Bjørnebye J. (eds.) 2015. The Moving City: Processions,
Passages and Promenades in Ancient Rome (London).
O’Sullivan, T. 2011. Walking in Roman Culture (Cambridge/New York).
Piccaluga, G. 1962. L’aspetto agonistico dei Lupercalia, SMSR 33, 51-62.
Piccaluga, G. 1965. Elementi spettacolari nei rituali festivi romani (Roma).
Reynolds, D. 2013. Empathy, Contagion, and Affect. The Role of Kinesthesia in Watching
Dance, in: Brandstetter, G. et al. (eds.) Touching and Being Touched: Kinesthesia and
Empathy in Dance and Movement (Berlin/Boston).
Romano, G. 2005. Les sodalités. Luperques, ThesCra 5, 89-91.
Rüpke, J. 1990. Domi militiae: die religiöse Konstruktion des Krieges in Rom (Stuttgart).
Rüpke, J. 2007. Religion of the Romans (Cambridge/Malden).
Rüpke, J. 2008. Fasti Sacerdotum: A Prosopography of Pagan, Jewish, and Christian
Religious Officials in the City of Rome, 300 BC to AD 499 (Oxford/New York).
Sarullo, G. 2014. Il Carmen Saliare: indagini filologiche e riflessioni linguistiche (Berlin).
Scheid, J., Granino Cecere, M.G. 1999. Les sacerdoces publics équestres, in:
Demougin, S. et al. (eds.) L’ordre équestre. Histoire d’une aristocratie. IIe s. av. J-C-
IIIe s. ap. J-C. (Roma), 79-189.
Sklar, D. 2008. Remembering Kinesthesia: An Inquiry into Embodied Cultural Knowledge,

in: Noland, C., Ness, S.A. (eds.), Migrations of Gesture, (Minneapolis/London), 85-111.
Sumi, G.S. 2002. Impersonating the Dead: Mimes at Roman Funerals, AJP 123:4, 559-85.
Tomei, M.A. 1999. Scavi francesi sul Palatino. Le indagini di Pietro Rosa per Napoleone
III (1861-1870) (Roma).
Torelli, M. 1990. Riti di passaggio maschili di Roma arcaica, MEFRA 102, 92-106.
Torelli, M. 1997. Appius Alce. La gemma fiorentina con rito saliare e la presenza dei
Claudii in Etruria, SE 63, 227-55.
Tortorella, S. 2000. Luperci e Lupercalia. La documentazione iconografica, in: Carandini,
A., Cappelli, R. (eds.) Roma: Romolo, Remo e la fondazione della città (Milano),
244-255.
Valli, B. 2007. Lupercis nudis lustratur antiquum oppidum palatinum. Alcune riflessioni
sui Lupercalia, Florentia 2, 101-54.
Von Rohden, H. Winnefeld, H. 1911. Architektonische römische Tonreliefs der Kaiserzeit
(Berlin/Stuttgart).
Wiseman, T.P. 1995a. The God of the Lupercal, JRS 85, 1-22
Wiseman, T.P. 1995b. Remus: a Roman Myth (Cambridge).
Wissowa, G. 1902. Religion und Kultus der Römer (München).
Wrede, H. 1983. Statuae Lupercorum habitu, RM 90, 185-200.
Ziolkowski, A. 1998-9. Ritual Cleaning-up of the City: From the Lupercalia to the Argei,
AncSoc 29, 191-218.

brill.com/grms
Dionysius of Halicarnassus on Rome’s Greek

Musical Heritage
Andrew Barker
University of Birmingham, Birmingham, United Kingdom
[email protected]
Abstract
Dionysius tells us that his main objective in writing the Antiquitates Romanae, his
massive history of Rome, was to convince his fellow-Greeks that the Romans were
by origin Greeks themselves, that in their customs they preserved central features
of the noble Greek culture they had inherited, and that the people under whose
regime the Greeks now lived were therefore not to be despised or resented as barbar-
ians. This paper examines some of the allusions to music scattered through the text,
considering the extent and nature of the support they give to this thesis, and asking
whether there is anything to be learned from them about the characteristics of the
culture which Dionysius regards as both admirable and essentially Greek, and which
he represents as manifesting itself among the Romans from the earliest times and
persisting among them to the present day.
Keywords
Dionysius of Halicarnassus – Antiquitates Romanae – Rome’s Greek musical heritage
Dionysius of Halicarnassus came from his native Halicarnassus to Rome

shortly after the end of the civil wars, that is, in about 30 BC, and spent the
rest of his life there; he seems to have earned his living as a teacher of rhetoric.
* An earlier version of this paper was presented at the Symposium Cumanum in June 2016.
I was unfortunately unable to be present, and I am very grateful to Angelo Meriani for deliv-
ering it on my behalf.

64 Barker
He is of course best known for his essays on the literary styles of the ancient
Greek orators and historians, and his fascinating study of prose style in the
De compositione verborum. But for all their admirable qualities, these writ-
ings are relatively short, and his great work was something very different, the
Antiquitates Romanae, a history of Rome from its beginnings—indeed rather
before its beginnings—down to the time of the first Punic War. It was in twenty
books, of which the first nine survive complete, together with most of Books X
and XI and scattered fragments of the rest. The business of researching and
composing this massive history occupied him, he tells us, for twenty years.
Dionysius took the view that no adequate history of early Rome had yet been
written, and that contemporary ideas about it were often badly mistaken. His
account, he tells us, is based on his thorough reading of all existing Greek and
Latin writings on the subject, including some which were very old, and he cites
by name a large number of earlier historians, most of whose works are now
lost. Clemence Schultze has argued1 that this is not just empty self-advertising
and that he made these claims in good faith, and I see no reason to doubt this
conclusion; he appears to have been a thorough, conscientious and dedicated
researcher. But his main reason for tackling the project is not merely that no
existing account was satisfactory. As he explains at some length in his pref-
ace, his central purpose was to convince his fellow-Greeks that the people by
whom they were now ruled were not a nation of barbarians, but worthy inheri-
tors of the culture which the Greeks had pioneered. In pursuit of this objec-
tive he undertakes to show that the founders and early inhabitants of the city
were in fact Greeks themselves, and that contemporary Romans still retained
many of the best elements of their forefathers’ traditions; and it is partly
because this conception of the Romans’ origins was so crucial to Dionysius’
agenda that he took his history back to the very earliest times.2 We should
notice immediately that the ‘Greekness’ which he attributes to the Romans
has a strong ethical component, carrying with it standards of both private and
public behaviour, ancestral patterns of Hellenic virtue, in some cases preserved
even more faithfully by contemporary Romans than by the Greeks themselves.3
I have said that Dionysius pursued his researches very thoroughly, but of
course it does not follow that his account of Roman history is always factually
accurate. Many of its early episodes, after all, belong to what we would regard
1 Schultze 2000.
2 For a close examination of Dionysius’ project see Gabba 1991. Wiater 2011, 120-225, presents
a sophisticated study of his approach to historical writing; on the Ant. Rom. in particular see
165-223.
3 See Peirano 2010, especially 39-48.

Dionysius of Halicarnassus on Rome ’ s Greek Musical Heritage 65
as legend rather than history, and he himself comments on various uncertain-

ties and contradictions in the sources on which he drew. But my main concern
here is not with the factual status of what he says about music or anything
else.4 It is to investigate the manner and the extent to which his musical allu-
sions play into the theme of Roman ‘Greekness’ and the ideological agenda
which I have sketched. I shall argue, in addition, that the musical allusions can
give us valuable insights into the nature of this ideological agenda itself. I shall
begin by drawing attention to certain things which—in the light of my com-
ments so far—we would probably expect Dionysius to do, but which in fact he
does not. It may seem a little perverse to set off in this negative way, but as it
turns out, I think we can learn quite a lot about his attitude to Greco-Roman
musical culture by reflecting on what he doesn’t do, as well as by examining
what he actually does.
First, music was at the heart of Greek social life from earliest times, and
the Greeks themselves reckoned their musical achievements among the great
glories of their civilization. We would therefore expect Dionysius to represent
it as one of the more important components of the Romans’ Greek cultural
inheritance; and yet he does nothing of the sort. With very few exceptions,
all his allusions to music are brief and undeveloped, apparently just incidental
or even trivial details, brought in on the coat-tails of more significant matters.
Readers are likely to receive the impression that music, in his eyes, is of no real
importance at all.
With this failure to exploit the fact of music’s prominence in Greek culture
in the service of his themes we may associate another surprising absence.
When we think of Greek music and its history, we probably have in mind such
things as the songs of the Homeric bards and the singing and dancing cho-
ruses depicted in the epics, the lyric tradition in all its forms, both monodic
and choral, and the music of elaborate and highly sophisticated genres such
as dithyramb, tragedy, comedy, the kitharoidic, auloidic, auletic and kitharistic
nomoi, professionally composed partheneia and paeans, and so on—in short,
the works composed for high-profile public performances, together with the
music of the symposium. But Dionysius discusses none of these prominent
public genres—he doesn’t even mention them—and his one allusion to the
symposium and its music occurs in an anecdote which certainly does not rep-
resent it in a favourable light. Nor does he mention any of the great composers
whose names appear in other accounts of Greek music in its most glorious
days, from Olympus and Terpander in the early period through to Timotheus,
Philoxenus and others in the late fifth century and the fourth. It seems that his
4 For an interesting discussion of these issues see Schulze 2000.
66 Barker
conception of the music that the Romans inherited from their Greek ancestors
has virtually nothing in common with the Greek music we learn about from
other Greek sources and our modern textbooks.
Now for an absence of a different sort. We are all familiar with Plato’s con-
tention that the kinds of music performed in a city, especially those used in
the context of education, have crucial effects on the moral health and well-
being both of individuals and of the community as a whole. This doctrine,
in its essentials, was shared by Aristotle, and in one form or another it exerted
an enormous influence on ethical and political writings in both Greek and Latin
throughout the Hellenistic and Imperial periods. As a moralizing political and
social historian, Dionysius could be expected to make a good deal of use of it
in his Roman history. But in fact any traces of it that we may be able to detect
are slight and at best implicit; there is no direct appeal to the theory of musical
ēthos anywhere in the surviving text. The nearest he comes to evoking it is in
an allusion to music as one of the components of traditional Greek paideia
(in his account of the education of Romulus at I.84.5); but the passage con-
veys no hint that it has any ethical significance. This is all the more surprising
in view of the obvious and very relevant (though unacknowledged) echoes of
Plato’s Republic in Dionysius’ account of the institutions of Romulus at II.18-19,
which we shall glance at later. His failure to exploit the opportunities that
Plato’s theories could have provided calls for some sort of explanation, per-
haps one connected with the explanation—whatever it may be—for the fact
that in this work he says nothing about Greek philosophy or philosophers, and
nowhere suggests that they exercised any influence on Roman culture. Or, to
be meticulously accurate, he mentions just one Greek philosopher, Pythagoras,
and this is only to deny, on chronological grounds, that he can have had any
influence on King Numa. Numa, he says, lived long before Pythagoras, and
hence Pythagoras cannot have been his teacher, as some other historians had
asserted (II.59).
The last item in my list of ‘missing ingredients’ in Dionysius’ work is not
in fact completely absent, but it certainly figures less prominently than one
might have expected. One of his central themes is the importance of cohe-
sion and mutual agreement between the individuals and classes that form the
citizen body; and a famous analogy that appears repeatedly in this connection,
especially from Plato onwards, is that between the attunement, the harmonia,
of high and low notes on an instrument’s strings and the ‘harmonizing’ of indi-
viduals and classes in the community. It seems to be a piece of imagery that
would have served Dionysius’ purposes well, as would the related metaphor
of musical concord, symphōnia, very commonly used in the same contexts in

a similar way. But Dionysius never applies the word symphōnia or its cognates
to political or social conditions. He often makes comments on such matters
into which this word would fit very naturally, but he always selects some other,
non-musical expression in its place; homonoia, ‘agreement’, is his usual choice.
One might even get the impression that symphōnia, for some reason, is a term
which in these contexts he deliberately avoids. As for the concept of political
or social ‘attunement’, it appears as an explicitly musical image just once, very
briefly, at II.62.5, where Numa is described as ‘having attuned (harmosamenos)
the whole people like an instrument’. Dionysius makes nothing more of the
musical resonances of this simile here, and there is only other place where
the notion of harmonia figures in such a context. That is at II.11.3, where Gaius
Gracchus is said to have destroyed the harmonia of the state; and in this case
we cannot even be sure that any musical allusion is intended at all. The idea
of an analogy between musical and socio-political harmonia seems to play at
most a very minor role among Dionysius’ themes.
At the end of this paper I shall say a little more about topics and concep-
tions which are surprisingly absent from Dionysius’ work, but we must now
move on to consider the text’s actual musical content. It may be that—aside
from anything else we may learn from them—the relevant passages will shed
some light on the question why these other matters are apparently ignored.
Let us begin with the passage in which Dionysius mentions music for the
first time.
Right down to my own time, Falerii and Fescennium were inhabited

by Romans, and preserved a few small glimmers of the Pelasgian race,
though previously they had belonged to the Sikels. In these cities there
remained many of the ancient usages which the Greeks once employed—
for instance the form of their weapons of war, such as Argolic shields
and spears; the fact that whenever they sent an army out beyond their
frontier, either to start a war or to resist invaders, certain holy men went
ahead of the others, unarmed, carrying proposals for a truce; the arrange-
ment of their temples, the images of their gods, their sanctifications,
their sacrifices and many other such things. But the most conspicuous
of all the reminders that the people who drove out the Sikels once lived
in Argos is the temple of Hera at Falerii, which was constructed just like
the one at Argos; here there was the same form of sacrificial ritual, holy
women serving in the precinct, the unmarried girl called the ‘basket-
bearer’ who initiated the sacrifices, and choruses of virgins hymning the
goddess in their ancestral songs (I.21.1-2).
68 Barker
The only allusion to music in this passage comes right at the end, but we need
to take note of the rest, too; the allusion is meaningless without its context. To
extend this context a little further, we have been told that these Pelasgians are
the second wave of Greeks to have arrived in central Italy at a very early date,
long before the Trojan War. There they joined the first group—whom Dionysius
calls ‘Aborigines’ and who came, he says, from Arcadia—and together they
expelled the barbarous Sikels, who were eventually driven off the Italian main-
land altogether. (Dionysius is evidently concerned to make it clear that these
genuine barbaroi played no part in the developments which led to the founda-
tion of Rome.) The Pelasgians originated in Argos, as we learn from the pres-
ent passage, but had not migrated to Italy directly from the Peloponnese; their
ancestors had moved long ago from Argos to Thessaly, and it was from Thessaly
that these Pelasgians travelled to Italy and settled in the regions we know as
Latium and Etruria. At this stage, according to Dionysius’ account, Rome did
not yet exist, and the towns he mentions here, Falerii and Fescennium, are
some 50 kilometres to the north of the site where it later grew up, so there is
no sense in which the Pelasgians of this period were Romans. But Dionysius
is not trying to pretend that they were; the claim he is making is only that
when the actual founders of Rome finally arrived on the scene, the population
of the area that received them and joined with them in their enterprise were
unambiguously Greek.
To return to our passage itself, Dionysius sets out a string of pieces of evi-
dence that the Pelasgians came originally from Argos. He doesn’t tell us explic-
itly where he found his information, but the reference in the first sentence
to ‘glimmers’ or ‘remnants’ (zōpyra) of these people which still existed in his
own time suggests that he may have confirmed aspects of the written records
or added some of the details on the basis of personal observation, as he occa-
sionally does elsewhere. But it is their existence in the distant past rather than
their survival into the present that really concerns him, as is shown by his con-
sistent use of the imperfect rather than the present tense in the remainder of
the passage.
We need to notice that everything mentioned in Dionysius’ catalogue of evi-
dence belongs in one or other or both of two categories—military (shields,
spears, armies) and religious (holy men, holy women, sacrifices and other
rituals, the temple of Hera and so on)—and most of the passages we’ll come
to later are similar in that respect. But in fact it’s slightly misleading, at least
in the present passage, to pair the military group with the religious group as
though they were of equal importance, since the only items on the list which
are unconnected with religion are the shields and spears; martial expeditions
are indeed mentioned, but only as the setting in which the truce-bearing

‘holy men’ perform their function, and it is the activities of the holy men, not
the operations of the army, which constitute the relevant piece of evidence.
The items and practices which reveal the Greekness of the ancient Pelasgians
are overwhelmingly of a religious sort, and it is among these that we find our
allusion to music: ceremonies connected with the temple of Hera at Falerii
included ‘choruses of virgins hymning the goddess in their ancestral songs’.
What are we to say about this allusion? It seems very slight, but there are
points worth noticing. There is first its religious context, which I have just
noted. Then there is the fact that the singers are young and female; they are
parthenoi, unmarried girls, regularly regarded as symbolizing innocence and
purity, and choral songs described as partheneia and performed by such
groups were widespread in archaic and classical Greece. Singing women are
also mentioned by Dionysius in other passages to do with religion in the early
period; there are examples, for instance, in I.31, I.55 and II.4. Thirdly, what
they sing are ōidai patrioi, songs belonging to the ancestral tradition of their
homeland, which must in this case be Argos, and references to ōidai patrioi
or hymnoi patrioi appear in several other passages too, notably in II.34, II.70
and III.32. There is no indication that the pieces so described were elaborate
compositions by eminent composers, like the partheneia of Alcman or Pindar,
for instance; on the contrary, the fact that they are unattributed, and described
merely as ōidai patrioi, strongly suggests that they were simple and traditional,
and probably anonymous, as were the songs described in the same terms else-
where. That would fit well, too, with the profile of other pieces of music which
Dionysius mentions, and I’m reasonably sure that it’s correct.
But there is a more important point to make about the adjective patrios,
which is that Dionysius’ deployment of it in the Antiquitates Romanae is a
major element in the repertoire of his ideological rhetoric. It appears over
90 times in the surviving parts of the text, most often attached to nouns such as
thysiai (‘sacrifices’), nomoi (‘laws’), ethismoi (‘customs’) and above all to nouns
designating political constitutions—referred to either by the general term
politeia,’constitution’, or as a constitution of a specific type, such as aristokratia.
The phrase patrios politeia, ‘ancestral constitution’, is of course much older
than Dionysius; it appears first, as far as I know, in the late fifth century, in
passages from the sophist and rhetorician Thrasymachus and the orator Lysias
which are quoted elsewhere by Dionysius himself.5 More generally, the adjec-
tive patrios is common in fifth- and fourth-century prose writings on politi-
cal themes, where it is attached most often to the noun nomos, ‘law’. In all
these occurrences, both in the earlier texts and in the works of Dionysius, it
5 Thrasymachus at Demosth. 3, Lysias at Lys. 32.
70 Barker
is much more than a descriptive expression; it confers weight and authority

on the law, custom or constitution that it characterizes, and at the same time
implies that the writer himself is someone who values the ancestral tradition,
thereby showing himself to be a person of austere moral character and sober
judgement. This tells us a good deal, then, about the authorial persona that
Dionysius wants to project in his history. To return to our present passage, the
overt purpose of designating the songs as ōidai patrioi is simply to indicate that
they belong to the tradition of ancient Argos; but the atmosphere of authority
and approval that suffuses the adjective in all its other contexts spills over onto
this occurrence too, and registers the honourable status of these songs within
the framework of a solidly respectable social ideology.
We may usefully bear these few points in mind as we move to another
passage. It is concerned with the next group of Greek settlers after the
Pelasgians to arrive in this part of Italy; according to Dionysius (I.31) it was
an influx of Arcadians under the leadership of Evander, sixty years before the
Trojan War. In I.33, after writing at some length about the typically Arcadian
forms of sacrifice and other religious ceremonies which these people brought
with them, some of which survived, he says, right down to his own time, he
continues as follows.
These Arcadians are also said to have been the first to transport into
Italy the use of Greek writing (grammata), which had recently made its
appearance, and music played on instruments, which are called lyres and
trigōna (harps) and auloi, the previous people [sc. in Italy] having used
no musical devices apart from pastoral Panpipes; and they established
laws, and transformed people’s way of life from its pervasive bestiality to
a civilized condition, and introduced crafts and occupations and many
other things that are beneficial to the community; and for these reasons
they met with great favour from those who had received them. This was
the second Greek race after the Pelasgians who came to Italy and shared
a place of residence with the Aborigines, settling in the best part of Rome
(I.33.4-5).
Dionysius represents these Arcadians as the bringers of a civilized culture to the

area in which Rome was founded, and the list of benefits which they brought
is obviously reminiscent of other Greek accounts of the transformation of
human life from a primitive to a civilized condition. We might think in particu-
lar of the gifts that Prometheus is said to have given to benighted humans in
the Prometheus Vinctus and in the ‘Great Speech’ put by Plato into the mouth

of Protagoras,6 where they are supplemented by attributes of a moral sort, pro-

vided by Zeus, without which cooperative social life is impossible. But there
is a difference. In those legends the skills, attributes and other items which
are necessary for a civilized human life are provided by a superhuman being,
and had previously been unknown to mankind; whereas in our passage these
benefits already existed among some human beings, constituting their own
pre-existing culture, and the action recorded is their transmission of them to
others. In this respect the story is more like accounts given by some of the
Alexander historians, and recorded by Diodorus Siculus and Flavius Arrianus
(Arrian), of tales told by the Indians about a great conqueror who came with
his army from the north, at a time when the Indians had no cities and were
living in what we might call a ‘state of nature’. The Greeks identified this con-
queror with Dionysus, though Diodorus also reports, rather paradoxically, that
he reigned over India for 52 years and then died, and was succeeded by his sons
(Diod. Sic. II.38.6). These people, we are told, taught the Indians agricultural
skills, and how to preserve fruit and make wine; they founded cities and intro-
duced laws, law-courts and religious observances in honour of the gods; and
they established everything else which a civilized life involves. Arrian adds that
among the religious rituals that they taught were those in honour of Dionysus
himself, in which they played cymbals and drums, and danced the satyr dance
known to the Greeks as the kordax (Arrian Hist. Ind. 7.8).
The passage in our text obviously belongs to the same family as all these
others, but from our present perspective its main interest is in the reference
to musical instruments. The syrinx or Panpipe is regularly treated as an instru-
ment of shepherds and others who live in wild places outside the centres of
civilization, typically in the mountains; it has no significant role in the life
of the city, and here is emblematic of the primitive social conditions under
which these Italians were living. The other instruments named in the passage,
the lyres, harps and auloi brought with them by the Arcadians, are equally
emblematic of a properly established civic community. But Dionysius’ allusion
to them here raises a significant question. I said earlier that he never mentions
music of the sort that was heard in high-profile public performances, in dithy-
rambs, kitharōidic nomoi and so on, but we might wonder whether the fact
that he now treats these instruments as essentials of a civilized life is, excep-
tionally, an implicit recognition that such performances have genuine cultural
importance. On the other hand, with one probable but only implicit excep-
tion, he says nothing anywhere else in this work to suggest that they have a
6 Plato, Prt. 320c8-322d5.
72 Barker
significant part to play in the idealized proto-Roman civilization he is describ-

ing. He mentions lyres in only one other passage (VII.72.5), and there the
message conveyed is entirely different. Auloi appear in that passage too and
in a handful of others, but none of them assign cultural significance to public
performances of a sophisticated sort.
How, then, are we to interpret the allusion to these instruments in the pres-
ent passage? Perhaps the mention of a third type of instrument, the trigōnos,
may give us a clue. Harps of various kinds were known in the Greek world
from early times, but in the archaic and classical periods were never performed
in public. Their place, as depicted both in literature and in vase-paintings, is
always indoors, either in domestic scenes where they are played especially
by women, or in the context of the symposium. We hear of a few public per-
formances on harps in Hellenistic times, but even then they are very much
the exception. So we might infer from this that the culturally essential music-
making that Dionysius has in mind when he mentions the three types of instru-
ment does not consist in elaborate performances presented in public, but
takes place in the home or in other relatively private settings. A suggestion of
this sort may be encouraged by a parallel with the contents of the Greek-style
education given to Romulus and Remus (I.84.5), which I mentioned earlier;
writing and music are the first items listed among its elements, just as in our
passage they are first among the elements of civilization. The concepts of edu-
cation and civilization are always very closely intertwined, and it seems at least
a reasonable hypothesis that when Dionysius represents these instruments as
pieces of equipment necessary for civilized living, their role is in the develop-
ment of the musical skills learned in the course of a child’s education, and in
the exercise and appreciation of these skills in a cultivated citizen’s private life.
But if what he has in mind is musical education, his views about it may
need to be qualified in the light of a story he tells later about Aristodemus of
Cumae, who engineered a coup d’état against the ruling oligarchs in 504 BC,
and made himself tyrant of his city. After killing most of his opponents and
banishing their children to the remote countryside, he set about ensuring that
the children of the remaining citizens would never be able to mount an effec-
tive rebellion against him.
In order that a noble and manly spirit would not spring up in any of the
other citizens, he decided to feminize the young men who were being
raised in the city through the manner in which they were brought up,
putting an end to gymnasia and to training in the use of weapons, and
transforming the way of life that the children had previously followed. He
commanded the boys to wear long hair like the girls, and to decorate it

with flowers, curl it, and tie it up in hair-nets, to wear embroidered robes
hanging down to their feet, with delicate, soft shawls draped over them,
and to spend their lives in the shade. When they went to the schools
of the dancing-teachers, aulos-players and other such lackeys of the
Muses, the paidagōgoi who accompanied them were women carrying
parasols and fans; and these same women bathed them, bringing combs,
alabaster pots containing perfumes, and mirrors into the bath-houses.
Through this kind of upbringing he went on ruining the young men until
they had completed their twentieth year, and from that time onwards he
allowed them to be counted as men (VII.9.3-5).
This passage too has an eminent ancestor; it is Herodotus, in his account of

the advice that Croesus gave to Cyrus about how he should tame the Lydians
he had conquered, so that they would not become a threat to his authority
(Hdt. I.155). Weapons are course banned, as in the scheme of Aristodemus.
Croesus says nothing about suitable coiffures or bathing arrangements, but he
does prescribe changes in the Lydians’ clothing, requiring them to wear tunics
under their robes, and he says that their sons should be trained to play lyres and
harps. If Cyrus lays down these rules, he says, he will soon see that the Lydians
have become women instead of men. The parallel between the two passages
is not exact, but Dionysius is evidently reactivating a topos which Herodotus
had established. Perhaps he is drawing on other classical Greek sources too.
Herodotus does not mention the parasols which figure in Dionysius’ tale, or
the injunction that the young men must spend their time in the shade—to
prevent them getting the tanned and weather-beaten complexions that are
marks of virile men; but parasols were certainly part of the Greeks’ image of
the soft-living Lydians on which Dionysius seems to be drawing, as we know
from vase paintings of effeminate aesthetes who affected Lydian manners
around 500 BCE, including Anacreon, the Oscar Wilde of his time.7
A musical training, as construed in this passage of Dionysius, is something
that contributes to the moral enfeeblement and feminization of young men,
and the word he uses to refer to the teachers of mousikē—mousokolakes, which
I have translated as ‘lackeys of the Muses’—is plainly contemptuous. Yet else-
where he recognizes music as an integral part of a Greek educational system
of which he approves. But there need be no contradiction here, though I don’t
think we should try to resolve it by appealing to distinctions, of the type well
known from Plato, between ethically beneficial and ethically deleterious kinds
of music. Nothing in Dionysius’ remarks suggests that the music taught in the
7 See e.g. Beazley 1963, 185 no. 32, and cf. Caskey and Beazley 1954, 57-61.
74 Barker
approved Greek curriculum and the music taught to the Cumaeans are differ-
ent in kind. What really makes the difference, I suggest, is partly that musical
education in the latter case is just one among a good many other things which
interact with one another to produce their effect, and partly that the Cumaean
youths are subjected to these feminizing influences for such a long time, right
up to the time when they complete their twentieth year. By the standards of
the major authorities on such matters, Plato and Aristotle, that is far too long,
and is likely to have disastrous consequences. According to Plato’s Socrates,
Whenever anyone lets music beguile his soul with its piping, and lets it
pour into his soul through his ears as though through a funnel the sweet
and soft and mournful melodies we have just been discussing, and when
he uses up his whole life humming, enraptured by song, then if he has
anything of the spirited element in him, to begin with this man will
temper it like iron and make useful what was useless and hard. But if he
persists in enchanting it without ceasing, he will eventually dissolve his
spirit and melt it till he pours it away, and cuts, as it were, the sinews from
his soul, and makes of it a ‘feeble warrior’ (Plato Rep. 411a-b).
That, of course, is precisely the result that Aristodemus of Cumae was aiming
for. But it would be rash to assume that Dionysius has Plato specifically in mind
at this point. As we can see from Herodotus’ story about Croesus and Cyrus
(along with several other fifth-century texts), the notion that excessive atten-
tion to music makes men soft and effeminate was present in Greek culture
decades before Plato wrote about it, and independently of any philosophical
baggage.
On the other hand, it’s clear from his other writings that Dionysius was famil-
iar with Plato’s dialogues, and there’s a passage in the Antiquitates Romanae
itself (II.18-19) which leaves little doubt that he had read the Republic and
absorbed a good many of its ideas. The context is his account of the admirable
institutions laid down by Romulus.
He realized that the causes of the good government of cities, which

all politicians chatter about but few succeed in establishing, are first,
the favour of the gods, whose presence brings all human activities to
greater success; next moderation (sōphrosynē) and justice (dikaiosynē),
through whose influence people do less harm to one another and are
more in agreement, and measure happiness not by the quantity of the
most disgraceful pleasures but by what is noble (kalon); and finally cour-
age in war, which also makes the other virtues useful to those that have

them. He understood that none of these good things arise by chance, for
he knew that it is good laws and the pursuit of noble enterprises that
make a city reverent, moderate, devoted to justice and courageous in
war (II.18.1-2).
The thoughts expressed here are very close to those of the Republic, despite the
fact that eusebeia, ‘reverence’ or ‘piety’, has supplanted Plato’s sophia, ‘wisdom’,
in the list of virtues. The substitution is wholly in line with the central impor-
tance of religious practices in Dionysius’ various portrayals of well constituted
communities, and with the complete absence from them of any suggestion
that advanced intellectual attainments have a significant part to play. We
might reasonably, on the other hand, argue that the contents of the passage are
too generalized, and in a sense too commonplace, to allow us to be sure that
Dionysius’ inclusion of them was prompted by his reading of Plato. In this
respect the sequel has more weight. After setting out the arrangements that
Romulus put in place for the proper worship of the gods, he continues:
But he threw out all the traditional legends about the gods which contain
blasphemies or evil-speaking against them, holding that they are wicked,
useless, unseemly and unworthy of good men, let alone of the gods; and
he trained the people to say and think the best things about the gods,
and to attribute to them no behaviour unworthy of their blessed nature.
[19] For among the Romans there is no story of Ouranos being cas-
trated by his own sons, or of Kronos destroying his own children for fear
of their attacks on him, or of Zeus putting an end to the rule of Kronos
and shutting his own father up in the dungeon of Tartarus, or indeed of
wars or wounds, or of gods being tied up or enslaved by humans (II.19.1).
The Platonic ancestry of these remarks can hardly be doubted. It is self-

evidently a rather highly coloured reprise of a well-known passage in Republic
Book II (377b-378e); and the next sequence of statements in Dionysius’ account
(II.19.2-5, too long to quote in full), which includes both direct and indirect
allusions to music, is equally in tune with views Plato expresses in both the
Republic and the Laws. Essentially, it makes three claims. First, no Roman festi-
val is conducted with women dressed in black, beating their breasts and sing-
ing laments for gods who have disappeared, as the Greeks do for Persephone
and other divine beings. Secondly, even in these decadent modern times,
the Romans do not go in for wild ecstasies or Corybantic raving, or Bacchic
rituals and secret mysteries, or all-night activities with men and women con-
sorting together in temples, or any other such excesses; all their ceremonies
76 Barker
are conducted with a dignity unequalled elsewhere, among either Greeks or

barbarians. Finally, Rome is flooded with people of innumerable different
races, who conduct their sacred rituals in their own peculiar ways. But unlike
other cities, Rome does not adopt them all as her own, and even when—in
obedience to certain oracles—it does adopt a few of them, it celebrates them
in ways consistent with its own customs and rejects all outlandish flummery.
Thus in the annual festivals of Cybele, for instance, the praetors perform sac-
rifices and hold games in the Roman way, but the priest and priestess of the
goddess are Phrygians, not Romans, and it is they who process through the city
wearing strange symbols on their multi-coloured clothes, beating drums to the
accompaniment of auloi. Romans are strictly forbidden to do any such things,
and they ‘detest all high-flown nonsense that lacks decorum’.
In music as in all else, what Dionysius dismisses as un-Roman are excesses
of emotional display, unrestrained ecstasies, elaborate ostentation and any-
thing that is outlandishly exotic; what he applauds is the maintenance of
ancient tradition, simplicity and restraint. He represents these characteristics
as fundamentally Roman, in some cases more consistently maintained by
Romans than by the Greeks themselves.8 But at the same time the unmistak-
ably Platonic overtones of everything in these two chapters are enough to
convey the impression that like so much else in the Roman repertoire, these
noble traditions are by origin Greek, even if contemporary Greeks have partly
abandoned them.
The institutions of Romulus, along with those of King Numa, are the ones
that Dionysius praises most of all among those of the early Romans. Excerpts
from another passage of Book II, again about Romulus, will help to confirm
some of the points I have been making. It depicts the celebrations after one of
his successful campaigns against the Sabines.
He led his army home, carrying the booty from those who had fallen in
the battle and dedicating the finest of the spoils as offerings to the gods;
and at the same time he made many sacrifices. He himself came last in
the procession, dressed in a purple robe with a laurel wreath on his head,
and riding in a four-horse chariot so as to preserve his royal dignity. The
rest of the army, both foot-soldiers and horsemen, accompanied him,
deployed in their various divisions, hymning the gods in their ancestral
songs, and praising their leader in improvised compositions.
8 See Peirano 2010, especially 39-48.

After describing the way in which they were received by their families and
other citizens, Dionysius continues:
That, then, was the style of the trophy-carrying procession of victory and
the sacrifice, which the Romans call a ‘triumph’, in the form in which it
was originally instituted by Romulus. But in our life-time it has become
very expensive, and is now an ostentatious display, whose theatricality is
designed to make a display of wealth rather than to register approval of
valour; and in all its aspects it has departed from its ancient simplicity
(excerpted from II. 34.1-3).
Dionysius is evidently trying to impress on his readers the conviction that the
glitzy and ostentatious ‘theatricality’ of Roman triumphs in his own time is to
be deplored, and that it would be much better to return to the simplicity of the
ceremonies in the time of Romulus; and the nature of the music included in
those early victory parades contributes substantially to their modest profile.
There are ‘ancestral’ songs in honour of the gods, ōidai patrioi once again, songs
rooted in ancient tradition, which are not sung by professionals or specially
trained choruses, but by the soldiers themselves. Similarly, their expressions of
praise for their general Romulus are not the elaborate works of sophisticated
composers or poets. The soldiers praise him poiēmasin autoschediois, that is, in
pieces that they extemporize on the spot. We cannot be sure whether or not
these poiēmata were musical in our sense of the word—that is, whether or not
they included tunes, invented ad hoc by the soldiers, to go with their extem-
pore verses—since when it is not further qualified, the noun poiēma allows
for either possibility. But when the same phrase reappears at VII.72.11, again
in connection with soldiers in a triumphal parade, it clearly refers to songs:
poiēmata aidousin autoschedia, ‘they sing extempore pieces’. Be that as it may,
however, the image is of a simple and informal celebration, far removed from
the pomp and swagger of Roman triumphs in later periods.
Dionysius’ statements here fit well into the ideological framework of the
other passages we’ve been considering. But his description in Book VII of the
large and colourful procession called the Pompa circensis which introduced
the celebration of the ‘Great Games’, the ludi magni, may seem to undermine
my characterisation of this ideology. Here, as quite often elsewhere, Dionysius
says that he is relying on the authority of Quintus Fabius Pictor (c. 200 BC); and
he claims to be describing the proceedings in the form in which they were insti-
tuted, by command of the senate, not long after the expulsion of Tarquinius
Superbus, who went to ground in Cumae with the tyrant Aristodemus around
the end of the sixth century. The topic is introduced in VII.70-71; details follow
78 Barker
in VII.72. By the standards of this text VII.72 is an enormous chapter, much too
long to be quoted or adequately discussed here; I shall merely summarize it as
best I can, and try to bring out some of its most significant features.
The procession is led by young men approaching military age, on horseback
and on foot, who are followed by a company of horse-drawn chariots. Then
come the contestants in the athletic games, naked except for their loin-cloths.
(Here Dionysius pauses to demonstrate, with quotations from Homer, that the
practice of wearing loin-cloths in athletic contexts belongs to the most ancient
Greek tradition. The Greeks themselves, he says, have abandoned it and now
compete completely naked, but the Romans still preserve the original Greek
custom.) After the athletes come troops of dancers, divided by age into three
groups: men, adolescents and boys. Their dancing is accompanied by auletes
who play small auloi of an antique design—as they still do, says Dionysius,
in our own time—and by lyre-players playing seven-stringed lyres veneered
with ivory, and what he describes as ta kaloumena barbita, ‘the instruments
called barbita’, which are the long-armed, deep-pitched lyres we see quite
often in sixth- and fifth-century Greek vase paintings, especially in scenes with
Dionysus and his satyrs. I suppose that Dionysius refers to them in this round-
about way because they may be unfamiliar to his Greek readers; though they
belong to the Greek tradition, he says, the Greeks no longer use them, but the
Romans still do so ‘in all their ancient ceremonies of sacrifice’.
After describing the dancers’ costumes, Dionysius identifies their move-
ments, in the quick, four-beat rhythm known as prokeleusmatic, as those of
a war-dance, and he identifies this, in turn, with the Greek war-dance called
the pyrrhichē—a very ancient Greek institution, he says, tracing it back to
either the gods’ celebrations after the defeat of the Titans or the noisy dances
of the Curetes when they were guarding the infant Zeus. Those alleged ori-
gins are at least of the right sort for a vigorous war-dance, but they suffer from
the disadvantage of being obviously mythical; and we are bound to judge that
Dionysius’ next attempt to prove its Greek origins, by quoting verses from
Iliad 18, is also a failure. The passages he quotes describe dance-scenes depicted
on the shield made by Hephaistos for Achilles, but the dancing shown there
could hardly be less like a war dance. It is in fact one of the emblems of a city
living happily in peace. I find it puzzling that he is content with such a feeble
way of supporting his claim that the war-dance has Greek ancestry, and it is
hard to believe that his contemporaries found it convincing. One might suggest
that he was simply unable to find evidence by which the Roman war-dance
could legitimately be linked to Greek antiquity, and hoped instead to bamboo-
zle his readers with legends of the gods and appeals to the authority of Homer,
but in that case it is hard to see why he was pushed into this corner; there is no

shortage of allusions to war-dances in classical Greek literature, any of which

would have served his purpose perfectly well.
Dionysius also finds evidence of Greek origins in the people who followed
the war-dancers. They were dressed as Silenoi and satyrs, and danced the Greek
satyr-dance called the sikinnis, and they made fun of the ‘serious’ movements
of the war-dancers by imitating them in a grotesque way. Dionysius asserts that
satyr-dancing and mockery of this sort takes place in a number of Roman set-
tings, including the triumph, in which soldiers sing rude songs about their vic-
torious general, and that it is done most of all at the funerals of wealthy men;
but its origins, he insists, are Greek. He cites only one rather loosely specified
Greek precedent, however, ‘those who ride in wagons in processions at Athens’,
and then adds that he’s afraid he would bore his readers if he added more
evidence to support this assertion, since it is a homologoumenon pragma,
something that is universally agreed to be true. This is the sort of thing that
writers are often inclined to say when evidence for their case is lacking, and we
are entitled to be a little sceptical about his sincerity.
After these satyric dancers came large numbers of lyre-players and auletes,
on whom Dionysius makes no comment, and they are the last musicians to
be mentioned in this passage. They are, in fact, almost the last members of
the procession, followed only by the people who carry the sacred objects—
censers containing burning incense, vessels made of silver and gold, and
finally the images of the gods, who are represented in the same way as those of
the Greeks. There are quantities of them, many of whom Dionysius identifies
by name; they include the twelve Olympians, their precursors in the previous
generation of gods, lesser divinities such as the Muses and the Graces who
came into being after the Olympians, and heroes like Herakles, the Dioscuri
and Pan, who were given divine status after their bodily deaths. It follows,
says Dionysius, that the Romans who founded this festival cannot have been
barbaroi, since plainly they worshipped all the gods of the Greeks and none of
any other nation.
Anyone who reads the whole chapter straight through is likely to get the
impression of a splendid and spectacular event, much more elaborate than
the others of which Dionysius approves, and one that has abandoned the
archaic simplicity of events like Romulus’ triumphal procession. But in one
important sense that is an illusion. If we consider each of the procession’s
components individually, none of them has features which Dionysius criti-
cizes elsewhere; such things as war-dances and satyric mockeries fit perfectly
well into his image of a simple ancient tradition, uncorrupted by luxury and
over-sophistication. In any case, his major theme in this passage is different.
The point that stands out most clearly is that he is using the description as a
80 Barker
major set-piece to underpin his thesis about the Greek origins of Rome and the
Romans. In the previous chapter, in fact, he explicitly tells us that this is his pur-
pose here, adding a significant twist. Some people might suppose, he says, that
the Greek elements found in Roman practices date only from the time after the
Roman conquest of Greece. Dionysius scoffs at the idea. One cannot believe,
he points out, that the Romans would have suppressed their own ancestral tra-
ditions and substituted those of an alien people they had recently conquered;
and we may be reminded of cases mentioned earlier, like the Phrygian rituals
of Cybele, in which they had explicitly refused to do so. But just in case anyone
remains unconvinced, says Dionysius, he will provide evidence from a period
long before the conquest of Greece. Claiming the authority of Quintus Fabius
Pictor is an integral part of his strategy, since Quintus is a historian who wrote
before the Roman conquest, and whose history ended with the second Punic
War. His writings therefore cannot have been contaminated by any knowledge
of post-conquest innovations.
Our survey has shown that Dionysius uses references to music for two
main purposes: to help in establishing the Greek lineage of the Romans and
their customs, and to promote the view that civilization is at its best when
it is characterized by simplicity and restraint; and these two objectives are
linked to one another by their shared appeal to the virtues of the ancestral
tradition.9 In the light of these conclusions, most of the initially surprising
absences which we noted at the beginning of my paper are easily understood.
Dionysius’ profile of the unassuming musical traditions brought to Italy from
Greece has no place for large-scale public genres such as dithyramb and drama
or the nomoi performed by professional soloists, or for allusions to the works of
famous composers, either of which would disrupt the impression he is trying
to create. I would suggest also that his reticence about Plato’s theory of musi-
cal ēthos, though he certainly knew about it, is similarly motivated, along with
the absence from his history of any explicit reference to Greek philosophers
or other intellectuals. Even when he is plainly echoing Plato, as in his com-
ments on the institutions of Romulus, he doesn’t say so; he presents them as
aspects of an ancient tradition, originating in the untutored good sense of an
honourable and morally upright leader. Sophisticated intellectualism of any
9 The survey is of course incomplete, but nothing in the remaining passages casts doubt on
these conclusions. The most important of them are II.70-71 and III.32, which present quite
detailed accounts of the activities of the Salii; the former is also the occasion of another of
Dionysius’ rather unconvincing attempts to show that a Roman practice has Greek origins.
Much has been written on the Salii; see most recently Alonso Fernández 2016, especially
314-319; Alonso Fernández 2017.

sort, like elaborate styles of music, is completely alien to the image of an admi-
rable Roman culture, firmly grounded in the patrioi ethismoi of the Romans’
Greek ancestors, which Dionysius is trying to impress on his Hellenic readers.
A brief postscript. If any readers are wondering how Aeneas and his Trojan
followers fit into Dionysius’ Hellenizing picture of Rome’s origins, I am happy
to inform them (on the authority of I.61-62) that the Trojans too are Greeks,
who like some of the other groups came originally from Arcadia.10
Bibliography
Alonso Fernández, Z., 2016, Choreography of Lupercalia: Corporeality in Roman Public

Religion, GRMS 4.2, 311-332.
Alonso Fernández, Z., 2017, Re-thinking Lupercalia: from Corporeality to Corporation,
GRMS 5.1, 43-62.
Beazley, J.D. 1963, Attic Red Figure Vases, second edition, Oxford.
Caskey, L.D. and Beazley, J.D., 1954, Attic Vases in the Museum of Fine Arts, Boston,
Part 2, London, 57-61.
Gabba, E., 1991, Dionysius and the History of Archaic Rome, Sather Classical Lectures
vol. 56, Berkeley.
Gruen, E.S., 1992, Culture and National Identity in Republican Rome, Ithaca NY.
Peirano, I., 2010, Hellenized Romans and Barbarized Greeks. Reading the End of Dionysius
of Halicarnassus, Antiquitates Romanae, JRS 100, 32-53.
Schultze, CE, 2000, Authority, Originality and Competence in the Roman Archaeology of
Dionysius of Halicarnassus, Histos 4, 6-49.
Vanotti, G., 1995, L’altro Enea. La testimonianza di Dionigi di Alicarnasso, Rome.
Wiater, N., 2011, The Ideology of Classicism: Language, History and Identity in Dionysius
of Halicarnassus, Berlin.
10 The idea is considerably older than Dionysius. For discussions of the Aeneas legend in
this context, see Gruen 1992, 6-51, and Vanotti 1995.
brill.com/grms
Music, Sexuality and Stagecraft in the

Pseudo-Vergilian Copa
Harry Morgan
St. Cross College, University of Oxford
[email protected]
Abstract
The Pseudo-Vergilian Copa (‘The Female Tavern-Keeper’) opens with the eponymous
character dancing ‘drunkenly’ and ‘sexily’ to the rhythms of the castanet. Her perfor-
mance, which is accompanied by several other musical instruments, sets the scene for
a brief, yet richly detailed, vignette describing the attractions of a rustic Roman tavern.
This paper examines how the poet uses music to (re)construct the Copa’s sensory world.
The dancing tavern-keeper is a complex literary creation, which incorporates influences
from both the elegiac and pastoral traditions as well as from contemporary visual cul-
ture. Moreover, her characterisation as an erotic, exotic entertainer invites comparisons
between the tavern, the dining-room and the theatre as interactive performance spaces.
The alluring dancers and musicians who performed in these venues inspired a number
of poetic depictions in the early imperial period, and the commonalities between these
depictions can in turn shed valuable light on our poem and its elusive protagonist.
Keywords
dance – tavern – sexuality – mime
Introduction
The Copa, a thirty-eight line elegiac poem preserved in the Appendix Vergiliana,
opens with the description of a rambunctious musical performance.1 Whisked
1 A preliminary version of this paper was given at the Symposium Cumanum 2016, held at the
Villa Vergiliana on the subject, ‘Music in the Time of Vergil’. I wish to thank the organiser,

Music, Sexuality And Stagecraft In The Pseudo-vergilian Copa 83
away from the hustle and bustle of Rome, we find ourselves standing at the
threshold of a country motel, populated by weary travellers seeking refresh-
ment and cheap thrills. As we peek behind by the curtains, we catch a glimpse
of the barmaid at work. She cuts an attractive figure, her hair neatly tied up
in a bandana and her voluptuous curves swaying sensuously as she moves.
Entertainment is her forte. A specialist erotic dancer, her mesmerising steps
and noisy castanets lure in punters from the roadside.
The Copa’s ‘smoky tavern’ ( fumosa . . . taberna, 3) is awash with sensory
enticements. There is an array of cheeses, fruits and nuts to satisfy the gour-
mand (17-22), perfumes and scents to excite the nostrils (bene olentia, 35), board
games to stimulate the mind and the touch (pone . . . talos, 37), and ‘supple girls’
offering kisses to randy strangers ( formosum tenerae decerpens ora puellae, 33).
Embedded within this rich sensorium is a dynamic soundscape. Music brings
the Copa’s world to life, from the pulsating rhythms of the ‘castanet’ (crotalo, 2)
to the piercing tones of the ‘reeds’ (calamos, 4) and the soothing melodies of
‘lyres’ (chordae, 7), the ‘pipe’ (tibia, 7) and the ‘pan-flute’ ( fistula, 8). In addition
to the five explicit references to musical instruments in the opening ten lines,
the tavern is pervaded by a polyphony of sounds from the natural landscape:
the tranquil hum of babbling rivers (crepitans rauco murmure rivus aquae, 12)
and the ‘incessant song’ of cicadas bursting through the trees (cantu crebro
rumpunt arbusta cicadae, 27).
This profusion of musical imagery is just one of many features which make
the Copa both an intriguing and perplexing subject of study.2 What inspired
a poet of the Roman Empire to write an elegy about a humble tavern and its
proprietor? And how would readers have responded to her depiction as a sexy
artiste? On the one hand, the portrayal of the dancing barmaid possesses an
intimacy and vibrancy that would seem to indicate a close familiarity with the
musical life of the ‘real’ Roman taberna. Many of the Copa’s modern readers,
to be sure, have sought to extrapolate from its descriptive details information
about the social and cultural practices of this elusive establishment. On the
other hand, the history of the taberna is (and was) never clear-cut. Our under-
standing is dependant largely upon the observations of a small clique of elite
Professor Timothy Moore, and all the participants at the conference for their invaluable com-
ments and suggestions.
2 The bibliography on the Copa is remarkably extensive. The main studies and editions
are helpfully listed by Tarrant 1992, 331, n. 1; further works to be included are Morelli 1912,
McCracken 1932, Zarri 1974, Franzoi 1988, Cutolo 1990, Kershaw 1992, Rosivach 1996, Grant
2001, Henderson 2002, Steures 2003, and Merkle 2005. I have adhered to the text of the Copa
edited by Clausen et al. 1966.
84 Morgan
authors who looked upon public leisure spaces with a marked ambivalence.3
As a focal Roman ‘pleasure zone’, the taberna was caught up in a myriad of
allusions, illusions, collusions and delusions, which, when combined, pro-
duce a distorting impression of what really went on within its walls. While
many looked down upon the tavern as the place where urban popular culture
expressed itself most viscerally—the stereotypical haunt of the plebs sordida,
to be avoided at all costs—others could imagine it as an idyllic haven where
one’s deepest fantasies were made real. Both of these conceptions, though
seemingly disparate, have a distinct role to play in configuring the Copa’s
sensory world.
Music and dance had been associated with the convivial pleasures of sex
and wine throughout antiquity, reaching back at least as far as the Homeric
epics.4 The Classical Athenian sculptor Lysippus was still being celebrated
in the first century AD for his depiction of a ‘drunken girl playing the pipes’
(temulenta tibicina), while a Greek dance known as the ὄρχησις παροίνιος
(‘dance to accompany wine’ or ‘drunken dance’), performed traditionally by
women at sympotic gatherings, is attested by both Athenaeus and Lucian.5
The sexual dynamics of musical performance found equally complex and
far-reaching expressions in Roman literary and material culture. If we are to
make sense of the Copa, we will therefore be required to look beyond the con-
fines of the tavern itself, to the dimly-lit dining-rooms of the private domus
and the bright lights of the public stage, where audiences rich and poor alike
were treated to the finest performers that the Empire had to offer. Like the
denizens of the taberna, these eye-catching entertainers (nearly always young
women) elicited from our elite authors a powerfully divergent set of responses.
In fact, as I will argue in this paper, the cultural intersectionality between the
tavern, the triclinium and the theatre in many ways holds the key to under-
standing the characterisation of both the Copa’s protagonist and her work-
place (tavern and tavern-keeper are not always easily distinguishable, as we
shall see). The enduringly popular theatrical genres of mime and pantomime,
which blurred the boundaries between reality and fiction, entertainment and
excess, will offer a particularly revealing frame of reference through which to
3 See especially Toner 1995, 73-77, and 2009, 109-10; Holleran 2012, 148-149.
4 cf., e.g., Od. 4.17-19, Il. 18.494-95.
5 Plin. HN 34.63; Athen. Deipn. 14.629f, Luc. De Salt. 34. An Attic krater from the fifth century
BC (British Museum E38; Barker 1984, fig. 15) depicts what appears to be a musical scene
from a symposium, in which a young man plays the auloi and an exotic-looking girl, wearing
an elaborate headdress and a skimpy outfit, dances while accompanying herself with the
crotala.

assess the relationship of the Copa to the erotic dancers depicted by contem-
porary (or near-contemporary) elegiac poets.
Music in the Copa: Interpretation and Reconstruction
The Copa first comes to our attention in the writings of two fourth-century
grammarians, Charisius and Servius, both of whom ascribe the poem to Vergil.
Servius locates the Copa more precisely in a list of Vergil’s juvenilia, along with
the other minor works now compiled in the Appendix.6 However, scholars have
long resisted the notion of Vergilian authorship, and have preferred instead to
regard the Copa as the work of an anonymous author who lived sometime in
the century and a half after the foundation of the Principate.7 Idiosyncrasies
of language, style and metre make it difficult to assign the poem to any specific
period. Richard Tarrant’s view that the Copa represents a product of the Flavian
or Antonine age seems to me the most reasonable suggestion; the works of
Martial, Statius and Juvenal will at any rate provide important comparanda at
various points in our investigation.8
Let us return to the tavern’s threshold, where we first meet our epony-
mous hostess advertising her wares. The woman’s name, we are told in line 1,
is Surisca, a Greek diminutive of the Latin feminine noun for ‘Syrian’ (Syra).9
Her foreign status is signified additionally by her ‘Greek headdress’ (Graeca . . .
mitella, another striking diminutive noun). She is a dancer (saltat, 3), and
a good one at that: ‘skilled at moving her quivering hips to the accompani-
ment of the castanet’ (crispum sub crotalo docta movere latus, 2), her perfor-
mance bears the mark of her ‘tipsy’ and ‘promiscuous’ demeanour (ebria . . .
lasciva, 3).10
6 Serv. Comm. in Aen. 1 praef., ed. Thilo 1878, 1: scripsit etiam septem sive octo libros hos: Cirin
Aetnam Culicem Priapeia Catalepton Epigrammata Copam Diras; Charis. Art. Gram. 1.63K,
ed. Barwick 1925, 79, l. 5: quamuis Vergilius librum suum Cupam [sic] inscripserit.
7 Israel Drabkin’s thesis in defence of Servius’ attribution, published in 1930, stands as a
final bastion against the communis opinio.
8 Tarrant 1992, 333.
9 As Goodyear (1977, 121) points out, Surisca may function as an ethnic rather than a nomi-
nal identifier, although the implication of Syrian nationality remains valid in either case.
For Syria’s association with taverns, cf. Lucil. fr. 123 Warmington (caupona hic tamen una
Syra), Mart. 5.70.2-3 (Syriscus | in sellariolis vagus popinis).
10 Henderson (2002, 260) notes the ‘clattering’ onomatopoeic effect produced by the accu-
mulation of cr and ct sounds in line 2.
86 Morgan
The copa’s actions in line 4 are frustratingly difficult to construe. H.R.

Fairclough in the Loeb edition renders the phrase ad cubitum raucos excutiens
calamos as “tapping against her elbows a noisy tambourine”, but this reading
is far from secure.11 If it is the copa herself who plays the crotalum in line 2—
as is surely the case—the notion of her performing on the ‘castanet’ and the
‘tambourine’ simultaneously presents obvious choreographic difficulties.
Wilamowitz posited a solution to the problem by interpreting calami as a syn-
onym for crotalum, meaning that the same instrument is to be inferred from
both lines, but this interpretation finds little support in the ancient sourc-
es.12 There is, however, another alternative. If we take ad not with its adjacent
word, cubitum, but with the final word of the line, calamos—thereby reading
the phrase as cubitum excutiens ad raucos calamos—we are able to envisage the
‘reeds’ being played by another musician as an accompaniment to the Copa’s
‘elbow-shaking’. Admittedly, rearranging the syntax in this way results in a
rather awkward dislocation of the preposition from its object (ad + calamos);
yet such a device was by no means anathema to Latin poets of the time.13 The
advantage is that the verb excutiens (‘shaking to and fro’; OLD s.v. ‘excutio’ 7b)
can thus be coordinated effectively with movere in line 2. In both instances the
phraseology specifies the energetic motion of a particular body part (elbow/
hip) in time with the music of a particular instrument (castanet/‘reeds’).
But what instrument exactly is denoted by the raucos calamos? Again,
the answer is not necessarily obvious. Textual and archaeological evidence
confirms the prevalence of reed-based materials in the manufacture of both
‘percussion’ and ‘woodwind’ instruments in antiquity.14 Fairclough’s sugges-
tion of ‘tambourine’ is seemingly substantiated by the recent discovery of
tambourine sticks from burial sites in the north-west of the Roman Empire,
which have been identified by the excavator as direct equivalents to the rauci
calami.15 However, the evidential basis for interpreting the calami as a type
of percussion is limited to say the least. It is far more likely, according to
11 Fairclough 2000, 439.

12 Wilamowitz-Moellendorff 1924, 2.311-315. All we have to go on is an entry in the Suda
(Bekker 629) describing the κρόταλον as a ‘split reed’ (σχιζόμενος κάλαμος).
13 Housman (1937, 105), in his edition of Manilius, gives a long list of examples in which
prepositions are ambiguously located; cf., e.g., Culex 205 (in); Manil. 1.245 (in); Hor. Carm.
4.1.19-20 (prope); Prop. 3.4.18 (subter); Tibull. 2.5.66 (ante); Lucr. 4.597 (per), 6.863 (prope);
Stat. Theb. 5.362-4 (super).
14 See West 1992, 82-86; Mathiesen 1999, 163 and 182-204.
15 Steures (2003, 213-14) oddly assumes that the tambourine stick “is the only real musical
instrument to be suggested so far”, while conceding that the instruments uncovered in the
excavation are actually made of wood rather than reed.

the majority of scholars, that we are dealing with a reed-pipe of some sort.
Housman surely went a step too far in imagining the instrument to be the
‘bagpipes’ (tibiae cum folliculo) cherished, notoriously, by the Emperor Nero.16
The more straightforward explanation is that the ‘noisy reeds’ are a poetic
substitute for the tibiae, the double pipes ubiquitous in Greco-Roman musical
culture and renowned for their distinctly piercing tone.17 We know that the
Copa’s tavern contains a tibia—the poet explicitly says so in line 7—and we
know, too, that calamus (in both singular and plural forms) was favoured as an
equivalent for tibia by a number of Augustan and post-Augustan writers.18
In fact, our sources emphasise the indispensability of the tibia in creating
the sexually-charged, alcohol-fuelled atmosphere characteristic of dance per-
formances in the Roman world. Horace’s Epistle 1.14 is a case in point. In a poem
contrasting the peace and quiet of the countryside with the heady sensuality
of the city, Horace imagines the scene of revelry inside a taberna. The life and
soul of the party is the meretrix tibicina, ‘the pipe-playing prostitute’, who calls
upon her listeners to leap about in time with her ‘noise’ (one is reminded of
Lysippus’ temulenta tibicina statue).19 Propertius’ Elegy 4.8, similarly, describes
a raunchy banquet-scene which features among its cast of entertainers a female
castanet-player (crotalistria) and a male tibicen.20 The two instrumentalists
are accompanied by a dancing dwarf, who ‘claps his stunted hands in time
to the boxwood pipes’ (iactabat truncas ad cava buxa manus, 4.8.42). The
Propertian epithet ad cava buxa is probably the model for the Copa’s (ad) rau-
cos calamos.21
Material culture gives a vivid impression of what these exuberant musical
cabarets actually looked like. A remarkable mosaic (Figure 1) discovered in the
early eighteenth century near Santa Sabina in Rome, dating from the second
16 Housman (1937, 105). Apart from Nero’s penchant for the instrument (cf. Suet. Ner. 54,
Dio Chrys. Or. 71.9), information about the form and function of the Roman bagpipes is
scarce.
17 On the raucous tone of the tibia, see Moore 2012, 53-54.
18 Cf. Catull. 63.22, Plin. HN 16.164, 16.169, Apul. Met. 11.9.6, Tert. Anim. 14; Goodyear 1977,
121-22.
19 Hor. Ep. 1.14.24-26: nec vicina subest vinum praebere taberna | quae possit tibi, nec meretrix
tibicina, cuius | ad strepitum salias terrae gravis. The word strepitum implies that she is
playing a percussion instrument, presumably the scabellum, rather than the crotala as
Mayer (1994, 209) proposes.
20 Prop. 4.8.39 (Miletus tibicen erat, crotalistria Byblis); cf. Macr. Sat. 3.14.4.
21 Prop. 4.8.41-42: Magnus et ipse suos breviter concretus in artus | iactabat truncas ad cava
buxa manus. Merkle (2005, 110-11) is the latest in a long line of scholars to draw a connec-
tion between the poems.
88 Morgan
Figure 1 Banqueting scene with dancers and musicians; mosaic from S. Sabina, Rome.
After Clarke 2003, 217, fig. 127.
century AD, shows two women in diaphanous dresses dancing to the accom-
paniment of tibicines with scabella (percussive clappers) attached to their feet.
The dancers sway their hips in seductive fashion while holding crotala in their
hands, much as the Copa herself is said to do in line 2. There is a large amphora
of wine pictured at the centre of the mosaic, served by a dwarf or a small
slave wearing a tunic. The scene bears a striking resemblance to a fresco from
the triclinium of the House of the Ephebe in Pompeii (Figure 2), in which
a female pipe-player is shown playing music for a pair of lovers. Behind the
couple there is a man carrying an amphora, and to the left of the painting we
see what appears to be a dancing dwarf or pygmy. A female figure stands to
the right of the lovers, stretching out her hands. According to some she repre-
sents a copa demanding money from her customers, but her movements could
just as convincingly be interpreted as those of a crotalistria providing rhyth-
mic support to the tibicina. The deterioration of the painting unfortunately
prevents us from making out what instrument (if any) she is holding. A male
tibicen, a pair of lovers, and an amphora of wine also figure prominently in a
fresco from the peristyle of the House of the Physician in Pompeii, where they
form part of a vibrant Nilotic scene showing pygmies fighting wild animals and
dining under a large canopy (Figure 3).
The opening lines of the Copa therefore conjure up a musical scene that
would have been recognisable, at least in essence, to a Roman audience versed
in the literary and visual arts. Our multi-talented hostess is the undisputed
star of the show. Supported by a (male?) tibicen, she executes a rhythmic,
probably up-tempo, number for the entertainment of her customers. But, as
John Henderson has underlined, her dance also prefigures and condenses the
copia, the copiousness, of the tavern itself.22 Not only does she procure wine,
22 Henderson 2005, 260-61: “As we inventory the dance, the sensuous body of writing shim-
mers—with figurality: it is not that the dancer presences her routine with prelusory

Figure 2 Pair of lovers accompanied by a tibicina; fresco from the triclinium of the House of
the Ephebe, Pompeii (I.7.10-12); in situ. After Varone 2001, 45, fig. 40.
Figure 3 Nilotic scene with pygmy couple, tibicen and amphora; fresco from the peristyle of
the House of the Physician, Pompeii (VIII.5.24); Naples Archaeological Museum,
inv. no. 113196. After Varone 2001, 29, fig. 24.
entertainment and sex for her clientele, but she herself embodies these sen-
sual pleasures through the physicality of her dancing—entertaining, intoxi-
cating and arousing in equal measure. Her performativity rubs off on the
world around her, as we discover when we finally get a glimpse of the taberna’s
promise of performative improvisation; her faceless preface figures the Image as the
dance of writing, the erotics of metaphor”. The term copa is etymologically related to
caupona (a synonym for taberna); cf. Priscian, Corpus Glossariorum Latinorum ii.209.6
Keil, on Lucil. fr. 123.
90 Morgan
interior. Having been promised wine and sex (both visually, ebria . . . lasciva,
and verbally, bibulo decubisse toro, 6), the first thing that the worn-out traveller
(defessum, 5) sees upon entry is a collection of ‘wine-jugs’ (cyathi), ornamental
furniture (topia et calybae . . . triclia) and musical instruments (tibia, chordae)
ordered in neat array. The tour continues at breathless pace, taking in all the
sounds, sights, tastes and smells that the Copa has at her disposal. Her tavern
is a “factory of sensuality” (as Henderson aptly puts it), “an oasis away from the
exhausting heat, dust, and sweat of brute mundane labour”, which caters for
all seasons (aestivo, 5 and 29; spring flowers, 13-16; autumnali . . . die, 18), for all
appetites (17-22), for party animals both human and asinine (asellus, 25-6).23
Music and Genre in the Copa: Country Ballad or Pop Anthem?
Since the publication of James McCracken’s article “The Originality of the

Copa” in 1932, the question of historical ‘realism’ has dominated scholarly
investigations of the poem, producing in turn some rather startling con-
clusions.24 In the 1960s, Wilhelmina Jashemski inferred from a study of the
archaeological remains of Pompeian tabernae that “it would not be surprising
to find a Syrian copa at Pompeii”.25 Westendorp Boerma took Jashemski’s argu-
ment a step further, suggesting on the evidence of the fresco from the House of
the Ephebe (discussed above) that the poet had actually “lived in Campania”
and “had seen with his own eyes a Syrian girl like ours dancing before a smoky
tavern”.26 Mark Grant, likewise, surmises from a survey of Pompeian graffiti
that the Copa was likely to have been written by a young man who regularly
23 Henderson 2005, 276.

24 McCracken (1932, 125) maintains that the Copa represents “a realistic depiction of some-
thing which the poet knew deeply”. His eagerness to find a modern parallel for almost
every detail in the poem has rightly come under scrutiny. In one of his more fanciful
analogies, the Copa Surisca is likened to the “strolling singers and mandolin players” of
the twentieth-century Italian trattoria, “who visit several inns at a single meal for the few
coppers given them by the diners”! Wilkinson (1965) attempts an equally imaginative,
yet arguably more justifiable, transposition of the poem “into terms of a modern pseudo-
Tudor road-house”.
25 Jashemski 1964, 347-8. Kleberg (1957, 117), likewise, cites the opening four lines of the
Copa as proof that “la musique et la danse faisaient partie des distractions offertes à la
clientele”.
26 Westendorp Boerma 1976, 660.

frequented taverns and was inspired by the lewd messages that he found scrib-
bled onto their walls.27
Of course, given the intractable problems of authorship and dating, the task
of contextualising the Copa is necessarily grounded in speculation. There is
undoubtedly much to be gained from adopting a methodological framework
which is sensitive to the poem’s cultural as well as literary influences; reductive
assessments to the effect that “the poet is not portraying reality, but the make-
believe world of his dreams” are clearly unhelpful and should be avoided.28 But
the concept of ‘historicity’ also has its limits. It goes without saying that we
know almost nothing about the musical life of the tavern beyond a few scraps
of literary evidence.29 Horace in his brief vignette of the tavern may hint indi-
rectly at the use of the tibia as a form of entertainment, but he gives no explicit
indication that the meretrix tibicina was actually a distinctive fixture of this
historical setting as opposed to a literary construct born out of elite cultural
stereotypes about popular leisure.30
The musical content of the Copa needs to be treated with the same degree
of caution. Even if we dismiss the fanciful idea that the poem was written with
a ‘Pompeian’ setting in mind, we are still left with the problematic assumption
that the poem represents “die Lebenswirklichkeit des Tanzes bis in die nie-
dersten Bereiche der Unterhaltung”.31 Whether or not ancient readers would
have been familiar with the social realities of the taberna (an issue that is hotly
contested in the scholarship), there is no reason to believe that they would
have identified such a large instrumental ensemble (crotalum, tibia, chordae
and fistula) as an axiomatic feature of this or indeed any other musical con-
text. We can accept easily enough the notion that the tibiae and chordae were
commonly used in combination to lend melodic direction to performances of
dance in the Roman world.32 But the additional inclusion of the fistula pres-
ents a considerable obstacle. The fistula is associated most prominently with
the quasi-mythical landscapes of Theocritus’ Idylls and Vergil’s Eclogues; hence
27 Grant 2001, esp. 131-32; for the kind of documents Grant is interested in, cf. CIL 4.8842,
9.2689, and 13.10018, 95.
28 As argued by Fairclough (2000, 375-6). Vincent Rosivach’s study, “The Sociology of the
Copa” (1996), represents something of a turning-point.
29 The relevant material is cited by Kleberg 1957, 117-18. For evidence of singing (and lyre-
playing) in the tavern, cf. Phil. Vit. Apol. 4.39, 4.42; Sidon. 8.11.3, vv. 49-54; note also Apul.
Met. 4.8.
30 See Toner 1995, 74.
31 Wille 1967, 199, making only passing reference to the Copa.
32 For the traditional combination of pipes and lyres in Roman musical ensembles, cf. Cic.
De Orat. 3.197; Dion. Hal. Rom. Ant. 1.33.4, 7.72.5; Quint. 1.10.20.
92 Morgan
the poet identifies it explicitly in line 10 as an instrument ‘of the shepherd’

(pastoris), played ‘in the rustic fashion’ (rustica . . . more).33 The pastoral
connotations of the fistula are further evinced by its setting ‘beneath the
Arcadian cave’ (Maenalio . . . sub antro, 9).34 This idyllic world of nymphs
(virgineo . . . Achelois, 15) and streams (rivus, 12; amne, 15) seems a far cry from
the gritty banality of the fumosa taberna which sets the scene in the open-
ing lines of the poem. Yet music and nature interact in tuneful harmony, the
raucos calamos of the tibia (4) echoing the rauco murmure of the stream (12).35
The allusion to the ‘virgin daughter of Achelous’ also carries with it a distinct
musical resonance. The association of the virgo and the fistula brings to mind
the tale of the beautiful nymph Syrinx, whose metamorphosis into a reed
gave the panpipes their name.36 One is reminded, too, of those nefarious crea-
tures the Sirens, begotten by the river-god Achelous himself, whose enchanting
singing lured unsuspecting mariners to their deaths. It is perhaps only fitting,
therefore, that the femme-fatale of our poem inflicts upon her listeners an
aurally-induced ‘death’ of their own (Mors aurem vellens ‘vivite’ ait, ‘venio’, 38).37
Noting these mythological allusions does not, of course, preclude the possi-
bility that the Copa’s readers actually did spend their downtime in the company
of scantily-clad dancers playing the castanets. There is, by way of comparison,
interesting evidence to suggest that the fistula played a role in the orchestral
accompaniment of the pantomime, which may be indicative of a wider usage
in Roman spectacle.38 But to interpret the character of the Copa as a product
or mirror of this ‘reality’ is to miss the essence of what her tavern is all about.
After all, how is it that the taberna can function in Horace’s worldview as the
33 The language of the Copa has particularly strong echoes of Vergil’s Second Eclogue: the
song of the shepherd Corydon contains an extended allusion to the fistula as an invention
of Pan (ll. 31-38); cf. also Ecl. 2.9, nunc viridis etiam occultant spineta lacertos, with Copa
28, nunc uaria in gelida sede lacerta latet, and Ecl. 2.12-13, cum raucis . . . sole sub ardenti
resonant arbusta cicadis, with Copa 27, nunc cantu crebro rumpunt arbusta cicadae.
34 Rosivach (1996, 609) elaborates on the significance of the Arcadian locus amoenus in rela-
tion to the Copa.
35 As Tarrant (1992, 337) observes, “down to the end of the Augustan period calamus or cal-
ami without qualification always refers specifically to pastoral song or poetry”.
36 Ov. Met. 1.689ff.; cf. Serv. Expl. in Verg. Buc. 2.32: . . . Syringam Numpham, quae fugiens eius
informitatem in calamum versa est seu fistulam et amorem suum cantu delectabat.
37 Cf. Ov. Met. 5.552; Serv. ad Virg. Georg. 1.9; Aen. 8.300. For the Sirens as daughters of
Achelous and Terpsichore, the Muse of Dancing, cf. Apoll. Rhod. Argon. 4.895-96. The
dancer with crotalum is a standard motif in funerary iconography: see Clarke 2003, 215-19.
38 Cf. Luc. Salt. 63, 68, 72; Arnob. ad Nat. 2.38; Apul. Met. 6.24. An imperial relief discussed by
Ciotti 1950 appears to represent a musician who played panpipes in a pantomime show.

prototypical centre of urban plebeian life while being elevated in the Copa to a
quasi-mythical pastoral paradise? Surely these two visions are irreconcilable?
Yet it was the very marginality of the tavern—its existence at the juncture
between public and private, political and social, lowbrow and highbrow—that
defined its place in the Roman literary imagination. Disavowed by many as
a den of corruption, drunkenness and debauchery, the tavern nevertheless
remained a constant source of fascination to those in the upper echelons of
society, who projected onto it their own fantasies and insecurities, hopes and
fears. This ambiguity of the taberna as both fumosa (2) and formosum (33), a
home for gods as well as mere mortals (Ceres . . . Amor . . . Bromius, 20; tuguri
custos, 3; Vestae, 26), made it not only a dangerous space, a space that needed
to be controlled and policed, but also a space to be fantasised about and
desired.39 The Copa activates both poles of the dichotomy. We never lose sight
of the distinctive ‘Otherness’ that makes our hostess a suspicious character—
a Siren, as it were—but we also eroticise and glamorize this ‘Otherness’ as a
metonymic symbol of the tavern’s exotic allure. Those who come to the Copa’s
establishment only to deplore its pleasures find themselves speedily evicted
from the premises: a pereat cui sunt prisca supercilia! (34)
Dancing like a Syrian Showgirl: The Copa’s Life beyond the Tavern
In a typically colourful anecdote from his Life of Nero, Suetonius describes

a bizarre ‘open-air’ spectacle devised for (or by?) the emperor as a means of
passing time whenever he set sail from Rome. As he stood on the deck of his
cruise-ship and looked out onto the surrounding vista, Nero was presented
with (what must been familiar to him as) a lavishly decorated stage-set. A row
of pubs (diversoriae tabernae), makeshift yet remarkably lifelike, stood at con-
venient stopping-points along the banks of the Tiber. Each one possessed its
very own copa, skilled temptresses whose sole responsibility was to seduce the
emperor as he sailed by with a cheeky wink here, a wolf-whistle there. They
were not real copae, of course; that would not be in keeping with the charade.
No, they were actors, and not just any actors: Nero had employed the most
39 Seneca (Vit. Beat. 7.3) associates wine and perfume (mero et unguento madentem) with
‘places that fear the aedile’ (loca aedilem metuentia); on the policing of the tavern, see
Robinson 1992, 135-8. For the tavern as a place of nostalgia and fantasy, see Purcell 1996,
202-3; Edwards 1997, 86.
94 Morgan
aristocratic ladies in the capital in the most degrading role he could dream up
for them.40
Suetonius’ anecdote includes the only other attestation of the copa in the
extant corpus of Latin literature. Certainly, the depiction of Nero’s matronae in
the act of soliciting sex (as the verb appelleret surely implies) conjures up an
image not dissimilar to that of the lasciva copa “offering herself for the passer-
by’s pleasure”.41 It is interesting, too, that the copae in Suetonius’ narrative are
linked with the prostitutes (scorta) and dancing-girls (ambubaiae) who, we
have just been told, were hired to serve as waitresses at Nero’s extravagant
dinner-parties in the Campus Martius and Circus Maximus.42 The association
of these three (exclusively female) professional groups—all actors in an elabo-
rate Neronian mise-en-scène—reflects in many ways the multifaceted nature
of the Copa’s own vocation: she is not only a dispenser of wine and food, but
also “an entertainer” who is “sexually available”.43
In fact, the Copa and the ambubaia have more in common than one might
think. Deriving their unusual name from the Syrian word for ‘pipe’ (Aramaic
abbub or abbuba), the ambubaiae titillated audiences across Rome with their
racy dance routines set to popular tunes from the East.44 Their starring role
at the Emperor’s parties says much about their desirability as both sex icons
and musical virtuosi. But it also says much about their notoriety as cultural
‘aliens’, who were held responsible by aristocratic moralists for importing the
‘effeminate’ (effeminata) musical styles popular at the time among Roman
audiences.45 Horace has the ambubaiae appear as mourners at the funeral of
the playboy musician Tigellius, in a cortege including such lowlifes as beggars,
40 Suet. Ner. 27: quotiens Ostiam Tiberi deflueret aut Baianum sinum praeternavigaret, dis-
positae per litora et ripas diversoriae tabernae parabantur insignes ganea et matronarum
institorio copas imitantium atque hinc inde hortantium ut appelleret.
41 Rosivach (1996, 612-613), noting that “[the copa] is, after all, running something akin to a
brothel”.
42 Suet. Ner. 27: cenitabatque nonnumquam et in publico, naumachia praeclusa vel Martio
campo vel Circo Maximo, inter scortorum totius urbis et ambubaiarum ministeria; cf. Suet.
Tib. 42.4 (nudis puellis ministrantibus cenaretur).
43 Purcell 1996, 204. On the cultural assimilation of tavern-keepers and erotic dancers, see
Wiseman 1998, 73.
44 Schol. Hor. Sat. 1.2.2: ambubaiae dicuntur mulieres tibicinae lingua Syrorum. Etenim Syris
tibia sive symphonia ambubaia dicitur. They are attested almost exclusively in the period
from Augustus to Hadrian; see Gowers 2012, 90-91 for more. Fear (1991, 79) regards the
Copa Surisca as “the most famous example” of an ambubaia.
45 Quint. Inst. Orat. 1.10.31; cf. Ov. Rem. Am. 751-5, Prop. 2.22.4-6.

drug-peddlers, mime-actresses and clowns.46 Petronius’ Trilamchio even uses

ambubaia as a derogatory term in mockery of his wife’s servile origins and
sexual promiscuity.47
By targeting the figure of the Syrian dancer-cum-prostitute, these authors
were tapping into deep-seated feelings of anxiety and resentment about Syrian
culture and its assimilation into Roman society.48 Juvenal famously bemoans
the fact that ‘the Syrian Orontes has for a long time now been polluting the
Tiber, bringing with it its language and customs, its slanting strings and its
pipe-players, its native drums too, and the girls ordered to offer themselves
for sale at the Circus’.49 It is generally assumed that Juvenal is referring here
to the presence of Syrian prostitutes in the brothels and taverns concentrated
around the Circus Maximus; no doubt this is partly the implication of the
phrase iussas prostare.50 But what about the Syrian girls who were to be found
inside the Circus itself? It was here, after all, that Nero would come to wit-
ness his ambubaiae perform. The travelling ludii (‘dancers’?) who appeared
at Augustus’ banquets also came ex circo.51 Juvenal alludes in another poem
to the appearance of puellae on the stage and, in a detail reminiscent of the
Copa, singles out their sexy gyrations and shimmies on the dance-floor as ‘a
stimulant to jaded desire and a prickly goad to the cock’ (inritamentum veneris
languentis et acres ramitis urticae).52 It is quite possible that these girls were
categorised more widely as ambubaiae: their hometown of Gades (the modern
Spanish town of Cádiz) was said to have been founded by Syrians.53
46 Hor. Sat. 2.1.1-3: ambubaiarum collegia, pharmacopolae, | mendici, mimae, balatrones, hoc
genus omne | maestum ac sollicitum est cantoris morte Tigelli. The Tigellius described by
Cicero (ad Fam. 7.24.2) as a ‘good pipe-player and a decent singer’ (bellum tibicinem . . . et
sat bonum cantorem) may be the same man, although there is some debate surrounding
this issue.
47 Pet. Sat. 74: “Quid enim?” inquit “ambubaia non meminit se de machina?”.
48 Isaac 2004, 336-37. For both Apuleius (Met. 8.24) and Lucian (Dea Syria 44), the crotala
were synonymous with the ecstatic rituals of the dea Syria cult. See also Naerebout (2007,
151-57) on responses to the Emperor Elagabalus as a ‘Syrian’ dancer.
49 Juv. Sat. 3.62-5: iam pridem Syrus in Tiberim defluxit Orontes | et linguam et mores et cum
tibicine chordas | obliquas nec non gentilia tympana secum | vexit et ad circum iussas pros-
tare puellas. | ite, quibus grata est picta lupa barbara mitra.
50 See, e.g., Braund 1996, 185.
51 cf. Suet. Aug. 74.
52 Juv. Sat. 11.162-64: forsitan expectes ut Gaditana canoro | incipiant prurire choro plausuque
probatae | ad terram tremulo descendant clune puellae; cf. Prop. 2.22.1-10.
53 A scholiast on Juv. Sat. 11.162 makes the Syrian connection explicit: id est, speras forsi-
tan, quod incipiant saltare delicatae ac pulchrae puellae Syriae, quoniam de Syris et Afris
Gades condita est. For the Roman identification of Carthaginians/Phoenicians as Syrians,
96 Morgan
By the time of Horace, the exotic attraction of the Gaditanae had become
almost proverbial.54 In Statius’ reimagining of a theatrical spectacle in the
Domitianic period, ‘the cymbals and jingling Gades’ (cymbala tinnulaeque
Gades) provide the soundtrack to the dancing of ‘Syrian troupes’ (agmina
Syrorum) and ‘buxom showgirls from Lydia’ (Lydiae tumentes); lest we be left
with any doubt about the sexual availability of these foreign starlets, we are
assured by the poet that their services are ‘easily bought’ ( faciles emi).55 The
razzle-dazzle of Flavian show-business left its mark on Martial, too. One of his
Epigrams is devoted to a slave-girl called Telethusa, who was ‘skilled at per-
forming lascivious gestures to Baetican castanets and dancing to tunes from
Gades’.56 We reencounter the (same?) dancer Telethusa on two occasions in
the Carmina Priapea, an anonymous anthology thought to have been com-
posed like the Copa sometime in the first or early second century AD.57 Here,
her bare buttocks (clunem tunica tegente nulla, 19.2) and sinuous thighs ( fluc-
tuante lumbo, 19.4) mark her out as one of the (in)famous girls making a living
in Rome’s red light district (nota Suburanas inter Telethusa puellas; 40.1).
There is another poem in the Priapea that commands our attention. It is
dedicated to a girl named Quintia, ‘the people’s darling’ (deliciae populi), who
had shot to fame as a dancer ‘in the great circus’ (magno . . . circo).58 In addi-
tion to the parallels evoked by the poem’s brevity and metre, Camillo Morelli
among others has noted how the formula praising Quintia’s skill in line 2, vibra-
tas docta movere nates, bears an uncanny resemblance to the second line of the
see Lucil. 15.540-1 (Syrophoenix); Juv. Sat. 8.159-61 (Syrophoenix . . . Syrophoenix); Stat. Silv.
4.5.29-48; Dio Chrys. Or. 33.41; Isaac 2004, 327-346.
54 cf. Hor. Od. 2.2.10-11, Juv. Sat. 10.1-2; Fear 1991.
55 Stat. Silv. 1.6.67-71: hic intrant faciles emi puellae, | hic agnoscitur omne quod theatris | aut
forma placet aut probatur arte. | hoc plaudunt grege Lydiae tumentes, | illic cymbala tinnu-
laeque Gades, | illic agmina confremunt Syrorum.
56 Mart. 6.71.1-2: edere lascivos ad Baetica crusmata gestus | et Gaditanis ludere docta modis;
cf. Mart. 14.203, 3.63.5, 5.78.26-28. A ‘dancer from Asia’ is the subject of a love poem attrib-
uted to the Augustan epigrammatist Automedon (Anth. Gr. 5.129); also notable is Anth.
Gr. 9.139.1-2, a late-antique epigram possibly modelled on the opening lines of the Copa:
Μαχλὰς ἐϋκροτάλοισιν ἀνευάζουσα χορείαις, | δίζυγα παλλομένοισι τινάγμασι χαλκὸν ἀράσσει,
‘the wanton, accompanying her dance with shrill shrieks and castanets, beats the brazen
clappers together with quivering movements’ (Loeb trans.).
57 Editions of the Priapea, with helpful translation and commentary, are available in various
languages: Goldberg 1992, Hooper 1999, Bianchi 2001, Callebat 2012.
58 Priapea 27, ed. Buecheler 1917, 142. For the identification with the Circus Maximus, see
Callebat 2012, 152-53.

Copa (and notably also to Martial’s phrase edere lascivos . . . gestus . . . docta).59

We get the same close-up of the dancer’s body, the same sensuous gestures,
even the same musical accompaniment: Quintia dedicates a set of castanets
(crotala, 2), as well as the cymbala and tympanum, as votive offerings to the
god Priapus, in the hope of guaranteeing her status as the crowd’s favourite
(ut semper placeat spectantibus orat, 5). The designation of these percussion
instruments as pruriginis arma, literally ‘the weapons of lasciviousness’, points
towards a further connection with the lasciva Copa, who is herself shown pay-
ing respect to Priapus and his ‘huge groin’ in lines 23-24 (est tuguri custos arma-
tus falce saligna | sed non et vasto est inguine terribilis).
One of the main platforms for erotic dancers like Quintia and Telethusa
was the mime, an enduringly popular theatrical genre throughout the repub-
lican and imperial periods with links to the ambubaiae (cf. Hor. Sat. 2.1.1-2,
above). Elaine Fantham and J.C. McKeown demonstrated several decades
ago the profound influence of the mime on the Augustan elegists.60 As well
as drawing attention to some important ancient testimonies for elegiac works
being adapted for performance on stage—Ovid himself proclaimed that ‘my
poems are also often danced before the people’ (Tr. 2.519-20)—Fantham and
McKeown identified an underlying commonality between elegiac narratives
and the plotlines of the so-called ‘Adultery-Mime’, which dramatized sexual
escapades in an often crude and outlandish fashion. Several writers, perhaps
significantly, ascribe to the mime the distinctive quality of lascivia.61 The mime
was also a highly musical art form. Cicero’s comment on the use of scabella to
signal the exitus mimi may hold the key to explaining the prominence of per-
cussion instruments in elegiac depictions of erotic dancers.62 The significance
of the Circus Maximus as the home of Quintia and Nero’s ambubaiae may also
59 Morelli (1912, 235) regards this as a parody of the Copa. Extant funerary inscriptions in
honour of female stage-performers often include a formula praising their artistic skill:
cf. CIL 3.10501 (artibus edocta), 6.10096 (docta erodita), 6.10127 (artis omnium erudita),
6.25808 (eruditae omnibus artibus), 9.3122 (quae me omnes artes docuit). Hemelrijk (1999,
83) has highlighted the importance of musical accomplishment and dancing as attri-
butes of the docta puella idealized by the Augustan poets: cf. Prop. 1.2.27-8, 1.3.42, 2.1.9-10,
2.3.17-20, 2.22.4-6; Ov. Am. 2.4.25-32, 2.11.31-2, AA 3.315-28, 3.349-52; Hor. Od. 3.9.10.
60 McKeown 1979 and Fantham 1989.
61 Cf. Ov. Fast. 5.331, Macr. Sat. 2.7.1, August. De civ. D. 6.7 (noting the role of Priapus in the
mime).
62 Cic. Cael. 65: mimi ergo exitus non fabulae: in quo cum clausula non invenitur, fugit aliquis e
manibus, dein scabilla [sic] concrepant, aulaeum tollitur. In the so-called ‘Charition Mime’
(P.Oxy. III 403), the sole surviving text of a Roman mime play preserved on an Egyptian
papyrus from the mid-second century AD, percussive ‘sound effects’ may have been
98 Morgan
reflect this venue’s association with the Ludi Florales, the annual springtime
festival held in Rome in honour of the goddess Flora. Her temple was located
in the immediate vicinity, and it was probably in the Circus that the festival’s
notorious female mime-artists gave their lewd performances.63
Closer examination of the mime reveals some intriguing points of connec-
tion with the Copa. For a start, the taberna seems to have been a common dra-
matic setting. Propertius’ Elegy 4.8, noted for being not only one of the Copa’s
prevailing models but also one of the works most strongly influenced by the
‘Adultery-Mime’, climaxes with two prostitutes escaping to a nearby tavern
(excipit obscurae prima taberna viae, 62; cf. 19).64 Furthermore, the Copa’s
headdress, the mitra, matches the one worn in Ovid’s Fasti (3.669) by Anna
Perenna, the goddess whose festival, like Flora’s, was renowned for its mimes.
The celebrated mimographer of the first century BC, D. Laberius, wrote a play
called Anna Perenna, of which the sole surviving fragment, conlabella osculum
(‘give me a kiss on the lips’), finds a notable parallel in the Copa’s line 33, formo-
sum tenerae decerpens ora puellae (‘tear with bites the beautiful lips of a supple
girl’).65 The title Surisca also calls to mind the name of Laberius’ great rival,
Publilius Syrus, ‘the Syrian’; Laberius himself is actually said to have dressed up
as a Syrian character in one of his productions.66
The mime’s ever-popular sister genre, the pantomime, may have less to tell
us, given that its dancers were more or less exclusively male.67 Yet the stage-
craft of the pantomimus accentuated the erotics and semiotics of the dance in
a uniquely conspicuous way, and for this reason it warrants brief consideration.
The two genres of mime and pantomime were, after all, closely related and as
such are often conflated in our sources. The pantomime was noted in particular
inserted into the drama at various points to accentuate moments of tension or climax;
Skulimowska 1966, 177.
63 Cf. Ov. Fast. 5.189-90; Wiseman 1998, 71 and 2002, 293. On Flora’s mimae, cf. Val. Max. 2.10.8,
Sen. Ep. 97.7, Mart. 1 Praef.; note also Lucr. 4.973-983.
64 McKeown 1979, 74-5. For similar ‘mimic references’ in the inn scene at Apul. Met. 1.7ff.,
see May 2006, 136-139. One of Ennius’ lost comedies, though not strictly a mime, bears the
title Caupuncula, ‘The Little Tavern-Keeper’ (Nonius 155.30, Warmington fr. 381).
65 Nonius 90.22 Mueller; cf. Gellius 16.7.10; Wiseman 1998, 72-3.
66 Macr. Sat. 2.7.4. On Pubilius Syrus, cf. Macr. Sat. 2.7.6-8. Characters in the comoedia pal-
liata, (and perhaps in the mime too) were often named Syrus or Syra: cf. Plaut. Bacch.
649-50, Merc. 808, Truc. 405-6; Ter. Ad. passim, Eun. 722, HT passim, Hec. 59; Pet. Sat. 52
(Syrum histrionem); Rosivach 1996, 606.
67 Our understanding of the pantomimus has been advanced particularly in the last ten
years following the publication of Lada-Richards 2007, Webb 2008, and Hall & Wyles
2008. Starks (2008) discusses possible evidence for female pantomimae.

for its use of large orchestral accompaniments, including the cithara, tibia,
fistula and scabella.68 Its mute soloist, moreover, attracted negative attention
for his overtly feminine style of performance, which many moralists equated
with the licentia of the genre as a whole.69 Quintilian, in a powerful exten-
sion of this idea, characterises the ‘effeminate’ (effeminatum) and ‘spineless’
(enervem) compositio of contemporary orators as the kind that ‘dances to the
very lascivious tunes of the castanet’ (lascivissimis syntonorum modis saltat).70
I would like to bring our discussion to a close by comparing one final lit-
erary passage, ostensibly inspired by the pantomime, in which the themes of
music, sexuality and stagecraft are closely intertwined. Book 10 of Apuleius’
Metamorphoses culminates in a narrative set-piece describing a dramatic pro-
duction of The Judgement of Paris in the theatre of Corinth. All the characters
in the production are played by dancers, with the role of Venus assigned to the
leader of the troupe:
There stood a mountain made of wood, crafted with exceptional work-

manship to resemble the famous Mount Ida of which the poet Homer
sang. It was planted with greenery and live trees, and from a hand-made
spring at its summit it spouted a stream of water. . . . After [Juno and
Minerva] another girl took to the stage, surpassingly beautiful in appear-
ance, her ambrosial complexion indicating that she was representing
Venus, as Venus looked when she was still a virgin (cum fuit virgo). She
flaunted her perfect figure: her body was naked and uncovered except
for a piece of sheer silk fabric which shaded her eye-catching private
parts. At one moment an inquisitive little breeze would waft this garment
aside in lascivious playfulness (lasciviens), so that it lifted to expose the
flower of her youth; at another moment it would gust exuberantly against
it, so that it clung tightly and graphically delineated her voluptuous
curves . . . Then, much to the delight of the spectators, Venus took her
position right in the middle of the stage. She smiled sweetly, exuding cha-
risma, and was surrounded by a crowd of the most cheerful little boys. . . .
Now a pipe with many holes played tuneful melodies in the Lydian mode
(tibiae multiforabiles cantus Lydios dulciter consonant); and while these
tunes were charmingly caressing the spectators’ hearts, Venus herself—
the most charming of all—started to move. Taking slow and gentle steps,
68 See note 38. On the musicality of the pantomime, cf. Macr. Sat. 2.7.18, Jerome Chron. Ol.
189.3, Lib. Or. 64.97, Cassiod. Var. 4.51.9; P.Flor. 16; Lada-Richards 2007, 41.
69 Lada-Richards 2007, 30; cf. Plin. Paneg. 46.4, Apul. Apol. 74.7.
70 Quint. Inst. 9.4.142.
100 Morgan
her spine undulating smoothly and her head nodding ever so slightly, she
began to walk forward, and to respond with subtle gestures to the soft
accompaniment of the pipe (mollique tibiarum sono delicatis respondere
gestibus). She gesticulated with her glances, sometimes gently languid,
sometimes sharply intimidating, and often she would dance with her
eyes alone (nonnunquam saltare solis oculis).71
The scene unfolds in familiar fashion, the interplay between music and dance,
sound and movement, shifting attention onto the sexualized female body.
More startlingly, the performance is narrated by Apuleius in such a way that
the realms of reality and fantasy are made to converge. The Homeric l andscape
of Mount Ida, replete with flora and fauna, functions as an elaborate stage-set
for a pantomimic spectacle in the heart of the city, but at the same time the
staging itself becomes Mount Ida—a ‘suspended reality’ within the ‘suspended
reality’ that is the Metamorphoses. The same is true of the dancer playing
the role of Venus: “no longer is a dancer dancing the role of a goddess so well
that she might be mistaken for her, but the goddess dances herself, in propria
persona”.72
The Copa exhibits its own metamorphic properties. Our shapeshifting nar-
rator constantly eludes our grasp, oscillating seamlessly between the autho-
rial voice and the persona of the Copa. So, too, our vision of the taberna is
filtered through the oppositional, yet complementary, lenses of its characters:
the passer-by, the customer, and of course the hostess herself. In this way, we
come to inhabit the Copa not just as readers but as a live audience, spectat-
ing from within the scaena like the vivae arbores on Apuleius’ Mount Ida. We
imagine the undulating figure of the dancer as though present before our very
eyes—the sense of immediacy heightened by the iteration of present-tense
verbs (sunt . . . est . . . est . . . sunt)—and we hear her music ringing through our
ears. Yet, try as we might to ‘own’ our ‘little Syrian’—to objectify and fetishize
her like Martial’s Telethusa or Juvenal’s pin-up girls, mere ‘playthings’ for the
well-to-do to enjoy as they please (nugas, Juv. Sat. 11.171; cf. Mart. Ep. 6.71.5-6)—
we encounter constant reminders of the scene’s artificiality, reminders of the
fact that this is all an elaborate mytho-literary façade constructed by and for
the titillation and gratification of elite Roman male readers.
By tuning into the Copa’s music, we also tune into the various levels of rep-
resentation and counter-representation which condition this ‘suspension of
disbelief’. To reduce the dancer’s performance to a historical re-enactment is
71 Apul. Met. 10.30-32.

72 May 2008, 350; also 353ff.

to deprive it of its vital significance as a literary creation, manufactured not so

much from lived experiences as from a complex synthesis of ideas, preconcep-
tions and misconceptions. More than simply a ‘dancer’, a ‘tavern-keeper’ or a
‘prostitute’, the Copa is a composite of all three, whose music takes on the poly-
valent cultural symbolism of both the taberna and the stage. Her cornucopia
of delights may be too good to be true, but it is precisely the thrill of imagining
ourselves among them, revelling in the sensory stimulation they provide, that
keeps us coming back to her country motel and clamouring for an encore.
References
Barker, A. 1984. Greek Musical Writings: I. The Musician and his Art (Cambridge).
Barwick, K. 1925 Flavii Sosipatri Charisii artis grammaticae libri V (Leipzig).
Bianchi, E. 2001. Carmina Priapea (Milan).
Braund, S.M. 1996. Juvenal: Satires Book 1 (Cambridge).
Buecheler, F. 1911. Petronii Saturae et Liber Priapeorum (Berlin).
Callebat, L. 2012. Priapées (Paris).
Ciotti, U. 1950. Rilievo romano e plutei medioevali ritrovati a Castel S. Elia, BA 35, 1-8.
Clarke, J.R. 2003. Art in the Lives of Ordinary Romans: Visual Representation and Non-
Elite Viewers in Italy, 100 B.C.-A.D. 315 (Berkeley, CA, and London).
Clausen, W.V., Goodyear, F.R.D, Kenney, E.J., Richmond, J.A. 1966. Appendix Vergiliana
(Oxford).
Cutolo, P. 1990. The Genre of the Copa, in: Cairns, F., Heath, M. (eds.) Papers of the Leeds
International Latin Seminar 6 (Leeds), 115-119.
Drabkin, I.E. 1930. The Copa: An Investigation of the Problem of Date and Authorship
with Notes on Some Passages of the Poem (Geneva and New York).
Edwards, C. 1997. Unspeakable Professions: Public Performance and Prostitution in
Ancient Rome, in: Hallett, J.P., Skinner, M.B. (eds.) Roman Sexualities (Princeton,
NJ), 66-95.
Fairclough, H.R. 2000. Aeneid VII-XII. Appendix Vergiliana. Revised by G.P. Goold
(Cambridge, MA).
Fantham, R.E. 1989. Mime: The Missing Link in Roman Literary History, CW 82, 153-163.
Fear, A.T. 1991. The Dancing Girls of Cadiz, G&R 38, 75-79.
Franzoi, A. 1988. Copa. L’ostessa: poemetto pseudo-uirgiliano. Introduzione, testo critico
e commento (Padova).
Goldberg, C. 1999. Carmina Priapea : Einleitung, Übersetzung, Interpretation und
Kommentar (Heidelberg).
Goodyear, F.R.D. 1977. The Copa: A Text and Commentary, BICS 24, 117-131.
Gowers, E. 2012. Horace: Satires Book 1 (Cambridge).
102 Morgan
Grant, M. 2001. The Copa. Poetry, Youth and the Roman Bar, PVS 24, 121-134.
Hall, E., Wyles, R. 2008. New Directions in Ancient Pantomime (Oxford).
Hemelrijk, E. 1999. Matrona Docta. Educated Women in the Roman Élite from Cornelia
to Julia Donna (London).
Henderson, J. 2002. Corny Copa: the Motel Muse, in: Spentzou, E., Fowler, D. (eds.)
Cultivating the Muse: Struggles for Power and Inspiration in Classical Literature
(Oxford), 253-301.
Holleran, C. 2012. Shopping in Ancient Rome: The Retail Trade in the Late Republic and
the Principate (Oxford).
Hooper, R.W. 1999. The Priapus Poems: Erotic Epigrams from Ancient Rome (Urbana).
Housman, A.E. 1937. M. Manilii Astronomicon. Liber Primus (Cambridge).
Isaac, B. 2004. The Invention of Racism in Classical Antiquity (Princeton, NJ).
Jashemski, W.F. 1964. A Pompeian Copa, CJ 59.8, 337-349.
Kershaw, A. 1992. Copa 5-6 Once More, CPh 87, 240-241.
Kleberg, T. 1957. Hôtels, Restaurants et Cabarets dans l’Antiquité Romain. Études histo-
riques et philologiques (Uppsala).
Lada-Richards, I. 2007. Silent Eloquence: Lucian and Pantomime Dancing (London).
Mathiesen, T.J. 1999. Apollo’s Lyre: Greek Music and Music Theory in Antiquity and the
Middle Ages (Lincoln, Neb.).
May, R. 2008. The Metamorphosis of Pantomime: Apuleius’ Judgement of Paris (Met.
10.30-34), in: Hall, E., Wyles, R. (eds.) New Directions in Ancient Pantomime (Oxford),
338-362.
May, R. 2006. Apuleius and Drama: The Ass on Stage (Oxford).
Mayer, R. 1994. Horace: Epistles Book 1 (Cambridge).
McCracken, G. 1932. The Originality of the Copa, CJ 28, 124-127.
McKeown, J.C. 1979. Augustan Elegy and Mime, PCPhS 25, 71-84.
Moore, T.J. 2012. Music in Roman Comedy (Cambridge).
Merkle, S. 2005. Copa docta, in: Holzberg, N. (ed.) Die “Appendix Vergiliana”: Pseude
pigraphen im literarischen Kontext (Tübingen), 91-115.
Morelli, C. 1912. Note sulla Copa, SIFC 19, 228-236.
Naerebout, F.G. 2007. Der Reich Tanzt . . . Dance in the Roman Empire and its Discontents,
in Hekster, O., Schmidt-Hofner, S., Witschel, C. (eds.) Ritual Dynamics and Religious
Change in the Roman Empire. Proceedings of the Eighth Workshop of the International
Network Impact of Empire (Heidelberg), 143-58.
Purcell, N. 1994. Women and Wine in Ancient Rome, in: McDonald, M. (ed.) Gender,
Drink and Drugs (Oxford), 191-208.
Robinson, O.F. 1992. Ancient Rome: City Planning and Administration (London).
Rosivach, V.J. 1996. The Sociology of the Copa, Latomus 55, 605-14.

Skulimowska, Z. 1966. Les instruments de musique dans le mime scénique grec en Egypte,
in: Bernhard, M.L. (ed.) Mélanges offerts à K. Michalowski (Warsaw), 175-179.
Starks, J.H., Jr. 2008. Pantomime Actresses in Latin Inscriptions, in: Hall, E., Wyles, R.
(eds.) New Directions in Ancient Pantomime (Oxford), 110-145.
Steures, D.C. 2003. Appendix Vergiliana Copa 4: Material Culture, Words and Sentence,
Mnemosyne 56, Fasc. 2, 213-217.
Tarrant, R.J. 1992. Nights at the Copa, HSPh 94, 331-347.
Thilo, G. 1878. Servii grammatici qui feruntur in Vergilii carmina commentarii (Leipzig).
Thomas, R.F. 1991. A Bibulous Couch ([Verg.] Copa 5-6)?, CP 86, 41-3.
Toner, J.P. 1995. Leisure and Ancient Rome (Cambridge).
Toner, J.P. 2009. Popular Culture in Ancient Rome (Cambridge).
Varone, A. 2001. Eroticism in Pompeii (Rome).
Webb, R. 2008. Demons and Dancers. Performers in Late Antiquity (Cambridge, MA, and
London).
West, M.L. 1992. Ancient Greek Music (Oxford).
Westendorp Boerma, R.E.H. 1976. The Copa Illustrated by Archaeology, in: Zadoks-
Josephus Jitta, A.N. Boersma, J.S. (eds.) Festoen opgedragen aan A. N. Zadoks-Josephus
Jitta bij haar zeventigste verjaardag (Groningen), 653-61.
Wilamowitz-Moellendorff, U. von. 1924. Hellenistische Dichtung (Berlin).
Wilkinson, L.P. 1965. Copa Today, G&R 12, 38-41.
Wille, G. 1967. Musica Romana. Die Bedeutung der Musik im Leben der Römer
(Amsterdam).
Wiseman, T.P. 1998. Roman Drama and Roman History (Exeter).
Wiseman, T.P. 2002. Ovid and the Stage, in: Herbert-Brown, G. (ed.) Ovid’s Fasti:
Historical Readings at its Bimillenium (Oxford), 275-99.
Zarri, G.P. 1974. Une étude quentinienne sur la tradition manuscrite de la Copa, RELO
1974.1, 1-16.
brill.com/grms
Ambrosian Hymns
Evidence for Roman Music of Late Antiquity?
Egert Pöhlmann
University of Erlangen
[email protected]
Abstract
After the abortive attempts of the bishop Hilarius of Poitiers, Ambrosius, bishop of
Milan, created with the metrum Ambrosianum the starting point for Latin Hymnody by
using a familiar pagan meter, the iambic dimeter, as the basic line. Combining four such
lines into a stanza he followed the type of the four-line stanzas of Horace. With eight
such stanzas he found a model for Christian hymnody for centuries. The text of four
of the innumerable Ambrosian hymns is attested for Ambrosius by Augustine. As the
ancient notation fell into disuse in the 6th century AD, the melodies of the Ambrosian
hymns were transmitted orally until the 10th century. They appear in the medieval
manuscripts with neumatic or alphabetic notation, but without rhythmical values and
adorned by rich melismata, which mirror the predilections of each monastic commu-
nity. Five of them are attributed to Ambrosius, from which this inquiry has to begin.
Keywords
Ambrosian Hymns – melismata – metre – aequalistic or quantitative rhythm –

notation – melodies.
Introduction
The last fragment of Ancient Greek Music, which is simultaneously the first
specimen of Christian Church Music, is the well-known Hymn to the Holy
* This paper was presented at the conference of MOISA on 30.07.2015 in Newcastle UK.

Ambrosian Hymns 105
Trinity, preserved on POXY 1786 (DAGM no. 59). It was written down on the
threshold between the third and the fourth century AD. Its metrical and musi-
cal peculiarities as well as its literary and structural properties can be fully
derived from pagan parallels of Ancient Greek Poetry and Music from the
2nd century BC to the 3rd century AD.1 We face in POXY 1786 the last exam-
ple of the type known as ʻcitharodic monodyʼ. The attempt of E.J. Wellesz to
understand it as the missing link between Syriac hymnody and early Christian
liturgic singing has been shown to be erroneous.2 Metrical Latin Hymnody
begins half a century later than the Hymn to the Holy Trinity, and its melodies
are transmitted, if at all, in manuscripts after 1000 AD. Its earliest specimens are
the melodies of the Ambrosian Hymns. In this paper I shall investigate the
question whether the Ambrosian Hymns can give evidence for Roman Music
of Late Antiquity.
1 The Metrics
After the discovery of the Psalmus responsorius3 it became evident that already
in the 3rd century AD Christian congregations were singing in rhythmic prose.
The Psalmus begins with a prayer (hypopsalma) in four lines. There follows
an ABECEDARIUS, of which 11 stanzas of different length are preserved. The
stanzas beginning with A, C-F and I have 7 lines, the stanza beginning with L
has 8 lines, the stanzas beginning with B, H and K have 9 lines, and the stanza
beginning with G has 11 lines. The number of syllables of the lines is variable.
Sometimes two lines are coupled by rhyme. The rhythm of the lines exhibits
the elements of the prosaic cursus.
But in the 4th century AD the traditionalism of the cultivated upper class
grasped also the Christian hymnody, forcing upon it inherited classical met-
rics. The first poet to write Latin metrical hymns (carmina), of which only
fragments are transmitted,4 was Hilarius, the bishop of Poitiers (315-367 AD),
according to the dictum of Isidore (Hymnorum carmine floruit primus5).
It is assumed that Hilarius was stimulated during his exile in Asia Minor to
1 West 1992b.
2 Wellesz 1945, see Pöhlmann 2016. For early Jewish and Christian liturgic singing see now
Stapert 2007, Franz 2013, Smith 2011, Leonhard 2014.
3 Roca-Puig 1965; Speyer 1989, p. 64-69; Herzog 1989, p. 220.
4 J.F. Gamurrini 1887, Feder 1916.
5 Isidore, De eccl. offic. 1,6.
106 Pöhlmann
compose Latin hymns following the example of oriental hymnody.6 This does
not hold good for the metres used: The three hymns which can be ascribed
with confidence to Hilarius use ancient Greek metres, which also found their
way to pagan Latin poetry. The first hymn (Ante saecula qui manes) is a muti-
lated ABECEDARIUS, 19 stanzas beginning with A-T. But Hilarius adopts for
it the fourth asclepiadeic stanza,7 which is used by Horace twelve times. The
second hymn (Fefellit saevam verbum) is again a mutilated ABECEDARIUS,
18 stanzas beginning with F-Z. Here Hilarius combines two iambic trimeters to
a stanza. The third hymn (Adae carnis gloriam) has 10 stanzas, each of which
consists of three trochaic tetrameters catalectic. Iambic trimeters and trochaic
tetrameters, with the names Senarius and Septenarius, are most familiar in the
poetry of the old Latin stage. Evidently the hymns of Hilarius were no success
in Poitiers, as Hieronymus remarks: Hilarius . . . in hymnorum carmine Gallos
indociles vocat.8 There are no melodies transmitted to these three hymns.
Hilarius of Poitiers had adopted for his hymns classical meters familiar to
him from his education in literature, which was obligatory for the upper class
in late antiquity. Ambrosius, bishop of Milan (339-397), found another way: by
combining four iambic dimeters to a stanza he created the so called Metrum
Ambrosianum. Eight such stanzas of four lines make up an Ambrosian hymn.
With the iambic dimeter Ambrosius adopted a metre which was flourish-
ing in his time. Horace used the iambic dimeter, if at all, only in epodic sys-
tems, thus following Archilochus. In Seneca the metre appears only once, in
a chorus, which consists wholly of iambic dimeters (Agamemno 759-774). The
iambic dimeter did not become really popular until the second century AD.
The last words of the emperor Hadrian (117-138) on his deathbed are a famil-
iar example: animula vagula blandula, / hospes comesque corporis, / quae nunc
abibis in loca / pallidula rigida nudula, / nec ut soles dabis iocos.9 Terentianus
Maurus (2nd-3rd century) and Diomedes (4th century AD) inform us that the
iambic dimeter was a favourite meter of the Poetae neoterici or Novelli poetae,
a group of modernist poets in imperial times.10 According to Terentianus a
6 Thus Schanz 1914, 227 without argument: „Den Anstoß zur Hymnendichtung erhielt
Hilarius sicher im Orient während seines Exils“.
7 Twice Glyconeus + Asclepiadeus.
8 Hieron. Comment. in Galatos II praef.
9 Historia Augusta, Spartian, Hadrian 25,9: et moriens quidem hos versus fecisse dicitur;
Büchner 1982, p. 169.
10 Terentianus Maurus GLK VI p. 383 ff., 2528; Diomedes GLK I p. 516, 25, Schanz Vol. III
(1959) 21-25.

Ambrosian Hymns 107
certain Alphius Avitus (2nd century AD) wrote several books (libros rerum
excellentium) about topics of Roman history in continuous iambic dimeters.
There are fragments transmitted from Books I and II.11 Another member of the
Novelli poetae was Septimius Serenus (2nd century AD), the author of Opuscula
ruralia in iambic dimeters, of which five fragments are transmitted.12 The iam-
bic dimeter could be used for occasional poetry also: Gellius tells us about
a friend who translated ex tempore an epigram of Plato (III Page) licentius
liberiusque, swelling it up to 17 iambic dimeters.13 A contemporary might be
a certain Marianus and his Lupercalia, from which five iambic dimeters are
transmitted.14 Ausonius (310-395) still uses the iambic dimeter for an epigram.15
In Greek poetry of imperial times however, the iambic dimeter is very rare.16
In Latin poetry of imperial times, the iambic dimeters show a tendency
towards isosyllabism. The poem of Hadrian about his fluttering animula with
its abounding resolutions is a special case, owing to its effortful pursuit of a
pictorial effect.17 The dactyl or anapaest very seldom appears instead of the
first or third iambus.18 Altogether, there prevails in these places the long anceps
with 82 % of 116 cases, by contrast with 18 % with short anceps.19 The pagan
iambic dimeters are used always in stichic series. There are no hints of the
formation of stanzas. According to the transmitted titles, the poems in iambic
dimeters were poetry for reading.
The treatment of the syllaba anceps in the iambic dimeters of pagan Latin
poetry is faithfully preserved by Ambrosius. In the first and the third iambus
of the hymns which are attested by Augustine for Ambrosius (see below p. 109
nr. 29-32), long anceps with 80 % of 256 cases prevails against 20 % with short
11 Terentianus Maurus GLK VI 2448: ut pridem Avitus Alfius libros poeta plusculos usus
dimetro perpeti conscribit excellentium; Marius Victorinus GKL VI p. 137, 31: apud nos metro
continuo Alphius Avitus libros rerum excellentium fecit.- Büchner 1982, 174 f.
12 Büchner 1982, pp. 175 ff.
13 Gellius 19,11; Büchner 1982, 171.
14 Büchner 1982, 175.
15 Ogygia me Bacchum vocat; XIII Epigrammata Nr. 12 Green.
16 West 1982, 165-167.
17 Historia Augusta, Hadrian 1: animula vagula; 4: pallidula rigida.
18 E.g. Incerti Odarium, Büchner 171, 6: cucurrit ad labeas mihi. 17: ad puerum <ut> intus
viverem; Alphius Avitus, Büchner 175, 4 f.: Exteraque muri ducere / spatiando paulatim
trahit.
19 Seneca: 25 against 7 cases; Incerto odarium: 28 against 4 cases; Alphius: 20 against 2 cases;
Marianus: 10 cases of long anceps; Serenus: 9 against one case; altogether: 95 against
21 cases.
108 Pöhlmann
anceps.20 We very seldom find the anapaest instead of the spondee.21 Thus, the
metrical technique of Ambrosius in his hymns is as good as identical with the
metrics of pagan poets in poems of iambic dimeters, with the sole exception
of the introduction of strophes of four lines, as we shall see immediately. Thus,
the Ambrosian line should be scanned like the pagan iambic dimeters, namely
as x – ᴗ – / x – ᴗ –. Following Martin West,22 we can transcribe this into note
values by representing the metrical short by a quaver, the metrical long by a
crotchet and the anceps in the first and third iambus by the head of a note only.
This means that the Ambrosian line is still a metre for reading. To transform it
into an effective metre for song the preponderance of the long ancipitia should
have been reduced, as the rhythmical vagueness of the ancipitia, which is eas-
ily tolerable in spoken poetry, should give way to a clear cut rhythm in song.23
The beginning of the Christian Hymnody is accompanied with new features,
which were aptly described by Bruno Stäblein: “As soon as sung Latin hymns
appear, two peculiarities appear. The first is the Iso-strophism, which means that
all stanzas have the same structure. The second is the Iso-syllabism, which means
that the number of syllables of every line as a rule is the same.”24
The use of the iambic dimeters by Ambrosius confirms this statement. Four
iambic dimeters, which use the metrical technique of the pagan model, are
strictly coupled to stanzas of four lines, which however do not appear in the texts
or in their metrics, but are evident from the transmitted melodies and the pre-
sentation of the texts in the manuscripts. With the restriction to stanzas of
four lines Ambrosius follows a standard of Horace, the so called Lex Meineke,
which claims that in all odes of Horace the respective sum of lines is always a
multiple of four.25
Evidently, the reason for the striking success of the Hymnus Ambrosianus
was not only the choice of a pagan metre popular in the time of Ambrosius, the
iambic dimeter. No less significant was the fact that Ambrosius couples iambic
dimeters to stanzas of four lines which made the repetition of a fixed melody
20 Intende qui regis: 52 against 12 cases; Aeterne rerum: 54 against 10 cases; Deus creator:
43 against 21 cases; Iam surgit: 58 against 6 cases; altogether: 207 against 49 cases.
21 Intende qui regis: verse 1: qui regis; verse 5: thalamo; geminae.
22 West 1992a, 137 f.
23 West 1992a, 137.
24 Stäblein 1957, 995: „Sofort mit dem Auftreten gesungener lateinischer Hymnentexte
sind zwei Merkmale, die bis heute Geltung behielten, gegeben: 1. Der Iso-Strophismus,
d.h. alle Strophen sind gleich gebaut und können somit auf dieselbe Melodie gesungen
werden . . . 2. Der Iso-Syllabismus, d.h. innerhalb der Strophen ist die Zahl der Silben fest
oder fast fest gegeben.“
25 Heinze 1917, 8 f. (except IV 8); see IV 8 and Heinze on 17 und 34.

Ambrosian Hymns 109
possible. Thus, he misused a Latin spoken metre for sung lyrics, following the
classic models of the four-line-stanzas of aeolic lyric, the asclepiadeic, alcaeic
and sapphic strophes, which were familiar to every cultivated Roman in their
transformation by Horace.
The next poet of Latin hymns after Ambrosius, Prudentius (died after
407) adopted the Metrum Ambrosianum, which spread quickly and became
canonical. Two thirds of all transmitted texts of hymns are Ambrosiani.26
The experiments of Prudentius with the fourth asclepiadeic or the sapphic
strophe (Cathemerinon 5 and 8) were never repeated. Thus, the Hymnus
Ambrosianus, having superseded all other attempts, eventually prefigured the
conception of the hymnus of the Christian congregation.27 Bruno Stäblein has
brought together much more than a thousand texts of hymns in the Metrum
Ambrosianum.28 Four of these are attested for Ambrosius in quotations by
Augustine, namely Deus creator omnium,29 Aeterne rerum conditor,30 Iam sur-
git hora tertia31 and Intende qui regis Israel, which is the first stanza of Veni
redemptor gentium, widely known in Martin Luther’s translation: Nun komm,
der Heiden Heiland.32 The authorship of Ambrosius for fourteen further hymns
is likely.33
2 The Melodies
It is much more difficult to assess the authorship of Ambrosius for his

melodies,34 the manuscript transmission of which begins no earlier than
1000 AD. Evidently the transmission of Ambrosian melodies from 38635
to 1000 AD was exclusively oral. This is puzzling, as the system of the Greek
26 Stäblein 1957, 996.

27 For extensive commentaries on Deus creator omnium, Aeterne rerum conditor, Splendor
paternae gloriae and Iam surgit hora tertia see Franz 1994, on Intende qui regis Israel,
Inluminans altissimus and Hic est dies verus Dei see Zerfass 2008.
28 Stäblein 1956, 663-679.
29 Confessiones 9,12.
30 Retract. 1,21.
31 De natura et gratia c. Pelagianos c. 63.
32 Sermo 372.
33 Schanz 1914, 231 f., Franz 1994, p. 17-25.
34 See Dreves 1893.
35 Ambrosian Hymns were sung in S. Ambrogio in Milan at 386, if not earlier (Schanz 1914,
229f.).
110 Pöhlmann
notation was known in Rome centuries earlier.36 The last witness is the learned
Boethius (480/85-524/26), who, following Gaudentius,37 explains erroneously
the invention of the notation by the veteres (παλαιοί) by their aim to replace the
full names of strings of the monochord (integra nomina, ὀνόματα) by signs (not-
ulae, σημεῖα) in order to save time and space (propter compendium scriptionis).
But from his anxious assertion that composers sometimes added the notulae
to their poems in order to preserve the melodies for posterity, it transpires that
the use of the Greek notation in musical practice was no longer familiar to the
readers of Boethius: ita miro modo repperientes, ut non tantum carminum verba,
quae litteris explicarentur, sed melos quoque ipsum, quod his notulis signaretur,
in memoriam posteritatemque duraret.38 A century later the Greek notation
was forgotten: Isidore of Seville (560-636) thinks that music is transmitted only
by memory: nisi enim ab homine memoria teneantur soni, pereunt, quia scribi
non possunt (Origines 3,15,2). Evidently the Greek notation ceased to be used
by practising musicians by about the fourth century.39
Investigation of the authorship of Ambrosian melodies finds some help in
the age, the provenience and the affiliation of manuscripts. The oldest manu-
script with readable musical notation is a hymnal in Kempten (before 1026 AD),
which offers 16 melodies with neumes without lines and 22 melodies with
alphabetic notation, of which 11 melodies have neumes without lines in
addition.40 Not the oldest, but the most important manuscript of hymns is
no. 347 of the Bibliotheca Trivulziana, which contains the stock of Milanese
hymnody in the 14th century with neumes on lines. From this hymnal
Bruno Stäblein has singled out an oldest layer of melodies,41 namely melody
1 (Aeterne rerum conditor), 3 (Splendor paternae gloriae), 6 (Iam surgit hora
tertia), 8 (Deus creator omnium) und 14 (Intende qui regis Israel). The texts of
these Hymns (with the exception of no. 3) are attested for Ambrosius also.42
The bold numbers of melodies are the numbers of Stäblein’s Monumenta
Monodica Medii Aevi (1956).
36 See West 1992a, 272 f. The first witness is Varro (fr. 282 p. 305 Funaioli). Quintilian gives
evidence for singing from scores (Inst. 1.12,14). Tropica (Tables of keys) are mentioned by
Marius Victorinus (GLAT VI 183, 23).
37 Boethius, De institutione musica IV 3; 15 f., Gaudentius 20.
38 Boethius, De instutione musica IV 3, 309 Friedlein.
39 West 1992a, 273.
40 Stäblein 1956 VIII and 578 f.
41 Stäblein 1956, 503 f.
42 See above p. 109, nn. 29-32; p. 110, n. 41 f.

Ambrosian Hymns 111
The problems of the transmission of the Ambrosian hymns can be easily

demonstrated by melody 14. This melody is transmitted in the aforesaid ver-
sion of Milan for the first stanza of Intende qui regis Israel. There follow the
stanzas 1-7 of Veni redemptor gentium. The same version in a transposition from
D to G, is transmitted in a Cistercian hymnal also.43 Melody 14 is transmitted
also in a hymnal of Nevers,44 which nevertheless omits the first stanza, Intende
qui regis Israel. Instead, after stanzas 1-7 of Veni redemptor gentium, there fol-
lows as an 8th stanza the doxology.45
The version of Veni redemptor with doxology became canonical, whilst the
original version of the beginning was forgotten. Bruno Stäblein communicates
six further melodies for the later version of Veni redemptor with doxology.46 Of
these six the melody 5031 from Klosterneuburg, and its predecessor, melody
5032 from Einsiedeln, became the model for Martin Luther’s Nun komm, der
Heiden Heiland.47 For text and melody of Ambrosius of Veni redemptor how-
ever the melody 141 of Intende qui regis Israel in the Hymnal of Milan remains
the best witness. Therefore, we shall use it as an example for the problems of
the evaluation of the melodies transmitted in manuscripts after 1000 AD.
By comparing the different realisations of the melody 14 in six manucripts
there appear some variants which are listed in Stäbleinʼs Kritischer Bericht.48
The melody is transmitted in the D-mode and the G-mode. A special problem
consists in the distribution of the melismata in the manuscripts after 1000 AD
(see fig. 1). Comparing the six realisations of the melody 14,49 it appears that
1) the melismata in a given manuscript are not set in the same pattern in the
43 Melody 142: Cistercian Hymnal, Heiligenkreuz Stiftsbibliothek 20, 12th /13th century,
Stäblein 1956, 30; 515. For the oldest source of melody 14 (not in Stäblein) see Zerfass 2008,
p. 138-140.
44 Melodie 143: Hymnal of Nevers, Paris BN nouv. acquis.lat. 1236, 12th century, Stäblein 1956,
81; 540.
45 Gloria tibi, Domine,/Qui natus es de virgine,/ Cum Patre et Sancto Spiritu,/ In sempiterna
saecula.
46 Melody 719, Hymnal of Gaeta, Rom, Bibl. Casanatense 1574, 12th century.; Melody
406, Hymnal of Worcester, Cath. Libr. F 160, 13th century.; Melody 5031, Hymnal of
Klosterneuburg, Stiftsbibliothek 1000, 1336 AD; Melody 5032, Hymnal of Einsiedeln,
Stiftsbibliothek 366, 12th century.; Melody 597, Hymnal of Dürnstein, St. Florian,
Stiftsbibliothek XI 407, 15th century.; Melody 703, Hymnal of Verona, Bibl. Cap. CIX (102),
11th century.
47 Encheiridion geistlicher Gesänge, Erfurt 1524; Stäblein 1956, 568, Zerfass 2008 p. 139-147.
48 Stäblein 1956, 501-624, esp. 507.
49 Intende qui regis: Stäblein 1956, 8, 30, 81; Sic ter quaterni: Stäblein 1956, 188; A patre, unige-
nite: Stäblein 1956, 221; Ut nox tenebris: Stäblein 1956, 384.
112 Pöhlmann
Nr.1: Aeterne rerum conditor

Syllable 1 2 3 4 5 6 7 8 sum
Milan (14. AD) 1 1 3 5
Cistercian (13. AD) 3 3 6
Klosterneuburg (1336) 3 3
Verona (11. AD) 1 3 4 8
metre x - ᴗ - x - ᴗ -
Nr. 3: Splendor paternae gloriae
Syllable 1 2 3 4 5 6 7 8 sum
Milan (14. AD) 1 1 1 1 4
Cistercian (13. AD) 1 1 1 2 5
Nevers (12. AD) 1 1
Klosterneuburg (1336) 1 1
Einsiedeln (12. AD) 2 1 1 4
Verona (11. AD) 1 1 2
Gaeta (12. AD) 1 1 2
Nr. 6: Iam surgit hora tertia
Syllable 1 2 3 4 5 6 7 8 sum
Milan (14. AD) 2 1 1 1 5
Cistercian (13. AD) 2 2 1 5
Nr. 8: Deus, creator omnium
Syllable 1 2 3 4 5 6 7 8 sum
Milan (14. AD) 1 1
Cistercian (13. AD) 1 1 1 1 4
Nevers (12. AD) 1 1
Worcester (13. AD) 1 3 2 1 1 8
Klosterneuburg (1336) 2 2 1 1 6
Verona (11. AD) 2 2 1 1 6
Nr. 14: Intende, qui regis Israel
Syllable 1 2 3 4 5 6 7 8 sum
Milan (14. AD) 2 1 1 2 1 7
Cistercian (13. AD) 1 1 1 2 2 1 8
Nevers (12. AD) 1 1 2 1 5
Worcester (13. AD) 1 2 1 2 2 1 9
Klosterneuburg (1336) 1 1 1 1 2 1 7
Verona (11. AD) 1 1 2 2 1 7
figure 1 Distribution of melismata on the eight syllables of the metrum Abrosianum.

Ambrosian Hymns 113
four lines of the stanza, but in different places in each of the four lines, 2) that
in different manuscripts the melismata mostly are not set in the same pattern,
and 3) that the melismata are set indiscriminately on each of the 8 syllables of
the iambic dimeter, without any preponderance of the anceps syllables or the
long syllables of the text. Comparing the use of melismata in the other hymns,
of which text and melody are attested for Ambrosius also, namely nr. 1 (Aeterne
rerum conditor), 3 (Splendor paternae gloriae), 6 (Iam surgit hora tertia) and
8 (Deus creator omnium), there appears the same use of the melismata in the
second dimeters, while the first dimeters tend to be plain syllabic melodies.
All in all, it appears that the melismata disagree with the metrics of the text.
Therefore, they do not represent old tradition, but belong to different aequal-
istic realisations according to the predilections of the different monastic con-
gregations, as represented in the respective manuscripts after 1000 AD.50 When
trying to reconstruct the original melodies of Ambrosius, the melismata have
to be stripped off.
3 The Rhythm
In accordance with the usus of the manuscripts and the usus of the Roman
church, Bruno Stäblein transcribes the melodies in the Monumenta in
an aequalistic version. This means that every note has the same value.51
figure 2 Monumenta Monodica I.14.
50 See Stäblein 1956, 503, 507, 514).

51 Stäblein 1959 XVI and n. 24.
114 Pöhlmann
We present therefore the melody nr. 14 (Intende qui regis Israel) first in Stäblein’s
aequalistic version with all melismata.
When the melismata are stripped off, there remain still several metrical
anomalies, which accumulate in the first stanza, where Ambrosius transplants
nearly word by word psalm 80, 2-3 in the version of the vulgata into iambic
dimeters.52 This can be seen by comparing stanza 1 and 2:
1. In – ten – de qui re–gis Is – ra – el,

2. Ve - ni, re - demp – tor gen – ti – um,
1. su – per Che - ru – bim qui se - des,

2. os – ten – de par – tum vir - gi - nis;
1. ap – pa–re . . . E - phrem co–ram, ex – ci – ta
2. mi – re–tur om – ne sae - cu –lum,
1. po – ten – ti – am tu – am et ve – ni.

2. ta – lis de –cet par - tus de – um.
figure 3 Metrical peculiarities in Nr. 14 stanza 1 and 2.
We find anceps resolved into anapaest (regis Israel) and scriptio plena of short
syllable (appare Ephrem; coram, excita; tuam et), which is to be replaced by eli-
sion. The following stanzas are flawless, except stanza 5 (anapaests at thalamo;
geminae), where Ambrosius again paraphrases a psalm (19,6).53 Use of previ-
ously coined material in quantitative poetry can easily raise metrical problems.
It is interesting also to see that Ambrosius is not intent on strict responsion
of the metrics of the respective stanzas, but uses the licenses in the treatment of
the anceps arbitrarily. Thus, we find opposed in the anceps-position Intende
/ veni; super / ostende; coram / omne; potentiam / talis; tuam / partus (short
anceps in bold in Fig. 3). This means that the rhythmic shape of a quantita-
tive realisation of the iambic dimeter would oscillate not only from anceps to
anceps, but morover from stanza to stanza. This vagueness is, as we have seen
(see above p. 108), tolerable in spoken poetry, but not for the chanting of great
congregations.
52 Psalm 80(79) 2-3 : qui regis Israhel intende/ . . . qui sedes super cherubin manifestare/coram
Effraim . . ./excita potentiam tuam.
53 Psalm 19(18) 6: procedens de thalamo suo/exultavit ut gigans.

Ambrosian Hymns 115
When comparing stanza 1 and 2 of nr. 14, it appears, finally, that Ambrosius
did not try to reconcile the word accent with a posssible ictus of the verse. As a
trivial consequence of the laws of the Latin word accent there is perfect coinci-
dence when the line is divided into a trisyllable, a disyllable and a trisyllable. If
not, ictus and word accent disagree. This demonstrates that Ambrosius wanted
to write quantitative lyrics, falling back on a model of the Roman classic, the
ʻStrophic Formʼ with which he was familiar from Horace. By inventing melo-
dies which were repeated for every strophe and might be transferred to other
texts also Ambrosius made it easy to learn his hymns by heart.
Already before 386 Ambrosius had introduced his hymns into the liturgy
in Milan.54 They were learned and sung by great multitudes, the populus,
as Ambrosius himself attests.55 When Iustina, the mother of the emperor
Valentinian, took sides in 386 with the Arians against Ambrosius, the bishop
together with the Athanasians occupied the basilica, where the congregation
held out singing day and night the hymns of Ambrosius, as Augustine attests.56
There are no clues to the origin of the melodies of the Hymni Ambrosiani.
Augustine attests only that the usus of singing psalms and hymns in Milan
before 387 followed the example of the Greek East of the Roman empire.57 The
melodies were evidently an invention of Ambrosius himself. But in what way
were his hymns sung?
For sung Hymni Ambrosiani Bruno Stäblein does not claim an aequalis-
tic realisation as the only possibility. He also recommends singing in a 4/4-
measure with eight syllables of equal length, or in a 3/4-measure.58 As 80%
of the ancipitia are long (see above p. 107 f.), the first proposal would lead to
clashes of metrics and rhythm of 20% in the case of the ancipitia, and 100% in
the case of the short syllables of the text. The second proposal would produce
clashes of metrics and rhythm of 80% only in the case of the ancipitia, which is
less disturbing. But did the congregations of Ambrosius still preserve a natural
feeling of the quantities of the spoken Latin?
54 Paulinus vita Ambrosii 13 hoc in tempore (Easter 386) primum antiphonae, hymni ac vigiliae
in ecclesia Mediolanensi celebrari coeperunt. See Leonhard 2014.
55 Ambrosius, Sermo contra Auxentium (386 AD) 34: Hymnorum quoque meorum carmini-
bus deceptum populum ferunt. Plane nec hoc abnuo . . . Quid enim potentius quam confes-
sio trinitatis, quae cotidie totius populi ore celebratur? Certatim omnes student fidem fateri;
Patrem et Filium et Spiritum Sanctum norunt versibus praedicare.
56 Augustine, Confessiones 9,7 (25th April 387): excubabat pia plebs in ecclesia mori parata
cum episcopo suo . . . tunc hymni et psalmi ut canerentur secundum morem orientalium par-
tium, ne populus maeroris taedio contabesceret, institutum est.ex illo in hodiernum reten-
tum multis iam ac paene omnibus gregibus tuis et et per cetera orbis imitantibus.
57 See n. 53.
58 Stäblein 1956, XVI and n. 24.
116 Pöhlmann
Paul Klopsch has pointed to the known fact that Latin (like Greek) under-
went from the 1st to the 4th century AD a shift from a language with natural
quantities of syllables and musical accent to a language with stress accent at
the expense of the quantities of syllables.59 In the 4th century the educated
classes could internalize the model of the classical quantitative Roman poetry
by extensive reading. But the populus no longer had any feeling for the quanti-
ties of Latin. Augustine, in his De Musica (III 3,5), which was begun after 387
in Milan, gives a lively picture of this situation: A pupil declares that he has no
idea of long and short syllables, a difference which only grammarians preserve
(syllabarum longarum et brevium cognitionem me non habere, quod a grammat-
icis traditur). He does not deny that he can hear rhythmical patterns, but does
not know which syllable must be lengthened or shortened, which can only be
gathered from examples (iudicium aurium ad temporum momenta moderanda
me posse habere non nego; quae vero syllaba producenda vel corripienda sit, quod
in auctoritate situm est, omnino nescio).60 Indeed many grammatical treatises
give examples and rules for the quantities of Latin words in poetry, the first of
which is Servius (4th century AD).61
Like the pupil of Augustine, the congregations of Ambrosius evidently had
no sense of the natural quantities of Latin. Therefore we might choose a reali-
sation of the five genuine Ambrosian melodies (nr. 1, 3, 6, 8, 14) in 3/4 measure
without any melismata (see Appendix), thus saving the iambic character of
the hymnus which Ambrosius evidently intended. Of course this holds good
also for the other presumable Ambrosian hymns and melodies, which are a
new start from Roman roots, superseding all other attempts by its simplicity
and coining the conception of the hymnus of the Western Christian congrega-
tion for the centuries to come.
Bibliography
Büchner, K. (ed.). 1982, Fragmenta Poetarum Latinorum, Leipzig.

Dreves, G.M. 1893, Aurelius Ambrosius, der Vater des Kirchengesangs, Freiburg
(= Ergänzungsheft zu den Stimmen aus Maria Laach Nr. 58).
Feder, A., (ed.). 1916, S. Hilari episcopi Pictaviensis opera, pars quarta, Vienna/Leipzig
(CSEL 65).
59 Klopsch 1972, 1-3.

60 Klopsch 1972, 4.
61 Klopsch 1972, 61-63.

Ambrosian Hymns 117
Franz, A. 1994, Tageslauf und Heilsgeschichte. Untersuchungen zum literarischen Text

und liturgischen Kontext der Tageszeithymnen des Ambrosius von Mailand, S. Ottilien.
Franz, A. 2013, art. Musik II (Vokalmusik), C. Jüdisch, D. Christlich, in: RAC 25, 264-284.
Gamurrini, J.F. 1887, Sancti Hilarii tractatus de mysteriis et hymni. Rome.
Green, R. 1991, The Works of Ausonius, Oxford.
Heinze (ed.), R. 1917, Q. Horatius Flaccus, Oden und Epoden, Berlin.
Herzog, R. (ed.). 1989, Handbuch der lateinischen Literatur der Antike 5, Restauration
und Erneuerung 284-374 n. Chr., München (HLL 5).
Klopsch, P. 1972, Einführung in die Mittellateinische Verslehre, Darmstadt.
Kurz, G. 2003, Intende qui regis Israel, der Weihnachtshymnus des Bischofs Ambrosius
von Mailand (hy. 5), in Jahrbuch für Hymnologie und Liturgik 43, 105-161.
Leonhard, C. 2014, Which Hymns were sung in ancient Christian liturgies, in:
C. Leonhard—H. Löhr (eds.), Literature or Liturgy? Early Christian hymns and
prayers in their literary and liturgical context in antiquity, Tübingen, p. 175-194.
Pöhlmann, E. 1970, Denkmäler Altgriechischer Musik, Nürnberg (= DAM).
Pöhlmann, E. 1996, art. „Hymnus“, in MGG 2. Auflage, Sachteil 4, 464-472; 505.
Pöhlmann, E.—West, M.L. 2001, Documents of Ancient Greek Music, Oxford (= DAGM).
Pöhlmann, E. 2016, Der Trinitarische Hymnus (POXY 1786) und sein Umfeld, in: Die
Septuaginta—Orte und Intentionen, WUNT 361, Tübingen, p. 835-850.
Schanz, M. and Hosius, C. 1959 (=1922), Geschichte der Römischen Literatur Vol. III,
München.
Schanz, M. and Hosius, C. München 1959 (= 1914), Geschichte der Römischen Literatur
Vol. IV. 1, München.
Smith, J.A. 2011, Music in Ancient Judaism and Early Christianity, Farnham.
Speyer, W. 1989, Der bisher älteste Psalmus abecedarius. Zur Editio princeps von
R. Roca-Puig, in: W. Speyer. 1989, Frühes Christentum im antiken Strahlungsfeld 1,
Tübingen, p. 64-69.
Stäblein, B. 1955, art. „Frühchristliche Musik“, in: MGG 4, 1036-64.
Stäblein, B. 1956, Monumenta Monodica Medii Aevi Vol. I Hymnen (I), Kassel and Basel.
Stäblein, B. 1957, art. „Hymnus“, in: MGG 6, 993-1018.
Stapert, C.R. 2007, A New Song for an Old World. Musical Thought in the Early Christian
Church, Grand Rapids, MI.
Wellesz, E.J. 1945, „The Earliest Example of Christian Hymnody“, CQ 39, 34-45.
West, M.L. 1982, Greek Metre, Oxford.
West, M.L. 1992a, Ancient Greek Music, Oxford.
West, M.L. 1992b, Analecta Musica V, The Christian Hymn from Oxyrhynchus. Greek
Music or Syriac?, in: ZPE 92, 47-54, reprinted in West (2013) 236-244.
West, M.L. 2013, Hellenica III, Oxford.
Zerfass, A. 2008, Mysterium mirabile. Poesie, Theologie und Liturgie in den Hymnen des
Ambrosius von Mailand zu den Christusfesten des Kirchenjahres, Tübingen-Basel.
118 Pöhlmann
Appendix
Five Ambrosian hymns without melismata in a iambic version.
Monumenta Monodica I 1

Ambrosian Hymns 119
120 Pöhlmann

Ambrosian Hymns 121
122 Pöhlmann

brill.com/grms
Book Review
∵
Leedy, D.
Singing Ancient Greek: A Guide to Musical Reconstruction and Performance,
eScholarship, University of California, 2014, 281 pp. Downloadable for free
(http://escholarship.org/uc/item/1rj4j3n1)
How was ancient Greek poetry sung in practice? The music of song that emerges
from the tiny remnants of notated documents represents only the visible tip
of the proverbial iceberg. The likely contours, if not the precise details, of the
submerged bulk can potentially be traced by combining metrical knowledge
of poetic texts, evidence of melodic structures as expounded in technical and
theoretical sources, and reasonable reconstructions of instrumental resources
such as lyres, kitharas, and auloi. Given the ubiquity of music in the ancient
Greek world, and the evident potential for it to affect the impact of songs that
today survive almost entirely as words without music, bringing together such
disparate sources of evidence to create practical realisations of sung music has
long presented a goal and a challenge to musicologists, musically-minded clas-
sicists, and enthusiasts of ancient Greek culture.
Douglas Leedy has risen to the challenge in this bold and valuable handbook,
which deserves to be widely read, pondered, and used as intended to guide
practice in the performing of ancient texts. The text is presented as a typescript
for free download, rather than a formally published and fully polished mono-
graph. Thus, nearly all the Greek script and musically-notated excerpts appear
in Leedy’s clear, neat hand rather than in print; and in many places the staff
notation illustrating the melodies of Greek texts gives way to melodic indica-
tions using modern pitch-letters placed above the syllables of words (similar to
the way ancient melodies were notated). Clearly these aspects of presentation
would have been upgraded and standardised in a finished publication. While
the frontispiece states that ‘the Department of Classics [at Berkeley, University
of California] is pleased to host this suggestive work by a UC Berkeley alumnus’,

124 book review
lack of official publication should not be imputed to the work’s alleged ‘sugges-
tiveness’ nor to any presumed scholarly deficit. To complete the task as it was
obviously intended would have been a time-consuming matter, so we may be
grateful that the author, an American composer, musicologist, and conductor
who died aged 77 in March 2015, was prepared to disseminate the book in its
current form.
Despite not being a professional classicist, Leedy’s knowledge of the ancient
Greek sources is thorough and accurate, his handling of the evidence authorita-
tive, and his use of Greek faultless. Building on the pioneering work of experts
in ancient Greek music (including Winnington-Ingram, Pearson, Anderson,
Wählstrom, West, Barker, and Danek and Hagel), he sets out to provide prac-
tical principles—and in many cases, illustrative musical scores—for singing
ancient Greek texts ranging from Homeric epic through to elegy, iambus, and
the full gamut of lyric poetry. In order to offer a properly comprehensive guide
to singing, he is obliged to go into detail about the numerous metres that one
encounters en route. It is regrettable, if understandable, that the metrical sys-
tems must be presented using the (to many) over-complicated ancient termi-
nology, which may have the effect of preventing some readers from engaging
as desired with the heard rhythms of the poetry. But the scope of Leedy’s task
is considerable, and the issues complex, and he has presented the material in
a commendably accessible manner with no sacrifice of scholarly rigour. As a
result, while the detail of his practical musical reconstructions remains specu-
lative, we may be confident that the melodic proposals for which he argues are
based on accurate citations and sound interpretations of the source material.
The crucial basis for Leedy’s project is that applied by Martin West in his
well-known 1981 article on the singing of Homer, the premise that melodisa-
tion was from the earliest times based on the natural word-pitch of Greek. On
this rides the further assumption that word-pitch in the archaic and classical
periods is accurately reflected by the accent marks introduced in Hellenistic
times. While neither position is incontrovertible and both undoubtedly lack
adequate nuance, they are sufficiently probable to justify Leedy’s proceeding
to make detailed proposals based on the texts of ancient poetry. The value
of his proposals is that, even if one is inclined to alter specific elements (the
working-out of melody for Homeric epic is, for instance, rather different from
that of Danek and Hagel), they produce unequivocal practical realisations for
a wide range of ancient song. These realisations demonstrate, among other
things, that the melodisation of strophic song can be consistent with a rea-
sonable degree of conformity to pitch accent. It would have helped Leedy’s
insistence (pace Martin West) on this sensible viewpoint to emphasise the pri-
oritisation of harmonia over melos in Greek musical theory and practice. If the

book review 125
aural basis of a song is a fixed set of notes such as that used for the tuning of
a lyre, the precise ordering of these notes into a melodic line will be of less
concern for the song’s identity than the repeated use of the same set of notes
from stanza to stanza.
While it would be misleading to view Leedy’s resulting written realisations
of the melodies of Greek poetry as scores, the proposed melodisations may be
usefully played, sung, heard, evaluated, pondered—and, of course, held up to
critique. Leedy recognises that what he offers can only be ‘clues’ to the sounds
of classical antiquity. He writes: ‘our reimaginings of the music of antiquity are
always provisional; we are always ready to try out a better idea, to find a truer
musical expression of a composition’ (11). This formulation betrays a tendency
to privilege the the way the music ‘worked’, when it is arguably no less impor-
tant to think about how the music was heard. Thus it may be the case that ‘the
Greeks nurtured one of the great musical cultures of the world’ (6), but that
does not mean that all their musical expressions counted as ‘great music’ (as
Aristophanes among others makes clear); while the cultural and musical abyss
between the Greeks and us is such that we could not expect to recognise what
would have been heard as right and beautiful to a Greek ear. Leedy says next
to nothing about ethnomusicological parallels, which indicate that in an oral
culture one should not expect, for instance, there to be any fixity of melody
(or even text) across performances, that a profusion of styles and registers can
coexist which operate in very different ways, and that the ideals of sound and
expression in developed Western classical music cannot be readily mapped
onto other traditions.
Nor, surprisingly, does Leedy subject any of the actual notated fragments
of ancient melody to detailed analysis. This could have drawn his attention
to expressive principles that bid to override conformity to pitch-accent (even
if that principle was present). Such expressiveness is evident in the fragment
of Euripides’ Orestes, which Leedy classifies as outside his frame of reference
in that it is an example of the New Music. Of that style he writes (3-4): ‘music
that had been straightforwardly built on the rhythms of the long and short
syllables of the words acquired new rhythmic freedoms, and new melodic
independence superseded what seems to have been a quite strict melodic limi-
tation to the tonal shapes of the words, given by the written accents . . . We can
hardly begin to imagine its nature and sound, and are forced to conclude that,
absent the revelation of substantial new material its reconstruction is beyond
our reach’. A sensitive listener such as Leedy could undoubtedly have elicited
something of the expressive musical idiom of the Orestes papyrus; but he can
be forgiven for choosing to limit his purview to the texts that seemed to offer
providential scope for melodic reconstruction on the principle of word-pitch.
126 book review
For providing musicians and classicists with such a comprehensive, exhilarat-

ing, and well-developed elaboration of that principle, Douglas Leedy deserves
our thanks and commendation.
Armand D’Angour
Jesus College Oxford


Greek and Roman Musical Studies 2017.1 by Greek and Roman Musical Studies

Uploaded by

Copyright:

Available Formats

Greek and Roman Musical Studies 2017.1 by Greek and Roman Musical Studies

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Greek and Roman Musical Studies 2017.1 by Greek and Roman Musical Studies

Uploaded by

Copyright:

Available Formats

Greek and Roman Musical Studies 5 (2017) 1-17

Is the Idea of ‘Musical Emotion’ Present in

Andreas Kramarz, Ph.D.

music – emotions – ethos – mimesis – psychology – magic

© koninklijke brill nv, leiden, ���7 | doi 10.1163/22129758-12341286

Greek and Roman Musical Studies 5 (2017) 1-17

7 See e.g. Kivy 2007; Patel 2008, 86.

Greek and Roman Musical Studies 5 (2017) 1-17

Next, we need to clarify the concept of ‘emotion.’ Based on current

Greek and Roman Musical Studies 5 (2017) 1-17

emotional spectrum may generate incongruences between both perception

Ancient Use of Musical Emotions

26 This is the definition employed in the OHME, see p. 10.

Greek and Roman Musical Studies 5 (2017) 1-17

The Debate about the Relationship between Music and Emotion

Greek and Roman Musical Studies 5 (2017) 1-17

Specifically Musical Emotions

Most of us would probably agree that the emotional experience of a Chopin

Greek and Roman Musical Studies 5 (2017) 1-17

In addition, the enjoyment felt in the aesthetical experience of music in its

the unpleasant end, ἐχθρός (‘hated, hateful’),57 or in Latin, compounds from

Greek and Roman Musical Studies 5 (2017) 1-17

Barker, A. 1984/1989. Greek Musical Writings. 2 vols. (Cambridge/New York) (cited as

Greek and Roman Musical Studies 5 (2017) 1-17

Meyer, L.B. 1956. Emotion and Meaning in Music (Chicago).

The Symphony of Temperance in Republic 4

Plato – musical imagery – temperance – symphōnía – choral practice – octave –

The reader of the Republic is presented on many occasions with thought-

© koninklijke brill nv, leiden, ���7 | doi 10.1163/22129758-12341287

(συμφωνία) or even a proper ‘tuning system’ (ἁρμονία).1 Socrates introduces

1 The Choir of Temperance: A Symphony of Natural Differences

Greek and Roman Musical Studies 5 (2017) 18-34

beginning of this discussion, great emphasis is put on the importance of pre-

provide a precise definition of sōphrosýnē but vaguely likens it to ‘some sort of

Ὁρᾷς οὖν, ἦν δ᾽ ἐγώ, ὅτι ἐπιεικῶς ἐμαντευόμεθα ἄρτι ὡς ἁρμονίᾳ τινὶ ἡ

Greek and Roman Musical Studies 5 (2017) 18-34

Do you see, then—I said—that we prophesied aptly before, saying that

Here the essential trait of temperance is identified in its inclusiveness: differ-

the well-ordered and temperate city is presented as a harmonious choir in

Greek and Roman Musical Studies 5 (2017) 18-34

2 A Symphony of Opposites: Choral Singing, the Octave and the

Greek and Roman Musical Studies 5 (2017) 18-34

Greek and Roman Musical Studies 5 (2017) 18-34

Why do people not sing corresponding melodies at the fifth? Is it because

Once again, the practice of singing antíphōna in octaves is related to a distinc-

Greek and Roman Musical Studies 5 (2017) 18-34

Why is it that the sýmphōnon is more pleasant than unison? Is it because

Greek and Roman Musical Studies 5 (2017) 18-34

Annas, J. 1981. An Introduction to Plato’s Republic (Oxford).

Greek and Roman Musical Studies 5 (2017) 18-34

Aristotle and Musicologists on Three Functions

Aristotle – musicologists – music – leisure

© koninklijke brill nv, leiden, ���7 | doi 10.1163/22129758-12341288

Greek and Roman Musical Studies 5 (2017) 35-42

enumeration, where he firmly distinguished amusement and leisure. In the

Greek and Roman Musical Studies 5 (2017) 35-42

Greek and Roman Musical Studies 5 (2017) 35-42

force us interpreters to reconstruct what his answer might, or should, have

This paper is an addendum to the article “Choreography of Lupercalia: Corporeality in

Roman dance – ludus – choreography – Lupercalia – Roman identity

© koninklijke brill nv, leiden, ��7 | doi 10.1163/22129758-12341286

7 See e.g. Kivy 2007; Patel 2008, 86.

26 This is the definition employed in the OHME, see p. 10.

© koninklijke brill nv, leiden, ��7 | doi 10.1163/22129758-12341287

© koninklijke brill nv, leiden, ��7 | doi 10.1163/22129758-12341288

© koninklijke brill nv, leiden, ��7 | doi 10.1163/22129758-12341289

39 Cf. McLynn 2008.

50 Cf. Alonso Fernández 2016a, 313.

© koninklijke brill nv, leiden, ��7 | doi 10.1163/22129758-12341290

4 For an interesting discussion of these issues see Schulze 2000.

5 Thrasymachus at Demosth. 3, Lysias at Lys. 32.

6 Plato, Prt. 320c8-322d5.

8 See Peirano 2010, especially 39-48.

© koninklijke brill nv, leiden, ��7 | doi 10.1163/22129758-12341291

11 Fairclough 2000, 439.

23 Henderson 2005, 276.