Greek Dance J. W. Fitton

CAMBRIDGE

UNIVERSITY FKESS

Oci^ %

Greek Dance

Author(s): J. W. Fitton

Source: The Classical Quarterly, New Series, Vol. 23, No. 2 (Nov., 1973), pp. 254-274

Published by: Cambridge University Press on behalf of The Classical Association

Stable URL: http://www.jstor.org/stable/638179

Accessed: 08/12/2010 08:21

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.j stor.org/action/sho wPublisher?publisherCode=cup.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

Cambridge University Press and The Classical Association are collaborating with JSTOR to digitize, preserve
and extend access to The Classical Quarterly.

STOR

®
http ://www.j stor.org

GREEK DANCE*

Many books have been written on Greek dance. The fault which bedevils a
large number of them is that their authors have tried to recreate the movements
of the dances from the artistic evidence without taking into account the con-
ventions of Greek vase-painting and sculpture. 2 Other books, and they are the
most useful, set out the literary and the artistic evidence without attempting to
reconstruct the dances. 3 Rarely, however, are the wider implications considered,
and it is these which I wish to discuss here. More analysis and discussion of the
evidence for many of my statements is no doubt required, but the place for that
is a book rather than an article which ranges over a comparatively large field.
In the first two sections I discuss the origin and nature of Greek dance, and in
the following three the effects of dance on the function, form, and rhythm of
Greek poetry.

I. The Origin of Greek Dance

The question 'Why did the Greeks dance ?' might seem to be superfluous and
to invite the glib but not entirely convincing reply 'Why not?'. The reply is
unconvincing because the way in which a practice becomes natural varies from
one historical period to another.

Historical analysis of Greek dance leads us to assume an early connection
with cult or ritual. Born in a later age, with different ideas of religion, we are
almost automatically involved in a prejudice about this. Dance is a pastime ;
religion means duty and devotion. What possible connection could there be
between them? The answer is that cult in a primitive phase involves both
prescribed and also playful behaviour. The two sides are shown by the Latin
word ludus which means both rite and play ; unfortunately Roman religion for
the most part failed to live up to this etymology.

The serious side of early ritual is concerned with the problems of the con-
tinuation of existence and of social solidarity. These problems are worked out
in a kind of play, and the group dance is a way in which the group can feel
itself working them out together. Hunting communities centre their attentions
on the particular wild animals they chase. Dressing up in the skins of these
animals, they dance a perfect hunt, or, alternatively, help to propagate the
animal species by miming their sexual union. The animal masquerade 4 is
frequently depicted on ancient Greek vases, and an animal disguise may lie

1 This article originated as a lecture vases, but his approach is vitiated by its
delivered to the South- West Branch of the interpretation of them in terms of modern
Classical Association at Exeter by the late dancing.

J. W. Fitton in 1963. The scrupulous 3 Of special importance now is L. B.

scholarship of Mr. J. H. Cowell, a former Lawler, The Dance in Ancient Greece (1964) ;

pupil of the author, who has revised the see also her monograph The Dance of the

text, incorporated parts of later drafts, and Ancient Greek Theatre (1964). Fitton would

added the notes, has made publication have welcomed T. B. L. Webster's approach

possible.— J. G. Griffith, F. D. Harvey. in The Greek Chorus (1970).

2 These books are, however, often valu- 4 For animal dances see L. B. Lawler,
able for their reproductions of the attitudes The Dance in Ancient Greece, ch. 4, and A. W.
of Greek dancers. Thus G. Prudhommeau Pickard-Cambridge, Dithyramb, Tragedy, and
{La danse grecque antique, 2 vols., 1965) pro- Comedy 7 ' (1962), 15 1-7.

vides a valuable record of the postures on

GREEK DANCE 255

behind the theriomorphic affinities of many Greek gods and goddesses. The
myth of Io in the form of a cow chased by Zeus in the form of a bull 1 is an
almost transparent relic of an erotic dance in animal disguise such as we see in
primitive cave drawings. Agricultural communities link their rituals with the
death and revival of the seasons, and hold dance festivals at various special
times of year, especially at harvest time when the work is over. Aristotle (Eth.
Nic. 8. n6o a 25) mentions that in early times sacrifices and assemblies took
place chiefly after the ingathering of the harvest, when there was more leisure.
And this was also presumably the time for dancing. Dancing frequently
involved a kind of communion with the living powers of nature through the
handling and carrying of foliage, fruit, and crops. Many Greek female cult-
processions are of this type. A particular dance movement connected with this
kind of seasonal festival is the fertility leap. In the Hymn of the Kouretes a band
of young men call upon their god to leap, and there is no doubt that they leapt
with him. The leap, they say, is for the wine-vats, the fleecy flocks, the fruit
crops, and the fulfilment of the soil. 2

Another type of ritual, which is often blended with masquerade and seasonal
festivals, is the initiation ceremony. Originally this marked the passing from
one age-group or status to another, including passing away altogether to the
world of one's ancestors. The direct representatives of this tradition in ancient
Greece, as in modern Europe, are the rites of marriage and death, and in
Greece they were accompanied by dancing. The case of the Greek wedding 3 is
clear, and Catullus' poem 4 gives the most vivid picture of it. An epithalamium
was a dance as well as a song. The case of the Greek funeral is more debatable,
but the rhythmical use of the hands in lamentation, which, as can be seen from
vase-paintings, became a stylized mime representing the beating of the head
and the rending of the cheeks, can without much apology be called a dance. 5
The more indirect, though flourishing, representatives of the tradition of
initiation are the rites of secret societies (the ^ivarripia) . According to Lucian
(Salt. 15), all iivoT-qpia included dancing. We are reminded of the Auin bush-
man who was asked what really happened at puberty ceremonies and gave the
short and sweet answer, 'We dance'. 6

Characteristic of the mystery rite was the maze-dance. One example is the
wanderings of torch-bearing initiates at Eleusis who were supposed to represent
the wanderings of Demeter in her search for Persephone. 7 The maze-dance
perhaps gives us the clue to the puzzle of the labyrinth. The labyrinth has been
traced in many parts of the world, and is associated with caves, temples,
palaces, and open floors. The explanation is that it was a dance consisting of
intricate maze-like movements. Its original function was probably to produce
amazement in the initiates and to prepare them for the mystic realities of their

1 On Io see A. B. Cook, Z eus ' l ( l 9 l $)> (above, p. 254 n. 4), 54-6, 82-3. For the
438-41. importance of hand-movements in Greek

2 For the fertility aspect of the hymn see dancing see below, p. 261.

M. L. West, J.H.S. lxxxv (1965), 149-59, 6 Curt Sachs, World History of the Dance,

who offers a new text, with commentary. Eng. trs. (1937), 68.

3 The wedding-dance is found as early as 7 For the wandering see L. R. Farnell,
Homer, Od. 4. 17-19 and 23. 131-49. The Cults of the Greek States iii (i9°7)> ^i-

4 Catull. 61. Hymenaeus is bidden to For the maze-like character of the dancing
perform: pelle humum pedibus, manu / pineam at Eleusis, W. F. Jackson Knight, Vergil:
quate taedam (14-15). Epic and Anthropology (1967), 226-7.

5 For funeral dances see Lawler, op. cit.

256 J. W. FITTON

new status. The Greek mysteries were a preparation for the after-life. Modern
scepticism about Orphic and Eleusinian traditions does not seem to have
seriously shaken this view. The mystics took care to carry on their dancing in
the other world. This, I believe, is the reason for the meadows and throngs of
dancers in the realm of the blessed. 1 The procedure of the mystery was pro-
jected into the after-life.

Another dance was the weapon dance. This seems to have been originally
part of an initiation ceremony in which mature youths danced round the
initiate, scaring him and scaring away evil spirits with the noise of their
weapons. 2 The rite was, however, combined with a seasonal celebration aimed
at fertility, as can be seen from the Hymn of the Kouretes. In neither of its two
functions — initiation or the securing of fertility — is there any intrinsic connec-
tion with the war-dance. A similar dance involving the brandishing of sticks or
swords was the ancestor of our Morris dance and the many others which are
related to it throughout Europe, and a comparison of the different forms of
Morris leaves no doubt that it was a seasonal mime of fertility.

The cult dance is frequently dramatic: it involves impersonation and a
myth. The relation of myth to cult is a vexed question. 3 There is the degenerate
myth which comes long after a practice has dwindled to senselessness, but
there is also the more dynamic myth which accompanies the cult act and makes
it something like drama. The cult dramas of the ancient New East and of
Egypt were provided with mythological commentaries explaining what was
going on, and these are often extant. 4 Unfortunately, in the case of Greek cult
the evidence is not so accessible. It also appears that in two cases at least — the
Crane dance at Delos and the Kouretic dance in Crete — the mythical plot was
a reinterpretation of the original dance. 5 The cult dance has within it the
embryo of drama; but a mythical plot is the mark of a change of function.

When a Greek god has an important role in what is evidently a cult drama,
and is also seen as the leader of a dance group, there is a good case for supposing
that originally the dance group and its leader were enacting the cult drama.
Thus Apollo, as leader of the choir of Muses or Graces, is the main antagonist

1 Themistius, Trepl foxi 5 a P- Stob. 4. 52. relationship between myth and ritual is very
49 = Plut. fr. 178 Sandbach, discussed by variable, and generalizations are unsafe.
G. E. Mylonas, Eleusis and the Eleusinian See now, however, W. Burkert in C.Q. n.s.
Mysteries (1961), 264-9. xx (1970), 1-16, and most recently Homo

2 The initiatory aspect of the Hymn of the Necans (1972), as well as the discussion of the
Kouretes and related myths forms the basis whole question by G. S. Kirk in Myth: its
of J. E. Harrison's Themis (191 2, 1927). meaning and functions (1970), esp. 12-31.

3 The point at issue is whether the action 4 See the texts in T. H. Gaster, Thespis
of a myth reflects the action of the ritual. (1950, 1961).

This line of interpretation has been developed 5 That is, the Crane dance, like other

by British scholars: see Myth and Ritual, ed. maze-dances, was part of an initiation cere-

S. H. Hooke (1933). Unfortunately this book mony, but was said to commemorate

is as much concerned with proving the Theseus' escape from the Cretan labyrinth

existence of a certain type of ritual as with (Dicaearchus fr. 85 Wehrli ap. Plut. Thes.

exploring the general relationship between 21 ; cf. the Francois vase, Arias-Hirmer—

myth and ritual. The theory has aroused Shefton, A History of Greek Vase Painting

violent criticism, often aimed at the par- (1962), pi. 43 with p. 288). In the mythical

ticular ritual proposed rather than the explanation of the dance of the Kouretes,

general question (see the criticisms listed Zeus was a new-born babe, not a youth

by Hooke in 'Myth and Ritual: Past and approaching puberty and the age of

Present' in Myth, Ritual and Kingship, ed. initiation.
S. H. Hooke [1958], 1-2 1). In fact the

GREEK DANCE 257

in the battle with the Python at Delphi. 1 Dionysus, leader of the Bacchae, is the
main participant in the drama of Pentheus' destruction. And we may add
Orpheus, the leader of the Muses, who is killed by the Maenads in what is
clearly a ritual act. 2

In early communities, dance is song-and-dance. At a folk or primitive level,
voice, music, and body-rhythm are inextricably connected. The development
of artistic culture lies in the gradual differentiation of the parts from this
whole. The practical result is that in the study of the origins of Greek poetry we
frequently find traces of dance in the development of non-dance forms. Thus the
epic poet is at first sight a reciter ; then when we see him, as in Homer, with a
lyre, he is a singer ; and then when we see him operating amid dancing youths,
or leading a chorus of dancing maidens at the Delian festival, we see him as the
leader of the song-and-dance. 3 Similarly we quite rightly stress the personal
quality of Sappho's poetry, yet on closer inspection her Ode to Aphrodite (fr. 1
LP) reveals itself as nothing other than a hymn. We have no clear evidence for
the way in which such a hymn would have been performed, but we may arrive
at some idea of the tradition lying behind it through the imagination of a poet
in the Palatine Anthology (9. 189, anon.) :

"EXOere TTpos relievos Tavp(i>7Tihos dyXaov "Hprjs,
AecrftlSes, aft pa ttoSwv firjiMad' iXicrao fxevai.

zvda kolAov GrrjGaade Qefj -)(op6v' u/x/xt 6° drrdp^ei
Uarrc/xh xpvaeirjv ^epalv e-^ovaa Xvprjv.

oXfiiai 6p)(7]dpLov TToXvyrjdeos' rj yXvKvv vjxvov
elaateiv avrrjs Sdfere KaXXioTrrjs.

It is difficult to infer what the Greeks in historical times did from the evidence,
such as it is, for the mythical dance. For instance, when Pindar represents all
the gods as taking part in a Bacchic revel in Zeus' house {Dith. 2, fr. 70b
Snell 3 ), the picture is clearly not a replica of something that happened in real
life. That the Greek imagination was caught up in the dance is of course an
indirect testimony to its power. For example, the depiction of Ares as a dancer, 4
the warrior as a dancer of war,s and the elaborate contrast in Euripides'
Phoenissae (226-49) between the revel throng of Ares and that of Dionysus, all
hang together as a tradition in which battle was seen imaginatively as a sort of
dance. And we can see a certain appropriateness in this when we realize that
the form of warfare depicted by Homer resolves itself into tournaments of
agility between rival champions.

1 Apollo and the Muses: Horn. 77. 1. Phemius and Demodocus. For the dotSds
603-4 5 Hymn. Horn. 3. 182-206; Paus. 5. 18. among dancing youths see Od. 8. 261-4,
4 (the chest of Cypselus). On a sixth- and leading the Delian maidens, Hymn.
century Attic amphora in Copenhagen Horn. 3. 166-73.

(N.M. Inv. 3241 = CVA Copenhagen 3, 4 According to a Bithynian story Ares was

pi. 102. 2), four Muses (if they are Muses a dancer before he became a warrior

and not Graces) walk playing castanets (Lucian, Salt. 21).

behind Apollo who is playing the kithara 5 A warrior could be called opxyorrip

as he leads them to Zeus. 7roAe/Ltoto, Nonnus, Dion. 28. 304; cf. 275,

2 The classic exposition of Pentheus' where the Corybantes are called opxrjaTrjpes
death is of course in Euripides' Bacchae. 'Evvovs. The Thessalians called their front-
For Orpheus' death see W. K. C. Guthrie, rank men and champions TTpoopxr)OTr}p€s
Orpheus and Greek Religion 2 (1952), 32-3; the (Lucian, Salt. 14). Athenaeus (14. 628F)
longest account is Ovid, Met. 11. 1-84. quotes a verse of Socrates: ot Se x°P°' s

3 The doiSot par excellence in Homer are /caAAiora Oeovs tl/jl&oiv, apLoroi iv iroXe/Mcp.

258 J. W. FITTON

Myth leads away from the actual local practice to the world of story- telling
and allegory. Sometimes an incongruity arises. When we see Heracles as a
leader of the Muses with a lyre in his hand, 1 we see not the hard-headed
philistine of popular tradition, but the leader of a group-dance in honour of
victory. Allegory may develop from actual practice. The god of Love and the
goddess of Victory are shown in art as dancers. 2 The Muses exhibit many
oddities if they are considered simply as fictions devised to emphasize the poet's
debt to tradition ; they are in the first instance a band of women who sing and
dance on the hillside, as is evident from Hesiod ( Theog. 68-70) :

at tot* ioav irpos "OAvfiTrov ayaAAo/xerai ottl KaXfj,
dfjuppoGLT) ixoXttt}- rrepl §' ta^e yaia fieAaiva

VfJLVeVGGLLS, ipCLTOS $€ TTOO&V V7TO OOV7TOS 6p<x)p€L,

In the case of the dances of Pan, and of Satyrs and Bacchants, there is no
doubt about their historical existence. Plato, in distinguishing desirable from
undesirable dances, firmly labels these as undesirable, 3 and he is clearly legis-
lating about actual practice, not fantasies. When a modern scholar, com-
menting on the head-jerking described by Pindar in a Bacchic dance, says
'Pindar may have drawn the epithet pa/javxevi from contemporary representa-
tions of the Maenads on Attic vases 5 , 4 it is reasonable to say that academic
blinkers have prevented him from seeing that Pindar did not have to consult
vases in order to recognize a characteristic motion in an actual dance. The
ecstatic hillside dance of Dionysiac cult, in which nervous exhaustion led to a
state of trance and the sensation of perfect peace, was real, and analogies to it
exist in other cults.

dv€jJLO€VTl, S 677 Oydll)
oAoAvyfJLGLTa TTOLVVVXIOIS V7TO TTCXp-
OeVWV 10L-)(€L TTOOWV KpOTOLGLV.

Euripides is describing not the cult of the oriental upstart Dionysus, but the
festival of the sober goddess Athene (Heracl. 781-3).

II. The Nature of Greek Dance

The two most frequent terms for 'dance' are x o P°s an d opxqvis. x°P°s> according
to Hesychius, 5 originally meant a round dance. It tends to mean the dance of
the group as opposed to the individual, the formation dance as opposed to the
improvised, opxrjvis shows the opposite tendency: this, not x°P°$> 1S what
tumblers do. 6 x°P°s> tne set-dance, is distinguished from preliminary move-
ments. For example, in Callimachus (Hymn 3. 240-3), the Amazons first dance

1 For Heracles' connection with the 4 L. R. Farnell, The Works of Pindar ii
Muses and lyre-playing see F. Boehm in (1932), 423, discussing Dith. 2 fr. 70b. 13
R.E. viii. 574-8, esp. 576-7, and R. Peter Snell 3 ; for a more open-eyed view, see E. R.
in Roscher i. 2, 2970-6, esp. 2975-6. Like Dodds, The Greeks and the Irrational (1951),
Apollo he was called Mouaayeras (I.G. xiv. 273 f., and his note on Eur. Bacchae, 862-5.
101*). He was taught to play the lyre by 5 He glosses xopos- kvkXos, are^avos.
Linos (Apollod. Bibl. 2. 4. 9). 6 Thus Athenaeus (5. i8od) uses

2 M. Emmanuel, The Antique Greek Dance, opxeioOat of the tumblers in Od. 4. 18-19
Eng. trs. (1927), 248-51. and II. 18. 604-6; cf. Hdt. 6. 129, of

3 Laws 7. 815 b-c; cf. G. R. Morrow, Hippocleides.
Plato's Cretan City (i960), 362-5.

GREEK DANCE 259

the weapon-dance and then set up the broad x°P°s I i n tne description of a
performance of Sappho's hymn in the Anthology (p. 257 above), the girls first
do a processional and then set up the x°P°$ f° r tne hymn ; in tragedy, there is
first a march into the orchestra and then the chorus sets up its formation-dance
or ardaifiov. The formation aspect is also seen in the extension of the word x°P°s
to mean 'dancing-place' and even 'curving row', as, for example, a row of
teeth. 1 On the other hand opxycns, according to Athenaeus (1. 21 a), was used
metaphorically for all sorts of motion. This agrees with the extant evidence, for
it is also used, for example, of hearts beating with fear and of an earthquake. 2
The distinction between x°P°s an< ^ opx 7 !® 1 * 1S not always clear-cut, however,
especially since op^at? was always the more general term. A similar distinction
exists in Medieval Latin and European languages: Medieval Latin: choreaj
ballatio; Italian: carola\danza\ Provencal: carola/dansa; Old French: carole\
danse ; German : reigenjtanz. Similarly the English word 'carol' originally meant
a dance in a ring to the accompaniment of song. 3

The protoplasm of Greek lyric poetry was a song-and-dance. The integrated
nature of the performance is reflected in the word pLohrr) which exhibits a basic
ambiguity. Sometimes, as with the acts of a warrior or tumbler, the reference is
clearly to action without song ; 4 sometimes just as clearly it means song ; 5 and in
many cases the word is coupled with other terms for song-and-dance and must
therefore refer to a song-and-dance performance. 6

The Greek scholarly and philosophical tradition recognized the importance
of dance in Greek culture. 'What is \xovaiKrf\ V asks Plato. The answer that
immediately suggests itself is 'Harp-playing, singing, and moving properly'
(Alcib. 1 108 c-d) . This is only natural, since plovglkt] comes from the Muses who
were dancers as well as singers. The distinction between novGLKr), in its widest
sense of 'culture', and yvjjLvaarLKrj is not a simple distinction between mind and
body. Plato, for example, in his analysis of (jlovglkti refers to the bodily ex-
pression of rhythm, 7 and derives the rhythmical nature of poetry from the
random exploratory movements of children (Laws 2. 653 d-e). Dance was a
thing of the soul as well as of the body. 8 Soul is the difference between a live
man and a dead man, and dancing is clearly a good symptom of the difference.

Dancing enjoyed high prestige from the period reflected in the Homeric
poems. In Homer nobles dance as an exhibition of skill (Od. 8. 250-65, 370-80),
and King Priam's sons were best at dancing if nothing else (//. 24. 261). We
hear that the early tragedians were called dancers not only because of their
'dancy' plays but also because they gave dancing lessons (Athen. 1. 22a).

1 x°P°s oBovrwv, Galen, De Usu Part. 11. 8 152, 23. 145.

Helmreich; cf. irpoaOioL xopoi of the front 7 Laws 2. 664 c; cf. 672 e: the branch of

teeth, Aristoph. Ran. 548. tiovaiKrj known as ^opeia combines rhythm,

2 Of fear: Aesch. Choeph. 167; Ion fr. defined as the ordering of motion, and
50N 2 and Snell ; of an earthquake : Callim. dpfxovia, the ordering of voice. Conversely,
Hymn 4. 139. yvfivaoriKrj affects the soul (Rep. 3. 410 b-d).

3 Curt Sachs, op. cit. (above, p. 255 n. 6), 8 This is because, according to Plato
269-75; P. Dearmer in The Oxford Book of (Laws 10. 896 a), the soul is the source of
Carols (ed. Dearmer and others, 1928), v-ix; movement. For emotional processes con-
Oxford English Dictionary s.v. Carol. sidered as movements of the soul see Athen.

4 Horn. //. 7. 241 : /ue'ATrccrflcu *Aprfi. 14. 628c (= Damon fr. 37 B6 DK), where

5 Eur. Ion 881-2: /zcAttcov rds KuOdpas ol -nepl Ad^icova are quoted for the opinion
ivoirdv. that ras w&ds Kal rds opxrjoeis avdytcrj

6 fioXirrj and song: Eur. Heracl. 780; ylveodau Ktvovfievrjs tto)S rrjs ^XH^'i ps.-Ar.
fjuoXirrj and dance: Horn. //. 13. 637; Od. 1. Probl. 919*26-37, 920 a 3~7.

2 6o J. W. FITTON

Sophocles danced in the nude to lyre accompaniment to celebrate the victory
at Salamis, and appeared in his play Nausicaa dancing and playing ball (id. i .
2of) . The tragedians were not only writing great odes ; they were also active in
organizing the dance of the chorus. Aeschylus invented many new dance
figures (id. i. 21 e-f), Sophocles wrote a book about the chorus (Suda s.v.
Uo(f>oK\7)s), and Aristophanes tells us that the tragic poet Phrynichus derived
his wonderful displays from the dance-pieces of the Mother Goddess (Aves
746-50) . I Dance was an important part of the song-and-dance that we call
choral lyric. Pindar refers proudly to his dancers (01. 14. 15-17), and, more
subtly, invests his poetry with the glamour of the dance (Pyth. 1.2: pdaus
dyXatas dpxd).

After the fifth century, dance underwent a steady decline in prestige. It
became an ungentlemanly thing, suitable only for the socially inferior, or, as
Cicero said, for drunkards : nemo fere saltat sobrius, nisi forte insanit (Pro Mur. 13).
For Greece with its wealth of dance forms was not peculiar : Italy also had a
great song-and-dance tradition, 2 but found only Plautus to raise it to the level
of art. Greece was, however, peculiar, in this as in other ways, in the use that
it made of its tradition.

The Greek dance was frequently vigorous and closer to folk-dance than to the
stylized motions of ballet or ball-room. Vase-paintings show the whirling
turn, in which full use was made of feminine clothing, the backward and
forward bending of the body in the Bacchic dance, and the crouching upon the
haunches that is now known as the Cossack dance. 3 Drawing the rough dis-
tinction used by Curt Sachs 4 between 'expanded' dance, a rebellion against
gravity, and 'close' dance, in which the dancers are more restricted in their
movements, we see that the Greeks were masters of both styles. This distinction
is often correlated with a sex difference : the woman is more restricted than the
man in dance movements. Plato is very anxious that men should not move like
women, or women like men. 5 Leaping, striding, kicking, and skipping in
'expanded' dance are, if the Greek evidence is any guide, male rather than
female activities. Bending, stretching, whirling, hand-gestures, and the carrying
of baskets and suchlike in procession 6 are 'closer' motions and as such more
feminine. But we must not press the distinction too hard, especially since Greek
women did not by any means always observe the decorum required by Plato.

Greek dance was not always poised or, to use the modern term, 'eurhythmic'.
The Maenad on vases who looks as if she is having a nervous breakdown is, to
use a recently fashionable catchword, 'sent'. 7 The satyr-dance involved the
mimicry of drunkenness (PL Laws 7. 815 c), and when Pentheus is told to lift
his right hand along with his right leg, he is being instructed to make an

1 In this context Aristoph. Vesp. 1498 is {Laws 7. 802 d-e) ; feminine gestures should
particularly revealing : et res rpaywSos not be assigned to verses composed for men
<f>r)oiv opxeioOau Ka\a>s ... (2. 669 c).

2 See the references to specifically Italian 6 For basket-carrying in processional
dancing s.v. Tanzkunst in R.E. ser. 2. iv. dances see L. B. Lawler, op. cit. (above,
2233-47, es P- 2239 and 2247, and in general p. 254 n. 4), 108-10.

G. Wille, Musica Romana (1967). 7 For representations of Maenads on

3 M. Emmanuel, op. cit. (above, p. 258 vases, and the interpretation of their move-
n. 2), 140-2, 171-4, 1 70- 1 respectively. ments, see L. B. Lawler, Mem. Amer. Acad.

4 Op. cit. (above, p. 255 n. 6), 24-37. Rome vi (1927), 69-112; M. W. Edwards,

5 The lawgiver must see that suitable J.H.S. lxxx (i960), 78-87.
songs are allotted to men and to women

GREEK DANCE 261

arhythmical movement like the capering of the typical drunken satyr on vase-
paintings. 1

Greek dance made an expressive use of arms and legs that would be un-
thinkable in traditional ballet. Hand-movement was clearly important;
Aristoxenus considered that the Mantinean dance was the best of all on this
account. 2 Arm-movements were trained by ball- throwing, which frequently
accompanied the dance in ancient Greece. 3 Whether the Greeks had anything
like the Hindu JVatya, a science of dance and drama concerned with torso- and
arm-movements, 4 is a difficult question to answer. Plutarch's dialogue on
dancing (Quaest. Conv. 9. 15 = Mor. 747A-748D) implies a tradition of gesture
but not necessarily a science. Gesture passed from the dance to the tragic actor
and thence to the public speaker: hence rpaywSetv and xeipo^ojuetV of orators,

Greek dance was 'mimetic'. 5 But what is mimesis? A recent inquiry 6 suggests,
after an exhaustive survey, that it was a dance word, and that the dance
involved was originally the cult dance of Dionysiac worship. It has not, how-
ever, by any means been proved that Dionysus was originally involved. What
is rather clearer is that the sense of mechanical copying was brought in by
Plato for his own philosophical ends. 7 When the word is used of cult-acts, then
clearly this is not imitation, for the worshipper did not imitate the god but
impersonated or acted the role of the god. As the word often occurs in ancient
discussions of imovoiktj, it is interesting that Roller's inquiry shows a notable
emphasis on the bodily expression of behaviour. This must surely be the point
of Plato's application of the word mimesis to the parts of Homer that were in
direct speech (Rep. 3. 3926-3930). It was not that Plato had some abstract
objection to direct speech, but that the epic reciter acted out the speeches of his
characters with mime and gesture. That mimesis involved bodily behaviour is
indicated by Plato's use of this and kindred terms in the Laws to refer to the
dance. For example he says casually, as if it were generally accepted, that
dances are the expression of character (iiLfirj^ara rpoirajv), both good and bad
(Laws 2. 655 d).

The view that dance is 'expression' (mimesis) is developed further in Plutarch's
discussion (cited above). Here, however, there is a shift in terminology.
Dance is made up of movements (<f>opal), holds or attitudes (cr^aTa) , and
gestures (Se^ei?). Movements 'indicate' like metaphors in poetry; attitudes
'express' like onomatopoeia; gestures 'demonstrate' as proper names do in
poetry. In this analysis, the term mimesis, 'expression', is reserved for attitudes,
and is not applied to movements or gestures. It is clear, however, that in a wider

1 Eur. Bacch. 943-4. A satyr on the 5 The dance itself, says Aristotle {Poet.
Pronomos vase is executing such a step, and 1447*28), can imitate character, emotion,
it has been suggested that he is performing and action.

the aiKiwis, the dance typical of the satyr- 6 H. Roller, Die Mimesis in der Antike

play (A. W. Pickard-Cambridge, The (1954); the book has been severely criticized:

Dramatic Festivals of Athens 2 [1968], 254 with see, for example, G. F. Else in C.Ph. liii

fig. 49). (1958), 73-90*

2 Athen. 1. 22B. On x €L P ovo l J '^ a see 7 On Pl ato ' s use °f wmwwmj and kindred
Pickard-Cambridge, op. cit. (above, n. 1), terms see, briefly, D. W. Lucas, Aristotle:
248-9. Poetics (1968), 260-1; at greater length,

3 For ball-games and ball-dancing see R. McKeon, Modern Philology xxxiv (1936-7),
Athen. 1. 14D-15C. 3-16; W.J. Verdenius in G. Vlastos (ed.),

4 See G. P. Kurath in Funk and Wagnall's Plato : a collection of critical essays ii ( 1 97 1 ) ,
Standard Dictionary of Folklore, Mythology and 259-73.

Legend ii (1950), 786.

262 J. W. FITTON

sense all three components are thought to be expressive. It is interesting that
the only purely stylistic criterion used in the passage is that of gracefulness,
which is applied to gestures ; the main emphasis is on the expressive power of
the dance.

How expressive, then, was Greek dance ? Showing-off dances, such as the
leaping of young men between the points of swords, referred to by Democritus
(68 B 228 DK), tend towards a circus act. Processionals seem to involve less
expressive movements, 1 but of course carriage could express high or low, manly
or womanly character. Curt Sachs 2 uses a distinction between image-dance and
imageless dance, and it is clear that the imageless dance which, according to
Sachs, is mainly circular, was practised by the Greeks. The imageless dance,
however, while not expressive in the way that an animal masquerade is, very
frequently uses the formation of a circle as a symbol of something like possession
or incorporation. Thus the circle-dance of the Furies around Orestes (Aesch.
Eum. 307-96) is intended to bind and dominate him. It is difficult to form a full
picture of the expressive powers of Greek dance, but apart from the general
argument from mimesis, it can be surmised that the influences of cult on the one
hand, and of drama and choral lyric on the other, tended to pull the dance in
the direction of impersonation or symbolism, and make it very different from
the dancing with which we are familiar.

III. Dance and the Function of Greek Poetry

The Greeks felt keenly the magical quality of poetry and, more generally, its
power of affecting the personality. In discussing the tradition that poetry is
magical, one of its aspects will be ignored, namely what may be called the
carmen-tradition of magical effects which goes back to the sacred words and
formulae of ritual ; we will concentrate instead on its more physical aspect.

The power of music to move objects is exemplified by the stories of Orpheus,
of Apollo at Troy, and of Amphion at Thebes. 3 The two last were both lyre-
players who played the stones into position for the building of the city.
A possible echo of this mythical business is the use of music by Epaminondas
of Thebes at the founding of Messene, and contrariwise by Lysander at the
demolition of the walls of Athens. 4 Euripides suggests a rationalization, by
saying that the music of Amphion at Thebes was to give facility to the hands of
the builders (Page, Gk. Lit. Pap. no. 10. 84-9 [Antiope]). This tradition of magic,
then, may well derive from the use of music to lighten and regulate labour.

The power of music over animals, illustrated by the stories of Orpheus, is an
extension of the role of the song-leader in leading and inspiring the group
dance. It is represented as an ability to organize animals into a band of
followers, and is really the power by which participants in the dance feel
themselves possessed. In Pindar's dithyramb to Thebes (Dith. 2. fr. 70b Snell 3 ,
19-23) the beasts that form the train of Artemis are said to dance, and to beguile
Zeus with their performance. In the Palatine Anthology 5 there is an anecdote
about a priest of Cybele who escaped from a nasty situation by the magical

1 L. B. Lawler, op. cit. (above, p. 254 n. 4), 4 Epaminondas : Paus. 4. 27. 7 ; Lysander :

99-100. Xen. Hell. 2. 2. 23; Plut. Lys. 15. 4.

2 Op. cit. (above, p. 255 n. 6), 62-77. s Anth. Pal. 6. 217, imitated in 218 and

3 See W. F. Jackson Knight, op. cit. 219.
(above, p. 255 n. 7), 1 18-19 with the notes.

GREEK DANCE 263

power of his drum-playing. When he met a ravening lion, he beat his holy
drum in fright, and the beast was filled with the goddess and rolled its head in
ecstasy. The priest dedicated to the goddess the lion that had taught itself to
dance.

The power of music over human beings is also the power to inspire the dance.
This can be seen from the cult of the Gorybantes, who danced round a person
to cure him of his malady. Soon he was possessed by the all-pervading rhythm
and, after passing through a trance-like state, emerged cured. 1

Dance involved a religious as well as a magical experience. To dance in a
chorus was to devote oneself to a god; hence the meaning 'devotee' or 'pupil'
which attached itself to the word xopevrrjs. 2 Sophocles makes his chorus say :

€t yap at roiaihe 7rpd%€is rt/uai,

ri Bet fxe xopeveiv; (0,T. 895-6)

Since the dance was a way of rejoicing and giving thanks, the implication is
that there is no occasion for rejoicing and thanksgiving, and therefore no
occasion for dancing. The religious dance strengthens the feeling of unity in the
group by overwhelming the newcomer and making him feel at home with gods
and men. A vivid example of this effect is seen in the case of a modern observer
of the dance ceremonies of Haiti. Though experiencing a culture quite alien
to her own, she was drawn to participate and to feel that the Voodoo religion
was somehow true. 3 In sixth-century Attica the aliens and wanderers sought for
a 'togetherness' which would compensate for their alienation from official cults
and brotherhoods. In choosing Dionysus, the dance god par excellence, as the
patron of their gathering, they chose wisely. 4

It is these magical and religious effects of the dance that we should bear in
mind when we consider the attack of Plato on the enchantments engendered by
music and poetry. 5 He is thinking not of words on a page, or even the words of
a reciter, but the infectious atmosphere of a popular festival. The magic was
embodied in the music and the dance ; the audience felt it working upon them.
And in the dramatic festivals of Athens there was participation in a real sense,
in that the dancers of the choruses were ordinary citizens.

The effect on character which the Greeks attributed to poetry was not
merely a matter of the content, but also of the style ; and this included the
apjAovla, or musical form. Some account therefore of these 'modes' is necessary. 6

1 For the healing effect of Corybantic poetic genres that produced the right kind
dancing see PL Laws 7. 790 d; for the cult of enchantments (eWSai) (Laws 2. 659 d-e;
in general, E. R. Dodds, The Greeks and the 7. 812 b-c). The acceptable types of musical
Irrational (1951), 77~9- enchantments are listed in Laws 3. 700 b.

2 x°P eVT ys of a devotee of a god : PL 6 There are many accounts of the Greek
Phaedr. 252 d; of a devotee of a philosopher : scales. R. P. Winnington-Ingram provides
Julian, Or. 6. 197D; of a pupil: Libanius, Or. a sound survey of recent work (Lustrum iii
54. 38. [1958], 3i~7)j and the literature there men-

3 Maya Deren, Divine Horsemen: the tioned provides copious references to earlier
Voodoo Gods of Haiti (1953). discussions. Unless the 'gapped' scales de-

4 L. Gernet and A. Boulanger, Le ginie scribed by Aristides Quintilianus (1. 9, p. 22
grec dans la religion (1932), 124-5; E. R. Meibom) are accepted as the fifth-century
Dodds, Euripides: Bacchae (1944, i960), note apfiovtai, and scholars disagree about their
on lines 421-3. authenticity, the structure of the dp/xoviou is

5 PL Rep. 10. 605 c-608 b. The poets in unknown. The later application of the term
his other ideal state had to compose, and dp^iovta to octave-species is in this respect
therefore the citizens could hear, only those misleading. In fact, the relationship between

264 J. W. FITTON

There were different ethnic traditions of song, indicated by ethnic names such
as Dorian, Phrygian, Ionian, and Lydian. 1 Some of these traditions fell away or
were absorbed in others. 2 The ethnic provenance was indicated by an adverb
ending in -tart, a rough-and-ready terminology in as much as the same words
were used in non-musical contexts. 3 These adverbs referred to a apfjuovla,
primarily a way of setting strings on a lyre. The mode was a melodic pattern
of notes. Gradually it was analysed, and its basic elements, such as fourths and
fifths, detected. We then have a system of keys (tovol), which frequently retain
the old modal names, sometimes with the prefix vtto- (below) or virep- (above)
to indicate the relative pitch. 4 But gradually the old modal style was lost,
either because it was not based on an exact pitch or because it was too frag-
mentary for a developed musical style. Thus Plato deals with the modes in the
Republic (3. 398d-39ga), though elsewhere he frequently mentions scales,
intervals, and so forth. Aristoxenus is concerned mainly with keys, not modes. 5
The author of the De Musica attributed to Plutarch, an archaizer, is still
interested in the modes, but he frequently uses the term tpottos, 'manner',
instead of ap/xcWa, to refer to them. 6 This is what we would expect if, as I have
suggested, the terminology was originally non- technical.

Greek music was the offspring of native folk-song wedded to an Eastern
instrumental tradition. 7 The Eastern tradition, as far as the evidence allows us
to judge, was also modal. The difficulty in giving precise pitch- values and
evolving a system of scales for Greek music also arises in the study of folk-song.
Folk-melody is frequently very restricted in its range. We read that c the scale of
a folk tune is nothing more nor less than the series of tones which it employs' ;
'flexibility and variability characterize the tune as well as the text of a folk
song . . . Singers in the same community give a different form to the melody' ;
and 'many details of melody, intonation and rhythm, often quite subtle, make
up a special phase of folk song, its manner of rendition 5 . 8 The song combines
melody, timbre, and rhythm : alter any one of these, and the whole character of
the performance is changed.

apjjLovtai, octave-species, and tovol is much dorian (Heracl. Pont. fr. 163 Wehrli, ap.

disputed, the height of scepticism being Athen. 14. 624 e-f).

reached by M. I. Henderson, who believed 3 p or example, aloXiarl also means 'in

that there was a complete break in the musi- Aeolic dialect'. The related verbs ending in

cal tradition at the end of the fifth century -i£a> often have an even more general sense,

(see her chapter 'Ancient Greek Music' in e.g. Aclkcovl&lv 'to behave like a Spartan'

The New Oxford History of Music i (1957), or 'to be (politically) pro-Spartan'.

336-403). The clearest statement of Professor 4 The tovol are arranged in the most

Winnington-Ingram's own views is now to be developed scheme in the system of Alypius

found in Grove's Dictionary of Music and Musi- (C. von Jan, Musici Scriptores Graeci [1895],

cians (1954) 5 , s.v. Greek Music (Ancient). 367-406).

5 For Aristoxenus' tovol see his Harm. 1.

1 Other styles of less importance were the 37 and [Gleonides] Isagoge 12.

Cretan, Carian, and Mixolydian. 6 He uses em tov Acoptov t P 6ttov for

2 Terpander wrote music in the Boeotian Acoplotl (de Mus. 17. 1136F). The usage is
style ([Plut.] de Mus. 4. 1132D; Suda s.v. as old as Pindar {01. 14. 17), who speaks
Moaxos). It was known to Sophocles (fr. of music iv Avh<x> Tpoiuo.

966 P), but does not occur in later writings ? Curt Sachs, The History of Musical

on music. Composers wrote in the Locrian Instruments (1940), 128, says 'no instrument

style in the time of Simonides and Pindar, originated in Greece'.

but it later fell out of favour (Athen. 14. 8 G. Herzog in Funk and Wagnall's

62 5E). The Aeolian seems for practical pur- Standard Dictionary of Folklore, Mythology and

poses to have been absorbed in the Hypo- Legend ii. 1041, 1037, 1041 respectively.

GREEK DANCE 265

Likewise the style of Greek music involved more than melody. The tunes
(voyboi) said to have been ^discovered' (more probably reorganized) by
Terpander have ethnic and rhythmic names, 1 and a tune of Olympus has a
cult-name [imTviiftiSios, Pollux 4. 79). The terms 'tense' [avvrovos:) and
'relaxed' [dvecfxevos:) were first assigned by scholars to high and low pitch, but
this led to certain contradictions. I suggest that the terms were concerned with
timbre, with the distinction between rich and thin voices. For example, negro
voices have a characteristically rich tone by comparison with the thin voices of
orientals. Primarily the Greek terms indicate the bodily state of the performer,
and for this reason rhythms too could be classified as 'tense' or 'relaxed'.
Presumably Ionics were 'relaxed' because it was the style of the Ionian to let his
body go, whereas Dorian and Cretan rhythms were 'tense' in that a muscular,
perhaps even rigid, style of performance was required. 2 Thus the musical modes
could not be very easily separated from the rhythms. The author of the De
Musica attributed to Plutarch says : 'Neither musical expertise nor the know-
ledge of various rhythms will decide the appropriateness of the elements. The
right character [ethos) is produced by a combination of music and rhythm. He
who knows the Dorian mode without being able to distinguish the fitness of its
use will not know what he is doing — in fact he will not preserve the character'
(ch. 33, 1143A-G, summarized). An analogy might help here. Terms like
'Irish jig', 'Highland fling', and 'Latin- American samba' convey an impression,
have character in as much as they combine music and rhythm, and both are
needed for the over-all effect. The ethos of the modes has often been discussed, 3
but not so much attention has been given to the rhythms. 4 This is odd, because
there is far more evidence from ancient scholars on the significance of rhythm,
and it is quite wrong to dismiss it as no more than an idea dreamt up by
Walter Headlam. 5 Moreover the extant body of poetry, which tells us next to
nothing about the music, still presents us with a rhythmical form.

Rhythmical character [ethos) is made visible in the movements of the dance.
When Plato and others grumbled about the musical revolution which had been

1 The list of his nomes in [Plut.] De Mus. pitched. The author seems to assume a
4. 1132D includes the Boeotian and Aeolian similarity in character between the melody
and the Trochaic. and rhythm associated with a dp^ovla ('if

2 Cretics were ovvtovol, Strabo 10. 4. 16. the melody is tense, then the rhythm is
The tenseness of Dorian rhythm depends also'). But of what dpfiovla is this true?
on the assumption that the dactylic hexa- Dorian rhythms might be tense, but there
meter, which according to Eustathius (Od. is no evidence that the melody was any more
P- 1899. 58-64) could go with tense move- than solemn. Nor is it always true of Ionian
ments, had a Dorian ethos: Aristotle (Poet. music, since Ionian melody could be tense
i459 b 34) uses the adjective ardatfios of the but the rhythm relaxed.— J. H. C]
hexameter and (Pol. I342 b i3) of Dorian 3 The basis for all discussion is still H.
music. [I have not tried to quote further Abert, Die Lehre vom Ethos in der griechischen
evidence for the statements about avvrovos Musik (1899), where the evidence is col-
and dv€Lfi6vos. The point made is basically lected. The most recent treatment is W. D.
sound : the primary, meaning of the former Anderson, Ethos and Education in Greek Music
is 'tense' and of the latter 'relaxed', and when ( 1 966) .

this is realized there is no contradiction 4 The only comprehensive collection of

between avvrovos Avotart and ocdrovos the evidence is G. Amsel, De vi atque indole

avvrovos (which I assume is the 'contra- rhythmorum (Breslauer philologische Abhand-

diction' referred to by the author), since lungen, Band 1 Heft 3, 1887).

both involve tense notes. But the evidence 5 'Greek Lyric Metre', J.H.S. xxii (1902),

for the ethos of the dpfiovtat shows that in 209-27. His views on this, as on so much

this scheme at least, avvrovos was applied else, have been expanded by George

to high-pitched music and dveifxivos to low- Thomson, Greek Lyric Metre (1929, 1961).

266 J. W. FITTON

taking place towards the end of the fifth century in Athens, it is clear from a
close analysis of texts that they had in mind the forms of dance that went with
the new music as well as the music itself. Similarly the early fifth-century poet
Pratinas in his attack on the new style of his day (fr. i = Page, P.M.G. no.
708) l is objecting to the turbulent dance-movements as much as the aulos-
playing. He wants a simple beat in the Dorian style, and points to the correct
movements of his own hand and foot. It is perhaps significant that when
Aristophanes parodies Euripides' style, and turns elMaaere into eUei-ei-eiXiaaere
{Ran. 1 3 14, cf. 1348), the word he chooses for musical perversion is a word
meaning 'whirl' — in other words, his criticism may have been directed just as
much at the whirling dance-movements of the singer.

The Pythagorean view that music had a great effect on character is well
known. What is interesting in the present context is that the Pythagorean
Damon traced this effect to a dance performance. In one fragment (37 B 6 DK,
ap. Athen. 14. 628c) his point is that in moving the body in a certain way the
the soul is moved as well. Hence gentlemanly motions produce a gentleman,
and so on. Another fragment (37 A 8, ap. Galen, De Plac. Hippocr. et Plat. 5. 453
Mtiller) relates that when Damon saw some young men acting wildly, he told
the flute-girl to strike up the Dorian tune, and they 'ceased forthwith from
their crazy movements'. Again, when Aristotle is talking about the effect of the
Phrygian mode upon the character, he argues from the Bacchic cult-dance and
similar types of movement {Pol. I342 b i~i2).

IV. The Effect of Dance on Greek Poetic Form

The multiplicity of rhythms which we find in Greek lyric poetry is a reflection
of the defining power of the dance. Multiplicity of rhythm, according to the
author of the De Musica (21. 1 138B), was a characteristic of the old lyric. Later
poets (by which he probably means those who wrote after the fifth century)
achieved a sense of variety by music rather than rhythm. Multiplicity of
rhythm was traditional. Aristotle has a shrewd crack at those who want to
regiment culture : earu p,ev yap ojs ovk eorai Trpouovcra ttoAls . . . wairep kolv et tls
ttjv avpL<j>coviav iroirjcreiev 6pio<j>coviav rj tov pvOpiov fidaiv pLiav {Pol. I263 b 33~5). If
this is read in the light of Plato's attack on the variety and change of rhythm in
poetry, and his espousal of simple, stately rhythms {Rep. 3. 3996-401 a), we
will see that Aristotle has his former tutor in mind. Such simplicity is quite
different from the indeterminateness in English blank verse and, to a lesser
extent, in Greek spoken iambics.

Choral lyric is more complex in rhythm than solo song, and the choral lyric
of drama exhibits more variety of rhythmical combination than choral lyric
outside drama. This is because dance meant greater complexity, and the
dramatic role of the dancers in Greek drama resulted in changing rhythmic
patterns.

One very important stylistic development, which spreads to all kinds of non-
dance poetry, is the balancing of two lines of roughly similar length. Poetry
composed in this way is usually said to exhibit "parallelism of members'. The
form arises from the set-up of the group dance in which one chorus answers

1 Hugh Lloyd-Jones has plausibly sug- and that the author of our fragment was
gested (Estudios sobre la tragedia griega [1966], a lyric poet of the late fifth century.
18) that there was more than one Pratinas

GREEK DANCE 267

another, or a leader answers and is answered by a chorus. 'Parallelism of
members' finds its formal expression in the metrical line divided into two
parts. In fact, this balancing and combination of two parallel 'members' into
one unified phrase is an extremely widespread practice in all kinds of ancient
and primitive poetry. Examples are the Greek dactylic hexameter, the Latin
Saturnian, the division of Sanskrit verse into two and four parts, the balancing
parts of the Old Persian Avesta line, the symmetrically two-part eight-
syllable line of Slavic lyric, and the two-part ten-syllable line of Slavic and
Russian epic. Of course in the study of metrical form there is so much room for
juggling with numbers that more detailed study is necessary before definite
conclusions can be reached.

Clearer and more persuasive are those examples of parallelism where there is
not merely metrical balance but also a balance of content. This sounds abstract,
but is admirably demonstrated by the Hebrew Song of the Well :

Spring up, O well ; sing ye unto it ;

The princes digged the well, the nobles of the people digged it,

By the direction of the lawgiver, with their staves. (Num. 21 : 17-18)

This is not in any way peculiar to Hebrew poetry : it is also found, for instance,
in ancient Egyptian liturgies. 1 Nor does it seem to be peculiarly liturgical,
though naturally it has a splendid dignity when so used :

From the blood of the slain : from the fat of the mighty
The bow of Jonathan turned not back : and the sword of Saul returned
not empty. (2 Sam. 1 : 22; cf. the whole passage, 19-27)

Parallelism, then, is a form in which metre and content coincide. The
hexameter became more sophisticated : the words often overrun the metrical
division, and lines which divide neatly into two sense-wholes, though not so
infrequent in Hesiod as in other poets, are censured by the Greek commen-
tators. 2 However, we can see the hexameter in its more primitive shape in one of
Sappho's wedding dance-songs :

ii/joi $r) to fJLeXaOpov deppere reKroves avSpes-
ydpLppos €ts to* 'Apevi- avopos /xcyaAcu rroXv pie^cov
rreppoxos <*>S or* aoioos- 6 Aecrftios dWoSdrroLatv.

(frs. 1 1 1 + 106 LP, refrain omitted)

1 Discussed by A. Erman, The Ancient dgova S' eirraTroS-qv p.dXa yap vv rot

Egyptians: a sourcebook (Eng. trs. 1966 = dpp.evov ovrat.

The Literature of the Ancient Egyptians 2 ), lx- el he Kev oKTairoSrjv, dtro /cat o<f>vpdv Ke

lxi; T. E. Peet, A Comparative Study of the rdp.oio.

Literatures of Egypt, Palestine and Mesopotamia (Hesiod, Works and Days 423-5).

(iqqi), 54-5, 63. Examples in J. B. Pritchard x „ , „ ,

edf), Ancient Mar Eastern Texts 2 (1955), ^ l ? MW l ™> € ™ S ™™' fio€ S ° ? KM

365-81. . T*™*' • ' a « ' "

a That is, lines divided by sense into two W*™ » amunprnaOai' irapa epya

parts, and separated by sense from the poeaaiv .

preceding and following lines. For example : (ibid. 453-4). Such lines are criticized by

». ^ / S / *y Dionysius of Halicarnassus, De Comp. Verb.

oXfiov fiev Tpnrobrjv rapveiv, vnepov be 7 > r

Tpi7rr)xw,

2 68 J. W. FITTON

Sense and metre go together in the traditional Swallow song from Rhodes :

kolAols ojpas dyovaa- kolAovs €vlolvtovs,
ini yaarepa Aeu/ca- em vchra jiieAaiva.
naAddav av 7T/)o/cu/cAef e/c ttlovos olkov
oivov T€ SeTraarpov rvpov re Kavvarpov. 1

The larger form of the choral ode must have been related to the general set-
up and evolutions of the dance. On the smaller scale there is the correlation of
dance-step with metrical foot. This can never be entirely certain, but at least
we can try to relate the two by considering the rhythmical pattern of the poetry
as an accompaniment to dancing. The first question is whether there was a
one-to-one correlation between metrical foot and dance step. Where the metre
allows plenty of substitution, it is unlikely that one syllable corresponded with
one foot movement. When the metrical feet go in twos, there is a presumption
that alternate feet are being used. Anapaests were marched to (see e.g. Carm.
Pop. 10 and 1 1 = Page, P.M.G. 856 and 857), and it seems that the feet made
a step with the strong beat of each anapaest. Lyric dactyls, which went in ones,
not twos, avoided substitution, and seem to embody a quicker motion than
anapaests ; so presumably they had a one-to-one relationship with the dance-
step. This means that they went

— W W — W W — WW — WW

LrliMrLrlZMr

The effect would resemble a waltz, though the music would not be in waltz
time.

Among other types of poetic organization which seem to come from the set-
up of the dance-song is the cadence. The musical rallentando, marked in Greek
poetry by the use of long syllables, 2 corresponds to the graceful slowing down
of the dancers. Of similar origin is the refrain, which was originally the chorus'
response to the soloist, but degenerated into nonsense syllables. 3 It comes in
with an accentuation of the main dance-rhythm after the performance of the
soloist, which may have been irregular in time, especially when improvised.
We may compare Irish usage, where 'lilt' means 'to sing nonsense syllables,
especially as accompaniment for dancing'. 4 In the work-song the refrain
accentuates the concentrated effort, as for example the 'Yo-heave-ho' of
British sailors and the pwrnraTTai of Athenian (Aristoph. Ran. 1073).

The connection between poetic form and dance is particularly evident in
Greek metrical terminology. Ancient scholars were always ready to use an
analogy — "iambic is like walking' 5 — but in such cases as trochaic (literally the
'running' metre) and dochmiac ('zigzag') the name itself referred to movement.

1 Carm. Pop. 2 = Page, P.M.G. no. 848, aTTayOtirai (line 12), and then some im-
1-9. The first line falls outside the pattern. provised patter equivalent to an epode
Note the rhyme between the last two phrases. stuck on to the dance-song.

2 Note the term avp/xa. Used by Dante 3 I cannot suppress a suspicion that
(De vulgari eloquentia 2. 10. 4) in the sense of *Iolkxos was about as significant as 'Whacko!'
'coda', i.e. cadence, it also meant 'trailing *■ Funk and Wagnall's Standard Dictionary
movements' (Mesomedes, Hymn, in Solem of Folklore, Mythology and Legend ii. 623.

23) and 'lengthened musical sounds' 5 Marius Victorinus (Grammatici Latini,

(Ptolemy, Harm. 2. 12). In the Swallow ed. Keil, vi. p. 44, 28) derives the term
Song, quoted above (n. i), we have ovk 'iambic' d-n-o rov livai fidSrjv.

GREEK DANCE 269

The smallest analytical terms are Toot' (novs) and 'step' (jSaat?) . The foot goes
4 up' and 'down' and consequently has a 'rising' (apais) and a 'putting' (Oecris).
The feet were organized into kojXol, literally 'limbs', and the kwXol were
organized into a -n-epioSos or 'going- round'. (A modern parallel is the term
rondo which now denotes a musical form, but was originally a type of dance.)
Finally the TreploSot were organized into strophe and antistrophe, literally
'turn' and 'counter- turn'. 1

The Greeks recognized the connection between dance and the form of song.
Pindar, contrasting his dithyramb with that of old (Dith. 2. fr. 70b Snell 3 , 1-5),
refers to the old sort as 'strung out like a rope' (ax OLVOT€V V s ) 9 an< ^ wnen ne savs
that his is set up properly (the allusion is to a set- or formation-dance) and
provided with strophes, it is clear that he associates the old form with a
straggling dance, perhaps a maze-dance.

The view that dance gave definiteness to music is often expressed by musico-
logists. In the study of folksong the distinction between songs sung in parlando
style or with rubato and those in tempo giusto led to the conclusion that the
regular, strict style is especially characteristic of dance-songs. 2

V. Dance and the Rhythm of Greek Poetry

In ancient Greek usage, rhythm is a property of both dance and poetry. Thus
Aristotle (Poet. 1447*21-2) says that poetry works through rhythm, word (or
sense), and tune; dancing, if unaccompanied, achieves expression by rhythm
alone. Aristides Quintilianus (1. 13, p. 32 Meibom) says that the things that
are 'rhythmized' are the movements of the body, the tune, and the words.

The Greek theorists realized that rhythm was different from metre. The
distinction is at least as old as the fifth century (Aristoph. Nub. 638-54). The
later Greek analysts were divided into two schools, one distinguishing metre
and rhythm, the other treating them together. Much of our theory of metre
comes from grammarians such as Hephaestion who were concerned with longs
and shorts and little else : with metre, not with rhythm.3 Modern theorists who
assume that Greek poetry is simply a matter of longs and shorts are thus at
a disadvantage from the start in that they have to ignore the opinions of the
ancients. 4 Great reliance is put on the definition of Aristoxenus (Rhythmica fr. 1

1 Definitions of these terms can be found more obscure than ever. Some believe that
in any handbook on Greek metre, e.g. D. S. it is part of a treatise by Aristoxenus.
Raven, Greek Metre (1962), 18-19, 24-5. For 4 Work on the 'rhythmical' structure of
a full discussion of the problem of termino- Greek lyric poetry at the present time is
logy, see L. P. E. Parker, Lustrum xv (1970), concentrated on the analysis of patterns of
48-58 ff. long and short syllables and the occurrence

2 G. Herzog in Funk and Wagnall's of word-breaks, caesuras, etc. See the writ-
Standard Dictionary of Folklore, Mythology and ings of A. M. Dale (e.g. The Lyric Metres of
Legend ii. 1041. Greek Drama 2 , 1968) and Paul Maas {Greek

3 Hephaestion (ed. Consbruch, 1906) is Metre, Eng. trs. 1962) and, on a more
probably the most important metrical popular level, D. S. Raven {Greek Metre).
theorist. Lists of metrical feet occur in most An exception to the trend is H. D. F. Kitto,
of the Greek and Roman grammarians in who argues for the doctrine of Aristoxenus
discussions that make no reference to and the rhythmici ('Rhythm, metre and black
rhythm. For the metrico-rhythmical magic', C.R. lvi [1942], 99-108). [The
approach see especially P. Oxy. no. 9 (vol. i. difficulty is that while we know some of the
pp. 14-21), republished as no. 2687 (vol. rhythmical 'rules', we do not know how
xxxiv. pp. 1 5-25) with considerable additions extensively they were employed, and modern
which clarify certain points but make others rhythmical interpretations are often heavily

270 J. W. FITTON

Westphal) that rhythm is an arrangement of times. But Aristoxenus did
not say what is being arranged into times : a clock might be said to be an
arrangement of times, but it is also a set of moving wheels and springs in a
case. A dance consists of times (quick and slow), but it is also the movement
of muscles.

Plato is concerned with providing a general point of view of poetic form, not
with examining the length of syllables. His use ofpvOfios is less technical and as
such more likely to be the ordinary usage. In the Laws (2. 653 d-e) he traces
pv9nos to two origins : the chatter of children, which leads to song, and their
ceaseless movement, which leads to dancing. Later in his argument (660 a-b)
he is dealing with three things — words, music, and dance. He then (664 c-
665 a) reverts to two — pvOfios (defined as the arrangement of movement) and
music (the high and low of a voice) . I am not concerned with Plato's incon-
sistency, but with its origin in the fact that pvdfios means both body-rhythm and
the rhythm of poetry. A similar confusion is found in Aristotle's Poetics. At first
poetry is divided into rhythm, word, and tune (1447*21-2: pvOpos, \6yos,
apfiovia), then into rhythm, song, and metre (i447 b 25 : pvdfios, /xe'Ao?, iierpov).
In the first definition, rhythm is considered abstractly, in itself, as opposed to
the sense of words or the order of notes. Then, in a less abstract way, we have
the division into rhythm (i.e. body-rhythm as shown in the dance), metre, in
which rhythm is combined with words, and song, in which rhythm is combined
with musical notes.

In the Philebus (17 c-d) Plato is discussing musical training. The two essen-
tials are of course melody and rhythm. The sentence runs :

eireihdv Adprjs . . . ev T€ rat? Kivqaeaw afirov craj/xaro? . . . ivovra 7rddrj yiyvo-

/xeya, a 817 oV apidyLcbv ^erprjOevra 8eiv av cf>aai pvOpovs kcll /xerpa €7rovo/xa-

t^iv . . . — orav yap avrd re Adprjs ovtoj, rore iyevov ao<f>6s.

Rhythm then is primarily body-movement and only secondarily the reflex
of body-movement on the words of Greek poetry. The Greeks, unlike modern
theorists, were able to see this because of the vigour of their song-and-dance
tradition.

This definition of rhythm is supported by two general considerations. First
there is the development of the sense of rhythm in the ordinary child. From
games and songs, which are inseparable from bodily activity, he learns to
participate in a sing-song in which the syllables correspond with a vigorous
muscular effort. The abstract sense of rise and fall which allows him to scan
Shakespeare or Keats comes at a much later stage. Secondly, we have in
corroboration the results of investigations into the psychology of music which
show the clear, though complex, relationship between rhythm and body-
effects. 1 Rhythm has a biological value in that it lessens the expenditure of
energy and produces more effective action and a feeling of satisfaction.
The feeling of power produced by rhythm, as by a dream of flying, is part of
a motor attitude : 'rhythm is never rhythm unless one feels that he himself is
acting it'. 2 Rhythm is not mere periodicity, for that would never make danc-
ing beautiful.

dependent on the bar-structure of western her to construct rhythmical patterns —

European classical music. Thus Miss Dale, J. H. C]

for example, realized that rhythm was more

than a cold analysis of longs and shorts, but x G. E. Seashore, The Psychology of Music

felt that the lack of evidence did not allow (1938), ch. 12. 2 Ibid., 142.

GREEK DANCE 271

The argument against the bodily basis of rhythm has recently fastened on the
Greek word pv9p,6s. The old derivation was from pew, c flow\ I am not com-
petent to judge whether this is correct, and in any case the new theorists have
not argued about its etymology. They provide us with a number of uses of the
word pvdfios and its compounds where the notion of movement is not present.
Hence, they say, pv9p,6s means pattern, not motion. 1 The weakness of this
argument is obvious. That ISea, 'form', was originally a visual form is quite
clear from its etymology ; that it came to be applied to abstract concepts does
not invalidate that etymology. Moreover, the few usages that have been quoted
do not in fact tell against the overwhelming majority of uses of pvdfios. In the
case of ^erappvdfjLL^eLv 'transform', and /xerapvcr/xow, 'reform', where there is no
question of body-motion, we have simply an extension of meaning from bodily
effects to mental effects. Souls could have their rhythm changed, 2 and the
change would result in a different style of action.

The practical result of this new interpretation of pv9p,6s is that Greek lyric
poetry, unlike any other, becomes static. But the very application of the word
'static' to the free and flexible forms of Greek lyric surely indicates a fundamen-
tal misunderstanding.

Part of the trouble is the failure to see Greek poetry for what it is — not a
mere set of words on a printed page, but a real-life performance involving the
whole personality. To a Greek, actions without dance were often pervaded by
the same 'rhythm' as the dance. Ordinary life had its rhythms, in the way that
a gentleman walked and the way that a slave walked, for example (Alexis fr.
263 K 1-3, ap. Athen. 1. 2 id). Oratory depended considerably on the acting
powers of the speaker, his gestures and bodily emphasis. 3 The epic poem, in the
hands of the rhapsodist, was as much a drama as a narrative. 4 All this was
usually taken for granted by Greek authors, though occasionally we have
a hint of it, as when the author of the Aristotelian Problems says (o,i9 b 26-37)
that what is heard has character (ethos) because it has movement.

If this view of rhythm is correct, much modern analysis of poetical form can
be seen to be operating with terms that imply a traditional but only dimly
understood metaphor. We hear of the movement of poetry, its grandeur, its
quickness, its stateliness, and so forth. We talk as if there were a person acting
out the lines. Clearly the words on the printed page do not possess these
qualities. Nor should it be thought that rhythmical effects reside solely in the
voice of the reciter. The quickness of rhythm is the quickness of movement, and
this can be hinted at by the voice of the reciter ; but it is not the same as
quickness of utterance.

To turn from the modern aesthetics of rhythm to ancient scholarship might
seem to be a descent. On the contrary, although the grammarians and writers

1 See esp. Arist. Met. A 4. 985*16; E. 2 Because, according to Greek doctrine,

Harrison in P.C.P.S. 1937, 11, and W. the rhythmical movements of the music

Jaeger, Paideia i 2 (1947), 126. It contradicts, were paralleled by the movements of the

however, ancient definitions of rhythm, for soul (see the fragment of Damon cited above,

example that of Plato, who defines it as 17 p. 259 n. 8), and the movements of the soul

Tijs KtvTJaeats raits (Laws 2. 6646-665 a). changed as the character of the music

For the older interpretation which was in changed.

keeping with the ancient definitions, see 3 r. g. Austin, Cicero: Pro Caelio* (i960),

E. A. Sonnenschein, What is Rhythm? (1925), p. 58. For the physical side of oratory see

15-16; and for the many meanings of pvdpos, Quintilian 1 1 . 3 . 65- 1 36.

E. Wolf, Wiener St. lxviii (1955), 99-119. + G. F. Else, Hermes lxxxv (1957), 34-5.

272 J. W. FITTON

on music did not always understand what they were saying, they do show a fair
grasp of rhythmical matters. For example, they thought spondees stately
because they went with a lengthy and stately movement, and short syllables
nimble because they went with nimble movements. 1

Another approach was that of nineteenth-century writers on Greek lyric
poetry who tried to reduce the words to the bar-structures of modern music.
Typical examples are the metrical analyses in Jebb's editions of the tragedies of
Sophocles. This attempt is largely rejected nowadays. However, the dynamic
beat is something which can hardly be left out of any musical analysis. Of course
the nature of this beat is much disputed. It appears that we often imagine a
beat where objectively there is none. But the imaginary beat is derived from
experience of actual beats in a similar position. It is indeed very difficult for any
musical person to think of music entirely without beat. We would therefore
assume that Greek lyric poetry, considered not as marks on a page but as a
performance with music, must have had a beat.

Yet this is a point most strenuously contested. 2 The main argument against
the beat is the classification of the Greek language as a language without stress.
The language, we are told, knew nothing but longs and shorts and high and low
tones ; there was no strong/weak pattern of the kind that is to be found in
English and modern Greek. What are we to make of this dilemma? We may
admit that it is very probable that ancient Greek was predominantly without
stress. What is insecure is the inference drawn from this. A stress language like
English clearly makes use of quantity when poetry is set to music, since the
syllables are given quantity and fitted into a bar structure. Conversely, non-
stress languages use a beat in lyrics. Modern Welsh uses tone and quantity
more often than stress, yet undoubtedly many Welsh songs, for example 'Men
of Harlech', have a strong beat. French too is said to be mainly a non-stress
language, but in French songs, especially dance-songs, a prominent beat makes
itself felt.

What is the evidence for beating time in ancient Greece ? It was done by feet
and hands, by instruments, and by other objects, including a wooden clapper.

First, feet and hands. In Homer we find a singer surrounded by a throng of
people who beat time with their feet. Thus the boy sings the Linos song while
the throng stamps together to keep time (//. 18. 569-72). The Greeks noticed
the satisfying effect of the beat as the dancer's foot hit the ground. Thus in
Hesiod a 'lovely thud' (iparos 8ovttos) arose in accompaniment as the Muses
sang (Theog. 70). Callimachus (Hymn 3. 246-7) describes a female chorus
making a noise like castanets with their feet. After their victory over the
Persians the Greeks sent up a hymn to Paean, and they made the beat in time
with high-hitting dance-measures (cru/x/xerpot 8' irre- j ktvtt€ov noScbv \ vijjiKporois
X opeiais [Timotheus fr. 15 ( = Page P.M.G. no. 791), 197-201]). Callimachus,

1 Thus spondaic songs accompanied be no theory of "ictus" in the sense of purely
solemn libations and proceleusmatics formed metrical stresses, since there is no evidence
the rhythm of lively pyrrhic dances (Aristides whatever for its existence in Greek.' The
Quintilianus 1. 15, p. 37 Meibom). opposite view is taken by L. Laurand, 'Sur

2 See, for example, A. M. Dale, op. cit. quelques questions fondamentales de la
(above, p. 269 n. 4), 5: 'There is no vestige metrique', Rev. de Phil. ser. 3, xi (1937),
of evidence that dynamic stress had any 287-9, who gives as one reason for believing
structural significance in Greek verse in 'ictus' the louder tone which is physio-
rhythm before the imperial period' ; and logically inseparable from the thesis of the
again in Lustrum ii (1956), 20 : 'There should dance.

GREEK DANCE 273

in a description of a festival, says that there are two groups — the men sing a
song of Olen while the maidens beat the ground with their feet {Hymn 4. 304-7) .
Or again, there is the invitation to the dance in Aristophanes :

aAA' aye Ko\xav 7rapa/Z7rt;/aSSe X € Ph ttoSolv re TrdSr)

a tls eAa<f>os- Kporov S' dfia ttoUl xopaxfreXrjTav. {Lys. 1316— 19)

When we find an obscure phrase in Theocritus (18. 7-8) describing dancers as
(literally) beating-in with interwoven feet into one tune {deihov S' dfia iraoai is
€v fieXos iyKpoTeoujai / ttoggi TrepLirXeKTOLs), the reference may again be to the
regularizing effect of the beat on the tune.

The use of hands as a rhythmical accompaniment to dancing is referred to in
a passage of Euripides {Suppl. 72-3), where the hands of the serving-maidens
are resounding and the beat is in time with (literally combined with) the song.
As in Callimachus, there are two choruses, one beating time for the other.

Some other ways of providing the beat are worthy of notice. Nausicaa leads
the song as her band of maidens throw a ball about {Od. 6. 100-1). The
exhibition dance of two young men at the court of Alcinous is accompanied by
the snap of forefingers from other youths standing by, according to Athenaeus
(1. 15 g-d, referring to Od. 8. 379). Vase-paintings give evidence of clapping,
stamping, and the rhythmical use of hands in the dirge. 1 Work songs, such as
the wine-press songs sung to the treading of the grapes, marching songs, and the
songs of children's games also seem to have had a beat. 2

Furthermore, the beat was sometimes provided by instruments. It appears
from Pindar (Pyth. 1 . 1-4) that the lyre was used for controlling the dance-step.
It gave the cue to the singer with its strumming. This effect is presumably
referred to in Horace's direction to the dancers : Lesbium servate pedem meique j
pollicis ictum {Odes 4. 6. 35-6). The strumming is rendered by the word ro-
</>XaTToOpar in Aristophanes {Ran. 1 285-95) . A character asks whether Aeschylus
got the <j)XaTTo6par from the songs of the well-drawer (loc. cit. 1296-7). I take
this to be a reference to the strenuous rhythm of a work-song in which the beat
would have been heavier and more monotonous than in the normal lyric.

Of wind instruments, the aulos was especially used to set the dance in
motion. According to Longinus, 3 it is a more 'dancy' instrument than the lyre,
and it forces men to move in rhythm. It was, however, accused by musical
conservatives in fifth-century Athens of jazzing up and ruining the tune and the
rhythm (Pratinas, fr. 1 = Page, P.M.G. no. 708). The panpipe is similarly
mentioned by Callimachus {Hymn 3. 242-3) as being played to keep the feet of
the dancers in time.

The drum was used extensively in the ecstatic dances of Dionysus and
Gybele. 4 The priest of Gybele, as we saw (pp. 9-10 above), made the lion dance
with the drum. Cymbals too were used in the rites of these two gods. Lest the
magical word 'oriental' be hurled at this instrument, we may call attention to

1 M. Emmanuel, op. cit. (above, p. 258 3 De Subl. 39. 2; cf. Horace, Ars Poet.
n - 2 )> 255 (clapping, stamping); L. B. 202-4: the aulos was useful adspirare et
Lawler, op. cit. (above, p. 254 n. 4), 54-6 adesse choris.

(dirges) . 4 xhe drum as the invention of Dionysus

2 For wine-press songs see Callixenus of and Cybele: <Eur. Bacch. 59 with Dodds
Rhodes ap. Athen. 5. 199A; cf. Longus 2. ad loc. In fr. 586 N 2 Euripides speaks of
36. The feet of marchers obviously provided Dionysus os av* "IBav \ ripirer at ovv /xarpt
a beat to regularize their songs. For the beat ^t'Aa / rvinrdvwv eV taga??. See also Diogenes
(Kporos = ? clapping of hands) accompany- trag. fr. 1 N 2 (= fr. 1 Snell) 3.

ing a game-song see Pollux 9. 123.

274 J- W. FITTON

the evidence of the rites of Zagreus in Crete and the worship of Demeter
Achaea, in both of which cymbals were used. 1 Castanets, by contrast, were
distinctively Greek. According to Dicaearchus (fr. 60 Wehrli, ap. Athen. 14.
636 c-d) the instrument was adopted and used very extensively by Greek
women to accompany dance and song. They were used in the worship of
Dionysus and Demeter. 2 In the Homeric Hymn to Apollo, the handmaidens of
the great god of the lyre are described as imitating a castanet-player giving time
to the dance {Hymn. Horn. 3. 162-3). They also seem to have been connected
with Artemis (frag. lyr. adesp. 37 — Page, P.M.G. no. 955). A cruder form of
castanets were the bits of pot used to give the beat to the dance of the Muse of
Euripides in Aristophanes' parody, 3 and Didymus tells us (ap. Athen. 14.
636E) that pots and shells were actually used to provide a rhythmical sound for
dancers. It sounds like a custom of poorer folk.

The use of other objects to make a steady jingling or banging noise is a
regular practice in folk-song, and the weapon-dances of the Greeks involved
the rhythmical clashing of shields or quivers. 4 A wooden clapper (Kpovire^a) ,
strapped to the aulos-player's foot, was used to give the beat for dramatic
choruses. 5 Pickard-Cambridge absurdly says that it was used to give the first
note of the choral song. 6 It is rather unlikely that anyone would strap a piece of
wood to his foot in order to stamp once, and once only. The function of the
clapper was surely linked to the need for organized movement in the chorus.
Finally, there is the aavlSiov, a wooden board or plank. In a papyrus dating
from the third century B.C. there is an attack on musical theoreticians who are
described as performing a musical experiment. 7 They hit the plank in time with
the sounds of a stringed instrument. They are, in other words, beating time.
The author says that they are excited and off the beat (irapa tov pvOpov) .

The evidence I have quoted is, I am well aware, inadequate as a full picture
of musical procedure. But it does show that the Greeks did beat time, that foot,
hand, instrument, armour, and clapper were used to impart a dynamic pattern
to song and dance. It is of course possible to draw theoretical distinctions
between what was done in the music and what was done in the poetry. If,
however, we believe that the song-and-dance which we call Greek choral lyric
was an integrated performance (and many pay lip-service to this without
seeing the consequences), then it is not too bold to say that rhythm in the
ordinary English and Greek sense of the word — the regular throb and pulsation
of bodily movement — pervaded it throughout.

University of Exeter J. W. FiTTONf

1 For cymbals in the rites of Dionysus see 4 Shields were clashed in the Persian
Aesch. fr. 57 N 2 6; in the rites of Cybele, dance (Xen. Anab. 6. 1. 10; cf. the whole
Diogenes trag. fr. 1 N 2 (= fr. 1 Snell) 4; passage 6. 1. 5-13); for clashing of quivers,
in the rites of Zagreus, Firmicus Maternus, Callim. Hymn. 3. 246-7.

De Err. Prof. Rel. 6. 5 (Liber = Zagreus) ; in 5 The evidence on the Kpovirela is con-

the rites of Demeter Achaea, scholiast to veniently assembled in Pickard-Cambridge,

Aristoph. Acharn. 708. op. cit. (above, p. 261 n. 1), 262 n. 4. For

2 For castanets in the worship of Dionysus the aulos-player beating time with his foot
see Eur. Cycl. 204-5; in the worship of in non-dramatic choruses see Lucian, Salt.
Demeter, Pind. Isthm. 7. 3-4 with the 10; cf. 63 and 83.

scholiast. Cf. also above, p. 257 n. 1. 6 Op. cit. (above, p. 261 n. 1), 262.

3 Ran. 1304-7. In Euripides' Hypsipyle the ? p. Hibeh 13. 27-31 (vol. i. pp. 45~ 8 )» dis "
heroine accompanied the song to her child cussed by W. D. Anderson, op. cit. (above,
with KporaXa (fr. 1. ii. 8-14 Bond). p. 265 n. 3), 147-52.