Rhythmical interweaving: epiplokē in ancient Greek metrical theory

In the frame of the PENELOPE project my aim is to explore instances of weaving imagery in archaic and classical Greek literature as revealing a distinctive mode of composition, one that reflects fundamental features of ancient textile technology: the structure of the weave acts in antiquity as a paradigm for pattern generation ‒ every geometric or figurative motif emerging on the fabric resulting from an ordered crossing and combination of discrete elements, the threads, regulated by numerical relationships. In early Greek thought, where the pattern created (or better ‘discovered’, as Indra Kagis McEwen puts it) by the craftsman was conceived of as an instantiation of the order of the universe (Greek kosmos), it represents no big surprise to find cosmic and poetic weaving as ways of describing and conceptualizing complex structures and their constitutive architecture (here we have tried to trace some ramifications of archaic Greek weaving imagery encompassing the cosmos, the poem/song and their reciprocal relationship).

There seems to be a persistent association in Greek literature between weaving (and the related techniques of plaiting, braiding, stringing) and poetry/music (these being two domains largely coextensive in the archaic and classical Greek ‘song culture’). The range of this interplay is surprisingly broad; it encompasses, for instance, i) perceived technological/mechanical/auditory analogies between the warp-weighted loom and string instrument, in turn reflected in terminological convergences (see here); ii) etymological and semantical derivation of key-terms designating the poem/song (or its structural features) from concepts of textile technology; iii) a sustained pattern of imagery coupling weaving and singing (mostly as singing while weaving); iv) a sample of metaliterary metaphors illustrating the performance and composition of a choral song in terms of weaving (the chorus singing ‘I am weaving a song’, or the like); v) textile imagery as reflecting the pragmatics of performance: woven garments were offered to deities during cultic choral songs, and choristers wore crowns and woven headbands that the poem presents as metaphor for the performed song.

So why weaving? What’s the complex structure or arrangement of elements that the analogy with weaving suggest being an underlying feature of archaic Greek choral poems? Is it the harmonic and synchronic integration of song, music and dance in the performance? Or the compositional architecture of the poem, where different motifs and themes are interlaced in complex patterns of rings and symmetries?

There is a further structural layer grounding the whole of ancient Greek poetry – the rhythmical pattern of meters combined into sequences: in the case of choral songs, this was closely connected to the (now lost) melodic and harmonic features of the music, and to the choreography of the dance. A first, fundamental characteristic that Greek meter shares with weaving is its binary nature: the opposition of long and short elements (reflecting syllabic quantity); metrical patterns are obtained through the combination of long and short elements in certain sequences. One possible association between ancient Greek meter and rhythm, song, and weaving has been explored by Anthony Tuck based on the practice, still attested in traditional weaving contexts, of mapping into rhythmical songs numerical information related to the patterns to be woven on the fabric. Thus, traditional rhythmical songs would encode pattern-related data for the weaver, and the pervasive association of weaving and singing in both archaic Greek poetry and the Rig Veda suggests, according to Tuck, the hypothesis that the origin of Indo-European metrical poetry may be connected to just such a tradition.

In this post I would like to point to another potential area of contact between weaving and the system of ancient Greek meters and rhythms. The association is in this case explicitly drawn by ancient metrical theory: metricians gave the name epiplokē (ἐπιπλοκή) ‘interweaving’ to a mode of meter-generation within specific rhythmical types. The input for this post originates in the material presented and discussed at the Summer School in Greek metrics and rhythmics, held yearly at the University of Urbino by scholars of Greek metrics and devoted in 2017 to the topic of epiplokē. In its original formulation, the theory of epiplokē arises from the observation of the formal relationship between pairs of meters (Greek metron, plural metra) falling within a common rhythmical genos (this being characterized by the ratio between the thesis [‘beating of the (dancer’s) foot’] and arsis [‘lifting of the (dancer’s) foot’]). Through an abstract mechanism of ‘interweaving’, effected by the addition or subtraction of one quantity at beginning or end of a metron, such metron is transformed into another – the two being different species (eidē ‘forms’) within a rhythmical genos (‘type’, plur. genē) .

Three are the fundamental genē of rhythm in ancient theory: the ‘even type’ (genos ison), where the ratio between thesis and arsis is 2:2; the ‘double type’ (genos diplasion), whose ratio is 1:2 or 2:1; the ‘type with a ratio of 1,5:1’ (genos hēmiolion), whose ratio is 3:2. The varieties of metra which form the cola (sing. colon , literally ‘members’, recognizable combinations of metra) represent but realizations of those rhythms: each of the nine fundamental meters (called metra prōtotypa in ancient metrical treatises, generally forming pairs of equivalent/specular combinations), generated by a sequence of long and short quantities, is in turn grounded on the opposition of thesis and arsis, and thus made to fit into one of the rhythmical types; the ancient metrical doctrine regards a long quantity (-) as equal to two shorts (⏑⏑), a fact that also allows for a long quantity to be substituted by two shorts (‘resolution’), and the opposite (‘contraction’, when a long takes the place of two shorts). Thus, for instance, the metra prōtotypa of the ‘even type’ (ratio 1:1 or 2:2) are the dactyl (- ⏑⏑) and the anapaest (⏑⏑ – ⏑⏑ -), the ionic a maiore (- – ⏑ ⏑, considered as 2:2 + 1:1) and ionic a minore (⏑ ⏑ – -); the ones of the ‘double type’ (ratio 2:1, 1:2) are the iamb (⏑ – ⏑ ­-), the trochee (- ⏑ ­- ⏑), the antispast (⏑ – – ⏑), and the choriamb (- ⏑ ⏑ -); the metron prōtotypon of the third genos (ratio 3:2, 2:3) is the cretic (- ⏑ -). The cola of archaic and classical Greek choral lyric are formed by combinations of these metra. Ancient colometry, i.e. the disposition of lyric sections in discrete sequences of cola (‘members’, significant combinations of primary metrical units, in turn forming the structure of verses and strophes) by the Alexandrine scholars (3rd century BC) preparing editions of archaic Greek lyric poems and drama, provides the layout of lyric sections we normally find in papyri and medieval manuscripts. This is often the earliest stage we can reach in the history of an archaic and classical Greek poetic text’s transmission, though the possibility that colometry may be traced back to the original author’s architecture of a text, and the crucial question of its relation with the musical score of the piece, are still matters of debate among philologists.

Back to the mechanism of epiplokē in ancient metrical theory: within a single rhythmical type (genos), each of the different eidē ‘forms’ of metra prōtotypa (with the exception of the cretic and its genos having a 3:2 ratio) is created through the transference (initial addition and subtraction, or internal transposition) of a quantity from one metrum to the other of a given pair. Ancient metricians describe three kinds of epiplokē, each encompassing different eidē of equal quantities: 1) the first kind comprehends the two metra of three quantities belonging to the ‘even type’ of rhythm, i.e. iambus ⏑-⏑- and trochee -⏑-⏑; the second type has the two metra of four quantities of the ‘double type’, i.e. dactyl -⏑⏑ and anapaest ⏑⏑-; the third kind explains the genetic relationship between four metra of six quantities of the ‘even type’, i.e. choriamb -⏑⏑­-, antispast, ⏑–⏑, ionic a maiore – – ⏑ ⏑, and ionic a minore ⏑ ⏑ – -. It is easy to visualize in abstract how epiplokē acts; the example given by the scholia to Hephaestion for the first kind of epiplokē is the following, taken from a fragmentary line by Aeschylus in iambic trimeters. Here, the first quantity (syllable) is subtracted to obtain a catalectic trochaic trimeter:

ἰὼ Καϊκε Μύσιαί τ᾽ ἐπιρροαί ⏑ – ⏑ – ⏑ – ⏑ – ⏑ – ⏑ –

ὼ Καϊκε Μύσιαί τ᾽ ἐπιρροαί    – ⏑ – ⏑ – ⏑ – ⏑ – ⏑ –

The mechanism of the third kind of epiplokē is worth being visualized:

– – ⏑ ⏑ – – ⏑ ⏑ – – ⏑ ⏑       ionic a maiore

                                                 – ⏑ ⏑ – – ⏑ ⏑ – – ⏑ ⏑ –     choriamb

                                                    ⏑ ⏑ – – ⏑ ⏑ – – ⏑⏑ – –    ionic a minore

                                                       ⏑ – – ⏑ ⏑ – – ⏑ ⏑ – – ⏑   antispast                       

But what is the practice of choral lyric poets? In what ways is epiplokē relevant to our interpretation of rhythmic and metrical patterns? In one way, cases of epiplokē alert us of the importance of the aesthetic and structural concept of ‘variegation’ (poikilia) in the rhythmic architecture of archaic Greek choral lyric: the juxtaposition of different and opposite eidē within a metrical period of strophe is generally avoided, as these are perceived of as conflicting sub-species of rhythm (descending -⏑⏑ vs. ascending ⏑⏑-). The epiplokē provides us with an interpretative tool to recognize instances of alternation/variegation in lyric versification, especially in cases where ancient colometry (i.e. the disposition of lyric section in papyri and medieval manuscripts) seems to reflect such poikilia: particularly in the choral lyric sections of drama, alternating sequences of rhythm ‘in epiplokē’ tend to be regularized by modern editors, who often incorporate successive cola into one single verse, or apply a rhythmically homogeneous colometry in place of an attested (at least attested by the manuscript tradition) rhythmical poikilia.

The principle of epiplokē can prove an interesting interpretative tool particularly in cases of sequences characterized by rhythmical ambiguity, i.e. when two competing (and both metrically acceptable) colometries may be applied to a sequence of verses, or when the colometry that seems to better follow rhetorical division shows an alternating pattern of cola (often aeolic and ionic) apparently antithetical – and part of the same epiplokē. A typical instance of this is the combination/interweaving of choriambic dimeters (- ⏑ ⏑ – – ⏑ ⏑ – or – ⏑ ⏑ – ⏑ – ⏑ -, i.e. choriamb + iamb) and ionic dimeters (catalectic, ⏑ ⏑ – – ⏑ ⏑ -, or acatalect ⏑ ⏑ – – ⏑ ⏑ – -, or as ‘anacreontics’ ⏑ ⏑ – ⏑ – ⏑ – -) in passages of choral odes of drama. As A. M. Dale (1968, 146-147) noted, ‘the poets in fact delighted in interweaving the two metres and passing from one to the other by way of ambiguous cola […] Where there is no doubt as to colometry and it is merely the attribution to ionic or aeolic which is uncertain a decision is perhaps not of great importance. Choriambic series and ionic series are in ἐπιπλοκή, and perhaps the best we can do is to call such passages aeolo-ionic, whether this be taken merely as a confession of our own ignorance or (as I believe) as a valid indication that such ἐπιπλοκή was objectively present in the actual choral rendering of the passages in question’. As the scholars teaching at the Summer school in Urbino have shown, epiplokē may be used by ancient choral lyric poets and playwrights for both stylistic (variegation) and dramatic (marking particular thematic motifs through a change of rhythm) reasons.

A reorientation of the concept of epiplokē as a sort of overarching principle of Greek versification in the form of a pattern operator of cyclic recurrence has been proposed by Thomas Cole (1988): for him, the iambo-choriambic (- ⏑ ⏑ – ⏑ – ⏑ -) dimeter and the anacreontic (⏑ ⏑ – ⏑ – ⏑ – -) should be considered as two instances, or better two potential way of demarcation, of a ‘octadic’ (‘of eight quantities/syllables’) cycle of epiplokē; thus, a sequence of the kind ⸥- ⸢⏑ ⏑ – ⏑ – ⏑ – ⸥ – ⸢ ⏑ ⏑ – ⏑ – ⏑ – ⸥ – ⸢ ⏑ ⏑ – ⏑ – ⏑ -⸥  may be marked off as either choriambic (aeolic) or ionic, depending on the pattern of verbal demarcation – however, it should be called a iono-choriambic epiplokē. In a forthcoming joint publication with Alex and Ellen, we start exploring potential analogies between this approach to epiplokē and the concept and language of Tidal Cycles, a pattern-operator of rhythmical cycles in electronic music.



Cole, Thomas. 1988. Epiploke: Rhythmical Continuity and Poetic Structure in Greek Lyric. Cambridge MA & London.

Dale, M.A. 1968. The Lyric Metres of Greek Drama. Cambridge.

McEwen, Indra Kagis. 1993. Socrates’ Ancestor. An Essay on Architectural Beginnings. Cambridge MA: MIT Press.

Palumbo Stracca, B.M. 1979. La Teoria Antica Degli Asinarteti. Roma.

Tuck, Anthony. 2006. ‘Singing the Rug: Patterned Textiles and the Origins of Indo-European Metrical Poetry.’ American Journal of Archaeology 110 (4): 539–50.

Leave a Reply

Your email address will not be published. Required fields are marked *