¹ The earliest terms ἄνω/κάτω, borrowed from ordinary language, are attested for example in Pl. Resp. 400b, Aristox. El. Rhythm. §17, 20, 25, 29, 10–16 Pearson. The technical terms ἄρσις/θέσις appear mostly in later texts: e.g. Aristid. Quint. De Musica 31.15–16 Winnington-Ingram (hereafter W.-I.); Bacchius, Ench. §98, 314.10–12 Jan; Lucian, Imag. 14.7, Harm. 1.9; Sext. Emp. Adv. Math. 6.60; but see already Herophilus apud Galen, Syn. Puls. 9.464.1–4, quoted in full in notes 44–5 below.

² Aristox. El. Rhythm. §16, 10.21–2 Pearson: ᾧ σημαινόμεθα τὸν ῥυθμὸν καὶ γνώριμον ποιοῦμεν τῇ αἰσθήσει, πούς ἐστιν εἷς ἢ πλείους ἑνός. See also Frag. Neap. §12, 28.14–16 Pearson. Rhythmical feet differ from metrical feet because the former are defined by the ratio between the number of rhythmical units (protoi chronoi) contained in their arseis and theseis (Aristid. Quint. De Musica 31.8–10, 33.12–13 W.-I.), while the latter consist of fixed sequences of long and short syllables (Aristid. Quint. De Musica 44.11–12 W.-I., App. Dionys. 332.7 Consbruch).

³ Cf. Aristox. El. Rhythm. §§20–1, 12–14 Pearson, with the Aristoxenian Psell. Rhythm. §12, 24.16–19 Pearson and Frag. Neap. §22, 30.25 Pearson.

⁴ Cf. Pl. Resp. 3.399e–400a, Leg. 2.670d; Arist. Pol. 2.1265b35–6, Met. 14.1087b36; ps.-Long. Subl. 39.2 and 41.2; Bacchius, Ench. §91, 312.19–20 Jan; Hermog. Id. 1.1.149–51, 1.6.261–9, 2.3.178–82, 2.12.123; Pollux, Onom. 2.200; Syrian. Ad Hermog. 18.4–9; and Planudes, Rhet. Gr. 5.454 Walz (μεταφωρικὴ δὲ ἡ λέξις ἀπὸ τῶν χορευτῶν· τὴν γὰρ ἐν χοροῖς βάσιν ὁρίζονται οὕτως οἱ μουσικοί· βάσις ἐστὶν ἄρσεως καὶ θέσεως ποδῶν σημείωσις· τὸ γὰρ αἴρειν τὸν πόδα, εἶτα τιθέναι, ἄρσιν καὶ θέσιν ὠνόμασαν· ἄρσις οὖν καὶ θέσις ἡ ἐν τῷ ἄρχεσθαι καὶ λήγειν τῶν χορευτῶν ὁρμὴ λέγεται). A different usage of the term βάσις is attested in late Greek and Latin metrical sources such as Choer. 211.17–212.2 Consbruch, Johannes Rhet. Ad Hermog. 6.239.6–9, and Marius Victorinus/Aphthonius 6.47.3–7 Keil; here βάσις indicates a sequence comprising two or more basic feet (πόδες), i.e. what ancient rhythmicians called a ‘compound foot’ (cf. Aristox. El. Rhythm. §26, 16.6–7 Pearson). As we will see in greater detail below, such terminological and theoretical differences between rhythmicians and metricians were very common; cf. Goodell, T.D., Chapters on Greek Metric (New York, 1901), 6–57 Google Scholar.

⁵ El. Rhythm. §2, 2.5–8 Pearson: περὶ τοὺς χρόνους ἐστὶ καὶ τὴν τούτων αἴσθησιν […] ἀρχὴ γὰρ τρόπον τινὰ τῆς περὶ τοὺς ῥυθμοὺς ἐπιστήμης ἐστὶν αὕτη. See also El. Rhythm. §§8, 11, 12 and 16, 8–10 Pearson, and Aristid. Quint. De Musica 33.12–13 W.-I.

⁶ E.g. Hom. Od. 8.379; Hes. Theog. 70; Pap. Hib. 13.ii.29–31; Lucian, Salt. 10; Ath. Deipn. 636d–e; Philostr. Imag. 1.10.4; Philostr. Jun. Imag. 6.3; Greg. Nys. Contra Eunomium 1.1.17.

⁷ Although particularly strong in English and German scholarship, this approach is by no means limited to these traditions: see, for instance, Maas, P., Greek Metre (Oxford, 1962)Google Scholar, 6; Heberden, C.B., ‘ Rhythmica ’, in Smith, W., Marindin, G.E. and Wayte, W. (edd.), A Dictionary of Greek and Roman Antiquities (London, 1890), 558–65Google Scholar; Williams, C.F., The Aristoxenian Theory of Musical Rhythm (Cambridge, 1911)Google Scholar, 27; Gevaert, F.A., Histoire et Théorie de la Musique de l'antiquité (Gand, 1881)Google Scholar, 2.18; Rossbach, A., ‘Griechische Rhythmik’, in Rossbach, A. and Westphal, R. (edd.), Metrik der griechischen Dramatiker und Lyriker nebst den begleitenden musischen Künsten, Theil 1 (Leipzig, 1854)Google Scholar, 24; Westphal, R., Die Fragmente und die Lehrsätze der griechischen Rhythmiker (Leipzig, 1861)Google Scholar, 99. For further bibliography, see Palmieri, V., ‘ Res metrica: il “piede” e il significato di ἄνω e κάτω’, in Talarischos—Studia Graeca Antonio Garzya sexagenario a discipulis oblata (Naples, 1987), 33–127 Google Scholar.

⁸ This passage seems to refer to a special kind of sandal, called κρούπεζα, which had a wooden clapper attached to the sole and was worn by aulos-players: cf. West, M.L., Ancient Greek Music (Oxford, 1992)Google Scholar, 123. Except for this fragment, the use of kroupezai is mentioned only in late Greek sources (e.g. Pollux 7.87, 10.154; cf. Schol. in Aesch. In Timarchum 126.20–5); by contrast, the Latin scabellum is widely attested both in iconographic and literary sources: e.g. Cic. Cael. 65.13, Suet. Cal. 54.2, Aug. De Musica 3.1.1; cf. Bélis, A., ‘ Kroupezai, Scabellum ’, Bulletin de correspondance hellénique 112 (1988), 323–39CrossRefGoogle Scholar.

⁹ For a discussion of this teaching device, cf. Camp, M.W., Teaching Piano: The Synthesis of Mind, Ear and Body (Los Angeles, 1992)Google Scholar, 5 and 77–82.

¹⁰ In general on this issue, see Kolinski, M., ‘A cross-cultural approach to metro-rhythmic patterns’, Ethnomusicology 17 (1973), 494–506 CrossRefGoogle Scholar and Kvifte, T., ‘Categories and timing: on the perception of meter’, Ethnomusicology 51 (2007), 64–84 Google Scholar; on Indian and African rhythms, see Clayton, M., Time in Indian Music: Rhythm, Metre, and Form in North Indian Rāg Performance (Oxford, 2000)Google Scholar and Agawu, V. Kofi, African Rhythm: A Northern Ewe Perspective (New York, 1995)Google Scholar.

¹¹ E.g. Heberden (n. 7), 558: ‘The groups into which a succession of sounds fall are clearly recognized only when a sound more intense than its neighbours occurs at equal intervals of time. This accentuated part of each group is called by Aristoxenus βάσις, by the earliest writers after Aristoxenus θέσις. The unaccentuated part is called ἄρσις […] All these terms originated in the fact that the accentuated portion of the group was marked by setting down the foot, the unaccentuated by lifting it up.’

¹² For experiments showing that listeners perceive a rhythmical beat even if no difference in intensity or pitch is introduced into metrically regular sequences of durations, see Grahn, J.A. and Brett, M., ‘Rhythm and beat perception in motor areas of the brain’, Journal of Cognitive Neuroscience 5 (2007), 893–906 CrossRefGoogle Scholar and Brochard, R., Abecasis, D., Potter, D., Ragot, R. and Drake, C., ‘The “Ticktock” of our internal clock: direct brain evidence of subjective accents in isochronous sequences’, Psychological Science 4 (2003), 362–6CrossRefGoogle Scholar.

¹³ Cf. the similar conclusions reached by West (n. 8), 133–4 and J. Barris, Silva, Metre and Rhythm in Greek Verse (Vienna, 2011), 18–23 Google Scholar.

¹⁴ On the role of accents in Indian music, cf. Clayton (n. 10), 37–42, 69–70 and 160–1, where the author shows how the use of dynamic accents is negligible when the surface rhythm of a musical piece corresponds to its basic rhythmical structure (tal). By contrast, dynamic accentuation ‘plays a far more important role in defining grouping’ when the surface rhythm runs contrary to the basic one: in this case, the first element of each ‘anomalous’ group is emphasized by means of dynamic stress. See also Deo, A.S., ‘The metrical organization of Classical Sanskrit verse’, Journal of Linguistics 43 (2007), 63–114 CrossRefGoogle Scholar, who demonstrates how Sanskrit metre may be understood better in the light of its chanted nature and argues for the presence of musical syncopes marked by dynamic accents on otherwise weak positions. On Japanese music, cf. Adriaansz, W., The Kumiuta and Danmono Traditions of Japanese Koto Music (Berkeley and Los Angeles, 1973)Google Scholar, 47 and 229.

¹⁵ Cf. W.D. Whitney, Sanskrit Grammar (Cambridge, MA, 1889), 27–34; T.J. Vance, The Sounds of Japanese (Cambridge/New York, 2008), 7 and 115–53. For a discussion of the relationship between the features of a language and the relative musical rhythms, see A.D. Patel, Music, Language, and the Brain (New York, 2008), 97–180.

¹⁶ E.g. C.E. Bennett, ‘What was ictus in Latin prosody?’, AJPh 19 (1898), 361–83, at 371–2: ‘I define ictus, therefore, not as stress, nor as accent, but simply as the quantitative prominence inherent in a long syllable.’

¹⁷ See M.R. Jones, ‘Musical time’, in S. Hallam, I. Cross and M. Thaut (edd.), Oxford Handbook of Music Psychology (Oxford and New York, 2009), 81–92, at 88–90; R.J. Ellis and M.R. Jones, ‘The role of accent salience and joint accent structure in meter perception’, Journal of Experimental Psychology: Human Perception and Performance 35 (2009), 264–80, especially 268 on the key concept of accent salience: in brief, while all kinds of accents attract our attention by means of a ‘local’ change in one specific dimension (duration, pitch or volume), their relative salience indicates the degree to which they actually do so and depends on different factors (e.g. the magnitude of the change and the number of simultaneous accents occurring at the same point in time).

¹⁸ For a partial and largely unsatisfactory attempt, see Palmieri (n. 7).

¹⁹ Cf. New Grove Dictionary of Music and Musicians, s.v. ‘Folk Music’ II.4 (‘Rhythm’).

²⁰ Just as many English heterophones are distinguished by a difference in the position of the stress (e.g. prèsent/presènt, rècord/recòrd), Greek homographs were differentiated by the shape and position of melodic accents: e.g. λῦσαι/λύσαι, βίος/βιός.

²¹ Cf. West (n. 8), 197–204, for a detailed discussion of the intimate correlation between melody and word-accents in non-strophic songs. See also G. Comotti, ‘Melodia e accento di parola nelle testimonianze degli antichi e nei testi con notazione musicale’, QUCC 32 (1989), 91–108; C.H. Cosgrove and M.C. Meyer, ‘Melody and word accent relationships in ancient Greek musical documents: the pitch height rule’, JHS 126 (2006), 66–81. With regard to strophic songs, the general consensus is that the same melody was employed both in strophe and antistrophe: cf. E. Pöhlmann and M.L. West, Documents of Ancient Greek Music (Oxford, 2001), 16. However, A. D'Angour (‘The New Music—So what's new?’, in S. Goldhill and R. Osborne [edd.], Rethinking Revolutions Through Ancient Greece [Cambridge, 2006], 264–83) presents a compelling case for the idea that using the same melody in corresponding strophes and antistrophes was a Euripidean innovation introduced with the staging of Medea in 431 b.c. It is worth noting that, in principle, this hypothesis is not incompatible with Pöhlmann and West's interpretation of Papyrus Vienna G 2315 (= DAGM 3), since this musical fragment belongs to a tragedy first produced in 408 b.c. (Euripides’ Orestes).

²² On the use of stigmai in Greek rhythmical notation, see Anon. Bell. §3, with West (n. 8), 268; R.P. Winnington-Ingram, ‘Fragments of unknown Greek tragic texts with musical notation: the music’, SO 31 (1955), 29–87; S. Hagel, ‘Ancient Greek rhythm: the Bellermann exercises’, QUCC 88 (2008), 125–38.

²³ Cf. El. Rhythm. §17, 20, 21.

²⁴ Cf. A. Barker, Greek Musical Writings 2: Harmonic and Acoustic Theory (Cambridge, 1989), 433 n. 153. Aristides’ debt to Aristoxenus’ treatise is clear from the very beginning of this passage, which presents a drastically condensed version of Aristoxenus’ analysis of the different meanings of the term ῥυθμός, originally offered in Book 1 of the Rhythmics. Cf. El. Rhythm. §1, 2.1–4 Pearson.

²⁵ Cf. Barker (n. 24), 434 n. 155, who draws attention to Aristides’ unusual characterization of arsis/thesis as πάθη of rhythmical durations; but this word was already employed by Plato to indicate precisely some structural features of rhythmical sequences that play a similar role to that of different intervals in a tuning system: cf. Pl. Phlb. 17d.

²⁶ E.g. Rossbach (n. 7), 25 n. 12; H. Feussner, De Rhythmorum Metrorumque Discrimine Poetarum Antiquarum (Leipzig, 1837), 15; J. Luque Moreno, Arsis, Thesis, Ictus: las marcas del ritmo en la música y en la métrica antiguas (Granada, 1994), 135–6.

²⁷ E.g. Eur. Bacch. 687, Cycl. 443–4; Archytas, fr. 1 DK; ps.-Arist. De Aud. 802b40 (ψόφοι become ‘hard’ only if they are ‘violent’) and 803b10–16; ps.-Arist. Pr. 11.19; Arist. De an. 422a23–5; Ptol. Harm. 7.16–19 Düring; Porph. In Ptol. 42.27, 77.29 Düring. For additional references, cf. E. Rocconi, Le parole delle Muse (Rome, 2003), 147 s.v. ψόφος.

²⁸ Arist. De an. 418a11–13: λέγω δ᾽ ἴδιον μὲν ὃ μὴ ἐνδέχεται ἑτέρᾳ αἰσθήσει αἰσθάνεσθαι, καὶ περὶ ὃ μὴ ἐνδέχεται ἀπατηθῆναι, οἷον ὄψις χρώματος καὶ ἀκοὴ ψόφου καὶ γεῦσις χυμοῦ. Cf. De an. 420a20–2 (ἔστι γὰρ ὁ ψόφος κίνησις τοῦ δυναμένου κινεῖσθαι […] ), Cat. 15b1 (ἔστι δὲ ἁπλῶς μὲν κίνησις ἠρεμίᾳ ἐναντίον), Ph. 226b9 (ἐναντίον γὰρ ἠρεμία κινήσει […] ).

²⁹ De an. 422a23–5.

³⁰ De an. 420a26–9: αἱ δὲ διαφοραὶ τῶν ψοφούντων ἐν τῷ κατ᾽ ἐνέργειαν ψόφῳ δηλοῦνται· ὥσπερ γὰρ ἄνευ φωτὸς οὐχ ὁρᾶται τὰ χρώματα, οὕτως οὐδ᾽ ἄνευ ψόφου τὸ ὀξὺ καὶ τὸ βαρύ.

³¹ According to Aristotle, differences in volume belong to the category of μέγεθος/μικρότης (De an. 422b29–30), which is common to all perceptive faculties and is not specific to the domain of hearing (De an. 418a16–20: τὰ μὲν οὖν τοιαῦτα λέγεται ἴδια ἑκάστης, κοινὰ δὲ κίνησις, ἠρεμία, ἀριθμός, σχῆμα, μέγεθος). Cf. also Gen. an. 787a2–10, where Aristotle explicitly distinguishes pitch height (τὸ βαρὺ καὶ τὸ ὀξύ) from loudness and softness (μεγαλοφωνία καὶ μικροφωνία), emphatically denying any unqualified identifications between low-pitched and loud sounds, or the opposite (συμβήσεται τὰ αὐτὰ εἶναι βαρύφωνα καὶ μεγαλόφωνα καὶ ὀξύφωνα καὶ μικρόφωνα· τοῦτο δὲ ψεῦδος).

³² De an. 420b5–33: ἡ δὲ φωνὴ ψόφος τίς ἐστιν ἐμψύχου […] φωνὴ δ᾽ ἐστὶ ζῴου ψόφος […] σημαντικὸς γὰρ δή τις ψόφος ἐστὶν ἡ φωνή. Cf. Aristid. Quint. De Musica 31.6–7 W.-I.: λέγεται [scil. ῥυθμὸς] καὶ ἰδίως ἐπὶ φωνῆς· περὶ οὗ νῦν πρόκειται λέγειν.

³³ E.g. Barker (n. 24), 434 n. 155, G. Moretti, ‘Il ritmo in Aristide Quintiliano’, Musica e Storia 24 (2006), 33–92, at 39.

³⁴ De Musica 30.14 W.-I. (δι’ ἧς εἰς ἠρεμίαν τὴν ψυχὴν περιάγομεν), 67.31–68.1 W.-I. (ὅσα μὲν τῶν αἰσθητῶν ἔς τε ἡδονὴν δελεάζει καὶ ἠρέμα διαχεῖν τὴν γνώμην πέφυκε, ταῦτα εἰς τὸ θῆλυ κριτέον), 69.17–18 W.-I. (ὅπου μὲν τὸ ἠρεμαῖον δηλῶν τῆς ἐπιρροῆς), 76.13–14 W.-I. (τούτων δὲ τὰ μὲν ἠρεμαίως προάγοντα τὸν ἀέρα κἀκ τῶν περὶ τοὺς ὀδόντας τόπων κέκληταί τε ψιλά), 119.30 W.-I. (πᾶν φασι σῶμα τάχει πολλῷ φερόμενον καθ᾽ ὁμοίου καὶ ὑπείκοντος ἠρέμα), 120.32–121.1 W.-I. (τὸ δὲ θῆλυ ὑγρὸν διειμένον ἠρεμαῖόν τε καὶ φυγόπονον).

³⁵ Anon. Bell. §§39–41: ἐν μὲν γὰρ τῷ διαστήματι λέγοιτ᾽ ἂν κινεῖσθαι ἡ φωνή, ἵσταται δὲ ἐν τῷ φθόγγῳ. ἄλλως οὖν λέγεται ἠρεμία φωνῆς παρὰ μουσικοῖς καὶ κίνησις, καὶ ἄλλως παρ᾽ ἄλλοις. ἐπίτασις μὲν γὰρ καὶ ἄνεσις κίνησις φωνῆς, τάσις δὲ ἠρεμία […]. Cf. Aristox. El. Harm. 17–18 Da Rios and Aristid. Quint. 5.19–6.7, with Barker (n. 24), 405 nn. 31–2.

³⁶ Cf. G. Moretti, ‘Il ritmo in Aristide Quintiliano’, Musica e Storia 24 (2006), 33–92, at 79 (‘percussione’); L. Colomer and B. Gil, Arístides Quintilianus—Sobre la Música (Madrid, 1996), 152 (‘golpe ritmico’).

³⁷ Barker (n. 24), 485 n. 156.

³⁸ Particularly telling is the evidence provided by the Anonyma Bellermanniana, where different melodic tropes are defined as types of krouseis (πρόκρουσις, ἔκκρουσις, ἐκκρουσμός: Anon. Bell. §2); cf. also §§18 and 68. For the standard use of the word krousis in musical contexts, see e.g. ps.-Plut. De Musica 1137b–d, 1138b, 1141a–b; Porph. In Ptol. Harm. 35.17–35 and 59.18 Düring. This usage was already established in the Classical age, as is shown by the Hippocratic De Victu 18.13–15 Littré: κρούεται δὲ [scil. τὰ διάφωνα καὶ ξύμφωνα] τοὺς φθόγγους ἄνω καὶ κάτω, καὶ οὔτε τὰ ἄνω κάτω κρουόμενα ὀρθῶς ἔχει οὔτε τὰ κάτω ἄνω. Cf. Rocconi (n. 27), 135 s.v. κρούω.

³⁹ De Musica 10.2–4, 23.21, 27.28, 31.26, 74.19 W.-I. Cf. also 43.10–16, where the term σύγκρουσιν refers to the blending of two syllables.

⁴⁰ De an. 420a30–1: τὸ μὲν γὰρ ὀξὺ κινεῖ τὴν αἴσθησιν ἐν ὀλίγῳ χρόνῳ ἐπὶ πολύ, τὸ δὲ βαρὺ ἐν πολλῷ ἐπ᾽ ὀλίγον.

⁴¹ De Musica 40.20–5 W.-I.: τινὲς δὲ τῶν παλαιῶν τὸν μὲν ῥυθμὸν ἄρρεν ἀπεκάλουν, τὸ δὲ μέλος θῆλυ· τὸ μὲν γὰρ μέλος ἀνενέργητόν τ᾽ ἐστὶ καὶ ἀσχημάτιστον, ὕλης ἐπέχον λόγον διὰ τὴν πρὸς τοὐναντίον ἐπιτηδειότητα, ὁ δὲ ῥυθμὸς πλάττει τε αὐτὸ καὶ κινεῖ τεταγμένως, ποιοῦντος λόγον ἐπέχων πρὸς τὸ ποιούμενον. Cf. Anon. Bell. §38, where the different natures of vocal movement and rest are likened to the role of ‘the producer’ (τὸ ποιοῦν) and ‘what is produced’ (τὸ ποιούμενον)—i.e. the exact same roles attributed respectively to rhythm and melody by Aristides.

⁴² De Musica 5.6–7 W.-I.: περὶ μὲν γὰρ μελῳδίαν ἁπλῶς ἡ ποιὰ φωνή, περὶ δὲ ῥυθμὸν ἡ ταύτης κίνησις, περὶ δὲ τὴν λέξιν τὸ μέτρον.

⁴³ Aristid. Quint. De Musica 31.9–14 W.-I.: καὶ τὰ τούτων πάθη καλοῦμεν ἄρσιν καὶ θέσιν, ψόφον καὶ ἠρεμίαν· καθόλου γὰρ τῶν φθόγγων διὰ τὴν ὁμοιότητα τῆς κινήσεως ἀνέμφατον τὴν μέλους ποιουμένων πλοκὴν καὶ ἐς πλάνην ἀγόντων τὴν διάνοιαν τὰ τοῦ ῥυθμοῦ μέρη τὴν δύναμιν τῆς μελῳδίας ἐναργῆ καθίστησι, παρὰ μέρος μέν, τεταγμένως δὲ κινοῦντα τὴν διάνοιαν. Cf. De Musica 40.20–5 W.-I., where the same outcome is attributed to the effect of rhythm on melody: τὸ μὲν γὰρ μέλος ἀνενέργητόν τ᾽ ἐστὶ καὶ ἀσχημάτιστον […] ὁ δὲ ῥυθμὸς πλάττει τε αὐτὸ καὶ κινεῖ τεταγμένως .

⁴⁴ Gal. Syn. Puls. 9.464.1–4: ὥσπερ γὰρ ἐκείνους [scil. τοὺς ῥυθμοὺς] οἱ μουσικοὶ κατά τινας ὡρισμένας χρόνων τάξεις συνιστῶσι παραβάλλοντες ἀλλήλαις ἄρσιν καὶ θέσιν, οὕτως καὶ Ἡρόφιλος ἀνάλογον μὲν ἄρσει τὴν διαστολὴν ὑποθέμενος, ἀνάλογον δὲ θέσει τὴν συστολὴν τῆς ἀρτηρίας […]. Cf. H. von Staden, Herophilus—The Art of Medicine in Early Alexandria (Cambridge, 1989), esp. 276–84, 354–6.

⁴⁵ Gal. Syn. Puls. 9.464.10–465.9: τούτοις εἰς δύο χρόνους τοὺς πάντας ὁ ῥυθμὸς τοῦ σφυγμοῦ μερίζεται, τόν τε τῆς αἰσθητῆς κινήσεως, ἡνίκα τὴν ἁφὴν ἡμῶν ἡ ἀρτηρία διαστελλομένη, καὶ τὸ λοιπὸν ἅπαντα συγκείμενον ἔκ τε τῆς ἐκτὸς ἠρεμίας καὶ τῆς μετ᾽ αὐτὴν συστολῆς, καὶ τῆς ἐπ᾽ ἐκείνῃ πάλιν ἠρεμίας καὶ τῶν πρώτων τῆς διαστολῆς, ἅπέρ ἐστιν ἀναίσθητα καὶ αὐτά. […] ἀλλὰ τὴν διαστολὴν ταῖς αἰσθηταῖς κινήσεσι γνωρίζειν, ἃς ἐκ τοῦ πλήττεσθαι τοὺς δακτύλους ἡμῶν διαγινώσκομεν, τὴν συστολὴν δὲ πᾶν τὸ λοιπὸν τίθεσθαι καθ᾽ ὃ κινήσεως οὐκ ᾐσθάνετο.

⁴⁶ Cf. Heberden (n. 7), 558, quoted in n. 11 above.

⁴⁷ Cf. D. Shanzer, A Philosophical and Literary Commentary on Martianus Capella᾽s De Nuptiis Philologiae et Mercurii Book 1 (Berkeley, 1986), 4; see also H. Deiters, Über das Verhältnis des Martianus Capella zu Aristides Quintilianus (Posen, 1881), L. Cristante, Martiani Capellae De nuptiis Philologiae et Mercurii liber IX (Padua, 1987).

⁴⁸ Aristid. Quint. De Musica 31.6–7 W.-I.: καὶ ἰδίως ἐπὶ φωνῆς· περὶ οὗ νῦν πρόκειται λέγειν.

⁴⁹ Cf. also De Nupt. §§939–40, 947–8 and 968–9, where Martianus associates the rhythmical subdivision into arsis and thesis with melodic modulatio and not with dance, precisely as Aristides did. On the musical nature of vocal accents, cf. §268: et est accentus […] anima uocis et seminarium musices, quod omnis modulatio ex fastigiis uocum grauitateque componitur, ideoque accentus quasi adcantus dictus est.

⁵⁰ This is further confirmed by Remigius d'Auxerre, who glossed the word efferenda as subleuanda in acumen and the word premenda as scilicet in grauitatem: cf. W. Seidel, ‘Rhythmus/numerus’, in H. Eggebrecht (ed.), Handwörterbuch der musikalischen Terminologie (Wiesbaden, 1980), 1–36, at 16. Cf. also Mar. Vict./Aphth. 6.40.14–17 Keil (quoted and discussed below) and the consistent use of elatio as pitch raise in later musical treatises, e.g. Fulgentius, Myth. 3.9, Iac. Leod. Spec. 6.69.4, Commentum Einsidlense 8.228.23 Keil, Bonifat. Ars Metrica 209.23 Gebauer-Löfstedt, Anon. Ecce modorum in H. Schmid, Musica et scolica enchiriadis (Munich, 1981), 3.183–4 and the late testimony of Ger. Plethon, Keph. Log. Mous. in C. Alexandre, Pléthon—Traité des Lois (Paris, 1858), 461.15–16: ἄρσις μὲν οὖν εἶναι ὀξυτέρου φθόγγου ἐκ βαρυτέρου μετάληψιν, θέσιν δὲ τοὐναντίον βαρυτέρου ἐξ ὀξυτέρου.

⁵¹ E.g. R. Westphal, ‘Griechische Rhythmik’, in A. Rossbach and R. Westphal (edd.) Theorie der musischen Künste der Hellenen (Leipzig, 1885), 1.106–8; Bennett (n. 16), 367–8; M.G. Nicolau, L'origine du ‘cursus’ rythmique et les débuts de l'accent d'intensité en Latin (Paris, 1930), 43–56; C. Del Grande, La metrica greca (Turin, 1960), 252; F. Crusius and H. Rubenbauer, Römische Metrik (Munich, 1967), 29–31; see also the bibliography quoted by Luque Moreno (n. 26), 130–40, and J. Luque Moreno, De Pedibus, De Metris: las unidades de medida en la rítmica y en la métrica antiguas (Granada, 1995), 143–5.

⁵² Cf. Heberden (n. 7), 558: ‘in Marius Victorinus, p. 2482, the two meanings are given, apparently without any sense of their incongruity’; cf. Westphal (n. 51), 107–8, 210–11; R. Westphal and H. Gleditsch, ‘Allgemeine Theorie der griechischen Metrik’, in A. Rossbach and R. Westphal (edd.), Theorie der musischen Künste der Hellenen (Leipzig, 1887), 3.1.155–6; Bennett (n. 16), 368–9; Nicolau (n. 51), 53–4.

⁵³ Victorinus enjoyed so large a reputation that a statue was erected in his honour in the forum: cf. Aug. Conf. 8.2.3 (who also mentions in commendatory terms Victorinus’ Latin translations of some Platonic dialogues); Jerome, Chron. ad ann. Abr. 2370.

⁵⁴ These accompaniments are described with terms such as ictus and crepitus, not arsis/thesis: e.g. Hor. Carm. 4.6.35–6, Ars P. 251–3, 274, Sat. 1.10.42–3; Plin. HN 2.209; Quint. Inst. 1.12.3–4.1 (ne pes quidem otiosus certam legem temporum seruat), 9.4.51–5 (tempora etiam animo metiuntur et pedum et digitorum ictu), 9.4.55 (oratio non descendet ad crepitum digitorum).

⁵⁵ Cf. West (n. 8), 123.

⁵⁶ See e.g. chapter 31 of the Natyasastra: this text describes complex patterns of hand gestures, including claps and hand waves, which were employed to mark different rhythms. On the independence of claps and dynamic accents in Indian music, see L.E. Rowell, Music and Musical Thought in Early India (Chicago, 1992), 193–6, and Clayton (n. 10), 61–2. On the comparable case of Roman music, cf. T. Moore, Music in Roman Comedy (Cambridge, 2012), 158–62, with Cic. De or. 3.182–6.

⁵⁷ This approach closely resembles Aristoxenus’ discussion of the relationship between different rhythmizomena and their subdivision into chronoi: cf. Aristox. El. Rhythm. §§3–10, 2–4 Pearson.

⁵⁸ This is not the place to enter into the complex debate on the nature of Latin accent, but it is important to note that interpreting the expression arsis/elatio uocis purely as a matter of stress or dynamic accentuation is inconsistent with the latest research in the field, which suggests that accented syllables were marked by a raise in pitch and in volume at the same time. Cf. A.L. Sihler, New Comparative Grammar of Greek and Latin (Oxford, 1995), 241; J. Herman, Vulgar Latin (University Park, 2000), 36; E.H. Sturtevant, The Pronunciation of Latin and Greek (Chicago, 1940²), §214. According to G. Sampson, Writing Systems: A Linguistic Introduction (Stanford, 1985), 107, a similar combination of stress and pitch may have characterized also the shift from pitch- to stress-accentuation in late Greek.

⁵⁹ Cf. Diom. 1.430.29–30 Keil: accentus est acutus uel grauis uel inflexa elatio orationis uocisue intentio uel inclinatio acuto aut inflexo sono regens uerba (‘An accent is the acute, grave or circumflex elevation of the speech, or the tension and depression of the voice that governs words with a high-pitched or circumflex sound’). Many other passages describe accented syllables as higher in pitch: e.g. Varro apud Sergius 4.525.21–526.1 Keil (natura uero prosodiae in eo est, quod aut sursum est aut deorsum: nam in uocis altitudine omnino spectatur […]. ab altitudine discernit accentus, cum pars uerbi aut in graue deprimitur aut sublimatur in acutum) and 4.523.16–19 Keil (quae notae demonstrant omnem acutam uocem sursum esse et grauem deorsum. ipsum etiam musicorum docetur diagrammate). See also Cic. Orat. 57–8 (mira est enim quaedam natura uocis cuius quidem e tribus omnino sonis, inflexo acuto graui, tanta sit et tam suauis uarietas perfecta in cantibus. est autem etiam in dicendo quidam cantus obscurior […] cum alter alteri obicit uocis inflexiones) and ps.–Priscian 3.519.25 Keil (accentus namque est certa lex et regula ad eleuandam et deprimendam syllabam unius cuiusque particulae orationis). The dynamic component of word accentuation was described with expressions like plus sonat: e.g. Pompeius 5.126.30-33 Keil, with W.S. Allen, Vox Latina—A Guide to the Pronunciation of Classical Latin (Cambridge, 1965), 84. However, this expression occasionally indicates a combination of stress and pitch: cf. Servius 4.426.16–20 Keil.

⁶⁰ See also Sergius 4.482.5–18 Keil.

⁶¹ Cf. D. Creese, The Monochord in Ancient Greek Harmonic Science (Cambridge, 2010), 224–5.

⁶² Later sources regularly associate higher pitches with physical height (altitudo), e.g. Priscian 2.6.17–22 Keil: nam si aer corpus est, et uox, quae ex aere icto constat, corpus esse ostenditur, quippe cum et tangit aurem et tripertito diuiditur, quod est suum corporis, hoc est in altitudinem, latitudinem, longitudinem, unde ex omni quoque parte potest audiri. praeterea tamen singulae syllabae altitudinem quidem habent in tenore, crassitudinem uero uel latitudinem in spiritu, longitudinem in tempore. On this passage, cf. E. Rocconi, ‘The development of vertical direction in the spatial representation of sound’, in E. Hickmann, R. Eichmann and A.D. Kilmer, The Archaeology of Sound: Origin and Organisation (Rahden, 2002), 389–92. For similar statements, see also ps.-Priscian 3.519.8–11 Keil, Sergius 4.525.22–8 Keil, Audax 7.357.14–26 Keil and Julian of Toledo in M. Maestre Yenes, Ars Iuliani Toletani episcopi (Toledo, 1973), 135.182–91.

⁶³ Diom. 1.430.28 Keil: accentus est acutus uel grauis uel inflexa elatio orationis.

⁶⁴ E.g. Att. Fortunatianus 6.281.3.4–7 Keil (hic est motus et ingressio, quam Graeci basin appellant: de sublatione constat et positione, quae et thesis dicitur. ‘Arma ui’; ‘ar’ sublatio est temporum duum, ‘ma ui’ positio temporum duum), Mar. Vict. 6.40.17–18 Keil (in dactylo uero tollitur una longa, ponuntur duae breues; in anapaesto contra [scil. tolluntur duae breues, ponitur una longa]), 6.45.18–22 Keil; cf. also Diom. 1.480.6–12 Keil, Mar. Vict. 6.45.3–5 Keil.

⁶⁵ In general on this distinction, cf. Goodell (n. 4) and West (n. 8), 137 n. 22, 141 and 245.

⁶⁶ For the consistent use of δακτυλικός in relation to the rhythmical genus 1:1, cf. Aristox. El. Rhythm. §30, 16.17–19 Pearson, §32, 18.1–4 Pearson; Psell. Rhythm. §12, 24.8–16 Pearson; §17, 26.15–16 Pearson; Frag. Neap. §13–14, 28.17–29.1 Pearson, Aristid. Quint. De Musica 35.3–26 W.-I.; for the notion of δάκτυλος κατ᾽ ἴαμβον (⏑⏔ ⁝ ⏑⏔), cf. P.Oxy. 2678 ii.3–6, 37 Pearson, and Aristid. Quint. De Musica 38.5–12 W.-I., who describes also other types of rhythmical ‘dactyls’ (δάκτυλος κατὰ βακχεῖον τὸν ἀπὸ τροχαίου, δάκτυλος κατὰ βακχεῖον τὸν ἀπὸ ἰάμβου, δάκτυλος κατὰ χορεῖον τὸν ἰαμβοειδῆ and δάκτυλος κατὰ χορεῖον τὸν τροχαιοειδῆ). All these rhythmical sequences are called ‘dactyls’ because the duration of their arseis and theseis is equal, independently of the specific sequence of syllables they comprise: cf. West (n. 8), 136.

⁶⁷ Cf. Dion. Hal. Comp. 17.55–6, Hephaestion 11.6 Consbruch, Scholia B 307.17 Consbruch, Trichas 367.1 Consbruch.

⁶⁸ E.g. Marius Victorinus/Aphthonius 6.40.6–13 Keil (de arsi ac thesi, id est de alterna syllabarum sublatione ac positione, quibus pedes in metris nituntur atque formantur), Diom. 1.474–5 Keil (metrum est pedum iunctura numero modoque finita. uel sic, metrum est conpositio pedum ordine statuto decurrens modum positionis sublationisque conseruans), Diom. 1.473.21–3 Keil (rythmus [sic] est pedum temporumque iunctura cum leuitate sine modo. alii sic, rythmus est uersus imago modulata seruans numerum syllabarum positionem saepe sublationemque continens). Cf. Quint. Inst. 9.4.46–50 and 9.4.55, where the author outlines a similar contrast between the precise syllabic sequences defined by metrical structures and the greater freedom of rhythms. Quintilian provides these sketchy observations on rhythmical feet only to reveal the superiority of prose and metrics over poetry and music (cf. 9.4.60), and his lack of interest in the specifics of rhythmical matters is reflected by his imprecise extension of the term sublatio to rhythmical sequences.

⁶⁹ Anon. Ambros. 215.21 Studemund: ἄρσις μὲν γὰρ καλεῖται ἡ ἀρχὴ τοῦ στίχου, θέσις δὲ τὸ τέλος; Choerob. 211.14 Consbruch = Scholia B 294.13 Consbruch: ποὺς τοίνυν ἐστὶν σύνταξις συλλαβῶν ἄρσιν καὶ θέσιν περιέχουσα. See also Anon. Ambros. 225.15–20 and 227.9–18 Studemund.

⁷⁰ Not appreciating this distinction, Deiters wrongly accused Capella of contradicting himself by employing two different meanings of the word arsis: ‘Martianus folgt hier dem Sprachgebrauche der lateinischen Metriker, in den späteren Abschnitten aber im Widerspruch mit sich selbst wieder dem des Aristides’ (Deiters [n. 47], 14).

⁷¹ This is consistent with the definition of metrical foot he provides at 6.43.9–10 Keil: pes est certus modus syllabarum, quo cognoscimus totius metri speciem, compositus ex sublatione et positione. Cf. 6.41–2 Keil: differt autem rhythmus a metro, quod metrum in uerbis, rhythmus in modulatione ac motu corporis sit.

⁷² This terminological shift most likely depended on Victorinus’ source, and his subsequent distinction between different rhythmizomena suggests that it was an excerpt from an Aristoxenian treatise on rhythmics: cf. Aristox. El. Harm. §§2–9, 2–6 Pearson.

⁷³ E.g. Sergius 4.522.26 Keil (scandere uersus), with OLD s.v. scando 6, and Lewis-Short s.v.

⁷⁴ Cf. W. Stroh, ‘Arsis und Thesis, oder: Wie hat man lateinische Verse gesprochen?’, in W. Stroh, Apocrypha: Entlegene Schriften (Stuttgart, 2000), 193–216, esp. 206–8; S. Bonner, Education in Ancient Rome: from the Elder Cato to the Younger Pliny (London and New York, 2012), 190.

⁷⁵ See also Quint. Inst. 1.8.13, Sext. Emp. Adv. math. 1.159–60 and Pompeius 5.118.6 Keil.

⁷⁶ See esp. Ter. Maur. 6.368 Keil, vv. 1420–3 on the division of amphibrachs (⏑ – ⏑) arsis hinc sumat necesse est tria priora tempora | et thesi relinquat unum: uel licet uertas retro, | arsis uno subleuetur, deprimant thesin tria. On the role of sublatio as marking the division of metrical feet, cf. also Ter. Maur. 6.350 Keil, v. 835: et pedum uel hinc uel inde subleuant discrimina.

⁷⁷ Cf. Ter. Maur. 6.393 Keil, esp. vv. 2253–5: scandendo [scil. necesse est] et illic ponere adsuetam moram | quam pollicis sonore uel plausu pedis | discriminare, qui docent artem, solent. The context of this quotation clarifies that Terentianus refers to metrical theseis. In order to explain why dactylic hexameters are measured by counting single feet while iambic trimeters are measured by groups of two feet, Terentianus specifies that this distinction depends on the position of the mora (i.e. ‘space of time’ or ‘foot section’, not ‘pause’—cf. Aug. De Musica 2.13.24) that is marked by clapping: if the marked mora appears already in the first foot, then the verse will be dactylic and each foot will comprise a long syllable in sublatio and two short syllables in positio; by contrast, if the marked mora is not present in the first foot but appears only in the second, fourth and sixth—i.e. where it is compulsory to have purely iambic feet in iambic trimeters—then the resulting verse will be iambic. Cf. Caes. Bass. apud Rufin. 6.555.24 Keil, Mar. Vict. 6.80–1 Keil, Aug. De Musica 5.11.24.

⁷⁸ Ter. Maur. 6.365–6 Keil, vv. 1345–6: parte nam attollit sonorem, parte reliqua deprimit | ἄρσιν hanc Graeci uocarunt, alteram contra θέσιν.

⁷⁹ Cf. W. Beare, Latin Verse and European Song. A Study in Accent and Rhythm (London, 1957), 61.

⁸⁰ See e.g. 6.493.24 Keil, where Sacerdos treats the words perspĭ́cere possit as a possible hexameter ending, regarding the short accented syllable as long, and 6.494.7–12 Keil on the complementary tendency for long unstressed vowel to be shortened. Cf. J.N. Adams, Social Variation and the Latin Language (Cambridge, 2013), 43–51 and 6.423 Keil.

⁸¹ Cf. Quint. Inst. 1.5.31: est autem in omni uoce utique acuta, sed numquam plus una nec umquam ultima. This passage refutes the explanation proposed by Stroh (n. 74), 208, according to whom Sacerdos referred to the existence of a secondary accent (cf. the German word Múttertàg). Cf. K. Zeleny, Itali Modi: Akzentrhythmen in der lateinischen Dichtung der augusteischen Zeit (Vienna, 2008), 64–5.

⁸² Zeleny (n. 81), 64–5; on the similarity between the letter a ( ) and t ( ), cf. E.A. Lowe, The Beneventan Script: A History of the South Italian Minuscule (Oxford, 1914), 93–4, 133–48.

⁸³ Cf. Adams (n. 80).

⁸⁴ Serg. 4.522.26 Keil: syllabas natura longas difficile est scire. sed hanc ambiguitatem sola probant auctoritatis exempla, cum uersus poetae scandere coeperis. See also 4.480.12–14, 4.482.4–18 Keil, where Sergius highlights again the artificial nature of metrical arsis/thesis and their original dependence on melodic accents. On the difficulty to identify syllable quantities in Late Antiquity, cf. Aug. De Musica 2.1.1 and 3.3.5.

⁸⁵ Cf. Ter. Maur. 6.393 Keil, quoted above; Aug. De Musica 2.10.18 and passim.

⁸⁶ 6.365–6 Keil, vv. 1342–7: una longa non ualebit edere ex sese pedem, | ictibus quia fit duobus, non gemello tempore. | breuis utrimque sit licebit, bis feriri conuenit, | parte nam attollit sonorem, parte reliqua deprimit | ἄρσιν hanc Graeci uocarunt, alteram contra θέσιν | una porro bis feriri quando poterit syllaba? Cf. Diom. 1.475.3 Keil: ergo una longa pedem non ualebit efficere, quia ictibus duobus arsis et thesis, non gemello tempore perquirenda est.

⁸⁷ E.g. Mar. Vict. 40.17–18, Aug. De Musica 2.14.26, where spondees, dactyls, proceleusmatics as well as anapaests are said to correspond to each other not only with regard to their duration but also with regard to their scansion.

⁸⁸ The use of melodic means to provide rhythmical cues is attested in many folk traditions, such as Sardinian music played on launeddas and Scottish pibroch piping; as I hope to show elsewhere in due course, several ancient sources suggest that this was the case also in aulos music. For a stimulating discussion of the role of instrumental accompaniment in the perception of ancient Greek rhythm, cf. A. Barker, ‘Heterophonia and Poikilia: accompaniments to Greek melody’, in B. Gentili and F. Perusino, Mousiké: Metrica, Ritmica e Musica Greca in Memoria di Giovanni Comotti (Pisa and Rome, 1995), 41–60.

⁸⁹ ps.-Arist. Pr. 19.38, 920b33–6: ῥυθμῷ δὲ χαίρομεν διὰ τὸ γνώριμον καὶ τεταγμένον ἀριθμὸν ἔχειν, καὶ κινεῖν ἡμᾶς τεταγμένως. Cf. Aristid. Quint. De Musica 31.10 W.-I. and 40.20–5 W.-I.