Hostname: page-component-586b7cd67f-t8hqh Total loading time: 0 Render date: 2024-11-25T04:04:12.306Z Has data issue: false hasContentIssue false
  • English
  • Français

Melody and word accent relationships in ancient Greek musical documents: the Pitch Height Rule

Published online by Cambridge University Press:  23 February 2012

Charles H. Cosgrove  and
Mary C. Meyer
Charles H. Cosgrove
Affiliation:
Northern Seminary, Lombard, Illinois
Mary C. Meyer
Affiliation:
Department of Statistics, University of Georgia
Get access Rights & Permissions [Opens in a new window]

Abstract

It has long been known from the extant ancient Greek musical documents that some composers correlated melodic contour with word accents. Up to now, the evidence of this compositional technique has been judged impressionistically. In this article a statistical method of interpretation through computer simulation is set forth and applied to the musical texts, focusing on the convention of correlating a word's accent with the highest pitch level in the melody for that word: the Pitch Height Rule. The results provide a sounder basis for judging evidence for the operation of this convention in specific pieces and a sharper delineation of its use in the history of ancient Greek music. The ‘rule’ was used by at least some composers from the late second century BC through the second century AD, but there is no certainty that it was used before or after this period. In some cases where previous scholars have discovered the rule's operation, statistical analysis casts doubt. Of special interest is the showing that one piece long judged as offering no evidence of the use of the rule probably displays an inversion or parody of the rule for rhetorical-musical effect.

Type
Research Article
Information
The Journal of Hellenic Studies , Volume 126 , November 2006 , pp. 66 - 81
Copyright
Copyright © The Society for the Promotion of Hellenic Studies 2006

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

1 A correlation of melody and word-accent was first observed by Otto Crusius in the Seikilos inscription. See Crusius, O., ‘Zu neuentdeckten antiken Musikresten: I. Nachträgliches über die Seikilosinschrift; II. Fragment einer Partitur des euripideischen Orestes’, Philologus 52 (1893) 173.Google Scholar

2 The following formulation of the rules is a summary based on the observations of Winnington-Ingram, R.P., ‘Appendix II: Melody and word-accent’, in Eitrem, S., Admundsen, L. and Winnington-Ingram, R.P., ‘Fragments of unknown Greek tragic texts’, SO 31 (1955) 1–87Google Scholar (Appendix II: 64–73). See also West, M.L., Ancient Greek Music (Oxford 1992) 199.Google Scholar

3 First edition: Pöhlmann, E. (ed.), Denkmäler altgriechischer Musik. Sammlung, Übertragung und Erläuterung aller Fragmente und Fälschungen (Nürnberg 1970)Google Scholar; revised edition: Pöhlmann, E. and West, M.L. (eds), Documents of Ancient Greek Music: The Extant Melodies and Fragments Edited and Transcribed with Commentary (Oxford 2001)Google Scholar (henceforth: DAGM).

4 Anderson, W.D., ‘Word-accent and melody relationships in ancient Greek musical texts’, Journal of Music Theory 17 (1983) 198.Google Scholar

5 Winnington-Ingram (n.2) 71.

6 Wagner, R., ‘Der Oxyrhynchos-Notenpapyrus’, Philologus 79 (1924) 205.Google Scholar So also Powell, J.U. and Barber, E.A. (eds), New Chapters in the History of Greek Literature, second series (Oxford 1929) 165Google Scholar; Abeti, H., ‘Ein neu entdeckter früchristlicher Hymnus mit antiken Musiknoten’, ZfMW 4 (1921/1922) 528.Google Scholar

7 Pöhlmann (n.3) 109; Reinach, Th., ‘Un ancêtre de la musique d'Église’, RMus 3 (1922) 19 n.2.Google Scholar

8 West found that of 2200 two-note sequences, 22% show a repetition, 47% show a move to the next nearest note (up or down), 19% show a move to the next nearest degree, and 12% show a move to some more remote degree. See West (n.2) 191–2.

9 The figure ‘1.37’ is obviously not an actual interval but a fraction resultant from averaging. We obtained this average by using only readings about which there is confidence. Keeping in view what matters for PH accord, we considered melodic movement as ‘syllable to syllable’ and counted only highest notes in cases of melisma.

10 Clashes in φερόπλοιο (1. 10) and Θνατοῖς (1. 20).

11 Clash in δικόρυφον (1. 2).

12 We think that composers did this shaping consciously because in certain respects the melody/accent rules differ from the significance of accents in speech (e.g. the voice falls on the grave in speech but is treated as high by the PH rule). The melody/accent relations may be a stylizing compositional technique — perhaps also archaizing if the tonal accent was being replaced in the Roman period by a dynamic accent. But it is also possible that the tonal accent persisted for a time as the dynamic accent developed. See West (n.2) 200.

13 In the case of one piece, we use a different alternative hypothesis in order to test whether there is evidence that the composer went deliberately against the PH rule. See the analysis of P. Berlin 6870 (11. 16–19) below.

14 Where stakes are high (e.g. drug trials), p < .01 is used.

15 Sometimes accent placement has an effect on chance accord, but usually it makes no difference.

16 Confident readings include words with no uncertain letters and notation, words where uncertainty about letters or notation does not affect judgements about PH accord, words where other possible readings (in the DAGM apparatus) show the same result for accord, and a very few cases, mentioned explicitly below where they arise, in which other factors warrant confidence about a PH accord judgement even with missing or uncertain notes.

17 See West (n.2) 192–3.

18 Fr. 13.ii shows perhaps 3 of 3 PH accord, but this sample size is too small for analysis. Note that the section of fr. 15.ii that contains legible notation is not part of fr. 15.i but belongs to another piece.

19 See the further comments on this piece in the discussion of the history of the PH rule below.

20 Crusius (n.2) 173 (speaking of his impression that design was at work in relations of pitch height to accent in this piece and in the compositions of Mesomedes).

21 We have included φαντάσματ' (3b) in the confident reading because it is hard to imagine that the missing note for the opening syllable was higher than a Z (g'), the highest note used anywhere in the fragment.

22 Winnington-Ingram (n.2) 56–7.

23 We follow Pöhlmann and West (DAGM, 124–5), who included these two notes that Winington-Ingram suggests have been deleted.

24 The word is uncertain but the accent placement is not. We have either φαρία (vessel) or ἲα (violets). In both possibilities there is a clash according to the note readings in DAGM.

25 See Johnson, W.A., ‘Musical evenings in the early Empire: new evidence from a Greek papyrus with musical notation’, JHS 120 (2000) 65.CrossRefGoogle Scholar

26 On 11. 1–18 and 11. 19–26 as separate items in P.Mich. 2958, see Pearl, O.M. and Winnington-Ingram, R.P., ‘A Michigan papyrus with musical notation’, JEA 51 (1965) 179–95Google Scholar; DAGM, 142.

27 Haslam, M.W., ‘3161. Texts with musical notation’, P.Oxy. 44 (1976) 63.Google Scholar

28 DAGM, 178.

29 Strophe and antistrophe carry the same melody. It would be very difficult to devise an antistrophe that carried accentual patterns suiting the melody of the strophe. See Anderson, W.D., Music and Musicians in Ancient Greece (Ithaca 1994) 95–6.Google Scholar

30 West (n.2) 199.

31 Pöhlmann (n.3) 81–2; cf. DAGM, 17.

32 Feaver, D., ‘The musical setting of Euripides' Orestes’, AJPh 81 (1960) 1–15.Google Scholar

33 P. Leiden inv. P. 510 (lines from Euripides, Iphigenia in Aulis); most of the fragments from P. Vienna G. 29825; the fragments of P. Vienna G 13763/1494; Epidaurus, SEG 30.390 (Hymn to Asclepius); P.Berlin 6870, I. 23 (but date uncertain).

34 See West (n.2) 320–1.