¹ F. is the Loeb translator, H. N. Fowler (1925): H. is Hackforth, R., Plato's Examination of Pleasure (C.U.P., 1945): G. is Gosling, J. C. B., Plato, Philebus (Clarendon, 1975).Google Scholar

² In musicological sources the word avp. has a precise technical sense, ‘concordant’ by contrast with ‘discordant’ (). For definitions see e.g. Euclid, Seclio Canonis 149.17–20, Cleonides, Eisagoge 187.19, Nicomachus, Enchiridion 262.1, Bacchius, Eisagoge 293.8, Gaudentius, Eisagoge 337.8 (all cited by pages and lines of Jan, Musici Scriptores Graeci). Within the octave, only the fourth, the fifth and the octave itself were reckoned ‘concordant’, not thirds or sixths (see e.g. Aristoxenus, El. Harm. 20. Iff.). Though the tuning of an instrument to a scale (), or the adjustment to one another of notes in a melody, involved the proper attunement of intervals other than these, particularly the smaller ones that subdivide the fourth, the tuning of the concords was fundamental, in three ways, (i) The so-called ‘fixed notes’, which provided the basic framework for any scale, stood in concordant relations to one another, (ii) According to Aristoxenus, the most important constraint on any melodic series was that as it descended or ascended by consecutive scalar steps from any given note, either the fourth note in order stood at a fourth from the given note, or the fifth note stood at a fifth, or both (El. Harm. 29.6–14, 53.33–54.21). (iii) In some important forms of tuning, e.g. that implied at Timaeus 35b–36b, every note can be found by movements of fourths and fifths up and down from a given initial note (the method of tuning ‘by concordance’, which is still in common use). See Aristox. El. Harm. 55.3ff., cf. Eucl. Sect. Can. prop. 17. Then in a passage like the present one it is possible to take in its strict sense, not just vaguely as ‘the melodically proper’, even though the process referred to (whether tuning an instrument or adjusting melodic intonation) required the formation of non-concordant intervals too. For a similar hint of the centrality of concords to a musical system see Rep. 531c3–4.

³ See the passages cited in LSJ s.v. 1.5.

⁴ E.g. Sachs, C., The History of Musical Instruments (London, 1942), p. 132Google Scholar

⁵ For a possible example see my remarks in ‘The Innovations of Lysander the Kitharist’, CQ 32(1982), 268.

⁶ The phrase is strongly reminiscent of Rep. 531 a4–b 1, which speaks of people , and seeking to identify by ear the . Both passages focus on the unscientific ‘empiricism’ of certain people's practices, those in the Republic being described as . In both cases what is ‘hunted’ is something purely auditory, not the correct length on a string (the Republic's investigators adjust the strings′ tension, , 531b3–4: they are not marking lengths on a monochord). But beyond this the passages are not parallel. The of the Republic are theorists, not practical musicians. They search for a minimal interval with a view to the formal representation of scale-systems (cf. Aristox. El. Harm. 39.4ff., and his references to his predecessors′ diagrams at 2.11–18, 7.31, 27.34–28.6: on this see Aristotle, Metaph. 1016b18–24, 1053a12–17). The in the Philebus is not an interval that acts as a common measure, but the correct pitch of each note, determined by the interval at which it stands from others in its musical vicinity (this need not, of course, be a minimal interval, nor is it the same size in every case); and those who search for it are real musicians, not theorists.

⁷ May I renew the familiar plea that the mistranslation ‘flute’ be abandoned? If needs to be translated rather than transliterated, perhaps we can settle for the harmless ‘pipe’.

⁸ I cannot produce an unambiguous case in classical literature where the noun must mean ‘note’ and cannot mean or imply ‘string’, though the later theorists quite often use and interchangeably, e.g. Arist. Quint. De Mus. 12.6. But a good parallel to the usage in the Republic is given by the phrase in Simonides, fr. 46 Bergk = Frag. Adesp. 947(b) Page.

⁹ At Aristox. El. Harm. 43.1–6, the tuning of the strings of kithara or lyra (which would be done, or at least checked and adjusted afresh for each performance) is treated as parallel to the drilling of the of the aulos (which was done once for all by the maker). Cf. e.g. Xen. Symp. 3.1: when an aulos and a lyra are to be played together, the tuning of the aulos is taken as fixed, and the lyra is tuned to it.

¹⁰ See particularly Aristox. El. Harm. 43.10–24: indications of the specialised techniques used by fourth-century auletes are given in Theophrastus′ account of reed-making, Hist. Plant. 4.11.1–7, especially 4–5.

¹¹ Aristox. loc. cit., arguing against those who base their account of correct scalar form on the physical structure of auloi (the spacing and arrangement of their finger-holes, etc.) says that it is astounding () that they do not abandon their view, since it is absolutely plain that every note produced by such an instrument . If the structure of any instrument is to be taken as authoritative in identifying the relations that correctly constitute (which it is not), at any rate the instrument cannot be the aulos, which is the least secure of all instruments in its pitching (). On the other hand it must be granted that the difficulties involved in tuning a stringed instrument would certainly have been familiar to any Athenian who had been taught to play the lyra as part of his elementary education.

¹² For a careful account of this see Aristox. El. Harm. 11.4–21.