¹ See, for example, the many distinguished names in the bibliography on pp. 120–4 of Kuipers, A. H., Phoneme and morpheme in Kabardian, 's-Gravenhage, 1960Google Scholar.

² See BSOAS, XIII, 1, 1949, 36–79Google Scholar, and BSOAS, XIII, 2, 1950, 381–8Google Scholar.

³ My short and very superficial acquaintance with Abaza was made in the company of W. Sidney Allen, who subsequently went on by himself to probe this language deeply and to rich effect. See especially his paper ‘Structure and system in the Abaza verbal complex’, TPS, 1956, which splendidly combines careful and detailed exposition of the phonetic data with important excursions into general linguistic theory.

⁴ These were the following: Kabardinskaya azbuka, Tiflis, 1906Google Scholar; Turchaninov, G. and Tsagov, M., Grammatika kadbardinskogo yazyka, I, Moscow, 1940Google Scholar; Yakovlev, N., Materials for the Kabardey dictionary, i, Moscow, 1927Google Scholar; Apazhev, M. L. and others, Kabardinsko-russkiy slovar', Moscow, 1957Google Scholar; Kuipers, A. H., Phoneme and morpheme in Kabardian, 's-Gravenhage, 1960Google Scholar.

⁵ Kuipers presents a very useful table of eight Kabardian alphabets alongside his own phonetic transcription on pp. 116–17 of the book cited above.

⁶ Catford, J. C., ‘The Kabardian language’, Le Maître Phonétique, 78, 1942, 15–18Google Scholar.

⁷ See pp. 103, 105–6.

⁸ His reasons for doing so are given on p. 21 of the work already cited.

⁹ See op. cit., 17, 27, 32–8.

¹⁰ op. cit., 21, 26–7.

¹¹ loc. cit. Kuipers's transcription renders them .

¹² op. cit., 22.

¹³ See p. 100.

¹⁴ art. cit., 16.

¹⁵ op. cit., 20.

¹⁶ This is, of course, not to say that such a feature may not be present and detectable in the live utterances of my informants, merely that, if present, I did not notice it at the time.

¹⁷ The first of these contrasts is treated by Kuipers in his very interesting phonological analysis as one of ‘laryngeal features’, the second as one of ‘mouth-resonator features’. For the variety of Kabardian he describes, palatalization is an important mouth-resonator feature. It appeared to be less so in the speech of my informants. See above.

¹⁸ Not all Catford's words could be identified by my informants, so that I was unable, for example, to obtain a recording of the full series of contrasts given in the footnote to p. 16 of his paper.

¹⁹ op. cit., passim.

²⁰ Yakovlev and Kuipers list many other interesting consonant combinations that I had hoped to record if there had been time.

²¹ The sound spectrograph used was a Kay Sonograph. The pitchmeter was the Frøkjær-Jensen Trans Pitchmeter. The inkwriter was the Siemens Oscillomink (Mingograph). The recordings and all the sonograms and mingograms used in this study were made by Mr. A. W. Stone, Chief Technician at the School.

²² It appears improbable that Soviet phoneticians can have overlooked such a promising field for research, and there may therefore be Soviet work on the subject of which I am ignorant.

²³ Catford, , ‘On the classification of stop consonants’, Le Maître Phonétique, 65, 1939, 2–5Google Scholar.

²⁴ ‘Sonogram’ is the usual term for a sound spectrogram made on the Kay Sonograph and will be used henceforth.

²⁵ cf. Kuipers, , op. cit., 19Google Scholar.

²⁶ See, for example, Lehiste, Ilse, Some acoustic characteristics of dysarthric speech, Basle, Karger, 1965, esp. p. 96Google Scholar.

²⁷ Strevens, P., ‘Spectra of fricative noise in human speech’, Language and Speech, III, 1, 1960, 32CrossRefGoogle Scholar.

²⁸ Strevens, , op. cit., 33Google Scholar.

²⁹ See p. 93.

³⁰ The formant structure of the following vowel may also be relevant. Clearly, a full-length study of Kabardian fricatives would require the examination of a much larger sample, taking into account all variants found.

³¹ A good example of how such distinctive patterns can be derived from the speech of a number of speakers is Peter Strevens's paper already cited. By ‘centre of gravity’ is meant the concentration of energy towards one end of the frequency spectrum rather than another.

³² The metalanguage of acoustic description in phonetics is still far from fixed and varies from writer to writer. Some writers for example, would use the word ‘formant’ in these descriptions instead of such clumsy expressions as the ‘region of greatest concentration of energy’, and so on. ‘Formant’ has, however, on the whole been reserved for the description of voiced sounds, and so is perhaps better avoided here, where the reference is to voiceless fricative sounds.

M's pronunciations rather than F's were used as being more stable. F sometimes hesitated or misread items. For a comparison upon which to build significant generalizations a wider range of speakers would be necessary.

³³ See p. 95.

³⁴ It should be noted that since relative intensity is shown on the sonograms by the relative blackness of the shading, judgements as to whether the energy present is ‘strong’ or ‘weak’ are bound to be subjective, and might not always agree with judgements made by another investigator.

³⁵ The recordings of such sounds are perceived as plosive-fricative sequences but there is often no clear plosive ‘spike’ preceding the fricative element on the corresponding sonograms. This seems to confirm Kuipers's observation (p. 19) that voiceless stops in this position are particularly weak.

³⁶ Comparable results have been obtained elsewhere. See, for instance, Fant, Gunnar, ‘Modern instruments and methods for acoustic studies of speech’, Royal Institute of Technology, Division of Telegraphy-Telephony, Report (Stockholm), 8, 1957, 29Google Scholar.

³⁷ In sonograms of F' s speech this feature was less noticeable than in M's, but was still discernible.

³⁸ Fant, , op. cit., 24Google Scholar.