An important task of specialists in any field is the elaboration of a clear and adequate terminology. The coining and popularization of well thought out technical terms both reflects theoretical advances and facilitates clarity of thought and expression.

If semiotics is the general theory of communicative signs (Morris, 1938: 80), then phonetics constitutes part of that theory, in the sense that phonetics tries to model the means by which communication is achieved through the use of spoken signs. The technical vocabulary of phonetics is thus a sub-set of the meta-language of semiotics, and one might reasonably expect that other terms and concepts from semiotic theory, such as that of INDEXICAL and ICONIC signs (Feibleman, 1946: 91), might be illuminating when applied to the subject-matter of phonetics.

The subject of these notes is the dialect of Mudau, a small market-town in the south-eastern Odenwald, 18 km N.E. of Eberbach (Neckar) and 19 km S.S.W. of Miltenberg (Main). The population numbers around 1700 and the main means of livelihood are shop-keeping, agriculture and local light industry.

In this journal a number of suggestions have recently been made as to possible improvements to the Association's alphabet. Some of these proposals are relatively minor, involving only the addition of symbols or diacritics; the acceptance of these would have little or no effect on the basic character of the alphabet. But others, specifically those relating to the re-organization of the Cardinal Vowels and the consonant chart, are more thorough-going. Here, important principles are involved which would seem to demand more discussion. The following remarks are intended as a contribution to this discussion rather than as an evaluation of specific proposals.

Ever since the appearance of the monograph On Defining the Phoneme (Twaddell, 1935), it has become customary to classify the views of the phoneme into the four categories: mentalistic, physical, functional, and abstract. The reader is also referred to Jones (1957, and 1967: 212 ff.) and Fudge (1970) for discussions on that basis. However, we can achieve additional insight into the problem of defining the phoneme if we classify the views of the phoneme, this time, on a populational basis. By this, we mean whether the phoneme is defined as a unit of (1) an idiolect, (2) a dialect, (3) a multidialect, or (4) a supralect.

1. In this article and the one to follow I shall be describing my approach to the problem of extracting a system of intonation from a corpus of unscripted colloquial Russian conversation, and discussing in detail the results obtained by this approach.

‘“… there is more near-ambiguity in language than it is always convenient to admit”; and in studying stress and juncture it would be especially valuable, in order to avoid too great a reliance either on introspection or on artificially contrived confrontations, to make a collection of ambiguities in action in the form of fully-documented instances of uncertainty of interpretation or of actual misunderstanding‘ (Sharp, 1960: 112–13).

Contrastive phonology has been preoccupied with the phoneme concept and the isolated word. The Contrastive Structure Series demonstrates this quite clearly. In Moulton's contrastive comparison of English and German sounds (Moulton, 1962), to take a representative example of the series and of contrastive phonological studies in general, 107 pages out of 145 deal with the consonant and vowel phonemes in English and German words and with their contrastive analysis. Only 31 pages treat stress (largely word stress), intonation and the inevitable juncture, which is the epitome of word phonemics in that it looks at the incidence of boundary signals at word and morpheme boundaries, e.g. antrat and das Kitzeln with strongly aspirated stops as against Landrat and die Skizze with weakly aspirated or unaspirated stops indicating different divisions between word final and initial consonants (p. 142). I would like to mention in passing that this opposition is artificially constructed in word material by a linguist on the basis of the equally artificial codification in Siebs (de Boor et al., 1969), where again hardly any tribute is paid to segmental sentence phonology (Kohler, 1970).

Windsor Lewis's recent article (1972b) admirably summarizes the case for a Gimson-type transcription as a reference tool for EFL students. As he states, his own ‘CPD’ (Windsor Lewis, 1972a) transcription clearly—and sensibly—derives from the system established by Gimson (1962, 1970) in ‘the standard up-to-date description of Received Pronunciation’. On general principles, there is little that could usefully be added to Windsor Lewis's (1972b) comments on three other transcription types that are increasingly coming to be regarded as inappropriate for EFL purposes:

(i) that used by Abercrombie (1964), which, as Windsor Lewis (1972b: 60–63) shows, has disadvantages in its choice of detail symbols (though in underlying philosophy this system is perhaps closer to ‘CPD’ than is the latter's acknowledged parent);

(ii) the long-established ‘EPD’ (Jones, 1917) system, with its inconsistent treatment of the monophthong phonemes;

(iii) the ‘simplified’ (Jones, 1956: 346 ff.) types of transcription first used in books by Scott (1942) and MacCarthy (1944, 1945) and taken up by Kingdon (West, 1965) and others, which although useful for the Ll speaker are grossly misleading for the EFL student in implying a solely quantitative distinction between such pairs as the vowels of seat and sit.

This article will largely take for granted Windsor Lewis's comments on those types, and will concentrate on some aspects of the system discussed by Windsor Lewis (1972b) and applied in his CPD and the new third edition (1974) of the Advanced Learner's Dictionary.

All common transcriptions of the Russian Cyrillic or Kirillitza [ki'rilitsa] into the Latin alphabet are unsatisfactory. They either have unsightly and untypable diacriticals (the Czech system) or else a superabundance of h's and y's (English systems). Both spoil the dignity of the original while scarcely correcting its notorious shortcomings, and thus they continue to hide the very information which is essential in order to pronounce the words!