1 Gleason, H. A., An Introduction to Descriptive Linguistics (New York, 1965) 1.12 (p.10)Google Scholar.

2 So Bloomfield, L., Language (London, 1935), p. 21 Google ScholarPubMed. It is interesting that Sapir, Language (New York, 1921), p. 7, does not include ‘vocal’ in his definition of language. Many European linguists would not do so either. Firth appears to reach the other extreme in defining ‘speech’ as ‘oral or written’, in “The Semantics of Linguistic Science,” Papers in Linguistics (London, 1957), p. 144.

3 From the bareness of this statement it can be seen that no particular theory of language is presupposed here. However, the distinction between grammatical and lexical form, or between a form and its meaning, and the classing of the units so distinguished under separate terms, seems an unnecessary complication, at least for the practical study of language.

4 E.g., Harris, in Methods in Structural Linguistics (Chicago, 1951), speaks of morphemic segments as “sequences of phonemes,” and phonemes “constituting” a morpheme (pp. 195-6 passim), yet on p. 184 he discusses order as a morphemic element. Such an element can only be “represented by” phonemes, it cannot “consist of” them. (See also note 5.)

5 Gleason, H. A., Introduction, p. 75 Google ScholarPubMed, following “Morphemes are generally short sequences of phonemes” and similar statements (pp. 51-3).

6 E.g., the material on the basis of which a phoneme is determined may range from a single idiolect, as in D. Jones, The Phoneme (Cambridge, 1962), Chapter 2, (note especially #28, p. 9) to a wide group of dialects, as in Trager and Smith, An Outline of English Structure (Studies in Linguistics, Occasional Papers 3, Washington, 1957), Section 1.

7 Phonemic transcription also, of course, gives an “imperfect and incomplete picture of actual speech” (see note 1), a fact not noticed by those who condemn traditional writing systems on this ground. The linguist will claim that all “significant” features of speech are recorded in phonemic transcription—but significant to whom and for what? (see note 6). Other features of speech, such as the “tones of voice” expressing anger, admiration, boredom, etc., are certainly significant to the hearer, although generally classed as “paralinguistic,” i.e., beyond the linguist’s concern.

8 This is true whether one holds the traditional view that it was genuinely alphabetic, or Gelb’s view in A Study of Writing (London, 1952), p. 147f., that it was originally a syllabic system.

9 E.g., Steele and other English scholars discussed by D. Abercrombie in Studies in Phonetics and Linguistics (London, 1965), Chapters 5 and 6.

10 Note also, besides the Indian tradition represented by Panini, the description of the vowels of Hebrew according to their position of articulation by Saadya Ga’on (d. 942) given in Skoss, S. L., Saadia Gaon, the Earliest Hebrew Grammarian, (Philadelphia, 1955) p. 27f Google Scholar., and the description of the consonants of Arabic on the same basis by Ibn Sina (Avicenna, d. 1037), which is preceded by a detailed description of the physical production of speech sounds, given in Semaan, K. I., Arabic Phonetics (Lahore, 1963)Google Scholar.

11 Even a child knows the intonation contours typical of his language before he begins reading. It is therefore unnecessary to mark them. It is interesting, in this connection, to compare Pitman’s “Initial Teaching Alphabet” (a semi-phonetic spelling designed for the teaching of beginning readers, and used with success in the U.K. and the U.S.A.) with a phonemic transcription, from which it differs in a variety of ways.

12 Contemporary spoken English differs dialectically in the number of phonemes and their use, as is shown for example, in the section on phonology in Trager and Smith’s Outline of English Structure, especially Sect. 1.33, p. 22f. Sound changes will induce similar differences in a given locality over a period of time. If phonemic transcription were used as a general writing system, several systems would be needed, no matter how broad a definition of “phoneme” was envisaged, even in North America, (cf. Gleason on the transcription used in the “English-for-Foreigners” programme of the American Council of Learned Societies [in Introduction, 19.17-18, p. 324 f.]), let alone the rest of the English speaking world. Furthermore, these systems would need periodic revision.

13 In CJL 9, (1964), p. 108 f.

14 Of course broad general rules for the determination of spoken forms from written forms can be given, and these often enable readers to recognize a morph, when they are familiar with its spoken, but not its written representation. The traditional “spelling out” method of teaching reading is based on this. These rules, however, cannot be systematized. See note 15.

15 The reader will, of course, normally state his recognition in terms of the spoken representative of the morph. However, because the connection between written and spoken form is only indirect, it is possible to “mis-recognize” a written form, even when the spoken counterpart is known, and even to the extent of including a new morph in the idiolect. E.g., misled is not uncommonly ascribed to a verb mizle or mizzle rather than mislead.

16 Cf. the history of the Chinese writing system. The “Arabic” numerals are also a case in point.

17 Thus the teaching of beginning readers by the “word recognition” method is unsatisfactory not because the theory on which it is based is wrong, but because there is no way in which the student can proceed on his own after acquiring basic principles.

18 This can be easily seen in the use of “punctuation marks.” E.g., the question mark can be said to be used in imitation of the rising inflection of speech. Its use would coincide with rising inflection in speech in a considerable proportion of cases, but by no means in all.

19 Noted first, perhaps, by de Saussure, Cours de linguistique générale (Paris, 1922), p. 57.

20 E.g., the many “abbreviations” of mediaeval MSS such as Dno for Domino, and writings such as U.S. $10.50 today. The convention that “letters represent sounds” is strong enough to produce spoken counterparts for a great many (e.g., N.A. T.O./neytow/), even where the combination of letters is impossible in English (e.g., N.F.C.U.S./nifkəs/).

21 This is not a new suggestion, see, for example, McIntosh in “The Analysis of Written Middle English,” TPS 1956, p. 40.

22 Language, p. 21.

23 Symbols such as c, q, and the capital letters have their own place in the conventional written representation of English (for capitals, see note 25). They cannot be considered superfluous any more than allophones can be considered superfluous sounds.

24 E.g., export can represent a noun /’ekspəHt/or a verb/ek’spəHt/.

25 E.g., in such pairs as tail: tale, Liberal: liberal. Cf. also #4.

26 MacLuhan, H. M., The Gutenberg Galaxy (Toronto, 1962), p. 238 Google Scholar, coupled with his even more surprising assertion that “the printing press made bad grammar possible” (p. 231); cf. note 27.

27 The language of the Bible has long formed the standard for Hebrew. The earliest record comes from Saadya Ga’on (d. 942) : “I likewise saw a number of Israelites who do not observe the simplest rules of our language, not to mention the more difficult ones, and they mispronounce it in their speech. . . .” (quoted in Skoss, Saadya Ga’on, p. 58).

28 This standard is now less admired than formerly. It is well illustrated by the writings of Shaw, G. B.. See a collection, including excerpts from “Pygmalion,” the most obvious example, in Tauber, A., ed., Shaw, G. B. on Language, (London, 1965)Google Scholar.

29 See Kahle, P. E., The Cairo Geniza (Oxford, 1959), p. 141 f., and references thereGoogle Scholar.

30 See, for example, Aristophanes’ fragment 198, on teenage slang, and 685, distinguishing city speech (effeminate) and country speech (vulgar) from “average Attic,” in J. M. Edmonds, The Fragments of Attic Comedy, vol. 1 (Leiden, 1957).

31 Examples can be found in T. Pyles, Words and Ways of American English (London, 1954), Chapter 11. His remarks show that in many (most?) cases the condemnation of a particular structure stems from a theoretic ideal, not actually adhered to in the written language. An example of a community speech form contravening the written ideal occurs in the story quoted from R. A. Hall (p. 204), showing a case where a man using “upper class” speech forms was not fully admitted to a community until he adopted its “lower class” forms. This sort of thing is, of course, parallelled by the necessity of using the “correct” slang in various groups, of which teenagers are a standard example.

32 Written language, of course, is also used in different forms in different situations : e.g., legal documents, newspaper reports, intimate letters.

33 An extreme example is G. M. Schramm, The Graphemes of Tiberian Hebrew (Berkeley, 1964), in which the graphemes, the supposed subject of the study, are transcribed (not transliterated, p. 32, note 1) into sound symbols before their distribution is analysed. Important features of the writing system, such as the “shewa sign” and word final “alef” and “he” are not marked in the transcription, and their distribution therefore not analysed, so that Schramm cannot be said to be analysing the writing system. The sound symbols used are not intended to represent phonemes (p. 32), so it would be difficult to say just what he is analysing, and it is difficult to see how anything but gain could have resulted from analysing the distribution of the graphemes themselves, and basing the phonological transcription on this information. There are, however, some notable exceptions to this tendency to treat writing as an intended record of speech, e.g., Halliday, M. A. K., “Secret History ot the Mongols,” The Language of the Chinese (Oxford, 1959)Google Scholar.

34 Opinions on the significance of the standard signs for the vowels of Biblical Hebrew ranges from “basically phonemic” (so S. Morag, The Vocalization Systems of Arabic, Hebrew, and Aramaic ¡The Hague, 1962], p. 29) to “basically phonetic” (so Rabin, C., “The ‘Small Vowels’ in Tiberian Hebrew,” in Studies in Honour of N. H. Tur-Sinai [Publications of the Israel Society for Biblical Research 8, Jerusalem 1960, p. 206]Google Scholar. There is a correspondingly wide variety of phonemic interpretations of the system.