Phonological theory: A brief comment

I was flattered (and more than a little puzzled) to see the length and detail of Chomsky and Halle's recent reply to my earlier article in this journal (cf. JL I(I965) 13-34 & 47-I38). I had hoped to take up a few of the points they raise and raise a few more of my own, but owing to exigencies of both time and space, I find I must be content to restate my three main points as clearly as I can, to reaffirm my belief in them, and to mention a few of my reasons. (i) I still do not like the use of such strong God's truth language, saying always 'such-and-such is X' instead of 'such-and-such may conveniently be described or looked at as X'. And even if I believed their assertions were as securely based as those a physicist may make about atomic nuclei, I would say that this makes no difference, and that a physicist who says what the atom Is must be speaking in some sort of figurative language. Incidentally, one of the difficulties of intuition as a check on theory is that the problem of the self-fulfilling prediction comes in. I remember well the beautiful 4-stress 4-pitch dialects spoken by some of the Foreign Service Institute personnel some years back; and there outside observers had access at least to the physical manifestations. I am not accusing anyone of subjectivity; I am merely wishing for more secure lines of communication. When I see red and you see red, even though I have no way of knowing that the quality of your perception resembles mine, I at least am able to count votes, and devise tests which will locate red-green blindness. But here the chances of comprehension are often very low, and the will to believe is not enough; real faith is required. (2) On my disagreement with Chomsky about traditional phonemes and phonemics, I would like to add a little. I now see that for this level a kind of uniqueness is important, and perhaps bi-uniqueness. The name 'taxonomic' is of course merely pejorative, not informative; the function of such phonemes as conceived by the pioneers of the thirties and forties was to explain the manner in which speakers can REPEAT (not just MIMIC) new words. This kind of repetition implied a systematic analysis of the sound system WITHOUT ANY POSSIBLE REFERENCE TO DEEP MORPHOPHONEMIC POTENTIAL (hence not Chomsky's systematic phonemic level) and without the kind of detail which permits international phoneticians to discuss similarities and differences between sounds in different languages (hence not Chomsky's systematic phonetic level).