¹ Franklin Edgerton, ‘Sir William Jones: 1746–1794‘, JAOS 66.230–9 (1946).

² Here and later, in material about the 1870's, I rely mainly on Holger Pedersen, Linguistic science in the nineteenth century, tr. by John W. Spargo (Cambridge, Mass., 1931).

³ Lausanne, 1916; 2nd edition, Paris, 1922.

⁴ The Hague, 1957.

⁵ Pedersen, op.cit.; J. O. H. Jespersen, Language: Its nature, development, and origin (London and New York, 1922), chs. 1–2; Leonard Bloomfield, Language (New York, 1933), ch. 1.

⁶ There seems to be a clustering of such scholars in England. Thus George B. Milner (uncertain rather than committed on the point) felt it appropriate to publish an article entitled ‘Notes on the comparison of two languages (with and without a genetic hypothesis)‘, Linguistic comparison in South East Asia and the Pacific (London, 1963).

⁷ See fn. 2.

⁸ The outer quotation is from Pedersen, op.cit. 293–4; the inner one is his (or Spargo's) rendition of Brugmann's ‘aus dem hypothesentrüben Dunstkreis der Werkstätte, in der man die indogermanischen Grundformen schmiedet … in die klare Luft der greifbaren Wirklichkeit und Gegenwart’.

⁹ Lg. 33.375–407 (1957); Indiana University Research Center in Anthropology, Folklore, and Linguistics, publication 12 = IJAL 26:3, Part 2 (1960).

¹⁰ For this reason, when Mary R. Haas published evidence pointing towards a genetic connection between Algonquian and Gulf, she felt it best to make overt that this was not to be interpreted as displacing any directions of relationship already suspected for either: ‘A new linguistic relationship in North America: Algonkian and the Gulf languages’, Southwestern journal of anthropology 14.231–64 (1958), p. 231.

¹¹ Op.cit. (fn. 5), ch. 20, esp. p. 354.

¹² The transformationalists have adopted ‘merely taxonomic’ as an epithet of opprobrium, which is all right except that they sometimes forget the ‘merely’. A taxonomy can be based on any criteria whatsoever, including criteria that prove to be of no relevance. But the search for, and development of, a significant taxonomy has played an important role in the history of most or all branches of science.

¹³ See fn. 2.

¹⁴ Jespersen, op.cit. (fn. 5) 166–7.

¹⁵ Saussure: op.cit. (fn. 3). Bloomfield: first in practice (without the word ‘phoneme’) in Tagalog texts (Urbana, 1917) = University of Illinois Studies in Language and Literature 3:2–4; then in various reviews; ‘A set of postulates for the science of language’, Lg. 2.153–64 (1926); ‘On recent work in general linguistics’, Modern philology 25.211–30 (1927). Jones: Outline of English phonetics ² (Leipzig and Berlin, 1922). Sapir: Language (New York, 1921); ‘Sound patterns in language’, Lg. 1.37–51 (1925); ‘La réalité psychologique des phonèmes’, Journal de psychologie normale et pathologique 30.247–65 (1933). Trubetskoy: ‘Zur allgemeinen Theorie der phonologischen Vokalsystem’, TCLP 1.39–67 (1929), and subsequent articles in TCLP, culminating in his Grundzüge der Phonologie (Prague, 1939) = TCLP 7. Jakobson's earliest phonological papers are in Russian or Czech, and thus inaccessible to me: see the bibliography, pp. 1–12, in Morris Halle et al., eds., For Roman Jakobson (The Hague, 1956). For Trubetskoy and Jakobson, see now also Josef Vachek, A Prague School reader in linguistics (Bloomington, 1964).

¹⁶ J. R. Firth, ‘The word “phoneme” ‘, Le maître phonétique, no. 46 (1934), reprinted in Papers in linguistics 1934–1951 1–2 (Oxford, 1957).

¹⁷ Of course, the new view stemmed also from what Bloomfield called the 19th-century ‘philosophical-descriptive’ tradition—op.cit. (fn. 5) 18 f.—and drew from a long and rapidly developing tradition of ‘practical’ grammar-writing by missionaries, travelers, and anthropologists, some of whom had no knowledge of developments in the main stream of linguistic science. A synchronic view was in itself not particularly new. Truly new was the altered view of the relation of the synchronic and the diachronic.

¹⁸ I remember conversations with Roman Jakobson in the mid-1940's, when he praised Saussure but at the same time characterized the historical linguistics of the latter (even of the Cours) as ‘purely neogrammarian’. This was clearly intended as disapprobation of that particular portion of Saussure's views.

¹⁹ Économie des changements phonétiques (Berne, 1955), incorporating the results of articles going back almost 20 years.

²⁰ See fn. 5. A reading of this chapter as far as the middle of p. 365, reinterpreting all references to ‘phonemes’ as though Bloomfield were speaking of distinctive features, yields a point of view on sound change almost identical with my own. But the part of the chapter that begins towards the bottom of p. 365 must be omitted: Bloomfield here turns, in my opinion, in a logically and empirically impossible direction.

²¹ ‘In at least some environments‘—because there can also be environments in which it does not matter whether the actual sound falls into one or another of a set of adjacent regions (in the limiting case, there are situations in which any loud noise conveys the crucial message; but this is perhaps not to be classed as language). But for each pair of regions that share a boundary, there must be at least one environment in which the contrast is maintained.

²² I phrase these assumptions using the word ‘sentence’ in its current popular sense, rather than the older term ‘utterance’. Many of the early phonologists, of course, worked largely or exclusively in terms of (isolated) words. I pass over the complications that arise from this defective frame of reference, as not germane to the present discussion.

²³ Not only his writings but also his conversations with colleagues and students revealed Bloomfield as perhaps more insistent on this point than anyone else. It was a matter of basic principle with him. On the other hand, when it came to applying the principle to specific languages, he gave many expressions of uncertainty. He certainly did not assume that discovering the ‘phonemic’ system of a language was an easy matter. His Menomini notes (in my hands) reveal various changes of treatment during the 25 years of his work on that language; the posthumously published description—The Menomini language (New Haven, 1962)—incorporates some modifications that must have been made shortly before his death, since he uses terms (‘juncture‘) that were just then becoming current.

²⁴ ‘Linguistic elements and their relations’, Lg. 37.29–53 (1961), esp. §1.

²⁵ Hermann Paul, Prinzipien der Sprachgeschichte (Halle, 1880).

²⁶ See the works cited in fn. 15; also his reviews in the period 1925–35, as listed in the bibliography (pp. 95–7) appended to the obituary notice by Bernard Bloch, Lg. 25.87–98 (1949). Of the reviews, probably the most important is that of Eduard Hermann, Lautgesetz und Analogie (Berlin, 1931), in Lg. 8.220–33 (1932). The best general statement of the recommended view is his ‘Language or ideas?‘, Lg. 12.89–95 (1936).

²⁷ Op.cit. in fn. 23.

²⁸ The phrasing is my own, but is intended to agree with Chomsky's various versions, of which the latest is in Noam Chomsky and George A. Miller, ‘Introduction to the formal analysis of natural languages’, in R. Duncan Luce et al., eds., Handbook of mathematical psychology (New York and London, 1963), vol. 2, ch. 11. The assumption most often made currently, that sentencehood is a sharp criterion, is admittedly (283–4) only a first approximation.

²⁹ In mathematical usage, a transformation is (most often) a mapping of a set, especially a space, onto itself. This bears only a vague resemblance to Chomsky's linguistic use. Closer is Rudolph Carnap's usage of the pair ‘formation and transformation’ in logical syntax: The logical syntax of language, tr. by Amethe Smeaton (London, 1937) [German original, enlarged in the translation, Die logische Syntax der Sprache (Vienna, 1934)]. I find, to my amusement, that in the morphophonemics part of my 1939 doctoral dissertation I speak of ‘transformations’, in a somewhat Chomskyan way, probably through the influence of Carnap: the thesis was published as ‘Potawatomi’, IJAL 14.1–10, 63–73, 139–49, 213–35 (1948), and the term appears on p. 4. In the summer of 1945, at the Linguistic Institute at the University of Michigan, I gave a luncheon talk on ‘The syntax of formation and of transformation’, still under Carnap's influence; this was never published. It is not at all my purpose here to claim priority—probably none of the above, except possibly Carnap's work (which is not linguistics) was known to Chomsky. The fact that there are usually forerunners for a breakthrough detracts not at all from the credit due those who definitively achieve the advance. Zellig S. Harris's work using the term ‘transformation’ is virtually contemporary with the beginning of Chomsky's, and is too well known to require bibliographical references here.

³⁰ A partial bibliography should be of some interest; I arrange it chronologically. Noam Chomsky, ‘Three models for the description of language’, IRE transactions IT2.113–24 (1956). S. C. Kleene, ‘Representation of events in nerve nets and finite automata’, Automata studies = Annals of mathematics studies 34.3–41 (Princeton, 1956). Chomsky, ‘On certain formal properties of grammars’, Information and control 2.137–67 (1959). S. Scheinberg, ‘Note on the Boolean properties of context free languages’, Information and control 3.372–5 (1960). Y. Bar-Hillel, Perles, and Shamir, ‘On formal properties of simple phrase-structure grammars’, Zeitschrift für Phonetik, Sprachwissenschaft, und Kommunikationsforschung 14.143–72 (1961). E. Shamir, ‘On sequential languages’, Technical report no. 7, Applied Logic Branch, The Hebrew University (Jerusalem, November 1961). S. Ginsburg and H. G. Rice, ‘Two families of languages related to ALGOL’, Journal of the Association for Computing Machinery 9.350–71 (1962). S. Ginsburg and G. F. Rose, ‘Operations which preserve definability in languages’, JACM 10.175–95 (1963). S. Ginsburg, ‘Some recursively unsolvable problems in ALGOL-like languages’, JACM 10.29–47 (1963). S. Ginsburg and E. H. Spanier, ‘Quotients of context-free languages’, JACM 10.487–92 (1963). Chomsky, ‘Formal properties of grammars’, in R. Duncan Luce et al., eds., Handbook of Mathematical Psychology (New York and London, 1963), vol. 2, ch. 12. See also a forthcoming book on mathematical linguistics by David G. Hays (title undetermined) and, for developments in rather different directions, my forthcoming monograph Language, mathematics, and linguistics.

³¹ Sydney M. Lamb, ‘MT research at the University of California, Berkeley’, in Proceedings of the National Symposium on Machine Translation, H. P. Edmundson, ed. (Englewood Cliffs, N. J., 1961); ‘On alternation, transformation, realization, and stratification’, Monograph series on languages and linguistics 17.105–22, Georgetown University Institute of Languages and Linguistics; ‘The sememic approach to structural semantics’, in A.Kimball Romney and Roy Goodwin D'Andrade, eds., Transcultural studies in cognition = American anthropologist 66:3 part 2 (1964); ‘Linguistic structure and the production and decoding of discourse’, to appear in Victor E. Hall, ed., Speech, language and communication (UCLA Forum in Medical Sciences); Outline of stratificational grammar, to be published by Georgetown University Institute of Languages and Linguistics. Related articles by other investigators: H. A. Gleason Jr., ‘The organization of language: A stratificational view’, Monograph series on languages and linguistics 17.75–95 (1964); Petr Sgall, ‘Zur Frage der Ebenen im Sprachsystem’, Travaux linguistiques de Prague (1964); James H. White, ‘The methodology of sememic analysis with special application to the English preposition’, Mechanical translation (1964).

³² Examples: Joseph Grimes, Huichol Syntax (Cornell Dissertation 1960; The Hague, 1964); Gary J. Parker, Ayacucho Quechua Grammar (Cornell Dissertation 1964), to be published (in Spanish) in Peru.

³³ The first paragraph of ‘On recent work in general linguistics’, Modern philology 25.211–30 (1927).

³⁴ ‘Mentalism in linguistics’, Lg. 40.111–23 (1964).

³⁵ Of course, my interpretation may not be at all what Katz intended: I am here reading Bloomfield as narrowly and Katz as liberally as possible. As I interpret his point, he has forerunners: for example, I should say that this was the point (perhaps not well expressed) of my ‘Biophysics, linguistics, and the unity of science’, American scientist 36.558–72 (1948), which he lumps with Bloomfield's own statements on antimentalism. Are we really fair to Bloomfield in accrediting him only with the first of the two points? This is very hard to say: the climate of opinion changes (though not so rapidly as does terminology), and our understanding of something written as recently as thirty years ago is all too apt to be colored by our emotional attitude towards the writer. To this I must plead as guilty as I believe Katz is. The ultimate decision must be left to the objective historian of our science.

³⁶ Were the attacks in print I should cite them. Since those that have come to my attention were largely made in informal conversation that is supposedly off the record, I believe it would be inappropriate to identify the individuals involved.

³⁷ Note well that there is much important work to be done in our science that does not require the mastery of mathematics. My severe remark in the text refers only to the technical follow-up of the fourth breakthrough.

³⁸ This has been reported to me as one of Chomsky's remarks in viva voce discussion; I do not find it in his publications, and he may, of course, disclaim it.

³⁹ Morris Halle, ‘Phonology in generative grammar’, Word 18.54–72 (1962). Chomsky's most extended discussion is pp. 944–73 of ‘The logical basis of linguistic theory’, in Horace G. Lunt, ed., Proceedings of the 9th International Congress of Linguists, Cambridge, Mass., August 27–31 1962 914–78 (The Hague, 1964). Of course Chomsky is right in asserting that in handling any one language one must refer to ‘phonetic universals’. The sources of error are two: the wrong mathematical formalism, and what can only be described as tin-ear practical phonetics. On the latter, note that Chomsky and Miller (op.cit. fn. 28) cite, as an example of a pair of English sentences differing only at the end, the following: The good candy came anyway / The good can decay many ways (280). In the light of the progress that has been achieved in the last two decades in American English phonology, this appears as a ridiculous instance of the baby-out-with-bath syndrome.

⁴⁰ Another thing that can surely happen is for a point on a most probable path, which is not (at a certain time) an allophone, to become (subsequently) a local maximum relative to other points on the path as well as relative to the rest of its neighborhood, thus becoming an allophone. This is the historical phenomenon of epenthesis.

⁴¹ Charles F. Hockett and Robert Ascher, ‘The human revolution’, Current anthropology 5.135–68 (1964).

⁴² That the so-called ‘comparative method’ of cultural anthropology is totally unlike that of linguistics becomes obvious upon the most superficial examination of the anthropological literature; for example, A. R. Radcliffe-Brown, ‘The comparative method in social anthropology’, in his Method in social anthropology 108–29, ed. by M. N. Srinivas (Chicago, 1958); Franz Boas, ‘The limitation of the comparative method of anthropology’, in his Race, Language, and Culture 270–80 (New York, 1940); Erwin H. Ackerknecht, ‘On the comparative method in anthropology’, in Robert F. Spencer, ed., Method and perspective in anthropology 117–25 (Minneapolis, 1954). Of course, our own use of the term is the unusual one; when we actually use a method more like the ‘comparative method’ of other social sciences, we say (following Benjamin Whorf's suggestion) that we are doing contrastive linguistics.

⁴³ Op.cit. fn. 3.

⁴⁴ The delivery of this address to the Society of course did not include the footnotes, and omitted the eleventh paragraph of §2 on reflexivity. It concluded with the following remark, here relegated to a footnote because its homiletic tone is hardly appropriate for the style of our journal; yet I feel that what was actually said should be on record: ‘The scholar earns immortality only of the sort he bestows on those who have gone before him. As we extend the power and flexibility of our new tools, let us always temper passion with humor; let us never favor, nor disfavor, the new simply because of its novelty; let us dedicate our talents to building on our heritage, not to tearing it down, praising our predecessors for their wisdom and ignoring their folly—replacing a nail here or a plank there when we must, but always with humility rather than Schadenfreude when a bright old idea must give way to a bright new one.‘