Published online by Cambridge University Press: 22 April 2026
Linguistic science is today in every sense of the word an international science. Few disciplines can lay better claim to this term than ours, in view of its universally and specifically human subject matter, as well as its bearing on the interrelationship and communication of nations. Even within our generation a vast expansion of linguistic study has taken place when compared with the preceding one. It is characteristic that around 1930 contributions to phoneme theory were being made by men as widely scattered as Trubetzkoy in Austria and Yuen Ren Chao in China. This was already a forward step over the much narrower field of Rask and Grimm, but we have seen a still more intense effort in the last two decades. From the occasional contributions of isolated professors we have in our own country proceeded to a concerted and eager program of linguistic research sponsored by a whole group of scholars working together in a vigorous and well-knit Linguistic Society.
1 Delivered as the Presidential Address at the meeting of the Linguistic Society in Chicago, 29 December 1950. The present form of the paper owes much to discussion with colleagues at the University of Wisconsin.
2 Cf. SIL 8.8 (1950), where the label ‘unscientific’ is applied to ‘much of the European structural studies’; ibid. 8.100: ‘the usual kind of European philosophizing on the basis of insufficient evidence’.
3 As honorable exceptions note Eli Fischer-Jørgensen, Travaux du Cercle linguistique de Copenhague 5.214–34 (1949); and W. E. Collinson, Lingua 1.306–32 (1948). The letter's survey of American contributions to metalinguistic discussion is outstanding.
4 Some of the more important are: Bloch and Trager, Outline of linguistic analysis (Baltimore, 1942); Harris, Morpheme alternants in linguistic analysis, Lg. 18.169-S0 (1942); Pike, Taxemes and immediate constituents, Lg. 19.65–82 (1943); Harris, Discontinuous morphemes, Lg. 21.121–7 (1945); Chao, The logical structure of Chinese words, Lg. 22.4–13 (1946); Harris, From morpheme to utterance, Lg. 22.161–83 (1946); Bloch, Studies in colloquial Japanese, JAOS 66.97–109 (1946), Lg. 22.200–48 (1946), JAOS 66.304–15 (1946), Lg. 26.86–125 (1950); Wells, Immediate constituents, Lg. 23.81–117 (1947); Voegelin, A problem in morpheme alternants and their distribution, Lg. 23.245–54 (1947); Pike, Grammatical prerequisites to phonemic analysis, Word 3.155–72 (1947); Hockett, Problems of morphemic analysis, Lg. 23.321–43 (1947); Bloch, English verb inflection, Lg. 23.399–418 (1947); Pike, Phonemics (Ann Arbor, 1947); Bloch, A set of postulates for phonemic analysis, Lg. 24.3–47 (1948); Harris, Componential analysis of a Hebrew paradigm, Lg. 24.87–91 (1948); Voegelin, Distinctive features and meaning equivalence, Lg. 24.132–5 (1948); Nida, The analysis of grammatical constituents, Lg. 24.168–77 (1948); Pittman, Nuclear structures in linguistics, Lg. 24.287–92 (1948); Nida, The identification of morphemes, Lg. 24.414–41 (1948); Wells, Automatic alternation, Lg. 25.99–116 (1949); Nida, Morphology, 2d ed. (Ann Arbor, 1949); Joos, Description of language design, Journal of the Acoustical Society of America 22.701–9 (1950).
5 Cf. D. D. Runes, Dictionary of philosophy, 4th ed. (New York, 1942); Charles Morris, Signs, language, and behavior 179 (New York, 1946); R. Carnap, Introduction to semantics 4 (Cambridge, Mass., 1942); id., Philosophy and logical syntax (London, 1935).
6 Cf. C. H. Borgström's use of ‘meta-research’ in this sense, Acta linguistica 5.1–14 (1945–49). L. Hjelmslev uses ‘metasprog’ in his work Omkring sprogteoriens grundlæggelse (see fn. 8).
7 George L. Trager. The field of linguistics 2 (Norman, Okla., 1949).
8 Copenhagen, 1943; hereafter abbreviated OSG.
9 Recherches structurales 1949: Interventions dans le débat glossématique (Copenhague, 1949).
10 Harris, Lg. 22.161; Hjelmslev, OSG 72. Cf. Joos, op.cit. 22.702: 'We must make our “linguistics” a kind of mathematics, within which inconsistency is by definition impossible.
11 Harris, Lg. 22.177 fn. 26; Hjelmslev, OSG 9. Cf. Nida, Lg. 24.437 fn. 40.
12 Hjelmslev, OSG 28; Wells, Lg. 23.92. Harris originally assumed a segmentation, and then worked his way up to more inclusive classes ‘from morpheme to utterance’ (Lg. 22.17S Wells showed how an analysis in terms of immediate constituents could be worked from the whole down, and pointed out that it matters little in which direction one works (Lg. 23.101 Hjelmslev grants the same when he defines ‘induction’ (working up) and ‘deduction’ (working down) as functions of one another; his insistence on the priority of deduction is basez on the sound observation that the linguist must start with the unanalyzed whole text.
13 Wells, Lg. 23.84; Hjelmslev, OSG 30.
14 Harris, Lg. 22.161 ff.
15 OSG 67.
16 Cf. Hjelmslev, OSG 59: 'For an understanding of the structure of language it is of the greatest significance to make it clear that this principle shall be extended to apply also to all the other invariants of the language, regardless of their degree or place in the system.' Harris writes of his morpheme analysis, Lg. 18.179: 'it shows that we can arrange alternants into units in exactly the same manner as we arrange sound types (positional variants) into phonemes.'
17 Journal of the Acoustical Society of America 22.701–8 (1950).
18 Nida distinguishes between identification and distribution of morphemes in somewhat the same way as here suggested; cf. Morphology 78.
19 Bloomfield called it ‘altering the word’; Trubetzkoy, Grundzüge der Phonologie 31–2 (TCLP 7; Prague, 1939), spoke of certain sounds as being Vertauschbar'; the term ‘substitution’ was apparently first used in America by Morris Swadesh, The phonemic principle, Lg. 10.124 (1934).
20 Lg. 23.83–4; cf. Pittman's Premise 1, which is here made the only principle for determining a satellite.
21 Lg. 24.287–92.
22 OSG 23.
23 Lg. 24.22–6; 26.8P.
24 Lg. 26.90.
25 Lg. 23.328.
26 OSG 71.
27 Norsk tidsskrift for sprogvidenskap 12.1–29 (1939).
28 Note that Wells describes classes established by distributional analysis as only ‘roughly coincident classes’ (Lg. 23.82), while Harris leaves it to the linguist to decide which ones he shall use in order to get ‘the most convenient total description’ (Lg. 22.177).
29 Trager analyzed American English in 1940 as having six vowels, in 1947 as having nine (Lg. 23.141); Swadesh analyzed American English diphthongs in 1935 as single phonemes, in 1947 as double (Lg. 23.137); Hockett analyzed the Chinese aspirated stops in 1944 as single consonants (in Spoken Chinese: Basic course), in 1947 as clusters with h (JAOS 67.258); Bloch's phonemic analysis of Japanese in 1946 had several features, including a phoneme q, which he discarded in 1950 (Lg. 26.112).
30 Cf. Hockett, Lg. 23.327 fn. 20.
31 Lg. 22.180.
32 Lg. 24.179.
33 Word 3.155–72 (1947).
34 Lg. 23.81.
35 OGS 7.
36 The field of linguistics 4.
37 It is interesting to note that even a Hjelmslev disciple like Eli Fischer-jørgensen is, in the final instance, driven back on phonetic similarity (‘la parenté phonique‘); cf. Travaux du Cercle ling. de Copenhague 5.227.
38 Op.cit. 22.708.
39 Lg. 23.341 fn. 38.
40 It will appear from the foregoing that I agree with the criticisms of the Hjelmslev theory advanced by Carl Hj. Borgström, The technique of lingusitic description, Acta linguistica 5.1–14 (1945–49), and A problem of methods in linguistic science: The meaning of its technical terms, Norsk tidsskrift for sprogvidenskap 14.191–228 (1947).
41 Cf. Hockett, A note on ‘structure’, IJAL 14.269–71 (1948).