No CrossRef data available.
Published online by Cambridge University Press: 22 April 2026
The purpose of this paper is to suggest a technique for determining the morphemes of a language, as rigorous as the method used now for finding its phonemes. The proposed technique differs only in details of arrangement from the methods used by linguists today. However, these small differences suffice to simplify the arrangement of grammars.
1 L. Bloomfield, Language 161 (New York, 1933).
2 M. Swadesh and C. F. Voegelin, A Problem in Phonological Alternation, Lang. 15.4 (1939).
3 Z. S. Harris, Linguistic Structure of Hebrew, JAOS 61.155 (1941).
4 Ibid. 159.
5 L. Bloomfield, Menomini Morphophonemics, TCLP 8.105 (1939).
6 R. H. Lowie, Z. S. Harris, and C. F. Voegelin, Hidatsa Texts, Indiana Historical Society Prehistory Research Series 1.192 fn. 38 (1939).
7 Cf. Bloomfield, Language 217, where it is shown that the relation between masculine and feminine adjectives in French can be most simply described by regarding the feminine forms as basic.
8 This excludes synonyms, i.e. morphemes of approximately similar meaning, which usually occur in the same positions: a fine youngster, a fine lad.
9 As in a score of voices, but twenty voices. However, we may consider that twenty occurs in the same position as score in a twenty ‘a $20 bill’.
10 Lang. 15.5 ff. (1939). The formulation presented here is a restatement, in terms of morpheme alternants, of their morphophonemic analysis.
11 Or if it is always short while each of its neighboring vowels is either always long or always short.
12 Unless it is next to a basically long vowel, in which position even a neutral vowel is always short.
13 But a neutral vowel next to one with basic length is always short.
14 Certain additional general statements involving /ˀ/, etc., must be applied before the statement about vowel length.
15 M. B. Emeneau, An Echo-Word Motif in Dravidian Folk Tales, JAOS 58.553–70 (1938); Echo Words in Toda, New Indian Antiquary 1.109–17 (1938).
16 Bloomfield, TCLP 8.105–15 no. 10–2.
17 What is called external sandhi, therefore, differs from internal sandhi merely in that the former contains statements which have word juncture as a necessary part of their determining environments, while the latter does not. In some languages, alternants next to word juncture may differ so much from those which are not, and differences determined by word juncture may have so many features in common, that it becomes convenient to arrange all statements involving word juncture environments together. In other languages, however, where many statements apply to environments both within words and across word juncture, it is simpler not to distinguish external from internal sandhi.
18 Bloomfield, TCLP 8.105–15 no. 18.
19 Ibid. no. 15.
20 Ibid. no. 18.
21 E.g. the complete difference between go and wen- exists only between these two sequences of phonemes, hence (barring homonyms) only in this particular unit. However, the partial difference between knife and knive- can occur between any two sequences of phonemes that contain /f/ and /v/.
22 Bloomfield, TCLP 8.105–15 no. 13.
23 Except for sequences of unique alternants (see last paragraph of §4.2). Such cases should be indicated in a special list of alternative possibilities.