1. Setting of the problem. When in 1937 Douglas Chrétien and I published Quantitative Classification of Indo-European Languages (Lg. 13.83–105) it aroused a little flurry of attention but no follow-up, except an extension of the method by ourselves to include Hittite (Lg. 15.69–71 [1939]). For some years I had looked on our essay as a dead limb of effort, especially since my coauthor came subsequently to doubt the fruitfulness of the method on the ground that most of the values found by us lacked statistical significance.

This study presents a summary of the phonological structure of the Standard Colloquial Bengali (SCB) of the educated people of Calcutta. It offers neither a mass of detailed data nor startling innovations of theory, but is intended to provide basic explanation for someone who wants to know the facts about the phonology of the language. There are astonishingly few straightforward descriptions of the phonologies of the various modern languages of South Asia, and this attempts to fill that need for one of them. Bengali is one of the major Indic (or Indo-Aryan) languages of South Asia; in number of speakers and general importance it is second only to Hindi-Urdu, and in quality and quantity of literary production it is second to none. It is spoken by over seventy million people, somewhat more than half of them in East Pakistan and the remainder in West Bengal, India. There is a standard literary form of the language, the sādhu bhāṣā, quite diverse regional dialects, and a Standard Colloquial, the calit bhāṣā, which is described here. Bengali is customarily written in a special alphabet of the Devanagari type, the Bengali alphabet.

The Proto-Semitic vowels, as usually reconstructed, are identical with those of Classical Arabic: short */i a u/, long /*ī ā ū/. In addition, a number of the long vowels are recognized, on the basis of morphophonemic alternation, as resulting from earlier sequences of short vowels and */y/ or */w/. For example, on the basis of the cognate set Heb. /dān/, Acc. /dānu/, Syr. /dān/, Arab. /dāna/ ‘he judged’, Gray reconstructs Proto-Semitic */dāna/. Since this long-vowel correspondence contrasts, however, with that of the set Heb. /šālōš/, Acc. /šalāšu/, Syr. /tәlāθ/, Arab. /θalāθun/, Eth. /šalās/ ‘three’, where he also reconstructs an */-ā-/, he posits a presumably pre-Proto-Semitic */aya/ for the first set on the basis of the imperfects Heb. /yāÐīn/, Acc. /ˀidīn/, Syr. /nәÐīn/, and Arab. /yadīnu/. The methodological error here is immediately apparent: if the long vowel of */dāna/ contrasts in its reflexes in the descendent languages with the long vowel of */θalāθu/, then the contrast existed in Proto-Semitic, and the morphophonemic alternation is irrelevant in determining it. A similar situation exists with respect to the verb Heb. /qām/, Syr. /qām/, Arab. /qāma/, Eth. /qōma/ ‘he stood’ (for the Acc. reflex cf. Acc. /kānu/ ‘he existed’, Arab. /kāna/ 'he was'), where Gray reconstructs */qāma/. To explain the Heb. reflexes one has to resort to an analogical formation from the forms with suffixes (/dantā/ 'you judged', /qamtā/ ‘you stood‘) where the vowel remained short, with subsequent stress lengthening of the reestablished short vowel in the simple forms. For Eth. one must recognize an */aw/ nucleus in the suffixed forms which went to /ō/ and was reestablished in the simple form. It is much easier, however, to recognize that */aya/ and */awa/ existed in Proto-Semitic.

Immediate constituent analysis is based on an implicit assumption that linguistic structures—especially syntactic structures—are layered structures amenable to analysis by progressive dichotomous cutting. This paper suggests an alternative assumption, viz., that some linguistic structures are layered while others are ordered like beads on a string. Negatively this alternative assumption leads to a rejection of conventional immediate constituent analysis along with its reluctance to recognize more than two constituents on the same level of cutting. Positively, it leads to the development of a string constituent analysis in which grammatical strings are discovered and described (with mention of any observable layering characteristics), and in which constituent substitution points within the strings are likewise discovered and described. It is important to note that, while string constituent analysis is not geared to the assumption of the relevance of layering at all points in the structure of a language, it does not rule out the observation and description of layering where this is relevant.