As with words, it is the practical usefulness of paradigms that accounts for their role in descriptions of grammatical systems. Comprehensive descriptions of languages with intricate morphological systems tend to enumerate exemplary paradigms and patterns at different levels of generality. This is particularly true of grammars prepared for pedagogical purposes, as Matthews (1991) observes:
Pupils begin by memorising paradigms. These are sets of words as wholes, arranged according to grammatical categories. This is not only traditional, it is also effective. They learn that different members of a paradigm are distinguished by their endings... They can then transfer these endings to other lexemes, whose paradigms they have not memorised... It seems unlikely that, if a structuralist method or a method derived from structuralism were employed instead, pupils learning Ancient Greek or Latin—or, for that matter, Russian, Modern Greek or Italian—would be served nearly so well. (Matthews 1991:187f.)
No practical alternative has been devised that replaces exemplary patterns by a more theoretical description in which, e.g., abstract underlying representations were related to surface realizations by general rules or constraints.1 In part, this is due to the fact that theoretical models since Bloomfield’s time have not been designed with practical (or comprehensive) description in mind, so that the output of an analysis may range from nontransparent to indeterminate (Karttunen 2006). The same general suspicion of practical notions that coloured Bloomfieldian attitudes towards the use of words for “the systematic study oflanguage” also plays a role in the attitude towards exemplary patterns.
The Bloomfieldian scepticism towards words and paradigms also derives from more fundamental assumptions. By reducing words to independent aggregates of smaller parts, Bloomfieldian models disrupt any properties of words and any relations between words that cannot be recast in terms of properties of their parts, or in terms of relations between their parts. The organization of words into classes, or the definition of relations over these classes clearly falls outside the purview of this type of model. Yet this ‘atomistic’ perspective rests ultimately on little more than a theoretical aesthetic that places a premium on what Bloomfield terms “scientific compactness”:
- 1 There have been occasional attempts to establish descriptive grammars on a more ‘theoretical’ basis. An instructive example is provided by the use of Tagmemics (Pike 1943, 1967) as the foundation for descriptive grammars published by the Summer Institute of Linguistics. The success of this effort can be gauged from the assessment that “Tagmemics chokes on its own terminological complexity” (Hockett
Word and Paradigm Morphology. First edition. James P. Blevins © James P. Blevins 2016. First published 2016 by Oxford University Press
The inflectional forms are relatively easy to describe, since they occur in parallel paradigmatic sets; the traditional grammar of familiar languages gives us a picture of their inflectional systems. It may be worth noticing, however, that our traditional grammars fall short of scientific compactness by dealing with an identical feature over and over again as it occurs in different paradigmatic types. Thus, in a Latin grammar, we find the nominative-singular sign -s noted separately for each of the types amicus ‘friend’, lapis ‘stone’, dux ‘leader’, tussis ‘cough’, manus ‘hand’, facies ‘face’, when, of course, it should be noted only once, with a full statement as to where it is and where it is not used. (Bloomfield 1933:238)
For Bloomfield, as for many of his successors, it is self-evident that the distribution of an inflectional exponent such as nominative singular -s should be expressed declaratively, by means of “a full statement”, rather than exhibited in traditional paradigm tables. Whatever the merits of this notion of ‘compactness’ for the preparation of written grammars (or for the transmission of an oral grammatical tradition), it has no established relevance to language acquisition or use. There is at present no evidence that the language faculty imposes memory demands that strain the storage capacity of the human brain, or that linguistic notions of ‘compactness’ would be relevant to reducing this load.
-  Or, as Bolinger (1979:110) puts it, “The human brain is not a vestigial organ”.