The Post-Bloomfieldian legacy
Arguably the most influential claim expounded within the Bloomfieldian tradition is the idea that morphological (and, indeed, linguistic) analysis is fundamentally a process of identifying and classifying minimal recurrent units, and that the classification of larger units is derivative of the analysis of these minimal units. Procedures of segmentation and classification are so deeply embedded in current linguistic practice that they are often perceived as intrinsic to linguistic analysis rather than as embodying specific assumptions about the grammatical role of part- whole structures. In the same way that phonemic and phrase structure analysis consists largely of parsing a string into component phonemes or phrases, morphological analysis is taken to involve breaking words down into minimal, individually meaningful, units, or ‘morphemes’.
The practice of parsing words into morphemes is so well established that introductory textbooks often present the identification of morphemes as the core task— or even as the sole task—of morphological analysis. To take just two examples, O’Grady and Dobrovolsky (1996:112) state:
The most important component of word structure is the morpheme, the smallest unit of language which carries information about meaning or word structure.
In a section titled ‘Morphemes: The Minimal Units of Meaning, Fromkin et al. (2010:81) likewise assert that “the linguistic term for the most elemental unit of linguistic form is morpheme”. What is noteworthy about these passages is that they are presented by their authors as uncontroversial statements of fact. Morphological analysis is construed solely in terms of processes of segmentation and classification that parse words into their ultimate constituents.
The idea that morphemes provide the basis for morphological analysis is also assumed in the problem sets that follow in the tradition of workbooks and textbooks such as Nida (1949) or Gleason (1955). These typically require the reader to identify and classify the morphemes and allomorphs that occur within a set of complex forms. Exercises of this type reinforce the impression that morphological analysis consists essentially of segmenting words into morphemes, and that other tasks, such as assigning words to paradigms or other larger classes, are merely convenient ways of presenting data or organizing it for pedagogical purposes. The primacy of morphemes is further reinforced by other practices, such as the convention of assigning ‘morpheme glosses’ to complex forms in linguistic examples. What is again striking about these descriptive practices and conventions is how they tacitly treat the morpheme as an indispensable component of morphological analysis in general, rather than as an element of a particular model of analysis.
Word and Paradigm Morphology. First edition. James P. Blevins © James P Blevins 2016. First published 2016 by Oxford University Press
The appeal of this kind of constructive approach owes much to the initial simplicity of a purely syntagmatic approach, particularly one in which individual features determine the forms that express them, and forms signal the features that they express. This conception is particularly well suited to introductory presentations of morphology and to the formulation of problem sets that isolate a single alternation or pattern within a language.1 But these practical tasks also reflect a more general view of linguistic structure, encapsulated in the hypothesis that the application of principles of segmentation and classification can reveal a biunique correspondence between ‘minimal units of meaning’ and ‘minimal units of form’. Classical WP accounts also adopt a version of feature-form biuniqueness but assume, based in part on the tangled patterns of exponence exhibited by classical languages, that words and paradigm cells are the smallest units that can be brought into this correspondence.
The cardinal achievement of the Bloomfieldian model was to reorient morphological analysis away from a classical word-based perspective towards a more atomistic approach, in which meanings were associated with sub-word formatives. From a contemporary standpoint, the most striking aspect of this reorientation is its persistence, long after morphemic analysis was repudiated or simply abandoned by initial proponents and redefined beyond all recognition in later accounts. This apparent conundrum can be resolved by the observation that the morphemic tradition is united by the use of the term ‘morpheme’ rather than any substantive claim about morphological analysis.