UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Evaluative study of the transformational-generative approach to the syntactic description of Thomas Deloney's prose Bowers, Fred


The development of generative models for the description of natural languages has led, in the past ten years, to a Transformational-Generative Model of great explanatory power. As a methodological tool of historical description, however, generative models have rarely been tested. The present study evaluates the efficacy of the Transformational-Generative Model, as revised in 1965, in describing the syntax of Thomas Deloney's novels, all written shortly before 1600. Certain modifications to the revised model have been made in this study: the complex symbol rewrite of major categories has not been used, because it has many limitations in a grammar which aims at the description of a particular corpus; instead, the distinctive syntactic feature analysis of nouns and verbs is presented as a series of Base rules, not as a lexical strict-subcategorization. The model, as modified, presents many advantages over traditional and structural descriptions. It establishes a hierarchy of systematized rules which clearly indicates the level at which aspects of Deloney's English usage varies from those of a Modern English user; it permits distinctive presentation of optional stylistic devices and necessary syntactic rules; most importantly, it reveals the latency at one period of certain structures which, at a later period, flourished as surface utterances. The model distinguishes between the few changes in syntactic structure which occurred between 1600 and the present, and the many morphological changes. The major syntactic changes are seen to be in the development of structures such as the expanded continuous aspects of the verb, in the gradual personalization of impersonal verbal structures, and in the transformations governing indirect object structures, negatives, interrogatives and imperatives. Stylistically, the model reveals Deloney's highly nominal prose writing and his preference for post-nominal modification. In terms of latency, the model shows that structures like that of the continuous aspect in verbs are within the deep structure of Deloney's English, although they rarely emerge as surface utterances. The loss of impersonal constructions is revealed as a gradual convergence of deep and surface structure, not as analogic change. Indeed, the model presents a diachronic description which reveals the unchanging nature of the deep structure of English; historic shifts are seen to be variations in the choice of interpretive transformations at different periods. Such changes are not all in the same direction: the personalization of impersonal structures is a return, as it were, to the deep structure, while nominalization and multiple pre-nominal modification is a departure involving many transformations. The conclusion reached by the study is that the Transformational-Generative Model is highly revealing of the system of language underlying diachronic change. None-the-less, certain limitations of the Model remain to be eliminated by further experiment: the position of the lexicon, and its relation to syntactic description is by no means clear; the generation of adjectives and adverbs in the present model is not really satisfactory and the possibility of deeper deep structures underlying elements in these categories would seem to require further exploration within the theory; lastly, the position of pronouns and of other indexical elements in a language is not well described in a sentence-grammar which has to ignore discourse in order to preserve formality. It seems certain, however, that the Transformational-Generative Model's explanatory power far outweighs its present limitations, in the syntactic description of a historic text.

Item Media

Item Citations and Data


For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use.

Usage Statistics