Book Description

The present dissertation examines first-person singular (1sg) subject pronoun expressionthe alternation between yo I and the unexpressed variant based on data (N=6,450) drawn from 13th- 20th century Spanish literary texts. Our results reveal no increase in expression rates over time and that, in general, the linguistic conditioning of variable 1sg subject expression has largely remained constant over time.First, effects of subject continuity are found, with pronominal expression disfavored in contexts of continuity (more accessible referents) and favored in discontinuous contexts (less accessible referents). Three measures of subject continuityswitch reference, human switch reference, and distanceare shown to be applicable. However, subtle signs of change arise by comparing their applicability over time. In particular, the measure of distance from the previous mention of the same referent in the preceding discourse shows a graded effect: as distance increases, assessed in the number of intervening clauses, so progressively does use of pronoun yo. Interestingly, evidence suggests that the graded effect of distance may be developing over time into a more local effect of switch reference with respect to the immediately preceding clause, with pronominal expression especially favored when the subject of the intervening clause has a specific, human referent. Further, the results of this study support the view that an approach merely focused on verb form ambiguity does not provide conclusive answers regarding tense effects on subject expression. Although I found a general favoring effect of morphologically ambiguous verb tenses (with syncretism between first and third person), this is only marginal in contexts of subject continuity. Thus, a synergetic dynamic operates between ambiguity in verb morphology and switch reference: ambiguous verb forms significantly increase subject expression with discontinuous referents. In addition, a mechanical priming effect is shown to be present in all the periods. Thus, the presence of a 1sg pronominal subject enhances the likelihood that the subsequent 1sg coreferential subject is also pronominal. Conversely, when the previous coreferential subject is unexpressed, the likelihood of subject expression diminishes, favoring a subsequent unexpressed 1sg subject.Finally, the widely reported effect of cognition verbs, strongly favoring 1sg subject expression in present day varieties, is absent in these data. A usage-based construction grammar approach allows analysis that considers particular lexical items and forms, which reveals the role of prefabs in explaining idiosyncratic usage patterns of yo with specific verb forms and their impact on broad categories (semantic classes). This analysis provides evidence that the cognition verb forms creo and s have undergone changes that can explain present-day usage patterns. The use of yo I with creo think, originally believe, has increased steadily across time as its meaning has bleached. In parallel, increasing grammaticalization of s I know with a complement clause is observed. The diachronic patterns suggest that the [yo + cognition verb] construction emerged gradually and was consolidated fairly recently. In sum, where change is observed is in particular expressions, while general discourse-cognitive effects on subject expression have been operative since very early in the history of our language.