Diachronica - Volume 26, Issue 1, 2009

This study explores the related concepts of parallel independent development and drift, highlighting in particular the challenge of quantifying isolation. I analyze the precisely synchronized spread of a sound change, the monophthongization of /aɪ/, across Pennsylvania German ‘speech islands’ in the American Midwest. A key finding is that the intensity and duration of interspeaker contact required to catalyze apparent parallel developments may have lower than expected thresholds. The significance of extensive yet low-intensity cross-migration patterns across these communities at particular points in their histories ultimately leads to an exploration of the minimal level of contact required for diffusion of a change and feeds into recent discussion on the social contexts for transmission and diffusion (e.g., Labov 2007).

The aim of this article is twofold: (a) To evaluate the learned source for the rise of personal infinitives in the history of Spanish, and (b) to address the issue of the origin of personal infinitives from the perspective of language acquisition. It is argued, by means of novel empirical evidence derived from the Corpus del Español, that the Latin Accusativus cum Infinitivo source is responsible for the personal infinitive in complement position in Old Castilian (in line with Pountain 1998). However, this learned source is not the source of the Modern Spanish personal infinitive in adjunct position. For the latter, an acquisitional scenario is put forward whereby changes in the primary linguistic data (cf. Lightfoot 1999) made adjunct personal infinitives possible.

In the history of English one finds a mixture of V2 and non-V2 word order in declaratives for several hundred years, with frequencies suggesting a relatively gradual development in the direction of non-V2. Within an extended version of a cue-based approach to acquisition and change, this paper argues that there are many possible V2 grammars, differing from each other with respect to clause types, information structure, and the behavior of specific lexical elements. This variation may be formulated in terms of micro-cues. Child language data from present-day mixed systems show that such grammars are acquired early. The apparent optionality of V2 in the history of English may thus be considered to represent several different V2 grammars in succession, and it is not necessary to refer to competition between two major parameter settings. Diachronic language development can thus be argued to occur in small steps, reflecting the loss of micro-cues, and giving the impression that change is gradual.