In lieu of an abstract, here is a brief excerpt of the content:

  • Introduction
  • Heather Goad and Yvan Rose

1. Introduction

A significant portion of research conducted on the acquisition of phonology has traditionally focussed on segmental development, on the acquisition of contrasts, and on processes commonly affecting target sounds. While early generative research was primarily concerned with first language (L1) learners (Smith 1973; Ingram 1974), an important body of work was devoted to second language (L2) learners (Eckman 1977; Flege 1980) and to individuals exhibiting delayed or disordered phonological development as well (e.g., Grunwell 1975; Ingram 1976). In the 1970s and early 1980s, the segmental focus of work in acquisition was largely due to influence from theoretical research in generative phonology at the time, which accorded no formal status to the syllable or to hierachical representations more generally (Chomsky and Halle 1968). With the onset of nonlinear phonology (Goldsmith 1976; Kahn 1976) and subsequent development of highly articulated models of feature organization (Sagey 1986), the acquisition research on segment structure shifted to investigating the ways that feature organization could predict paths in the emergence of contrasts (Rice 1996; Brown and Matthews 1997) and appropriately constrain the rule components of developing grammars (Hancin-Bhatt 1994; Levelt 1994; Dinnsen 1998).

The new focus on highly structured representations in theoretical phonology led to the development of nonlinear approaches to prosodic representation (Selkirk 1980; Clements and Keyser 1983). With this, some attention was turned to the acquisition of syllable structure, stress, and prosodic constraints on word shape in [End Page 139] both first (Spencer 1986; Fikkert 1994) and second language acquisition (Broselow 1984; Archibald 1993). Researchers also began to systematically investigate the role that prosodic factors play in segmental acquisition (e.g., Macken 1992).

While nonlinear phonology standardly assumes that both segments and prosodic units are highly articulated, they constitute independent modules of the grammar. These modules can, of course, interact, but the architecture of the model predicts their behaviour to be largely autonomous. In the more recently developed Optimality Theory (OT; Prince and Smolensky 1993), the opposite view is taken as a starting point. In OT, where the principal source of explanation is a set of universal, rankable constraints, a single constraint can contain both segmental and prosodic information, and constraints which express segmental generalizations can be inter-ranked with those that express prosodic generalizations. In short, segmental-prosodic interaction is expected to be the norm.

The interactive approach that defines OT is supported by work on positional effects in end-state grammars, namely, that the realization of segmental contrasts is often dependent on position in the string: a greater range of contrasts is observed in those positions that are strong, while loss or lack of contrast is observed in contexts that are weak. Strong-weak asymmetries are conventionally tied to structural considerations, that is, to prosodic licensing (Itô 1986; Goldsmith 1990; Harris 1997; in OT, see, e.g., Zoll 1998). Recently, however, other sources of explanation have been sought; strong-weak asymmetries have been attributed to the relative strength of perceptual cues (e.g., Steriade 1999; Côté 2000) or to psycholinguistic considerations (e.g., Beckman 1997).

Regardless of the source of explanation for position-sensitive effects, an appeal to markedness enters into most accounts of the asymmetries observed. Specifically, faithfulness to segmental contrast is less likely to be respected in weak positions where markedness considerations take priority. Turning to acquisition, there is widespread recognition that in first language development, children's early productions are unmarked (Jakobson 1968; Demuth 1995; Gnanadesikan in press). In second language acquisition, where most research has focussed on the acquisition of aspects of grammar where the L1 is a subset of the L2, markedness has been shown to play an important role in shaping interlanguage (IL) outputs as well (Eckman 1977; Broselow, Chen, and Wang 1998).

These two factors, the focus on position-sensitive effects and on the role of markedness in shaping grammars, serve to contextualize the theme of the current issue: segmental-prosodic interaction in phonological acquisition. Our goal has been to include articles that reflect different points of view on the formalization of this interaction. While some contributions discuss the effects of prosodic factors on learners' development, others appeal to alternative sources of explanation for...

pdf

Share