Computer Simulations of Developmental Change: The Contributions of Working Memory Capacity and Long‐Term Knowledge

Abstract

Increasing working memory (WM) capacity is often cited as a major influence on children's development and yet WM capacity is difficult to examine independently of long-term knowledge. A computational model of children's nonword repetition (NWR) performance is presented that independently manipulates long-term knowledge and WM capacity to determine the relative contributions of each in explaining the developmental data. The simulations show that (a) both mechanisms independently cause the same overall developmental changes in NWR performance, (b) increase in long-term knowledge provides the better fit to the child data, and (c) varying both long-term knowledge and WM capacity adds no significant gains over varying long-term knowledge alone. Given that increases in long-term knowledge must occur during development, the results indicate that increases in WM capacity may not be required to explain developmental differences. An increase in WM capacity should only be cited as a mechanism of developmental change when there are clear empirical reasons for doing so.

Does phonological short-term memory causally determine vocabulary learning? Toward a computational resolution of the debate

“…As we have emphasized in the present work, we take these to be key characteristics of nonword repetition, ones that go to the heart of what makes the task a working memory task. Accordingly, the Jones et al (2008) model characterizes working memory in a very different way than in the present model. In addition, that model does not appear to actually produce output at all, and the behavioral performance it simulates appears more akin to nonword recognition than nonword repetition/production.…”

“…In addition, that model does not appear to actually produce output at all, and the behavioral performance it simulates appears more akin to nonword recognition than nonword repetition/production. For these reasons, we regard the present work as considerably different from that of Jones et al (2008), and as making a substantial additional contribution.…”

“…One recent computational model has sought very directly to examine the relative roles of working memory and vocabulary knowledge in nonword repetition and nonword learning, i.e., phonological vocabulary learning (Jones et al, 2008). This interesting work employs the computational framework of EPAM, a discrimination net (Simon & Feigenbaum, 1964).…”

“…Although there have been numerous computational accounts of immediate serial list recall (e.g., Botvinick & Plaut, 2006;Brown et al, 2000;Burgess & Hitch, 1992, 1999Gupta, 1996Gupta, , 2009Page & Norris, 1998) as well as of word and nonword repetition (e.g., Gupta, 1996Gupta, , 2009Gupta & MacWhinney, 1997;Hartley & Houghton, 1996;Vousden, Brown, & Harley, 2000), they have either not addressed phonological vocabulary learning, or have not addressed it in a manner that resolves the critical question (but see Jones, Gobet, & Pine, 2008 for a recent approach). The issues can, however, usefully be framed in computational terms, and in particular, in terms of three questions: First, what is required computationally for repetition of a novel word form?…”

Evidence for a Non-Lexical Influence on Children’s Auditory Repetition of Familiar Words

“…Several computational models of nonword repetition in the literature have addressed these effects (e.g. Gupta and Tisdale 2009;Jones et al 2008), and our account should not be seen as a competitor with these models. The nonlexical route described here is designed solely to provide an empirically based estimate of the overall strength of an individual's nonlexical route (the nl parameter) in order that the contribution of the nonlexical route to children's real word repetition can be assessed.…”