The central hypothesis in the Pierce, Genesee, Delcenserie, and Morgan article is that phonological memory is key to explaining the relationship between early language experience (more specifically, less and more optimal ends of the language-experience continuum) and language learning outcomes. One piece of evidence offered is that phonological memory skills are enhanced by bilingualism, with bilingualism representing enriched experience. Here, I propose that data from bilingual children may contradict Pierce et al.’s central hypothesis, rather than support it.

The findings of the so-called bilingual advantages for phonological memory are relatively rare in the literature. A large number of studies documented bilingual disadvantages on phonological memory tasks (e.g., Buac, Gross, & Kaushanskaya, Reference Buac, Gross and Kaushanskaya2016; Masoura & Gathercole, Reference Masoura and Gathercole1999; Michas & Henry, Reference Michas and Henry1994; Thorn & Gathercole, Reference Thorn and Gathercole1999). When advantages are found, they appear to be limited to very specific subsets of bilinguals (i.e., ones who are in the process of gaining proficiency in their second language) and to a very select set of tasks (i.e., those that load more heavily on the central executive rather than the storage component of the working memory system; Kaushanskaya, Gross, & Buac, Reference Kaushanskaya, Gross and Buac2014; Kudo & Swanson, Reference Kudo and Swanson2014; Yoo & Kaushanskaya, Reference Yoo and Kaushanskaya2012).

The bilingual literature therefore highlights the need to carefully consider the terminology one uses when talking about phonological memory, because conflation of working memory and short-term memory measures (the way Pierce et al. do) is deeply problematic both for theory building (most models/theories of working memory distinguish between the two; Baddeley, Reference Baddeley2003; Engle, Tuholski, Laughlin, & Conway, Reference Engle, Tuholski, Laughlin and Conway1999; Kail & Hall, Reference Kail and Hall2001; Kane, Bleckley, Conway, & Engle, Reference Kane, Bleckley, Conway and Engle2001), and for interpretation of research findings (the easier storage-only tasks behave differently than the more difficult storage+processing tasks, at least for bilinguals).

However, even if we accept the premise that bilingual experience may enhance phonological memory function, the argument that early experience influences language outcomes through phonological memory still falls short. This is because whatever the influences of bilingualism are on phonological memory, bilingual experiences do not appear to enhance children's language outcomes. Bilingual children have repeatedly been found to score below their monolingual peers (e.g., Bialystok, Luk, Peets, & Yang, Reference Bialystok, Luk, Peets and Yang2010) or below monolingual norms (e.g., Fernandez, Pearson, Umbel, Oller, & Molinetmolina, Reference Fernandez, Pearson, Umbel, Oller and Molinetmolina1992; Pearson, Fernandez, & Oller, Reference Pearson, Fernandez and Oller1993; Umbel, Pearson, Fernandez, & Oller, Reference Umbel, Pearson, Fernandez and Oller1992) on single-language vocabulary measures in one or both of their languages. Of course, when vocabulary knowledge in both languages is taken into account, bilingual/monolingual differences in vocabulary can disappear (Pearson et al., Reference Pearson, Fernandez and Oller1993), but they do not always (Gross & Kaushanskaya, Reference Gross, Buac and Kaushanskaya2014). However, for the Pierce et al. argument to work, it has to be the case that enhancements in phonological memory are associated with enhancements in language, and this is simply not the case for bilinguals (unless one defines enhancements in language as metalinguistic in nature, thus conflating language and executive-function tasks).

Another set of findings in the bilingualism literature, the link between phonological memory performance and language ability (or proficiency), also goes against the hypothesis that phonological memory is at the root of the effects that language experience has upon the language system. Phonological memory tasks by their very nature are nearly always language-specific, and as Pierce et al. point out, bilinguals’ proficiency in the target language is consistently associated with performance on such tasks (e.g., Masoura & Gathercole, Reference Masoura and Gathercole1999; Michas & Henry, Reference Michas and Henry1994; Thorn & Gathercole, Reference Thorn and Gathercole1999). However, it cannot be the case that the relationship between phonological memory and language proficiency stems from the effects of phonological memory on language proficiency, at least in sequential bilinguals. Instead, it must be that language knowledge constrains phonological memory performance. This is because in bilinguals who are in the process of acquiring their second language, the phonological memory system is already in place (via native language acquisition). Thus, unless one were to propose formation of two different phonological memory systems each supporting acquisition in bilinguals’ two languages, the relationship between second language proficiency and performance on phonological memory tasks administered in the second language must stem from the influences of language knowledge upon phonological memory performance (and not vice versa).

Such a conceptualization of the relationship between language knowledge and phonological memory is consistent with theoretical views of working memory other than Baddeley's. For example, Cowan (Reference Cowan1988) suggested that working memory involves activation of information in long-term memory, and that selection of specific information within the activated space is accomplished via an attention-based process. However, even Baddeley's model presupposes participation of language knowledge (stored in long-term memory) and attention (through the central executive) in performance on phonological working memory tasks. Viewed in this way, data from bilingual children (as well as from children with less optimal language experiences and the corresponding language and phonological memory difficulties) may be interpreted to suggest that experience influences the formation of linguistic representations and attentional mechanisms, and it is the interaction between them that gives rise to performance on phonological memory tasks.