Introduction

Over the past century, researchers have examined data from diverse languages in order to discover and potentially explain similarities and differences in language systems. One primary goal has been to discover characteristics that occur across languages ('universals') versus those that pertain specifically to a given language (e.g. Bybee, 2006; Chomsky & Halle, 1968). The major area of focus has been adult language, but with some attention to children's language development, both typical and protracted. Because children's early language has reduced complexity compared with the adult language, acquisition researchers suggest that developmental data may highlight basic universal limitations (constraints) on language, with protracted development further affording opportunities to observe a greater variety of constraints operating over an extended period (Ingram, 1989; Jakobson, 1968). Some researchers suggest that such constraints are innate (e.g. Chomsky & Halle, 1968; Jakobson, 1968), but others argue that the constraints derive from limitations in cognition, brain and body (e.g. Bates & MacWhinney, 1982; Bernhardt & Stemberger, 1998; Hayes, 1999; Pierrehumbert, 2001). Regardless of the source of constraints, crosslinguistic developmental data may reveal both frequent patterns across languages and language-unique patterns, the former suggestive of universal tendencies (Locke, 1983) and the latter of language-specific influences (e.g. Ingram, 2012; Pye et al., 1987). Pye et al. (1987) observed that acquisition of a relatively complex phonological element, the affricate /tʃ/, was generally earlier in Quiché Mayan than in English. In Quiché Mayan, the affricate is very frequent (one of the five most frequent consonants in the language) and has a high functional load in the language, whereas in English it is infrequent (one of the five least frequent phonemes). Thus, universal tendencies can be tempered by frequency and functionality.

Such data emphasize the crucial importance of crosslinguistic studies for addressing issues of universality in language, yet very few studies have been conducted using equivalent methodology across languages.