Substrate Influences

Substrate languages are defined as the local language(s) or the native language(s) of the nondominant population. While Bickerton acknowledges that substrate languages influenced the development of Hawaiian Pidgin English (HPE), he does not believe that they played any role in the creation of Hawaiian Creole English (HCE). This is in keeping with his assumption that the bioprogram does not require outside social influences in the acquisition of language. However, over the last twenty years, other researchers have questioned both his methodologies and the accuracy of his conclusions.

While Corne (1984:192) agrees that Bickerton's bioprogram may explain protocreole language genesis, he states that Bickerton undervalues contributions from sub and superstrate languages in his general approach. Another way of explaining the structural similarities among creoles is by diffusion. McWhorter (1997:29) has proposed that English creoles developed from a pre-existing pidgin in Africa, and then spread to the Caribbean along with imported slaves. With regards to the incorporation of substrate features into a creole or pidgin, studies of second language acquisition indicate that imperfect language transfer (or interference) is a common factor (Mufwene 1990; Wekker 1996; Siegel 1998).

Regarding the acquisition of HCE, Goodman (1984:193) and Siegel (2000:230) state that Bickerton based his conclusions upon interviews with old Japanese, Filipino and Korean immigrants. According to Goodman (1984) when the first Japanese immigrated in 1888, there was already a fairly fixed form of pidgin established in Hawaii. Secondly, Koreans and Filipinos did not arrive until the early 1900s (Siegel 2000:230). According to Goodman, the Hawaiian demographic data indicates that for decades, colonial born slaves formed less than 10% of the population (ibid.). Goodman (1984:194) believes that "[T]he creole languages of the New World arose almost entirely as lingua francas among African born slaves rather than among those who were locally born".

In the 1990, Alleyne (1996:112-118) examined the development of several French based creoles (Haitian, Reunionnais, Mauritian and Seychellois). He found that the treatment of states and processes were neither due to a cognitive nor linguistic blueprint... it may be a function of the world view of the populations among whom these language developments took place. States are derived from processes, not perceived neutrally or abstractly, independent of the processes which give rise to them.

Roberts (1998:34) supports the idea that HCE emerged among the native population but contradicts several aspects of Bickerton's hypothesis. These were as follows (Roberts 1998:34-35): 1) the complex features of HCE took several generations to develop; 2) the first locally born generation of children did acquire the language of their parents, which was widely in use until the 1920s. These children were more than likely bilingual, since they would have learned English in school; and 3) it was the second generation of children who were the first monolingual speakers of HCE, and it was they who were credited with the later grammatical innovations of HCE.

Bickerton presents the following arguments against the influence of substrates (1977, 1981, 1984, 1996): 1) it does not seem possible that language could be made up of a mixture of features from other languages; 2) creole rules do not correspond to substrate rules, nor do they have a similar distribution; 3) there is no clear evidence as to how substrate features are actually incorporated into pidgins or creoles; and 4) no principles have been suggested to explain why some substrate features are chosen over others.

Lefebvre (1996:155) believes that creolization is due to relexification. The relexification hypothesis states that the substratum languages contributed to almost all of the syntactic and semantic properties of creole (ibid.). In contrast to Bickerton (1981), who claims that in situations where creole languages are created, children are deprived of an adult model for language, Lefebvre (1996:156) believes that it is the relexified language which is presented to children who learn creole as their native language. According to Posner (1984:205), it is uncertain whether creolization differs from other linguistic changes in nature, or merely in degree. Additionally, she states that sociohistorical conditions are more involved in these processes than Bickerton acknowledges (ibid.).

Wekker (1996:140) believes that creolization is a more gradual process of language formation that involves a period of bilingualism in which substrate features will be transmitted. Contrary to Bickerton (1981), Wekker (1996:141) states that there is no reason why creoles should develop more rapidly than other human languages, apart from the fact that in their formative states creoles are created under pressure, by people who know the grammar of at least one other natural language.

Bickerton (1996:40-41) believes that all of the critiques of his bioprogram fail to appreciate the logic of his arguments. He refutes any claim that the presence of other natural languages could have any effect upon the genesis of pidgin or creole languages, since the input sources would be too impoverished (ibid.). However, there is no evidence to prove Bickerton's claims that something either language or species specific might be at work in language creation or expansion.