From Early Melanesian Pidgin to Solomon Islands Pijin, Bislama and Tok Pisin. A diachronic, comparative analysis of the emergence of individual varieties of Melanesian Pidgin English

Abstract

The present study analyses when and how the contact languages Tok Pisin (TP), Bislama (BIS) and Solomon Islands Pijin (SIP) developed out of Melanesian Pidgin English (MPE). Even though it has been previously assumed that the end of the labour trade represents the most decisive factor for the individual development of the three varieties (cf. Mühlhäusler 1978; Clark 1979; Jourdan 1985; Keesing 1988; Baker 1993; Jourdan & Keesing 1997; Sankoff 2021+), the current scientific knowledge still raises questions about the concrete time period in which the varieties localised, as well as about the linguistic features involved. To date assumptions have been predominantly grounded on extralinguistic events. A comparative study that is based on historical linguistic data and that combines qualitative and quantitative-statical methods has not been conducted so far but is necessary to understand the linguistic development of the varieties. To fill this academic void, the present study is based on historical linguistic data which was collected inter alia in the German Colonial Archives, the Pacific Research Archives, the Western Pacific Archives, and the Pacific Manuscript Bureau, and it uses qualitative and quantitative-statistical methods to learn when differences developed in the three varieties and whether, and if so, to what extent these differences can be traced back to the end of the labour trade. The collected linguistic data covers the time from the earliest detectable attestations until 1950. In order to identify morphosyntactic differences that existed between the three early varieties, a morpheme-by-morpheme analysis was conducted and four linguistic categories – demonstratives, relative clauses, modality markers and prepositions – were selected for further analysis. To visualise how the features developed across time, boxplots were generated in RStudio (RStudio 2019) and timelines were created using Excel. To investigate whether the year of attestation had an impact on the realisation of a feature in the varieties, conditional inference trees (ctrees) were used. It was also tested whether other factors, such as the text type and author, had an impact on the choice of form as well. While initially 41 significant time-based splits were identified with the help of ctrees, this number decreased when further predictor variables were considered so that when including all possible predictor variables, only nine dates remained significant. In addition, the results of the analysis showed that we cannot generalise per se when a divergence and/or stabilisation of the three MPE varieties took place. Even if the ctree analyses revealed that most of the statistically significant changes in the three varieties dated to the first half of the 20th century, which may be an indicator that the changes in the choice of feature forms might have been propelled by the end of the labour trade, the significant dates, in the first instance, show significant changes in the data. They do not specify per se whether the respective changes involve a reduction of variants and/or the introduction of new, diverging forms. Furthermore, the study showed that a differentiation between the notions of divergence and stabilisation is required. Though most attested forms only began to stabilise after the end of the labour trade with overseas plantations, the attestation of forms such as might (dubitative modality), with him (comitative), catch him (terminative), close up long (adessive), and where (relative clause particle), which were exclusively attested in SIP and BIS and not in TP, made clear that the labour recruitment years (and not the end of the labour trade) seem to have had a major impact on the emergence of diverging forms. In addition, the study revealed that, as in non-pidgins and non-creoles, the forms used to encode the individual morphosyntactic features developed and stabilised at different points in time. This shows that the divergence and stabilisation was a gradual rather than abrupt process. In sum, the qualitative and quantitative-statistical analysis of historical linguistic data revealed that the labour recruitment histories and the concomitant impact of overseas plantation pidgins, the influence of mission varieties, substrate reinforcement, and the degree of exposure to the lexifier can be regarded as the major reasons for diverging forms in the three varieties. In contrast, the end of the labour trade, the spread and use of the varieties to further domains in the home areas, and substrate influence represent major reasons identified for the stabilisation of each of the individual varieties and thus, for the reduction of – or at least a preference of – particular variants. The findings pose an important contribution and advance for the field of creolistics since they provide information about the processes involved in the origin and development of contact languages, and about grammaticalisation processes and language universals. The results suggest that it is likely that several of the mechanism which have been put forward in individual theories were involved and interacted during the emergence and stabilisation of contact varieties and that none of the existing theories alone is sufficient to explain the complex development of the MPE varieties.