The Gulf of Guinea creoles (GGCs) exhibit a number of cases of initial vowel agglutination to etymologically consonant-initial words in the lexifier, Portuguese. This property is especially common in Lung’ie (Principense). Comparing agglutinated items in the four GGCs not only sheds light on their diachronic development, it also shows the linguistic compromise made between the different strata that gave rise to this feature. It will be argued that prosthetic vowels are phonetically calqued on the Portuguese definite article system, which bleached and became generalized as something else than a gender/number system, whereas the African contribution consists of creating vowel-initial items that are guided by rules of vowel harmony.
In this article I attempt to approach the question of what effect the incorporation of a large number of KiKoongo vocabulary items by the creole languages of Surinam had vis-a-vis notions of simplification or complexification. So, we have the situation of a language lacking an extensive noun-class system with number marking having to absorb many words possessing such a system. An examination of the KiKoongo borrowings belonging to three classes is performed: 3/4, 5/6, 7/8. Each class appears to display a different constellation of singular and plural prefixes. 3/4 appears to represent only singular forms, either with an explicit suffix, or with no suffix. 5/6, where the KiKoongo dialect which seems to be best represented in Surinam has no explicit prefix in the singular, has a large preponderance of plural prefix forms. 7/8 has mixed results, possibly indicating KiKoongo dialect mixture. Most items occur without a prefix – presumably representing singulars, while a smaller number have explicit singular or plural prefixes. A surprisingly large number of forms display the wrong prefix. The common prefixes /ma-/ (pl.cl. 6) and /mu-/ (sg.cl. 3) seem to be involved frequently. I hypothesize that KiKoongo speakers initially used the relevant prefix in words borrowed from KiKoongo. This is however redundant given that the definite article is marked for number. Fongbe speakers would not know the correct number agreement, but would be able to recognize KiKoongo morphemes because of word length. In class 5/6 the preferred option was for a prefixed form (plural). The next stage is the loss of knowledge that KiKoongo loans contain a prefix at all. The loss of number distinctions, then gender distinctions, then parsability into discrete morphemes takes place in the lack of any meaningful function for the prefixes. The number distinctions presumably initially encoded by KiKoongo speakers would be redundant in any case. Gender distinctions play no role in the larger Surinam Creole lexica. And meaningless prefixes cease to have any role. The complexity that is lost in this corner of the lexicon never played a significant role in the creoles.
Old Tibetan shows extraordinary complexity in its syllable structure as well as highly complex or rather opaque verb morphology. The syllable structure (CCC)CV(CC) has broken down completely in the modern Central Tibetan dialects to CV(C), while the opaque alternations of prefixes, consonants and vowels in verb stem formation were levelled out and replaced by regular systems of periphrastic construction in the western and central varieties. Both developments can be described as processes of simplification that were triggered in a linguistic contact situation, where Old Tibetan served as a lingua franca for various non-Tibetan peoples.
In a number of French-related creoles a distinction is made between long and short forms of verbs. We argue that the alternation is a reflex of French inflectional morphology that has survived the creolization process, showing that the result is a long-short opposition of verb forms, similar to the formal variation in the learner varieties, and therefore ultimately due to learner strategies in the acquisition of French as a second language. We further discuss the potential role of substrate and argue that the alternation started out as a phonological/prosodic phenomenon (as it basically still is in Haitian Creole). We conclude that the alternation (or verb allomorphy) can be seen as a morphological reflex of the Spell-Out domain at the vP-level.
This paper considers the historical context in which Berbice Dutch was formed before turning to the significance of the presence in that language of function words derived from the Eastern Ịjọ substrate. The view that transfer of Eastern Ịjọ grammatical properties took place in the formation of Berbice Dutch, is subjected to detailed scrutiny for tense-mood-aspect marking and negation. Despite similarities, important areas of divergence or discontinuity between Berbice Dutch and its substrate are identified – areas which point to reanalysis of substrate-derived functional material in the genesis of Berbice Dutch. This runs counter to the view that Eastern Ịjọ speakers played a central role in the formation of Berbice Dutch, and suggests that ‘the invisible hand’ in its genesis must have been another group, possibly the mixed progeny of the plantation population, which included Dutch, Eastern Ịjọ and Arawak speakers.
In Honiara, capital city of the Solomon Islands, speakers of the local variety of Pijin are making extensive usage of the transitive suffix -em (and its variants -im and -um) to transform prepositions into prepositional verbs: daon /down/ becomes daonem /to lower/; ap /up/ becomes apum /to raise/; insaet /inside/ becomes insaetim /to insert, to take inside/; aot /out/ becomes aotim /to remove/, etc. Looking at data gathered in Honiara since 1981, this paper will hypothesize that the formation of prepositional verbs in Solomon Island Pijin (SIP) is best understood as an instance of paradigmatic regularization that is also present in other parts of the morphosyntax. The paper will argue that: (1) simplification and complexification are not the only types of linguistic changes affecting the life of PCs (Pidgin and Creole language); and (2) that regularization is internally-induced, and may not be linked to any substrate or superstrate effect.
The process of article incorporation early in the genesis of Mauritian Creole (MC) resulted in the occurrence of a bare nouns in argument positions with ambiguous interpretations between (in)definite, singular, plural and generic. A new determiner system gradually emerged, but MC continues to admit bare nouns in argument positions. It is argued in this paper that the process of article incorporation triggered a shift in noun denotation from predicative in French to argumental in MC. Like English bare plurals, MC nouns are of an argumental kind-denoting terms that do not require a determiner in argument positions. The MC singular indefinite article enn and the plural marker bann serve to cderive instances of kinds. The differential behaviour of MC count and mass nouns is attributed to the number feature which must be checked for count nouns, and provides evidence for a phonologically null definite determiner which is licensed in subject position by the specificity marker la.
Afrikaans seems to have lost the Dutch independent demonstrative dat ‘that’ as well as the pronoun het ‘it’ and the attributive element deze ‘this, these’, while independent dit ‘this’ seems to have taken over the functions of dat and het and while attributive die ‘that, those’ has acquired a proximate reading (Afr. dié week ‘this week’). In the present paper it is argued, however, that the weak pronoun het was bound to disappear anyway, that dat only underwent phonological change (so that dat and dit couldn’t be distinguished any longer) and that the changes in the system of attributive demonstratives are due to developments in Cape Dutch Pidgin.
This paper examines several instances of phonological and structural complexification in Mindanao Chabacano, a predominantly Spanish-lexifier creole of southwestern Mindanao, which have arisen as the result of the interaction of elements of Philippine (especially Tagalog and the Bisayan languages), Spanish and English origin in the language, which have given rise to contact-induced change in the language’s structure.
This paper presents data from negation in Sri Lankan Malay (SLM), a language whose grammar has converged typologically on the grammar of Sri Lankan Muslim Tamil, and to some extent of Sinhala. SLM negation exhibits greater inflectional complexity than its lexifier, by encoding finiteness and tense features. SLM has also developed independently of the other Sri Lankan languages. It has developed a strict morphological contrast between all non-finite negation contexts on the one hand (infinitives, participles, and imperatives), and finite negation contexts (tensed verbs). This effectively circumvents the Dravidian constraint blocking co-occurring tense and negation morphology, in order to emphasize a contrast which is salient in the discourse structure that is common to the languages in the sprachbund.
The aim of this paper is to present a view of contact language formation in which language creation in multilingual ecologies follows the same principles as language maintenance in monolingual ecologies, i.e. selection and replication of features available to speakers in a given environment. In order to do so, I introduce the foundations underlying an evolutionary framework to contact language formation and the views they offer for our understanding of language contact and change. The view of grammar as an evolving system, I believe, can be best appreciated in a functional-typological theory of language. For this reason, I first introduce the basic functionalist, usage-based linguistic theories required for an evolutionary framework. I then synthesize a view on language contact and change in evolutionary terms based on Croft (2000, 2006a) and Mufwene (2001). Finally, I apply the views presented here to a case of contact language formation, namely the evolution of case markers in a variety of Sri Lanka Malay. These are particularly interesting as, from a classic or orthodox view, they might be seen as ‘complex’, ‘marked’ or at least ‘unexpected’ instances of contact-induced change. The evolutionary framework however can explain these as natural acts of linguistic replication in multilingual settings, thus avoiding exceptionalist explanations. Instead, an evolutionary framework offers an integration of socio-historical and functional-typological observation, something that our current approaches to language change still largely lack (Croft 2006b). Among the advantages of the framework applied here, as discussed in the concluding section, is the suggestion that overall structural complexity, however defined, does not change as a result of contact language formation: a new grammar is simply the result of a recombination of grammatical features of the input languages.
This paper explores the idea that in contact situations, there exists a process whereby so-called morpho-syntactic simplification is correlated to semantic complexification. In examining this process in the verbal and nominal domains, we show that a given morpheme may actually carry in Cape Verdean Creole (CVC) and Guinea-Bissau Creole (GBC) a cluster of semantic feature values where the European language only has one. In order to measure degrees of simplification versus complexification, this paper uses Kusters’ (2003) complexity evaluation metrics involving inflectional morphology specifically. This paper also shows to what extent two sister creoles such as CVC and GBC, assumed to have emerged from the same source languages, display similarities and distinct differences in their morphological properties.
This paper demonstrates that the notion of simplicity as often used in creole studies is completely irrelevant to the understanding of the structure, as well as the genesis, of creole languages. This is because creole languages are linguistic hybrids in the biological sense. They emerge from the recombination of linguistic features from different languages. Given this perspective, it appears that what could be of some relevance to the study of language change is rather the notion of complexity. Within the framework of Competition and Selection as proposed in Mufwene (2001ff.), and adopted in this paper, creole languages develop opaque syntactic and semantic features. These could not have arisen solely in the context of their source languages. Accordingly, the common claim that creoles are simplified versions of their sources is a fallacy, just as it would be to claim in biology that hybrids are genetically simplified children of their parents.
This paper addresses the issue of complexity in language creation and the time it takes for ‘complex’ structures to emerge in the history of a language. The presence of morphological material is often equated to a certain degree of complexity or is taken to signify a certain time-depth in the history of a language (e.g. Dahl 2004; McWhorter 2005). Though this assumption may be seen as trivial in the absence of a theoretically-based definition of complexity (Muysken 1988), or even misleading (Aboh and Ansaldo 2007; Farquharson 2007), we here put it to a test by looking at morphology in a relatively ‘young’ language, namely Sri Lanka Malay (SLM). SLM is a mixed language which shows considerably more morphological material and other signs of old age than ‘prototypical’ creoles. We explain this by arguing (a) that structural output in language genesis is closely motivated by the typology of the input languages and (b) that our understanding of rate of change needs to be revised to take into account ecological matters.
The Gulf of Guinea creoles (GGCs) exhibit a number of cases of initial vowel agglutination to etymologically consonant-initial words in the lexifier, Portuguese. This property is especially common in Lung’ie (Principense). Comparing agglutinated items in the four GGCs not only sheds light on their diachronic development, it also shows the linguistic compromise made between the different strata that gave rise to this feature. It will be argued that prosthetic vowels are phonetically calqued on the Portuguese definite article system, which bleached and became generalized as something else than a gender/number system, whereas the African contribution consists of creating vowel-initial items that are guided by rules of vowel harmony.
In this article I attempt to approach the question of what effect the incorporation of a large number of KiKoongo vocabulary items by the creole languages of Surinam had vis-a-vis notions of simplification or complexification. So, we have the situation of a language lacking an extensive noun-class system with number marking having to absorb many words possessing such a system. An examination of the KiKoongo borrowings belonging to three classes is performed: 3/4, 5/6, 7/8. Each class appears to display a different constellation of singular and plural prefixes. 3/4 appears to represent only singular forms, either with an explicit suffix, or with no suffix. 5/6, where the KiKoongo dialect which seems to be best represented in Surinam has no explicit prefix in the singular, has a large preponderance of plural prefix forms. 7/8 has mixed results, possibly indicating KiKoongo dialect mixture. Most items occur without a prefix – presumably representing singulars, while a smaller number have explicit singular or plural prefixes. A surprisingly large number of forms display the wrong prefix. The common prefixes /ma-/ (pl.cl. 6) and /mu-/ (sg.cl. 3) seem to be involved frequently. I hypothesize that KiKoongo speakers initially used the relevant prefix in words borrowed from KiKoongo. This is however redundant given that the definite article is marked for number. Fongbe speakers would not know the correct number agreement, but would be able to recognize KiKoongo morphemes because of word length. In class 5/6 the preferred option was for a prefixed form (plural). The next stage is the loss of knowledge that KiKoongo loans contain a prefix at all. The loss of number distinctions, then gender distinctions, then parsability into discrete morphemes takes place in the lack of any meaningful function for the prefixes. The number distinctions presumably initially encoded by KiKoongo speakers would be redundant in any case. Gender distinctions play no role in the larger Surinam Creole lexica. And meaningless prefixes cease to have any role. The complexity that is lost in this corner of the lexicon never played a significant role in the creoles.
Old Tibetan shows extraordinary complexity in its syllable structure as well as highly complex or rather opaque verb morphology. The syllable structure (CCC)CV(CC) has broken down completely in the modern Central Tibetan dialects to CV(C), while the opaque alternations of prefixes, consonants and vowels in verb stem formation were levelled out and replaced by regular systems of periphrastic construction in the western and central varieties. Both developments can be described as processes of simplification that were triggered in a linguistic contact situation, where Old Tibetan served as a lingua franca for various non-Tibetan peoples.
In a number of French-related creoles a distinction is made between long and short forms of verbs. We argue that the alternation is a reflex of French inflectional morphology that has survived the creolization process, showing that the result is a long-short opposition of verb forms, similar to the formal variation in the learner varieties, and therefore ultimately due to learner strategies in the acquisition of French as a second language. We further discuss the potential role of substrate and argue that the alternation started out as a phonological/prosodic phenomenon (as it basically still is in Haitian Creole). We conclude that the alternation (or verb allomorphy) can be seen as a morphological reflex of the Spell-Out domain at the vP-level.
This paper considers the historical context in which Berbice Dutch was formed before turning to the significance of the presence in that language of function words derived from the Eastern Ịjọ substrate. The view that transfer of Eastern Ịjọ grammatical properties took place in the formation of Berbice Dutch, is subjected to detailed scrutiny for tense-mood-aspect marking and negation. Despite similarities, important areas of divergence or discontinuity between Berbice Dutch and its substrate are identified – areas which point to reanalysis of substrate-derived functional material in the genesis of Berbice Dutch. This runs counter to the view that Eastern Ịjọ speakers played a central role in the formation of Berbice Dutch, and suggests that ‘the invisible hand’ in its genesis must have been another group, possibly the mixed progeny of the plantation population, which included Dutch, Eastern Ịjọ and Arawak speakers.
In Honiara, capital city of the Solomon Islands, speakers of the local variety of Pijin are making extensive usage of the transitive suffix -em (and its variants -im and -um) to transform prepositions into prepositional verbs: daon /down/ becomes daonem /to lower/; ap /up/ becomes apum /to raise/; insaet /inside/ becomes insaetim /to insert, to take inside/; aot /out/ becomes aotim /to remove/, etc. Looking at data gathered in Honiara since 1981, this paper will hypothesize that the formation of prepositional verbs in Solomon Island Pijin (SIP) is best understood as an instance of paradigmatic regularization that is also present in other parts of the morphosyntax. The paper will argue that: (1) simplification and complexification are not the only types of linguistic changes affecting the life of PCs (Pidgin and Creole language); and (2) that regularization is internally-induced, and may not be linked to any substrate or superstrate effect.
The process of article incorporation early in the genesis of Mauritian Creole (MC) resulted in the occurrence of a bare nouns in argument positions with ambiguous interpretations between (in)definite, singular, plural and generic. A new determiner system gradually emerged, but MC continues to admit bare nouns in argument positions. It is argued in this paper that the process of article incorporation triggered a shift in noun denotation from predicative in French to argumental in MC. Like English bare plurals, MC nouns are of an argumental kind-denoting terms that do not require a determiner in argument positions. The MC singular indefinite article enn and the plural marker bann serve to cderive instances of kinds. The differential behaviour of MC count and mass nouns is attributed to the number feature which must be checked for count nouns, and provides evidence for a phonologically null definite determiner which is licensed in subject position by the specificity marker la.
Afrikaans seems to have lost the Dutch independent demonstrative dat ‘that’ as well as the pronoun het ‘it’ and the attributive element deze ‘this, these’, while independent dit ‘this’ seems to have taken over the functions of dat and het and while attributive die ‘that, those’ has acquired a proximate reading (Afr. dié week ‘this week’). In the present paper it is argued, however, that the weak pronoun het was bound to disappear anyway, that dat only underwent phonological change (so that dat and dit couldn’t be distinguished any longer) and that the changes in the system of attributive demonstratives are due to developments in Cape Dutch Pidgin.
This paper examines several instances of phonological and structural complexification in Mindanao Chabacano, a predominantly Spanish-lexifier creole of southwestern Mindanao, which have arisen as the result of the interaction of elements of Philippine (especially Tagalog and the Bisayan languages), Spanish and English origin in the language, which have given rise to contact-induced change in the language’s structure.
This paper presents data from negation in Sri Lankan Malay (SLM), a language whose grammar has converged typologically on the grammar of Sri Lankan Muslim Tamil, and to some extent of Sinhala. SLM negation exhibits greater inflectional complexity than its lexifier, by encoding finiteness and tense features. SLM has also developed independently of the other Sri Lankan languages. It has developed a strict morphological contrast between all non-finite negation contexts on the one hand (infinitives, participles, and imperatives), and finite negation contexts (tensed verbs). This effectively circumvents the Dravidian constraint blocking co-occurring tense and negation morphology, in order to emphasize a contrast which is salient in the discourse structure that is common to the languages in the sprachbund.
The aim of this paper is to present a view of contact language formation in which language creation in multilingual ecologies follows the same principles as language maintenance in monolingual ecologies, i.e. selection and replication of features available to speakers in a given environment. In order to do so, I introduce the foundations underlying an evolutionary framework to contact language formation and the views they offer for our understanding of language contact and change. The view of grammar as an evolving system, I believe, can be best appreciated in a functional-typological theory of language. For this reason, I first introduce the basic functionalist, usage-based linguistic theories required for an evolutionary framework. I then synthesize a view on language contact and change in evolutionary terms based on Croft (2000, 2006a) and Mufwene (2001). Finally, I apply the views presented here to a case of contact language formation, namely the evolution of case markers in a variety of Sri Lanka Malay. These are particularly interesting as, from a classic or orthodox view, they might be seen as ‘complex’, ‘marked’ or at least ‘unexpected’ instances of contact-induced change. The evolutionary framework however can explain these as natural acts of linguistic replication in multilingual settings, thus avoiding exceptionalist explanations. Instead, an evolutionary framework offers an integration of socio-historical and functional-typological observation, something that our current approaches to language change still largely lack (Croft 2006b). Among the advantages of the framework applied here, as discussed in the concluding section, is the suggestion that overall structural complexity, however defined, does not change as a result of contact language formation: a new grammar is simply the result of a recombination of grammatical features of the input languages.
This paper explores the idea that in contact situations, there exists a process whereby so-called morpho-syntactic simplification is correlated to semantic complexification. In examining this process in the verbal and nominal domains, we show that a given morpheme may actually carry in Cape Verdean Creole (CVC) and Guinea-Bissau Creole (GBC) a cluster of semantic feature values where the European language only has one. In order to measure degrees of simplification versus complexification, this paper uses Kusters’ (2003) complexity evaluation metrics involving inflectional morphology specifically. This paper also shows to what extent two sister creoles such as CVC and GBC, assumed to have emerged from the same source languages, display similarities and distinct differences in their morphological properties.
This paper demonstrates that the notion of simplicity as often used in creole studies is completely irrelevant to the understanding of the structure, as well as the genesis, of creole languages. This is because creole languages are linguistic hybrids in the biological sense. They emerge from the recombination of linguistic features from different languages. Given this perspective, it appears that what could be of some relevance to the study of language change is rather the notion of complexity. Within the framework of Competition and Selection as proposed in Mufwene (2001ff.), and adopted in this paper, creole languages develop opaque syntactic and semantic features. These could not have arisen solely in the context of their source languages. Accordingly, the common claim that creoles are simplified versions of their sources is a fallacy, just as it would be to claim in biology that hybrids are genetically simplified children of their parents.
This paper addresses the issue of complexity in language creation and the time it takes for ‘complex’ structures to emerge in the history of a language. The presence of morphological material is often equated to a certain degree of complexity or is taken to signify a certain time-depth in the history of a language (e.g. Dahl 2004; McWhorter 2005). Though this assumption may be seen as trivial in the absence of a theoretically-based definition of complexity (Muysken 1988), or even misleading (Aboh and Ansaldo 2007; Farquharson 2007), we here put it to a test by looking at morphology in a relatively ‘young’ language, namely Sri Lanka Malay (SLM). SLM is a mixed language which shows considerably more morphological material and other signs of old age than ‘prototypical’ creoles. We explain this by arguing (a) that structural output in language genesis is closely motivated by the typology of the input languages and (b) that our understanding of rate of change needs to be revised to take into account ecological matters.