Mittwoch, 17.30 Uhr
According to this paper, the assumption that Functional Categories (FC)may be first lost, then (partially) reconstituted during creolization reflects an erroneous view of this process and of the status of FC.
Adhering to a strictly mentalist view of language, one must conclude that FC are never lost in the minds of the speakers, even though the latter may pass through a stage (the so-called pidgin) where no phonological forms are available for some or all of the FC they have in their innate and/or acquired competence, depending on whether they are children or adults.Should a creole ensue, however, the FC are soon realized - NOT reconstituted, as they never disappeared.Realization may occur in two ways: (a) lexification or relexification; (b) zero morphemes being unmarked members of contrastive oppositions.
Furthermore, creolization should involve no loss in FC between the lexifier or substrate language and the creole. FC that seem to be discarded in the process can be shown to be either subcategories of major FC (e.g., Gender, a subcategory of Number, or Mood, a subcategory of Tense), or FC that probably have no autonomous reality to begin with (e.g., AGR or Case). Major or 'real' FC such as Tense (T), D, Foc, Neg, etc. are never lost.
This may have two implications: (i) only [+interpretable] FC have syntactic reality, whereas [-interpretatble] FC do not exist apart from their morphological realization; (ii) there is no variation among languages as to the number and the type of FC present in the grammar. Human language contains just those FC that are required for its functioning, and nothing can be lost or recreated. Whether our current inventory of FC is the right one is of course a different matter.
The foregoing reflections are empirically anchored in a study of Number in the Portuguese-based Creole of Guinea-Bissau, aka Kriyol, in comparison with the lexifier language Portuguese and the substrate languages Mandinka (Niger-Congo, Mande), Manjaku, and Wolof (both Niger-Congo, Atlantic).
Kriyol seems to have lost Gender and Number at a stage in its evolution. As for Gender, this is a fact. But Gender is lexical in Portuguese, hence not an FC. In Manjaku and Wolof, it is also a lexically assigned feature that is only manifested in association with the FC Number, through the so-called 'noun class' system. In Mandinka, there are no noun classes, but the cognitive category [+/-animate] - not an FC - interacts with Number marking: animate nouns are either [-Plural], morphologically unmarked, or [+Plural], morphologically marked unless cardinality is indicated separately by a numeral or a definite quantifier. Inanimate nouns, in contrast, are [+Singular] or [-Singular]. Neither is morphologically marked, but [+Singular] is generally denoted by a quantifier, while [-Singular] is interpreted as plural if cardinality is indicated by a numeral over 1 or a quantifier implying multiplicity; otherwise, it is interpreted as nonplural, i.e. as collective. There is thus a Number FC and a NumP projection in Mandinka syntax, but its content and interpretation depend on the cognitive classification of the noun.
What happened in Kriyol is that the system has changed (and is still changing) from a system calqued on Mandinka to a system more like Portuguese where, by and large, Number is only related to the 'real world' cardinality of the entity. At no point in this evolution has the FC Number ever been lost.
zum Programm der AG 6
zur alphabetischen Übersicht der Abstracts
zur zeitlichen Übersicht