In the english literature in linguistics, the term Gradience seems to appear for the first time in 1961 with Dwight Bolinger in his book Generality, Gradience, and the All-or-None ([Bolinger1961]).
Gradience usually refers to categorial vagueness and reflects the fact that gradient phenomena are observed `at all levels of linguistic analysis, from phonology, morphology and lexis on the one hand, to syntax and semantics on the other' ([Aarts2004]) (see also in particular [Bresnan2003], [Keller2000], and [Sorace2004]).
In philosophical terms, the problem raised by the gradience is one of categorisation.
Bas Aarts, for instance--and philosophers such as Aristotle before him--discusses in [Aarts2004] the fact that in many cases items can't be classified properly because there isn't any class for which the item to be classified meets exactly all the membership requirements.
An all-or-none kind of policy would ignore such an item and fail to classify it.
Bas Aarts is even more specific and distinguishes between Subsective Gradience for `degree of resemblance to a categorial prototype' and Intersective Gradience for `degree of inter-categorial resemblance'.
Subsective gradience has to do with how close an item is from a prototypical item within the category it belongs to.
Intersective gradience has to do with the relative position of an item between two (or more) categories.
Typically, the problem of intersective gradience is one of identifying a threshold for assigning the item to one class or another.
The metaphor used by Bas Aarts in this regard is the one of a heap of sand:
Does one grain make a heap? Evidently not. Do two grains make a heap? No. Do one hundred grains make a heap? Yes. Where does one draw the line? (...) when do we call a collection of grains a heap? Is there a cut-off point such that n grains of sand form a heap, but n-1 do not?
Note that in such an approach one still keeps categories with well-defined boundaries. A 'gradience-is-everywhere' kind of approach, as Aarts calls it, would make a different assumption.
Please send comments or queries about this web site
to jpprost@ics.mq.edu.au
|
|