|Title:||Empirical Linguistics and Mixing the Levels|
|Description:||In his article ''Cognitive versus Generative Linguistics: How commitments influence results'', Lakoff writes the following:
''Empirical linguistics, in itself, makes no a priori commitment as to whether these [syntax, semantics, pragmatics] are separate subfields, but takes it as an empirical matter as to whether syntax is autonomous, or whether the generalizations governing the distribution of grammatical morphemes, categories, and constructions involve aspects of semantics, communicative function, or other aspects of cognition." (1991: 53)
Isn't one of the basic tenets of Bloomfieldian/post-Bloomfieldian empirical linguistics the prohibition of mixing levels, wherein we cannot use syntactic or morphological information in defining e.g. a phoneme - its description must strictly remain within the scope of phonological analysis.
What do others think, am I completely missing the point? Thank you very much for your help.
For context, there is a free copy of Lakoff, 1991:
|Linguistic Field(s):||Linguistic Theories|