Are the "operators" of descriptive grammar a) word allocation and b) subffixes; while the operators of explanatory grammar a) semantics and b) pragmatics?
Frank,
Describing a language requires the accumulation of facts. Explaining language requires abstracting from the data and proposing principles that bind the data.
For instance, in English, we might observe that:
1. articles come before nouns (the cat).
2. demonstratives come before nouns (that cat).
3. possessive pronouns come before nouns (my cat).
4. adjectives come before nouns (black cat).
However, we might want to group #1-3 under the category of "determiner" because they share the functions of only having one determiner before a given noun (not the my cat, etc.) and always coming before the adjective(s) that come before a noun (that black cat, not black that cat). The list above is a description. Abstracting a category and predicting order in terms of categories becomes part of the explanation. The rules derived from this process are an explanatory grammar.
Of course, the notion of "noun," "adjective," etc. is a step toward description and explanation, but examining the interactions among basic and abstracted categories increases understanding of how the language works. By discovering these principles, we can then compare languages in systematic ways and propose principles that might be universal to language.
Glenn
To my mind, description is often as far as we get, but without any explanatory insights there is no point in linguistic enquiry. Plus, if "DESCRIPTION_PERIOD" was enough (or at all possible), would there exist different (and often highly conflicting) schools or paradigms? I tend to believe "descriptive grammar" does not quite exist. Best regards.