Is there any consensus among historical linguists on what statements such as the following mean?
"English is a Germanic language [and not a Romance/Hellenic/non-IE language]"
"Hungarian is a Uralic language [and not an IE/Turkic/etc. language]"
"Tok Pisin is an English Creole language [not a Papuan/IE/etc. language]"
The only criteria I know of for making these kinds of statements are the Swadesh list and similar lists of "core" vocabulary/morphemes. There seem to be sets of words and other morphemes that are more resilient than other words/morphemes (though never perfectly resilient, that I know of) when they are passed on from person to person.
However, Swadesh-type sets don't constitute anything like the totality of the language spoken by any individual (I don't use only the Swadesh list of English words, people in Hungary don't use only the Swadesh list of Hungarian, etc.). What's the rationale, then, for classifying these totalities in an exclusive way, such that English can only be considered Germanic and not also Romance (despite the vocabulary English has acquired through French, etc.), Hungarian can only be considered Uralic and not also Turkic (despite the vocabulary Hungarian has acquired through Turkish), and so on?
Or, are exclusive classifications primarily a relic of earlier periods in historical linguistics, and not felt to have any validity (except as a practical tool) by most researchers today?
I'm not assuming that the majority of historical linguists think in these terms, but some of the most common terminology still in use (inherited word, loanword, family, cognate) seems to imply exclusivity, and I wonder if this terminology doesn't sometimes affect the thinking of the people who use it.