As someone correctly pointed out earlier on, it's exactly like arguing against species, and the discussion always bogs down in the excluded middle. In the end, there is no distinction between a language and a dialect, there is just a continuum of dialects, some of which have achieved more than local utility, and are called languages.

But thare is certainly a difference when looked at over a wider scale- no (human) one could fail to notice the difference between English (and Norn Iron, and Scots, and Geordie, and Jamaican, and North Calina) and say Japanese.

As for pre- literate speech groups, yes, the criteria have to be different. I suppose the main test would be whether you could get sensible (and truthful) answers about the boundaries of usage. So many of the ancient national and tribal names seem to go to the roots "the people", "our mates" or "the ones who talk properly", and the words for outside groups as "not proper people", "wild men", "babblers" etc. Group cohesion is improved by deliberately excluding outsiders.

And of course many groups have more than one speech usage. Pidgin is (as I understand it) a trading language, used in places with many speech communities, none of which is dominant. Each speaker woukld be expected to use the exchange medium in public, and their own group's speech in domestic or community situations.

Can anyone comment about the current status of, say, Dutch? There, they have the full works, literature, legal usage, newspapers, hsitories etc.- but also whole important functions in e.g. higher education are conducted in English, so a scientifically educated Dutch person has to be fluent in English. I believe that this happens in most smaller European countries. Does it make Dutch "less of a language"?