It has been made vivid to me (thx JB) that the languist/linguist distinction that I have often invoked has been taken to reflect a disregard for GGers who study Gs in detail. This was not how what I intended. Rather I wanted the languist/linguist distinction to demarcate two different attitudes concerning the primary object of linguistic inquiry. Let me elaborate a touch.
Languists take the fundamental object of inquiry to be language. For a languist, languages are ontologically primary and their cognitive underpinnings secondary, the latter being (pale?) reflections of the surface properties of the former.
Just the opposite for linguists. In particular, linguists understand language to be (at best) an idealization (one generally defined by army/navy size). What’s real is the G of the individual speaker (though the idealization (i.e. literally false assumption) that every “language” has the same underlying “G” is a standard, useful one), and such inquiry aims at descriptive adequacy (i.e. describing that G that the native speaker of the relevant language has internalized).
Putting this another way: languists take surface facts and patterns to exhaust all there really is and so not attending to them amounts to not doing science. Why? Because the generalizations one finds within and across languages are at best ontologically secondary and at worst simple fabrications/distortions. What’s real are the surface facts, the generalizations being abstractions over these. For languists these generalizations are a bit like averaging, they loose information. This is a good place to think of Greenberg Universals.
For linguists, in contrast, underlying patterns reflecting grammatical structure are the name of the game. Where, for languists the primary ontological reality lives on the linguistic surface, for the linguist it is at most a hint of what is really going on below. Importantly, the linguist assumes that G organization is often obscured at the surface, or even hidden by surface forms. Coherence for the linguist is to be found at the abstract level, not at the surface. In this sense what’s ontologically primary is the hidden unobserved (and directly unobservable) underlying structure, the surface patterns being merely the visible forms, the footprints, of this underlying reality. Gs are real and so describing Gs is real. The lingusitic tokens one hears or sees are porducts of this underlying reality and so are ontologically secondary.
Note that saying that that surface forms/patterns are ontologically secondary does not imply that they are epistemologically so. Assuming that mental structures are more ontologically basic than visible forms is simply to say that the former cause the latter. All GGers believe that Gs are casually more fundamental than sentences for every GGer believes that sentences are the products of Gs, not the other way around. Languists take the opposite view. Visible patterns are real. Gs are causally secondary in being abstractions of these surface patterns.
Thus, for the linguist the ontological bed rock is the mentally represented G, not the surface forms that are their products. Here’s a good place to think of Chomsky Universals.
Given this version of the distinction, as I see it, virtually every GGer is a linguist for virtually all work in GG centers on the structure of Gs. Languists, unfortunately, are everywhere with us. Here’s a possible surprise (not): languists tend to be Empiricists and linguists are necessarily Rationalists (whether they know this or not and whether they care or not (which many, sadly, do not)). It’s a short step from thinking that what’s real is the visible surface data to the view that linguistic competence is largely a matter of correctly mirroring what one “sees.” Similarly, the step from studying the properties of underlying Gs to the conclusion that FL has non-trivial structure is also short.
Second to last point: languists can do useful work, but for this work to be useful to linguists it will almost always need to be reinterpreted in Gish terms. Again, think of how GGers understand the importance of Greenberg Universals; they are interesting to the degree that they can tell us something about the underlying structures of Chomskyan Universals. Put another way: they are targets of explanation, the goal being to explain these patterns in terms of fundamental properties of FL/UG.
Last point: just as being a languist does not preclude one from doing useful work, being a linguist does not guarantee that one’s research is interesting. Higher order commitments can be useful in guiding research especially if self-consciously attended to. However, as I’ve noted many times before: science is hard and there are no sure fire methods. Too bad eh?
 Here’s a project for the numerically inclined E-languist: calculate the navy+army size necessary to take a dialect to a language. A good follow-up study would calculate the breakeven point between an idiolect and a dialect. Have fun.
 As noted below, languists tend to be Empiricists and as Jerry Fodor has noted, Empiricists tend to confuse metaphysics and epistemology, or, at the very least, treat the epistemologically accessible as the metaphysically basic. This is precisely what GGers deny given their primary interest in descriptive adequacy, viz. the description of a native speakers actual G.