I'd like to talk a bit about a question.

There's a phenomenon called "negative inversion" where initial (fronted) negative constituents trigger subject–auxiliary inversion in the main clause.

> He found not one of them useful. (no inversion)

> Not one of them did he find useful. (inversion)

In the second example, the negative constituent not one of them has been "fronted" (moved to the front of the sentence).

This triggers subject–auxiliary inversion, which I've highlighted in italics (did he).

Of course, because the original sentence has no auxiliary (just the lexical verb found), we have to add the dummy (meaningless) auxiliary do, turning found into did find.

That makes it possible for the subject he to swap places with an auxiliary: did he rather than he did.

This is sometimes called "negative inversion", but what's interesting about it, something that has come up several times recently on ELL, is that we find similar inversion with fronted constituents that aren't technically negative.

Initial constituents with only, for example, are similar both semantically and grammatically. In terms of grammar, they're similar in that they also trigger inversion:

> He has seen her only once. (no inversion)

> Only once has he seen her. (inversion)

In this example, we didn't add auxiliary do because it already has an auxiliary, have (in the finite form has). He changes places with has.

In terms of semantics (meaning), it's similar to negation because it implies that he hasn't seen her more than once. So as we discussed briefly before, I've been grouping these examples under the negative-inversion tag.

Now that we've talked a bit about negative inversion, let's get back to the OP's question.

The OP appears to know that little/few are approximate negators and trigger negative inversion when they're moved to the front of the sentence. They've come up with an interesting pair of examples that contain both initial only and a little / a few (which are, by the way, grammatically distinct from little/few without a), and they'd like to know if they work the same way.

This is clear because of two things:

1. The question body makes it clear they're talking about negative inversion. They don't even mention the difference between little and few.

2. The question title gives one example with inversion, and one example without.

Unfortunately, one of their examples used few and the other little, so people jumped on that and wrote about the count / non-count distinction, ignoring the actual question the OP asked.

That's understandable because the question title wasn't clear as written, although if they read the question body they should have understood that they weren't addressing the OP's main point.

So I've gone ahead and edited the title to remove the few/little distinction, after some thought.

The question still talks about few and little, and the answers can still address that.

But the main point is the negative inversion, and the answers need to address that to be valid. They already needed to address that point before my edit, so I don't believe my edit invalidates the existing answers.

To the extent that they are now invalid, they were already invalid.

I thought I'd explain a bit about that because I thought it might be a somewhat controversial edit, and I also thought it was a good chance to discuss the topic of negative inversion.

I'm going to link to this portion of the chat transcript from the question now, in case anyone would like to discuss it.