4

Here are nonnegative sentences where "at all" looks OK. What is licensing it?

  1. It is outrageous that they have been subjected to a trial at all.

  2. Maps are often outdated assuming the links still work at all.

  3. Youngblood's vocals are mixed so low that the ear struggles to make them out at all.

  4. The terrible history of what had happened at these sites haunted me, as did their material remains, but so did the troubling decision I made to be there at all.

LizErsh
  • 43
  • Also, do these sentences look grammatically correct? It's surprising they made a decision to get a dog at all. It's surprising they made a decision to go swimming at all. – LizErsh Aug 10 '22 at 15:22
  • 3 is not an affirmative context: it's saying there are almost no vocals. – Stuart F Aug 10 '22 at 15:38
  • The sentences in your comment are fine, but please don't ask questions in comments. Restrict yourself to one main point in your question post. (And "Is this correct?" is proofreading, which we don't do.) – Andrew Leach Aug 10 '22 at 15:39
  • 1
    The answer is that there are "negative triggers" in your sentences, as explained by John Lawler in "At all" vs "Not at all" in negated sentences. The fact that his answer only contains a "not totally complete" list makes me think this works as its own question: Can we come up with a more comprehensive list? – Laurel Aug 10 '22 at 15:51
  • 1
    "At all" is an adjunct so it is not licensed by anything. It's an NPI, completely excluded from positive declaratives.. – BillJ Aug 10 '22 at 16:35
  • @BillJ But the OP gives perfectly grammatical examples in positive declaratives, though. – Araucaria - Him Aug 10 '22 at 21:34
  • @laurel JL has talked about "negative triggers", but how do you know OP's examples each involve negative triggers? There are other, potentially better, explanations available! – Araucaria - Him Aug 10 '22 at 21:49
  • 1
    @Araucaria-Nothereanymore. In 1. I'd say that "at all" is semantically non-affirmative: you feel that they should not have been subjected to a trial at all". – BillJ Aug 11 '22 at 16:53
  • In positive sentences at all invites comparison with the negated version of a proposition which describes what one might expect to be the case or have been the case. Example (1) implies one might have expected them not to be subject to a trial, (3) suggests that the speaker would expect not to be able to hear the music, (4) suggests that one might expect the speaker to have decided not to be there. (2) is different, in that it is specifically licensed because the NPI occurs within a conditional adjunct. These are well-known to be NPI contexts. – Araucaria - Him Aug 12 '22 at 14:37
  • @BillJ Don't disagree (Not sure why your message didn't ping me ...) Still no news? – Araucaria - Him Aug 12 '22 at 15:41

3 Answers3

5
  • It is outrageous that they have ever been mentioned.
  • Maps are often outdated, assuming the links ever worked at all.
  • His vocals are mixed so low that I struggled to make any of them out.
  • It haunted me, but so did the decision I made to ever go there again.

Note that these contexts also license other NPIs, besides at all.
Thus they are not "affirmative contexts".

John Lawler
  • 107,887
  • 1
    I think OP's question amounts to "Why is this a negative context", though. – Araucaria - Him Aug 10 '22 at 21:32
  • 1
    Well, negation is a mosh of logic, meaning, and syntax. Anything that implies, presupposes, entails, invites the inference, or conversationally implicates the possibility of negation can license NPIs. The references at the end of my encyclopedia article go into it in great detail, which may be more than they want to know. – John Lawler Aug 10 '22 at 22:02
  • How about a summary of that article in the answer? Otherwise this becomes a "link only" answer, with the link in comments. – DJClayworth Aug 11 '22 at 12:25
  • The article is two pages long. It's already a summary. The second page is all bibliography, with links where available. I am not responsible for prejudice against link-only answers. In the official answer above, note that there is no link. – John Lawler Aug 11 '22 at 14:18
  • "Says something positive" is not a useful description, semantically speaking. If clauses and hypotheticals generally, like questions, imply the possibility of negation and that's good enough for some NPIs (not all; there are degrees of neg-neediness, as Haj Ross calls it). At all is an edge case; it's an emphasizer, of the "minimal object" variety, but it can only emphasize something negative. It's considered a "parasitic" NPI in the trade because it often appears with (and probly helps license) other NPIs. – John Lawler Aug 11 '22 at 15:19
  • You say "at all" can only emphasize something negative. How about "If he achieves ultimate happiness at all he will do it today"? – DJClayworth Aug 11 '22 at 22:42
  • If clauses always license NPIs because they always entail the possibility of negation. How good are his chances of achieving ultimate happiness, anyway? – John Lawler Aug 12 '22 at 14:55
  • 1
    @DJClayworth JL won't get to read that because you didn't ping him! In your example, the negative trigger is a condition adjunct. These explicitly allow for two possibilities, P or Not P. [They do that because they are in fact polar interrogative clauses. You heard it here first]. – Araucaria - Him Aug 12 '22 at 15:10
  • 1
    Actually, the system figured it out and pinged me. So "interrogative clause" is now a thing? What happened to questions? This reminds me a bit of Mandarin "shi bu shi" expressions (lit 'be not be'). – John Lawler Aug 12 '22 at 15:13
  • 1
    @JohnLawler. My basic point is that the reality seems very much more complicated than your very short answer indicates. Try "When he achieves ultimate happiness at all it be be with the help of meditation techniques." – DJClayworth Aug 12 '22 at 16:40
  • You explain the appropriateness of using a NPI by the presence of a trigger which licenses it, and in many cases that explanation seems plausible. @DJClayworth has, however, challenged that explanation by presenting some cases in which there is no trigger. Your response to that challenge seems to be that a trigger is implied in these cases. The problem with that response is that, if a trigger (which is not otherwise identifiable) can be implied by the presence of a NPI, then the concept of a trigger cannot explain anything about NPIs. – jsw29 Aug 12 '22 at 16:46
  • 1
    That's right, and that's why NPIs are used, like chemical indicators, in sentences to detect the presence of negation in their meaning. This was all discovered and written up back in the '60s and '70s. Nowadays there are lots of books and dissertations about negation and polarity, if you want to look it up. The references, as I've mentioned, are linked in the bibliography to the encyclopedia article. Start with Larry Horn's stuff; he discovered most of it. – John Lawler Aug 12 '22 at 17:07
  • 1
    @JohnLawler Well, closed interrogative and open interrogative clauses are things now for some writers, (e.g. Aarts 2011, or CGEL). They reserve 'questions' for the semantic/pragmatics, and use 'open' and 'closed interrogative' for the syntactic categories. Kind of works. Ish. – Araucaria - Him Aug 12 '22 at 17:30
  • 1
    Is there a reason all this isn't in the answer? – DJClayworth Aug 12 '22 at 21:06
  • 1
    In my thesis I suggest that many NPS's (most NPIs?)exist to override pragmatic constraints that a listener might otherwise apply when interpreting an utterance. – Araucaria - Him Aug 13 '22 at 09:07
  • 1
    Certainly some of them do. At all overrides the possibility that there might be a little thing we hadn't noticed, which is usually true. – John Lawler Aug 13 '22 at 14:09
2

There are plenty of uses of "at all" with a positive context.

He was ecstatic to have been given a trial at all.

If the link works at all it will give you all the information you need.

If you can make out Youngblood's vocals at all you will be impressed by his power and range.

DJClayworth
  • 25,795
  • If these are positive contexts, and that's doubtful, the presence of at all (which @BillJ says is the NPI) certainly puts a downer on things. – Andrew Leach Aug 10 '22 at 17:52
  • What's negative about "you are lucky"? – DJClayworth Aug 10 '22 at 17:53
  • 1
    It implies "It almost never happens." The import is most definitely negative. – Andrew Leach Aug 10 '22 at 18:17
  • I guess if you think that if something "almost never happening" is negative then you would be right, since that is the definition of "at all". But I don't think so. How about "The army suffered no losses at all". – DJClayworth Aug 10 '22 at 18:20
  • 1
    Again, the NPI accords with no. – Andrew Leach Aug 10 '22 at 18:22
  • What about "He was ecstatic about getting a trial at all". What's the NPI in that, other than saying "at all" is the NPI, which makes the whole thing a circular argument. (You can only use "at all" with an NPI and "at all" is an NPI). – DJClayworth Aug 10 '22 at 18:46
  • That's a good example. I'll think about how to change my answer. – Andrew Leach Aug 10 '22 at 18:48
  • 1
    Professor Lawler doesn't really say any of this in his answer. Maybe he could expand on it? That would certainly address the issues that many skilled native English speakers have where "at all" is used in a seemingly positive way. – DJClayworth Aug 10 '22 at 23:17
  • 1
    @DJClayworth Conditional adjuncts are well-known NPI environments. It's clear why, of course. They allow for both the P and Not P propositions. Your examples include two of them. In truly positive environments at all still relies on negative propositions in that it specifically invites comparison of the proposition with its negative counterpart which often represents what might be expected. At all in your first example specifically implies that one might have expected he wouldn't be given a trial. – Araucaria - Him Aug 12 '22 at 21:03
  • 1
    This is all good stuff, but why is it in the comments instead of the answer? – DJClayworth Aug 12 '22 at 21:05
  • @DJClayworth Which answer? [In any case, I'm not answering these days. Not till I finish my degree. Too dangerous.] – Araucaria - Him Aug 12 '22 at 21:06
  • 2
    Sorry. I mean John Lawler's, not mine. – DJClayworth Aug 12 '22 at 21:07
  • 1
    I think he just fancied a short answer this time! – Araucaria - Him Aug 12 '22 at 21:09
  • 1
    Actually, I've discussed Negative Polarity here and elsewhere pretty frequently. Anybody who wants to know more about what I think is welcome to read it. – John Lawler Aug 13 '22 at 14:14
1

Your sentences do have a negative feel, so at all is well used there.

At all used affirmatively (both in form and 'feel') has mostly died out, but it was there in the beginning, and does survive in some parts of the world:

Before the negative, conditional, and interrogative usages came into being, “at all” was used in affirmative statements to mean “in every way,” “altogether,” “wholly,” and “solely,” the OED says.

The dictionary’s earliest example, from about 1350, is

  • I þe coniure & comande att alle” (I thee conjure and command at all).

The affirmative use of the phrase has died out in common usage, however, and now survives only in some regional dialects of American and Irish English.

A 1945 article in the journal American Speech says this affirmative use “lives on in Irish dialect and in colloquial speech in certain parts of America, especially after a superlative.”

The article, which gives

  • We had the best time at all

for an example, says the usage was reported in Georgia, Kentucky, North Carolina, Oklahoma, elsewhere in the South, and the Midwest.
See more about the affirmative use of "at all" on Grammarphobia

fev
  • 33,009