Skip to main content.
May 29th, 2004

Vague Adverbs

We have lots of heuristics to tell whether a predicate is vague. If it admits of borderline cases, or if it is Sorites susceptible, it’s probably vague. But predicates aren’t the only vague terms. Predicate modifiers, like adverbs, can be vague. How should we test adverbs for vagueness?For concreteness, let’s focus on a particular (intuitively) vague adverb, ‘quickly’. How could we test/argue for the vagueness of ‘quickly’.One idea is that ‘quickly’ is vague iff for some verb V, V quickly is vague. But that won’t do, because V quickly might be vague in virtue of V being vague.Here’s a Saturday night live guess at what we might do. If we can find a verb V such that there’s a sorites sequence for V quickly such that each term is a clear instance of V, then quickly is vague. That condition seems necessary and sufficient at first pass. Or so it seems to me. Counterexamples and counterproposals welcome!

Posted by Brian Weatherson in Uncategorized

9 Comments »

This entry was posted on Saturday, May 29th, 2004 at 9:36 pm and is filed under Uncategorized. You can follow any responses to this entry through the comments RSS 2.0 feed. Both comments and pings are currently closed.

9 Responses to “Vague Adverbs”

  1. djc says:

    Here’s an artificial counterexample. Let’s say that X speeds doubly iff someone going half as fast as X would still be speeding (likewise for any verb associated with quantitative degrees and thresholds). Then if “speed” is precise (threshold at 65mph, say), “speed doubly” is precise (threshold at 130mph). But if “speed” is vague, “speed doubly” is vague. So it looks like any vagueness in “V doubly” is inherited from vagueness in V. So it looks like “doubly” is not vague. But it passes your Sorites test.

  2. Michael Kremer says:

    djc: I’m sorry. I’m slow today (doubly slow, in fact). Can you explain to me how “doubly” passes Brian’s Sorites test? That is, can you provide a verb V and a sorites sequence for “V’s doubly” such that every instance of the sequence is a clear case of V-ing? (Please do not speed doubly, or even singly, through the explanation.)

  3. djc says:

    Let’s say that “speed” is vague, that 60mph is determinately not speeding, that 70mph is determinately speeding, and that in between are some borderline cases. Then [120mph, 121mph, …, 140mph] is a Sorites sequence for “speed doubly”, all of whose members are determinate cases of speeding.

  4. Kenny Easwaran says:

    I take it then that you want to avoid considerations of vagueness based on the way that a single adverb can mean different things applied to different predicates or verbs? If so, then this might avoid this “speed doubly” objection, because it relies on the particular behavior of an adverb with a particular verb. “Doubly” doesn’t even seem well-defined for most verbs, and it seems almost that we should consider “speed doubly” a separate lexical item from either “speed” or “doubly”, rather than a compound of the two. Of course, at this point it seems almost more a matter for linguists than philosophers.

  5. Matt Weiner says:

    It seems to me that the problem that arises with “doubly” is that “doubly” isn’t a factive adverb*—that is, “V doubly” does not imply “V.” Brian’s proposal might work for factive adverbs—for instance, “barely” is factive, and even if V is not vague “barely V” will be, and “barely” is probably an adverb you want to count as vague.

    A problem with this is that we want to be able to account for vague non-factive adverbs. For instance, “almost” is anti-factive, and it’s also clearly vague (even if “speed” is non-vague at 65 mph, “almost speed” is vague).

    A simple suggestion is that a-ly is vague iff for some verb V, V a-ly is vague and V isn’t—since that the test we seem to be using!

    *Don’t know the real term for this.

  6. Matti Eklund says:

    Matt’s suggestion seems correct. However, the suggestion doesn’t address what seems to be the underlying issue behind Brian’s question about adverbs: whether vagueness should be properly characterizable as sorites-susceptibility. A revised proposal more in line with Brian’s original one would be: a-ly is vague iff for some verb V, V is not sorites-susceptible, but V a-ly is.

    [Notice that both this suggestion and Matt’s, like many suggested tests for when an expression is vague rule out immediately that vagueness should be an “emergent” feature – that a complex expression could be vague without any of its constituents being so. I would certainly agree that it’s odd that the vagueness of a complex expression should be emergent. But even so, what, exactly, justifies ruling this out by definition?]

    To me the revised suggestion in terms of sorites-susceptibility seems fine (although I wouldn’t be surprised if there are clever counterexamples) – but Brian: there are other things you have said which, if correct, might present problems. Specifically (in the entry Eklund on Vagueness…) that ‘tiny number’ is vague but not sorites-susceptible. (Because 3 might be the only borderline case of being a tiny number.) First, it is at least odd to focus on sorites-susceptibility in the case of adverbs if we already take ourselves to know from elsewhere that vagueness cannot in general be characterized by appeal to sorites-susceptibility. Second, although I haven’t been able to come up with an example, it seems that if Brian is right about the ‘tiny number’ case, then there should be adverbs which exhibit the same phenomenon…. [In the spirit of artificial counterexamples, let ‘tinies’ be a verb such that ‘n tinies’ is true just in case n is a tiny number. Then can’t there be a (non-factive) adverb ‘x-ly’ such that ‘tinies x-ly’ is sorites-susceptible?]

    I happen not to be convinced by Brian’s ‘tiny number’ example myself, but that is a different story.

  7. Matt Weiner says:

    I realized after posting that “doubly” is factive as applied to “speed”; if applied to other verbs analogously it probably won’t be factive though. However, we could just stipulate that “f-doubly” is just like doubly but factive (A`V’s f-doubly iff A V’s and A V’s doubly); then Dave’s example goes through just the same.

    A less artificial example: “Twice.” If V is sorites-susceptible you can get a sorites series for “A V-ed twice”: case 1: A determinately V’s and determinately doesn’t V
    Case 2: A determinately V’s and then does something a little closer to V-ing than in case 1
    Case 3: A determinately V’s and the does something a little closer to V-ing than in case 2
    etc.

    All these cases will be cases in which “A Ved” is determinately true, because of the first time. Yet “Ved twice” is definitely inheriting its vagueness from V. (And I think this argument goes through no matter how vagueness is defined—if V has a borderline case, then V-ed twice will have a borderline case that’s not a borderline case of V-ed.)

    But we may well want to distinguish temporal modifiers from adverbs!

  8. Matt Weiner says:

    About emergence—“acceptably” might be interesting here. There will probably be some situations where “X V-ed acceptably” is not vague, because clear guidelines have been laid down for what is and is not acceptable V-ing in the circumstances. There will be lots of other situations in which “X V-ed acceptably” is vague (off the top of my head, “X sang the national anthem acceptably” would be one).

    Is it possible that, in these latter cases, the vagueness is emergent? I’m not sure. I’m tempted to say that in this case it doesn’t make sense to say that “acceptably” is or is not vague in and of itself—in some situations it leads to vagueness, in others it doesn’t, and there’s an end on’t. But that may not be an acceptable way of resolving the problem.

    (I also haven’t established any cases in which “V acceptably” is vague but “V” isn’t.)

  9. Matti Eklund says:

    The ‘acceptably’ example is rather nice.

    One worry concerns whether the issue of whether vagueness can be emergent is just one of theoretical bookkeeping. (I know, I was the one who brought up the issue…)

    Suppose we have a case where one theorist says that x-ly isn’t in itself vague but still, for some non-vague V, Vs x-ly is vague – and an opponent insists that this means that x-ly is vague: it is only that the vagueness shows up only in certain linguistic contexts. What’s the cash value of their dispute, if any?