After all this time I am still confused about scopal and non-scopal modifiers.
All explanations which I find are very DELPH-IN-centric, in particular the difference is usually explained in terms of qeqs. This wiki page does give some data, but it is too terse for me.
It seems that non-scopal modifiers, unlike the scopal ones, do not allow certain ambiguity. I can see how The dog barked loudly is not ambiguous; loudly modifies barked. But is The dog probably barked actually ambiguous? What is the ambiguity?
Try: That athlete probably won every medal. Every can scope above or below probably (because probably is scopal), giving two readings:
- For every medal, it is probably the case that that athlete won it.
- It is probably the case that that athlete swept the competition (received all the medals).
Note also that scopal/non-scopal is actually a simplification of a larger space. I can’t quickly turn up a discussion, but there’s a distinction between intersective, subsective, … and scopal, I think.
Hmm I must admit I can’t really understand the difference between 1 and 2 :(. I mean, maybe, there is some statement about each medal vs. there is some statement about an athlete?.. But doesn’t it actually mean the exact same thing ultimately, in terms of the state of affairs?
Ah. I think I can feel the difference if I put emphasis on that athlete vs. every medal. So, information structure is probably involved. In 2, the athlete is the topic of the sentence, but in 1 it is the medals that are the topic (looks like).
No, it’s not about the relative scope/emphasis on that athlete and every medal, but between every medal and probably. Is it a statement about the probability of winning each medal independently or about the probability of winning them all?
That I can understand in theory of probability terms but am really not sure the difference is salient in the sentence… I’ll keep trying.
…And yes, I cannot find much on the broader discussion either.
I think the term intersective is used a lot for adjectives; perhaps that’s where I should look.
I think I have some references to scopal semantics in my thesis, as well as a section in the difference between intersective, subjective, and non-intersective semantics. If I recall correctly, Partee introduced those terms. Scopal semantics is sometimes discussed alongside those items, but I think scopal semantics is more of a different sort of thing, as @ebender suggests, because of the presence of scopal ambiguity. So, you may find more info on scopal modifiers in scopal ambiguity literature.
Ambiguous sentences can take time to think about because people disambiguate so automatically. So another way is to think about different sentences:
- That athlete definitely never won a medal. (I’m certain they won nothing.)
- That athlete never definitely won a medal. (There are no wins I’m certain of.)
The order of non-scopal modifiers doesn’t affect truth conditions (even if there is sometimes a preferred order):
- That athlete slowly, steadily won many medals.
- That athlete steadily, slowly won many medals.
I tried to post this yesterday but the site went down again… I’ll put it anyway, but I think T.J. and Guy have already covered many of the same points.
Back to your original example, The dog probably barked., imagine it as an answer to these questions:
The trick-or-treaters ran away without their candy. What happened?
(probably scopes over the)
Do you think the dog barked or begged at the trick-or-treaters?
(the scopes over probably)
As for intersective adjectives, in my dissertation (Section 5.3.3, ~p74) I gave the following non-scopal adjective classes and examples:
- intersective: white cat
- subsective: cognitive scientist
- intensional: fake gun
Also this paper by Ulrich Reichard looks nice: https://philpapers.org/archive/REIIAG.pdf
I just skimmed it but it has some examples with set descriptions of the phenomena. Note that what I called “intensional” Reichard calls “non-subsective”, and also discusses another term, “privative”. It doesn’t seem to cover scopal adjectives, though.
I think the only way I can see the difference is still by placing different emphasis on different constituents (i.e. assume different information structure). But in the example with the athlete, Emily disagreed with that interpretation. Perhaps she meant to say, it is about the relative scope on probably and every medal (not that athlete and every medal)—but the information structure difference still may be with respect to that athlete and every medal?..