What is having sufficient evidence?

Attention conservation notice: Large swathes of epistemology are basically foreign to me, and though I did (I promise!) scan at the SEP article titled "Evidence", it's not as if that's much of a qualification. Also the mentioned scanning took place in July, and then I wrote this in early August, and then I forgot about it for a month, and then I revisited it without having revisited the SEP, so the following is acknowledged in advance to be essentially amateurish. On the other hand, this means that I speak with the voice of The Folk and must be heeded on those grounds at least a little bit.

Anyway. I began reading Matt Weiner's paper "On Chum and Chumps" because he posted a link to it, and was brought up short by this bit here:

Ichikawa puts forward a broad account of positive epistemic obligation, which I call the Strong Theoretical Principle:

(STP)

If one considers the question whether P, and has sufficient evidence for P, one ought to believe that P (Ichikawa 2024, p. 83).

STP is arguably too strong. Consider the question of whether there are less than twenty words in that formulation of STP, beginning with “If” and ending with “P.” Your visual evidence is enough to settle that question, and you’re now considering it, but it seems unlikely that (before I tell you the answer) you are obligated to form the belief that there are less [sic] than twenty words in the formulation. It’s just not important for you to count the words in that sentence.

Here is what seemed striking to me about this. The basic form of the argument here is that although I do have sufficient evidence to believe that P, I simultaneously am not able to conclude that P on the basis of that evidence. If I were able to so to conclude, there would be no question about the importance of counting the words. I wouldn't need to. Since I do need to, not only am I not obligated to believe that P, really I am obligated to remain agnostic about P.¹ The question then is, if this is what "having evidence", much less "having sufficient evidence", comes to, why would Ichikawa believe that principle?

Obviously there are ordinary senses of "having evidence" in which it is clearly not correct. I might have evidence regarding your thoughts about all sorts of things in that you have handed me a banker's box full of your papers, but clearly my coming into possession of the box doesn't mean I'm in a position to conclude anything about what you think, especially since (surprise!) your papers are written in Finnish, a language that I could learn but have not learned. The police may have evidence in their evidence locker incriminating so-and-so but they would not be justified in believing in so-and-so's culpability until somewhat actually puts the pieces together. They may not even know that some of the relevant evidence pertains to the case! The evidence in itself jointly contains sufficient information to make a conclusion, but evidence must be made use of; it isn't a talisman. One can perhaps think of it practically: the police "have sufficient evidence" for P in a first-potentiality sense, in that the evidence suffices, and it is possible for them to avail themselves of it, but they aren't even aware that it is evidence—they would have to do substantial work to convert that evidence into something practicable, just as I have the first potentiality to read Finnish (and am thus in a sense able to avail myself of the evidence about your beliefs I physically possess), in that I could learn it, and then have the second potentiality/first actuality, in that I would be in a state of being able to exhibit the second actuality (reading a Finnish text when presented with it).

Weiner's case seems somewhat like that: the second potentiality/first actuality. After all, I can count! The situation is rather as I were sitting in a room next to one of the walls, and you came in and said to me: "I know all the walls in this room are either plaster or drywall, but I can never remember which is which. Is that wall", pointing now at the wall I'm sitting by, "plaster, or is it drywall?" and it were alleged that I have sufficient evidence to settle the question, but am not obligated to do so, because it's not important for me to rap on the wall and listen to the sound it makes. I have the suspicion that "visual evidence" is playing some important role that I can't quite make out, as if because the answer is already in some sense contained within what I'm looking at, I really do have the answer, somehow. But it's hard for me to see how the situation when I'm actually looking at the formulation is substantively different from my merely knowingly having Weiner's paper on my person, and not in my visual field at all. When I do go to count the words, a number doesn't simply precipitate out of what I'm looking at; I change what I'm looking at, looking at each individual word in turn, in a way very different from what I do when reading a sentence. The fact that I'm already looking at it means there's less required of me than if I had to get the paper out and turn to the right page and then look at it, but in either case I have to go beyond what is then-and-there present to me. It isn't even a matter of converting my visual evidence at the moment of being asked into a number, since, as I said, my manner of looking changes too. In both these cases I want to say that it is really, really easy for me to get sufficient evidence about P—I mean evidence in a form I can use²—(the word count, the composition of the wall), but that I don't yet have it, which certainly seems plausible about the wall, since I don't seem to have anything evidence-shaped yet. However, if it admittedly seems odd to say that I have evidence, now, about the composition of the wall simply because I am next to it and know how to exploit the dispositions of plaster and drywall to sound and feel different ways when rapped, and am physically able to rap on the wall, because (I suppose) "evidence" is thought of as a thing, even if a transient thing, and not a disposition, I say that in each case I am able to act on something, either by physically touching it or by refocusing my attention on it, so as to gain information which I am able to use—so that we might say that the wall is the second potentiality of evidence, along with the paper, in some reasonable sense. Of course we may not wish to say that have the (potential) evidence—I don't have the wall—which was the whole question: what is "having sufficient evidence"?

(If I were sitting next to a wall and you were actively rapping on it while asking "is this plaster or drywall, do you think? I know it's one or the other", then surely, if you have brought me to consider the question at all, I ought to believe the wall is made out of the material the sound indicates. Right? The evidence, that your rapping is producing a hollow thuddy sound (say), arrives in a form that requires no further processing to be usable to conclude that it's drywall.)

It seems to me that it's reasonable to use "have sufficient evidence for P" in all these ways—to describe the unwitting or not immediately exploitable possession of material or techniques sufficient to answer the question whether P, to describe the witting or immediately exploitable, but not exploited, possession &c, and to describe the actual presence of the information that exploiting the &c makes available. Weiner seems to mean the second. I assume Ichikawa means the third; certainly, if he means the second, Weiner has his number, but the failure of the principle would seem to be pretty obvious once one spells out what "having sufficient evidence" actually means. But if Weiner means the second and Ichikawa the third then they are talking past each other a bit. On the other hand, it's nice to imagine philosophical argument that seems to imply, even if inadvertently, that if you're sitting at your computer or have your phone within reach and some trivial question occurs to you, you are actually obliged to pull up wikipedia. Perhaps one that works is just around the corner.