What is it to wonder? I’m going to defend an analysis that I think might be right: to wonder is to want to know. This should be of interest especially to epistemologists, given wondering’s central role in inquiry.There seems to be something of an inquisitive turn going on in epistemology, continuing a rich tradition that goes back to Plato and Aristotle. See, e.g., recent papers by Jane Friedman ((2013) "Question-directed attitudes”, (2017) "Why Suspend Judging?, (2019) "Inquiry and Belief”, (forthcoming) "The Epistemic and the Zetetic”), and Christoph Kelp ((2014) "Two for the Knowledge Goal of Inquiry" (2019) "Inquiry and the Transmission of Knowledge”, (2020) "Theory of inquiry”).
Why Analyze?
Before getting into the analysis, a quick methodological preamble. I agree with those who think epistemologists shouldn’t get too hung up on finding analyses of the various states and relations they are interested in: knowledge, belief, evidence, etc.E.g., Williamson (2000) Knowledge and Its Limits and Pasnau (2017) After Certainty. What we really want are good descriptive and normative theories of these things, or whatever things in the ballpark turn out to be most interesting or important. And we shouldn’t presume that the good theories will make significant use of reductive analyses. Fortunately, most epistemology going on these days doesn’t presume this, and doesn’t focus too much on finding analyses.
But this isn’t to say that if there is some true reductive analysis involving important epistemic notions that it isn’t worth knowing. Analyses make for nice theoretical connective tissue, letting you port over independently motivated views about the analyzing notions to the analyzed one and vice versa. They also at least give you the option of claiming a kind of parsimony: Defined Thing, you might say, just is Defining Thing, so we no longer need to treat these as separate ontological commitments. Or if you prefer lush to arid landscapes, you have the option of saying that we have given a real definition, revealing something’s essence. Analyses aren’t the end-all, be-all, nor are they necessary preliminaries to theorizing, but they’re good if you can get it’s ’em.
The Analysis
Okay, so here it is:
WTK Theory of Wonder S wonders Q iff S wants to know Q.
For instance, I want to know who will win the election, so it follows that I wonder who will win the election. I wonder how many degrees of warming there will be this century, so I must want to know how many degrees of warming there will be this century. It’s not the case that I want to know exactly how many blades of grass there are in the nearby park, so that’s not something I wonder. And I don’t wonder whether it’s raining in Topeka, so it must be that it’s not the case that I want to know whether it’s raining in Topeka.
Results so plausible as to seem trivial. Perhaps this is why no philosophers have defended this analysis.As far as I know. Please tell me if I’m missing some literature on this. Note also that some linguists—Kartunnen (1977), Guerzoni and Sharvit (2007)—have suggested a semantics for wonder along these lines. Uegaki (2016) argues that on this analysis, plus some assumptions about want’s presuppositions and their projection, one can explain why we can’t say that someone wonders that p. This is a cool result, but see Theiler, Roelofsen, & Aloni (2019, §A.2) for reasons to think this won’t generalize to other clause-embedding predicates that don’t take declaratives. They also raise a problem to do with presupposition projection which I might address at some later date. But in fact the analysis is inadequate as it stands. We need to get clearer on what kind of wanting is involved, which will help us avoid some apparent problems for the analysis. We also need to deal with a problem about metacognition.
What-things-considered?
Here’s a class of potential counterexamples: sometimes we find ourselves wondering something when we don’t really want to know. Suppose you’re wondering how many years you have left to live. But you also think that if you could somehow know the answer to this, it would cast such a pall over the rest of your life that any benefits to knowing would be outweighed. If someone offered you a glimpse into the crystal ball, you would turn it down, because you don’t want to know. Nevertheless, you could still be wondering how many years you have left to live. So it looks like we have wondering without wanting to know. Another analysis bites the dust, it seems.
Not so fast! Distinguo: all-things-considered wanting and some-things-considered wanting.
Suppose you invite me to play tennis on Tuesday at 2:00. I would really enjoy this, need the exercise, and would be happy to have an excuse to catch up, but I’m scheduled to teach at that time. As Wayne Davis points out,Davis (1984) "The Two Senses of Desire”: p. 239. I can truly describe my mental state in a pair of contrasting ways:
- “I want to play, but I have to teach.”
- “I don’t want to play, since I have teach.”
The former reports that I some-things-considered want to play, whereas the latter reports that I all-things-considered want not to play.
This distinction allows us to avoid the counterexample. Wondering, we should say, is wanting to know some-things-considered. When you wonder when you will die, you want to know some-things-considered, even if all-things-considered you’d rather not know. That seems right to me.
Activity
Here’s another concern about the WTK Theory.Friedman (2013, p. 154) objects along these lines. Thanks also to Zoltán Szabó for raising it in conversation. The verb to wonder seems to express an activity, whereas to want to know expresses a state. They have different aspectual classes.For a classic division of verbs (or really, verb phrases) by aspectual class, see Zeno Vendler (1967) Linguistics in Philosophy: Ch. 4.
One mark distinguishing these classes of verbs is that statives don’t take well to the progressive, whereas non-statives do. Picture a scene: your friend is sitting at their desk, pulling the Thinker pose. Contrast the following as descriptions of what’s going on.
- “They’re thinking that expressivism might be true.”
- “They’re believing that expressivism might be true.”
The latter sounds terrible! This is because believe is a stative verb, in contrast with think, which is non-stative. Belief is a state you can be in, whereas thinking is some kind of activity—something that happens.
Now let’s try this test for wonder and want.
- “They’re wondering whether expressivism can make sense of normative inquiry.”
- “They’re wanting to know whether expressivism can make sense of normative inquiry.”
Again, the latter sounds terrible! It seems that want is a stative verb and wonder is not. Desire is a state you can be in, whereas wondering is some kind of activity—something that happens.
This looks bad for the WTK Theory of Wonder. A successful analysis of wonder will have to identify wonder not with a state like desire, but with something non-stative, some kind of activity or happening.
Here’s one alternative proposal: trying to find out. This gets many normal cases right. When I’m wondering what I should cook for dinner, I am trying to find out what I should cook for dinner by thinking through what ingredients I have, how long various options would take and how much effort they would involve. And it’s not easy to think of cases where one is trying to find something out but doesn’t wonder the relevant question.
That said, I don’t think it gets all the cases right. It has trouble, first of all, with the merely some-things-considered wanting to know cases. You can be wondering when you will die while purposefully not trying to find out the answer to that. Second, it has trouble with cases of wondering what you realize you’ll never know. I sometimes wonder what Mozart would have composed had he not died young, but I’ve never tried to find out, since it’s so transparently hopeless.
I think a better way to go here preserves the WTK theory, again by making a distinction between types of wanting. Distinguish, as Goldman does,Goldman attributes the distinction to Alston, but no doubt it goes back further, in some form or other. between occurrent and non-occurrent wants. Here’s how he introduces the distinction:
John is concentrating on finishing the lawn by six o’clock. He is giving all his attention to mowing it as quickly as possible, with the thought of getting it done by six. During this period, whether it be a few seconds or a whole minute, John has an occurrent want to finish mowing the lawn by six o’clock. The thought of finishing the lawn by six occurs to hum, occupies his attention, fills his consciousness. It is a datable event or process. During the same period it would also be correct to say that John wanted to be president of his company. But his wanting to be president of the company is not an occurrent want during this time, for John has not been thinking about the company or his future at the company. Nevertheless, we could say that John wants to be president of his company because he has a disposition to have occurrent wants to this effect, and this disposition is present while John is mowing the lawn, though it is not being manifested. I shall say that John has a standing want to be president of his company.
Pinning down what precisely this distinction amounts to is no easy task,See Bartlett (2018) "Occurrent States" for one recent attempt. but I take it to be clear enough as is for my purposes. Note that this kind of want is not just a philosopher’s postulate. We do have words for at least some such wanting events. In English: longings, cravings, hankerings, lustings, yearnings. It would be a bit over the top to say that John was yearning to finish mowing the lawn, but it is something that could be said if he wanted it badly enough (and perhaps met some other conditions required for yearning).
What’s important is that there is a kind of wanting that is a “datable event or process”, a non-stative kind of thing. We can naturally use the progressive here. John is occurrently wanting to finish mowing the lawn by six, craving chocolate cake, etc.
We can solve the non-stative problem for the WTK Theory, then, by saying that the kind of wanting it identifies wondering with is occurrent wanting. To wonder is to occurrently want to know. Modulo the exaggeration and other connotations, we might say it is to crave or yearn for knowledge.
Besides getting the non-stative-ness right, this also works for the merely-some-things-considered and the hopeless wondering cases. Craving something but not wanting it all-things-considered is very familiar, as is yearning for what you know will not happen.
Why knowledge?
I’ve spent all this time talking about the kind of wanting involved in wondering, but what it is supposed to be a want for? Why think it’s knowledge? To be honest, this is mainly just a hunch. There are various alternatives worth considering.
Maybe all we want is belief, as Peirce claims.Peirce (1877) "Fixation of Belief”: §IV. Peirce is talking about ‘doubt’, but it seems to me that he has wondering in mind. Or maybe true belief, or justified true belief, or stable credence, or rational stable credence, or some other non-knowledge-entailing state. But I don’t think so. Peirce observes that one stops wondering once one has settled one’s opinion into a belief. But one also stops yearning for what one mistakenly believes one has attained. The desire is not satisfied in such cases, but merely dissipates. This is what strikes me as going on in cases where one stops wondering but doesn’t know.
This is not yet a substantive argument against these other options, but I suspect one can be given (maybe in another post, if I can think of an interesting one).
I’m less confident that the object of a wondering desire shouldn’t be taken to be something something that entails knowledge, but involves more. Perhaps wondering Q is wanting to learn Q or to find out Q. The usual way one goes from not knowing to knowing is through learning or finding out, so it wouldn’t be surprising if most cases of wanting to know also involve wanting to learn or find out. But it does seem possible to come to know without learning or finding out. Sci-fi examples come to mind: one may come to know through some kind of direct uploading without learning or finding out. I will leave trying to figure out if one’s wondering is satisfied or merely dissipated in such cases to another time. Perhaps in the end we should favor the WTF (Want To Find out) over the WTK Theory.
These are still cases of coming to know through coming to have a belief. Another way one might come to know without learning is going from having a belief that is not justified to one that is, or going from Getteried justified true belief to un-Getteried justified true belief. It’s somewhat unclear whether one can really want to undergo such a change, for Moorean Paradox-ish reasons, but if one can I suspect we wouldn’t want to call it wondering. If so, I think the WTK theory can handle this with appeal to some form of contrastivism. Something like wanting to know Q as opposed to remaining in suspended judgement about Q.
For now, then, I’m inclined to accept something along the lines of
WTK Theory of Wonder (elaborated) S wonders Q iff S occurrently, some-things-considered wants to know Q rather than remain in suspended judgement about Q.
Metacognition Without Cognitive Concepts
One more potential counterexample. If Samson the puppy lacks the concept of knowledge (plausible), he can’t have attitudes towards contents about knowledge. So he can’t have desires about knowledge. So he can’t, e.g., want to know stuff. But he does seem to wonder. He wonders what made that sound, which is why he’s going to the other room. But apparently he can’t want to know what made that sound. So wondering can’t be wanting to know.
More generally it seems that we’re comfortable describing various non-human animals as wondering, but we wouldn’t be so comfortable ascribing to them a concept of knowledge. Such creatures lack the capacity for metacognition, and so must lack the capacity for metacognitive desires. So wondering can’t be a metacognitive desire like wanting to know.Friedman (2013, §3) makes this argument, as does Carruthers (2018) "Basic Questions”. Whitcomb (2010) "Curiosity was Framed" makes the same point about curiosity, building on Kvanvig (2003) The Value of Knowledge and the Pursuit of Understanding: pp. 143ff.
Attributing to non-human animals mental states with particular contents is fraught with philosophical difficulty, so I’m not confident what the best view is here. However, I’m unpersuaded that the analysis is refuted.
One reason I’m unpersuaded stems from how I myself feel about the cases. In most contexts, I’m totally happy to say that Samson does want to know what made that sound. “Why’s Samson trying to go there?” “He wants to know what made that sound.” This just seems true to me.
Does that mean I think he possesses the concept
In other contexts—like ones where philosophical issues about animals’ mental states are raised—I admit I am hesitate to say he really wants to know what made that sound. But in those same contexts I’m similarly hesitant to say he wonders what made that sound. Regardless of context, answering the question about why Samson is going to the other room with “He wants to know what made that sound” seems about equally objectionable as “He wonders what made that sound”. This is as the WTK Theory would predict.
The other reason I’m not persuaded is more theoretical. I think there are good reasons to think that at least some kinds of metacognition are less cognitively demanding than the skeptics make it out to be, and that Samson and other non-human wonderers do represent their own knowledge states in some of their goal representations. But that gets into issues about psychosemantics best left for their own post.
In the meantime, email me if you’ve got other counterexamples to the WTK Theory of Wonder.