Here’s a fun case. I was walking around the Science Museum in Buffalo with Simon yesterday, and I came across this odd passage in the dinosaur exhibit. Here is a picture of it, but I’ll type out the passage so you don’t have to click away.
Triceratops, the Fighter
Paleontologists believe that Triceratops used its three horns for fighting off other Triceratops, as well as other meat-eaters. How do they know this? Some Triceratops skulls show facial wounds that appear to have been caused by horn points. They were probably caused by fights with other Triceratops to establish dominance within a particular group or territory.
Here are some reasons I’m interested in it.
- A belief ascription is followed up by a knowledge ascription.
What I find odd about this, that belief attribution in a context like this is generally used to avoid giving the appearance of attributing something as strong as knowledge. - The knowledge ascription is in a context where the possibility of error is salient.
The question seems to anticipate the reader having doubts about paleontologists’ ability to know what Triceratops used their horns for. I’m assuming that the museum curator, or whoever wrote this, was not unaware of the possibility of error when he/she wrote this. - The grounds actually cited provide (at best) a very low degree of justification.
I’m not claiming that these are not adequate grounds for believing that Triceratops fought other Triceratops, but what I am saying is that the knowledge ascription, taken at face value, presupposes a low-level justification requirement on knowledge.
I’m inclined to think that this is an infelicitous use of ‘knows’, but it’s a fun philosophy exercise to pretend that it’s not. Suppose this were a knowledge ascription that we felt pressure to capture. What epistemic theories would have problems with it?
Contextualism
Some varieties of contextualism that hold that speaker awareness possibility of error is sufficient to shift the meaning of ‘knows’ to some property that we don’t satisfy. Assuming (2) above is true, then these versions of contextualism have a problem with this case.
Traditional Invariantism
Folks with views of the word ‘knowledge’ that are more traditional will be OK with this case, however, not all traditionalists will be happy. The bar for knowledge that seems presupposed by the above ascription is set pretty low. If you think knowledge requires a high level of justification, then this case is trouble.
Subject-Sensitive Invariantism
SSIers do alright. The practical stakes in this case for being wrong about whether or not Triceratops fought with their horns is presumably pretty low. Assuming you’re an SSIer that’s comfortable with a low evidential bar on knowledge, then you should be OK with this case.
Anyway, this was more of a fun little game. As I said, I’m inclined to think that this is an infelicitous use of the word ‘know’. However, if I’m wrong about that – we have a prima facie problem for certain versions of contextualism and high-level justification invariantists.
Yeah, that’s probably infelicitous—maybe just an oversight. Then again, the use of the “believes” might not conflict so much with the knowledge ascription given that belief is ascribed to purported experts. I’ve seen this phenomenon elsewhere too. People will fairly naturally slide between saying “all the scientists believe” and “we (or they) know.” I assume this is because something like this is in the background: if all the experts think it, then they probably have good evidence, etc., so they must know it.
But then there’s the issue of seemingly low justification. But maybe that’s not much of an issue if the person is trying to be brief on a little sign in reporting the evidence and is maybe only reporting part of it. The author says “some” of the heads had the holes. That’s consistent with there being a lot of them. And maybe there was a single pattern that followed so closely to the horns, etc. etc.
But it does seem odd no matter what to use “know” here. It seems unnecessary for the point that’s trying to be made. Cool case either way!
Hmm. Here are two comments:
(1) What versions of contextualism have the strong result that you suggest? I don’t think I know any contextualists who think that mere awareness of some possibility of error is enough to make knowledge attributions that difficult.
(2) I don’t see why you think it’s an infelicitous utterance. You think paleontologists don’t know why the triceratops had three horns? Why?
I agree with Jonathan regarding contextualism. It seems like the usual result would be similar to the SSI analysis.
I think Lewis would say that mere awareness of an error possibility precludes a correct knowledge attribution. See his Rule of Attention, “Elusive Knowledge,” p. 559.
Contextualism is often cited as offering a solution to the skeptical paradox (including possibility of error style skeptical arguments), and a way to capture infallibilist intuitions. The versions of contextualism that do this are the ones that spell out the semantics for ‘knows’ so that the meaning shifts to some property that requires impossibility of error in contexts where the possibility of error is made salient.
Regarding the bit about whether the knowledge ascription was infelicitous…I was just reporting that I had an inclination to think that. I’m not sure what to think about it yet.
I can try and guess why I’m inclined to think it is. One, it may just be the dialectical weirdness of following up a belief ascription with a knowledge ascription. I suspect when people deviate from certain conversational norms it can give the illusion of infelicity. I suspect a better explanation of my inclination is that the cited reason for ascribing knowledge was something relatively weak. If you thought knowledge required more justification than what was cited, you might be inclined to think that the knowledge ascription was infelicitous. Those are just some guesses.
So, assuming that the story given about the paleontologists’ evidence is correct and complete, do you think this is true?
“Paleontologists don’t know to what use triceratops put their horns.”
I’m inclined to think that is true. But let’s clarify what I take to be the complete and correct evidence.
(A) Some triceratops skulls have marks of such-and-such a shape
Merely knowing that there are dents in the skull that have a shape with a certain appearance (and nothing else) would be evidence that some Triceratops had wounds that we’re caused by horns. Perhaps sufficient for justification (aside: I am inclined to think it’s pretty easy to get justification).
Now suppose we grant that paleontologists know that these marks were caused by other Triceratops horns. Presumably additional information would be brought to bear to support that the horn marks were caused by an act of offense or defense rather than say dino-foreplay or something….
Now we can appeal to other evidence about the behavior of other species in the animal kingdom (with horns) and perhaps gather some additional inductive evidence to support the thesis that Triceratops used their horns to fight others…but that sort of evidence is not stated in the passage.
Note this is evidence that I think paleontologists probably have…I agree with Joshua’s point above that there probably is a lot more evidence that paleontologists rely on that is not stated in the passage.