Apparently my ‘book’ is already linked on
some websites.
This is dangerous territory. I do hope people don’t take it to be the last
word. For what it’s worth, the discussion of epistemicist theories of higher
order vagueness now seems to me to be seriously misguided. Hopefully a
correction will be posted shortly. This is not to say the other bits do not seem seriously misguided!

In
better news, Brown email is back working again, and apparently now with virus
scanning installed. I was rather miffed at the time about the delays (and if there
was time-crucial email I missed I might still be) but this was one of the
quicker repairs to a major (and I mean major)
technical problem at a large institution I’ve seen. So well done Brown techies.

Sunday’s edition of the “papers blog”:http://www.brown.edu/Departments/Philosophy/Opp/ is online, but the only new paper is my APA paper, and I’ve already plugged it here several times, so there’s barely any need to read it.

“Michelle”:http://www.platonicrelationship.com/2004_03_01_blogarchive.html#107912617199758442 doesn’t like some of the trends in contemporary epistemology.

bq. We as humans are not brute memorizing machines…we do not aim only to collect the most true information possible (which is probably a reason that we don’t memorize the phone books). What we are are __rational__ beings who seek reasons and explanations for things. We want to make sense of the world, and we do that by understanding it…. I find it shocking that contemporary epistemology doesn’t seem very interested in what I take to be a fundamental drive that we all, as rational creatures, possess. Why are lottery paradoxes more important than trying to sort out the nature of understanding and explanation? Why have word games and brain teasers become standard fare in epistemological debates, while far more fundamental issues are ignored?

There is a large field of study of explanation – for some of the best current work see “here”:http://www.stanford.edu/~strevens/research/overview.html#expln – but for some reason it doesn’t seem to be classified as part of mainstream epistemology.

I often don’t understand the classifications of people into various areas of philosophy. This has a few quirky consequences. For one thing, when people ask what I do, I have to stutter in place of answering because I don’t have a nice classification. For another, when I’m hiring I don’t pay as much attention as some people to the areas in which the department is nominally hiring. My usual rule is that if you advertise for an X, you hire the best philosopher who can teach undergraduate classes in X. Occasionally I look for people who can teach grad courses in X too, if it’s really important that those courses get taught and no one else can teach them. But normally I just have some initial constraint of making sure the person can cover some needs in area X, and then look for the best philosopher. This leads to a bias towards generalists, which of course I’m perfectly happy with.

Returning to the topic of understanding and explanation, I’ve been trying to think of a decent way of working the marginal revolution into a blogpost ever since Wednesday’s seminar and not having much success. It seems to me the marginal revolution should be a nice case study for theories of explanation, but I can’t quite figure out how all the pieces fit together.

On the one hand, it seems to me that in some pretty important sense that by 1880 we understood why diamonds were more expensive than water, and before 1870 we really didn’t understand that. We just didn’t __know__ why cost did not equal average utility. On the other hand, it’s not like the marginalist story is entirely unproblematic. It involves all sorts of idealisations, and even with those idealisations it might not work in every single case (e.g. we might have to complicate the story in the labour and capital markets). And although the pre-marginal explanations of why diamonds are more expensive than water don’t always make a lot of sense, they can probably be reconceptualised as idealised, not-fully-general, causal explanations. If I ever get around to writing a massive paper on explanations in social sciences, I want to say __something__ about this, but I’m a long way from knowing exactly what.

Getting directly back to lotteries, I think it’s a little unfair to lump that in with other ‘word games’. (I of course enjoy word games, but I enjoy all sorts of frivolities.) I think thinking about lotteries is a good way to get insights into the role of probabilities in epistemic concepts. And I think that’s true whether we care primarily about knowledge or about explanation.

Just what should the status of probabilistic reasoning be in explanations? Presumably this depends on what the probabilities are. If we know that there is a physical chance x of some event being caused by y, then we shouldn’t expect an explanation of the event that causally entails its happening, just an explanation that says “y happened, and that caused x to happen.” But if the probabilities are not primitive physical probabilities, but some other kind of probability, then perhaps things are trickier.

If we have a lottery that is strictly speaking deterministic, but which we couldn’t possibly know in advance how it will turn out, or even rationally assign higher probabilities to some outcomes rather than others, is it sufficient to explain why a particular ticket lost that it was highly probable that it would lose? If not, does this show we can never use anything other than physical probabilities in explanation? I think the answers are _no_ and _no_, but getting a theory of probabilistic explanation that accounts for both those answers will be non-trivial. So I suspect lotteries will be helpful even for those who care about less momentous matters than the intension of the English word ‘knows’. Maybe it won’t be, but I’d suspect it will.

While on epistemology, which I sort of was, “Scepticism, Rationalism and Externalism”:http://www.brown.edu/Departments/Philosophy/homepages/weatherson/sre.htm got accepted for the “Inland Northwest Philosophy Conference”:http://www.class.uidaho.edu/inpc/7th-2004/index.html. I’m quite happy about this because it’s one of my better papers; one of my rare efforts to really attack a central philosophical problem rather than knocking off peripheral puzzles as they arise. In general I think the periphery is where philosophy has most contact with the outside world, so I don’t feel too bad that I don’t spend much time on core puzzles. But it is nice to dive back into central issues from time to time.

There’s no truth at all to the rumour that Cornell’s “commencement speaker”:http://www.news.cornell.edu/releases/March04/Cornell.Clinton.2004.lgk.html asked for the chance to speak at Cornell when he heard I was moving there. Well, not much truth to it at least.

There’s no truth at all to the rumour that Cornell’s “commencement speaker”:http://www.news.cornell.edu/releases/March04/Cornell.Clinton.2004.lgk.html asked for the chance to speak at Cornell when he heard I was moving there. Well, not much truth to it at least.

Geoff Pullum is a syntax nerd, but don’t let that stop you reading “his latest post”:http://itre.cis.upenn.edu/~myl/languagelog/archives/000580.html!

But just to illustrate that even nerds can make mistakes, he’s wrong when he says that “macaques have knowledge how but not propositional knowledge”:http://itre.cis.upenn.edu/~myl/languagelog/archives/000525.html. This is not to say that Andrew Carstairs-McCarthy, the target of Geoff’s attack, is right. It’s pretty clear that the macaques in question have never said anything with propositional content. But since, as Jason Stanley and Tim Williamson argued (“Knowing How”, __Journal of Philosophy__, **98**, 2001: 411-44) all knowledge-how is propositional knowledge, it is impossible that the macaques could know how to get food without having propositional knowledge.

Stanley and Williamson argue that we can directly construct a proposition that a creature who knows how to blah knows: __thus is a way to blah__, where __thus__ picks out demonstratively the method they use to blah. I won’t repeat Stanley and Williamson’s argument here, instead I’ll run a secondary argument that can bolster their conclusion. (UPDATE: Actually, this is one of the arguments they run as well. Oops. Think of this whole post as being expository rather than setting out something that might be construed as, er, new.)

‘Knows’ is not ambiguous between (on the one hand) a relation that holds between a knower and a proposition and (on the other) a relation that holds between a knower and a type of action. If it were, (1) would be problematic, when actually it is fine.

bq. (1)   Alex knows which places sell beer this time of night and how to get to the nearest one.

‘Knows’ is ambiguous between the relation denoted in (1) and a relation that holds between two people as in (2).

bq. (2)   Alex knows Shea.

We can argue for that by noting that (3) is odd, in a way that (1) isn’t.

bq. (3)   ??Alex knows which places sell beer this time of night and Shea, who runs a couple of them.

So the standard test for whether a term is ambiguous seems to work here – it correctly returns the verdict that ‘knows’ is at least two-ways ambiguous – and it says that it isn’t ambiguous between a ‘knowing-how’ reading and a ‘knowing-that’ reading.

(I know that some languages have different words for these two meanings. That’s no proof that English ‘knows’ is ambiguous. After all, most European languages have different words for female cousin and male cousin, but that’s no proof that English __cousin__ is ambiguous between female cousin and male cousin. The ambiguity test seems to show fairly definitively the English word is not ambiguous.)

From the fact that the English word is unambiguous, we can get the conclusion that knowledge how is propositional knowledge in a couple of ways.

First, we can try a sledgehammer approach. ‘Propositional knowledge’ is a term of art. It’s not like we have these independently defined things, propositions, and we can then ask whether a macaque or whatever knows one of them. Propositional knowledge is just the relation that a creature stands in to something when a sentence “x knows y” is true, x refers to that creature, and ‘knows’ means what it means in paradigm instances of propositional knowledge claims. So saying “That macaque knows how to get food” just is to say that the macaque has a piece of propositional knowledge.

If you don’t like being that blunt, we can try more subtle methods. The best explanation for the fact that there’s no ambiguity in English at this point is that there’s no difference in meaning for English to latch onto. Since many competent speakers think there’s an ambiguity, the best explanation of that is that there’s only one relation here for the word(s) to latch onto. So we again conclude that propositional knowledge also encompasses knowledge-how.

So I think Geoff is wrong to say macaques don’t have propositional knowledge, though he’s right to say they don’t express that knowledge.

To conclude on a more conciliator note, Geoff is entirely correct about which things “should be taught in high school but aren’t”:http://itre.cis.upenn.edu/~myl/languagelog/archives/000548.html.

Geoff Pullum is a syntax nerd, but don’t let that stop you reading “his latest post”:http://itre.cis.upenn.edu/~myl/languagelog/archives/000580.html!

But just to illustrate that even nerds can make mistakes, he’s wrong when he says that “macaques have knowledge how but not propositional knowledge”:http://itre.cis.upenn.edu/~myl/languagelog/archives/000525.html. This is not to say that Andrew Carstairs-McCarthy, the target of Geoff’s attack, is right. It’s pretty clear that the macaques in question have never said anything with propositional content. But since, as Jason Stanley and Tim Williamson argued (“Knowing How”, __Journal of Philosophy__, **98**, 2001: 411-44) all knowledge-how is propositional knowledge, it is impossible that the macaques could know how to get food without having propositional knowledge.

Stanley and Williamson argue that we can directly construct a proposition that a creature who knows how to blah knows: __thus is a way to blah__, where __thus__ picks out demonstratively the method they use to blah. I won’t repeat Stanley and Williamson’s argument here, instead I’ll run a secondary argument that can bolster their conclusion. (UPDATE: Actually, this is one of the arguments they run as well. Oops. Think of this whole post as being expository rather than setting out something that might be construed as, er, new.)

‘Knows’ is not ambiguous between (on the one hand) a relation that holds between a knower and a proposition and (on the other) a relation that holds between a knower and a type of action. If it were, (1) would be problematic, when actually it is fine.

bq. (1)   Alex knows which places sell beer this time of night and how to get to the nearest one.

‘Knows’ is ambiguous between the relation denoted in (1) and a relation that holds between two people as in (2).

bq. (2)   Alex knows Shea.

We can argue for that by noting that (3) is odd, in a way that (1) isn’t.

bq. (3)   ??Alex knows which places sell beer this time of night and Shea, who runs a couple of them.

So the standard test for whether a term is ambiguous seems to work here – it correctly returns the verdict that ‘knows’ is at least two-ways ambiguous – and it says that it isn’t ambiguous between a ‘knowing-how’ reading and a ‘knowing-that’ reading.

(I know that some languages have different words for these two meanings. That’s no proof that English ‘knows’ is ambiguous. After all, most European languages have different words for female cousin and male cousin, but that’s no proof that English __cousin__ is ambiguous between female cousin and male cousin. The ambiguity test seems to show fairly definitively the English word is not ambiguous.)

From the fact that the English word is unambiguous, we can get the conclusion that knowledge how is propositional knowledge in a couple of ways.

First, we can try a sledgehammer approach. ‘Propositional knowledge’ is a term of art. It’s not like we have these independently defined things, propositions, and we can then ask whether a macaque or whatever knows one of them. Propositional knowledge is just the relation that a creature stands in to something when a sentence “x knows y” is true, x refers to that creature, and ‘knows’ means what it means in paradigm instances of propositional knowledge claims. So saying “That macaque knows how to get food” just is to say that the macaque has a piece of propositional knowledge.

If you don’t like being that blunt, we can try more subtle methods. The best explanation for the fact that there’s no ambiguity in English at this point is that there’s no difference in meaning for English to latch onto. Since many competent speakers think there’s an ambiguity, the best explanation of that is that there’s only one relation here for the word(s) to latch onto. So we again conclude that propositional knowledge also encompasses knowledge-how.

So I think Geoff is wrong to say macaques don’t have propositional knowledge, though he’s right to say they don’t express that knowledge.

To conclude on a more conciliator note, Geoff is entirely correct about which things “should be taught in high school but aren’t”:http://itre.cis.upenn.edu/~myl/languagelog/archives/000548.html.

Saturday’s edition of the “papers blog”:http://www.brown.edu/Departments/Philosophy/Opp/ is online with new papers by Josefa Toriba (on perceptual experience), Harvey Brown and David Wallace (on the measurement problem) and Andrew Holster (on quantum measurement theory).

Soon I have to work out what’s happening with the papers blog over summer while I’m flitting around the world like a homeless philosopher. Ideally someone will set up something like the “PhilSci Archive”:http://philsci-archive.pitt.edu/ in the next three months, but on the off-chance that doesn’t happen I need to think of a plan.

The NY Times reports that my department just acquired a new fictional graduate.

bq. “Orders Come From a Talking Lion (Made of Wax)”:http://www.nytimes.com/2004/03/12/arts/television/X12HEFF.html?ex=1394514000&en=0d7f1ed9316e4d01&ei=5007&partner=USERLAND

bq. Jaye [the main character in __Wonderfalls__] lives in a tricked-out trailer, which makes her seem resourceful; she also has a degree in philosophy from Brown. And in the second episode we learn that she can write.

I would like to think that when we learn she has a degree in philosophy from Brown, we thereby learn she can write, but I’m not sufficiently down with the requisite fictional conventions to tell for sure. I do think it’s cute that saying a character is a Brown grad is a way of placing them in American fiction. I don’t know how exactly many other schools have fictional stereotypes associated with them, though obviously there are a few.

“Here”:http://www.etherfarm.com/etherblog/archives/20040224225238.html. It seems the Midwest does extend all the way to Montana!