Tuesday, November 22, 2005

Bunny wants a unicorn. What'll Kripke do?

Her desire for unusual pets extending beyond chickens and seahorses, Bunny McIntosh wants scientists to make a unicorn for her. This makes a problem faced by the Kripkean semantics for natural kind terms even more dramatic.

According to Kripke, natural kind terms like chemical formulas and species names refer to some object at the beginning of the causal chain that started when the word was first used. Anything that has the same essential features as that object is a member of the kind. Since unicorns don't actually exist and nonexistent objects can't cause anything, no proper referent stands at the beginning of the causal chain that underlies the course of the word "unicorn" through the world. Unicorns don't end up merely being nonexistent, they come out to be impossible objects, since there are no such things as the essential features of unicorns. Unless there actually were unicorns at some point in the past, and they stand at the beginning of our unicorn-talk, unicorns cannot exist. While one could create a horse with naturally growing horn, perhaps with wings, a kindly disposition, great intelligence, and an affinity for virgins, it still wouldn't count as a unicorn -- nothing could.

Now here's what Bunny says:


I want a pet unicorn. Maybe one with wings, even if it can't really use them or anything. It doesn't even have to be pink, necessarily. I'd settle for white. Or black.

Avian flu schmavian flu. Get to work on this unicorn thing. That way, when kids are in high school doing their biology assignments and they ask "when am I ever going to need to know this?" The teachers can say "I don't know if you've heard, but scientists have recently been working on genetically modifying a horse to have a horn. That's right, they're making Unicorns, and they're getting paid millions of billions of dollars for them. If you want to learn how to make your own from the privacy of your own state funded lab, you're going to need to shut up and pay attention."


She not only wants a pet unicorn, but believes that scientists can use genetic modification to generate one. So she believes that unicorns are not only possible objects, but possible in the future of this world. I don't think she's speaking loosely here -- it seems that she considers the creature that would be generated a genuine unicorn. (Her desire for one seems hard to explain otherwise. Would one wish in this way for a unicorn-knockoff?)

Here I'd push for a rigidified descriptivist cluster theory on which the nonexistence of a proper referent doesn't make an object impossible. If the cluster element that involves the original causal connection goes unsatisfied, having the right descriptive features is enough to make the species term refer properly to some possible object.

10 comments:

Dennis said...

From a happily categorically-inclined mathematical perspective, I'm quite pleased to support terms defined by descriptive features only. There are many structures indeed that are defined by what we call "universal properties" -- a sufficiently thorough description to characterize something as certainly as "water" is "H_2O." These descriptions are often enough to characterize something sufficiently well to work with under the assumption that it exists -- actually making the requisite object isn't required (though of course it is if you want to prove anything non-tautologous). For instance, I can safely conclude that a unicorn could poke me with a sharp appendage and that a horse couldn't without actually knowing that either exist. I might even say that an AIDS vaccine is a liquid which, when administered through syringe, prevents the later acquisition of AIDS, as even though there currently is no such thing, were one to come into existence it would be an AIDS vaccine. So yeah, we love descriptive definitions.

Julian Elson said...

In the medeival bestiary that I read, the description of "unicorn" listed "rhinoceros" as a synonym. So, maybe the original conception of the unicorn is based on heavily (very heavily) distorted reports of rhinos.

Anyway, if Bunny gets a unicorn, I'm getting a succubus. As for what Japanese tentacle-oriented anime pr0n fans and furries want to do with genetic engineering, it's best left unimagined.

Neil Sinhababu said...

Math is actually regarded as one of the best places for descriptivism, since many philosophers think that the objects of mathematics are abstracta, which don't do any causal work. So causal theories of reference don't really work there.

Natural kind terms and especially proper names are regarded as the places where the causal theory is especially strong.

Neil Sinhababu said...

Julian, the rhino thing makes me wonder how that stuff about virgins got going!

I'd want to say to a Kripkean: if you want to say that rhinos aren't unicorns, you probably have to rely on the descriptive content implicit in the term "unicorn," which rules out rhinos even if they're at the beginning of the causal story. So why not go all the way with the cluster-theory and say that satisfying this descriptive content is one way of being a unicorn?

dadahead said...

Just one of the many problems with referential semantics.

The meaning of a word is its use in language!

Aidan said...

The unicorn myth goes back way before the first rhinos were spotted. Marco Polo recounts first finding unicorns on Java, but he noted the ones he saw were somewhat crappy examples of the species having 'the hair of the buffalo' and feet like those of an elephant. He continued:

'It has one horn in the middle of the forehead very thick and large and black. And I tell you that it does no harm to men and beasts with its horn, but only with the tongue and knee, for on its tongue it has very long spines….
It has the top of the head made like a wild boar…It is a very ugly beast to see and unclean. And they are not so as we here say and describe, who say that it lets itself be caught in the lap by a virgin girl: but I tell you that it is quite the contrary of that which we believe that it was.'

By the way Neil, I doubt it's so clear that a causal theory of names is so obviously inappropriate for abstract (and so causally inert) objects. One could hold that what determines that the reference of a proper name, 'Neil' say, on some particular use is yourself is some appropriate causal chain going back to an initial baptism. Do we need the causal chain to reach back to the object itself? As Kripke presents it, yes, but Sainsbury has argued that's just a bias in Kripke.

In any case, there's no reason to think that the failure of the causal theory would push us towards descriptivism - there are plenty of non-causal theories which aren't descriptivist theories.

Jonathan said...

Yeah, boo Kripke!

bunny mcintosh said...

since there are no such things as the essential features of unicorns. Unless there actually were unicorns at some point in the past, and they stand at the beginning of our unicorn-talk, unicorns cannot exist. While one could create a horse with naturally growing horn, perhaps with wings, a kindly disposition, great intelligence, and an affinity for virgins, it still wouldn't count as a unicorn -- nothing could.


Mine doesn't necessarily need an affinity for virgins. In fact, so I hope that it doesn't care that the flower of my virginity has been long since trampled upon.

And, just so you know, right now I am eating a Grapple. It is a combination of an apple and a grape. Basically, it's a grape flavored apple, and it's delicious. From now on, I want to invent a new sort of Kosher diet for myself. I only want to eat the following things:

1) food that is a manifestation of the power of human intelligence and that has been severely and creatively genetically modified.

2) predators

3) fixins

4) anything from the sea


xox,
bunny

Neil Sinhababu said...

If I were to reinvent the unicorn mythos, there would different sorts of unicorns, so that girls of all kinds would have unicorns that loved them. The prospect of a unicorn that would eat only from the hands of girls who had participated in gangbangs is amusing to me.

bunny mcintosh said...

Unicorns only like pervs.