Tuesday, September 23, 2014

Putnam / Cats / Robots From Mars

Here are two versions of a thought-experiment suggested by Putnam. I can't remember whether he suggests both versions, or just one of them:

1.
We discover that every feline that has ever existed has been a robot planted by Martian scientists to observe and record activity on earth. No such creature has ever been an animal/biological. It's always been the robot thing.
Under such conditions, which of the following is true:
(CR) Cats are robots from Mars.
(NC) There are no cats. (There are only robots from Mars.)
?

2.
We discover that everything is as we thought it was until, say, 50 years ago, when the Martian robot program was instituted. For about 50 years, then, it's been the robot thing.
Under such conditions, which of above options is true?

6 Comments:

Blogger Protagoras said...

I find it so hard to believe that people think there's just a right answer to questions like this. If it happened, a choice would be made; one choice might be more likely than the other, but neither would be mandatory. There would probably be momentum behind whichever proposal was made first. In some cases, one choice might be so much more practical than the other as to be essentially inevitable, but I don't understand what tempts people to the view that it comes down to anything but choices.

11:06 AM  
Blogger Winston Smith said...

Thanks Aaron. I'll wait until some others (I hope) chime in before throwing out my $0.02 on the matter.

11:13 AM  
Anonymous John Plato said...

Blowing the dust off a mental copy of my old Philosophy of Mind textbook, I would say that, philosophically, statements can have different varieties of meaning, and that the difference between extensional meaning vs. intensional meaning makes all the difference here.

The phrase "There are no cats," would be true in an extensional sense -- there are no existing members of the set of creatures we once knew as cats -- but false in the intensional sense, because the idea of cats will always remain, even if such creatures have never existed.

7:51 PM  
Anonymous Hypatia said...

It's not clear what you mean by "under such conditions", but here are some responses:

Scenario 1- All things ever called 'cats' were robots from Mars.

Firstly, let's distinguish between the hypothetical world, which I'll call 'M-world', and the actual world, which I'll call 'the actual world'.

In both cases, cats are not robots. The word 'cats' as used by me and by you in the question refers to the living, flesh-based animals we keep as pets in the actual world (since neither of us is an inhabitant of M-world). We can truly say, using our word 'cat', that cats are not robots and that there are no cats in M-world.

A different question is whether 'cats are robots' expresses a truth in the language spoken by M-world inhabitants. Presumably they learned the word 'cats' through many different interactions with the robot-cats (which are not cats, btw). Thus, their word 'cats' refers to the robot cats. So, to answer the new question: yes, 'cats are robots' does express a truth given what the words means in that world.

Scenario 2 - about 50 years ago Martians replaced our cats with cat-like robots without our realizing it.

This situation is a bit trickier since given any theory of reference you'll need to make a choice about at which point the reference of a word changes. I think it's fair enough to say that this might be a problem for vagueness. That is, it's not clear how long a word needs to be used incorrectly in associating with a demonstrative picking out an object that is not the referent before that object becomes the referent. Moreover, I don't think the decision here matters too much provided we're willing to say that at some point the reference does, in fact, change (see Evan's style Madagascar cases for reasons to think that the reference does, in fact, change).

So, returning to the two relevant questions, in this scenario, the first answer is the same: cats are not robots. The answer to the second question (whether 'cats are robots' expresses a truth in the language spoken by inhabitants of that world) is: it depends. If you think 50 years of looking at robot-cats and calling them 'cats' is enough to change the reference, then the answer is yes. If you think not, then the answer is no.

Crucially, none of these answers depend on the inhabitants of these scenarios making a choice about how to use language. So the idea that it just depends on what the people in those situations choose to do is misguided. In fact, the people in that world need not even know that the objects they call 'cats' are in fact robots. We can still come to perfectly legitimate conclusion about the situations based on a comprehensive theory of reference.

5:09 AM  
Anonymous Anonymous said...

It's not clear what you mean by "under such conditions", but here are some responses:

Scenario 1- All things ever called 'cats' were robots from Mars.

Firstly, let's distinguish between the hypothetical world, which I'll call 'M-world', and the actual world, which I'll call 'the actual world'.

In both cases, cats are not robots. The word 'cats' as used by me and by you in the question refers to the living, flesh-based animals we keep as pets in the actual world (since neither of us is an inhabitant of M-world). We can truly say, using our word 'cat', that cats are not robots and that there are no cats in M-world.

A different question is whether 'cats are robots' expresses a truth in the language spoken by M-world inhabitants. Presumably they learned the word 'cats' through many different interactions with the robot-cats (which are not cats, btw). Thus, their word 'cats' refers to the robot cats. So, to answer the new question: yes, 'cats are robots' does express a truth given what the words means in that world.

Scenario 2 - about 50 years ago Martians replaced our cats with cat-like robots without our realizing it.

This situation is a bit trickier since given any theory of reference you'll need to make a choice about at which point the reference of a word changes. I think it's fair enough to say that this might be a problem for vagueness. That is, it's not clear how long a word needs to be used incorrectly in associating with a demonstrative picking out an object that is not the referent before that object becomes the referent. Moreover, I don't think the decision here matters too much provided we're willing to say that at some point the reference does, in fact, change (see Evan's style Madagascar cases for reasons to think that the reference does, in fact, change).

So, returning to the two relevant questions, in this scenario, the first answer is the same: cats are not robots. The answer to the second question (whether 'cats are robots' expresses a truth in the language spoken by inhabitants of that world) is: it depends. If you think 50 years of looking at robot-cats and calling them 'cats' is enough to change the reference, then the answer is yes. If you think not, then the answer is no.

Crucially, none of these answers depend on the inhabitants of these scenarios making a choice about how to use language. So the idea that it just depends on what the people in those situations choose to do is misguided. In fact, the people in that world need not even know that the objects they call 'cats' are in fact robots. We can still come to perfectly legitimate conclusion about the situations based on a comprehensive theory of reference.

5:10 AM  
Blogger Winston Smith said...

H,

No, if you read again carefully you'll see that it's clear what's meant by "under such conditions."

Thanks for the comments, though this was really a post aimed at people who don't know their way around this problem. A lot of us know our way around this stuff well, and so it's old hat. I posted this to get people to think about the problem before looking at a more politically contentious one where such questions are relevant...though I can't for the life of me remember what it was now...

I agree, of course, that none of the answers depends on the inhabitants making a choice, but I didn't suggest otherwise.

Also, actually 'cats are not robots' is not one of the options with respect to the first question. Maybe you mean "there are not cats."

As to whether the question about truth can be prised away from the question about facts...well, I've gone back and forth on that, but haven't thought seriously about it in a long time. But I didn't want to raise that issue here, which is why I left it out.

Thanks for the comment!

12:28 PM  

Post a Comment

Subscribe to Post Comments [Atom]

<< Home