Sunday, November 12, 2006

For the RG on modality: properties that are neither necessary nor contingent (by themselves)

About ten days ago Sònia asked in a message for the reading group of modality: "How intuitive is it that the very same property can be essential for some, and only accidental for others?" Though then I didn't give a public answer now I'll take advantage of this blog. My answer is that there are indeed examples of properties that are necessarily possessed by some objects but contingently possessed by others. Just consider the following examples:
  • (a) The (sort of) shape is an essential for the statue but not for the piece of matter from which it is made (since the statue cannot have a completely different shape.)
  • (b) 'having a body that contains (atoms of) gold' (or just:' containing gold') is a property necessarily possessed by a bar of gold, but not by a table that just contains 8 atoms of gold.
Let's say that properties that can be necessarily or contingently possessed are "modally neutral properties". These are some questions raised when we consider possible examples of such properties:
  1. Which are plausibles examples of modally neutral properties? Are there any other examples of this kind? (In Sònia's e-mail she suggests that Linsky and Zalta are committed with the view that "being abstract" would also be a modally neutral property, but that's a weird example.)
  2. Which is the class of properties for which that's true, and why this properties behave like that?
  3. If Pb is the claim that the object b has the modally neutral property P then we might fail to known a priori that (If Pb, then Necessarily Pb). So, once we know that Pb is true, what else must be known in order to conclude that Pb is necessary? (This is relevant for the question of the a priori passage in modal rationalism)
  4. Any further interesting implication of the existence of modally neutral properties?

18 comments:

Dan López de Sa said...

Interesting… Both (a) and (b) assume controversial doctrines, though—well, at least substantial: I guess it is mereological essentialism what would be regarded by most as controversial ;-)!—, but I guess this need be in the nature of the issue, no?

As to further candidate cases, I seem to remember Mario’s ‘Rigidity and Essentiality’ contending something similar for redness, in order to defend his view that rigidity for predicates is for them to be “essentialist” from the charge that ‘is red’ is clamed to be rigid by Kripke.

Oscar: Which would be the rationale for your claim 3?

Oscar Cabaco said...
This comment has been removed by a blog administrator.
Oscar Cabaco said...

You mention that (a) and (b) assume controversial doctrines, and that “it is mereological essentialism what would be regarded by most as controversial”. Well... (b) only assumes the negation of mereological essentialism, which is the thesis no object can lose or gain parts. Would you describe the negation of mereological essentialism as something controversial? Isn't mereological essentialism more controversial than its negation?

Secondly, you mention redness as a further candidate. It's true that I considered this property as a candidate of modally neutral property. The reason is that a red painting/picture is essentially red, but a red car is just contingently red. But I'm not sure about the connection between this an Mario's thesis. Prima facie I would say that what I say entails that 'is red' is not an essentialist predicate, though apparently we would like to say that it's rigid. Since I don't have access to Mario's paper from home, could you explain a little more what he says and which is the connection of his theory with this topic? In particular, could you tell us which is his definition of “essentialist predicate” and what does he say about the predicate of being red?

And about the rationale for claim 3: suppose that pointing to a table I claim that “this contains atoms of gold”. Do you know a priori that this is necessary if true? No, if the table just contains few atoms the sentence would be contingently true. If the table is made with gold (and cleverly painted to hide this fact) then the sentence is necessary. But you don't know a priori whether the table is made of gold or it just contains few atoms of gold. So this might be a case in which the sentence is necessary a posteriori but you fail to know a priori that it's necessary if true.

Dan López de Sa said...

Re the first: yeah, this is precisely what I said, sorry if I haven't been clear or something!

Re the second: honestly, I don't remember well. It might be that he thinks that 'is red' is primarily essentialist, and only applies secondarily to the ordinary objects we thought it does.

Re the third: interesting! If I got you straight here, however, the relevant feature is one that can be exhibited by sentences regardless of the issue of "modally neutral properties", no? I toss a coin, which as a matter of fact lands head. Consider the sentence:

'Either the coin actually landed head and water is H20 or it actually landed tails and my name is 'Dan.''

It looks as if this would be a aposteriori necessity, but is not apriori that it is necessary if true. Is this supposed to challenge the 2D insight that aposteriori necessary sentences are entailed by apriori necessary sentences plus aposteriori contingent ones?

Andi said...

Ok, let me have a quick go. Oscar seems to acknowledge that I can chop off the ear of a statue and still have that statue (that's what I read into the "sort of" in his post). Chop off everything exept the ear, and you presumably don't have that statue any more. If my methodology is too crude, let's chop off single atoms, one after one. Obviously there is not that one fatal atom where the statue becomes something else when I remove it. So we have a predicate that is both vague and, as you claim, essential.

Now, in the thought-experiment where I imagine being someone who can make sense of essentialism, this would give me a headache. It seems to me that whatever essential predicates should be, they shouldn't be vague.
I remember asking Dan about something like this last year at the Kripke workshop, and he said he was in the process of thinking about it. Did you come up with something to soothe my headache?

ps I'm aware that this could hardly be an innovative point; as I hinted at, I haven't ventured far into this jungle yet, as I lack the basic intuitions to guide me through. Thanks for your patience :)

Dan López de Sa said...

Andi,

Sorry, I don't remember exactly what you asked me then.

One think I have been thinking concerns predicates for natural kinds such as ‘is a cat’—i.e., involving what people probably regard as “essentialist” predicates—being possibly vague. We had some very interesting discussion at the GAF Blog, on this issue, in Spanish though.

In brief, I do think they are, for reasons analogous to your point concerning statues: they seem to be Sorites-susceptible—and this is sufficient for vagueness.

Is this related to what we talked about?

Oscar Cabaco said...

Well... first I'll address Dan's questions:

About the second issue... apparently he says that 'is red' is not an essentialist predicate, just as I said (see p. 254 ) But I still fail to the the conection of what we have been talking about and Mario's theory. Maybe we should stop discussing this unless someone clarifies the relevance (if any) of what we are talking about to the question of the rigidity for predicates.

With respect the last point: yes, you are right, though you give an unnecessarily complicate example. Williamson gave a simpler example that can be found in this paper by Edgington (p. 11). Here I quote the relevant passage:

Timothy Williamson gave (in discussion) a very simple example which shows that one cannot hold in general that, if a proposition is metaphysically necessary, it is knowable a priori that if it is true it is necessarily true. Just consider the disjunction of an a posteriori identity statement, a=b; and an a posteriori, clearly contingent statement, c is F: We can’t know a priori whether the disjunction is true. Nor can we know a priori that, if it’s true, it’s necessarily true; for it might be true in virtue of the first disjunct, or in virtue of the second.

This doesn't challenge the two-dimensionalist insight you mention, and possibly this is just indirectly related to the question about the a priori passage (I exaggerated in the original post so to move people to post too.)

The claim about the a priori passage can be summarized as the claim that any necessary statement (like 'Water is H2O') follows (a priori) from contingent a posteriori truths (like 'the watery stuff is H2O'.) For me, one related idea that directly addresses questions in modal epistemology is the claim that any modal statement (any statement involving explicitly or implicitly modal operators) follows a priori from non modal ones. So, for instance, we can conclude that 'necessarily a=b' just from the premise that 'a=b'. This would explain how/why we know modal truths at all in spite of the fact that we have only access to the character of the actual world. But modal statements with modally neutral properties would show that the path from the non modal to the modal is not so direct and simple (for some necessary statements we fail to know that the are necessary if true.)

I don't know if this makes sense, or whether you would think that these are two completely unrelated ideas. Anyway I do not think that this poses any problem whatsoever for 2D (so don't feel threatened!)

* * * * * * * * * *

And now, about Andi's question:

Could you explain why essential predicates (or better: essential properties) cannot be vague? At first this strikes me as implausible, because there seems to be plenty of bona fide examples of vague essentialist predicates (for example: being human.)

I can can imagine a philosopher arguing as follows. Vagueness is a linguistic or epistemological phenomenon, not a metaphysical one. So, in the real world there are no such things as vague objects or vague properties. A fortiori, there are no vague essential properties. But this doesn't seem to be your reasoning because (i) you just talk about essential properties, not about properties in general, and (ii) if that was the problem it would have an straightforward answer: though that might be true about properties, we have been talking about predicates (linguistic items).

So... could you develop further your question?

Andi said...

Never mind last year, my memories are hazy as well :)
But: Don't you share my gut-feeling that vague essential predicates are odd? That the grumpy 'could not fail to possess!' and the libertine 'maybe yes, maybe no, who knows?' could not be very happy together?

Andi said...

now if that's not proof that this is a living blog... oscar and me writing comments to the same post at the same time, and that time being 2am!
I think I was talking about predicates as well, but I will look at this again after a good round of sleep. Maybe it all just comes down to the fact that I find essential properties/predicates/what-have-you strange...

Oscar Cabaco said...

Andi:
I would see a problem if someone tells me both that 'b is essentially F' and that 'b is a borderline case of been F'. Then I would have the same "gut-feelings": b could not fail to possess F, but... if b is a borderline case it's unclear whether b is F at all!!

But, what's the problem with objects determinately having vague properties that are essentially possessed by them? (that would be the case if b is essentially F, F is a vague property, and F is not just a borderline case of being F.)

Of course, there is a related problem concerning transworld identity, vague essential properties and borderline cases: you can exemplify Salmon's four worlds paradox with any essential vague property. but this is a whole new discussion.

Dan López de Sa said...

Yeah, this is indeed the longest thread ever! Should we submit it for the next Philosophers’ Carnival ;-)?

In parts.

(A) Mario
Ok, I take the point that the relevance of my bringing this issue here is at best disputable, and I am happy to postpone discussion of it for another occasion.

(B) Williamson
Didn’t know about it, but it’s quite cool to come out with an example that is similar to his :-)! Mine was:

(1) Either the coin actually landed head and water is H2O or the coin actually landed tails and my name is ‘Dan.’

His scheme gives in effect a simplification of (1), just removing the coin stuff conjuncts:

(2) Water is H2O or my name is ‘Dan.’

One nice feature of mine, though, is that one knows apriori that (1) is true, it’s only that one doesn’t know apriori whether it is true necessarily. (If you are picky about it, just replace the first conjuct of the second disjunct by ‘it is not the case that the coin actually landed head’.) Not so, as Edgington observes, for (2).

What would people think of the following—where Harry is borderline wrt ‘is bald’?

(3) Harry is actually bald or Harry is not bald.

Is this one apriori knowable truth such that it is indeterminate whether it is necessary?

(C) Goliath

I am afraid I still don’t see the tension between a predicate being essentialist and vague. Say that a predicate is essentialist iff if it definitely applies to something, necessarily it is not false that it does. Consider Andi’s sorites removing tiny bits of statue Goliath—or mine going from cat Tibbles to pet-robot Tama, replacing organs by prostheses. Presumably, those who regard ‘is a statue’ or ‘is a cat’ as essentialist would hold that, in the borderline area, it is indeterminate whether Goliath and Tibbles still exist. (For what it is worth, the cat case seems to me much more palatable: statues just seem phases of lumps of matter, no?) Hence a way to combine essentiality and vagueness, what do you think? (This might be close to Oscar’s response, but without invoking the obscure doctrine of vagueness in rebus.)

Dan López de Sa said...

I said: "One nice feature of mine, though, is that one knows apriori that (1) is true, it’s only that one doesn’t know apriori whether it is true necessarily."

That's bull-shit, of course, sorry. Nothing to be said for (1), viva (2)!!

Andi said...

Dan, let me get clearer about the cat. In it's sad life (this, by the way, is the second time this week that a LOGOSer suggests cutting up cats...) three events seem to occur:

1. Tibbles ceases to be essentially a cat.
2. Tibbles ceases to be a cat.
3. Tibbles ceases to exist.

Now, what would you say in which order these events take place? And if any of these events occur simultaneously, do they do so necessarily?

Oscar: Wouldn't you say that if "to be human" is an essential property than anything that is human is essentially so? If not, do those humans who aren't necessarily so have one essential property less than the others, or do they have another one instead (eg, "to be something between a human and an ape")?

And what do you think of Dan's
"Say that a predicate is essentialist iff if it definitely applies to something, necessarily it is not false that it does" ?
This kind of sounds weaker than what I understood from you and Manolo ("not false" instead of "true").

Lots of questions, no answers... sorry about that, I'll try to write something more committed later on the weekend.

Dan López de Sa said...

I was suggesting on behalf of the "essentialist" wrt 'is a cat' that 1-3 are actually different sides of one and the same change, and hence they occur necessarily simultaneously.

BTW, the cat-mind might survive all the same, no? Wasn't it the whole point of RoboCop?

Andi said...

Good one, that Robo-Cop remark! Now Oscar seems to be in doubt about the essentiallity of "being human" (at least he was at the GRG-dinner last night)!
OK, back to the cat. So you hold that this is not one of the cases when we at some time have a cat that is not essentially a cat (always assuming "being a cat" is an essential property). So could you think of a case in which this happens? I mean, if there is no such case, then what is the point of concentrating on the non-borderline cases of "being a cat"?
And if there is no point in that maneuver, then we are back to the headache that Oscar admitted to share, no?
Once again, almost all my sentences end with a questionmark, so let me end with a proposal: how about we just pretend we define "essentially F" as "F and not a borderline case of F"; that's not what you're after, I know, but what could you do with the first that the second couldn't be used for? (damn, questionmark again!)

Dan López de Sa said...

Andi, I'm not sure about following you here, sorry. I thought the original question was whether there could be a predicate that is both essentialist and vague, and I think we do have a model now of how this might be.

(It is important to distinguish sharply here between predicates and properties, and essentialist predicates and essential properties, as Oscar suggested. On the view that I would favor, there is a number of cat-candidate properties, each of which having an equal claim to be the (essential, we might suppose) propery that predicate 'is a cat' signifies, and it is always determinate whether a (precisely specified) object exemplifies them. Borderline cases with respect to 'is a cat' do determinately expemplify some but not all of these (essential, we might suppose) properties.)

As to the last suggestion, it is not clear that 'it is F but borderline F' is assertible, and this might make 'it is F but not borderline F' odd. In any case, there are clear cases wrt 'is red', but this does not make, pace Mario, the predicate an essentialist one, no?

Oscar Cabaco said...

First I’ll try to answer an earlier question raised by Andi “Wouldn't you say that if "to be human" is an essential property than anything that is human is essentially so?

Indeed I think that there are properties that are always essential properties of those objects that instantiate them. Examples? Being identical to Dan, being an inanimate object made of wood, being a number, etc. For other properties, like being human, it’s controversial where will they fall in this classification. If you think that the person is identical to the animal, then we are essentially humans. In contrast, if you instead think that the person is still alive when you transplant her brain into a machine, then we are not essentially humans (assuming that a human brain doesn’t make a human.)

And now, let’s address the other question. As you point out, we don’t want to define “is essentially F” in terms of “F and not a borderline case of F”. And then you asked “but what could you do with the first that the second couldn't be used for?” Well... that’s like asking why do we need to distinguish between essential and contingent properties (and that's too general for a question!). But... Are you sure you don’t you have any essentialist intuition whatsoever?

After some beers I remember having told you that believing in physical necessity (necessity according the natural/physical laws) already counts as believing in metaphysical necessity (since there are already philosophers, like Edgington, Shoemaker, A. Bird, Swoyer, Fales, etc., who defend that natural laws are necessary.) Isn’t that sufficient for you (assuming that you accept, as I think you do, physical necessities)? What else are you asking for to accept a non-linguistic necessary/contingent distinction? Would you finally be satisfied if I tell you, for instance, that persistence conditions also allow us to maintain that distinction?

The problem with very general sorts of scepticism is that you don’t know how to start dealing with them...

Dan López de Sa said...

It has happened again, two of us commenting simultaneously on the same post :-)! Good sign…

Second thoughts about my last comment: perhaps there is after all a problem with supposing that each of the candidates is indeed an essential property, assuming the view of vagueness as semantic indecision I was mentioning. But still the predicate can be both essentialist and vague, no?

I agree with Oscar that it is surprising that people normally give 'is a human' as an easy case of an essentialist predicate. As a predicate of bodies or organism might be OK, but applying to persons is quite controversial, as illustrated by RoboCop, Star Trek, and so many others!