John West wrote:

John's fine, Tim:

However, I think the premise that determinables exist is not so difficult to establish. Crimson <I>is</I> a color and <I>is</a> a shade of red. And if it is true that every shade of red is red, then crimson is also red. But then why is crimson said to be a shade <I>of</I> red?

Well, I think the statements “Every shade of red is red”, and “Crimson is red” are true. I don't, however, think that predicates stand in one-to-one relations with properties.

First, I don't think that properties stand in one-to-one relations with predicates. For instance, there are almost certainly properties for which human language has no terms. Hence, these properties can't stand in one-to-one relations with predicates.

Going the other way. I don't think that predicates stand in one-to-one relations with properties. If, for instance, every disposition predicate stands in a one-to-one relation with a disposition, we can quickly generate an infinity of dispositions of a being by replacing the dash in "is disposed to—":

is disposed to dissolve
is disposed to dissolve when placed in water
is disposed to dissolve when placed in water on Sundays or Mondays,
is disposed to dissolve when placed in water and is such that snow is white,
etc.

In other words, word-making is not world-making. So, I would say that statements like “Every shade of red is red” and “Crimson is red” are true in virtue of the totality of (things that are) shades of red, but that there are only specific shades of red among the furniture of the world.

This is a standard position among contemporary metaphysicians, including the powers theorists. One cool result is that the problem of attributes turns out to have been based on a simple mistake of conflating predicates with properties. This, I think, is part of what Aquinas was getting at when he applied the doctrine of analogy to it.

Hello John, and thank you for your exceptionally clear response.

I think I have two difficulties with what you said above. The first, as to your example, is that I'm not seeing why further qualifying a disposition would or should be problematic, or raise difficulties, so long as the qualifications are implicitly contained. So, e.g., X has a disposition to dissolve; X has a disposition to dissolve in water (or perhaps X just is a disposition, say, to dissolve; and this, in water; and this, within a certain temperature range, etc.).

Second, as regards the relation to predicates and properties, I think this would involve difficulties insofar as we are now predicating this of properties themselves or of predication itself. Something tells me this might be problematic.

Edit: Added: In regards to the first difficulty regarding qualifications of (e.g.) disposition prediates, I think there would necessarily be a limit, so long as we are willing to grant excluding what is or might be accidental when qualifying the disposition that is being predicated.

Last edited by Timocrates (3/07/2016 8:02 pm)

John West wrote:

Timocrates wrote:

The first, as to your example, is that I don't believe further qualifying a disposition would be problematic, so long as the qualifications are implicitly contained. So, e.g., X has a disposition to dissolve; X has a disposition to dissolve in water (or perhaps X just is a disposition, say, to dissolve; and this, in water; and this, within a certain temperature range, etc.).

I think you're right. The upshot of this reply, however, is that predicates don't stand in a one-to-one correspondence with dispositions. That's all the “is disposed to—” argument seeks to prove.

The reply also fits well with the view that all dispositions are completely specific, determinate dispositions. That is, once we unpack all the implicit qualifications in a disposition, what's left is a determinate disposition.

Second, as regards the relation to predicates and properties, I think this would involve difficulties insofar as we are now predicating this of properties themselves or of predication itself. Something tells me this might be problematic.

I can't see what that would be. The predicates still apply in virtue of relevant constituents of the world (ie. the crimson property). So long as the truthmakers make the statements true in every possible world in which those truthmakers exist, there should be no problem[1].


[1]No ontological commitment to such worlds is intended.

Thank you John.

I still find the not one-to-one emphasis problematic. A cow is an animal. It is obviously true that a cow is not a planet (not a property of cow); however, I am finding a difficulty in the idea that a true predication is not a one-to-one with a property or truthmaker (regardless of qualification or specification). If animals do not exist, then neither do cows; and if cows are real, then so are animals. Indeed, if a cow is real, then animal is real, too (I would think). To be sure, identity predication is problematic because presumably identity is not a property of the thing being identified (since the idea of property means 'belonging to' something; but in identity, this impossible - how can the identity of the thing be itself a belonging to the identity of the thing? We'd have to go on forever). Hence, determinables must be real for determinates to be real.

If I say, X is a ship, then all that it is contained in ship will be contained in X (assuming saying X is a ship is true). It would not be wrong to further say that X is a warship or a merchant ship or whatever.

Thank you John for your patience in explaining this to me. There is certainly a lot here to consider.

Dennis wrote:

Incessable, while I don't mean to speak for John (since I'm not sure how he would react) and I know you have many other things to answer and say. But I think that's exactly the point, what genuine metaphysical problem is the addition/subtraction of googol (10^100) atoms bringing to the table? 

With what purpose do Phenomenalists run their ontology? To reduce it? This seems' wrongheaded to begin with.

To repeat something I said earlier, the interesting thing about the principle of parsimony: just about everyone agrees that the principle is valid in the abstract, but when we start discussing specific versions of it, that initial agreement evaporates.

One line of argument goes as follows: Parsimony can be expressed in mathematical terms through Kolmogorov complexity. (i.e. if we produce a precise description of the universe as a finite string of integers, e.g. record the position and other attributes of every subatomic particle, what is the length of the shortest computer program which generates that exact string of numbers as output – a program for an idealised computer which has arbitrarily large (yet finite) storage space and time in which to execute the program.) If we adopt that mathematical formalisation of complexity, then the principle of parsimony becomes Solomonoff induction (i.e. we prefer the theory with the lower Kolmogorov complexity). Now, in a chaotic universe, especially one with any degree of irreducible quantum indeterminism (or even a purely deterministic universe yet with incomputable determinism), a googol extra atoms would almost surely lead to a substantial increase in Kolmogorov complexity of that universe, hence in this formulation of parsimony we should prefer the theory with a googol less atoms.

(Of course, being incomputable, we cannot actually calculate the Kolmogorov complexity of any particular theory; but, in some cases, even if we cannot know the values of K(A) and K(B), there can be good arguments that it is very likely that K(A) is significantly greater than K(B).)

Now, you can resist the urge to mathematically formalise the principle of parsimony. But, mathematical formalisation has served us so well in other areas of human thought, it seems entirely appropriate to try that same approach here too. And, if we are going to try to mathematically formalise that principle, I don't see any major competitors to Solomonoff's approach.

Last edited by Incessable (3/11/2016 8:06 pm)

Dennis, if I understand what you are saying, you are saying that the principle of parsimony serves a particular purpose, and we should therefore adopt the version of it which best serves that purpose.

Now, I don't necessarily disagree with you there. But, I'm not entirely sure what purpose it serves. I guess I would say that, like all valid principles of reasoning, it brings us closer to truth, but that's not very specific; indeed, it fails to state how it brings us to the truth, or why we should be believe that it actually does. I guess, the best argument for adopting some version of the principle of parsimony, is that its acceptance has arguably has contributed to the success of the natural sciences in discovering truths (and clearly useful truths) about the world. However, that doesn't tell us by itself which version of it we ought adopt – although, maybe we should adopt whichever version (we think) is used in the natural sciences, or whichever version (we think) best serves the natural science's needs? But, some will object, that the best version of the principle of parsimony to choose for the natural sciences may not be the same as the best version to choose for metaphysics; maybe that is right, but it's really unclear to me how we then determine which version is the best to use for metaphysics.

In an earlier post, I drew an analogy between the principle of parsimony and the prohibition of murder; just about everyone agrees that the former is an important rule of rationality, and the latter is an important rule of ethics; but as soon as we turn to specific versions of those rules, the initial agreement disappears. To take the analogy further, we can note that people can agree on some ethical rule yet vastly disagree on the reasons behind it; for example, a utilitarian and a divine command theorist might both agree that "murder is wrong", yet give widely differing accounts of why murder is wrong. Further than that, many people agree that murder is wrong, yet are unsure of why it is wrong. In the same way, people can agree that the principle of parsimony is a valid principle of reasoning without agreeing on why it is a valid principle of reasoning, i.e. what reasons justify it; a person can even accept parsimony as a valid principle of reasoning yet be unable to give an cogent explanation of why they do so. To at least some extent, I am one of those people.