Sunday, January 25, 2009

Philosophical Questions

I think that Williamson's arguments for the conclusion that philosophical questions are not always about (either implicitly or explicitly) language are (for the most part) sound. It seems clear to me that the original question:

Was mars always either dry or not dry?

is not question about language or concepts. However, I do think that this is a question that is best answered by thinking about our language or concepts. In fact, here are three theses, each of which seems rather plausible to me, about the nature of philosophical questions:

T1: Philosophical questions are questions that can be answered only by analyzing our concepts or language (if they can be answered at all).

T2: Philosophical questions are questions that can be best answered by analyzing our concepts or language (if they can be answered at all).

T3: Philosophical questions are questions that are appropriately answered by analyzing our concepts or language (if they can be answered at all).

Williamson doesn't really address any of these theses. Although, some of the things he says may suggest that some of these theses are true or that some of them are false. However, each of these theses seems plausible to me and might be a way of holding onto the view that (in some sense) philosophy is primarily conceptual. Perhaps a defense of one of these theses is put philosophy right back into the concepual turn.

That being said, although I find these theses plausible, I do have worries about them. One worry is that some philosophical questions can seem to be answered by methods that are just as easily done from the armchair as philosophical analysis, but which are not methods of philosophical analysis. For example, Lewis attempts to answer the philosophical question "Are there any possible worlds?" by weighing the pros and cons of the various answers to that question against one another. It seems that if the philosophical question "Are there possible worlds?" can be answered by Lewis' method, then T1 is be false. Moreover, if there is no better way to answer that question than by Lewis' method, then T2 is false. Finally, if it is perfectly appropriate to answer that question by Lewis' method, then T3 is false.

As I see it, one can try to show how a Lewisian kind of argument can be turned into a argument involving conceptual analysis or one can give up on theses T1-T3. I do not know what the best response to this kind of worry is. In any case, it seems that these theses might be appropriate ways of characterizing the thesis that philosophy is concpetual. Moreover, even though the lewis example provides a decent case against theses T1-T3, it is not clear to me that these theses are false.


Blogger rock* said...

Hi Joshua,

T1-T3 all speak of "analyzing" our concepts or language. But it's not at all clear to me what you mean by that. What is it to analyze our concepts? What is it to analyze our language?

Also, do you think that, for instance, the questions that arise when discussing the paradox of undetached parts are counterexamples to T1-T3?

6:03 AM  
Blogger Christian said...


I'm inclined to think there is an ambiguity here. Concepts are vehicles through which objects, properties, and states of affairs are grasped or considered. But the object of analysis is not the vehicle. The objects of philosophical analysis are the properties et al.

For example, when I ask whether doing x is intrinsically morally equivalent to allowing x, I am asking a question about properties, and my answer adverts to them. Why isn't this true for much of philsophy? We are doing metaphysical analysis and grasp the objects of our concern through concepts.

5:33 PM  
Blogger Joshua said...


I am not sure what I meant by "analyzing concepts or language". My thought was that there might be something like a linguistic analysis of triangular into constituent concepts of being three in number and having angles or something like that. I might have just been thinking of typical analyticity (in which case Williamson seems to be addressing my suggestion in the next couple of chapters).


I was thinking of concepts as the vehicles through which objects (such as properties) are grasped. Generally I tend to reserve the word 'concept' for the vehicle through which a property is grasped and 'property' or 'feature' for the things that are grasped.

Christian and Greg,

I have a general worry about analysis that relates to both of your comments. Here is a dilemma:

1. If there is such a thing as philosophical analysis, then either the object of an analysis is a concept or it is a properties.
2. The object of an analysis is not a concept.
3. The object of analysis is not a property.
4. So, there is no such thing as philosophical analysis.

The key premises in this argument are (2) and (3). Here is why someone might believe (2): Most concepts are conceptually simple. For example, the concept vixen doesn't seem to have any conceptual proper parts. It is very different from the concept superhero which seems to have the conceptual parts super and hero. But, if the concept vixen is conceptually simple, then it cannot be analyzed into conceptual parts. But, if the concept vixen, cannot be analyzed into conceptual parts then (since that concept can be philosophically analyzed if any can) the objects of philosophical analysis are not concepts.

But, someone might believe (3) because of the following. If a property is subject to philosophical analysis, then we should be able to tell whether a property is complex or not. But, we cannot tell whether a property is complex or not because the linguistic or conceptual structure of the words or properties that stand for such properties does not reflect the structure of the property. So, the objects of philosophical analysis are not properties.

Now, it seems clear to me that Christian is right that the objects of philosophical analysis (if there are any) are properties rather than concepts. That means that (3) must be false. But, if (3) is false, then our method of finding the structure of properties (or method of philosophical analysis) must be a kind of indirect method. We do not see through the clear light of reason that the property of being a vixen is composed of (in some sense) or importantly related to other properties. Rather we discover the structure through indirect means.

I guess I am inclined to think that thought experiments are rather rudimentary indirect means of discovering the structure of properties. We can also discover structures by performing well constructed language use experiments. But, either way, the method of philosophical analysis turns out to be pretty different from what philosophers have traditionally thought. Indirect discoveries (which we might be able to make empirically) do not seem to be the kinds of discoveries that we traditionally think of when we think of philosophical analysis.

2:52 PM  

Post a Comment

<< Home