Metaphysics Interlude

The word "theory" is frequently used among scientists. I'm sure you've heard of the "Theory of General Relativity" or perhaps the "Germ Theory of Disease" or "Evolutionary Theory." Scientists propose theories often to do two things: (i) Explain existing data; (ii) Make predictions. General Relativity was a unified explanation of much apparently incompatible data; General Relativity makes predictions about, say, how light moves through space. Similarly, the Germ Theory of Disease provides an explanation for a diseases spreading through populations, among other things; the Germ Theory of Disease predicts, say, that if you let someone ill with the flu cough on you, then you'll likely get the flu too.

Sometimes scientists propose theories to explain data that lack any obvious empirical predictions. Murray Gell-Mann suggested the existence of quarks as underwriting the existence of electrons, protons, etc. Other physicists were not exactly kind to this proposal. They criticized Gell-Mann for not providing a theory with empirically testable consequences. Some physicists went as far as to find way to test for the existence of quarks. Turns out, there are good reasons to think there are such things on empirical grounds. Gell-Mann's pet theory was empirically confirmed, and he eventually was awarded many honors for the discovery.

The important thing to note about the preceding history lesson is that scientists propose theories to explain and predict. If explanations are poor, the proposed theory might not seem so good. If predictions are incorrect, the proposed theory might be considered falsified.

So much for how "theory" is used in empirical sciences. This is philosophy! Philosophers engaged in theory building aren't often making empirical predictions. This is especially true among those studying metaphysics. Metaphysics is a branch of philosophy that - roughly speaking - seeks to explain the fundamental nature of the world. A consequence of this approach to theory building is that metaphysicians don't often have the benefit of making empirical predictions. You might take this as a significant cost to metaphysics in general. Why care about theories that can't be empirically falsified?

Because we nevertheless want explanations for the way things are, and empirical sciences don't seem equipped to provide answers to every legitimate question. Suppose we find the "One, True, Scientific Theory." This theory answers any question we might have about the empirical world. Here's one question it won't be able to answer though:

(W) Why is the "One, True, Scientific Theory" true, rather than some alternative theory?

This question outstrips the empirical world. Go ahead, try and think of a way to test (W)…

In any event, it's not like metaphysical theories aren’t falsifiable, at least in some sense. Metaphysical theories still have to explain data, even if they don't make empirical predictions. If they do a poor job, then they're no good. How do we decide? Some cases are more obvious than others. Suppose I want to explain everything by simply saying "God exists and made everything." That's a very simple theory; it explains everything. That doesn't mean it's a good theory; it doesn't make clear how it explains everything…

There are other ways to dispute theories. Theories are comprised of axioms. These are the building blocks. You've likely seen such things before. Euclid's Elements starts off with axioms, from which he builds a theory of geometry. Axioms are often primitive assumptions of a theory. You don't argue for them (imagine trying to argue for the existence of a point…), but that doesn't mean you get them for free. Axioms earn their keep based on how much they explain. A good theory explains whatever domain you want to explain, and is composed of axioms that show very clearly how that explanation works. Euclid describes a geometric theory, and each axiom is used to explain precisely how that theory works. There are no surprises.

Sartre's Brand of Existentialism

You might wonder why I'm talking about metaphysics in a class on existentialism. Well, today we're talking about Sartre's Existentialism is a Humanism, where Sartre is responding to three objections to existentialism in general. His main line of response is that his objectors are confused. His strategy for responding to them is to clarify existentialism as a theory. He does this by outlining the axioms of his brand of existentialism. Sartre is engaged in theory building for the sake of clarification.

We will take the axioms in turn, and see if Sartre has adequately responded to his objectors. I list the axioms here:

(A1) Existence precedes essence
(A2) The only standards of evaluation are those created by agents
(A3) There are standards of evaluation
(A4) For any agent S, action A, and context C, if S does A in C, then S creates a standard of evaluation in C, namely, that any agent S' do A in C

Consider (A1):

By "existence" he simply means the brute fact that something exists.
By "precedes" he means, I think, "explanatorily prior". To say x is explanatorily prior to y is to say that any explanation of y involves x. Suppose you and I are having a conversation. That you and I exist partially explains the conversation. We say our existences is explanatorily prior to the conversation. Similarly, the set {John} exists because of me. I am explanatorily prior to the singleton. Atoms are explanatorily prior to tables. The list goes on and on…
By "essence" he means function or purpose. Let me unpack these further. "Essence" is Aristotle inspired jargon. For Aristotle, we can provide four different sorts of explanation for why something exists. Consider a table. One explanation is material, i.e. the wood explains why the table exists. Another is formal, i.e. the shape of the table explains why the table exists. A third is efficient, i.e. the producer of the table or artisan. The last is the final explanation, i.e. the point of having a table. Sartre is using the word "essence" as roughly synonymous with "final cause" or purpose.

Altogether, (A1) says the brute fact of human existence is what explains whatever purpose agents have, rather than the other way around.

Consider (A2) and (A3):

A knife has a function and is presumably designed to satisfy that function. But in a world with no objective values, there are no objective standards of evaluation. If there are any standards, then, they must come from us, for agents impose standards. So, if a knife can be better or worse, then it must be better or worse according to some standard imposed on it by agents. Since there are values and they can't be objective, they must come from us.
What goes for knives goes for us too, though in a different way. Knives are material objects and can acquire functions, but we are also material objects that acquire functions. The difference here is that we create functions for ourselves.

Consider (A4):

This axioms bears weight in the theory, as we will see in a moment, since it's needed to respond to the more serious objection offered against existentialism.
Before we turn to that though, one things is worth noting…As it stands, (A4) may be either too weak or too strong. To see how, suppose I have to decide whether to save a child from drowning or not, and I'm fully able to do so without any difficulty. Suppose I do it. Then according to (A4), I'm prescribing in my action that anyone else - S' - save the child from drowning - A - in similar contexts C. But what does "similar contexts" mean? Let's think about two extremes, trading on the distinction between types and tokens. Tokens are not repeatable; types are repeatable. In a classroom of students, there are many token students, but one student type, of which they're all instances; similarly, there are many token humans, but one human type. Let's apply this to contexts. That is, we can think of contexts as either types or tokens.

Suppose "C" in (A4) means a token context. Then it can't be repeated. Then (A4) is trivial, since I'm the only person who will ever be in that context, so of course everyone ever in that token context should do what I did; I did it!

Suppose "C" in (A4) means a type of context. Then it can be repeated, but it's not clear where to draw boundaries on contexts. Suppose I'm unable to swim and my trying to save the drowning child will ultimately kill us both. If context is understood broadly enough as a type to mean something like "a child is drowning and there's one person there who can help" then (A4) will say I should save the child in this new context. But that seems like the wrong result. Surely Sartre doesn't mean to say anyone should do the same as I or you or he does. Saying otherwise makes (A4) vacuous. Clearly, "C" needs to be spelled out in more detail.

But don't let me trick you here! It's true (A4) needs some work, but let's give Sartre whatever specification of context he needs to get his responses to objectors off the ground. This is permissible, because - remember - axioms are assumptions about theories; they earn their keep based on whether they can explain what they're designed to explain. If we want to dispute (A4), we need to think hard about whether this axiom - which I've claimed is very important in Sartre's defense - can do the work he claims it can.

Can it? Sartre considers three objections in this essay. I'm only concerned with the last, that existentialism is subjectivist about values.

(Q) Using the resources (A1)-(A4), can Sartre adequately respond to the charge that existentialism leaves us with only subjective values

My answer: No. It's a half measure at best. I think Sartre is biting the subjectivist bullet here, and trying to soften the worry by offering (A4). This is common when building a theory. We're not engaged in theory building - whether metaphysicians or empirical scientists - to provide perfect theories of the world. Perfection is too high a standard. Every theory has counterexamples; good theories provide explanations for those counterexamples, and do a good job of explaining why they're not so bad. This is just to say, simply because there is a counterexample to a theory, it doesn't mean the theory is dead. A theory may survive given its explanatory power, for instance.

This is not however to say I think Sartre's theory as characterized by (A1)-(A4) survives. Of course, you may not be moved by the preceding worry against existentialism. You may agree with Sartre that there are only subjective values. As a matter of fact, I agree with him…

Still, there are many ways to kill a theory. If you can show the axioms of a theory are inconsistent, or conflict with each other, then you can be sure that theory is false. I'm going to try and make the case that (A1)-(A4) are inconsistent, with the following argument:

The Parental Objection to Sartre's Existentialism

(1) SUPPOSE: (A1)-(A4) are true
(2) SUPPOSE: Sartre decides to and does have a daughter - Jessica - according to (A1)-(A4), who in turn raises a daughter - Sally
(3) If (1) and (2), then Sartre's action of raising Jessica according to (A1)-(A4) prescribes that Jessica raise Sally according to (A1)-(A4)
(4) Hence, Sartre's action of raising Jessica according to (A1)-(A4) prescribes that Jessica raise Sally according to (A1)-(A4)
(5) If (4), then Sartre creates a standard of evaluation on Jessica by raising her as he does
(6) Hence, Sartre creates a standard of evaluation on Jessica by raising her as he does
(7) If (6), then there is at least one agent - Jessica - whose essence precedes existence
(8) Hence, there is at least one agent - Jessica - whose essence precedes existence
(9) Hence, (A1) is false

I'll let you decide whether this is sound, i.e. whether the premises logically entail a true conclusion on line (9). Here's some advice. You might think the conclusion is true. But here I'm asking you whether the conclusion is true because of (1)-(8). That means you have to think about (1)-(8). As a general strategy, when evaluating arguments you can ignore lines that start with "SUPPOSE". Additionally, any line that starts with "Hence" follows from the preceding lines by logic. We're not going to dispute the logic in this class any further, so you can ignore those lines. That leaves lines (3), (5), and (7). If the argument fails, then at least one of those lines is false. To show a line is false, you try and provide a counterexample. As further guidance, I think (3) and (5) are plausible. If I were to dispute the argument, I'd try to undermine (7)…

Appendix

Sartre argues quickly that intuition cannot guide one's decision to do some action A over B. This is his discussion of the soldier. It’s very quick; I wanted to make sure how he’s arguing is clear. Here's the argument as I understand it:

Sartre's Objection to Intuition as a Guide to Decision

(1) SUPPOSE: Intuition guides my decision to do A rather than B
(2) If intuition P guides by decision to do A over B, then my intuition P to do A is stronger than my intuition Q to do B
(3) Hence, my intuition P to do A is stronger than my intuition Q to do B
(4) For any intuitions w, x, and actions y, z: w is stronger than x for me if and only if I actually do y rather than do z
(5) Hence, I actually do A rather than do B
(6) It is impossible to be guided by an intuition to do A and do A at the same time
(7) Hence, intuition cannot guide my decision to do A rather than B

(1) is either true or false. Sartre is supposing (1) to show this leads to contradiction. Since there are no contradictions in the world, if (1) leads to contradiction, then (1) cannot be true, so it must be false. Sartre claims (2) is true. (3) follows from (1) and (2) by logic. Sartre claims (4) is true. (5) follows from (3) and (4) by logic. (6) seems true based on the meaning of ‘guide’. It doesn’t seem plausible that intuitions can play a guidance role if they’re only useful at the time of action. That’s like saying the following is a warning: Someone yells “Watch out!” to someone else just as they’re being hit in the head by a stray basketball. That doesn’t seem much of a warning...

Textual aside: Sartre says "feeling is formed by the decisions I make" but that's not strong enough for the conclusion he draws about feelings here. For this argument to work, he needs to be saying instead that "feeling is decisions made". If he allows for temporally precedent formation, then this argument fails. This is why I believe he's making the stronger claim.

To evaluate this argument for soundness, i.e. for whether (7) follows from the premises, you need to check whether the premises are true. Ignore (1), (3), and (5), for the reasons given above. That leaves (2), (4), and (6). Can you construct counterexamples to any of these premises?