John Martin Fischer

John Martin Fischer (University of California--Riverside) is "presenting" his paper "The Direct Argument." The two invited commentators for his paper are Randolph Clarke (Florida State University) and David Widerker (Bar-Ilan Univesity).

Download john_martin_fischer.pdf
Download clarkes_commentary_on_fischer.pdf
Download widerkers_commentary_on_fischer.pdf

Comments

[Joe]

Wonderful papers!

I have one worry, though, regarding a comment made by John and endorsed later by Randy, who writes: "Fischer does suggest (n.d.: note 19) that we should recognize different families of direct argument, only some of which rely on a modal transfer principle" (6). He continues: "Pereboom's (2001: 110-17) four-case argument and Mele's (2006: 188-94) zygote argument are examples of this sort that don't employ a transfer principle" (6).

John also says that other "direct arguments do not employ modal transfer principles; rather, they invoke notions such as 'sourcehood,' 'origination,' and so forth" (14, fn. 19).

I take it that John is referring to the ultimacy argument, or Galen Strawson's basic argument. But this argument is discussed by David in the paper that John critiques. David claims that it is committed to this transfer principle:

Transfer NU: from NU(p) and NU(if p, then q) deduce NU(q),

where NU(p) =df. “p and no one is, or ever has been the ultimate source of its being the case that p” (Widerker 2002, 322). David does not refer to the ultimacy argument as a version of the direct argument, for it does not utilize a non-responsibility transfer principle, but it does seem to presume the truth of a "modal transfer principle."

Likewise, underlying the four cases in Pereboom's argument is what he calls the "fundamental incompatibilist principle":

"... if one’s action results from a deterministic causal process that traces back to factors beyond one’s control, to factors that one could not have produced, altered or prevented, then one is not free in the sense required for moral responsibility" ("Determinism al Dente" (1994), reprinted in Pereboom (ed.) Free Will (Hacket 1997), 249; see also 246 and 252).

I've talked with Derk about this in correspondence and, although there is some debate about how to understand this principle, it seems clear that it is a transfer principle of some kind.

Thus, it seems that ALL arguments for the incompatibility of determinism and moral responsibility (or determinism and free will) presuppose a transfer principle of some kind.

[Randy Clarke]

Here's the zygote argument from Mele (2006: 189):

1. Because of the way his zygote was produced in his deterministic universe, Ernie is not a free agent and is not morally responsible for anything.

2. Concerning free action and moral responsibility of the beings into whom the zygotes develop, there is no significant difference between the way Ernie's zygote comes to exist and the way any normal human zygote comes to exist in a deterministic universe.

3. So determinism precludes free aciton and moral responsibility.

No transfer principle in sight. Is one hidden somewhere? Not that I can determine.

[Derk]

Joe quotes my claim:

"... if one’s action results from a deterministic causal process that traces back to factors beyond one’s control, to factors that one could not have produced, altered or prevented, then one is not free in the sense required for moral responsibility"

I think that this principle is not employed by my manipulation argument as a premise. Rather, it is drawn from the argument as a general conclusion, and it is in this sense that Joe is right to say that it underlies the argument. Without presupposing this claim, interested parties are asked to reflect on the manipulated agents, to ask whether they are intuitively morally responsible, and if not, whether a there is a difference between the manipulated agents and the ordinary determined agent that can explain a difference in responsibility. This procedure does not presuppose the truth of the quoted principle, and one would not want to craft a manipulation argument against compatibilism in such a way as to presuppose the truth of the key incompatibilist claim.

[Eddy Nahmias]

If the zygote argument is trying to derive support for transfer principles rather than *use* such principles to reach its conclusions, then what would make premise 1 true (i.e., intuitive in the way it is supposed to be)? As far as I can see it would be because Ernie is manipulated by an external agent and such manipulation is intuitively responsibility-undermining. (It can't be because Ernie was produced by factors over which he had no control, factors that were sufficient for all his actions, because that brings back in a transfer principle, right?)

OK, but if premise 1 is intuitively true for the reason that Ernie is manipulated, then premise 2 looks to be false. There *is* a significant difference between Ernie and Bernie (the normally created person)--Ernie is a manipulated agent. And I would suggest that *that* difference is precisely what made premise 1 intuitively true.

The upshot seems to be that the zygote argument (and other manipulation arguments, I think) either works only by using a transfer principle or it is unsound.

[Joe Campbell]

I agree with Eddy that premise (1) of the zygote argument needs some kind of transfer principle for support. Or, as he says, there is no (non-question begging) reason for thinking that the argument is sound. Certainly, there are no explicit transfer principles used in the zygote argument, as Randy presents it. But that doesn't mean that its soundness does not depend on some transfer principle or other.

I would add that any argument for incompatibilism is going to start with some assumptions about fixed propositions -- propositions about the past and/or about the laws of nature and/or about the causal structure of the world -- and then try to transfer that fixedness on to all other propositions. If you can show me an argument that makes no assumptions of the first kind -- no assumptions about the past or laws of nature or causal structure or development of the world -- then I might begin to believe that there is an argument for incompatibilism which does not depend on some kind of transfer principle.

In Randy's version of the zygote argument, for instance, the first premise is supported (to the degree that it is supported) by the assumption that the key features of Ernie's world were manipulated by someone else. That is, they were propositions that were not up to Ernie, or propositions about which Ernie had no control, or propositions about which Ernie had no choice, etc. Unless these assumptions are made, I see no non-question-begging reason to suppose that "Ernie is not a free agent and is not morally responsible for anything."

As for Derk's point, first I think that the assumptions drawn from the first few cases of your argument presuppose the truth of some kind of transfer principle, for reasons stated above.

Second, it makes little difference if a transfer principle plays the role as a premise in the four-case argument, or if "it is drawn from the argument as a general conclusion." If the transfer principle is false, then there is something wrong with the argument either way. Consider:

T, therefore C
C, therefore T

If the transfer principle (T) is a premise, as in the first example, then clearly the conclusion of your argument (C) depends on T; if T is a logical consequence, as in the second example, then clearly C depends on T, too. For if T is false (or invalid) -- in the second example -- then C is false, also.

What I meant -- and should have said -- in my first post is this: ALL arguments for the incompatibility of determinism and moral responsibility (or determinism and free will) rely or depend upon a transfer principle of some kind.

[Randy Clarke]

If some transfer-of-nonresponsibility principle is valid, then stating it as a conditional gives us a necessary truth, one that follows as a logical consequence from any proposition. But the propositions it follows from needn't depend in any interesting way on it.

What would make premise 1 of the zygote argument true is that Ernie isn't responsible for what he does, because of the way his zygote was produced. The preplanning of his life by Diana is supposed to get us to see this. Then, with premise 2, we're supposed to see that subtracting the preplanning makes no moral difference. Once we see this second point, we might want to return to premise 1 and reject it. We needn't call it question-begging; we can just say we disagree. (I think we should be humbled, though, that our responsibility is no deeper than Ernie's.)

[Fritz Warfield]

Joe closed (above):

"Thus, it seems that ALL arguments for the incompatibility of determinism and moral responsibility (or determinism and free will) presuppose a transfer principle of some kind."

I'm not really sure what the rules for playing this game are, but here's an argument:

P1. Determinism rules out alternative possibilities.
P2. Moral Responsibility requires alternative possibilities.
C1. So, Determinism and Moral Responsibility are incompatible.

Sure, the argument is compressed, and there will be serous controversy over whether "alternative possibilities" is used univocally in the premises. And serious controversy over whether each premise is true. But I doubt there will be serious controversy over whether the argument "presupposes" a transferNR principle. Tell me what "presupposes" means so that the argument even arguably does "presuppose" such a principle and I bet I'll be able to see it....

***

Also, though John and others have done it (and the discussion of the issue is ongoing and interesting) denying the transfer principle that is equivalent to the claim that not-even-partly morally responsible transmits through *entailment* is at least not something that is simple and uncontroversial. Randy correctly points to the "avalanche" type overdetermination cases as the best attempts but I don't think these are decisive. After all, it's worth remembering that, for any relation, closure under entailment is strictly equivalent to the conjunction of closure under equivalence and closure under conjunction elimination. These aren't easy to deny for not-partly-responsible. John (working with Eleonore Stump) has rejected closure under conjunction elimination by piggy-backing on *their* intuitions about the overdetermination type counterexamples to transferNMR. But this move is at least controversial. Direct arguments using closure under entailment as their transfer principle at least have a chance still, or so it seems to me. [The overdetermination candidate counterexamples can be resisted in the old-fashioned way -- by being a bit fine grained about the individuation of events].

[Joe Campbell]

Fritz,

I wonder if you wrote this before or after my last post. I said there that I should have used 'rely' or some other word rather than 'presuppose.'

I did not mean to suggest anything profound. It was just in response to several claims made by John and Randy such as Randy's point that: "Fischer does suggest (n.d.: note 19) that we should recognize different families of direct argument, only some of which rely on a modal transfer principle" (6).

I do not believe that the best way to respond to incompatibilist arguments is to deny the corresponding closure principle. I was just making a point about the structure of those arguments.

[Joe Campbell]

Randy,

First, the latter point that I made -- about the connection between the four-case argument and Derk's principle -- was a fallback position, something to note if it turns out that the other criticism (Eddy’s point) didn't work.

Second, what you're pointing out is that every necessary proposition is logically connected -- dependent in an important way -- on every other proposition. We all know this. To say that some, possibly contingent transfer principles have a similar relation to propositions that comprise various incompatibilist arguments is not insignificant.

Nor do I want to make too much of this observation. Certainly it doesn't translate into any clear advantage for the compatibilist, as Fritz notes. For the record, I think that van Inwagen's BETA is (likely) valid. Although I think that some non-responsibility transfer principles are compromised due to counterexamples noted by Widerker and Fischer, I can't accept all of these counterexamples, for I am critical of the Frankfurt-style examples on which they depend. I'm just interested in understanding the deep logical structure of arguments for both kinds of incompatibilism.

Third, I think that the real action is with Eddy's point. For instance, you write: "What would make premise 1 of the zygote argument true is that Ernie isn't responsible for what he does, because of the way his zygote was produced. The preplanning of his life by Diana is supposed to get us to see this."

But how, exactly, does it get us to see this? One story is that Ernie’s actions result from causal processes that trace back to factors beyond Ernie’s control -- factors that were put in place by Diana, not Ernie. What work is being done with the appeal to Diana’s preplanning if it isn’t something along these lines? Yet if it is something along these lines, then we have an explicit appeal to Derk’s fundamental incompatibilist principle. That strikes me as a rather significant observation. Given the similarity between the zygote argument and Derk’s four-case argument it seems that Derk might be wrong about the role that his principle is playing in his own argument. That seems to be a significant observation, as well. Not earth-shattering, compatibilist establishing significant but significant nonetheless.

Fourth, I am a big fan of humility (which might be surprising to those who know me well) and I’m humbled by the fact that neither the zygote argument nor the four-case argument are easy to dismiss. Does either make me believe that I lack free will? No.

Fifth, the real importance of these issues is in another area, e.g., the area of Frankfurt-style examples. Why do we think that crucial agents in Frankfurt-style examples could not have done otherwise? What is the underlying argument that leads to that result? Suppose it turns out that that argument depends on the very principles and premises that make up the consequence argument. What conclusions should we draw in that case?

[Randy Clarke]

Joe, you raise some interesting questions about the use of cases in philosophical argument. Often judgments about cases are appealed to as providing support for general principles. One might demand justification for those judgments, and one might accept as justification nothing less than arguments employing general principles. But in fact (I think) judgments about cases are sometimes epistemically advantaged: they are sometimes more evident to us than are the relevant general principles.

[Joe Campbell]

That might be right. But one could also say that common sense beliefs -- like the claim that we have free will -- are epistemically advantaged over judgments about cases.

[Randy Clarke]

I have a question for John. I gather there's something in the Direct Argument that you reject. And I'd guess it's not the premises, since you don't reject the premises of the Consequence Argument. So, I guess, you find the transfer principle invalid. Is that right?

[John Fischer]

Randy,

Yes, I would reject the Transfer Principle (in its various formulations).

I believe that the Transfer of Non-responsiiblity principle can be rejected, whereas the Transfer of Powerlessness Principle ( with the "powerlessness" interpretation of the relevant modality, as employed in the Consequence Argument) is harder to reject. I discuss my reasons for rejecting the Transfer of Non-responsibility Principle in Chapter Eight of MY WAY.

I am grateful for the thoughtful papers by Randy and also David, and I will certainly benefit from them, and from the discussion, in revising the paper.

[John Fischer]

I agree with Fritz that denying Transfer of Non-Responsibility is not uncontroversial, and also that the examples I (and Mark Ravizza) have invoked are not decisive. I still think there's a strong case, however.

Also, just as I don't think Direct Argument needs TNR, I don't think the Consequence Argument needs the Transfer of Powerlessness. (I do not deny TP, I just don't think it is necessary for the argument.) Here again Fritz and I have a disagreement, and I admit that I haven't managed to adduce a knockdown argument.

[David Widerker]

Joe,

I agree with Fritz that arguments for the incompatibility of moral responsibility and determinism (or for the incompatibility of free will and determinism) needn't appeal to Transfer Principles of sorts. Like John, however, I am very suspicious of the validity of Transfer NR. Reason: So far, I have not seen a good response to any of the counterexamples to Transfer NR I have suggested in the past, including the one I mention in my comments on John's paper.

To friends of the Direct Argument I wish to report about the following interesting result that I came across when writing my comments:

IF there a valid version of Transfer NR,
THEN one can show that determinism rules
out free will,

(or something very close to the consequent of of this conditional). Less formally put:

IF determinism rules out moral responsibility, it also rules out free will.

[Joe Campbell]

Thanks, David! I should have been more careful and said something like: For any English language argument for the incompatibility of moral responsibility and determinism (or for the incompatibility of free will and determinism), there is a way of adequately expressing its content which makes reference to a Transfer Principle of some sort. You might still disagree with this but it is at least closer to the truth.

For instance, van Inwagen's Second Argument (1983, 78-93) is expressed in symbols from first-order logic and makes no explicit reference to a Transfer Principle of any sort. Yet the very same English language argument (the Consequence Argument, 56) can also be expressed in two other ways each of which makes explicit reference to a Transfer Principle. I find this interesting though clearly many think it isn't very.

Rather than wasting so much time pressing general principles of questionable interest, I should have just made a simple comment about the Four-Case Argument and the Zygote Argument, e.g., one can describe either of those lines of reasoning in a way that makes an explicit reference to a Transfer Principle of some sort. This still seems to conflict with statements made by John and Randy.

Note that the Direct Argument began its life as a formal version of an English language argument that was, to my knowledge, never fully expressed. It is analogous to van Inwagen's Third Argument. Both the Four-Case Argument and the Zygote Argument began as English language arguments. I see them as analogous to the Consequent Argument.

[John Fischer]

Joe,

Thanks for your helpful contributions to this discussion. You say:

"...one can describe either of those lines of reasoning in a way that makes an explicit reference to a Transfer Principle of some sort. This still seems to conflict with statements made by John and Randy."

But I'm not sure that the point does conflict with what I (or Randy) would be inclined to say. I don't think I would deny that one can formulate (reformulate?) the arguments--perhaps even ALL Direct Arguments--so that they make explicit reference to some sort of Transfer Principle. But certainly this doesn't show that such reference is necessary, or that all Direct Arguments rely on such a principle. Similar issues come up in the debate about whether all adequate versions of the Consequence Argument rely, at some basic level, on a modal transfer principle (van Inwagen's Principle Beta or the Transfer of Powerlessness Principle".

One can always formulate an argument so that it imploys a given principle (perhaps the formulation will be awkward or redundant or whatever); but the interesting question is whether the principle is needed.

[Joe Campbell]

Thanks, John!

I can admit that it was wrong to suggest that all direct arguments require transfer principles. But I also think that it is wrong to suggest that there are "different families of direct argument, only some of which rely on a modal transfer principle" (John, fn. 19; Randy, 6). Specifically, I object to the use of the term 'families' in this quote.

As I see it now, there are at least three families of direct argument -- where a direct argument is one that argues directly for incompatibilism about determinism and moral responsibility without first establishing that free will is incompatible with determinism.

1/ Manipulation arguments like the four-case argument and the zygote argument.

2/ Ultimacy arguments like Strawson's basic argument and THE ultimacy argument (McKenna, Widerker, Pereboom, Shabo).

3/ Other English language arguments like the one underlying van Inwagen's original direct argument (1980).

Any of these arguments may be characterized in formal ways and each of them may be characterized -- correctly and adequately -- by appeal to a modal transfer principle.

I was wrong to suggest that this was the best or only way to characterize the arguments but it is just as wrong, I think, to suggest that the "families of direct argument" divide up between those that rely on a modal transfer principle and those that don't. All of them may be formalized in this way, though none of them must be formalized in this way.

Of course, you might want to classify the families of direct argument differently putting, say, the zygote argument and Strawson's basic argument in one pile and van Inwagen's original direct argument in another. I would say that this is mixing apples and oranges since the former two are English language arguments and the later is a formalization of an (as yet unarticulated) English language argument.

[Fritz Warfield]

John,

About this:
"Also, just as I don't think Direct Argument needs TNR, I don't think the Consequence Argument needs the Transfer of Powerlessness. (I do not deny TP, I just don't think it is necessary for the argument.) Here again Fritz and I have a disagreement, and I admit that I haven't managed to adduce a knockdown argument."

You and I are actually on the same side of this issue (and against van Inwagen's published position). We are on the same side for different reasons - what I've rejected in the past is your argument for this claim, not the claim itself. So at least our fight over this one is an internal conflict for a change. Is that progress?

[John Fischer]

Yes, thanks Fritz, that's a kind of progress.

It is amusing, isn't it, how we philosophers find ways to disagree, even when we agree on much. Al Mele and I agree on lots of stuff (not everything), but I have managed to take issue with his agnosticism about compatibilism (about determinism and responsibility). So here is the situation. I am a compatibilist about determinism and responsibility, but I agree that I do not have a knockdown argument, and reasonable people can disagree. I think the evidence is good enough for me to accept compatibilism, but it is short of decisive. Al Mele is an agnostic, thinking that the evidence is not quite good enough for him to accept compatibilism (but also he does not accept incompatibilism). Now this is, in the scheme of things, a pretty small disagreement; but we philosophers can still hone in on it like a laser!!

[Eddy Nahmias]

Am I right to add to John's example of disagreeable agreements that van Inwagen and Mele, who disagree in many ways, agree in an interesting way? Mele thinks that either (or both) compatibilism or (his version of) libertarianism is more plausible than skepticism about free will, so he holds that FW and MR exist in one form or another. Meanwhile, van Inwagen thinks MR exists for sure (against the skeptic), and while he thinks it's likely to require FW (of a sort incompatible with determinism), if determinism turns out to be true, then he thinks compabilism is true. So, they both are agnostics, though of different sorts, about the truth of compatibilism, but they both take these views in part to oppose skepticism about MR...

[Buy Online Rx]

I would be really interested to hear from anyone in the agile community who was successful in gaining a spot on the core PMBOK v5 committee. I hope there are some agile proponents represented. So please drop me a line if you were successful.

[mens health]

I just don’t see how the right solution can be so tightly disciplined to a dollar figure. By all means, recognize that a problem has a cost attached to it, and therefore the solution must be limited by that financial context.