Steve Hays indicates that he can understand how I was “thrown off by his elliptical syntax and this serves as a basis for my supposed misunderstanding. The problem is that I wasn’t thrown off by Hays’ syntax but by the factual errors and conceptual confusions. The definitions Hays gives are mistaken and the conclusions are not implied by the premises.
An example of factual errors is the “thumbnail” definitions of various positions that he gives. My problem wasn’t with the fact that the definitions were “thumbnail” but that they were simply wrong. Hays wrote,
“ii) This raises the question of what, if anything, would count as evidence for LFW even if LFW were true. There are three logical alternatives:
a) Hard determinism: We are not free to do otherwise even if we wanted to do otherwise.
b) Soft determinism: We are free to do otherwise if we want to do otherwise—although we are not free to want to do otherwise.
c) Indeterminism: We are free to want to do otherwise.”
If libertarianism is true, that is if there is libertarian free will, then Hard Determinism and Soft Determinism are false. Moreover if libertarianism is true then indeterminism as a theory of causation is also true since it is a necessary condition for libertarianism. Hard Determinism is a species of incompatibilism, that freedom and determinism logically both cannot be true. Consequently Hays is just wrong to say that if there was libertarian freedom then Hard Determinism is a logical possibility. Hays is also wrong to say that if libertarian free will existed then Soft Determinism is a logical possibility. Soft Determinism is the thesis that determinism is true and is compatible with freedom, though not with libertarian freedom.
I wasn’t committing a “word=concept fallacy” as Hays maintains. I stated what the basic idea was and what I wrote was sufficient to pick out the ideas in question. Hays brings forward ideas that are contradictory to the ideas he is attempting to define. I wasn’t claiming that the definitions that I gave included everything, but they certainly included what I wrote and they don’t include what Hays said they did. It is certainly possible to define General Relativity without having to explain all of the consequences and supply every model. But it is not possible to define it by saying that it includes concepts that it doesn’t and that are contradictory to it.
Conditional Analysis. Hays thinks that I begged the question by stating that alternative possibilities are the exclusive domain of libertarians since it is possible to give a conditional analysis. Compatibilists have argued that statements about the ability to do otherwise such as “A could have done x instead of y” can be reconciled with determinism.
Conditional analysis is constructed to show that an agent could be said to have the ability to do otherwise given the truth of determinism. Determinism requires that certain past circumstances together with the laws of nature are sufficient to bring about present states of affairs. If the past or the laws of nature had been different, then present circumstances would in fact be different. The Compatibilist argues that we can understand statements concerning the ability to do otherwise in light of different antecedent states of affairs.
Conditional analysis (CA) proposes that we understand statements such as “x could have done y” to mean that if prior causal states had been different, then the agent would have done y. The fact that the agent would have done otherwise is how we are to understand the ability to do otherwise. Suppose that some agent had an intention I1 to perform some act D1. But she might have had a different intention, I2 to perform some other act D2 had different antecedent causal states obtained. I2 would be sufficient to bring it about that an agent did do otherwise (D2) than she actually did (D1). In other words, statements such as “x could have done y” should be interpreted as “x would have done y, if x had willed, desired, chosen, or intended differently than she did.”
This way of reading “could have done otherwise” statements may not seem initially plausible. Imagine that determinism is true and that you act from the desires that are brought about by antecedent states of nature plus the laws of physics. It still appears to be true that you are doing what you desire or intend to do. It seems that you are not in any way constrained by antecedent causal states. For all intents and purposes, the compatibilist notes, you act freely provided that there are no obstacles or impediments to carrying out the desires or intentions you have. If you had desired something else or intended to do something else you would have performed acts in accordance with those desires or intentions, again assuming that there were no impediments to your carrying out the said act. Given different antecedent conditions you could have done otherwise and therefore determinism is compatible with AP.
Historically such debates in philosophy are not so easily won. Needless to say incompatibilists were not immediately won over by a conditional analysis of “can” statements. J. L. Austin argued against conditional interpretations of “can” statements. (See Gary Watson, ed., Free Will, Oxford) He argued that they fail to suffice as a necessary condition for the possession of powers and abilities on the part of the respective agents. In his criticisms of conditional interpretations of “can” statements Austin asked a searching question. Do statements about what agents “can do” or “could have done” involve a hidden or disguised “if” clause? Take statements (1) and (2) below.
(1.) You could have done otherwise.
(2.) You could have done otherwise, if you had willed or chosen or wanted to do otherwise.
According to CA, these statements are logically equivalent. But according to Austin, this is mistaken because it requires that the presence of an ability or power to carry out something be contingent upon the agent’s willing, choosing, etc. to exercise or implement the ability or power. This would mean that one would only have abilities or powers to perform actions at only those times in which they chose to exercise them. A conditional analysis of the above statements should cash out “can” statements such as “you could have done otherwise” in terms of “you would do otherwise if.” So a CA analysis of (1) should now look like (2’)
(2.*) You would have done otherwise, if you had willed or chosen or wanted to do otherwise.
Here is a new problem. It is possible that, for a given set of possible actions available to an agent, the successful carrying out of any of those is not guaranteed by the antecedent conditions. The upshot is that a conditional analysis of “could have done otherwise” statement is too strong, because it implies that the antecedent conditions guarantee the successful completion of the act. To see this take Austin’s famous example of shooting a three-foot putt in golf. Suppose I am out on the green and I attempt to make a three-foot putt, and as luck would have it, I miss. Now it is true that I desired and intended to make the shot. So the conditions supposedly sufficient to accomplishing the task are present. It is also true that I could have made the shot even though I failed. The power to do X does not imply that one would do X on every occasion when one exercised the relevant power or ability. This implies that “would/if” statements are not a necessary condition for the truth of “can” statements. Hence a conditional analysis of everyday statements of “could have done otherwise” fails to capture the notion of ability. What a person can do is not exhausted by what that person does do.
There are other reasons for thinking that a conditional analysis of “could have done otherwise” statements is inadequate. Roderick Chisholm has given an influential line of criticism against a conditional reading of “can” statements. (See Gary Watson, ed., Free Will, Oxford) Consider the propositions below.
(a) He could have done otherwise
(b) If he had chosen to do otherwise, then he would have done otherwise
If a conditional analysis were successful, then (a) is equivalent to (b). Chisholm thinks that the problem with CA is with the very claim that (a) says nothing more or less than what (b) says. To see why this is a problematic claim we only need to think of ways in which (b) could be true while (a) is false. Imagine a person who, if he had chosen to do otherwise, would have done so. Take, for example, Jim, where Jim is a person of shadowy repute who decides to defraud his neighbor for personal gain. We may easily imagine that if Jim had chosen to do otherwise he then would have done otherwise than defraud his neighbor. But from this alone it does not follow that Jim could have actually done otherwise. Suppose that, unbeknownst to Jim there is an angel who with divine permission has it within his power to cause intentions to be formed in Jim’s mind. On this particular intention, the angel is set upon causing the intention on Jim’s part to defraud Bob and he is successful in doing so. In this case it is still true that if Jim had not chosen to defraud Bob, he would have not done so. But it is false that Jim could have not chosen to defraud Bob. He could not have done otherwise. We can easily imagine other cases in which Jim is forced to decide to defraud Bob. Perhaps Jim has some overriding motive that drives him to this decision, a starving family, say, or a mother in need of an operation. To infer that an agent could have done otherwise from the fact that if they had chosen otherwise, then they would have done so is a mistake.
The Compatibilist’s initial claim that statement (b) says all that (a) does is in trouble. Things might be such that if an agent had chosen to do otherwise, then he would have done otherwise, but also such that he lacked the ability to choose any other act than the one he did. What the compatibilist needs is an additional statement indicating that the agent could have done otherwise. With the added statement the conditional interpretation would look something like the following,
(b) If he had chosen to do otherwise, then he would have done otherwise.
(c) He could have chosen to do otherwise
(a) Therefore, he could have done otherwise.
We need proposition (c) to legitimately get from (b) to (a). This is the only way to rescue the analysis. The problem is, however that the analysis was supposed to show that statements such as (a) were equivalent to (b) without an appeal to statements such as (c). It is clear from Chisholm’s argument that they are not. In contrasting compatibilism with incompatibilism Chisholm has aptly remarked,
“They [compatibilists] are simply asking: Do we ever bring about the things that we intend to bring about? But our [incompatibilist] question might be put by asking: Are we free to will the things that we do will?”
In conclusion, the conditional analysis of “could have done otherwise” statements fails. It does so because it captures neither the sufficient nor necessary conditions of the ability to do otherwise. This is why Hays is simply wrong that it is coherent to “conditionalize” alternative possibilities.
Hays regards compatibilism as a special case of determinism but this too is a mistake. Compatibilism is not a type of determinism or a particular instance of it but rather a thesis about the relation between freedom and determinism, namely that logically they can both be true. Moreover, Incompatibilism is not a “special case” of indeterminism because Hard Determinism is a type of Incompatibilism. Incompatibilists fall under two general types, Libertarians and Hard Determinists.
My distinguishing between causation and agency was quite appropriate since one can believe in the existence of the former without also believing in the latter. In a world without agents you can still have causes. It is quite true that agents can be causes but what causation is, is not agency per se because there are causes that are not agents such as forces in nature. Indeterminism is a thesis about what causation is and not about what agency amounts to. Most libertarians are indeterminists about causation but the converse is not so.
Frankfurt Cases. Hays seems to not grasp what I was doing in discussing Frankfurt cases. When I gave the Kane/Widerker objection to Frankfurt’s counter-examples, I was not laying out other “interpretations” of Frankfurt cases but articulating objections to them. Kane & Widerker argue that contrary to Frankfurt’s claims, his case does presuppose determinism and thereby beg the question against the libertarian. For this reason, Frankfurt cases do not show, contrary to what Frankfurt claimed, that freedom doesn’t imply the ability to do otherwise. In any case, Frankfurt Cases are not examples of Hard Determinism because they are not cases where the subject is determined. That is how it undercuts Steve’s argument because Frankfurt cases then can’t be deployed as instances where someone is determined and yet thinks they are free because Frankfurt constructed them with the subject being free. Hays isn’t giving a Frankfurt case but rather a Cartesian case where a subject is deceived in believing x and is deceived in believing that y supports x by a covert controller. Hays is free to modify Frankfurt cases and then put forward the new thought experiments but then they wouldn’t be Frankfurt cases. To put forward a case where the agent is in fact determined as a Frankfurt Case is just to display Hays’ ignorance of the thought experiment and the philosophical literature on free will and determinism for the past 40 plus years.
Epistemology. Whether the evidence for libertarian freedom is equally explainable in a determined world is irrelevant to the truth of libertarianism. In my rebuttal I wrote that it is quite true that someone could think that they had libertarian freedom and not actually have it. But this is true for any belief in cases of covert controllers thereby showing that the problem isn’t with libertarianism but with our epistemological limitations or theories. If belief in an external world might possibly be illusory because of an evil demon, this is hardly a basis for thinking that the idea of an external world is false or unsupported by any appeal to experience. It would be a reason for thinking that our justification was inadequate. Skeptical arguments undermine the justification of our beliefs and not their truth.
Moreover it is not clear that freedom is a belief like a belief in an external object but more like a belief in one’s own consciousness. How does one know that one is conscious? Does one make inferences to establish this? No. One’s access to their consciousness is direct and since libertarian freedom is intrinsically part of one’s consciousness, or so Libertarians would argue, by virtue of the fact that free choices are intentional, a belief in freedom seems a lot more like a belief in one’s own consciousness rather than the red coke can over on the table. The appeal to experience by Libertarians is therefore direct and not acquired through inferrential steps.
What Hays is trying to show is that the evidence from experience for libertarian freedom doesn’t imply that we are justified in believing that we have libertarian freedom. The evidence is just “no good.” As I noted before, Hays is presuming that our belief derived from experience for libertarian freedom is formed like beliefs we form from sensory experience. And he is supposing that our beliefs from experience are made on the basis of inferences from the senses to beliefs. That is, he is assuming an internalist theory of knowledge. The case he gives is supposed to show that the inference from the evidence isn’t sufficient to produce knowledge. But as I noted above, we don’t make inferences to our belief that we are conscious from an experience of consciousness. Likewise we don’t make inferences from our sensations to beliefs. The connection between our beliefs and sensory data is causal rather than inferential and this is what skepticism is trying to teach us. (See John Greco’s, Putting Skeptics in their Place, Cambridge, 2000). To be an internalist is all well and good (well maybe not good) but a lack of justification for libertarianism shouldn’t turn on an internalist theory of knowledge.
Arguments against libertarianism shouldn’t turn on the truth of epistemological internalism. The mere fact that evil demon cases show that the evidence is consistent with a belief only shows at best that the belief on an inferential basis fails to provide justification for the belief, not that the belief is false. What is more, evil demon cases show that any evidence fails to provide justification for a belief indicating that the problem is not with the evidence for libertarianism but with the epistemological theory to which the objection is applied.
The problem of God determining someone to believe that they are a true believer and elect is a distinctive problem for Calvinism for the simple reason that while self deception is a possibility in other theological systems as Hays notes, the deception is attributable to the freedom of the agent and inattentive use of their cognitive and moral faculties rather than to God planning and rendering their deception inevitable. While every theological system has a distinction between real and nominal believers, that is certainly not co-extensive with the idea that God’s determining activity is the explanation for why people end up in one group or another. The latter idea is fairly distinctive to Calvinism (Even contradicting the 2nd Council of Orange). So on distinctively Calvinist grounds Hays has the same kind of defeater for his views on assurance or just about any other belief as the libertarian has for his beliefs concerning freedom.
The fact that on a Calvinist reading the same God who predestines people to think that they are elect when they really aren’t also determines that some people have grounds for thinking that they are elect when they are is of no help. If the agent in question is determined to think that they are elect when they are not then no evidence can function as defeaters to that belief, making cases of people having grounds for assurance and being elect indistinguishable to the agent from cases where people have apparent grounds for assurance and think that they are elect when in fact they are not.
The question is not whether God is conferring the same experience on both groups, elect and reprobate. The question is how is one to discriminate between being only apparently elect and actually being so. If God determines you to think you are elect, then it is inevitable that you think so. If God determines you to take your experience as confirming that you are elect, when in fact you are not, then it is inevitable that you think so. How then, if your beliefs are determined by God can the evidence function to discriminate between the two cases? Simple. It can’t. This is why we have a parallel case to the one that Hays provided-an agent has a belief but is in fact determined to be deceived such that the evidence for their beliefs doesn’t imply the truth of those beliefs and hence fails to amount to knowledge. There doesn’t have to be the same experience, just a covert controller and the distinction between appearance and reality.
For example, when I was Reformed I knew many people who were sure that they were elect. They were quite dogmatic on the point. Most of them ended up falling away and/or dying in unbelief to my knowledge. They thought that had confirming evidence of their own election. They thought that they had the self attesting witness of the Spirit. Now can someone be wrong and think that they have the self attesting witness of the Spirit and not actually have it? I see no reason to think that this isn’t possible. If God determines them to have that belief then this is the same circumstance as the individual who thinks that they have free will when in fact they don’t. Under the influence of covert controllers the justification of any belief can be undermined. It would be a silly mistake to take this as an indication, as Hays seems to do, that the problem is with the belief or the evidence for it. The problem is with the epistemological theory which falls prey to skeptical arguments. Hays is confusing epistemology with metaphysics.
Hays claims that the argument for predestination unlike the argument for libertarian freedom is not an argument from experience but from revelation. Certainly for some that is so, but for example it wasn’t so for Luther, who availed himself of arguments from Stoic determinism and other venues to prove his necessitarianism and Luther was not alone in making these kind of arguments from experience of the physical world and philosophy to establish his predestinarian views. Moreover, one can make a case for libertarian freedom from revelation and libertarians have historically done so. Very simply put God has libertarian freedom and we are made in God’s image and granted a measure of that freedom. The same language that is applied to God regarding choice and making judgments is equally applied to humans and other moral agents. There is no reason to think that the latter have some other *kind* of freedom than God even if God has it in greater measure.
I agree with Hays that the exegetical case for Calvinism has been made repeatedly. But not much follows from this fact. What Hays needs to show or at least refer to, are cases where it has been successfully made. Plenty of advocates of positions can refer to exegetical cases made for their positions. Just pick up a slew of commentaries on any particular book of the Bible and one will see exegetical cases made for a variety of positions.
I gave a rough sketch of the argument one could make for libertarianism from the Bible. That is hardly a display of my theological method. It is quite true that I didn’t make any effort to exegete various texts to support the basic argument, because as I said I was giving a sketch. I was trying to motivate my readers thinking more than produce a full length treatment of the subject. The exegetical case has been made elsewhere. It shows rhetorical flare to say that I use a “Biblical category as a cipher to plug in a totally extraneous concept” but it does no argumentative work. Hays has yet to show that the concept of libertarian freedom pre se is not a Biblical concept and so he is begging the question. Certainly, plenty of Calvinists historically have thought it was attributing it to at least God. At best Steve’s rhetorical comments beg the question and certainly his own tradition has ended up on my side on that point.
I certainly believe in seeing how various biblical writers employ terms to help grasp the meaning of the text. But again, I was giving a rough sketch of the argument and not giving supporting evidence for the premises. The example Hays gives is of no help because it doesn’t exemplify the swath of cases of biblical usage but only one case, which is hardly adequate. Certainly the phrase “the image of God” is used in other contexts, specifically Christological ones. What is worse is that the example that Hays gives doesn’t actually give a lexigraphical basis for understanding the phrase in a specific way but rather falls back on a sociological analysis in place of exegesis. If meaning is use then appealing to how the Hebrews saw their kings, especially when many of the key texts were written long before Israel even had kings, let alone a priesthood, doesn’t tell us how the text uses the term. At best it tells us how they saw the text in light of their cultural views. Hays doesn’t give us what he promised, namely an example of the “right way” to “interpret the key term” because the example isn’t doing exegesis.
Furthermore, Hays’ own source indicates that it is by and large the assumptions that one brings to the text that have determined what people took to be the meaning of the term. But no, this couldn’t possibly the case with Hays source! Say it ain’t so! One only has to ask why the author uses sociological data to tell us what the text means without giving any support from the text itself to see the presuppositions at work.
Certainly if covert controllers can render knowledge claims concerning freedom problematic then there is no special reason why they cannot also function in the same way for any other knowledge claim. It is irrelevant that an argument given for libertarian freedom is derived from experience. To be determined to believe something falsely one doesn’t have to derive the belief from experience. If an agent is determined to believe falsely and determined to mistakenly believe that there is data that supports this false belief, how exactly is one to find out if one is in this situation or not in a deterministic world? The evidence for any belief is compatible with the falsity of that belief if the agent is determined to believe falsely.
Time travel. It is true that Hays argument was predicated on the impossibility of time travel and that is exactly the point. Libertarians agree that time travel is impossible and libertarianism doesn’t imply otherwise. Just because libertarians believe in the ability to will otherwise that does not commit them to the idea that time travel or retrocausation is possible.
Replay argument. Hays says that if we replay the past a million times it might be “barely” possible that the agent will make the same choice every time. I don’t know what “barely” possible is. If it is possible, then its possible. Hays was clearly wrong in saying that if the agent has the power to do otherwise then “necessarily” they will do otherwise. First it doesn’t follow that just because I have a power that I will employ it. Second, there is nothing necessary about it which is why its called “free” will.
Hays then likens replaying the past up to the point in which a libertarianly free agent makes a choice to chance happenings as in gambling. But this is a false analogy since libertarianly free choices are intentional and goal directed and unlike the toss of a die or the draw of a deck. I can have good reasons for always making the same choice and never the alternative. If someone asks me to kill my little girl, then can ask as many times as they like and I will never do it. I have no good reasons to do it and I have plenty of good reasons not to.