Saturday, December 31, 2005

Clinging to Deterministic Freedom, Postponing Causal Explanation

Blake has offered the first step of his planned argument, asking for more clarity on my assumptions about determinism and causality. I’ve only read the Stanford Encyclopedia (SEP) link he offered so far, but I don’t think I need to read the other to respond to his comment.

In my first couple of posts related to this topic I was content with the term ‘determinism.’ I think I started—perhaps ill-advisedly, I am not yet sure—using the term ‘causal determinism’ after reading Blake’s Dialogue article. After reading the SEP article I am content to leave ‘causal’ out of it—I don’t know that any of my posts depend on it, though I would be glad to know how, if they do—and revert to naked ‘determinism’ as defined by the SEP to take as my premise:
Determinism: The world is governed by (or is under the sway of) determinism if and only if, given a specified way things are at a time t, the way things go thereafter is fixed as a matter of natural law.
I am also content with the statement near the top of the article that “there is no agreement over whether determinism is true (or even whether it can be known true or false), and what the import for human agency would be in either case.” I see this statement as consistent with the idea that taking determinism as a premise and seeking to work out a compatibilist view of freedom is a reasonable project. That is to say, there is no consensus that such a project is necessarily doomed to failure.

I don’t think the existence of eternal intelligences in LDS thought entails any unavoidable ‘space invader’ problems that invalidate this SEP statement. One can always maintain faith that relativity (or something like it), or a non-toroidal topology of space, prevents the action of intelligences as space invaders (objects of unbounded speed affecting a system at say, t=0, whose arrival was not logically entailed by states of the world prior to t=0).

Further, one can posit that the intelligences contemplated by LDS tradition do not have libertarian freedom, in the sense of having genuinely open futures; but instead arguably enjoy what I might call autonomous freedom: they possess a meaningful independence of action from certain other objects in the world—characterized by the capacity for scenario formulation and selection according to their own internal rules and values—while still having specific material aspects of their future (e.g. which kingdom of glory they will inhabit) that necessarily and uniquely follow from the infinite regressions of their own internal prior states and those of God.

I am rather hazy on the whole issue of causation, which is why I am basically punting on that aspect of Blake’s comment. I suppose that as a physical scientist my general sympathies lie with notion that efficient causation—as I understand it, the laws or regularities that describe the mechanics of how things happen—has some sort of primacy I am not presently equipped to describe well. (Nor can I offer a detailed efficient causal account of how what I have called autonomous freedom arises in humans, but think it plausible that such an explanation is possible.) I don’t even remember the specific names of Aristotle’s other forms of causation. I suppose my prejudice is that other forms of causation may give ‘insight’ and ‘meaning’ in human terms, but are not ultimate or fundamental “pushy explainers” that “make things happen in certain ways” (to use phrases from the SEP article).

Finally I suggest that while LDS scripture and doctrine declare humans to be ‘free,’ they are not sufficiently philosophically precise to distinguish between what I have labeled ‘libertarian’ freedom and ‘autonomous’ freedom. It is also not clear to me that they rule out the sufficiency of efficient causation as the ultimate “pushy explainer”—which sufficiency I tend to assume on the basis of parsimony, taste, and (admittedly limited) experience.


Christian: You really ought to read the other paper before you respond since it is the one that contains the arguments upon which I rely. As it stands, your affirmation of determinism is a mere article of faith without any sufficient reason to adopt it. The reason that you rejected libertarian accounts was that they did not allow for a sufficient explanation but left the choices of agents to mere chance or luck. However, it now turns out that you do not believe that the principal of sufficient reason applies to your views. You have no sufficient reason to explore determinist as opposed to libertarian views of freedom. Indeed, it truly appears that your choice in favor of determinism is random.

So your position is a form of special pleading. As such, your engagement of discourse is unfair -- you say in effect, "I demand that your view meet criterion X to explain accountable choices, but my own view doesn't have to me that criterion."

However, if what you adopt is a form of universal determinism without causes, then what is the explanation of our choices? You cannot appeal to sufficient causation because all you now have are general covering laws and a statement that given these laws (that remain unspecified and vague) every moment is determined to be what it is by the prior moments. That is not an explanation but a magical waive of the wand -- to use Jeff's felicitous phrase from Dennett.

However, you now run into another major difficulty. When you describe the freedom of intelligences you revert to a view where there is no time prior to T0 at which intgelligences who are not already in causal relations could enter into causal relations. However, given that you have a view of efficient causality it turns at that the possibility of interaction and space invaders cannot be avoided. That view is open only to someone who claims that there are no temporal causal relations at all.

I will also argue (later) that your terms "scenario formulation" and "selection according to their own rules" are loaded and presuppose an agent causal view of intelligences. In fact, what you seem to assume is a form of teleological cause (one of Artistotle's other four causes) because scenario formulation assumes that intelligences contemplate possible futures and then, on the basis of their own values, choose among these futures. Where do these possible futures come from and why believe that they are determined? More importantly, how can these values be their own? Why not accept that the choice among which possible futures to contemplate (in the sense of deliberation) is a creative interaction that requires agent causation? It seems to me almost certain that the "selection" among which futures haved been formulated is a libertarian choice since it assumes that the choices among which the selection is made are open in the moment of decision or selection. In other words, once again what you describe is an agent causal power to formulate scenarios and an agent causal power to select or choose among them.

However, I believe that your move is the wisest and best that can be done in this desperate view. The determinist ought to give up on any theory of causality and leave the explanatory-causal-connections vague and unexplained. I consider that to be a major libertarian step in the right direction because the major motivation for accepting determinism is the supposition that everything must be capable of explanation -- and you have now given up that assumption. Your prior assumption was that the prior causes provided a sufficient explanation for just this particular action (I wouldn't call it a choice) and there was no other action open at that noment given the past. You have also given up the view that the scientific world-view requires a preference for determinism. So the major reasons usually given for prefering determinism have been jettisoned and accepted as a mere article of faith without any sufficient support.

Now the rest of the discussion can proceed because I believe that there are good and sufficient reasons to accept the libertarian view and to reject the deterministic view outside of science and the presumption of sufficient causal explanation. That is all that I really wanted to achieve with this first argument anyway.


Comment by Blake | 12/31/2005 01:47:00 PM  

Christian: I wanted to seperate this comment from the prior one on the status of causal explanation because it addresses a different but closely related topic. You adopt a view of laws that you say entails determinism -- natural laws as pushy explainers. However, if you adopt a view of natural laws along the lines of Earman, Hume, Mill and other non-necessary causal connection theorists (who are by far and away the majority now-days) then your view of laws does not entail determinism. You accept a view of laws whereby these laws are "pushy explainers," but without more substance than you have given there is no explaining going on. Why reject the view that laws are merely descriptions of our constant conjunction generalizations that are subject to all kinds of exceptions? Why not accept the view of laws of nature that they are merely best systems regularities that systematize all events in history but there is no threat to libertarian free will? Indeed, this latter view seems to be the much more natural view to me for someone who accepts that there are eternal intelligences. The laws of nature then include every human agent's act as part of the universe-wide pattern of events that then form part of the laws. The laws thus don't explain human acts so much as include them as part of the explanation as to which laws obtain. If intelligences are eternal, it seems that this is the way we must formulate the laws of nature because intelligences are just as basic as the laws of nature -- or given in the way things are. That is the way I took in my book.

On such a view, I suggest that we re-frame the entire discussion without the assumption of determinism because our view of intelligences is so radical that it requires a fresh perspective without assuming an old perspective that grew out of the view that God created the world ex nihilo and controlled everything that happened with universal natural regularities thereafter. It seems that the much bolder view is to reject the notion of eternal laws as they have traditionally been discussed (pushy explainers) and accept that there may be no necessary and/or universal laws of nature like Cartwright, van Fraassen and Dupre. Given such a view, we can explore anew what it means to be a free agent having a choice between life and death, good and evil, and that we are accountable for such choices.  

Comment by Blake | 12/31/2005 02:19:00 PM  

Blake, I will read the other article, but I will say a couple of things quickly before doing so.

Do you dispute the following statement from the SEP article: “there is no agreement over whether determinism is true (or even whether it can be known true or false), and what the import for human agency would be in either case”? If you don't dispute it, I am puzzled as to why you seem to argue with such vehemence against the exploration of views that take determinism as a premise.

You have no sufficient reason to explore determinist as opposed to libertarian views of freedom. Indeed, it truly appears that your choice in favor of determinism is random. 

Come now, surely it's not as bad as all that. Let's make sure we're on the same page on "sufficient reason" (apologies for my philosophical inexperience). According to that same SEP page, "sufficient reason" refers to the idea that "everything that is, has a sufficient reason for being and being as it is, and not otherwise"?

It's true that I have real doubts about the universal applicability of this principle of sufficient reason; that is to say, I expect there are some ultimate features of the world that simply are, with no reason for them to be that way and not otherwise. I believe or hope there are only a few such 'ultimate features,' and that these few unexplained fundamentals do give sufficient reasons for a much, much larger set of observed phenomena to be as they are and not otherwise.

Now determinism might be one of these unexplainable ultimate features, one without "sufficient reason"; but I fail to see why that provides a basis for calling my faith in determinism "random" as you assert. To the contrary, the observed regularities of our experience seem to constitute good (if not sufficient?) reasons to make that choice. For example, out of the unimaginably large number of keystrokes made in typing arguments about free will, I am not aware than anyone's finger has ever passed through a key rather than depressing it. If fingers sometimes passed through keys and sometimes depressed them, then I would agree with you that belief in determinism would be a random choice. But such is not observed.

Goodness, I haven't even got past the first paragraph of your first comment on this thread. But that's all I have time to write at the moment. 

Comment by Christian Y. Cardall | 12/31/2005 03:20:00 PM  

Christian: Of course I don't dispute that there are reasonable and bright people who believe that determinism is true and that free will is compatible with determinism. That is simply an empirical fact. What I deny is that within the context of LDS thought such a view can be reasonably held. It is a question that hasn't been explored much -- but I am confident that the prevailing view of libertarian free will is required by LDS views of agency, accountability, ability to choose between good and evil, repentance so that the past is not determiner of the future, etc.

Moreover, it seems clear to me that there is no compelling reason to adopt determinism. If it conflicts with our experience of choosing among alternatives, we have good prima facie reason to question determinism until shown otherwise.

As for regularities -- I see no reason whatsoever to adopt a determinist point of view to explain why regularities like not having my finger press through the button rather than on it prevail in our experience. In process thought, novelty and creativity arise with complexity. Material conglomerates have very little creativity and simply mimic their past -- tho as a physicist you recognize that there is a non-zero probability that my finger just might press thru the key! In biological sciences, the emergence of new types of properties leads to greater creativity and novelty. A dog is more creative than a clump of clay and humans are more creative and novel in acting than dogs. This fact must be explained. How does a deterministic view do that?

If you "press your wife's buttons" you'll get a far greater divergence of responses than if you press the button on your computer. That is because humans are not mere things and are not merely the upshot of the past. Certainly the types of reafferance that occur in scenario formulation and choosing among alternatives takes a different kind of explanation than the reason why my finger doesn't press through the button. 

Comment by Blake | 12/31/2005 04:39:00 PM  

A little more on Blake's first comment on this thread (I'll get to subsequent ones later)...

On the third and sixth paragraphs, it's not that I have abandonded hope for a causal explanation---it's that this distinction between causation and determinism is new to me and I'm just now feeling my way forward. I tried to regain my feet a bit in the latter part of the post by supposing it is efficient causation that I have in mind. I have attempted to outline some higher-level elements of what an efficient causal explanation would include: scenario formulation and selection. The fact that it cannot presently be shown how these processes arise from (for example) particular patterns of brain activity does not logically exclude the possibility that this is in fact what happens. Since it is known that systems with simple determinstic rules can exhibit complicated behavior, it seems parsimonious to press this view to see how much it might account for.

I did not understand the fourth paragraph. Depending on conditions at T0, two billiard balls that have not yet collided at T0 may yet collide at some point in the future. Why not the same for intelligences?

On the fifth paragraph: I don't see why "scenario formulation" and "selection according to their own rules" imply an open future. As I have mentioned in past posts, expert chess programs do both of these activities, and I don't think you would disagree that they are acting deterministically. 

Comment by Christian Y. Cardall | 12/31/2005 08:18:00 PM  

Christian: I wanted to just comment on your last paragraph in the last post. You have modified the last requirement subtly but dramatically. You have changed it from "selection according to their own VALUES" to "selection according to their own RULES." I assume that you made this change because you recognize that a computer doesn't have values and so can't play a chess game according to such values -- although the game of chess has rules so it could mimic the rules that have been programmed into it. This change is dramatic because only humans, only persons with libertarian free will, can have values that guide their choices.

Your response is enlightening in another way. Do you really claim that human scenario formulation and choosing among the possible futures imagined is like a computer's mathematically formulating possible moves and playing a game according programmed rules? Can a computer program have imagination? Can it respond novelly to a scenario that is outside of its programming? I suggest that you might want to reconsider your position. Would we ever hold a computer formulating chess scenarios and selecting a move according to mathematical rules morally accountable? Would you blame the computer program or the computer programmer for a mistake made in selecting a move?

I suggest that we are more like computer programmers than computer programs. We can choose which programs will run our lives. Or we can reprogram ourselves. Or we can create entirely new programs. We can respond accountably to scenarios that we have never contemplated and act in ways that are entirely different than our past.

The fact that you would suggest that computers can perform the same kinds of functions as you believe determined individuals can is very revealing and instructive. If I am to be accountable, there must be some point at which I am the programmer rather than the computer running a pre-fabricated program. That is how agent causation works. We have a basic power of imagination and we also have values and we can choose among these values. For example, I may have a motivation to eat. But if it's fast Sunday, my values may override that bodily desire -- or it may not. I get to decide how important this value is to me. Or I could change my mind. No computer can decide what its values are nor can it change its mind about its values. That's the difference bewteen a determinist view of the world and the LDS view. It may be fine for a Dennett to see us as mere hardware and software. I take it that LDS cannot for we are like God creators and co-creators.  

Comment by Blake | 1/01/2006 11:55:00 AM  

Blake, just to let you know where I'm at... I'm studying the other paper of Hoefer's in order to respond more completely to your second comment on this thread (and indeed this first stage of your plan). It may take a couple days or so. I am willing to do so by way of clarifying the nature of my premise of determinsism and possible associated causal basis. But as I said previously, I'm not particularly interested in arguments against determinism, but rather in the question of why the views of freedom and responsibility that follow from the assumption of determinism are incompatible with Mormonism. 

Comment by Christian Y. Cardall | 1/02/2006 11:18:00 AM  

Christian: Just a matter of housekeeping. It seems that my post on the problem of equating the notion of sufficient causes for action with determinism (as I believe you have done) ought to start this thread rather than your response to it. Could it be moved to the beginning of the thread so that this discussion is more self-contained?  

Comment by Blake | 1/02/2006 12:18:00 PM  

Ideally I would respond to the comments in sequence---as I said I'm still on the first and second comments on this thread---but I'll forego that strict order and try to respond briefly to Blake's last (non-housekeeping) comment.

I'm not saying that humans and expert chess programs are equivalent. I have stated some ways in which they are different in this  post and this comment on that post (and do not claim the noted differences to be exhaustive). In particular I do not dispute that humans have moral responsibility that expert chess programs lack. But I am saying that expert chess programs behave deterministically, and also that these programs can formulate future scenarios and select among them; and therefore, that observation of a capacity for scenario formulation and selection does not constitute proof that determinism does not govern an entity exhibiting said capacity. Qualities unique to humans, and evidence that they are not subject to determinism, must be found somewhere other than a capacity for scenario formulation and selection.

I agree that humans can be computer programmers while expert chess programs cannot. But I hold forth the possibility that humans may be something like computer programs that can reprogram themselves as a response to their interactions with their environment; and further, that both the execution of an individual's current "program" and the process of self-reprogramming unfold deterministically.

Of course I do not claim there is either sufficient empirical evidence or a sufficiently detailed causal account to compel all reasonable people to accept this picture. But I do question whether such a picture is necessarily incompatible with Mormonism. I expect this last question will begin to be explored more fully when I discuss Blake's electrode scenario---which I will do after responding to Hoefer's other article.

On the housekeeping issue, I do think it would be good to have your main points be posts. Let me think about what to do. In terms of undoing what's been done, I'm more constrained with Blogger than I would be with WordPress, but I'm sure we can do something.  

Comment by Christian Y. Cardall | 1/02/2006 12:45:00 PM  

Blake, I'm sending you an invitation to be a blogger here so you can put major sections of your argument up as separate posts as you see fit. Let me know if you have any difficulties. Or if you prefer, you can email me any title and text you want to be a separate post and I can post it in your name.

Because this is an individual blog rather than a group blog with many simultaenous active threads on multiple subjects, I don't think long discussions need or ought to be confined to a single post. Indeed I dislike threads with dozens or even hundreds of comments; they seem to deteriorate fairly quickly and evolve in such a way as to make it more difficult to go back and find useful information later. New posts, and responding post-to-post when appropriate rather than comment-to-comment, allow readers to follow major trends in a long discussion without following all the details of all comments. I also think it helps improve the overall quality of the discussion, since authors are more likely to take care with a post than with a comment, and try harder to advance the argument by saying something they haven't said before. (A more pragmatic consideration is that an individual blog always needs fresh content; new posts catch much more attention than new comments.)

There is some loss of self-containment this way, but necessary connections can be made with short summary statements and key hyperlinks---which, again, help improve the quality of the discussion anyway. Moreover, related posts can be collected in one place and linked on the sidebar to achieve a higher level of contained organization.

So I would like to leave this post of mine as a post rather than demote it to a comment, but would be happy to see your comment it responds to---on the problem of equating the notion of sufficient causes for action with determinism---elevated to a post of its own. Go ahead and make it a post, and after you do I will reset the date on this one to make it appear after yours (I thought that was no longer possible in Blogger but now I see that it still is after all). Or if you're happy with your original comment just as it stands, I'll elevate to a post for you, but I'd need a title. 

Comment by Christian Y. Cardall | 1/02/2006 03:22:00 PM  

Christian: I'll wait until you resond to the causation-determinism entailment on this post before moving on in the dicussion. 

Comment by Blake | 1/02/2006 10:22:00 PM  

Blake, I finished a first reading of Hoefer's paper this evening. I think I will need a second reading so things may take another couple days. 

Comment by Christian Y. Cardall | 1/03/2006 09:05:00 PM  



<< Home