A War for Chance?

Posted by

Jeffrey Thayne

Agency is arguably one of the most crucial doctrines of the restored gospel. Richard Williams, in his paper “The Freedom and Determinism of Agency,” claimed that “there is a good case to be made for the centrality of agency to any understanding of the meaning and purpose of the Plan of Salvation and the Atonement of Jesus Christ which is its essential core. After all, if we were not moral agents, no atonement would be necessary because we could not really sin.”1 Now, I do not intend to defend this claim in this series, but rather I would like to devote this post to a discussion of what agency is… or, more precisely, what it is not.

Many defenses of the philosophy of agency have failed, Williams argues, because they associate the doctrine of agency with the philosophy of indeterminism. Mechanical or necessary determinism is the idea that all events have mechanistic causes, such that the event was inevitable given the circumstances. Its opposite, as usually understood, is indeterminism, which is the idea that events are completely random and unrelated to events that precede it.

Either view, however, makes for meaningless events, either because they are inevitable or because they are random. Neither allows for agency or meaningful actions. I agree with Williams that agency and indeterminism are not only very different philosophies, but also incompatible philosophies. In other words, contrary to popular rhetoric, agency is not indeterminism, nor is it the freedom to act capriciously or arbitrarily.

Why is agency so often associated with indeterminism? Because at least one form of determinism that is popular in today’s scientific world is inimical to agency. This is the kind of determinism found in biological and reductive naturalism. In my post “Shackled by Determinism,” I show how scientists have applied this kind of mechanical determinism to human behavior. This kind of mechanical determinism (a widespread assumption in modern science), when applied to human beings, is clearly incompatible with agency. “We cannot embrace both reductive, biological naturalism and human agency,” Williams claims—at least when applied to human behavior. Williams continues, “Since both conceptual dimensions [agency and indeterminism] are anchored on one end by determinism, it is common to conflate freedom and indeterminism since they both find themselves in opposition to determinism.”4

This tendency to associate agency with indeterminism can be seen in various places inside and outside of Latter-day Saint scholarship. For example, Henry Eyring*, a famous Latter-day Saint chemist, explains:

Recently, we have been obliged to give up the old determinism of classical mechanics … Mechanical determinism meant that if one were given the sate of the universe at any instant of time, a sufficiently expert mathematician could calculate the state of things at all times to come. This left no place for the great religious principle of free will. Then quantum mechanics brought with it the uncertainty principle. This principle eliminates the possibility of predicting the future exactly, and tends to confirm that fundamental Christian tenet that man enjoys free agency as a divine gift.2

Whether he intended to or not, it is clear that Eyring treats agency, indeterminism, and unpredictability as though they are the same thing—or at least as though agency requires a kind of indeterminism to work. Sir Henry Jeans, a philosopher of science at Cambridge University early last century, made the connection even more direct:

There were … two schools of thought—the determinists who maintained that all events, including human acts, were causally determined and so compelled by past events and acts, including such events as those of heredity, environment, acquired habits and so forth; and the indeterminists who maintained that human acts are not entirely determined by the past, but that at every moment we can exercise a certain amount of guidance through a fiat which is our own.

On a determinist view, a man’s actions would of course be completely predictable in principle by one who had a sufficiently intimate knowledge of his nature, of his past and of the character he has acquired in the past. On the indeterminist view, this is not so; a man can falsify all predictions by a capricious, and so unpredictable, choice.3

Thus, Jeans clearly associates free will with both unpredictability and capriciousness. It is clear to see why, in a science that is dedicated to finding ways to predict events in the world, agency or freedom would seem very unscientific. For this reason, many psychologists ignore the reality of agency in their research.

Jeans, however, explains why this indeterministic point of view is unsatisfactory when used to account for agency:

If, as most people would say, [someone] moves … in a particular direction because ‘he chooses to’, the question is why he chooses this direction rather the other. If something determines his choice, we are back to determinism; if nothing, he acts from pure caprice, and this leads to a free-will which is neither of the kind we want to find nor of the kind we feel we do find. … Capricious indeterminism [does not] give us a free-will at all resembling that of our experience or imagined experience.3

That is, to the extent that our choices are random, arbitrary, or not meaningfully connected with antecedent conditions, they are meaningless. I do not want to imply that all choices are predictable, only that they are not random or capricious. It is only when we equate unpredictability with randomness and arbitrariness that we fall into the trap of indeterminism. For this reason, Williams also believes that indeterminism cannot be a proper account of agency:

Indeed, many textbooks in the social sciences—if they deal with agency at all—deal with it as a species of indeterminism. This makes agency unscientific, mystical, and indefensible, totally incompatible with an empirical scientific psychology. And I must agree that they are right about that. This is also why agency has been so hard to defend—and thus why my side loses so many arguments about agency. Indeterminism is an indefensible position. It must ultimately hold that there are no causes of actions or events – that is, no strong ties between events and their antecedents. Any consistent indeterminism must hold, therefore, that events are ultimately just random.

This equation of agency with indeterminism is problematic on two accounts. First, even casual observation is sufficient to persuade us that events in the world are not seem random. There are indeed strong relationships between events and their antecedents. It also seems like a weak argument to hold that while most or all of the physical/observable world is deterministic, human action is not. Our observations, not to mention our reflective sense of ourselves tells us that our actions are indeed tied to meaningful antecedents.

This leads us to the second problem. It is that, even if it were true that indeterminism rules in the sphere of human action, this would not allow for any meaningful agentic human action. If human actions have no antecedents, then they have no context, they reveal no order, and tend toward nothing and have no meaning. This is not the sort of agency that any rational human being might want. If, as our scriptures and doctrine suggest, a war was fought in heaven over the issue of agency, it was a senseless war if the outcome is merely the right and privilege to behave randomly and without regard for history, context or even desire.1

Indeed, I do not believe that agency is the freedom to make choices that are completely unattached to antecedent conditions. This view of agency makes agency very hard to defend in the scientific arena, and puts it in the same position as the “God of the gaps” perspective of divine intervention—when we cannot account for a human action in terms of biological, hereditary, or environmental factors, only then can we invoke agency. In other words, if we treat agency as indeterminism, the more we explain, the more we learn about biological and environmental influences on human action, the less agency we have.

Thus, Williams and others have proposed that we discard the notion that agency is a form of indeterminism, and re-conceptualize how we talk about agency. Williams still maintains that “mechanical and biological links are clearly destructive of agency, as are stimulus-response links governed by environmental forces requiring no active participation by an agentic person.”4 However, he invites us to reconceptualize the alternative as a different kind of determinism (rather than indeterminism), a kind that will preserve agency in all of its richness in human experience. In a later, I will explain how he proposes to do this.

*This is not referring to Elder Henry B. Eyring, the Counselor in the First Presidency.
1. Richard Williams, “The Freedom and Determinism of Agency,” Brigham Young University. Emphasis added.
2. Henry Eyring, “Science and Faith,” included in Science and Your Faith in God (Salt Lake City: Bookcraft) 1958.
3. Sir James Jeans, Physics and Philosophy (Cambridge University Press) 1942.
4. Richard Williams, “Agency,” included in Jackson, A. P. and Fischer, L. (eds.) Turning Freud Upside Down (Provo: BYU Press).


  1. Just a terminological note. The way the term “indeterminism” is now used is different from how James is using it. It’s usually used to simply mean not-determined. Thus it can embrace ontological randomness, agent libertarian free will, and many other possible positions.

  2. I assume you are talking about William James? I am aware that his use of the word is somewhat different. William James’ views on free will, however, is a little outside the scope of this post.

    I’m aware that the term can be used in various ways, and that is perhaps the point of my post. It is clear that ontological randomness is often believed to open doors for free will, and as both are sometimes referred to as indeterminism, the conflation becomes even more likely. I believe this conflation of ontological randomness with meaningful, purposeful action is problematic, and perhaps we ought to disambiguate the term “indeterminism” so that we avoid the problem. Richard Williams approaches the problem by defining indeterminism as the philosophy that events “have no meaningful antecedents” and necessary determinism as the philosophy that events “have antecedents given which the event is inevitable.” In a future post, I’ll explore his version of minimal determinism, which holds that events “have meaningful antecedents, without which they would not be what they are” but which do not make the event inevitable.

  3. I think that ontological randomness is attacked via luck arguments. I’m pretty convinced by Mele that these don’t ultimately work, if only on epistemological grounds.

    The problem is that we have to be careful not to get caught up in a hidden false dichotomy. Between randomness and agent libertarianism there may be other choices. (Indeed I tend to think there are – and even pick one for my likely answer to the problem although I’m not exactly committed to it)

    The problem of invoking meaning is legion. But I’ll stay out of that issue. Typically in my view “meaningful” entails “corresponds to intuitions I like.” I’m not sure that’s fair, although most debates about free will end up hinging upon competing intuitions.

  4. Williams would say (at least, he has said when he’s guest lectured in my classes) that by “meaningful antecedents” he means “relevant antecedents.” That is, all human events have antecedents that are relevant to those events. Thus, he does not mean by “meaningful” something corresponding “to intuitions I like.”

    By the way, as long as we’re trying to clarify what folks mean here, what do you (Clark) mean by “agent libertarianism”? My experience has been that this term is used in various (and often ambiguous) ways.

  5. I think that in the distant past it was ambiguous. I think the past decade or two has clarified it. Roughly it is the idea that there is some irreducible agent that must do the choosing. The agent must be a source of cause but can not have those causes fully determined by anything other than the agent’s original choice. Further the choice can not be a matter of chance or luck. Finally the agent must have alternative possibilities. (Meaning that if the clock was replayed a different result may occur)

    The irreducible agent part is the tricky bit. It’s no problem if you believe in a Cartesian mind/soul. If you don’t though it’s kind of difficult. One new alternative from O’Conner and someone else whose name escapes me at the moment is the idea of radical or ontological emergence. That is akin to typical emergence (i.e. water from the laws of physics and hydrogen and oxygen atoms) except that in radical emergents the properties of the emergent substance can’t be explained by reduction to the constitutents plus the laws of physics. I’m pretty skeptical of that. Blake Oster ends up adopting a variation of radical emergence although his is different from O’Conners due to adopting a fairly complex ontology out of process theology.

    Regarding meaningful. Thanks for that clarification. I’m still not sure what relevance means here. I assume he means something beyond causal interactions. It seems this may be a bit of a muddle. We can, after all, talk about something being relevant hermeneutically. But relevance is always a tricky endeavor. (Heck, even causality ends up being more problematic than we think)

  6. Over on “Interpreting Sexual Attraction,” Capt. Moroni cited an interesting article that I thought I’d post here, since it’s more pertinent to this topic:

    John Conway & Simon Kochen, “The Free Will Theorem,” Princeton University, Department of Mathematics, accessed on the Cornell University Library website (arXiv.org), 2 Apr. 2009.

    One passage was very pertinent to Jeff’s article:

    11.2. Free versus Random? Although we find ourselves unable to give an operational definition of either “free” or “random,” we have managed to distinguish between them in our context, because free behavior can be twinned, while random behavior cannot (a remark that might also interest some philosophers of free will). Bassi and Ghirardi remark that it follows from Gisin’s theorem that their “hits” must involve a stochastic element in order to make the GRW theory relativistically invariant. We have shown that what the hits really need is some freedom (to be precise, that they must be at least semi-free). It is for reasons including these that we prefer to describe our particles’ behavior as “free” rather than “random,” “stochastic,” or “indeterminate.”

    In the rest of the article, I’ll admit that much of the physics jargon and formulas were over my head, but I got the gist of it. The basic premise (as summarized by this physics dunce) seems to be, if we assume the scientists studying the particles are free enough to choose various experimental conditions, why not assume the particles themselves are free enough to choose their reaction to the experimental conditions?

    As Capt. Moroni pointed out, this idea has some interesting parallels to a talk by W. Cleon Skousen, “A Personal Search for the Meaning of the Atonement.”

Leave a Reply

Your email address will not be published. Required fields are marked *