Hylomorphism and the Problem of Mental Causation (Part 1)

The problem of mental causation is a central problem in the metaphysics of mind, but hylomorphism implies an elegant solution to it.

Here’s a version of the problem I discuss in Structure and the Metaphysics of Mind:

  1. Actions have mental causes.
  2. Actions have physical causes.
  3. The mental causes and the physical causes of actions are distinct.
  4. If actions have multiple causes, then they are overdetermined.
  5. Actions are not overdetermined.

Claims (1) – (3) imply that actions have multiple causes, and claims (4) – (5) imply that they do not. The five claims are thus jointly inconsistent: they cannot all be true, and yet there are well rehearsed reasons to accept each.

Because hylomorphists endorse the existence of emergent properties with the characteristics described in my earlier post, they cannot reject claims (1) or (3), nor can they reject claim (2) because they also claim that higher-level behavior never violates lower-level physical laws. If the lower-level constituents of a structured whole were to lose their powers, they would become incapable of composing the activities of the whole. Those activities depend on lower-level items retaining and manifesting the powers they have.

By analogy, it is only because bricks and timbers retain their shapes under compression that they can be recruited as components of buildings. Similarly, it is only because lower-level materials retain their distinctive powers that structured individuals can recruit them as components for their own activities. This is one thing that sets the hylomorphic view apart from those classic emergentist theories such as Roger Sperry’s which claim that higher-level powers trump or nullify the powers of lower-level things.

That leaves claims (4) and (5). Between them, hylomorphists target (4). Claim (4) is true, say hylomorphists, only on the assumption that the mental causes and physical causes are causes of the same type. But hylomorphists reject the assumption that there is only one kind of causal relation, and they take explanations of human behavior that appeal to thoughts, feelings, and perceptions, on the one hand, and explanations of human behavior that appeal to physiological mechanisms, on the other, to correspond to different kinds of causes.

Thoughts, feelings, and perceptions contribute to actions in one kind of way, physiological mechanisms in another. The former contribute insofar as they are ways in which human behavior is structured; the latter contribute insofar as they are things whose activities get structured. Structure in general is different from materials or events that get structured—that’s what it means to say that structure is a basic ontological and explanatory principle.

Accordingly, reason-based explanations of human behavior and mechanistic explanations of human behavior pick out different kinds of causal relations on the hylomorphic view. As a result, they cannot overdetermine actions. Overdetermination in the relevant sense implies that the overdetermining causes contribute to their effects in the same way—paradigmatically each is by itself a sufficient cause. But that is not the case here. By analogy, high blood-alcohol levels and faulty brakes both contribute to car crashes, but they are not overdetermining causes of crashes; the reason is that they contribute to crashes in different ways. There is thus no threat of actions being overdetermined by thoughts, feelings, and perceptions, on the one hand, and by physiological mechanisms, on the other. Claim (4) is thus false on the hylomorphic view.

This distinction among kinds of causal relations enables hylomorphists to rewrite claims (1) – (5). To appreciate how, it will be helpful to coin some terms to express the different ways in which mental causes and physiological causes contribute to behavior. Let us say that thoughts, feelings, and perceptions rationalize actions, and that neural events trigger the muscular subsystems involved in actions. Claims (1), (2), and (3) can then be rewritten as follows:

(1′) Actions are rationalized by thoughts, feelings, and/or perceptions.

(2′) Muscular contractions are triggered by events in the nervous system.

(3′) Rationalizing causes and physiological triggers are distinct.

These claims are jointly consistent, and hylomorphists can endorse all three. Claim (5) can be rewritten in a variety of ways. Here are two possible formulations:

(5′) An action does not have more than one rationalizing cause.

(5′′) An action does not have more than one physiological trigger.

Both of these claims as well as their denials are consistent with (1′)–(3′). Consequently, whether hylomorphists endorse (5′) and (5′′) or deny either, they end up solving the problem.

One thing to note about the hylomorphic solution: it implies that the arrow diagrams made popular by Jaegwon Kim for discussing the problem of mental causation are misleading (Figure 1).

Screenshot 2016-05-12 08.43.44

Kim’s diagrams have been enormously influential, but according to hylomorphists, they misrepresent human action in at least two ways. First, they suggest that there is only one kind of causal relation which is represented by the arrows. Second, the arrows suggest that the kind of causal relation involved is a triggering relation; they do not capture the distinctive character of rationalizations. A better representation of actions depicts physiological conditions, along with social, psychological, and environmental ones, as necessary factors that contribute to the occurrence of an action in different ways (Figure 2).

Screenshot 2016-05-12 08.43.56

On the hylomorphic view actions are complex structured phenomena which comprise many different causal factors that contribute to their effects in many different kinds of ways. No one of these factors can qualify as the exclusive cause of the action. Consequently, thoughts, feelings, and perceptions, on the one hand, and physiological mechanisms, on the other, do not compete to occupy a single causal role: the role of being the one and only cause of an action. As a result, there can be no real worry about overdetermination even though the view comprises many causal factors.

In the next post I consider one of the most commonly heard objections to the hylomorphic view.


  1. I cannot think of a single example of a real-world event that does not have multiple causes. Everything is simultaneously interacting with everything else. So I cannot see why proposition 4 would *ever* be considered true for actions, since actions are a subset of events.

    Also I’m not clear on the distinction you are making between mental causes and physical causes. Earlier you said your view was substance monist. Here you seem to be implying that you in fact accept a substance dualist view and that mental causation is a different species of causation because mind and matter are somehow different. How can you justify the distinction otherwise?

    On the other hand if you are going to allow for analysis of different types of causation, I see no reason at all to unify “thoughts, feelings, and perceptions” as a single type of causative events.

  2. William Jaworski

    Hi Jayarava,

    Thanks for your comments. There are several things to say here. Let’s start with this statement: “I cannot see why proposition 4 would *ever* be considered true for actions.”

    Claim (4) is an implicit assumption of most discussions of the problem of mental causation. It’s motivated by the further assumption that causes are sufficient conditions of some sort. Given this assumption, saying that an event has multiple causes implies that the event has multiple sufficient conditions. Now, suppose further that in the case of action the causes are independent of one another (a supposition that many take to follow from (3) given reasonable assumptions). In that case, it follows that the action has multiple independent sufficient causes, and if that’s true, then the action is overdetermined. This combination of assumptions, or something very much like them, is why many people take (4) for granted.

    It’s great to hear you’re on board with causal pluralism: “I cannot think of a single example of a real-world event that does not have multiple causes.” I’m not exactly sure how to take the next statement however: “Everything is simultaneously interacting with everything else.” It’s not evident to me how, say, Julius Caesar could be interacting right now with anything that currently exists, nor in general how it could be true that all causal interactions occur simultaneously. Perhaps I’m missing something.

    I don’t altogether understand this comment either: “Here you seem to be implying that you in fact accept a substance dualist view and that mental causation is a different species of causation because mind and matter are somehow different.” I’m not sure I understand where the substance dualistic implication is supposed to come from.

    Suppose mental causes and physical causes are distinct. If causes are events, this implies that mental and physical events are distinct. If events are individuals having properties at times, then event A = event B exactly if A and B comprise the same individuals, properties, and times. E.g. a’s being F at t = b’s being G at t* exactly if a = b, F = G, and t = t*. Consequently, the distinction between mental and physical causes also implies that mental and physical properties are distinct.

    What’s difficult to appreciate is how the distinctness of mental and physical properties is supposed to imply the further distinctness of mental and physical substances or individuals. Dual-attribute theories of all sorts, whether they be emergentist, epiphenomenalist, or hylomorphic, reject any such implication. All of them claim that numerically one individual can have mental properties and distinct physical ones.

    You suggest that without a commitment to substance dualism there is no way to support the distinction between mental and physical causes. You ask, “How can you justify the distinction otherwise?” I’ve answered this question in part already: given the aforementioned assumptions about causes and events, the distinctness of mental and physical causes could be established by demonstrating the distinctness of mental and physical properties. There are many arguments that purport to do this without any recourse to substance dualism.

    Probably the most common arguments of this sort appeal to multiple realizability (for a survey of the literature see http://www.iep.utm.edu/mult-rea/). But there’s also Kripke’s argument, a variety of qualia-based arguments – even Davidson’s argument for the anomalism of the mental and Max Black’s objection to the psychophysical identity theory can be parlayed into arguments for property dualism. None of these arguments imply a commitment to substance dualism.

    For hylomorphists, the most obvious arguments for property dualism appeal to things like the DNA molecule discussed in the previous post. These are things that manifest certain properties irrespective of their environments, and that manifest different properties when they are parts of structured wholes. There are thus structure-dependent properties and structure-independent properties. Properties that are typically classified as mental, such as believing that two and two are four, feeling ecstatic, and seeing the half-full glass of bourbon, are all structure-dependent properties. Their structure dependence is one of things brought out by the squashing example: the squashed contents of the bag can’t think, feel, or perceive.

    There are, in short, many ways in which dual-attribute theorists can support property dualism (and a corresponding distinction among events) without any recourse to substance dualism.

    The discussion so far has been framed in terms of a mental-physical distinction, but in point of fact, one implication of the hylomorphic framework is that there is nothing canonical about this distinction. I argue in the book that how we choose to draw the distinction – or whether we choose to draw it at all – is entirely a function of our descriptive and explanatory interests.

    Hylomorphists look to understand our activities and capacities – including thought, feeling, and perception – in terms of the structured manifestations of powers. Whether we decide to call some of these powers ‘mental’ or ‘nonmental’, ‘physical’ or ‘nonphysical’ is orthogonal to the project of understanding what they are, and why and how they operate as they do. Consequently, hylomorphists are free to accept or reject one of any number of different ways of defining ‘mental’ and ‘physical’.

    Philosophers such as John Searle (Rediscovery of the Mind, pp. 26, 54-5) have claimed that the way to solve mind-body problems is to reject the mental-physical dichotomy. It’s somewhat disappointing, then, that when it comes to formulating his own theory, Searle employs the same dichotomy he disparages. Yet this should really come as no surprise: without developing a metaphysical framework that implies an alternative approach to understanding human activities and capacities, neither Searle nor anybody else has much of a choice when it comes to accepting the mental-physical dichotomy.

    Hylomorphists nevertheless do have a choice because theirs is a metaphysical framework of exactly the needed sort, one that implies an alternative approach to understanding human activities and capacities that doesn’t depend on the mental-physical dichotomy.

  3. “…then the action is overdetermined. ”

    You don’t say why this is a problem that would cause you to reject it as a possibility.

    “Suppose mental causes and physical causes are distinct. ”

    I’m reading Richard H. Jones’s excellent introduction to the reductionism/antireductionism divide in Western philosophy, “Analysis & the Fullness of Reality.” So, I get that you are arguing for substance reductionism and structure antireductionism, i.e. that you are arguing that structures are real and thus capable of being effective causes.

    I’m not convinced that the legacy concepts of “mental” and “physical” are as distinct as you want them to be. And indeed, you seem to waver, arguing both for an against them in your response to me.

    All those so-called mental properties are strongly correlated, in every case, to the physical structure and functioning of the brain. The mental/physical distinction is a matter of perception, not of being; i.e. epistemic, not ontic. Any causation attributed to the mind is in fact due to the physical operation of the brain, but the perspective on events varies depending on which senses we use to perceive events (some we think of as mental and some physical, but in reality it’s all just our brain working). This isn’t just a theoretical problem, we have plenty of evidence to help understand it. In theory there could be an ontic distinction between mental and physical (allowing for dual causation), but in practice there is no ontic distinction. So we’re back to one kind of stuff and perhaps a handful of kinds of causation at a few different levels: fields, particles, atoms and molecules, naked sense objects, planets and galaxies, cosmic.

    Which is not to say that structure antireductionism is not required to understand the universe – else the distinctions between quantum field, quantum particle, classical, and relativistic descriptions of the universe would not be necessary.

    To argue that every attribute required it’s own *level* of description would not simplify matters. If every complex object manifested fundamentally unique properties that could not be understood in terms of the level of reality that it existed at, then we would need infinite numbers of descriptions of the universe because each object and each event would be wholly unique. Crucially no two people would be covered by the same description of reality – each, being a unique structure, would manifest distinct forms of being and causation. Every instance of being would require its own unique ontology. Far from producing a dual-attribute theory, once you open the can, it turns into an infinite-attribute theory and all useful information gets averages out to zero, because anything divided by infinity is zero.

    This is not the case, however. We can treat the universe as having a finite number of levels of organisation (somewhere between 4 and 10 I think) with some blurring at the boundaries (such as a sixty-carbon buckyball molecule exhibiting quantum particle behaviour).

    Another problem with the idea of “mental” causation, is that we barely understand what it means in our own case; but we have almost no idea what it means as we look at other animals with brains. And it gets less clear the smaller the brain gets. Where is the cut off for “mental causation”. Does it require consciousness? Or not? Most of our actions are motivated by unconscious events, and most of the time we operate with multiple conflicting desires (everything is over-determined to use your phrase).

  4. tnmurti

    I understood the argument upto a point only.
    What I understood was that
    1.Structures exhibit causal powers.
    2.There are lower level functional structures and higher level organisational structures
    which are experienced as thoughts etc.within brain.
    3.Changes in organisational structures constitute thought movements.
    4.Same organisational structure is cause to action as well as experienced mental cause.
    I miss to understand the ability and independance of mental cause.

  5. William Jaworski

    Hi tnmurti

    Thanks for your comments. It can be really difficult to grasp a complicated theory when all you’re given to look at is a blog post. The view is presented with much greater patience and clarity in the book.

    With that in mind, let me try to redirect things a bit since the points 1-4 you list don’t really reflect the hylomorphic view.

    Re 1: Structures don’t exhibit powers. Structures ARE powers. The things that exhibit powers are individuals. Individuals are the things that HAVE powers, and hence they are the things that manifest or exhibit powers.

    Re 2: The view doesn’t draw any distinction between functional structures and organizational structures. A structure just is a form of organization: ‘organization’ and ‘structure’ in this context are synonymous. In addition, there is no claim to the effect that a structure would be experienced in the brain. It’s not even clear what such a claim could mean within a hylomorphic framework.

    Re 3: Changes in structures don’t constitute thoughts. Thoughts are activities in which individuals like us engage. We engage in those activities by imposing an order on the way our parts manifest their powers. In that sense, thinking is no different from throwing a baseball. We engage in the latter by coordinating the way our muscles and nerves along with things in the environment (such as the baseball) manifest powers they have. Thinking is the same kind of activity: we engage in it by coordinating the way our parts and things in the environment manifest powers they have.

    Re 4: Structures aren’t standalone causes of actions; neither are thoughts and feelings. Rather, actions are composite events – structured activities of the sort I described just now. Each action is composed of subactivities of our parts coordinated in a specific way. When we say, “Gabriel went to the store because he wanted to ensure he had enough wine for the party,” the action we’re describing is composed of Gabriel’s parts and things in Gabriel’s environment manifesting powers they have in a coordinated way.

    • tnmurti

      I thank you for the reply.
      I understand my views did not fit into the scheme described.But they appeared nearer to
      Basically I think neuronal Structures are self-organising entities.These structures are
      made-up of our parts and environmental things represented in brain.They have causal powers
      which are physical in nature.This causal power is felt as mental cause.So no cause originates
      in mind.
      On further readings I may get a different understanding.
      Thanks once again.

Comments are closed.