This question represents the second I have asked in relation to an argument I posted here approximately a year ago, which aims at a proof for the impossibility of free will.

It is inappropriate to attach this question to the initial post, as it would constitute a distinct third question (find the second question here), confuse the answer stream, and miss out on the potential consideration of many users who have already visited the original post over the past year and are unlikely to encounter any addition to it. This question also seems to constitute an interesting question in its own right, and to therefore warrant this dedicated post.

The premise at stake is:

  1. A decision is an act. Therefore, in order for a decision to be voluntary, a person must decide to decide it.

(An act is defined here as "A thing done").

I am motivated to ask this question after email communication with an internationally-noted philosopher of mind (not identified here, as I haven't obtained consent to do so). I presented the argument in question, and they happened to agree that - if only in their opinion - it was sound, but remarked that to define a question as an act is philosophically controversial. I'm honestly astonished they took the time to respond to my email and I don't want to pester them with follow-up questions. Regardless, I'm raising the question here because I'm interested in the diversity of (informed) response that Philosophy Stack enables, and primarily because any strong refutation of decision as action would likely prove fatal to my (long-cherished) argument.

The question is:

Where an act is defined as 'a thing done', how might a decision be defined if it is to be defined as something other than an act?

Kitajima & Toyota (2013) state, "Decision-making is the act or process of choosing a preferred option or course of actions from a set of alternatives".

McCall (1987) raises this very question, but I don't have access to the entire article to examine any answers he provides.

"Is a decision an action?", he asks. "If so, what sort of action? Must decision be preceded by deliberation?".

Any insight into this realm would be most welcome, whether as answers, quotes and/or references.

  • 1
    Decisions may well be mental acts, but the argument still fails. The problem is with your first premise that encodes what is called volitionism: "In order for an act to be voluntary, a person must decide to perform it". Hacker, following Ryle and Wittgenstein, disagrees:"When one utters a sentence, every word is spoken voluntarily, but it would be ridiculous to claim that one consciously performs successive acts of will, one for each word (or phoneme?) an instant before utterance".
    – Conifold
    Aug 12, 2022 at 14:17
  • 1
    The conclusion of your argument is - "insofar as free will requires the ability to make voluntary decisions, free will is impossible". It may well require such ability and be possible if some acts do not require exercising that ability. The regress only comes if all voluntary acts are alike, but psychological studies tell otherwise. We obviously make decisions sometimes and then act, sometimes we just act, and sometimes we then confabulate "decisions" after the fact. Perhaps decisionless voluntary acts draw on our ability to decide on maxims, as Kant thought. Or perhaps it is chicken-and-egg.
    – Conifold
    Aug 12, 2022 at 14:56
  • 1
    As Hacker's example shows, "if "sometimes we just act", we act involuntarily" is very plausibly false, and false presuppositions do not contribute to arguments. The volitionist regress fails for the same reasons as Wittgenstein's regress of interpretations, Ryle's regress of thoughts, and similar examples, and not because decisions are non-acts. If you discard the first premise there will no need to look for a fault in the second.
    – Conifold
    Aug 12, 2022 at 15:25
  • 2
    Not if you want to pinpoint the flaw in the argument, but perhaps I misunderstood the question's motivation. Btw, McCall concludes that decisions are mental acts after identifying four tests: can answer what is done, can be tried, can be used in imperatives and combined with adverbs. Few others qualify:"Besides decide, choose, and deliberate, which form a cluster, I have been able to find only one other rather small and unimportant cluster, calculate, count and compute, together with one large and important one containing verbs of imagination, creativity and inventiveness".
    – Conifold
    Aug 12, 2022 at 15:48
  • 2
    Your original argument is similar to Strawson's regress argument, of which SEP has some discussion with references. See also O’Shaughnessy's paper in Mental Actions. He also identifies "willing" as a mental act, but it "differs from bodily action in a fundamental respect... If I voluntarily talk (inwardly) to myself or imagine raising my arm, there is no distinction between my act of willing and an event my willing produces. Rather, in such cases, the willing just is the acting".
    – Conifold
    Aug 12, 2022 at 16:16

3 Answers 3


Here is a selection of references to make them more visible than in the comments.

That decisions are mental/intentional acts is a very common position. McCall's paper in the OP argues for it, and Shepherd in Deciding as Intentional Action: Control over Decisions says so before presenting a challenge and an alternative:

"Common-sense folk psychology and mainstream philosophy of action agree about decisions: these are under an agent's direct control, and are thus intentional actions for which agents can be held responsible. I begin this paper by presenting a problem for this view. In short, since the content of the motivational attitudes that drive deliberation and decision remains open-ended until the moment of decision, it is unclear how agents can be thought to exercise control over what they decide at the moment of deciding. I note that this problem might motivate a non-actional view of deciding—a view that decisions are not actions, but are instead passive events of intention acquisition."

The regress argument that motivates the question is old, and goes back at least to Spinoza, see What counters are there to Spinoza's argument that acts of free will create infinite regress? and Hobbes. In recent times, a similar argument was proposed by Strawson, see SEP and Davidson, with opposite aims. Strawson argued against free will, and Davidson against the "separate act of willing" premise, see Pettit, Deliberation and Decision:

"Does every action originate in a decision to perform that action? It cannot do so if decision is itself an intentional action: if it is a mental act, as some have taken it to be, in which an agent resolves uncertainty about what to do in a given context. For if action were always supposed to originate in decision, and decision were itself an action, then we would face a regress. The regress would be similar to that which Donald Davidson (1980) invoked in criticism of the idea that every intentional action must originate in an act of will, repeating – without apparently being aware of it – a point that Thomas Hobbes (1994: 125) had made in 1640: “a man can no more say he will will, than he will will will, and so make an infinite repetition of the word will.”"

Even authors who take decisions to be mental acts distinguish some of them from bodily acts in ways that block the regress. It has the same effect as denying that all voluntary acts must be preceded by decisions, as Hacker does, following Ryle and Wittgenstein. In both cases, decisions to act (mentally) can be events identical with the acting itself. For example, see O’Shaughnessy's paper in Mental Actions volume:

"Brian O’Shaughnessy’s “Trying and Acting” argues that there is at least one species of mental action that differs from bodily action in a fundamental respect. According to O’Shaughnessy, when we assert ‘A did x’, where x is a bodily action, we imply that there was an event which was “the active generation of x,” an act of willing or trying which is not identical to, but rather the cause of, A’s x-ing. Something similar holds for certain sorts of mental action: if I try to remember a name, and succeed, then my remembering the name is presumably an event caused by my trying to remember.

But, O’Shaughnessy maintains, there are also kinds of mental action to which this analysis does not apply. If I voluntarily talk (inwardly) to myself or imagine raising my arm, there is no distinction between my act of willing and an event my willing produces. Rather, in such cases, the willing just is the acting. So, O’Shaughnessy concludes, not all willings are tryings-to-produce; we must leave room for a form of willing which is internally, non-productively active."

  • 1
    @Futilitarian Not if we say the decision is one way to describe an aspect of the event, namely the ascription as one of my own volition. The philosophical issue here is that a plurality of discriptional layers doesn't make a plurality of ontological realities. A decision is, maybe, really nothing more than a rationalisation (linguistic representation) of certain events, after the fact.
    – Philip Klöcking
    Aug 12, 2022 at 17:34
  • 1
    @Futilitarian No, you pretty much nailed it. That is, not necessarily since we could play that game further and say that since no domain of discourse can claim inherent superiority in terms of epistemological access to what or what not really is the case, all we end up with is different domains of discourse. If we take this seriously, everything that has practical consequences (shapes how we interact with our environment) is to be called 'real' with equal credence. No matter whether we talk about religion, free will, science, or whatever.
    – Philip Klöcking
    Aug 12, 2022 at 17:49
  • 2
    @Futilitarian It is a common theme between Ryle, Wittgenstein, Davidson, O’Shaughnessy, etc., etc., whether it is wrapped into special kinds of acts, non-acts or voluntary acts w/o decisions. I have the same impression, for what little it is worth. That action has to be stamped by deliberations to be voluntary strikes me as old school over-rationalism, and that decisions stand in the same separable relation to mental acts as they do to physical acts, as regress arguments assume, strikes me as loose analogizing. I doubt people's ability to track "deciding to decide" introspectively to tell.
    – Conifold
    Aug 12, 2022 at 17:58
  • 1
    @Futilitarian In the pragmatic philosophy, everything that has practical consequences is real. It makes a difference to you whether God exists or you got free will? They obviously bear some reality. Does that make reality solipsistic, everyone having their own one? No, it is and never was anything more than a social construct. What we have to stop thinking is that there was any ultimate measure of ultimate 'capital-concepts' like Truth and Reality. There is human life interacting with its environment. Full-stop. All we can do is trying to frame that in words that make sense to us.
    – Philip Klöcking
    Aug 12, 2022 at 18:03
  • 1
    @Futilitarian I think we do both. I can think of many philosophers, and not, adopting positions, to stay consistent, that are very likely at variance with their preferences, or perhaps they sacrifice some preferences to others. And I think that free will, to some, is such an intuitively preferred stance that arguments against it become reductios of their premises.
    – Conifold
    Aug 12, 2022 at 18:20

(instead of a comment)

The question and its related questions center around an assumed split between "decision to act" and the "act" itself.

First of all, there is no a priori reason for a decision to act and the act to be fundamentally different and not aspects of the same thing nor be time separated and not happen simultaneously.

To elucidate let's perform a thought experiment:

  1. we agree that a mental thought is also an act.
  2. try to spot a difference or split between deciding to think of something and actually doing it, thinking it.

The previous gedanken experiment suits the preconditions of the question while making clear that decision to act and the act are not necessarily fundamentally different nor are necessarily time separated.

If decisions are indeed real, this would be a natural way to be since if otherwise could lead to unnecessary difficulties and infinite regresses.

The problem of infinite regress is present even if decision is taken as unconscious and not willed. Again if decision to act and the act itself are fundamentally separated, the Impossibility lurks. So regardless of free will, decision to act and the act cannot be fundamentally separated

That being said, when speaking we usually confuse the meaning of the word "deciding" with that of "deliberating". "Deciding" is at the same time the initiation of the action. What happens before is "deliberating".

Is "deciding" an "action"? According to this post however one might call "deciding" it is inseparable from an "action". Whether it is the same as the action, or whether it is simply a different aspect of the decision/action pair, or whether it simply cannot be separated in any meaningful sense from the associated action, I will leave it to the reader..

PS: The same problem of infinite regress plaques a standard interpretation of QM, ie collapse of the wavefunction. At some point t there is a state of affairs between collapsing the wavefunction at that point or not. But this state affairs itself can be formulated via another wavefunction which now has its own collapse to handle, ie either collapsing to collapse or not, via a third wavefunction and so on.. Obviously this interpretation accepts that collapse happens without regress and without higher-order wavefunctions. It happens irreducibly to further mechanisms.

  • 1
    Yes. As has been pointed out by others across some of my questions, I need/ed to do more reading on volition to get a better understanding of different theories of decision/action/will etc. I'm not sure about your thought experiment though. "Try to spot a difference or split between deciding to think of something and actually doing it" is not very helpful giving the potential time periods involved. I still struggle to think of a decision being part of the action that is being decided upon, but maybe that's just an argument from ignorance on my part. Not sure. Dec 29, 2022 at 2:28
  • 1
    Here is a related link Conifold sent me.Deliberation and Decision by Petit. Dec 29, 2022 at 2:48
  • For infinite regress to happen, two premises must be satisfied: a) decision must be such that can be meaningfully separable from its assoiated action and b) that this separation has a time delay. if not both are satisfied then no infinite regress can happen or even if it does it is harmless. If decisions are indeed real then there is absolutely no reason for any time delay between taking a decision and starting to act, that would be the natural way, simultaneity. If this is the natural way, then we come to the conclusion that they cannot be separatable in any meaningful sense.
    – Nikos M.
    Dec 29, 2022 at 8:49
  • The problem of infinite regress still lurks if decision to act and the act are fundamentally separated, even if decision is unconscious and not willed. Again we come to an Impossibility. So decision and act cannot be fundamentally separated, regardless of free will. (updated answer)
    – Nikos M.
    Feb 14, 2023 at 20:12
  • For example the same problem of infinite regress plaques a standard interpretation of QM, ie collapse of the wavefunction. At some point t there is a state between collapsing the wavefunction at that point or not, but this state itself can be formulated via another wavefunction which now has its own collapse to handle, ie either collapsing to collapse or not, via a third wavefunction and so on.. Obviously this interpretation accepts that collapse happens without regress and without higher-order wavefunctions. It happens irreducibly to further mehanisms.
    – Nikos M.
    Feb 15, 2023 at 7:01

Decision-making is a mental process, not a physical one.

A decision is the result of that mental process. A decision is knowledge about what the deciding agent is about to do.

A decision is not an action. Actions are physical. The mind makes decisions, the muscles perform actions.

  • 2
    So.. following from your answer, how would you define a decision? As simply as "a mental process"? Remember, a decision is demonstrably not equivalent to "knowledge about what the deciding agent is about to do", because a person can change their mind after a decision and act against what they decided to do. Aug 12, 2022 at 14:23
  • A decision is exactly knowledge about what the agent is about to do. Decisions can be discarded and replaced by a new one before they are implemented. A decision is the result of the mental process. Changing one's mind is just a part of that process. Aug 12, 2022 at 17:08
  • Your first sentence contradicts your second. Aug 12, 2022 at 17:09
  • No contradiction. A decision is not complete before it is implemented. A decision is exactly knowledge, not exact knowledge. A decision is just a plan for action. Aug 12, 2022 at 17:24
  • 2
    A decision is not implemented by itself. It is implemented by the action that was decided upon. As you state in your final sentence, "A decision is just a plan for action". Aug 12, 2022 at 17:28

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .