Atypical Rational Agency
Atypical Rational Agency


A competent agent exercises her capacity for self-directing autonomy by acting in ways that are explainable by appeal to her reasons. Central features of such rationalizing explanations are brought to light by way of a comparison with the sort of understanding that is afforded by merely causal accounts of one’s behaviour. The resulting observations are then applied to a recent decision by the Supreme Court of Canada, in which a psychiatric patient was deemed to be capable of making decisions about his own treatment. I conclude that when we make sense of someone’s treatment decision by seeing it as an expression of his reasons, we thereby commit ourselves to the possibility that an atypical, seemingly strange treatment decision may nonetheless be an expression of genuine, rational autonomy.

Table of contents

    1. The Capacity for Autonomous Decision-Making

    In respecting one’s autonomy I acknowledge her capacity to determine her own values and shape her life in accordance with them. To have this capacity, one must have some stable grasp of one’s values and be able to articulate projects on the basis of them. Those who lack such skills are deemed to be ‘incompetent’, or incapable of truly autonomous agency.

    In order to determine whether one possesses the requisite abilities, we are to focus not simply on a decision but, instead, on how it is reached, for we want to see whether the cognitive prerequisites of autonomy are evident in this decision-making process. We need to determine whether this person has the cognitive skills necessary for autonomous decision-making. We are not supposed to be addressing the quite different question of whether the choice itself is a good one. While such evaluation of the choice is relevant to moral and legal questions, it is not germane if our task is to determine whether the choice expresses genuine autonomy. Otherwise, there could be no bad autonomous decisions.

    2. Reason in Action

    Our aim is thus to achieve an understanding of the agent’s decision, an account that will show her choice to be rational in the light of her values and projects. We need not agree with the choice, for we may not share these ideals. What is required, rather, is that her choice should appear to be rational if one starts from her ideals.

    This sort of understanding is not supplied by a purely causal account of her choice; for the beliefs and values that make sense of her choice do so by means of prescriptive, normative standards rather than simply by means of the descriptive, nomological principles that sustain a mere causal explanation. To elaborate, when I give a mere causal explanation of an event, I subsume it under law-like generalizations, the implication being that the event occurred because things like it just do typically follow from those initial conditions. As John McDowell puts it, in this sort of causal account, “one makes things intelligible by representing their coming into being as a particular instance of how things generally tend to happen.” (McDowell 1985, 389) By contrast, when I make sense of an action by rationalizing it, my objective is not to portray the act as how people just do tend to behave in such conditions. Rather, I aim to portray the action as what the agent rationally ought to do given her values and other attitudes. As McDowell says of such normative accounts, they are “explanations in which things are made intelligible by being revealed to be, or to approximate to being, as they rationally ought to be.” (McDowell 1985, 389)

    The distinctive nature of rationalizing accounts can be appreciated by juxtaposing them with mere causal explanations. Thus, suppose I want juice and believe that I can most readily satisfy this desire by getting the drink from the fridge. It is then rational for me to get the juice from the fridge, since the statement that I ought so to act is the conclusion of a practical syllogism in which the premises express the contents of the belief and desire in question. (Anscombe 1957) So the account of my action by appeal to this belief-desire pair does double duty as both an explanation and a rational justification that presents the action as being rationally appropriate.

    By contrast, suppose that this same belief-desire pair were regularly followed by the motion of one’s left hand one millimeter to the right. In that case, one could causally explain this hand motion by appealing to my desire for juice and my belief about how best to obtain that drink, together with the (ceteris paribus) law that links these attitudes to such a motion. Here, the contents of my belief and desire play no central role in accounting for the explained behaviour; after all, one can imagine the same sort of nomic link connecting that hand motion to different beliefs and desires, and (unlike in the case of rationalizing explanations) this variation in the attitudes’ contents would subtract nothing from the explanatory work that is accomplished by appeal to such states

    In this second, mere causal account, the explanation works because of the described nomic pattern, a relation that leads us to expect that the hand motion just will typically follow the onset of that belief-desire pair, without any implication that it is rationally appropriate for it to do so.

    3. Further Distinctive Features of Rationalizing Explanation

    It has long been recognized in psychiatry that there are two such distinct modes of explanation. This is due largely to the influence of Karl Jaspers. Jaspers adopted from Max Weber and others the distinction between understanding an action from the agent’s perspective (Verstehen) and giving a causal account of the bodily motions that constitute the action (Erklaren). In his version of this distinction, Jaspers stressed that unlike the laws of nature, the rational principles that help to make sense of an action do not require confirmation by supporting cases in order to do their explanatory work. Whatever explanatory work is to be achieved by such rationalizing explanations does not await the discovery of a nomic pattern connecting the reasons to the action that they rationalize but is, instead, already there to be grasped just by understanding the belief-desire contents and their rational connection to the action.1

    In his Blue Book, Wittgenstein likewise contrasts mere causal explanations with rationalizations. He notes that in the former case, the claim that an action resulted from a particular cause is a hypothesis, and adds that this hypothesis relies upon confirming instances which show “that your action is the regular sequel of certain conditions which we then call causes of the action.” (Wittgenstein 1933/1965, 15) He contrasts this way of explaining an action with an account of the act in terms of the agent’s reasons, where “no number of agreeing experiences is necessary.” (Wittgenstein 1933/1965, 15) Here, again, rationalizing explanations are independent of empirical confirmation.

    Suppose that the general principles at work in a rationalizing account do not await empirical confirmation. Is it the case that they also retain their explanatory force in the face of disconfirmation? Yes, for they purport only to be normative principles, not true descriptions of actual patterns. Thus, for example, consider the case of Joan of Arc. We can explain (by rationalizing) her heroic actions against the English by appealing to her ideals even if many of her compatriots shared her ideals without acting on them as she did. This can be so even if Joan of Arc herself had not previously shown any greater tendency towards heroic deeds than her contemporaries. In this case, the statement that one who holds such ideals really ought to ‘stand up for them’ and oppose the enemy is not generally followed in the relevant population, but this is no obstacle to explaining or making sense of Joan of Arc’s actions in terms of those such ideals.

    4. Starson’s Capacity for Rational Autonomy

    Let us now examine issues concerning rational autonomy in the context of individuals who suffer from psychiatric illness.

    Some of these people continue to exhibit rational patterns in their decision-making to greater or lesser degrees. It is difficult in such cases to determine to what extent such patterns must be present in order for one to be capable of exercising genuine autonomy in determining the course of her own health care.

    A case of this nature was recently heard by the Supreme Court of Canada. (Supreme Court of Canada [SCC] 2003) The case involves Scott Starson, who was charged with issuing death threats to his neighbours. He was found to be not guilty by reason of his mental illness but was detained in a psychiatric hospital on the grounds that he posed a threat to others.

    Starson refused to take medications that had been prescribed by his doctors, who then claimed that Starson was not capable of making his own treatment decisions and should therefore be required to follow the prescribed treatment. Rejecting this determination, Starson appealed to the courts. After appeals to various courts, the Supreme Court ruled that Starson was competent to make his own treatment decisions.

    Starson’s case attracted much attention because of his intellectual accomplishments, which include co-authored publications in physics. Indeed, one prominent physicist (Pierre Noyes of Stanford University) says that Starson has done “exciting” work that has stimulated some of his own thinking about the theory of relativity. While Starson has not published a scientific paper since the 1980’s, he believes that his thinking about physics is a central source of meaning in his life. It is this dimension of his life that would, he believes, be extinguished by the medication. He bases this concern on previous experience with another anti-psychotic medication (Haldol), which dulled his mind to the point where he could no longer pursue his intellectual work.

    In explaining the Court’s ruling, Justice John Major did not deny that it may well be in Starson’s best interests to take the medication. He adds that respect for capacity derives not from the concern for another person’s best interests but, rather, from the duty to respect autonomy. Says Major, “The right to refuse unwanted medical treatment is fundamental to a person’s dignity and autonomy.” (SCC 2003, para. 75) According to Major, one’s autonomy must be respected even at the cost of one’s well-being.

    Granted, but was Starson competent to make autonomous decisions? In the relevant jurisdiction, Ontario, the legal standard for competency, or ‘capacity’, is as follows:

    A person is ‘capable’ with respect to a treatment … if the person is able to understand the information that is relevant to making a decision about the treatment … and able to appreciate the reasonably foreseeable consequences of a decision or lack of decision. (SCC 2003, para. 12)

    According to Major, the first part of this standard, the ‘understanding’ condition, “requires the cognitive ability to process, retain and understand the relevant information.” (SCC 2003, para. 78) The second part, the ‘appreciation’ condition, requires that “the patient be able to apply the relevant information to his or her circumstances and to be able to weigh the reasonably foreseeable risks and benefits of a decision or lack thereof.” (SCC 2003, para. 78)

    Starson met these conditions, as is evident from his reasons for his choice. To wit, he knew that the medications were intended to slow his ‘racing thoughts’, and it was for that very reason that he rejected them. He rejected the risk of having his mind dulled to the point where he would be unable to pursue the central project in his life, his physics research.

    As in the above example of Joan of Arc, to make sense of Starson’s choice in this way is not to regard it as a typical choice, or as one that is statistically normal. More specifically, when we see his choice as being rationally motivated by his projects we do not thereby assume that most people would make the same choice as he did. Hence, we can take his choice to have issued from reasons that support it, and we can thereby regard his choice as an expression of rational autonomy, while at the same time seeing it as an atypical choice. Indeed, we can take Starson to be quite unlike most rational agents, to be quite odd in comparison to them, without this compromising our view of him as a rational agent who is capable of exercising genuine autonomy. It is not even required that we see Starson’s choice as one that most people would make if they shared his goals and values, just as we need not take Joan of Arc’s choices to be the most likely ones for someone who shared her ideals. We can, in other words, allow for disagreement among rational people.

    This is because a rational agent’s perspective typically encompasses a host of competing interests and convictions. Thus, Starson, while wanting to pursue his work in physics, at the same time recognized that his symptoms led him into conflict with others, and also desired to be released from the hospital in which he was detained. These countervailing concerns could equally rationalize a decision to comply with the prescribed treatment (just as a concern for self-preservation could rationalize a decision by Joan of Arc not to confront the English). It is for this reason that opposing choices can equally be seen as expressions of a rational, autonomous self, the implication being that we should not see just one choice, the ‘normal’ choice, as the sole candidate for being an expression of rational, autonomous agency.


    1. Anscombe, G. E. M. 1957 Intention, Oxford: Basil Blackwell.
    2. Jaspers, Karl 1923/1963 General Psychopathology (2nd ed.), trans. J. Hoenig and Marian W. Hamilton, Chicago: The University of Chicago Press.
    3. Supreme Court of Canada 2003 Starson v. Swayze, 1 S.C.R. 722; 2003 SCC 32 (June 6, 2003), Available at
    4. Wittgenstein, Ludwig 1933/1965 The Blue and Brown Books, New York: Harper Collins.
    In Jaspers’ words, “Frequency in no way enlarges the evidence for the connection. Induction only establishes the frequency, not the reality of the connection itself…. A poet, for instance, might present convincing connections that we understand immediately though they have never yet occurred.” (Jaspers 1923/1963, 304)
    Paul Raymont. Date: XML TEI markup by WAB (Rune J. Falch, Heinz W. Krüger, Alois Pichler, Deirdre C.P. Smith) 2011-13. Last change 18.12.2013.
    This page is made available under the Creative Commons General Public License "Attribution, Non-Commercial, Share-Alike", version 3.0 (CCPL BY-NC-SA)


    • There are currently no refbacks.