Defining Desires

Defining Desires

In his book Rationality in Action John Searle asserts, “the single most remarkable capacity of human rationality…is the capacity to create desire-independent reasons for action” (p. 168). This is a strong and somewhat daring claim, for it assumes what a majority of philosophers have long denied: that such desire-independent reasons are even possible. Searle’s project throughout Rationality in Action is to rebut what he calls the “Classical Model” of practical reason - a general paradigm that includes, among other things, a view often labeled internalism: “to a rough approximation, the idea that reasons for action must invoke desires” (Millgram 6). Searle rejects the whole of this proposition, but most importantly, he does away with the broadly defined notion of “desire” contained in it. His defense of desire-independent (or external) reasons turns largely on his claim that the Classical interpretation of desire is too broad to be meaningful.

Internalism and Instrumentalism

In his noted defense of internalism, Bernard Williams defines desires as those things which comprise an agent’s “subjective motivational set” (Williams 78). This set “can contain such things as dispositions of evaluation, patterns of emotional reaction, personal loyalties, and various projects, as they may be abstractly called, embodying commitments of the agent” (81). On this view, a desire can be broadly understood as any internal concern capable of motivating the agent to action. Practical rationality, for Williams, takes the following form (for agent A contemplating action B ): “A has a reason to B only if there is a sound deliberative route from A’s subjective motivational set to A’s B -ing” (91). For Searle this thesis is quite in tune with the Classical Model, on which one’s desires are given, and reasoning is about means alone. As Bertrand Russell claimed, “reason has a perfectly clear and concise meaning. It signifies the choice of the right means to an end that you wish to achieve. It has nothing whatever to do with the choice of ends” (qtd, Searle 11).

What Searle labels “the Classical Model” encompasses several different claims about rationality. Since this model does not represent a single system developed by a single philosophic movement, we should not assume up front that the various beliefs Searle ascribes to it are logically interdependent. In fact, I will argue that one of the biggest flaws in Searle’s account of this model is the absence of a distinction between internalism - the claim, “there cannot be any desire-independent reasons for action” (26) - and what has been traditionally called instrumentalism - which would include Russell’s assertion, “reason…has nothing whatever to do with the choice of ends.” While these two views have their similarities (i.e. both deny the existence of external reasons), their connection is probably more historical than logical: those who believe that practical reason is fully internal have traditionally denied that ends can be rationally evaluated.

It is unclear to me precisely why Searle has run these two views together. However it does seem somewhat easy to slide from internalism to instrumentalism by simply tweaking certain term-definitions. For instance, we could use the following logic: if reasons for action must invoke desires, desires must in some sense precede reasoning. This might be further interpreted as saying that the motivational set is pre-rational - in other words, not subject to rational deliberation. So from this construal of internalism, we might get the impression that “the choice of means to our ends can be more or less rational, but our ends themselves can’t” (Fehige 49). However we have only arrived at Christoph Fehige’s definition of instrumentalism through semantic finagling. In the first place, we have ignored an obvious difference between internalism, as stated by Williams, and instrumentalism: internalism makes no negative claim about the rationality of ends. Williams’ claim deals with a certain kind of reason, a reason for action, but it does not preclude the possibility that other sorts of reasons - perhaps even reasons for choosing between ends - might exist as well. But as I said, the real problem here is in the definitions. If the claim, “reasons for action must invoke desires” does entail that the motivational set is pre-rational, it is only because the relevant meaning of “rational” here is “rational in action.” This amounts to claiming that reasoning about actions is preceded by the desires, and does not include reasoning about desires (which may in fact be possible). My reason for making this distinction is contained in the parenthesis: it is in what it does not claim, that is in leaving open the possibility for rational deliberation of ends, that internalism deserves special attention as separate from the instrumentalist assumptions of the Classical Model. Just how relevant this distinction is will play out further on, after we have examined Searle’s construal of external reasons.

Desire-Independent Reasons

Searle sees the denial of desire-independent, or external, reasons as the root problem of a certain aspect of the Classical Model - what I will now call the internal-instrumentalist quagmire. His account of the problem goes as follows: if desire-independent reasons for action are impossible, then “at any given point in one’s life no matter what the facts are, and no matter what one has done in the past or knows about one’s future, no one can have any reason to do anything unless right then and there, there is…a desire, broadly construed, to do that thing.” The consequences of this thinking might play out in the following way:

This sort of scenario runs counter to what we would expect from a rational agent, and so presents an intuitive dilemma to instrumentalism. However, this counter-intuitive quagmire does not up front present a logical objection to the internal-instrumentalist view. On that view it most certainly is rational for the agent not to pay; reasons for action are internal - no matter how strange, not to mention immoral, the action may seem. As Hume claimed, “tis not contrary to the dictates of reason to prefer the destruction of the whole world to the scratching of my finger” (qtd. 28).

But Searle does see the agent’s refusal to pay as fundamentally absurd, for reasons that, he claims, lie outside the motivational set. His claim that such desire-independent reasons are possible stems from his account of commitments: “a commitment is the adoption of a course of action or policy (or other intentional content; one can for example, be committed to beliefs and desires) where the nature of the adoption gives one a reason for pursuing the course” (175). In this case, “the nature of the adoption” is that it points to some motivational factor independent of the agent’s desires. To follow the above example, “by ordering the beer and drinking it when brought, I have already intentionally created a commitment or obligation to pay for it, and such commitments and obligations are species of reasons” (187). That commitments do not reduce to desires is illustrated when the agent, even if s/he does not want to pay in any sense of the word, recognizes an obligation to do so and acts accordingly. But what exactly is it that makes such a commitment a reason for action?

Searle answers the first question by exploring the nature of an assertion, which he classifies as a certain kind of commitment:

Thus agent A may give the simple reason for uttering sentence s, “s is true.” And that reason is entirely external to the agent. The internalist might object that A uttered s for ultimately subjective (internal) reasons; perhaps A desired to make the statement, and that desire was linked to a broader end in the subjective motivational set. But Searle is trying to reverse the causal chain of that kind of reasoning by introducing the concept of conditions of satisfaction. The desire to utter the statement is satisfied by the utterance, but if in wanting to utter it, the agent really wants to mean what the statement says, that intention-in-meaning is only rationally satisfied if the statement is true. Thus, A has reason to say s because A wants to say s; but A has reason to want to say s because s is true. If, conversely, s is false, then A does not have reason to want to utter it, and therefore has no reason to do so.

The further claim here is that an assertion is a type of rational commitment to an external entity (the truth of the assertion), and in fact all external commitments are similarly formed - “through imposing conditions of satisfaction on conditions of satisfaction” (175). If a truth-claim is made, and the speaker actually means the content of the claim, the speaker must be committed to the claim’s validity: “the speaker has now created a reason for accepting the logical consequences of his assertion, for not denying what he has said, for being able to provide evidence or justification of what he has said, and for speaking sincerely when he says it” (173). In participating in assertion-making, then, the speaker is imposing even further conditions of satisfaction on conditions of truth-satisfaction, and is further accepting responsibility for meeting those conditions.

How, though, can prior acts or assertions make us rationally responsible for future ones - especially if we do not want to do the future acts? Because we act in a social context, and therefore rational acts are associated with what Searle calls institutions. An institution is a mode of social interaction that provides “ways of representing the fact that that one’s present action creates a reason for future action” (203). Thus, to use the beer example, if I order a beer, then it ought to be understood that I am operating within the social institution of beverage-ordering, which has as one of its requirements that I pay for what I have ordered. Thus in performing the act of ordering I am committing myself to paying, and it is rational for me to do so.

Searle anticipates one objection to this view: “why are such commitments, obligations, and responsibilities binding on the agent? […] Why are they not social constructs like any others” (176)? The answer Searle gives is that institutions are social constructs, but that is simply how we operate; we need to associate our actions with social institutions if we are going to make them meaningful. “Reason does not derive from the institution, rather the institution provides the framework, the structure, within which one creates the reason. The reason derives the fact that the agent binds her will through a free and voluntary act” (204). And in so binding her will the agent creates a desire-independent reason for action - a reason binding on her in the future, whether or not she desires to follow through with it.

The Nature of Desires

So where does the internal-instrumentalist claim stand? Is it yet possible to see what Searle calls external reasons as really reducing to internal ends? Searle does admit, “in some very broad sense of ‘want’ or ‘desire,’ every intentional action is an expression or manifestation of a want or desire to perform that action” (168). However, this broad definition of desire

But the question remains, if all intentional action can be explained on a broad definition of desire, how can that same definition incorporate reasons that are in any sense external? In other words, if all reasons for action can be said to invoke a desire how can any of those reasons, even the recognition of a commitment, still be desire-independent? Of course, by “desire-independent,” Searle means desire in the limited sense, the sense in which the agent desires chocolate ice cream. But how can it be that Searle, who wants to make distinctions between desires and desire-independent reasons, can allow for even the broadest interpretation of the claim, “every intentional action is an expression…of a want or desire”? If that statement is true, and there are desire-independent reasons, then even on the broad definition of desire there would have to be desires somehow external to the agent. And what could this possibly mean? Perhaps it would entail that when I say, “I desire x” I am really saying, “my social obligations desire x”–where I actually mean that they require some action represented by x. But my social obligations are external to my self on Searle’s view, so how could “I desire x” possibly be construed as “my obligations require x”? This would only be possible were my obligations internal to my motivational set.

While Searle does not explicitly address this dilemma, he does draw a psychological connection between internal and external entities on the level of motivation. It is important that he do this, as a major internalist criticism of external reasons is that such reasons are not capable of motivating. John Robertson has captured the content of this view as follows: “A consideration C is a reason for an agent to B if and only if the agent would be motivated to B if she were rational and aware of C” (Robertson 130). Searle does not dispute this; rather he claims it is possible that factors outside the subjective set of desires can motivate the agent. He appeals once again to the way we handle truth statements:

Obligations follow a similar pattern. If I order a beer, then by recognizing that I have engaged in the institution of beverage-ordering I recognize that I have reason to pay for what I have ordered–therefore, for wanting to pay for it. And that reason-for-wanting is an external motivation because it is assumed that if the agent is rational, s/he will find such a reason to be a sufficient motivator.

On the Williams thesis, however, what Searle calls external motivation would reduce simply to belief, which, if it is to motivate, must have internal origins: “What is it that one comes to believe when he comes to believe there is reason for him to  if it is not the proposition, or something that entails the proposition, that if he deliberated rationally, he would be motivated to act appropriately?” (Williams 85). And the propositional content of such a belief could only be an internal reason statement. In the beer example, the internalist would say that the motivation to pay springs from a value that holds the institution of beer-ordering to include paying for what has been ordered. How else would the payment be binding on the agent? If there were no internal value, then all the agent is doing in ordering a beer is performing an action that society generally expects to be associated with a further action. But surely this is not what Searle means by a binding commitment! If I do not accept that my ordering the beer obligates me to pay for it, how does that alone make me irrational? What is capable of obligating me to pay other than that it is in my internal motivational set to view my prior action as a commitment to a future one?

The argument from assertion-making can be similarly addressed. There are a number of reasons why I might make an assertion such as, “it is raining.” Perhaps I only want to communicate a true fact; in that case I could say, “I said s because s is true.” But what if s is false, and I am lying? It does not make sense to say, “I said s because s is false.” There is no longer a clear external entity motivating the utterance of the statement - and yet the motive for lying can surely be rational. For instance, were I living in Germany during the late 1930’s, and I had decided to harbor Jewish refugees in my home - if I were asked by the Nazis, “do you have Jews in your home?” would it be irrational for me to lie? Further, would I have reason to tell the truth at that moment, simply because it is the truth - even if truth-telling would contradict my deepest-held internal sympathies? The answer to both questions is, of course, no, and this brings us to a further conclusion: the truth of a statement is only a reason for uttering it if, in the moment of utterance, we are concerned with truth-telling. And such a concern can only be internal to the agent.

What the internalist must still account for is what it means to have reason to do something that, in the moment of decision, one might not wish to do. Searle solves this problem by introducing the concept of reasons for desiring - claiming that such reasons are ultimately external to the agent. Is it still possible, then, that a reason for wanting something is tantamount to having an external reason for action? If there can be reasoning about ends, does that entail that such reasoning must be desire-independent?

Reasoning About Ends

What I have called the internal-instrumentalist quagmire - the counter-intuition that, in an example such as beer-ordering, the agent who refuses to pay his bill is rational - will turn out to be less of a problem for internalism when internalism is considered separately from instrumentalism. The real issue, it seems to me, is that if the rational agent truly is the slave of her passions she cannot be held responsible for her motivations; thus, when she is arrested for not paying her bar tab, all she can tell the police is that she acted rationally. There was nothing in her motivational set that would have rendered paying the tab rational, and thus in refusing to do so she was only meeting the requirements of reason. Without getting too much into ethics at the moment, I would like to point out that it is a disturbing quality of classical instrumentalism that it does not allow for a deliberation of ends. If for instance there was a person who had as his primary motivation to kill me - and all other motives sprang from that end–if this person were to face me with a gun and say, “give me one good reason not to pull the trigger,” there would be nothing I could say in my defense. I could appeal to no moral value, nor to the fact that he could face severe penalties if he was caught and tried for murder; those sorts of considerations are secondary to his primary aim of seeing me dead. And for the instrumentalist, reason “has nothing whatever to do with the choice of ends.” Thus, were I an instrumentalist, I would have to admit the rationality of him killing me.

Of course, the internalist cannot ultimately explain why the person in the scenario should spare my life if such an action truly serves no component of his motivational set. However, since internalism leaves open the possibility for deliberation of ends, it allows at least for the agent to be seen as responsible for organizing and reasoning about his motivations. While the agent may be rational on the level of action - he may have engaged in sound means-end calculation - it remains open whether or not he has rationally handled his motivational set; and it is my contention that the two types of reasoning are different. Reasoning about means is not the same as reasoning about ends.

A number of internalists are now offering alternative models to means-end instrumentalism. Aurel Kolnai has suggested that the set of ends actually contains a number of independent entities that are desires in their own right - meaning they do not simply point to some higher desire but often contain incompatibilities and contradictions. “A man with dominant spiritual interests may control his penchant for gluttony not merely because in a consequential sense it is apt to interfere with his studies but because he is pained by a sense of essential incompatibility between these two passions” (Kolnai 265). Searle wants to agree with this: “rational decision making is typically about choosing between conflicting reasons for action” (Searle 31). David Schmidtz has conceived the idea of “maieutic” ends, or ends that we adopt to fulfill a certain psychological or pragmatic need. For example, an agent might fall in love, not because she finds someone who meets the qualities she looks for in a lover, but because she has a need to love someone (Schmidtz 246). None of these theorists agree that there can be such a thing as a set reasoning process by which ends can be rationally created. But they all seem to want to hold that reasoning about ends is qualitatively different from reasoning about means (although Williams, interestingly enough, seems to reject this thesis). Kolnai goes so far as to suggest that all deliberative reasoning is of means, and that means-end reasoning is simply mechanical computation.

I do not wish to get into the specifics of ends-deliberation theory here. I wish to simply reiterate my opening point: a valid criticism of externalism does not invariably lead us to instrumentalism. Rationality-in-desiring may consist in simply the ability to balance one’s various ends and seek consistency between them—in the hopes of optimally fulfilling the motivational set. But even if there were such a thing as a desire-independent reason for motivation, that does not entail that it could be a desire-independent reason for acting. We have seen in such cases as truth-telling and the acceptance of social institutions, the two cases Searle sees as most directly lending themselves to an externalist interpretation, that the motivating factors cannot exist independently of the desires.

Works Cited

Searle, John. Rationality in Action. Cambridge: MIT Press, 2001.

Millgram, Elija. Practical Reasoning: the Current State of Play. Varieties of Practical Reasoning. Ed. Millgram, Elija. Cambridge: MIT Press, 2001. 1.

Williams, Bernard. Internal and External Reasons. Millgram 77.

Fehige, Christoph. Instrumentalism. Millgram 49.

Robertson, John. Internalism, Practical Reason and Motivation. Millgram 127.

Kolnai, Aurel. Deliberation is of Ends. Millgram 259.

Schmidtz, David. Choosing Ends. Millgram 237.

Moved from Essay & Theses