I

First, however, I want to outline the argument that Szasz puts across, starting with the parallels he draws between phlogiston and psychiatry, which can be represented best as a sort of matrix (see table 1).

Science, claims Szasz, uses the same rules to describe different phenomena—for example, why something will burn under some conditions but not others, why some planes crash and others don't, and so on—but psychiatry is unreliable because it uses different principles to explain different phenomena. This “congenital epistemological error of psychiatry” leads us to “attribute acceptable, `rational' behaviours to reasons”, quite alien to scientific explanation, “and unacceptable, `irrational' behaviours to causes”,⁴ which have—we think—no moral gravity. However, even the schizophrenic who kills is an actor: killing, for whatever reason, is an act, as opposed to something like a convulsion, which is an event. Schizophrenia does not tell us why someone killed when he did so; a schizophrenic still has reasons for his actions, and as such he remains a moral agent. Therefore the idea that mental illness serves as an a priori mitigating circumstance in moral or legal discourse is false.

This, then, is how I think Szasz' argument boils down:

P₁ “Sane” behaviour can be explained in terms of (moral) reasons, not (scientific) causes

P₂ It is erroneous to posit a substantially different explanation for a particular instance of “sane” and “insane” behaviour

C Therefore any particular behavioural instance must be explained in terms of reasons, not causes.

Just because someone might be deluded, we have no reason to think that they are acting without reason, any more than we might think that a person who uses a wrong timetable is acting without reason for his or her action. Someone who hears voices and obeys them wants to do so, and this is the reason for—not the cause of—his behaviour, notwithstanding that this desire is disavowed.⁵ The belief that the schizophrenic is susceptible to occult causes in a manner that is not true of the sane is to apply differing standards to behaviour according to whether or not we approve of it; as such, the onus is on psychiatry to rid itself of this inconsistency in the same way that chemistry rid itself of the phlogiston dogma. This means that behaviour, the bread and butter of psychiatry, must no longer be seen in terms of illness.

II

As I have already said, I think there is rather a lot going for Szasz's argument, and I agree, broadly, with the conclusion. Nevertheless, however brutally it has been caricatured it here, Szasz's argument faces an enormous set of problems deriving from the fact that it suppresses too many “ifs”. If the potted version presented above is an accurate representation, then the conclusion does follow from the premises: but the expectation that “sane” behaviour is explicable in terms of reasons and not causes carries certain metaphysical assumptions that are not as easily cashed out as we might hope.

Importantly, there seems to be an implicit and rather Kantian assumption that reasons are the key factor in motivations rather than causes. Kant places the locus of action “not in the nature of man nor in the circumstances of the world in which man is, but . . . solely in the concepts of pure reason”.⁶ The moral phenomenon of behaviour must be free of factors from the outside world: those that are caused by alien influences rather than generated by the agent in question are not autonomous, and so do not come under a moral heading.

Szasz's argument seems to play into this particular worldview. For what his claim amounts to is that “insane” behaviour is not, as we would perhaps like to think, simply caused by outside factors, but that any particular behaviour must be traceable, ultimately, to the subject in question. This must be what he means when he claims that schizophrenia cannot cause one to engage in any action, since one's schizophrenia bookends any and all of one's actions. For some reason, the schizophrenic decides to push his victim “in front of a subway train”⁵—and it is this reason, not the schizophrenia as such, that should be the object of our explanatory inquiries. The actual pushing of the person under the train is open to moral scrutiny, irrespective of whether or not there is a diagnosis of schizophrenia.

This contrasts with the traditional position that certain behaviours are caused—in effect, that it is in the nature of “insanity” to rob the individual of his capacity as a moral agent. The early psychiatrist was known as the alienist; the “mentally ill” are often considered not to be autonomous beings. These are not unrelated: the common picture we have of mental illness is that, in effect, it is a force from outside (alien, hetero); accordingly, what it has us do is not, properly speaking, attributable to us. Rather, “If Hamlet from himself be ta'en away, And when he's not himself does wrong Laertes, Then Hamlet does it not”.⁷ But this view, Szasz argues, simply does not fit the details of our views concerning those whom we do not consider to be mentally ill. If psychiatry is a science, then it must account for the dysfunction of the mind according to the same rules as the function of the mind. This means abandoning the causal model of mental illness. Hamlet did wrong Laertes, uncomfortable as that truth might be.

So, a person's behaviour always takes the form it does for a reason, and this reason is not something that can be accounted for in terms of cause: presumably, then, reasons are the sorts of things that lie outside of the sorts of empirically causative chains that do explain things like the interactions of billiard balls.⁴ Schizophrenia might roughly equate to a worldview, but a worldview is not a cause. Indeed, taking at face value Szasz's own contention that a killing-behaviour is an act rather than an event, it seems difficult to escape the attribution of action to a Kantian autonomous moral-behavioural faculty⁸—something that at all times remains immune from the principle of causation.

The problem that Szasz's argument faces here concerns just how seriously we should take the assumption implicit in the first premise about the uncaused nature of acts. To take it as it is does seem to imply presuppositions about agency; but it is far from clear why we should accept this premise. Szasz's argument that psychiatry finds its roots in a desire to medicalise “badness” and the critique he presents based on this claim plausibly suggest that his aim is, above all, to retain a moral aspect to all behaviour. As such, his early admission of his “wish only to maximise the scope of voluntaristic explanations—in other words, to reintroduce freedom, choice, and responsibility into the conceptual framework and vocabulary of psychiatry”⁹ is wholly consistent with his latest claim that “[e]rroneous explanations . . . lead to . . . false expectations of the human condition, to moral catastrophes”.¹⁰ But this demand to bring the psychiatric back into the realm of the moral, which motivates much of Szasz's work, is itself a moral demand, and arguably lacks epistemological or metaphysical backing.

The absence of metaphysical or epistemological ammunition is suggested by the way that it is difficult to see what would be wrong with turning the interaction of the first premise and the conclusion inside out. This done, the argument would look like this:

P₁ “Insane” behaviour can be explained in terms of causes, not reasons

P₂ It is erroneous to posit a substantially different explanation for a particular instance of “insane” and “sane” behaviour

C Therefore any particular behavioural instance must be explained in terms of causes, not reasons.

This form of the argument, too, would satisfy the demand that the same standards be applied to mental function and dysfunction. What it would not obviously do, of course, is provide the scope that Szasz demands for the behaviour of the “mentally ill” to be seen as appropriate for moral evaluation through a delineation of act and event; and, on the basis that we have to be consistent between the “mad” and “sane”, it does not appear to provide any scope for attributing moral epithets to sane behaviour—although I'll come back to this in a little while.

III

Szasz's argument means that the “mentally ill” are, though we might not like to admit it, no less autonomous than the rest of us, even though their autonomy might be expressed bizarrely. Inverting the argument can give pretty much the same result—a moral equivalence of the “insane” and “sane”—but on the basis not that the “mentally ill” are no less autonomous than the healthy, but rather that the “mentally healthy” are no more autonomous than the ill.

The advantage that this approach gives is that it brings psychiatry closer to the model represented by antiphlogistian chemistry. Instead of the rather ad hoc position of an immaterial principle of behaviour, we retain for all behaviour a principle—some form of causation—which is open to the possibility of empirical (psychological) study. This looks to be much more scientifically defensible. The intuitive objection, of course, is that the inversion apparently deprives us of self determination in our actions—indeed, it takes the whole moral concept of an action away from us, and levels the whole of the population off as being in exactly the same non-autonomous boat as that which we conventionally ascribe only to the lunatic. However, this sort of objection, it seems, represents a desire to make the world a moralisable place regardless of its nature; and this is backwards thinking. But even if this charge is misguided, I do not think the inverted model really does deprive the world of moral value.

Let us return to the intuitive model of the autonomous moral agent. The mentally “healthy” are the masters of their own movements and subject to none. However, as was pointed out by Plato, “`master of oneself' is an absurd phrase. For if you're master of yourself, you're presumably also subject to yourself, and so both master and subject. For there is only one person in question throughout”.¹¹ So autonomy, on this account, collapses right from the start into antimony. Not only is it a matter of moral gymnastics that one be master and servant of oneself; but also, in so far as one is servant at all, then one is not auto-nomous by definition. If autonomy is a condition of morality, it seems nevertheless to be conditioned by some sort of heteronomy—which is absurd. The flip-side of giving a law is taking a law, and this is something that the autonomy account of behaviour, I suspect, tends either to forget or to push quietly to one side.

Of course, it may be that this argument, if it succeeds at all, succeeds only by dint of accident of grammar, and the peculiarities of reflexive verbs. But it does seem to feed into what I think is a stronger argument based on Donald Davidson's account of the relationship between reasons and causes.¹² Davidson begins by making the trivial point that it is obviously possible that “a man may all his life have a yen, say, to drink a can of paint, without ever, even at the moment he yields, believing it would be worth doing”.¹³ This flies directly against Szasz's point that all behaviours are based in some sort of reason. Davidson's paint-drinker seems to be able to give no reasonable account at all for acting as he does—it is just something about the way he is put together. This make up causes him to drink the paint.

According to the Szasz argument, if we follow this line, we are simply pathologising, quite illegitimately, behaviour that we do not understand. But I do not see how this is a pathologisation. After all, a cause need not be a mechanical cause. Perhaps the cause of the paint-drinking was psychological: simply a desire to drink paint. The paint-drinker's reason for his action was that his desire caused him to carry it out. This does not imply an occult force working on him. “Why on Earth,” asks Davidson, “should a cause turn an action into a mere happening and a person into a helpless victim?”¹⁴ We tend to think of causes being external, but this could be a slip of language; the cause of a behaviour could be internal, and as such the worries of those who fear a nihilistically mechanical world should be allayed. “Some causes have no agents. Primary among these are those states and changes of states in persons which, because they are reasons as well as causes, make persons voluntary agents.”¹⁴ It is sufficient that it is me that causes my behaviour for it to come under the moral purview; a reason is simply one type of cause—the type that can be articulated.

Key here is the idea that an action may be both voluntary and caused—the claim that an action is voluntary if and only if I am its cause. Indeed, the whole concept of a voluntary action appears, on examination, to be dependent on its having been caused by its perpetrator. If this causal condition did not pertain, the remaining options would be that it was caused by an outside agent or that it was utterly uncaused. I think that an uncaused behaviour would not be voluntary—it would supervene upon me in an arbitrary or random manner in just the same way as would a convulsion or sneeze. Therefore any behaviour not caused would be an event, not an act at all. The only other possibility would be utter inertia.

So whereas for Szasz reasons and causes for a certain behaviour are incompatible—to the extent, one might think, that any deed that is caused is not, strictly, a behaviour so much as a simple event—it is possible for the situation to be quite different. Reasons and causes are entirely compatible; one's reason for doing something may be sufficient cause for one to do it. Importantly, though, it is no less a cause for that. It was the paint-drinker's desire to drink paint that caused him so to do; it does not follow automatically, though, that drinking paint is something that is forced upon him.

The same applies to more mundane, “sane” patterns of behaviour. I go to the freezer because (Middle English, by cause) I want a bowl of ice cream. My desire causes my action—but it is still my desire. There is no immediately obvious reason why we should try to say any more. My reason for behaving in such and such a manner is the determining factor of that behaviour. It is an account I can give for that behaviour, but need not be more than that.

IV

However, there does remain this problem: the explanation of whatever it is that causes me to have a particular desire. That is—granted that my desire to eat ice cream, drink paint, or push someone under the train is sufficient to explain my behaviour at a certain point, what is it that causes this desire? The dilemma is between an apparent infinite regress on one hand, or else an arbitrary causal starting point on the other.

The problem, I think, is still one of grammar. It arises from the notion that, if something is to count as an action, ultimately it must be a form of behaviour that is caused by, and logically predicated on, an actor. Much the same consideration, I think, underlies Szasz's approach, and as it is this that prompts him to reintroduce the “insane” back into the moral arena, since any and all behaviour, he thinks, arises from an agent. So far, what I've tried to suggest is that we could plausibly see mental health and illness alike in terms of causes, allowing that a reason for doing something is simply a type of cause, although it would still be the agent at the start of the causal chain. However, a critique of this view might be possible.

The approach outlined still does leave open a distinction between a functional and a dysfunctional mind. “Mental illness” and “mental health” might be equally caused, but there would still be room to talk in a manner that would correlate to “good” and “bad” causes. This distinction between function and dysfunction is one with which neither Szasz nor I are wholly happy, not least because “[t]he problem with defining [mental] disorder in terms of dysfunction . . . is that dysfunction itself requires a definition”.¹⁵ For Szasz, if this definition rests solely on social (un)desirability, then talk about “illness” is misplaced. I accept this, but would like to go further: mental dysfunction presupposes a functional norm; but I think that a demand for consistency must entail a demand that we examine whether it is sensible to ditch the idea of mental “illness” but retain that of mental “health”. Szasz seems to be committed to the possibility of mental health—it is the illness that is phlogiston—but perhaps it is just as mythical as mental “illness”.

From a number of perspectives, the ideal of “mental health” must be seen as an essentialist mistake. For the purposes of this paper, I'll restrict myself to cherry-picking the thought of Jean-Paul Sartre, though I think that a Heideggerian approach could do much the same thing. From the Sartrean point of view (in which “choice is nothing other than the being of each human reality”¹⁶ and “[m]an . . . exists only in so far as he realises himself. He is therefore nothing else but the sum of his actions”¹⁷), it is not the case that one makes autonomous decisions with reasons, or that one is overcome by irresistible impulses that make one do something either against or irrespective of one's will. Rather, it is the case that the actor is secondary to the act. An actor is a factor we attribute to behaviour; not the cause of an action, so much as the product of activity.

If this is a plausible account—all I want to do here is raise the possibility, and I leave the details of any such account unformed (although RD Laing seems to have made a start, however fumbling¹⁸)—then it means that neither mental health (and the attendant picture of behaviour based in reasons) nor mental illness (and the attendant picture of behaviour based in causes) is accurate. Szasz thinks that the mind cannot get ill because the mind is a metaphor for a moral, not a medical, faculty, and it is thereby paradigmatically different from the sort of things that could be ill (“[i]t would be foolish—and fruitless—to search for the `etiology' [sic] of speaking French,” he memorably suggests¹⁹). All I am suggesting is that we might want to go further, and suggest that the mind cannot be ill because it simply does not exist independently of what it does: it is a word we use to stitch together behaviours.

On this account, the actor is not overcome by aberrant mental processes—he arises from those mental processes, which therefore cannot be aberrant, simply because there is nothing from which to err. The suggestion is that it is a false cultural expectation rather than a genuine diagnosis which leads us to suggest that someone is mentally “ill” or behaving in a manner that is not autonomous and so not open to moral scrutiny. The idea of causation fails as a mitigating factor in the psychological and psychiatric fields because there is nothing upon which a cause could act: and it therefore is not a cause at all. A system of desires, then, might cause a certain pattern of ice cream obtaining or paint-drinking behaviours, but these are not behaviours in which I engage by virtue of a logical priority. On the contrary: “I” simply denotes a gathering together of patterns of behaviour. I am not overcome by causal mental processes; “I” am nothing but various processes, among which we might count paint-drinking, ice cream snaffling, passenger-pushing behaviour. For this reason, a mental “illness” is not a hetero-, alien factor, but is rather that which generates the auto-, and is therefore equiprimordial with “health”.

SUMMARY AND CONCLUSION

The intention of this paper has been to take issue with Szasz's arguments in his recent paper concerning the non-existence of mental illness. The argument has been that Szasz's own paper represents only a partial account of the issues he raises about mental illness. The possibility has been suggested of two arguments that would yield conclusions that sit more or less comfortably with those yielded by Szasz, but which would cover those aspects which he misses.

It is inconsistent to explain mental illness using one model and mental health using another; it is perfectly coherent to suggest that the mentally “ill” are just as responsible for their actions as the “healthy”. It is just as coherent, however, to reverse the demand that we regard the behaviour of the mentally “healthy” as being uncaused; inspired by a Davidsonian argument, it could be claimed that behaviour is caused—caused by psychological factors which, after the event, we call reasons, admittedly, but caused none the less. Indeed, such an account might actually turn out to be more coherent, since it does not rely on metaphysical assumptions about a “free will” utterly removed from the influences of the outside world.

Even if one is satisfied with Szasz on this point, however, it remains the case, it has been argued, that an argument about causes and reasons is itself only possible because of assumptions about the ontology of actors and the idea that mental function is possible, even if dysfunction isn't. What has been suggested here is that these assumptions—that all actions need an actor—are not a priori true. A Sartrean approach would subordinate the actor to his actions, and thereby would open the way to a claim about the equiprimordiality of mental “health” and “illness”: for if one is simply the aggregate of one's behaviours, attitudes and so on, then it is hard to see how one could be either healthy or unhealthy mentally. This approach would accommodate the claim that mental illness is a myth, but it would be able to go a step further also and claim that the same applies to mental “health”.