Intentionality without rationality more

intentionality without rationality by Lisa Bortolotti abstract It is often taken for granted in standard theories of interpretation that there cannot be intentionality without rationality. According to the background argument, a system can be interpreted as having irrational beliefs only against a general background of rationality. Starting from the widespread assumption that delusions can be reasonably described as irrational beliefs, I argue here that the background argument fails to account for their intentional description. I any have argued that there is a rationality constraint on the ascription of intentional states. In particular, Davidson (1985) and Heal (1998) endorse the view that there is a necessary link between rationality and belief ascription, in spite of the fact that cognitive psychologists have highlighted numerous cases of blatant irrationality in any believer’s reasoning repertoire. The best defence of the rationality constraint from the implications of the empirical evidence is what I call the background argument. Only within a system that is largely rational can intentional description be granted to behaviour that falls short of the fundamental norms of rationality. An intentional system might occasionally make reasoning mistakes or exhibit inconsistencies, but these departures from rationality would not be possible unless the right logical and causal relations between the system’s beliefs and the external world were in place and unless the system largely conformed to the fundamental norms of rationality. Davidson (1985, 196) is explicit about the role that the background of rationality plays in interpretation: ‘It is only by interpreting a creature as largely in accord with these [basic] principles [of rationality] that we can intelligibly attribute propositional attitudes to it, or that we can raise the question whether it is in some respect irrational.’ He also mentions some examples of basic principles: the principle of consistency, the principle of continence, the logic of the sentential calculus and M 386 lisa bortolotti the principle of total evidence. When an intentional system violates a basic standard of rationality, it departs from its own norms and has no good reason for doing so. For Davidson, the only explanation of why an intentional system could suffer from, say, an inner inconsistency, lies in its compartmentalisation in sub-systems. The conflicting beliefs are not found in the same sub-system, otherwise the inconsistency would be detected, but are in different ones. When the boundaries of the compartments break down, the inconsistency becomes explicit and there is no further excuse for the system that fails to recover from it. In general, if the system has the capacity to revise its beliefs as to restore conformity to the standards of rationality where these were violated, it does show that the breaches of rationality were not an open rejection of the standards themselves, but an effect of compartmentalisation. The acceptance of the principle in question on behalf of the system is necessary for intentional description and is manifested if the system ‘is disposed in the appropriate circumstances to conform to it’ and such circumstances include ‘explicit Socratic tutoring’ (Davidson 1986, 203). Elsewhere I argue that at least some delusional states can be legitimately described as beliefs that do not conform to standards of rationality (Bortolotti 2002). In this short paper I shall assume that this is the case. In Section II, I shall introduce the Capgras syndrome as an example of monothematic delusion and discuss one reason why some subjects affected by it cannot be viewed as largely rational. Delusional subjects might fail to recover from violations of fundamental norms of rationality (such as the principles of consistency and total evidence) even when they are made aware of such violations (Breen et al. 2000). In Section III, I shall explore another reason to doubt that the background argument can account for the intentional behaviour of delusional subjects on the basis of the differences between those cases in which the background argument is most successful and the case of delusions. The fact that the background argument cannot account for the intentionality of delusional subjects constitutes a serious limitation for the Davidsonian theory of belief ascription and it is bad news for the general view that intentionality and rationality are necessarily linked. intentionality without rationality II 387 In order to defend Davidson’s belief ascription theory from the charge of failing to account for the intentionality of delusional states, one can claim that the background argument does explain the ascription of beliefs to delusional subjects. All that one needs to show is that the behaviour of delusional subjects is largely rational in the relevant sense and that it can be interpreted via the ascription of beliefs that are largely true and consistent. If delusions like the Capgras syndrome could be seen as temporary and local breaches of rationality, then they would not differ from occasional reasoning mistakes for the purposes of intentional ascription. There is some independent support for this view. In the case of monothematic delusions, the majority of the beliefs are not ‘infected’ by the delusion and the behaviour of delusional subjects can be perfectly rational when the topic of the delusion is not raised. Moreover, evidence suggests that delusional subjects often appreciate the implausibility of the content of their delusion even if they accept it as true. The attempt to treat monothematic delusions as occasional deviations from rationality has some plausibility, but there are powerful considerations against the view that subjects affected by delusions such as the Capgras syndrome can be regarded as largely rational. What is it exactly that the background of rationality requires? One might define a system as largely rational if most of the subject’s beliefs have the right causal and evidential relations with the world and are in the right causal and logical relations to each other. But, as we saw, it is also important to consider how often and to what extent instances of behaviour can deviate from the standards of rationality and whether rationality can be restored. Suppose an undergraduate student in psychology, Julia, volunteers to participate in an experiment on reasoning. She is asked to read the description of a person and then assess the probability of a number of sentences about the occupation of that person. Julia’s assignment reads as follows (Tversky and Kahneman 1983): Linda is 31, single, outspoken and very bright. She majored in philosophy. As a student, she was deeply concerned with issues 388 lisa bortolotti of discrimination and social justice and also participated in antinuclear demonstrations. What of the following is most likely? (a) Linda is a bank teller (b) Linda is a bank teller and active in the feminist movement. As do the great majority of experimental subjects, Julia decides that (b) is more likely than (a). She commits what is known as the conjunction fallacy, the basic reasoning mistake people make when they regard the conjunction of two events as more likely than one of the conjuncts. There is some debate on the significance of this experimental result but one common interpretation is that the subjects feel that Linda is more likely to be a feminist than a bank teller and take (a) to mean that Linda is just a bank teller, a professional choice that is hardly consistent with her profile. Now compare Julia with another fictional character, Mike. Mike is a retired lawyer and comes to believe that the person who lives in his house and looks identical to his wife is not his wife, but a clone of his wife. Mike is affected by the Capgras syndrome. Typically subjects believe that one of their closest relatives or their spouse has been replaced by an impostor and dismiss overwhelming evidence that the people who look identical to their loved ones are who they say they are. According to a standard account of the formation of the delusion (Stone and Young 1997), the delusion arises when the affective component of the face-processing module is damaged, leaving recognition unimpaired. Subjects see the loved one and recognise their face, but form the belief that the person they are looking at is not really the relative or the spouse. On this interpretation, the delusion is an attempt to explain why the face seen, which appears identical to the familiar face, does not feel right. On most accounts, apart from an abnormal experience, a reasoning bias or a reasoning deficit is also necessary for the acceptance and the maintenance of the delusion (Davies and Coltheart 2000). As a consequence of their delusional beliefs, subjects often become hostile or aggressive towards the alleged impostors. Now suppose Mike and Julia are otherwise perfectly rational. Is Mike just as rational as Julia? Yes, if we adopt a simplistic intentionality without rationality 389 reading of the background of rationality. They are both largely rational, because they deviate from norms of rationality only in relation to, respectively, the identity of Mike’s wife and the likelihood of the conjunction of events. Otherwise they both have true and sensible beliefs. I suggest that we reject the simplistic reading of the background of rationality and have a closer look at the behaviour that Julia and Mike are likely to exhibit in relation to their irrational beliefs. Julia might easily give up her belief that Linda is more likely to be a feminist bank teller than a bank teller as soon as someone reminds her of how probability works and explains to her how she was ‘misled’ into giving the wrong answer by the description of Linda. Julia will be disposed to recover from her failure of rationality when this is explained and pointed out to her. But, if Mike is anything like the typical Capgras patient, he won’t give up his delusional belief, no matter what his psychiatrist, his friends and his family might say to persuade him that the woman who looks identical to his wife is in fact is wife. He might concede that what happened to his wife is almost incredible but this will not affect his attachment to the content of the belief. Julia and Mike are not (ir)rational in the same way because Mike’s belief revision processes seem not to be operating as well. Of course, that experimental subjects will be disposed to recover from the conjunction fallacy is an empirical claim that needs to be supported by evidence about debriefing. However, it is safe to claim that delusional beliefs are resistant to revision to a greater extent than the beliefs that lead most subjects to commit reasoning mistakes. As we saw in the first section, within the conditions that Davidson imposes on intentional systems, the capacity to restore rationality plays a central role. Let me remind you that for Davidson, it is a condition on intentional systems that they are disposed to revise their beliefs in order to conform to norms of rationality, once their deviation from rationality has been brought to their attention. It is hardly consistent for the supporter of the background argument to appeal to the phenomenon of recovery in the specification of the standards that an intentional system has to meet and then to endorse the view that delusional subjects meet those standards. Something has to give. Delusional subjects are typically very resistant to 390 lisa bortolotti anything that might count as evidence against the content of their delusional state and this should be a reason to doubt that the background argument can account for the intentional description of their behaviour. III By considering the cases in which the background argument is most successfully advocated, one might find another reason to challenge the thesis that delusional subjects are largely rational believers. We feel that the background argument has a point when we can make sense of an irrational belief by appealing to the other true and sensible beliefs the subject has. In the previous section, we explained why Julia commits the conjunction fallacy in the context of a psychological experiment. Julia’s other beliefs (about what Linda is likely to become given her interests and education) can be used to rationalise her fallacy. As Heal (1988, p. 99) says: When a mistake is agreed to have been made we will often look for, and find, a reason why it was made, not just in the sense of cause or regularity in its making but in the sense of some excuse which reconciles the mistake with the idea that, even in making it, the perpetrator was exercising his or her rationality. By developing the background argument philosophers can explain those scenarios in which people commit ‘explicable errors’ and end up endorsing a false belief as a consequence of not grasping or mastering a concept (Davidson 1974). Think about the case of Bert who goes to the doctor and claims he has arthritis in his thigh. Bert has a pain in his thigh and does not realise that arthritis is a condition of the joints only. He does not completely master the concept ‘arthritis’ (or his concept ‘arthritis’ is not the same as his doctor’s concept ‘arthritis’). In this case, an interpreter can apply the principle of charity and make sense of the fact that Bert is talking about a pain in the thigh. This is possible because Bert would not be resistant to revise his statement if an expert were to explain to him what ‘arthritis’ really means. intentionality without rationality 391 The background argument is especially well-suited to account for cases such as this, but cannot be applied to the case of delusions. Which true belief of Mike can help us rationalise his conviction that the person who looks identical to his wife is an impostor? If Mike’s belief can be rationalised, it is by reference to his unusual perceptual experiences related to face recognition, but that does not go far in explaining why Mike hangs onto the belief. Moreover, trying to reinterpret Mike’s belief in order to restore rationality seems hopeless. As a charitable interpreter, I could take Bert as saying: ‘I have a pain in my thigh’, instead of ‘I have arthritis in my thigh’. But how could I reinterpret Mike’s belief? If I interpreted Mike as having the belief that he is suffering from the Capgras delusion, I could not make any sense of his hostile behaviour towards his wife. My charitable ascription would not be helpful in attempting to explain Mike’s behaviour in intentional terms. If the background argument is plausible at all, it is plausible in those cases in which the intentional system’s failure is due to a misunderstanding. Julia misunderstands the reasoning task and Bert misunderstands what ‘arthritis’ means. In most other cases, the idea that we are better off when we take the believer to endorse something true and sensible seems not to be supported by our everyday experience and by the psychological evidence. We do ascribe false and irrational beliefs to others if that allows us to make sense of their behaviour in intentional terms. IV In this paper I have challenged the background argument. As interpreters, we ascribe beliefs to those delusional subjects whose behaviour can be successfully explained within the intentional stance. Delusional subjects can behave as believers when they act on their delusion, when they offer tentative arguments for it and when they relate the content of their delusion to the other beliefs they have. Although they can be ascribed beliefs, delusional subjects cannot be regarded as rational. Their beliefs are always strenuously resistant to change and some delusional subjects are known to fail to respond to norms of rationality although the 392 lisa bortolotti violations of these norms are made explicit to them. The typical behaviour of delusional subjects suggests that they do not meet the criteria for intentional characterisation in accordance with the best reading of what counts as a background of rationality. The case of delusions provides a reason to challenge the accepted view that there cannot be intentionality without rationality. Centre for Social Ethics and Policy School of Law, University of Manchester Lisa.Bortolotti@manchester.ac.uk References Bortolotti, L. 2002: Marks of Irrationality. In S. Clarke and T. Lyons (eds.), Recent Themes in the Philosophy of Science. Dordrecht: Kluwer. Breen, N. et al. 2000: Towards an Understanding of Delusions of Misidentification: Four Case Studies. In M. Coltheart and M. Davies (eds.) Pathologies of Belief. Oxford: Blackwell, 75–110. Davidson, D. 1974: On the Very Idea of a Conceptual Scheme. In Inquiries into Truth and Interpretation. Oxford: Clarendon Press, 1984, 183–198. Davidson, D. 1985: Incoherence and Irrationality. In Problems of Rationality. Oxford: Clarendon Press, 2004, 189–198. Davidson, D. 1986: Deception and Division. In Problems of Rationality. Oxford: Clarendon Press, 2004, 199–212. Davies, M. and Coltheart, M. 2000: Introduction. In M. Coltheart and M. Davies (eds.) Pathologies of Belief. Oxford: Blackwell, 1–46. Heal, J. 1998: Understanding Other Minds from the Inside. In A. O’Hear (ed.), Current Issues in Philosophy of Mind. Cambridge: Cambridge University Press, 83–100. Stone, T. and Young, A. W. 1997: Delusions and Brain Injury: The Philosophy and Psychology of Belief. Mind and Language, 12, 327–364. Tversky, A. and Kahneman, D. 1983: Extension versus Intuitive Reasoning: The Conjunction Fallacy in Probability Judgment. Psychological Review, 90, 293–315.