The European Rationalist – In Reason We Trust

What is it to “know your own mind”? In ordinary English, this phrase connotes clear headed decisiveness and a firm resolve but in the language of contemporary philosophy, the indecisive and the susceptible can know their own minds just as well as anybody else. In the philosopher’s usage, “knowing your own mind” is just a matter of being able to produce a knowledgeable description of your mental state, whether it be a state of indecision, susceptibility or even confusion. What exercises philosophers is the fact that people seem to produce these descriptions of their own mental lives without any pretence of considering evidence or reasons of any kind and yet these descriptions are treated by the rest of us as authoritative, at least in a wide range of cases. How can this be?

Most of the philosophers exercised by this problem would regard the English phrase “knows his own mind” and its connotations as a mere distraction, as the product of a theoretically unhelpful ambiguity. But to some, the English phrase suggests a fruitful approach to the philosophical problem of self-knowledge. In the twentieth century, Wittgenstein, Sartre and Austin all explored the idea that such mental phenomena as thinking, reasoning, deliberating are, in an important sense, activities which culminate in deeds, in the making of cognitive commitments. Moran sets out to refurbish this tradition, to revive the notion that self-knowledge is special because it is a matter of actively making up your own mind, rather than of passively apprehending it.  

Moran’s approach looks most promising when applied to the cognitive parts of our mental lives: our beliefs, judgements, the more intellectual of our desires and emotions. Yet the contemporary discussion of self-knowledge focuses at least as much on experiential states like pain and visual sensations. Moran says at the outset that he thinks these questions about experience require a quite different treatment (p. xxxiii) and focuses his attention on the cognitive. Some of his opponents might regard this as an admission of defeat but the range of phenomena Moran does cover is sufficiently impressive to make his view worthy of very serious consideration. In this article, I shan’t try to discuss Moran’s account of belief, desire and emotion. Instead I will concentrate on his view of our knowledge of our own intentions and of what are doing (or shall do) to execute them.

The Epistemology of Agency

People in a normal frame of mind usually know what they are doing and they know this without having to observe their own behaviour. In Anscombe’s example, which Moran adopts, someone pumping water may know that they are operating a pump without observation but not know that they are creating a clicking noise except by hearing the noise (pp. 124-7). In Anscombe’s view, this is because making a noise is no part of what they are trying to do. Anscombe acknowledges that one can be wrong in thinking that one is operating a pump (the pump one means to operate might be an hallucination); her point is that our knowledge of our own actions is groundless, not that it is infallible. We can (and often do) know what we are doing without the evidential backing needed for knowledge of someone else’s pumping.

As both Anscombe and Moran observe, this epistemic authority extends to our own future agency (p. 88). A person in a normal frame of mind can come to know that they will go to London tomorrow simply by deciding to go to London tomorrow. Again this way of knowing the future is hardly foolproof: an unexpected rail strike or weakness of will may intervene, and empirical evidence as to the reliability of the railways  etc. is needed before the decision can be taken. Anscombe’s point is that a normal person can know that they will go to London in this way without amassing inductive evidence of their own resoluteness, of their propensity to stick to and execute their own decisions, whereas if they were asked whether someone else who has decided to go to London will actually go, they would need evidence not only about the state of the railways but also about how resolute the person in question is. For Anscombe, the fact that we (directly) create this action gives us a special, evidence-independent way of knowing of it. We lack groundless knowledge of what other people are doing or will do precisely because we don’t (directly) create their actions, rather we apprehend them as objects in the empirical world.

In expounding these points, I confined my attention to “people in a normal frame of mind”. An example devised by David Velleman shows why this qualification is needed.[2] I have agreed to meet an old friend in order to patch up a quarrel. As the meeting unfolds, I start to become petulant, raise my voice, provoke my friend into saying some unconscionable things and we part in anger. I later infer that I must have decided, before the meeting, to end the friendship, an objective which I skilfully attained. But at no stage either prior to or during the meeting would I have announced such an intention and even afterwards I unearth it only in an effort to explain what happened. I may end up convinced that this was my intention but only qua observer of myself.

I take it Moran has this sort of case in mind when he says that:

Even within a psychoanalytic explanation, it will normally be the case that the contrary thoughts and attitudes which explain the subject’s blocked awareness of the intention will themselves be reasons for ambivalence in his overall intention; that is the intention itself will not be a wholehearted one. Ignorance in such a case will not be mere ignorance, not only because it will be irresistible to look for a motivation of sorts to explain it, but because the motivation we then impute to the person must qualify the ascription of the original intention (as conflicted or partial). (p. 57) 

Applying this to Velleman’s example, I don’t fully intend to end the friendship because I can’t, to use Moran’s favoured expression, avow this intention and I can’t avow this intention because I am aware of good reasons for not ending the friendship. In general, Moran thinks that a subject is an epistemic authority about the content of his intentions and intentional actions only where the subject’s intentions seem well grounded to the subject himself. To the extent that an intention seems to him misguided, the subject’s privileged epistemic access to it will be compromised, as will his sense of being in control of behaviour guided by this intention. To put it another way, he will start to wonder whether this intention is, in the full sense, his.[3]

Moran’s epistemology of agency takes off from the fact that we create our actions, we don’t just observe them. But why, someone may ask, does the mere fact that we create our actions make us an authority about them? I am unsure whether Moran offers to answer to this question. He might regard it as a brute fact that everyone is an authority about the character of their own creations (including their actions). Alternatively, he might be offering us an explanation for this fact, an explanation which goes as follows: when we act, we act for a reason and the character of our action is determined by the character of our reasons. Subjects (in a normal frame of mind) know the character of their reasons – they know what practical considerations they find convincing and persuasive – and so they know the character of their action. A subject is an authority about what they are doing just in so far as they are an authority about their reasons.

If Moran is telling us the latter story then he must be operating with a highly rationalistic notion of agency. A moderate rationalist would require that truly intentional action be motivated by a reason, by the agent’s awareness of some respect in which the outcome intended would be desirable, without requiring that the agent act on (what he would think of as) his strongest or most powerful reasons. But this moderate rationalism does not suggest a substantial epistemology of agency: we can hardly explain how an agent knows what they are doing by supposing that they know what reasons they are acting on. To avoid this difficulty we are pushed towards a more extreme rationalism which insists that truly intentional agency is agency determined by the strongest reasons which the agent knows they have. Since knowing what reasons you have is not synonymous with knowing what you are doing, there is a substantial epistemology of agency on offer here. Yet such hyper-rationalism runs up against the fact that people intentionally do things they know they have sufficient reason not to do.

A hyper-rationalist can acknowledge that various forms of behaviour approximate to fully intentional and responsible agency but, for him, agency occurs in its pure form only when it accurately reflects the subject’s reasons.  I am unsure whether Moran is invoking this rationalist notion of agency because I am unsure whether he means to offer us a substantial epistemology of agency. This uncertainty is connected with another interpretative issue: how should we understand Moran’s discussion of practical irrationality and, in particular, his notion of akrasia?

Practical Irrationality

On the usual understanding of these terms, akratic or incontinent action is action which is performed even though the agent judges that it would be better for him to be doing something else. Akrasia is a very familiar phenomenon. If I eat more biscuits than I know I’ll be able to digest comfortably or remain slumped in front of the TV when I realise I ought to be making lunch, my eating and my lounging are things I do, things which I do deliberately, freely and intentionally, things for which I am fully responsible. In these respects, akratic action is no different from fully continent agency: it is agency par excellence.

Understanding akrasia in this way, I find some of the things Moran says about it rather puzzling. For example

when I know that I am akratic with respect to the question before me, that compromises the extent to which I can think of my behaviour as intentional action … Nor does a person speak with first-person authority about such conditions. (pp. 127-8)

Moran also says that the akratic does not “identify” with their action and that his knowledge of it is empirical and not “ordinary self-knowledge” (p. 67).[4] The implication is that while the akratic may be able to predict that he is likely to behave akratically and then exercise a sort of self-control by putting obstacles in the way of akratic behaviour (e.g. placing a time-lock on the drinks cabinet), he will lack that first-person knowledge of and control over his behaviour which the continent possess.[5]

Moran’s description fits Velleman’s psychoanalytic example very well but as a commentary on everyday akrasia, it is a bit overstated. My partiality to lunch time TV hardly qualifies as an obsession or a compulsion even though it tempts me to do what I know I should not do. I am not overwhelmed by the desire to watch, I choose to indulge it. If I am honest, I won’t deny that it is me who decides to remain on the sofa. Furthermore, I know perfectly well why I take this decision: the lunchtime soap opera is genuinely entertaining and diverting (though, even in my own eyes, other considerations are more pressing). There is no failure of first-person access either to the motivation for my akratic behaviour or to what it gets me to do.

Reflection on such everyday cases of akrasia puts pressure on Moran’s notion of an avowal. Moran says that there are two elements to avowal: first, an authoritative awareness of the state of mind avowed and secondly an endorsement of it (pp. 91-2). Consider a Catholic woman, described by Jackson, who has become pregnant after rape.[6] She judges that she ought not to have an abortion but akratically decides to have an abortion nevertheless and makes plans to attend the abortion clinic next week. Can she avow this intention? She can do more than report this intention in the way she might report a third parties’ intention but she can’t go as far as to endorse this intention. What she can do is to affirm that she is set on having an abortion, that she has resolved to have an abortion, that she is committed to having an abortion. Moran’s notion of an avowal seems to include the idea of endorsement (p. 67) so perhaps he would maintain that the woman can’t avow her intention. But if that is how the notion of an avowal is to be understood, it looks as if what manifests a distinctive knowledge of (and control over) intention is not our ability to avow these intentions but rather our ability to affirm them.

Until now I have been assuming that Moran is using “akrasia” in its standard sense but this may be a misreading. In several places, he is at pains to draw a distinction between what might be called pure evaluation on the one hand and decision-making on the other (where the notion of “deliberation” applies only to the latter)

the mere appraisal of one’s attitudes, however normative, would apply equally well to past as well as to current attitudes, and indeed may have just the same application to another person as to oneself. In itself such an assessment is not an essentially first-personal affair. Rather “deliberative” reflection as intended here is of the same family of thought as practical reflection, which does not conclude with a normative judgement about what would be best to do but with the formation of an actual intention to do something. (p. 59)

The implication of this paragraph is that Moran is not really interested in the relationship between someone’s judgement of their reasons and their action;  such judgement interests him only in so far as it involves the formation of an intention so to act. What really concerns Moran is the relationship between intention and action. If that is right we should focus our discussion of practical irrationality not on akrasia but rather on what I shall call irresolution.

Philosophers sometimes apply the labels “weakness of will” and “incontinence” to both akrasia and irresolution but they are not the same thing.[7] Akrasia is a matter of failing to intend and act in accordance with your practical judgement, your judgements about what you should (i.e. have most reason) to do. Irresolution is a matter of not sticking to intentions once formed, of giving in to the very ‘inclinations’ which the intention was formed (however akratically) to resist. As Jackson develops his example, his Catholic woman is by turns akratic and irresolute. She is akratic when she forms the intention to attend the abortion clinic but later on her scruples get the better of her and she irresolutely refuses to leave the house on the day of her appointment. We need not take a stand on the issue of whether this woman’s akrasia or her irresolution are symptomatic of irrationality. The present point is simply that they are not the same thing.   

I suspect that Moran tends to equate akrasia with irresolution (p. 81) and that this goes some way to explaining his view of akrasia. Moran argues convincingly that a person who knows they are irresolute will, as a matter of conceptual necessity, find it very difficult to form intentions (pp. 77-83, pp. 94-8). If past experience of her own behaviour convinces Jackson’s woman that she won’t attend the abortion clinic even if she decides to do so, it is hard to see how she can even decide to attend the clinic.  Saying “I intend to have an abortion but predict that I won’t” is rather like saying “I believe that it will rain but it won’t”. Both sentences express a state of mind that is more than irrational, it is paradoxical.[8]

Of course, an analyst might convince someone that they had a belief which they wished to disavow and the subject might express this by saying “I believe that my brother drove my parents to an early grave even though he didn’t”. But here the subject seems entitled to enter a qualification: this is not his belief in the full sense, he knows about it only via the analyst and he cannot assume full responsibility for it. It is plausible to say the same of intentions that the agent believes they won’t execute. The analyst may convince me that I still intend to proposition my childhood sweetheart someday even if I know perfectly well that I’ll never bring myself to do it. But, I can plausibly insist, this intention is mine only in an qualified sense. Fully self-conscious irresolution is indeed paradoxical. None of this applies to everyday cases of akrasia. There is nothing paradoxical about the statement “I know I ought not to have an abortion but I shall”. And there is little pressure to say that such a statement evinces a division of the person, or a diminishment of the person’s control over or responsibility for the intention.

Reading Moran’s discussion of practical irrationality as concerned with irresolution, he maintains that a person is an authority about what they are doing because they are an authority about the intention with which they act and they are an authority about the intention with which they act because this is something they have created, something which is itself an expression of their agency. Here it is the subject’s ability to affirm their intention, rather than their ability to endorse it, which ensures that they have knowledge of (and control over) what they are doing. But, read in this way, Moran is not offering the substantial epistemology of agency I suggested he might. Rather he is assuming that an agent has direct knowledge of both of what he is doing and of what he intends to do simply because he is choosing to do it.

To sum up, Moran thinks that there is close tie between what he calls the standpoint of practical deliberation and the possession of first-person epistemic authority: one is an authority about what one is doing and why just in so far as one is occupying this standpoint (p. 127). To occupy this standpoint is, by definition, to both make judgements about what you have reason to do and to implement those judgements in decision and action (pp. 63-4, pp. 94-5, p. 131, pp. 145-6).[9] Now it is clear that these two things can come apart: one can make practical judgements which one does not implement and one can take decisions or perform actions which do not reflect one’s practical judgements. The question for Moran is this: when that happens, is first-person authority necessarily compromised? Is a subject who fails to do what he thinks he ought to do ipso facto less well placed to know what he is doing and why? My discussion suggests a negative answer.

Conclusion    

I have focused on what I take to be a central theme of Moran’s book but my treatment has been far from comprehensive. For example, I have had to ignore his interesting discussion of belief, desire and emotion and I have failed to mention his illuminating criticisms of superficially similar positions (like Shoemaker’s). Two other features of the book are particularly worthy of mention. First, Moran offers us an interpretation of the work the early Sartre which enabled the present writer to read parts of Being and Nothingness with great profit. Secondly, I strongly recommend the fascinating final chapter in which Moran explores some virgin territory: the moral psychology of the first person.

Reading a work of academic philosophy, one often finds oneself slogging through chapter after chapter of critical commentary on the work of others, to be rewarded with only the vaguest sketch of an alternative view. Yet (to paraphrase Feyerabend) no sensible person abandons a theory because of a counterexample, only for a better theory. Moran has taken Feyerabend’s maxim to heart. His prose is elegant and engaging. He lays out his view in enough detail to expose its weaknesses as well as its strengths. Occasionally, I would have wished to know more about why he thinks his position should be preferred to alternatives. Nevertheless Moran has said enough to encourage others to build up the theory’s defences and highlight its advantages over more familiar approaches to the problem of self-knowledge.

 

 [1] Richard Moran, Authority and Estrangement (Princeton NJ: Princeton University Press, 2001), xc + 201 pp. Page references are to this work.

[2] D. Velleman – The Possibility of Practical Reason (Oxford: Oxford University Press, 2000) pp. 126-7.

[3] See also Moran discussing of emotions which one can’t avow on p. 93.

[4] On p. 131 Moran connects claims about alienation to claims about control and responsibility, saying that (certain forms of) the latter are compromised by such alienation (see also pp. 117-8).

[5] Moran says similar things about the desires which motivate action (pp. 116-20).

[6] This example occurs on p. 4. of F. Jackson – ‘Weakness of Will’ Mind (1984) 93, pp. 1-18

[7] R. Holton – ‘Intention and Weakness of Will’ Journal of Philosophy (1999) 96: pp. 241-62.

[8] This reading makes good sense of the parallels Moran draws between akrasia and self-deception (p. 67, p. 87)

[9] This may be too strong. Perhaps Moran intends only that the deliberative standpoint be one in which it is believed that one’s decision will reflect one’s practical judgement. I don’t think this will help. Someone who does not believe but merely hopes that they will take what they regard as the right decision in a difficult situation may still know, in the usual first-person way, what they end up doing (or deciding to do).

 

KNOWING YOUR OWN MIND[1], David Owens  University of Sheffield, United Kingdom

• Leave a comment... (0)

The anthropological literature on religion is diverse and voluminous, but there is one common perspective that pervades virtually that entire body of work, and that is the conviction that the epistemological principles of the scientific method cannot and/or should not be applied to the content of religious beliefs, on the grounds that nonempirical phenomena are necessarily beyond the purview of empirical science. Evans-Pritchard offers a familiar formulation of the position in Theories of Primitive Religion:

He [the anthropologist] is not concerned, qua anthropologist, with the truth or falsity of religious thought. As I understand the matter there is no possibility of his knowing whether the spiritual beings of primitive religions or of any others have any existence or not, and since that is the case he cannot take the question into consideration (Evans-Pritchard 1965:17).

Whatever personal convictions anthropologists may hold as individuals, the overwhelming majority have agreed with Evans-Pritchard that, as anthropologists, they either cannot or should not investigate the truth or falsity of religious beliefs. In virtually every major anthropological work on religion, and in most if not all introductory textbooks in cultural anthropology, the question of the truth or falsity of religious beliefs is evaded, ignored, or de-emphasized in favor of questions concerning the social, psychological, ecological, symbolic, aesthetic, and/or ethical functions and dimensions of religion. (see note 1)

Thus, for example, Anthony Wallace, who affirms that religion “is based on supernaturalistic beliefs about the nature of the world which are not only inconsistent with scientific knowledge but also difficult to relate even to naive human experience” (Wallace 1966:vi), nevertheless chooses to “ignore the extremes of fundamentalist piety and anticlerical iconoclasm” and to regard religion as “neither a path of truth nor a thicket of superstition, but simply [as] a kind of human behavior…which can be classified as belief and ritual concerned with supernatural beings, powers, and forces” (Wallace 1966:5). Similarly, Edward Norbeck, who recognizes that “religious beliefs and acts are created by man on the basis of his life” (Norbeck 1974:7), nevertheless explicitly restricts the anthropological study of religious beliefs to “interpretations of their role in human life and of the factors that have molded the customs into their particular forms” (Norbeck 1974:3). Clifford Geertz (1973:89), who defines religion as a system of “sacred symbols” which functions “to synthesize a people’s ethos…and their world view,” is completely unconcerned with the question of whether any particular religiously-supported world view is true or false. And Marvin Harris, who has long been one of anthropology’s most persistent critics of irrational modes of thought, nevertheless declares that he “can readily subscribe to the popular belief that science and religion need not conflict,” since science, he argues, “does not dispute the doctrines of revealed religions as long as they are not used to cast doubt on the authenticity of the knowledge science itself has achieved” (Harris 1979:6).

In short, a common element of the anthropological perspective on religion can be summarized in a simple syllogism:

1. The essential defining feature of science is empiricism (i.e., the belief that the only reality which exists is the reality amenable to the five senses, implying that reliable knowledge of that reality can be obtained only through the five senses).

2. The essential defining feature of religion is supernaturalism (i.e., the belief that there is a reality which lies beyond or somehow transcends the reality amenable to the five senses, implying that reliable knowledge of that reality can be obtained by means other than the five senses).

3. Therefore, science cannot be used to determine whether religious beliefs are true or false, since empirical epistemological procedures cannot be applied to supernatural phenomena.

Despite its virtual ubiquity in anthropology, that argument is unsound, for the simple reason that both of its premises are false. The essential defining feature of science is not empiricism, and the essential defining feature of religion is not supernaturalism. The conclusion that religion is or should be immune from scientific scrutiny is thus wholly unwarranted; moreover, that conclusion is also ethically objectionable. Considerations of disciplinary integrity, public welfare, and human dignity demand that religious claims be subjected to anthropological evaluation.

My position, then, is that anthropological science can and should be applied to the content of religious beliefs. My goal here is to establish three points: first, that rationality rather than empiricism is the key element of science; second, that irrationality rather than supernaturalism is the key element of religion; and third, that anthropologists have an intellectual and ethical obligation to investigate the truth or falsity of religious beliefs. The first point concerns the nature of science; the second involves the nature of religion; and the third, obviously, is a question of value.

• Leave a comment... (0)

A Problem for Goldman on Rationality
Francis Jeffry Pelletier
Department of Philosophy
University of Alberta 
 

The central concern of Knowledge in a Social World is to restore the notion of Truth to the rightful place of glory that it had before the onslaught of those pragmatic, cultural-studying, social constructing, critical legalistic and feministic postmodernists (PoMo’s, for short). As G sees it, these PoMo’s have never put forward any “real” arguments for their veriphobia; and, well, how could they, since their position is committed to the “denial of Truth” and hence committed to denying that there is any such thing as a valid or a sound argument. Instead, according to G, they have artfully inculcated various types of considerations so that a defender of Truth now must swim against the entire flow of academe to even stand a chance of making his or her voice heard.
 

G finds six overlapping currents in this wash of anti-veritas ideas, listing them in his §1.3 and devoting a further section to each of them in turn (§1.4-§1.9). These currents spring up in various later parts of the book for further discussion, but even in these places G refers the reader back to the material in Chapter 1 as refuting them. (Insofar as a current that is not based on any consideration of Truth, validity, or soundness is susceptible of refutation.) I think it is fair to say that G does not really intend that his comments about the anti-veritas ideas will be taken as definitive by the veriphobes, but rather that he believes the strength of his positive account in the rest of the book (which accords Truth a central role) will sway all neutral readers into a belief of the usefulness of Truth in an account of social knowledge. Still, I think many readers…even veriphiles…will find the considerations that G brings against the six currents to be disappointingly short.
 

I want to consider some of the currents in the stream of PoMo ideas. As I said, G’s six currents overlap at various points, and each of the currents has sub-flows. This means that we might put together sub-flows from different currents in such a way as to find a somewhat different current in the PoMo stream than was found by G. To lay my cards on the table at the outset: I am a veriphile, I am an anti-PoMo, I am a believer in individual epistemology and hold that social epistemology is entirely reducible without remainder to the knowledge of individuals. Except possibly for this last point I am in agreement with G. My task in this short commentary should be seen as trying to bolster one small part of his overall position by making one PoMo consideration as strong as possible and asking G to show that it still fails. Of course, no answer of G’s will convince a PoMo because they do not think that considerations of truth, validity, consistency, and the like are relevant to what should be believed. But perhaps it will help with a few of G’s neutral readers.  
 

The Argument from Bias:
PoMo’s in general, at least in the version given by G, deny the reality of rationality. The various currents that G identifies all have the goal of relegating the desire for Truth and the desire for the search for Truth into the backwaters of human endeavor. The underlying message one gathers from G’s presentation is that PoMo’s not only deny that there is any such thing as Truth but additionally they attribute any belief in it either to ignorance or to sinister motives. The sixth of the PoMo currents detected by G is that
 

Truth cannot be attained because all putatively truth-oriented practices are corrupted and biased by politics or self-serving interests.
 

G’s rebuttal of this current in PoMo thought is simple: he claims that it depends on two very implausible assumptions, namely (a) that all belief is driven by motivational biases, and (b) that these biases are in conflict with the Truth. G claims that (a) is wrong, in general, because one does not have the voluntary choice to believe or disbelieve something and hence it cannot be due to motivational biases. He claims that (b) is wrong because it is simply not true that knowing the truth runs generally against one’s interests (although speaking the truth might run against one’s interests).
 

In §1.9, where this argument is discussed, there is very little more said about either it or about these two assumptions, except for a discussion of the role of case studies in the history of science. His analysis of the issue in this section of the book is that all the plausibility of The Argument from Bias comes from its application to political discussion, where a conflict might be seen between one’s interests and speaking the truth. But, G continues, such a generalization is not correct for knowledge and truth in general. There are, of course, researchers who think that case studies in the history of science show that “any theory can always be retained as long as its defenders have enough institutional power to explain away potential threats to it”. In his discussion of these case studies, G acknowledges that politics and interests have played a crucial role in the development and acceptance of scientific ideas-something no one has ever questioned-but he denies that this is at all relevant to the issue of Truth. To be relevant, he says, would require that politics and interests were causally efficacious for the scientific beliefs and not merely temporally coincident with them. And this would require establishing the counterfactual thesis that the beliefs would not have occurred had the politics or interests of the situation been different. G denies that the case studies have ever established this. And anyway, even if it could be established, this wouldn’t be enough to show that the acceptance of these ideas-especially over a long period of time-stemmed from these same motivations.
 

This last item, that long-term acceptance is independent of motivational bias, is a theme that I believe G should have emphasized more. My view of long-term acceptance is like the following, and although I believe this is in accord with G’s general position, I wish he had more explicitly addressed the issue. Imagine a flat plane upon which various theses about the world are inscribed. Imagine further that the choice of which of these theses are accepted by an individual (or group of people) is indicated by the position of a marble (or all of the group’s individual marbles) on this plane. Certain factors deform the plane: sometimes a heavy weight is placed at a location, making that point lower and the marbles more likely to roll to that place; or maybe the plane is forced upwards at some location and the marbles tend to roll away from that location. A person’s (or group’s) beliefs over time are rather like rolling the marble from one point on the plane to see where it will eventually end up…perhaps with the occasional shake of the plane to represent violent new starts. My view is that individual’s motivational biases are best represented as weights (”attractors”) on this plane that make the marbles tend to roll towards them. And there perhaps are other motivational features that function as “repellors”. The individual’s marble will roll towards the attractors and away from the repellors, as will the group of marbles representing the group or social beliefs. But as I see it, the greatest attractor is Truth; and after a sufficiently long time all (or close to all) the marbles will end up in the deformity caused by the great weight of Truth. The marbles will have escaped from all the minor attractors and repellors, and will have been drawn toward the Truth.
 

Of course, the world of beliefs is not really a plane with various theses inscribed at different locations, and biases and the Truth are not really heavy weights placed on the plane to deform it into the third dimension. No: the world of beliefs is really a high-dimensional hyperplane and biases and the Truth are really attractors/repellors that deform the hyperplane in a higher dimension. However, our beliefs, especially our beliefs over time, are just marbles that roll around in this hyperplane and are attracted/repelled by these various forces. But again, the factor that accounts for the majority of the effect is the super-attractor, Truth.
 

Gigerenzer/Cohen/Cosmides
 G continues his discussion of the PoMo current which uses The Argument from Bias in his Chapter 8, and devotes §8.3 to a further rebuttal. In this Section, G concentrates on specific biases that might be mentioned by PoMo’s (although I think they never have been) but which have been the subject of a dispute in Cognitive Science, concerning the nature of rationality and what it means to be rational.
 Biases of the sort that affect one’s acceptance or rejection of “factually-oriented” statements come in two flavors: hot and cold. The hot biases are the ones caused by emotions, desires, special interests that “blind” one to “the facts”, and the like. Such a bias will make one believe that one’s spouse is faithful, despite all the clear evidence to the contrary; they will make one “refuse to even consider” the benefits of gun registration, and the like. But it is not these hot biases that I wish to consider here; instead, cold biases are those which are due to the inbuilt mechanisms of human cognitive functioning. Thus, even in cases where one is attempting to “dispassionately investigate” some phenomenon, one’s inbuilt cold biases may lead one to “false” answers. PoMo’s could use this fact (if it is a fact) to argue that the notion of Truth is inevitably bound to be dependent on these cold biases and to have no independent use or validity. So far as I know, this exploitation of cold biases has not been followed up on by PoMo’s, but it is clear that it could be, possibly with even more plausibility than their employment of the similar case of hot biases. After all, with hot biases we want to say that they can be overcome by “cooling down” or “stepping back” and letting one’s underlying rationality come to the fore. But cold biases offer no such method, for the biases are built into the very structure of our rational nature.
 

 

Long ago it was discovered that many (most) people are “remarkably bad” at simple logic problems and that they similarly are “distressingly unable” to perform even the most simple tasks of estimating probabilities. Although such a discovery comes as no surprise to teachers of elementary logic and elementary probability theory, the detailed psychological study of this “shortcoming” took off in the deductive logic realm with the publication of Wason (1966) and in the probabilistic reasoning realm with the works of Kahneman & Tversky (there have many joint works, see Nisbett & Ross 1980 or Evans 1987 for summaries). In 1981 Cohen challenged these findings partially on the grounds that rationality was defined to be what people’s reasoning competence allows them to do. Therefore, he concluded, it cannot be true that people are “systematically biased”. Members of this “optimistic camp” (optimistic because they believe that people are inherently rational) have since that time been engaged in the exercise of explaining how the original experiments are in some way wrong. G is a member of this optimistic camp, and he employs some of the standard “explanations” with an eye to showing that the PoMo’s are wrong in their appeal to this sort of bias as a justification of denying Truth. His general conclusion is that cold biases (a) do not happen as often as PoMo’s would require, nor (b) are they necessarily detrimental to the use of Truth when they occur.
 

Here is a cross-section of the optimistic suggestions concerning how and why optimists think the alleged biases do not really count against rationality as a route to Truth. We’ll see that all of them, including the one advocated by G, are variations on the same theme; and I will indicate why I think this reply to PoMo’s just can’t be successful. Cohen (1981) postulated that the probabilistic reasoning that was actually demonstrated by subjects – the “bad reasoning” which Kahneman & Tversky had uncovered – was in fact the correct way to reason. This way of reasoning was formalized as a “Baconian probability calculus” which Cohen then put forward as the correct way to reason probabilistically (as opposed to the Baysean calculus that was assumed by Kahneman & Tversky when they made their judgment that people were inherently biased). A variant on this sort of position is taken by Gigerenzer (see for example his 1991, 1994, 1998). Gigerenzer believes that the flaw in the Kahneman & Tversky experiments was that they were trying to get subjects to give the probability of single events (such as the probability that Linda is a feminist bankteller) instead of the frequency of a type of event happening within a larger space of events. In a series of very compelling experiments Gigerenzer has shown how to make (most) of the alleged biases disappear by suitably rephrasing the experiments in terms of frequency judgments instead of single-event probabilities. The difference between Cohen’s and Gigerenzer’s strategies is that Cohen wanted to do a wholesale replacement of Baysean probability theory, whereas Gigerenzer agrees with the predictions of Baysean theory in those cases where there is a frequency interpretation of the case. He only disagrees with a certain portion of Baysean theory, viz., the probability of single events; and his claim is (a) that in the good portion of Baysean theory (where there are no single-event probabilities) people will agree with Baysean theory if the material is presented to them in a frequency manner, (b) even in the bad portion of Baysean theory, if the problem can somehow be rephrased as a frequency problem then people will agree with the Baysean answer, and (c) all the remaining cases are those where subjects are asked to find the probability of a single event and this is simply not a well-defined notion (so it is not surprising that subjects cannot do it coherently). Like Cohen, Gigerenzer is challenging Bayseanism as being the normatively correct theory of probability.
 

In the realm of deductive reasoning a similar battle is being waged. Wason’s (1966) results show that subjects do not follow classical logic in their assessment of simple reasoning involving conditionals and negations. So far as I am aware no one has really taken up the type of answer that Cohen gave for the probabilistic reasoning case – to construct a new deductive logic that obeys the results obtained in the experiments and claim that this is the normatively correct logic. (Although there have been some mutterings about “fuzzy logic”, “relevant logic”, and the like.) Instead, the main force in the literature has been more akin to Gigerenzer’s strategy: find particular types of problems that subjects can perform well on (well, as measured by classical logic) and then claim that this is the normatively correct notion of (deductive) rationality. The types of problems that people do not perform well are then relegated to the realm of “not well-defined problems”, much as the probability of single events is disparaged by Gigerenzer. In the deductive realm this course is championed most feverently by Cosmides (1989, see also Cosmides & Tooby 1987, 1994). Her claim is that people can perform well on the Wason task so long as the logic problems ask about “social interactions”, particularly about whether some “cheating” is going on. If we give people this problem:
 

From “No one who has not paid their dues can enter the meeting-house” and “Joe is allowed to enter the meeting-house” can we infer “Joe has paid his dues”?
 

then everyone will correctly answer “yes”. Yet, if the question is put:
 

From “No non-F’s are G’s” and “a is a G” can we infer “a is an F”?
 

we find that most people will answer this incorrectly (according to classical logic). Cosmides’ position is, like Gigerenzer’s, that these “abstract” questions simply do not have any “rational” answer, because the notion of (deductive) logical rationality is only defined for problems of the “cheating detection” variety. (And she has an evolutionary-psychology justification for this.) The background logic of her theory agrees with classical logic in those cases where it is defined. But there are other cases where it is not defined and people’s performance on these cases cannot be used to argue for or against human rationality.
 

According both to Gigerenzer and Cosmides, (a) there is an objective normative standard in the fields of probabilistic and deductive reasoning, and (b) people’s underlying rational competence matches this standard. The (a) aspect reiterates their commitment to veritism and the (b) aspect demonstrates their commitment to optimism. And it is to this picture that G cleaves. But can it really be used as an argument against the PoMo current about biases? It seems to me to have at least the following problem. The main idea of Gigerenzer and Cosmides is to presuppose the existence of objective rational standards, and then to use human performance to show the restrictions on the scope of these standards. It is true, they say, that people do not perform logical operations in accordance with the dictates of classical logic (e.g., they misapply contraposition when problems are presented in “abstract formats”) nor do the perform appropriate estimations of probabilities in accordance with Baysean theory (e.g., they misapply information about “base rates” if the information is given to them as a probability rather than a frequency). But this does not show that there are biases, they say; for the correct conclusion is to deny that human rationality extends to these cases. (Such cases are instead seen as some sort of “theoretical extension” of the basic theory of human rationality in the same way that imaginary numbers constitutes a theoretical extension to the ordinary notion of number.)
 

I do not see how adopting the Gigerenzer/Cosmides viewpoint can serve as an argument against PoMo. For, the main idea of Gigerenzer/Cosmides is to presuppose there to be Truth and to then use this to show the reach of human rationality, whereas the PoMos are using biases to show the non-existence of Truth. As an analogy, consider arguing with someone who claims that everyone believes in a supreme being. You proceed by pointing to those who have denied believing and have said so in their writings; and you might even point to yourself. But just when you think the opponent is bound to give in (after all…who better than you to know whether you believe in a supreme being?), s/he takes the “whatever you hold to be most important-that’s what you hold to be a supreme being” move. And just as we find this an intellectually unsatisfying maneuver, so too should PoMo’s find unsatisfying G’s definition (or the Gigerenzer/Cosmides definition) of  ‘rationality’ as whatever aspect of performance agrees with some pre-given normative standards. After all, PoMo’s have not yet even admitted there to be any such normative standards. I would like to see G address the “pure issue” of the possibility of (cold) biases making it impossible ever to know whether Truth has been attained without presupposing the existence of Truth. G recognizes, p. 234, that he has “not refuted the heuristics-and-biases approach root and branch.” But I would say that even if he could refute the Wason/Kahneman/Tversky tradition, his methodology presumes that there is a normative standard of Truth, and therefore this refutation would not (should not!) dissuade the PoMo’s.  
 

The Denial of Epistemic Privilege
 Another current G finds in the PoMo river is the denial of epistemic privilege:
 There are no privileged epistemic positions, and no certain foundations for beliefs. All claims are judged by conventions or language games, which have no deeper grounding. There are no neutral, transcultural standards for settling disagreements.
 

 

Having introduced my main worry with this aspect of G’s work, let me carry it a bit further into this other current of PoMo thought. In §1.7, where this current is considered, G spends much of his time discussing Rorty’s (1979) arguments against “privileged representations”, rightly remarking that even if Rorty’s claims here were right they would not tell against any of the other versions of epistemology on the market, especially G’s own “reliabilism”. A second sub-flow of this current concerns the idea that claims are always judged by conventions, perhaps the conventions that are part of the “language game” being played by the agents or by the agreement of others in the culture. It would follow from this that there can be no independently-justified cross-cultural standards. Once again G is swift in his dismissal of the claim (p. 29)
 

Careful reflection on judgments of justification suggests-as reliabilism maintains-that a belief is considered justified if it is arrived at by processes or practices that the speaker (or the community) regards as truth conducive. …Thus, judgments of justification are not without grounding, nor are they purely conventional. They are grounded precisely on appeals to truth conduciveness. Furthermore, there seems to be nothing arbitrary in a concept of justification tied to truth conduciveness.
 

The reason there is “nothing arbitrary” in this is that “people have an interest in truth” and those beliefs “formed by truth-conducive processes are more likely to be true.” A final sub-flow of this current concerns the notion of “agreement”. G thinks that PoMo’s believe that veriphiles require agreement for their notion of rational justification. G also thinks that PoMo’s believe such agreement to be impossible because they also believe there to be no transcultural principles for settling disagreements among different communities. G tries to calm this sub-flow by distinguishing the “ultimate aim” of epistemic practices from the specific methods adopted in pursuit of these aims. He claims that it has always been the pursuit of Truth that different cultures have aimed at, although these different cultures adopt many, many different and radically opposed methods in their pursuit of Truth.
 

Here’s a problem for this picture. I might also mention that it is a problem for my view that I expressed as the simile of marbles rolling on a plane. Both G’s view and the one I mentioned require that Truth be a causal factor. We have just seen that G makes Truth be the cause of a pursuit by people of any culture at all; and in my simile, Truth played the role of deforming the plane so that the marbles representing people’s beliefs would ultimately roll toward that place on the plane. In my picture I also had the biases and motivations be deforming features of the plane. Now, it’s pretty clear how biases and other motivations can be causally efficacious in producing belief. But the same cannot be said about Truth. It is, I think we would all agree, a non-natural property. It certainly is not a property that can easily be thought as exerting some mysterious force that attracts people’s mental states. Indeed, it seems to be qualitatively or categorially the wrong sort of property to do this. It is one thing to say that the cause for 82% of the people asked thinking that Linda is more likely to be a feminist bankteller than to be a bankteller is due to some mental short-circuit or bias to ignore conjunctions (or whatever). But what will then be the cause for the other 18%? Don’t we need some explanation on the same “level” as bias to explain their performance? How can we just simply appeal to the truth of the matter as the operative causal factor? (This is an adaptation of the “strong programme” requirement of symmetry of explanation-see Barnes & Bloor 1982 for further explanation. Although G mentions Barnes, Bloor, and Barnes & Bloor, he does not discuss this aspect of the strong programme.)
 

I would be very grateful to G if he could give a convincing answer to this problem.   
 References
Barnes, B. & D. Bloor (1982) “Relativism, Rationalism, and the Sociology of Knowledge” in M. Hollis & S. Lukes (eds.) Rationality and Relativism (Cambridge: MIT Press) pp. 21-47.
Cohen, L.J. (1981) “Can Human Irrationality be Experimentally Demonstrated?” Behavioral and Brain Sciences 4: 317-370.
Cosmides, L. (1989) “The Logic of Selection: Has Natural Selection Shaped how Humans Reason? Studies with the Wason Selection Task” Cognition 31: 187-276.
Cosmides, L. & J. Tooby (1987) “From Evolution to Behavior: Evolutionary Psychology as the Missing Link” in J-P. Dupré (ed.) The Latest on the Best: Essays on Evolution and Optimality (Cambridge: MIT Press) pp. 277-306.
Evans, J. (1987) Bias in Human Reasoning: Causes and Consequences. (Hillsdale, NJ: Lawrence Erlbaum).
Gigerenzer, G. (1991) “How to make Cognitive Illusions Disappear: Beyond ‘Heuristics and Biases’” European Review of Social Psychology 2: 83-115.
Gigerenzer, G. (1994) “Why the Distinction between Single-event Probabilities and Frequencies is Important for Psychology (and Vice Versa)” in G. Wright and P. Ayton (eds.) Subjective Probability (NY: John Wiley & Sons), pp. 129-161.
Gigerenzer, G. (1998) “Ecological Intelligence: An Adaptation for Frequencies” in D. Cummins & C. Allen (eds.) The Evolution of Mind (NY: Oxford University Press), pp. 9-29.
Kahneman, D., & A. Tversky (1973) “On the Psychology of Prediction” Psychological Review 80: 237-251.
Kahneman, D., P. Slovic, & A. Tversky (1982) Judgment Under Uncertainty: Heuristics and Biases. (Cambridge: Cambridge UP). 14
Nisbett, R. & L. Ross (1980) Human Inference (Englewood Cliffs: Prentice-Hall).
Rorty, R. (1979) Philosophy and the Mirror of Nature (Princeton: Princeton Univ. Press).
Wason, P. (1966) “Reasoning” in B. Foss (ed.) New Horizons in Psychology London: Penguin Books).  

• Leave a comment... (0)