The (Im)Possibility of Rationality

“The important point, however, is that there must be some normative principles bearing on PAs [propositional attitudes] to the effect that we ought to modify our PAs because of how they are. Rational normative principles are of the form that if one has such and such PAs then one ought to modify them in such and such ways. Unless there are such principles there is no rationality and no reasoning.”  – Nick Zangwill, “The Normativity of the Mental”

This is the simplest, most direct argument for normative ‘realism’ I’ve found. It is refreshingly plain in that it wears its transcendental structure on its sleeve: believing, wanting, and the like must be normatively governed psychological processes if rationality and reasoning are to be possible, and rationality and reasoning are not only possible but abundantly actual.  Moreover, it’s a wonderfully, boldly assertive philosophical mouthful, just the way I like it; when it comes to the root philosophical issues (and there is nothing more root than the normativity of the mental; everything of a peculiarly human interest stems from this) they invariably involve at their core simple, stark choice points.  Capturing those choice points in simple, direct language takes philosophical talent, and Zangwill brings plenty. But what makes philosophical disputes so thrillingly interminable is that virtually no two philosophers conceive the choice points in precisely the same way: God, the Devil, and each and every Man is bound up in his own peculiar details.  More on this presently.

Zangwill’s directness forces us to confront the chasm across which realists and nihilists about normativity self-confidently shake their fists.  Hume (spiritually a nihilist though not immune to the normative demands of the flesh) famously claimed that (providing one could actually make sense of ‘ought’-assertions) it seemed “altogether inconceivable” how any assertion of what ought to be could be derived merely from assertions about what is or happens to be (or was or will be).  Nor has anyone since managed to make such a derivation uncontroversially conceivable (certainly not in any form that doesn’t cry out to be relativized, contextualized, or personally monogrammed).  Yet the normative principles Zangwill is certain exist require that there be such a derivation.  Merely from knowing what you believe or desire I can know what else you ought to believe or desire (and without having to know you).  The very activity of reasoning and rational interpretation demands we move from is to ought routinely.  The corollary is that Hume’s perplexity (and, what’s probably worse, that so many of us persist – more than 260 years and counting – in being impressed with Hume on this matter) is really the altogether inconceivable bit.

When we consider the sorts of examples of normative principles governing PA’s that Zangwill mentions – the very same jejune examples with which the recent scholarly literature on the topic is replete – we might initially think the normative reifier has a point.  If we desire to X then (to that extent, at least) shouldn’t we intend to X?  If we desire to X  and we believe that we can X only if we Y, then (focusing on those two facts) shouldn’t we desire to Y?  If we believe that p, and also we believe that if p then q, then (considering those beliefs alone) we ought to believe q, right?

The italics of the preceding paragraph were of course meant to emphasize the intended modality of the forgoing principles, the very modality Hume thought “shou’d be observ’d and explain’d.” We’ve just observ’d it; as for explain’ng it, presently I will do no better than to say that such principles it would be best (is ‘better’ enough?) to conform to but there is no guarantee that we will.  In other words, the principles lead us to rationality but they can’t make us think; we can buck at the reigns of rationality and so be irrational.  And this is what’s really important, that our theory of the mind allows for and explains how we can be irrational. The raison d’être of philosophy itself is to be found in the fact that we are nature’s irrational animals.

The hitch, however, is that there is irrationality and then there is irrationality.  The first kind is familiar enough and seems fairly easy to explain.  Sometimes we desire to X but we are feeling so blue that we don’t bother to intend to (obtain) X.  Sometimes we desire to X and believe that we can X only if we Y, but find Y to so forbidding or distasteful that we can muster no desire for it.  Other explanations of similar examples would follow the same structure, which essentially involves appeal to mental states, conditions, or circumstances other than those specified in the principles (and so going beyond those principles ‘pro tanto’ status).  Being so explainable, however, renders ‘irrational’ little more than a label: given the explaining conditions, our ‘failure’ to intend to X or desire to Y is so understandable that we are tempted to say that (pro tanto, of course) we shouldn’t intend to X or that we ought not desire to Y.  This sort of irrationality is cheap.

It’s the second kind of irrationality that caused Socrates to cross swords with Protagoras, Kant to reject Hume, and Zangwill to be dissatisfied with Davdison. The latter member of each tandem has no explanation for this kind of irrationality.  Neither does Socrates, Kant, Zangwill or anyone else.  But what puts these philosophers at odds is the explanation they offer for their lack of explanation.  Protagoras, Hume, and Davidson simply don’t acknowledge the possibility of the kind of irrationality so close to the hearts of Rationalists of all stripes, an irrationality that would have us failing to intend to X or desire to Y even in the absence of the sorts of explaining conditions appealed to in the first kind of irrationality (and you wonder why they can’t explain it?).  What Zangwill and every ‘realist’ about normative principles bearing on PA’s insists is that it is possible to desire to X but have no intention to Y, or to believe both that p and that if p then q, and yet fail to believe that q, not because some other psychological factors are interfering in these cases of intention formation and belief fixation, but simply because of our…well…naked irrationality.

I’m with Protagoras, Hume, and Davidson (though I don’t think understanding rationality ‘interpretationally’ is required or even the most fruitful way to conceive of it): I don’t believe there are any values for the variables in these putative normative principles that would yield a plausible case of ‘pure’ irrationality. I have yet to encounter any real person – including myself – who self-consciously believed that p, believe that if p then q, but did not believe q, for some univocal values for p and for q.  But that isn’t because I run with an especially rational crowd.  In any case, consider the question for yourself, and to test whether you find it at all possible to self-consciously violate any of these ‘norms’.  If you cannot, then we might well say that the principles in question aren’t normative, in the standard, philosophical sense, but rather capture what takes place in normal circumstances (when there are no other psychological factors sufficient to disrupt and detour ‘reasoning’).  The ‘shoulds’ and ‘oughts’ of these principles would then be best explain’d as ‘shoulds and oughts of expectation.’  But if that is so, then the sort of ‘rationality’ Zangwill has in mind – not the cheap and easy sort that we possess (from some perspective) no matter what we do, but one so rich that we have to go out an earn it, that sort of rationality – might well be altogether inconceivable and impossible after all.

Advertisements

11 comments

  1. Are you suggesting, Joe, that the proper epithet for someone who does not reason according to modus ponens is not ‘irrational’ but rather something less normative? Something clinical, something like ‘broken’, perhaps? Your suggestion seems to be that this is the most substantive sort of normative claim we can make and that it is so minimal as to warrant being called something else in order to make clear that substantive normativity is not in the offing. Have I got it right? I would – you’re not surprised, even after all these years – say that this is surely a normative claim, that it is among the most basic of normative claims, and that we can derive more substantive (read: more philosophical) normative claims from it.

    In other news, how are you? Haven’t talked in ages…..

  2. Dave! Good to hear from you. What I’m suggesting is that it is impossible to “not reason according to modus ponens.” And I suspect this not on the basis of any Davidson/Dennett-like assumptions about the possibility of interpretation (at least I don’t think I’m making such assumptions, though what I am basing that claim on might not be that far removed from them), but rather from first-person considerations about belief-fixation. It seems as though I cannot intentionally, or self-consciously make myself such that the self-ascription of violating modus ponens fits. To the extent that I ‘accept’ the charge made by another, I’m am very likely to claim that I hadn’t thought of either p or q (or both) ‘in that way’.

    Here’s the thought I’m suggesting in another way: any putative failure to reason in accordance with modus ponens I believe to be in principle explicable. But if that is so, wouldn’t the conditions and circumstances appealed to in the explication ‘rationalize’ the piece of reasoning in question? That is, given those conditions, didn’t I reason precisely as I ‘should’ have? But if that is so, what is the normative force of the ‘should’ (here or in the ‘normal’ case)?

    1. Dave, I want to take another crack at clarity. I believe that among the issues that divide those disposed to think there is a robust (re: extra-psychological) normative standard to which each individual’s psychological processes are subject and those that don’t, is whether or not there are examples of irrationality that on some fundamental level indict us as individuals (if not as a species), because we are, somehow, capable of avoiding them (and so being rational) but we culpably do not. Now I am understanding that this sort of irrationality is distinct from the kind I referred to in the post as ‘cheap,’ irrationality explained-away, so to speak, by reference to other psychological conditions and circumstances (e.g., depression, addiction, distraction). Two different considerations lead me to (or perhaps rationalize my prejudice) doubt the possibility of such irrationality.

      1. The First-Person Perspective Irrationality Paradox

      An awkward mouthful, yes, but try to intentionally be (instrumentally) irrational. If you ‘succeed’, that is because you have taken sufficient (and possibly necessary) means to your irrational ends, which sounds very rational. If, however, you can’t; if being irrational is something we only do inadvertently, when we aren’t looking where we’re going, why isn’t such irrationality best understood as the cheap kind?

      (This consideration, by the way, strikes me as having the closest affinities with the Davidson/Dennett interpretational stance on rationality, though I’m not at all convinced (at present, anyway) that they amount to the same thing or that our occurrent self-understanding as necessarily rational
      is a special case of the intelligibility constraint on the interpretation of others.)

      2. I can’t (yet, anyway) give up my commitment to the Principle of Sufficient Reason

      Cheap irrationality is explicable irrationality, and being explicable it is understandable, but in being understandable we come to see how it might have been ‘rational from the inside’. Rich irrationality strikes me as (having to be) inexplicable and, as such, I just don’t understand what it is or why it is a failure or what the point is of being concerned and preachy about it.

      So, where do I go wrong (by your lights!)?

  3. When I turn on my coffee pot and the coffee pot fails to make coffee, I say the thing is broken- this because it is not doing what it is supposed to be doing. If Amy points out that I’ve accidentally unplugged it, I retract my normative claim. So the coffee pot ought to make coffee and if it doesn’t then it is not acting as it ought, and this normative claim retains force just in case there is not some cause external to it that is responsible for its failure to do what it is supposed to do. This is my kind of normativity, and I think we can apply it to persons as well as coffee pots. If a person fails to do what he ought and if this failure cannot be attributed to causes external to the person, then the person is failing to act as he ought. If a person fails to believe or think as he ought and if this failure cannot be attributed to causes external to the person, then the person is failing to believe/think as he ought– he is irrational.

    What does it mean to fail to act/believe/think as one ought? That’s the million dollar question, of course, but let’s table that for the moment. Do you think this standard of irrationality constitutes the ‘cheap’ kind or is it sufficiently substantive to threaten your normative nihilism (if, of course, we can make sense of it)?

    DS

    1. What do you mean by ‘external cause’ in this context (and what would constitute an ‘internal cause’, especially in relation to a coffee pot)?

      If a coffee pot is ‘broken’, because you dropped it on the floor or because the wiring has become fried and no longer carries an electrical current, then (if we were to oddly insist on using the term ‘ought’) we would say that the pot ought not to make coffee when you turn it on. Moreover, I believe this statement and the statement that the unbroken coffee pot ought to make coffee are both naturally understood as conveying expectation. When expectations are not met, we look for reasons why. When we find them, we ‘retract’ the normative claim, which amounts to adjusting expectations to the the revised context (“of course it didn’t make coffee, it was unplugged! Indeed, the shocker would have been if it had.”).

      As for humans, how ‘ought’ they to act? Beyond human expectations and hopes? In any case, for those who don’t act as they ought, if that ‘failure’ cannot be attributed to ‘external’ causes, wouldn’t you then attribute it to ‘internal’ causes? And wouldn’t this amount to giving ‘reasons’ for the failure, reasons that make the failure fitting? It’s this that makes normativity cheap in my mind.

  4. Joe,
    I think we can make sense of talk of normativity, though I suspect that in the end the sort of normativity I believe in is the sort you think is cheap. (You always did like nicer things than I did….). I also think we can make sense of talk of irrationality, though I think the argument for this is different in interesting ways from the argument for normativity simpliciter. Let me try to work some of this out…

    If the coffee machine does’t make coffee, it’s broken: it’s function is to make coffee and it is failing to perform its function. It should be heating water and pumping it through the coffee grounds, but it’s not doing so. There’s a simple normativity, and it obtains even when (especially when!- more on this below) we don’t know the cause of the failure. We employ normative language when there is a failure of proper function.

    This isn’t restricted to artifacts. If I can’t see out of my left eye, I go to the doctor because my eye’s not working as it should. We tend not to employ talk of ‘broken-ness’ here, but the case seems pretty similar to the coffee maker case. The eye has a job to do, it’s failing to do the job- it should be”‘allowing me to see” things that are in this part of my distal environment but it’s not doing so. (Indeed, if were to unpack the shudder quotes in the previous sentence, we would say that the retina in the left eye should be receiving light waves in such-and-such a way…. but it is failing to do so; the more mechanistic language makes the parallel with the coffee maker that much stronger.)

    This is, you want to say, a cheap normativity, and one that can’t be applied to persons and to the more meaningful contexts of moral, epistemic, or aesthetic normativity because the proper function model doesn’t work on those levels. I’ll not say much about aethetics, though I do think that Dan Kaufman has (had?) an argument for aesthetic value that fits nicely with the argument I am developing here. (No surprise, his is a virtue-theoretic account of aesthetic value and so is similar to my proper function account of normativity.) I’ll not say much about moral normativity either, because I am so much less confident in my own views on it- though I do think that a version of what I have to say about epistemic normativity will work for the case of morality….

    If we take your case of the agent who accepts A and accepts [A->B] but who does not accept B, then I think we have a case of someone who is failing to reason as he ought, that whatever part of the brain is responsibile for inferential reasoning isn’t functioning properly. That we don’t know people who actually fail to make such inferences complicates matters because, as you say, any such failure has got to be caused by something…. but notice that the failures in the cases above can be caused by something and we still want to use the language of normativity. Why is the case of epistemic normativity different?

    I think this is interesting and I’m not sure what I make of it: in the coffee maker/eye cases, knowing the cause of the malfunction doesn’t change our normative assessment (it’s still broken, and damn it Joe, it’s because you dropped it!) but in the epistemic case, it does- if the agent isn’t reasoning properly because of a brain tumor pressing down on her frontal lobe or because she is on drugs, then we excuse the irrationality and say that the failure is due to the illness or the drug. We seem to restrict the epithet ‘irrational’ to inexplicable cases of failure to reason as one ought, to inexplicable failures of proper function. And you want to say, since such failures always have a reason, ‘irrational’ becauses a case of empty reference.

    Here’s where things get messy- I think this is wrong and I’ve got a few ideas about why it might be wrong. I want to say that causes and explanations are not the same things, and that even if something has a cause, it is not thereby explained– so a faiure to reason properly might have a cause but isn’t therefore explicable. I want to say also that maybe our restriction of the term “irrational” to cases of inexplicable failure is problematic, grounded in some meshugah, Enlightenment notion of autonomy and the infalibility of the agent as rational being (if there is something wrong, it must be explicable and external to the agent).

    As you have suggested, however, the case of the agent who fails to obey modus ponens is a troubled case because, really– who fails to do that!? If we take someone who refuses to believe the evidence in front of her, however, I think we’ve got a more interesting case. Consider someone who refuses to believe that she is going to fail my class even though she has failed all of the exams going into the final and who has not demonstrated any understand of or competence in the material – Is she irrational? Suppose she says she believes she will pass because she has worked really hard and she just knows that she will do better on the final. She’s got a reason to believe,to be sure, but she is utterly deluded. Do you really want to say that this is not a case of a failure of proper inferential function? It’s explicable, to be sure, but I don’t see how this matters. As I stand in front of my broken coffee maker willing it to work and refusing to believe that it could really be broken, it still is broken. How are the cases different?

    DS

    1. Wait, she was in your class also?

      I find this case (the ‘deluded’ student) more interesting as well. This is the kind of case that we encounter and which troubles us most (not the straightforward failure of reasoning according to modus ponens; I think philosophers who focus on mp-failure put themselves at a rhetorical disadvantage, not to mention needlessly cloud the issue of irrationality-no wonder we we profit so little from such discussions). And we naturally find ourselves calling such a person ‘irrational’. I am not objecting to that. What I’m questioning is just what such a predication amounts to. You use the phrase ‘failure of proper inferential function’, as though we might liken it to the failure of a the inner workings of the the coffee pot to pump hot water through the grinds. But I’m dubious about the analogy. The kind of step-by-step inferential moves we claim to be ‘proper’ are generally (exclusively?) logical (modus ponens, disjunctive syllogism, conjunction, etc.); as such, they are formal. What strikes me as unnatural about the way philosophers often talk about irrationality is there framing of their discussions in formal terms: I don’t think we come across any indisputable cases of failures on the formal level. The problems become more tractable at the level of content. The student in your example doesn’t just believe p and q and r, but a whole number of specific things. Putting aside the possible doubts she harbors but is not sharing with you (perhaps in part to better persuade you and herself), she believes she has studied and believes her studying will guarantee success. She also strongly desires the success. I suspect that if we were to assign a weighting to these various attitudes, we would accurately predict her actual belief that she will do better on the final. In other words, I suspect her ‘inferential’ capacity is working as would be expected, given the inputs to the system.

      Our problem with the student is really with the inputs. Why does she discount her past failure? Why doesn’t she take ownership of her incomprehension? Why does she believe, seemingly for no good reason at all, that this time it will be different? It seems to me that you’ve got to say that these problematic attitudes are the result of ‘failure of proper inferential function’. But then I’m going to suggest that these also are going to be ‘rationalized’ by other inputs to her system. And so on. At some point we just have to say this is how this particular system takes on information and weighs its relevance. Justification has to begin somewhere.

      I’d like to hear more about the distinction you want to draw between causes and explanations. That distinction would seem to block one of the argumentative moves I am making, but I’m not sure I see the how you can sustain it.

      Finally, I’m happy to apply ‘cheap’ normative claims to persons in moral, epistemological, and all sorts of contexts. Indeed, I think that is precisely what we do. Being cheap, such normativity is ubiquitous. What I’m saying is that this is very likely the only kind of normativity there is.

  5. Sorry for the radio silence, Joe– things got busy here. I agree with your assessment of our deluded student’s reasoning process and I think your treatment of it allows me to pinpoint at least one place where we disagree. Your write: “It seems to me that you’ve got to say that these problematic attitudes are the result of ‘failure of proper inferential function’. But then I’m going to suggest that these also are going to be ‘rationalized’ by other inputs to her system.” Fair enough, but the passive voice construction masks something important– she is going to rationalize her situation. And I have no doubt that she will; she doesn’t think she is being irrational. But she *is* being irrational. It is from the outside that we recognize her irrationality, from the outside that we see her reasoning process as a process of rationalization rather than clear thinking.

    Given what we know about her circumstances, we want to say that she is reasoning poorly and that she ought not to reason in such a way given her cognitive abilities and her evidence. (If her cognitive abilities were impaired – see above re brain tumor – we would excuse her.) Now, you say she is reasoning quite well given her abilities and the evidence-as-she-understands-it. And I want to say that she should understand the evidence more or less as I do, that this failure to see things for what they are is the root of her irrationality. And that this is the case even if she steadfastly denies that she is seeing things wrongly.

    I need some arguments here, I know, but I want to pause for a second because I think this highlights two important differences in our approaches to the matter. I think we can blame her for her mis-apprehension of the evidence and so accuse her of irrationality, while you think she is reasoning well given her apprehension of the evidence and so cannot be considered irrational. Your’s is a process-oriented or procedural conception, mine is…. something else. In addition, mine is a third-person conception of irrationality while you’re employing a first-person conception. Something to think about there, perhaps….

    She should understand the evidence diffierently– she is foolishly committed to the conceit that she will master material that she her heretofore been unable to even understand. I press her on this in office hours– why on earth do you think you are going to be able to do it? If she gives me a good answer (‘I finally figured out X and everything else clicked into place’) I might recant my normative assessment. If she doesn’t give me an good answer, though, then I say that she is deluded- that she is refusing to see what is evident to everyone else. (We use language like this when it is obvious to everyone except the poor husband that his wife is cheating on him– he’s making excuses for her long absences and pregnancy when they haven’t slept together in months and his friends tell him that he is refusing to see what’s happening….’). That she has a reason for not wanting to see what is evident doesn’t make her less irrational. That she is telling herself some story ro help her in this refusal to recognize the facts (rationalizing) doesn’t make her less irrational. If you like, we can say that her belief is rational-for-her but irrational-for-everyone-else, but this is just another way of saying that she is rationalizing.

    DS

    1. Given what we know about her circumstances, we want to say that she is reasoning poorly and that she ought not to reason in such a way given her cognitive abilities and her evidence. (If her cognitive abilities were impaired – see above re brain tumor – we would excuse her.) Now, you say she is reasoning quite well given her abilities and the evidence-as-she-understands-it. And I want to say that she should understand the evidence more or less as I do, that this failure to see things for what they are is the root of her irrationality. And that this is the case even if she steadfastly denies that she is seeing things wrongly.

      I need some arguments here, I know, but I want to pause for a second because I think this highlights two important differences in our approaches to the matter. I think we can blame her for her mis-apprehension of the evidence and so accuse her of irrationality, while you think she is reasoning well given her apprehension of the evidence and so cannot be considered irrational. Your’s is a process-oriented or procedural conception, mine is…. something else. In addition, mine is a third-person conception of irrationality while you’re employing a first-person conception. Something to think about there, perhaps….

      Our differences are indeed a matter of ‘approach’ and little more (whether such differences will eventually cede to something substantial is the philosophically most interesting question). The ‘first person/third person’ distinction is apt and one I’ve been thinking in terms of for a while. That It is with respect to this distinction that any useful account of normativity will ultimately rest is precisely what I am arguing. Thinking (in all its forms, from imaginative play to rigorous logical analysis) is a process, and on its own terms can’t intelligibly understood to proceed rightly or wrongly; it merely proceeds in a manner relative to the circumstances of its operation. In this ratiocination differs in no way from respiration. The difference is that the former is ‘intentional.’ Thinking is always about something and often is about thinking but by necessity other thinking. And nothing precludes the system engaged in the thinking process from finding (i.e., forming the judgment, the ostensible function of the thinking process) the thought it is thinking about wanting. Here is normativity, as real and robust as we are ever likely to find it: it governs and regulates thought but only by virtue of its being from thought.

      This is why I think normaivity is ‘cheap’: it’s ubiquitous, inescapable, and endlessly iterative. Of course we often want to rest content with a judgment made, and perhaps the more so when it concurs with those made by others. But nothing requires this (either of ourselves or others) nor can we appeal to any authority other than ourselves in support of the verdict that our judgment is correct.

  6. Thinking is a process, sure, as is rationalizing. But it does’t follow from this that rationality is a process, or that is is process-oriented. I want to say that someone is irrational just in case she fails to evaluate and act on the available evidence in the right way. And I want to say that what counts as the right way is socially determined in some fairly deep fashion (it’s not just about doing as the Romans do when it Rome, but about following accepted epistemic norms). I have in mind here something like Rorty’s ethnocentrism. Go back to the broken coffee-pot example– I say it is broken and you cheeily place some papers under it and remark ‘it’s not broken at all; it makes a great paperweight!’ I get to say that you’re wrong– it may make a great paperweight but it is still broken because it’s not performing its proper function. And if you question me on why we should assume that making coffee is its proper function, well- I get to punch you in the nose. Because it is a *coffee maker* and there are certain things it is expected to do, about which there is a tremendously high degree of intersubjective agreement, etc. I want to say the same thing, more or less, about the deluded student.

    1. Pretty much agreed. The ‘right way’ to form intentions and fix beliefs, in light of the available information, is certainly ‘socially determined’ – what’s the alternative? Societal norms, epistemic, moral, or other, are the product of the interplay between individuals and the groups of individuals to which each is required, by the prevailing circumstances, to respond. And most of us would say the student is being irrational, which really doesn’t amount to much more than claiming that if we were her we’d arrive at a different belief (i.e., I’m going to fail’; ‘It would be best if I drop the course’). Of course, if we were her, we’d arrive at the very belief she did and would likewise be put off by our student-hating professors who believe otherwise.

      As for paperweights and proper functions, these, too, are much a matter of what you can get away with and what is allowed. Whether it is a paper weight, a coffee maker, a urinal or art is, as you say (and as I said in the original post), a matter of expectations. Such expectations are often entrenched and widely shared, but they are also endlessly open to challenge and occasionally change.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s