The Intelligence of the Passions – A Meditation on a Passage from Hume

 

                                                             David Hume (1711-1776)

“'Tis not contrary to reason to prefer the destruction of the whole world to the scratching of my finger.  'Tis not contrary to reason for me to choose my total ruin, to prevent the least uneasiness of an Indian or person wholly unknown to me.  'Tis as little contrary to reason to prefer even my own acknowledged lesser good to my greater.”

-David Hume

 

“Reason is and ought only to be the slave of the passions, and can never pretend to any other office than to serve and obey them.”

-David Hume

 

People often have reasons for their morally relevant actions – not only their actions, but attitudes, beliefs, and even sense of self.  Their reasons may be more or less well thought out and authentically expressive of the person; they may be admirable and deep, or selfish and superficial.  In any case, when such reasons appeal to our sympathies and values, we will find them convincing, which in this respect makes them logically similar, for example, to our tastes in music and literature.

At the same time, other people's actions and the reasons why they acted as they did may – and often do – conflict with other people's values.  From Plato on, this common occurrence has been deeply disturbing to western philosophers, who have thus looked for a reason at a level which every human being must acknowledge independently of their sympathies and values. This philosophical viewpoint directs attention to the search for this reason-beyond-value which is to judge all values, and despite many attempts to formulate it (appealing to Nature, God, or Reason) no agreement has ever been reached.  This fact alone should make one wonder whether the whole project is not somehow flawed at the root.   This indeed was David Hume's position, who thought it guilty of a logical-semantic confusion.  Hume's point was not to call morality into question – the specter of moral anarchy (“anything goes”) is itself a bogey which arises only within the confused perspective he cogently criticizes.  His point is rather that the emotional response (“the passions”) is the only proper and possible way to see moral reality, to which reason is blind. And this insight can be best appreciated by examining it in the light of the critique of Hume's great philosophical opponent, Immanuel Kant.

               Hume's discovery of the “naturalistic fallacy” hit on something which goes very deep in morality. That no rationally binding prescriptive statement is logically entailed by any possible set of descriptive statements points to the seat of morality in the human heart.  It is not that knowledge plays no role in morality – to the contrary, ignorance about the facts of one's life or one's moral situation can be fatal. But even the most accurate description of a situation will not logically imply an action as the conclusion of a practical syllogism; nor will it imply any particular change of direction in one's life. For example, knowing that my friend's wife is ill and that she can no longer do certain things for herself does not logically imply he should tend to her in the way 2 + 3 = 5.  To call his tending to her a “rational response” to her situation strains the language. He tends to her because he loves her in a very particularizing way, because she needs tending to, because he is able to tend to her. These reasons are only intelligible within his overall response to my wife, which is essentially emotional – all the “facts” he knows about her are only meaningful in the context of a very specific relationship of love. There is no ground or foundation “behind” this to which a philosopher could appeal to make his tending seem like a rational response to an underlying reality which can be identified and known independently of his emotional response, as is possible, for example, in the sciences: my friend may want the global warming hypothesis to be wrong because he is afraid or because he doesn't want to change his consumption habits, but the truth of the matter is quite independent of his wants. Not so in moral matters.

              Hume's appeal to custom and habit as a guide to life does not undermine morality as Kant and other critics have feared, arguing that morality requires the positing of a free subject, and a free subject requires the existence of a moral ground beyond the passions. It is true: behaviorism – the proposition that we act and respond as we do, that we are who we are because we are so conditioned – is antithetical to morality. Some have argued that Hume's appeal to custom and habit imply behaviorism. On this view, my friend tends his wife because he lives in a culture that teaches him to do so; he loves his wife because he has been taught that “love-behavior” is the norm, etc.  Something like behaviorism is one implication that Kant feared from Hume's thought, but this fear was based on a metaphysical assumptions about nature which Hume was right not to share.

            According to Kant, the character of morality requires that my friend be thought of as a free subject, and a free subject must be thought of as being able to appeal over his feeling, emotions (Kant: “inclinations”) to some independently identifiable, rational principle of action.  Morality can't be thought of as merely a result of whatever emotions happen to rule in a person.  Even more, given that human psychology – and thus emotion – is part of nature, and nature is causally determined, morality must be thought of as seated in a dimension beyond nature, in so far as it requires a free subject acting according to a principle which transcends a given emotional-psychological state.  The precondition for the possibility of morality thus implies a transcendental-metaphysical picture of ultimate reality in which morality is seated in a rational subject inhabiting a human body-in-the-world, the former being free to intuit and follow purely rational principles, the latter exposed to the messy causality of phenomenal (biological, psychological, sociological, etc.) nature.  The assumption is, of course, that morality is conceptually dependent on responsibility, and responsibility is conceptually tied to freedom – a tornado that destroys a house is not morally responsible.

            Hume is more skeptical about reason's ability to make pronouncements about transcendence than Kant, though of course Kant shares Hume's skepticism concerning traditional metaphysical questions.  But for Kant, two features of morality are evident which Hume's thought seemingly cannot account for: the unconditional nature of moral response, and its sometimes antagonist relation to our wishes and desires. We need to look at these two features and ask ourselves whether Hume's or Kant's account make more sense out of whatever might be true in them.

            The first thing about Hume which disturbed Kant concerned the relation of the emotions to moral action. Kant pressed the perfectly true point that since the right thing to do is often not what we are emotionally predisposed to do, the former cannot be derived from the latter.  Perhaps my friend after hours of tending to his ill wife felt like taking a break, doing something for himself like reading a book, but the situation of his wife didn't allow it; contrary to his “inclinations,” he stayed and tended to her.  But because the right thing to do cannot be reduced to our inclinations, it does not follow that the right thing to do (or the best way to live) is conceptually independent from our conscious, value-soaked, emotion-laden (“empirical”) inner life in the way the sun's position to the earth is independent from our perception.  A deep confusion is at work here. My friend's care for his wife is unintelligible apart from his love for her; because of his love, he cares for her, wishes her well, desires to help if he can; his love places any inclination he might have to read a book.  Thus not selfish emotions verses duty as determined by independent reason is the correct picture in this case, but selfish emotions verses benevolent emotions – which corresponds to Hume's account.  Why are the loving, benevolent emotions unconditionally preferable?  Why ought my friend not read a book when is wife needs him?  Hume might say, 'well, he just ought to.  It would be out of place for me to tell you a metaphysical or religious story to explain why he ought to.  That would be to distort what morality is, as Kant should have known had he been true to his insight that morality is unconditional.'

            Of course, sometimes we must simply do our duty without having the slightest emotional resources to motivate us: trivial duties like paying bills come to mind, but also perhaps for some people giving first aid in an emergency to a person he has no emotional ties with (e. g. for many, a homeless person).  But such cases are the exception, not the rule. And to make such cases the norm seriously distorts morality.  Take the lovely example of Christopher Cordner of two fathers who both read to their respective daughters in the evening.  The Kantian father has no inclination whatsoever to read to his daughter – say, he'd much rather be reading for himself The Critique of Practical Reason.  But out of a pure sense of duty he overcomes his true desires and reads to his daughter a fairy tale which bores him to tears – a purely, perfectly moral act for Kant because performed according to a rational assessment of what duty required and not according to a contingent emotional response.  The Humean father reads to his daughter because he has pleasure from it; he has pleasure from it because he loves his daughter; has joy over her particular life; relishes her reactions to the stories and observing her development; and a host of other intangibles. Now according to Kant, all this emotion is irrelevant for the moral worth of the Humean father's reading – in a way; he is selfish for doing merely what pleases him, whereas the Kantian father is altruistic.  I think: a father who reads only from a sense of duty is the morally deficient one, and the quality of his reading will be very much worse than the reading of the father who is not only doing his duty.  A Kantian father would be to that degree a deficient father.  And the claim that a bad father is a moral father whereas a good father is amoral is absurd.  Indeed, nothing better shows the absurdity of detaching emotional response from morality than this example.

            How we feel cannot be said a priori to be – as Kant thought – the result of some biological hardware plus culturally determined psychological software.  How we feel, how we respond emotionally, is a function of who we are morally, and what lies beyond this is beyond knowledge, period.  And who we are morally is partly a function of the quality of our loves, of our attitudes towards others. Of course, our emotional lives are intimately connected with reason in many ways; how artificial to separate them.  Of course, the picture of the absolutely free subject existing in transcendental space cannot be reconciled to the self of emotionally laden moral response, which does contain an element of contingency (though subject to a degree of choice as well).  But “purity” from the contingency of emotional response does not ground morality, unless the responses of automata can be considered moral.  Kant wanted to make morality a simple operation of reaching the “correct answer” (duty) which anyone with minimal intelligence could perform, giving morality an objective, even logical character; Hume wanted to describe our moral reality as we experience it, having drawn limits to speculative philosophy.

            Now to the second point concerning the unconditional in morality, which motivated Kant to reject any contingency and thus the relevance of moral response.  Kant, I think, is right so far: an unconditional element does inhere in moral judgments. The moral quality of my friend's tending to his wife in her illness has something unconditional about it:  it would be base of him to leave in her in dire need while he went out to a bordello or to drink with the boys.  I am aware the details would need to be filled in to make the baseness fully evident, but I have no doubt that under particular descriptions such an act would be simply base.   (Some thinkers believe Hume confused matters here with his distinction between “facts and values.”  Hume never claimed that factual descriptions were value-free; he did claim that reason won't give us an independent ground why we shouldn't do  for example base things.)  To say the moral ought is unconditional means that non-tautological reasons cannot be given why an action or attitude is base: to say, for example, that if my friend neglected his wife in a moment of suffering to go out drinking with the boys, his conduct is base because his wife is in pain, because the consequence will be that she might die, because he is neglecting his obligation towards her, or because he is violating his love – these because-clauses don't really give a reason why the behavior is base so much as to simply spell out what the baseness means in this particular case.   The moral ought often implies unconditional value judgments; that doesn't imply that people may not ignore them, in which case it makes more sense to think of them as base, for example,  rather than as irrational.

            Or the point may be illustrated by Wittgenstein's distinction between absolute and relative value judgments in his “Lecture on Ethics.”  As an example of a relative value judgment, Wittgenstein gives a tennis player who is playing badly.  Upon being told he is playing badly he responds that he simply doesn't feel like playing well today – perhaps he is preoccupied with problems at work or his lover has just broken up with him.  But if this person is behaving beastly to someone, and upon being told that he is being beastly he responds that he just doesn't feel like being decent to others today, then according to Wittgenstein (and I think he is right) the matter would be fundamentally different from the tennis case: we should at the very least want to say that he ought to behave decently, that the reasons he might give for not wanting to be decent – though perhaps understandable – would not excuse him from this ought. 

            And this is a very Humean point: judgments implying the moral ought are prescriptive, which is to say express something unconditional, sui generis; that is, they are very different from the sorts of propositions which may be independently determined to be true or false, this set of propositions being identified with the range of “reason” for Hume and Kant alike; they entail a very different relation to reasons than relative value judgments, as the tennis example makes clear.  Thus it is not the case that Hume is unaware of the apparent fact of the unconditional in morality.

            Kant's problem with Hume on this score has its source in Kant's metaphysical understanding of science and the universe – it is a metaphysical, not a moral disagreement.  That Hume makes moral response dependent on emotions like sympathy and benevolence disturbs Kant because of his preconceived notion that nature – including human nature – is subject to a mechanical causality; in other words, that nature is nothing but that which is explainable by Newtonian science.  Emotional response is part of nature, and so subject to the mechanical law of cause and effect; therefore, emotional response is radically unfree, eliminating its possible connection with morality (which of course presupposes a free subject).  It follows that emotional response (which Kant reductively calls “inclination”) cannot be the ground of morality, the implication being that the emotional aspect of my friend's attending to his ill wife is no more morally relevant than the tornado's destruction of the house – both being natural events subject to causality.  In other words, within the parameters of Kant's metaphysics, Hume can only be thought of as a behaviorist.

            But Hume would rightly not only find Kant's metaphysical picture ultimately confused; in fact, he positively refuses to tie morality to any possible metaphysics – and he was surely right not to do so.  Kant's belief that nature was just what Newtonian science shows it to be is, ironically, not part of science, since statements about reality as such (or the true nature of Being) can never in principle be a possible scientific hypothesis subject to public verification or falsification (Popper).   Without being aware of it, Kant accepted a metaphysically reductive picture of reality. Our experience of morality is confused with our everyday perceptions like “the sun rises,” which science had shown to be wrong; the common sense belief that the emotional response of my friend to his wife is a very important part of moral reality as well as the everyday language of morality in which this common sense belief is expressed give way to “science.” Kant might say:  “now we know such ways of thinking and speaking are naive, since science shows us that all phenomenological reality is subject to the iron law of causality; even though we still go on saying 'the sun rises' and 'my friend's care is part of morality,' we know that the latter is really nothing but the kind of thing that behaviorist psychology (or some other reductive science) describes more accurately.”  This is the threat Kant felt he had to respond to; this is why he had to locate the seat of morality in a transcendental self beyond the world with access to a noumenal reality again beyond the world.  Hume's wonderful skepticism about knowledge of “the ultimate nature of reality” (most famously expressed in his thoughts on causality in the Treatise) meant that he felt no need to metaphysically justify the reality of moral response; he didn't see matters from the “view from nowhere” as did Kant.

             Hume's compelling critique of causality was ironically more Kantian than Kant; he showed merely the limits of reason, remaining metaphysically agnostic.  What does the unconditional nature of morality say about the ultimate nature of reality?  Well, we can't know, because such matters can never be an object of experience.  Reason can't uncover a ground of morality in the universe if there is one. Kant's transcendental philosophy is the wildest speculation.  Nothing about my friend's response to his wife depends one way or the other on such speculation, as moral certainty does not depend on an impossible (and conceptually unintelligible) metaphysical certainty. Moreover, transcendental philosophy seriously distorts the character of my friend's response. Hume's is a philosophy of common sense – one almost wants to say, of reason.  In the latter expression of Wittgenstein, Hume, too, leaves things as they are, and attempts merely to describe them as well as he can.

            Kant was bothered by Hume's agnosticism towards foundations, towards philosophy. Hume seemed to undermine the two things Kant cares most about: science and morality. But Hume didn't undermine anything. The practice of science goes merrily on despite his critique of causality just as morality is still in force despite its lack of metaphysical foundations.  Kant couldn't stop with his fundamental insights:  that morality has an unconditional element and that emotional responses were not morally self-authenticating.  He felt the need of traditional metaphysics to give morality a foundation in Being which is an object of human knowledge. He recognized traditional metaphysics failed because it did not distinguish between scientific and logical kinds of knowing on the one hand, and metaphysical kinds of speculation on the other. He believed his transcendental philosophy escaped his critique of traditional metaphysics in that it did not make positive statements about that which transcends experience, but rather directed inquiry at the metaphysical conditions for the possibility of that which we do experience – especially morality.  A brilliant move. However, although the route is different, more indirect, the result is the same: a metaphysical principle of morality, outside of the world, independent of experience, according to which worldly moral experience must be measured.

            Hume was deeper in his mistrust of reason to come up with an independently verifiable formulation, which was to serve as a procrustean bed on which to make moral reality in the world conform.   As soon as morality is given an abstract foundation, it is distorted.  Its reality involves a freedom closely tied to contingency and the lack of foundations – which leaves the unconditional force of the moral ought alone.  As Simon Blackburn characterized Hume's moral thought: for Hume, there are only reasons in moral reflection – no Reasons. The move to philosophical relativism is an equally misguided way of understanding morality in terms of an abstract schema independent of experience.  I think Hume would have only shaken his head at most of moral philosophy since Kant, including the behaviorist and relativistic directions into which his own thought was pressed.   Hume's thought undermines not science or morality: it undermines philosophy – a fact still largely unrecognized or ignored, even by modern admirers of Hume.