How a critical Humean naturalism is possible: Contesting the Neo-Aristotelian reading

McDowell’s anti-Humeanism

Ethical naturalists or neo-Aristotelian virtue theorists have always found fault with Hume’s practical philosophy. Some have called this philosophy subjectivist in a disparaging sense, others have criticized his theory of moral motivation for being psychologistic, yet others have called his variant of naturalism shallow or narrowly empiricistic. ¹ Naturally, these criticisms combine and feed into each other. Still, in the following, I want to concentrate on the last point, namely Hume’s seemingly narrow variant of naturalism. I think it is foundational for some of the other criticisms and therefore deserves pride of place. As John McDowell has taken an influential lead in blaming Hume for suggesting an ill-conceived naturalism, I will begin by briefly reconstructing his approach to Hume. My general claim is that McDowell misses important stands of Hume’s naturalism and overlooks that elements of his own account of second nature are compatible, in certain respects, with a well-understood Humean naturalism.

McDowell’s essay ‘Two Sorts of Naturalism’ aims at developing a concept of nature that might serve as basis for an ethical project shared with philosophers such as Philippa Foot and aptly labelled ‘ethical’ naturalism. ² What would be the best understanding of nature in an attempt to develop a naturalistic ethical vocabulary? Apart from many similarities of approach, the problem McDowell sees in Foot’s ethical naturalism as well as in many other ethical accounts is the lingering presence of a scientistically reduced concept of nature that sees nature as wholly disenchanted and emptied of meaning, an ‘ineffable lump, devoid of structure or order’. ³ McDowell is convinced that this picture of nature is wrong and that Hume and his successors should not have entertained it.

One rather simple reason why it cannot be correct is the fact that the empiricist school itself insisted on conceptualizing disenchanted nature as a realm of law (not devoid of order at all) to be studied successfully by modern natural science. In fact, this sort of naturalism becomes the measure of what is real and objectively true and thus supplies the credentials for claims raised by theoretical and practical reason. As McDowell says, our broad acceptance of a scientistic understanding of nature shows in our acceptance that ‘a putative operation of the intellect can stand up to reflective scrutiny only if its products can be validated on the basis of the facts of nature, conceived in the disenchanted way that is encouraged by modern science’. ⁴ Disenchanted nature thus acquires foundational status not only for theoretical reason but also for practical reason; ethical recommendations or moral imperatives only deserve to be studied and taken seriously if they spring from natural sources or accord with natural facts about human beings. Otherwise they simply will not be granted truth-status or be taken as supplying us with genuine reasons for assent. The problem McDowell has with some variants of ethical naturalism is that they seem tempted to find their ethical recommendations on facts of nature that can be independently studied and validated, that is, independently of (already) taking for granted the validity of specific ethical or moral norms.

Hume, for McDowell, is ‘the prophet par excellence’ of the tendency to adopt the scientific outlook as measure of truth, reality and objectivity. ⁵ For McDowell, Hume’s acceptance of the scientistic outlook shows in at least two important respects. First, whatever meaning or intelligible order is accepted by Hume derives from the operations of the mind; nature as such is not meaningful. Second, while Hume does not ‘officially recognize a practical employment of reason’, he is still able to formulate correctness conditions for practical thinking: this thinking is correct, it appears, if we can ground it ‘in individual wants and likings, conceived as brute unassessable facts’. ⁶ Hume’s notion of practical rationality is thus subjectivistic, its ability to formulate valid moral judgements seems to rest on facts about subjects, on something, in other words, that must be true about the judging or judged subject. Neo-Humean positions, for that matter, do work with a more substantive and less subjectivistic notion of practical rationality in suggesting that there are facts about (disenchanted) human nature that give guidance as to what beings of such a nature truly (and not just subjectively) need. Both approaches, however, rest on the desire to ground morality in some facts about nature (subjective desires or objective needs), and that is what makes them exemplars of a still dominant scientistic naturalism. As long as the grounding capacity of these facts is to be understood independently of established ethical practices, they seem to be the facts of a law-governed and meaningless nature.

As is well known, McDowell thinks that another concept of nature is available, namely the concept of a second nature. Much discussion has gone into this concept but for lack of space I cannot go into detail. Suffice it to mention the central idea: subjects acquire a second nature through being educated into a culturally bound practical logos. They thus learn to see the world in particular ways and acquire ethically relevant reasons for acting. Basing second nature on the idea of a practical logos is important for it appears that not just any ethical education will adequately open the subject to the situationally relevant ethical reasons but just those educational processes that allow critical scrutiny of the reasons acquired. As McDowell says, in ‘imparting logos, moral education enables one to step back from any motivational impulse one finds oneself subject to, and question its rational credentials’. ⁷ A logos precluding this ability to step back from whatever impulses are pressing, would not, I take it, deserve to be called a second natural logos, for second nature, in McDowell, does not stand for an acquired unreflective habit or custom. This much Aristotelian teleology, then, is present in McDowell: it is natural for rational animals like us to develop critical capacities that allow us, if a second nature ‘of the relevant kind’ has been acquired, to evaluate the very practices that have produced these capacities in the first place. ⁸ Having acquired a second nature of this kind thus loosens the impact of either unreflective or simply first natural impulses. Even if we cannot fully ignore first nature (we cannot, e.g. ignore the need to nourish ourselves), we can, having adopted the right kind of second nature, try to reconfigure the established ways of satisfying first natural impulses. We can ‘break out at any time’ ⁹ ; thus, even if it is true in general that wolves hunt in packs, it does not follow for each individual (supposedly rational) wolf that he has to follow the natural rule: ‘Having acquired reason, he can contemplate alternatives’. ¹⁰

Historical naturalism

How convincing is this reading of Hume’s naturalism? In what follows, I claim that McDowell’s picture overlooks central elements of Hume’s approach and that parts of McDowell’s account of second nature are fully compatible with Hume’s variant of naturalism as I reconstruct it. To be sure, Hume’s philosophy has frequently been called naturalistic and no consensus exists as to what naturalism in Hume might actually mean. ¹¹ However, most of the interpretations shy away from rubricating Hume as a blunt empiricistic or scientistic naturalist. For want of a better term, I call my variant of Hume’s naturalism historical naturalism. Nietzsche, in his ‘On the Uses and Disadvantages of History for Life’, has famously announced that every ‘first nature was once a second nature and…every victorious second nature will become a first’, and this nicely captures the dynamic aspect of my reading of Hume’s naturalism. ¹² Historical naturalism appears close to what influential interpreters such as David Wiggins have called Hume’s vindicatory genealogy of morals. ¹³ A closer look would reveal, however, that the genealogy of morals as envisaged by Wiggins is meant to strengthen or vindicate social virtues such as benevolence in treating them as deeply ingrained (first) natural impulses to act in opposition to one’s similarly natural self-oriented motives. ¹⁴ The genealogical method of analysing the evolvement of principles for judging of character traits such as benevolence thus serves to prove the natural basis of morality and not to historicize it in a Nietzschean sense. Consequently, even the more artificial (second natural) virtues such as the virtue of justice are merely seen as redirecting or extending ethical impulses ‘already latent within our human nature’ and thereby sustain Wiggins’ account of naturalizing the basis of our typical judgements of moral right and moral wrong. ¹⁵ Historical naturalism, in contrast, sees even propositions based on an account of first nature as open to historical analysis and do not necessarily aim at vindicting moral precepts or judgements through proving their natural ordinariness or inevitability.

It will be helpful to take up Nietzsches’s use of the distinction between first and second nature in order to better grasp the notion of historical naturalism and apply it to Hume. What would a Humean version of first nature look like? The question is tricky and leads right into heart of my discussion. Deleuze suggests that one of the central questions for Hume is this: ‘How does the mind become human nature?’ ¹⁶ What this question implies is that there is no fixed object, call it human nature, that could be studied without knowledge of the forces and factors that fixed or objectified it. For Deleuze this means that ‘one must be a moralist, sociologist, or historian before being a psychologist’ – a claim that captures part of the spirit of what I call historical naturalism. ¹⁷ At the same time, it should be clear that there are natural conditions for any kind of a developed second nature. Even McDowell is explicit about this: ‘Of course first nature matters. It matters…because the innate endowment of human beings must put limits on the shapings of second nature that are possible for them’. ¹⁸ Others speak of the ‘natural, enabling capacities’ that Bildung or culture presuppose. ¹⁹ Hume, to be sure, time and again proffers first nature statements. Take the beginning of the chapter on justice:

Of all the animals, with which the globe is peopled, there is none, towards whom nature seems…to have exercis’d more cruelty than towards man, in the numberless wants and necessities, with which she has loaded him, and in the slender means, which she affords to the relieving these necessities. ²⁰

The necessities Hume mentions in this quote must be first natural necessities such as our need to be fed or to require (in most climates) housing. Even granted that how we feed ourselves is highly contingent on environmental, cultural and historical factors, it cannot be denied that we must feed ourselves. That is just a first natural necessity, and I don’t see any reason why we shouldn’t call it a natural law and even a natural fact.

But is it, in McDowell’s terms, one of the facts of nature or one of the grounds that determine the correctness of practical thinking? What should these natural facts generate but very unspecific natural imperatives like ‘feed yourself’ or ‘build yourself a house’? Hume, in continuation of his statement about the infirmity of the ‘savage and solitary condition’ of ‘man’ says this: ‘Tis by society alone he is able to supply his defects, and raise himself up to an equality with his fellow-creatures…By society all his infirmities are compensated’. ²¹ One might wonder just what is meant by ‘society’ in these statements. After all, the move from the rather primitive necessities of nature to society is impressive. Would a group of five savages bonding together to pool forces constitute a society? And is it really necessary to create a society to better cope with our natural infirmities? Necessary in the same sense as it is clearly necessary to feed ourselves? Hume’s answer to this last question is, I think, ‘No’. Yes, we do create society to cope with our natural infirmities, but this creation cannot be natural as opposed to artificial. ²² Society is artificial, it is an invention and justice is the virtue that should structure this invention if it is to succeed and fulfil its functions. Society, then, is natural in being one of the regular means of coping with first natural deficiencies, but this regularity is an observed regularity, it stems from having experienced the advantages of living in society. Thus, when Hume calls the artificial rules of justice ‘laws of nature’, he opposes what he takes to be ‘natural’ to what is supernatural or unusual and not to what is artificial. ²³ The naturalness of observed regularities is not first natural in McDowell’s sense, and as we will see most of the facts Hume draws on in proposing norms for practical rationality (if he does so at all) are of just this kind. They are, as we might say, second natural facts. Obviously, this is a significant point that should lead us to drop McDowell’s interpretation of Hume’s naturalism as a shallow variant of a thoroughly empiricistic naturalism.

How far second nature extends into first nature in Hume can be seen if we take a look at another seemingly first natural fact discussed by Hume. Hume claims that the ‘natural appetite betwixt the sexes’ constitutes the ‘first and original principle of human society’. Further, he speaks of a ‘natural affection’ that parents bear to their children. ²⁴ For Hume, these are important points, as he thinks that the willingness to enter or form society will not be instigated in humans ‘by study and reflexion’ alone. ²⁵ In primitive conditions, strong natural drives for company support whatever arguments one conjures up to make the option to enter society seem attractive. As I said, these drives, ‘appetites’ or ‘affections’ seem to be first natural facts about human beings. They are not invented and cannot be treated as mere conventions. But are they really first natural? Philippa Foot, in her Natural Goodness, discusses the case of reproduction and suggests that lack of capacity to reproduce is a defect in a human being. However, choice of childlessness and even celibacy is not thereby shown to be a defective choice, because what is good for humans is not the same as what is good for plants or animals. ²⁶ Thus, it is a defect of first nature if reproduction is biologically impossible; it is not, however, a defect of first nature if reproduction is possible but not chosen. If this is a defect at all, it is a defect only in cultural contexts in which reproduction is expected or taken to be the norm. Consequently, natural here does not mean determined or predetermined. It means something like ‘typical’, ‘regular’ or ‘normal’, that is, frequently to be observed. Seeming regularities, then, are not regularities because they are natural; they are natural because of their experienced regularity. ²⁷

Of course, from a different perspective, one could say that humans must reproduce if they want their species to survive. And this ‘must’ seems to be a first natural must. As McDowell has convincingly shown, however, for rational creatures, the validity of general statements concerning species-wide needs cannot be transferred to its individual members (1998, 172). Rationality just means being able to question seemingly natural imperatives and absolve oneself from blindly following them. Thus, while it may be the natural rule that wolves hunt in packs, nothing of normative relevance follows from this general fact about wolves for individual wolves. If they prefer to hunt alone, their behaviour should not be seen as irrational or unnatural even if it remains true that if too many wolves follow their example, species-survival may be at risk. In rational creatures, general facts about the species cannot determine forms of behaviour of individual members of the species or implement species-wide rationality standards. The general fact, then, cannot serve as justificatory ground for normative statements concerning the question ‘What should I do?’

What all this means is that Hume’s ‘facts’ are much more complicated than McDowell suggests. Calling them or their generation scientistic or empiricistic does not do justice to them and misses that most of the practically relevant facts Hume has in mind are second natural facts, even many of those he calls (first) natural. Of course, calling them second natural does not answer all questions, and it will now be necessary to be more precise. What do I mean when I call Hume’s facts second natural? So far, it was said that the ‘natural’ of second nature indicates observed regularity. Further, the regularity we observe is not of the deterministic kind observed in natural phenomena. People do not enter society because they have to, they enter it because it is difficult not to see the interest-related advantages of living in society. The regularities we observe are not law-like and their obtrusive character suggests genealogical reasons ‘primitive’ people might have for entering society. Natural, then, appears to mean no more than usual or normal. For humans, it is normal (natural) to enter society and to live according to its rules. Living outside society, however, does not rob one of all typically human characteristics, it just means that one does not follow the rules mostly or most commonly observed.

But is this really all one can say about Hume’s variant of naturalism? A thin layer of very general first natural imperatives is supplemented by a much thicker layer of artificially established rules, norms and institutions that generate morally supported expectations of normality? Haven’t I just replaced ‘natural’ by ‘normal’ without much loss in meaning? In other words, when Hume says that ‘a man naturally loves his children better than his nephews, his nephews better than his cousins, his cousins better than strangers’, is he just saying that this is normal? ²⁸ Not first naturally true (there may be men and women who do not love their children or who prefer their cousins to their nephews) but typically true or simply true according to observation?

I don’t think that this is all one can say. Let me take a short detour to explain why. In discussions of Hume’s epistemology and his theory of causality, it is sometimes suggested that ‘nature’ supplies the firmness of belief that could not be generated on the basis of the given evidence alone. In other words, we constantly draw causal inferences without having the necessary evidence to do so. Here is David Pears’ take on this: ‘The belief in necessity which is manifested in a readiness to make audacious causal inferences is pre-theoretical. It is an unanalysed belief, implanted in our minds by nature.…Where reason fails us, nature supports us’. ²⁹ Pears’ use of the concept of nature is difficult to evaluate. What exactly is nature supposed to mean in this context? Why not simply make use of Hume’s notion of habit and custom to fill the gap between evidence and belief? Our readiness to make causal inferences beyond the given evidence is not implanted in our minds by nature; rather, it stems from repeatedly experiencing similar pairs of events in immediate contiguity to one another. Of course, one can always claim, as Pears does, that this impact of habit is itself implanted in us by nature. Commenting on how Hume generates his theoretical concepts by closely observing his own mind, Pears says that ‘nature has so formed his [Hume’s] mind that, confronted by the first member of such a pair, he cannot avoid believing that the second member will follow immediately’. ³⁰ In other words, whatever experience of necessity we have, it is not evidence-based but based on, well, nature. We do not really know why we think that event b will follow event a but given the right circumstances we certainly will believe that event b follows event a, in fact we cannot help but do so. Nature, in this sense, serves as guarantor of a felt necessity we lack epistemic evidence for. The problem with this reading is twofold. First, the genuine force of habit and custom to shape our minds is downplayed if one simply naturalizes their effects. Hume, it is well known, has often exemplified his notion of necessity by citing social relations: ‘A man, who gives orders for his dinner, doubts not of the obedience of his servants’. ³¹ Stipulating that nature has so formed our minds that one giving orders cannot help but believe that the one ordered will obey is, obviously, not very satisfying in explanatory terms as more factors are involved in order to adequately grasp the give and take of orders. Emphasizing the force of habit, on the other hand, allows one to study the various factors involved and thus opens up the analytical frame in a way that the category of nature does not. In fact, the Wittgensteinian naturalism of Pears implements the category of nature to end further analysis and risks to dehistoricize and desocialize the various causal factors responsible for the production of feelings of necessity. This, then, is the second problem of Pears’ reading of Hume (closely linked to the first). In substantializing nature as he does (‘nature’ as force of implantation), Pears runs the risk of impoverishing its description in exactly the way McDowell criticizes in Hume. The felt effects of nature are, it seems, lawlike even if we don’t fully understand how they produce the effects they produce. A large part of Hume’s naturalism, however, revolves around expectations of normality that have a more or less clear normative frame.

Let me explicate this in going back to Hume. In stating that ‘a man naturally loves his children better than his nephews’, Hume does not appear to lack any evidence to raise the claim. The evidence, it appears, is there, to be seen by an unbiased willingness to acquire knowledge through observation. As such, the evidence seems to be merely descriptive, we observe the world and learn that most men and women love their children better than their nephews. ³² Our observation is based on a concept of love, of course, that guides us in recognizing what kinds of being can feel attached to what other kinds of being, but Hume has a concept of love, in fact, a large segment of the second part of the Treatise is dedicated to discussing just this concept. I take it that ‘natural’, in this context, just means usual or normal. At the same time, however, it seems to introduce a further dimension, namely a pro-attitude toward parents who love their children. This dimension may be difficult to detect, but I suggest that Hume prefers to call the love that ‘men’ feel for their children natural and not normal or usual because it is taken to be the accepted way of treating children. Further, it is accepted because it is natural in not just being normal or usual but in being the way things should be. Baier says, commenting on Hume, that ‘we approve of normality’, and it is this approval-dimension that appears to be present in Hume’s use of the word ‘natural’. ³³ However, we do not first observe normality and then infer that what is normal is good; rather, we define normality through the lens of what we take to be good. To be sure, this goes against the grain of Hume’s own interpretation. Hume famously suggests that ‘our sense of duty always follows the common and natural course of our passions’. ³⁴ He thus suggests that we move from premises about what is normal in humans to conclusions about what should be the case. But this picture of deriving normativity from non-normative facts is wrong as the facts at stake are already perceived through the lens of normativity. What allows us to overlook this lens is the systematic use of the concept of nature as the central force of normalization. Put differently, what is natural is more than just what is normal; what we take to be normal is guided by assumptions about what is natural. We embellish normality with naturalness to infer moral precepts from it.

If this somewhat dense interpretation of Hume is correct, Hume’s ‘is’ is shot through with ‘oughts’, a fact that makes it easy for him to derive an ‘ought’ from an ‘is’ in the first place. It is already there or, in Humean parlance, it is projected onto the ‘is’. The main instrument of projection or of overlooking projection is the concept of nature or, as we might add, of first nature. Earlier, I have defended an innocuous notion of first nature, and I do not see any reason to retract from this notion. What is at stake in the present discussion, however, is a more complex notion of first nature, namely a notion that serves (as Wiggins was well aware of) vindicatory purposes and enters as such the heart of almost all the second natural practices and institutions humans develop. Whatever facts have normative weight – they derive this weight from our tendency to treat them as if they were first natural facts. This, then, is the reason why it makes sense to call second natural facts second natural. Transforming their naturalness into normality would overlook to what extent we make use of the concept of the (first) natural to construct our concept of normality. ³⁵ What habits and customs mostly naturalize for us are frequent conjunctions of events that are seen as necessarily associated (like giving orders and receiving obedience) in response to what is taken to be natural. Calling these associations second natural requires to closely analyse the legitimation and power structures that produce the force we feel to move from one event to another, apparently closely related event.

The question of critique in Hume

McDowell, it was stated, introduces a critical component into his version of second nature. Hume, if my reading is plausible, could do so as well even though one might not find much textual evidence for a critique of, say, false virtues. But some evidence there is. An obvious example would be Hume’s critique of the so-called monkish virtues. In the Enquiry Concerning the Principles of Morals, Hume castigates seeming Christian virtues that he himself, in his final judgement, then reconceptualizes as vices:

And as every quality which is useful or agreeable to ourselves or others is, in common life, allowed to be a part of personal merit; so no other will ever be received, where men judge of things by their natural, unprejudiced reason, without the delusive glosses of superstition and false religion. Celibacy, fasting, penance, mortification, self-denial, humility, silence, solitude, and the whole train of monkish virtues; for what reason are they everywhere rejected by men of sense, but because they serve to no manner of purpose; neither advance a man’s fortune in the world, nor render him a more valuable member of society; neither qualify him for the entertainment of company, nor increase his power of self-enjoyment? ³⁶

This passage leaves no doubt that Hume treats traits of character that allow for membership in society (useful to others), fortune in the world (useful to self), entertaining company (agreeable to others) or self-enjoyment (agreeable to self) as positive qualities to be sought for. It also leaves no doubt that Hume’s variant of naturalism allows for a critique of perceived ills. The realm of the natural or normal is surrounded by the realm of the unnatural or abnormal that must be kept at bay. We clearly should be beings striving for the character traits that allow us to achieve the positive good of commercial sociability. True, instead of directly proposing these traits to be inherently good, Hume bases his claims on the verdicts of ‘men of sense’ who judge these ‘monkish’ virtues to be of no avail whatsoever. But it seems clear that Hume accepts this rejection and thus treats it as fully justified. The philosopher Hume, we might say, is voicing what ‘men of sense’ everywhere judge to be the correct way of acting in a world of sociable beings. ³⁷

But does this judgement also rely on experienced regularities? Have we seen or somehow experienced that, say, monks cannot be valuable members of society? Living according to the monkish virtues does not appear to be empirically impossible, so Hume’s idea must be that even if many people decide to adopt these virtues (maybe at the price of not being able to enjoy the fruits of full-fledged society), it would be wrong to do so. But why? To answer that question, it might help to remember that the perspective he takes is that of ‘natural, unprejudiced reason’. It seems to be this reason that generates the ‘is’ from which all virtue claims are derived. For Hume, Baier claims,

all virtues, natural or artificial, get their status from the fact that these traits are ones that we in fact welcome in human persons, once we have the relevant facts, take up a special impartial point of view, exercise our capacity for sympathy, correct its natural bias, and then finally let our reflective feelings pronounce judgment. ³⁸

But what are these ‘relevant facts’? What we let our reflective feelings pronounce certainly is in no way naively natural. It should also be obvious that the facts alluded to cannot be reduced to the individual wants and likings that McDowell takes to be the ‘brute unassessable facts’ that serve as ground for Hume’s conception of practical thinking. The ground is constituted by a complex interplay of moral judgements and morally relevant feelings. It is, we might say, a thoroughly moral ground and not a realm of (natural) law. But if this is so, whatever is called natural in this context is not first nature natural, as it were. What is natural is what we (who is ‘we’?) take to be normal. Calling it natural normalizes it for us and embellishes it with normative patina.

Can this critique of the monkish virtues also be used to further specify the historicity of Hume’s naturalism? But why ‘historical’? I think that the positive traits allowing human beings to joyfully live among other human beings are not taken to manifest eternal truths or inherent anthropological teleologies. In other words, I don’t think that Hume’s rejection of the monkish virtues exposes the neo-Aristotelean rock bottom of Hume’s social philosophy, the realm of ‘prior truths’ about what it is for human beings as such to do well. Rather, Hume’s truths are still post hoc truths, that is, truths that have historically evolved and must be studied accordingly. What evolves, however, is not completely contingent. Human beings are in need of social support at all levels of their development and this is not an a priori fact but a fact to be carefully studied on an individual and collective level. This natural (i.e. normal, regular, typical) sociability manifests itself in various historical guises and it appears that the more refined and intricate it gets, the better it manifests itself. Referring to the enlightened Europe of Hume’s time, Christopher Finlay argues that ‘sociability…was now a fact of human nature, even if it was not necessarily an original one’. ³⁹ The fact, then, that living according to the monkish virtues has not sufficed to establish full-blown modern society betrays something about what we might label the sociability potential of these virtues. The monkish virtues are unnatural in not reflecting or supporting those character traits that turn humans into agreeable partners, friends, market agents or citizens. In this sense, then, the valuations and aspirations Hume exhibits in his implicit or explicit normative judgements are thoroughly special to his historical or cultural situation. Consequently, I take what Hume calls natural as opposed to miraculous, unusual or rare to be of a non-teleological historicized kind. There is no place beyond history and experience that allows us to get to know the possibilities of human development or, for that matter, human nature. Again, both the monkish virtues and the more sociable virtues seem to be historically possible; but if history has shown that the monkish virtues do not allow the creation of large-scale commercial societies, that fact in itself calls for certain conclusions concerning human nature as it presents itself in given historical and social stages.

In conclusion of this section, I hope to have shown that Hume’s naturalism is not scientistic but historically and culturally enriched. I have also shown that it does contain some normativity in suggesting that what has historically evolved has not just evolved in arbitrary ways. There is progress in human development, and the monkish virtues are not part of it. However, progress does not amount to fulfilling a pregiven or a priori set of teleological principles but presents itself in those morally relevant collective reactions and emotionally codified judgements of a given culture that cannot be given up without altogether transforming the face of just that culture. Normality thus can be paraphrased as those reactions and judgements shared within a cultural frame that seem ineluctable and carry the force of what appears natural.

Critical naturalism?

But let me carry the question of critique a step further in discussing Annette Baier’s readiness to call Hume’s naturalism critical. Let us assume a deeply sexist culture. Could such a culture reform or abandon its deeply ingrained sexism? Baier sees two interlocking instruments of reform. First, Hume’s notion of sympathy as basic carrier of morality. If sympathy, according to Hume, is what makes us enter into the sentiments of others and is somehow made possible by ‘the great resemblance among all human creatures’, ⁴⁰ it should in principle (certainly as guided by the ‘general point of view’) be able to overcome seemingly natural biases in favour of one’s own kin, sex, race or culture. Second, sympathy will only be able to do this if the norms and values generated on the basis of sympathetic reactions institutionalize or materialize in what Baier calls ‘equitable artifices’, that is, conventions, general rules and norms that help to overcome seemingly natural biases. ⁴¹ The central site of reform in Hume is thus his concept of artifice for whatever is artificial is invented or authorized by us, Hume’s prime example being the rules of justice. And whatever is made by us can be undone by us, can be transformed and changed (and not just redirected).

But how is this going to happen if our prejudices are strong and are backed up by a long history of discrimination and differential treatment? Baier follows Hume in suggesting a second natural account of the forces of reform. Remember, Hume said that ‘nature provides a remedy in the judgement and understanding for what is irregular and incommodious in the affections’. ⁴² Baier takes up the cue: ‘It is’, she writes, ‘our natural understanding, along with our natural sympathy, that enables us not just to design social artifices but eventually to prefer those which are more “equitable”’. ⁴³ But this appears to be wishful thinking (as some of the other protofeminist readings Baier produces of Hume). What kind of understanding is this ‘natural understanding’? The one thing we can say on the basis of the larger context of Hume’s quote is that he takes this natural understanding not to be an ‘uncultivated’ kind of understanding. ⁴⁴ The question then is: what forces cultivate the understanding to the extent that it is capable of recognizing the injustice of present inequities? Baier’s answer to this question is vague but she suggests that the very same reasons that make us enter into artificial agreements with others in the first place, namely our deep-seated (natural) interest in social stability, will also open our eyes to possible disruptions of social stability stemming from the victims of a patriarchal or sexist status quo. ‘Hume’, Baier writes, ‘supposed that equitable artifices would be more stable because they are less prone to provoke morally justified rebellion from within or resentful attack from without by those excluded from the “sweets” of the society they produce and monopolize’. ⁴⁵ The problem is that Hume himself, in a rather infamous passage of the Enquiry Concerning the Principles of Morals, mentions the possibility of what we might call total exclusion, that is of ‘creatures’ who are not even able to make their resentment felt. ⁴⁶ In a strict sense, these ‘creatures’ are not taken to be members of society and suffer from what we today might call recognitional invisibility.

Part of the problem here is a one-sided progressive reading of Hume’s idea of the creation of social artifices at the price of neglecting the continuing influence of prejudice even in the act of inventing seeming solutions to the ‘natural’ partiality of our affections. Baier’s reading is compromised by this problem despite the fact that she is somehow aware of it. ⁴⁷ She is simply over-optimistic with reference to rebellion from within or resentful attack from without. Given my own terminology, we might also say that she underestimates the force custom and habit exert in creating second naturally unchangeable laws, that is, laws that appear unchangeable as they are taken to be first naturally unchangeable laws. As I mentioned earlier, Hume regularly mentions the power of custom and habit to associate frequently experienced ideas or impressions and create the feeling of necessary causality. Furthermore, he frequently socializes his account of causality, that is, he chooses the instances which he gives of the operation of causality from social relationships such as parent–child or master–servant relationships. ⁴⁸ Thus, very early in the Treatise, Hume exemplifies what he means by the relation of cause and effect by extending his account to social structures of inequalities in power. In master–servant relationships, for example, the master is a cause in having the power to direct the servant, that is, in having the power to produce effects in the servant or, more radically put, to produce the servant, simply by exerting his will. ⁴⁹ Hume leaves no doubt that relations of power are constituted by the ‘inseparable connexion’ subjects imagine to hold between an active force potentially issuing orders and a more passive force accepting these orders. Moreover, this connection is seen as natural the stronger the imagined causalities are taken to be. Habits and customs are nothing but imagined naturalized relations between objects, they are second nature disguised as first nature.

Overlooking the prejudicial forces of habit inherent in many social institutions and conventions may stem from the fact that Hume’s account of artifice suffers somewhat from not being historically and psychologically specific enough. It does not explicitly react to, say, habitually ingrained partialities, and if it actually is meant to do so, it leaves entirely unclear how these could possibly be broken by artificial conventions. It thus abstracts from elements Hume is usually aware of, namely the powerful psychological forces leading humans to produce affective partialities through naturalizing imagined causal connections and thus seeing them as unchangeable. As in other concepts of habit, habit is a second nature that has forgotten its secondariness or its history and the interesting question is: why is this the case? Artifice, to be sure, cannot be in itself the guarantor of social progress, on the contrary, the artifices we ourselves set up can be a central factor of discrimination and exclusion. This, or so I claim, we can learn from Hume, though it does not reflect what most interpreters usually garner from his texts and may not be true to his intent. All I want to do is to weaken alternative readings of Hume’s naturalism. Thus, it is certainly possible to highlight Hume’s account of the natural virtues as opposed to the artificial virtues and claim that they are natural in a way not yet accounted for in my presentation of the question. However, as some have claimed, their naturalness may have been exaggerated. ⁵⁰ In fact, according to my interpretation, references to naturalness may be a motor of discrimination and differential treatment.

One could take this thought a step further in assessing the critical potential of Hume’s naturalism. Maybe moral equity is better secured by orientation to facts not of our own making than by artificially set up institutional measures meant to overcome given inequities. Let me go back to Hume’s concern about the ‘great resemblance among all human creatures’ that allows us to enter ‘into the sentiments of others, and embrace them with facility and pleasure’? ⁵¹ Hume knows that this is not a description of any given collective practice among humans. Does this mean that talk of resemblance among humans is all ideological or, less negatively speaking, merely utopian? Not quite. As we saw, Hume is aware of egoism, partiality and prejudice and even finds them natural in one of the senses used by him. As should be obvious by now, what I don’t agree with is his account of nature’s own remedy to these biases. Still, couldn’t we say that the presumption of resemblance introduces a natural presumption of morality or a presumption of resemblance necessary to any moral theory of potentially universal stripe? Sabina Lovibond, in her Ethical Formation, says this about the relation of first to second nature:

The process from which human intelligence emerges – the construction of a ‘second’ nature on the basis of the first – does not amount to an absolute supersession of ‘nature’ by the ‘culture’ with which it is standardly contrasted: rather, ‘nature’ continues to operate within ‘culture’ through those common (though involuntary) patterns of response which allow us to understand one another – when we do – by experiencing one another’s signs and gestures as immediately meaningful. ⁵²

Earlier in the same section, she speaks of a like-mindedness ‘not of our making’ as precondition for understanding one another or for responding to the virtuous person’s sensitivity to what is required in a morally relevant situation. I suggest that when Hume talks about our great resemblance he talks about just such a fact ‘not of our own making’. It is true, the reality of this resemblance will have to be experienced, and since we never fully experience it, it takes on an element of idealization. But it may be necessary to make use of such an idealization if progress is to be possible, if single instances of partiality and prejudice are not to have the last word. In a certain sense, we need to presume the practices of our making to be founded on practices not of our making in order for our moral endeavour to get off the ground and attain more egalitarian levels; after all, discriminatory partiality is always of our making. Awareness of this may be present in Hume’s distinction between natural and artificial virtues. However, in my reconstruction of this thought, I have insisted on this being a necessary presumption of our particular moral practices and not a natural fact to be scientistically discovered. This kind of first nature matters, but it matters only to the extent that we take it to matter.