Making Sense of Moral Perception

Abstract

The purpose of this paper is to demonstrate that Francis Hutcheson’s moral sense theory offers a satisfactory account of moral perception. I introduce Hutcheson’s work in §1 and indicate why the existence of a sixth sense is not implausible. I provide a summary of Robert Cowan and Robert Audi’s respective theories of evaluative perception in §2, identifying three problematic objections: the Directness Objection to Cowan’s ethical perception and the aesthetic and perceptual model objections to Audi’s moral perception. §3 examines Hutcheson’s moral sense theory, focusing on his discussion of benevolence, the desire for the happiness of others. I deal with the unresolved issues in Hutcheson’s account by recourse to Charles Darwin’s evolutionary perspective on the moral sense in §4, arguing for the moral sense as the second-order faculty for judging benevolence. I return, in §5, to the objections, showing that moral sense theory solves all three problems and therefore offers a satisfactory account of moral perception.

1 Introduction

Moral sense theory (MST) has been largely ignored in modern ethical inquiry. C.D. Broad mounted the only defence of the position in the twentieth century,¹ and was followed by Joseph Duke Filonowicz in the 21st.² There are several reasons for this lack of interest, the first of which is a conflation between MST and British sentimentalism. “Sentimentalism”, broadly construed, locates the foundation of morality in social affection, and advocates included the Third Earl of Shaftesbury, Hutcheson, Hume, and Adam Smith. The origin of the concept of “moral sense” is attributed to both Thomas Burnet (1697: 5) and Shaftesbury (2001), but Hutcheson is the only philosopher to have advanced a systematic MST, and I shall concentrate on his conception exclusively. A second reason for the lack of interest is conflation within Hutcheson’s oeuvre. His first treatise was published in 1725, and he developed and revised his philosophy until his death in 1746, both in later editions of earlier publications and in later publications. Confusion is exacerbated by the posthumous editing of his works, and the posthumous publication of two new works, one of which was written during the 1730s. For the purposes of this paper, I shall restrict my discussion to Hutcheson’s two most comprehensive and significant works: An Inquiry into the Original of our Ideas of Beauty and Virtue, In Two Treatises (first published in 1725; revised in 1726, 1729, and 1738)³; and An Essay on the Nature and Conduct of the Passions, with Illustrations upon the Moral Sense (first published in 1728; revised in 1730 and 1742).⁴ I shall employ the final edition of each published in his lifetime.

Hutcheson considered the senses of beauty and morality as a sixth and seventh counterpart to the five ‘external Senses’ of sight, hearing, smell, touch, and taste (2008: 8), but MST will be a non-starter if one adopts a folk psychology view of sensory modalities. Typically, such a perspective requires a dedicated sense organ and exteroception, i.e. sensitivity to external stimuli. To these two criteria one could add the condition that if the dedicated sense organ is unimpaired, it will standardly provide veridical information about the external stimuli. There is clearly no organ dedicated to the moral sense, and what precisely the moral sense would perceive requires elucidation. As Fiona Macpherson points out, however, these three criteria would make equilibrioception (the sense of balance) a sixth sense (2011: 16–17). Furthermore, if a sense must meet the criterion of exteroception, then—as Louise Richardson notes—the orthodox view amongst philosophers is that smell is not a sense (2013: 401–402). If exteroception is dropped as a criterion, then proprioception (the sense of the position and movement of the parts of one’s body) is an uncontroversial candidate for a seventh sense. Mohan Matthen’s observations on flavour suggest that there may not be a discrete organ for gustation as the sweetness in a chocolate mint is detected by transducers on the tongue, but the chocolate is detected by retronasal olfaction (Matthen M (forthcoming) Individuating the senses. In: Matthen M (ed) Oxford handbook of the philosophy of perception). Macpherson’s conclusion is that both the folk psychology view and the standard theoretical views (which individuate the senses in terms of representation, phenomenal character, proximal stimuli, and sense organs) fail to provide a convincing categorisation or support Aristotle’s view of five discrete senses (2011: 28–29). If moral properties can be perceived, there is therefore at least a possibility that this perception is made by means of a discrete sensory modality.

Although the evaluative perception debate is relatively recent, the association of morality with perception extends back to Aristotle. John McDowell⁵ and Martha Nussbaum (1990) have both emphasised the Aristotelian view, and it is particularly important in Nussbaum’s work. For her, moral perception is ‘the ability to discern, acutely and responsively, the salient [ethical] features of one’s particular situation’ (1990: 37). The kind of perception with which Nussbaum is concerned is the cognitive processing which links the ‘concrete perceptions’ of particular situations to general principles about behaviour and responsibility, i.e. perception as understanding rather than perception as awareness (1990: 37). Justin McBrayer refers to the former sense of moral perception as the virtue sense, and the latter as the epistemic sense (2010: 308–309). My concern is with the latter alone, and epistemic moral perception is ‘akin to sense perception, and […] occurs when a subject becomes aware of a moral property via a perceptual process’ (2010: 308). The tension between virtue moral perception (VMP) and epistemic moral perception (EMP) runs throughout the contemporary debate, just as it did in Hutcheson’s MST. Advocates of moral perception new and old desire to have the features associated with sensory perception—such as being universal to human beings and ordinarily veridical—but equally realise that moral perception is quite unlike visual, aural, and tactile perception.

Robert Audi makes a similar distinction in his work, employing vision as his paradigm and identifying ‘perceptual seeing’ and ‘intellective seeing’ (2013: 9). The former is a sixth sense and the latter is described as the way in which seeing an American flag displayed on someone’s porch is an indication—and therefore observation—of that person’s political viewpoint (2013: 9).⁶ This is similar to W.E.S. McNeill’s distinction between perceiving that one’s fuel tank is half empty by looking into the tank and perceiving that one’s fuel tank is half empty by looking at the fuel gauge. McNeill refers to the two types of knowledge gained as ‘primarily perceptual’ and ‘secondarily perceptual’ respectively (2010: 579). McNeill’s paper concerns the perception of anger, and he maintains that it is not implausible that one’s knowledge of another’s anger can be non-inferential (primarily perceptual) rather than inferential (secondarily perceptual). Audi uses the very same example—seeing anger—as a means of explaining moral perception (2013: 41). I shall argue that the moral sense is a sense of EMP which produces non-inferential knowledge.

2 Ethical & Moral Perception

Robert Cowan summarises his position on ethical perception as:

Ethical Perception (EP): normal ethical agents can and do have perceptual experiences (at least some of which are veridical) as of the instantiation of ethical properties (2015: 166).

He maintains that the existence of high-level perception is a sufficient reason to take ethical perception seriously, and high-level perception refers to ‘perceptual experiences which represent “high-level” properties (H-properties) such as natural kind properties and causal relations’ (2015: 168). Cowan notes that ethical perceptual representation could be either hard-wired or acquired by humans and suggests the latter as the most plausible explanation (2015: 175–176). He states that cognitive penetration is the most convincing account of this acquisition:

a visual experience, e, is cognitively penetrated if the representational content and phenomenal character of e are altered by states in the cognitive system, e.g., beliefs, concepts, desires, emotions, memories, imaginative states, intuitions, and where this does not merely involve those cognitive states having effects on the subject’s visual attention (2015: 176).

Cowan concludes that ethical perception is made possible by cognitive penetration and that ‘at least some ethical perceptual experiences confer non-inferential justification and knowledge’ (2015: 187).

He addresses several challenges to his account of ethical perception: there is no way that h-properties look, therefore they cannot be perceived (the Looks Objection) (2015: 168–175); there can be no perceptual experiences of things that cannot be observed (Unobservables Objection) (2015: 180–181); a perceiver must be perceptually counterfactually sensitive to the presence of what is perceived, but is not in the case of ethical perception (Sensitivity Objection) (2015: 183–186); and that in order to count as perception, the relation between the visual experiences and what they represent must be sufficiently direct (Directness Objection) (2015: 181–183). With the exception of the last, Cowan sees off all of these objections in a convincing manner. The motivation behind the Directness Objection is that—unlike sources of knowledge such as memory and testimony—perception offers non-inferential knowledge through ‘direct epistemic access’ (2015: 181). The objection is thus that the representation of ethical properties through the mechanism of cognitive penetration is not sufficiently direct to qualify as EMP. The Directness Objection holds for cognitive penetration in both ethical perception and the perception of protons (e.g., by observing a vapour trail in a cloud chamber); in both cases the mechanism is insufficiently direct to enable perception, and neither proton nor ethical perception therefore exists (2015: 182).

Cowan’s initial response is that if cognitive penetration is both widespread and indirect, then the idea that perception is ‘direct or unmediated’ should be rejected (2015: 182). In fairness, he immediately presents a second option: to accept that cognitive penetration is not sufficiently direct to enable ethical perception, but that the objection is itself insufficient to falsify ethical perception: ‘so long as there is veridical phenomenal representation of ethical properties, this is enough to make EP true’ (2015: 182). Both these responses involve dropping the directness requirement. My concern is that if this requirement is dropped, it will be difficult to distinguish between Cowan’s EMP and Nussbaum’s VMP, and that a clear distinction is crucial to the contemporary theories of evaluative perception.

Audi claims that ‘moral cognitions, such as moral judgments, can constitute perceptual knowledge but depend epistemically, though not inferentially, on non-moral elements’ (2013: 61). He advances an aesthetic model of moral perception by employing an analogy between aesthetic perception, cognition, and properties and moral perception, cognition, and properties, and draws attention to the relationship between the h-properties (high-level, high order, or perceptual properties), such as balance and unity, and the l-properties (low order, base, or observable properties), such as shape and colour, in aesthetic perception (2013: 35–37). Audi maintains that there is a similarly strong supervenience relation between h-properties and l-properties in moral perception:

Specifically, no two actions or persons can be alike in all their non-moral properties and differ in their moral ones, and no two paintings or other artworks can be alike in all their non-aesthetic properties and differ in their aesthetic ones (2013: 107–108).

This claim is, however, incorrect.

Kendall Walton (1970) demonstrated that the same non-aesthetic properties could have different aesthetic properties if a work was regarded as belonging to a different category of art. In his famous “guernica” example, he proposed a new art form in which every art work replicated the shapes and colours of Picasso’s Guernica exactly, but differed in their bas-relief dimensions. Walton showed that Guernica would be perceived as possessing the opposite aesthetic properties (“lifeless”, “bland”) to those usually identified (“dynamic”, “violent”), if the work was categorised as a guernica rather than a painting. For paintings, flatness is a standard property, and the shapes and colours on the canvas variable properties; for the “guernica” art form, shapes and colours are standard. Perceived as a guernica, the flatness is therefore expressive, hence lifeless and bland; perceived as a painting, the shapes and colours are expressive, hence dynamic and violent. Walton concludes that the aesthetic properties of a work are dependent not only on its non-aesthetic properties, but upon whether these properties are standard, variable, or contra-standard (1970: 338).⁷ Walton’s point can be understood in Cowan’s terms as follows: alternate cognitive penetrations of the l-properties of the painted canvas result in different h-properties.

Jonathan Dancy makes a more fundamental objection to Audi, to his use of a perceptual model to explain moral knowledge:

Both Audi and I have written as if the primary form of moral knowledge is, or at least might be, perceptual. But in fact (and I am sure he would not disagree with this) moral deliberation, decision and judgement normally occur prior to action. Since the whole question is whether to act or not, and how to act if one is going to act at all, we cannot be dealing here with anything analogous to perception, because there is nothing yet to perceive (2010: 115).

Audi’s aesthetic model of moral perception thus faces two objections: the lack of analogy between aesthetic perception and moral perception, and the inadequacy of the perceptual model of moral knowledge.

3 Moral Sense Theory

I shall argue for a version of the moral sense which draws on both Hutcheson’s initial theory and Darwin’s subsequent discussion, focusing on the continuity between the two in order to present a case for my claim that MST accounts for moral perception. Darwin had read the work of two of Hutcheson’s sentimentalist successors, Hume and Smith, and knew of Hutcheson through James Mackintosh’s The Progress of Ethical Philosophy (1837: 405), but the degree of his familiarity with Hutcheson’s work in general and MST in particular is not known. Hutcheson’s approach to arguing for the moral sense is the first indication of the compatibility of his theory with Darwin’s reflections on the moral sense in The Descent of Man, and Selection in Relation to Sex (1871).⁸ In contrast to the moral sense, Darwin does not consider the ‘Sense of Beauty’ (1871: 63) either controversial or unique to human beings (1871: 63–65), and most of the second part of the work is devoted to the operation of this sense (understood in terms of sexual selection) across different species. Hutcheson similarly regards the (internal) sense of beauty, order, harmony, and design⁹ as less contentious than the moral sense. He admits that the moral sense ‘may appear strange at first View’ (2008: 9), but states that if his readers accept that there is an internal sense, then ‘it will be no difficult matter to apprehend another superior Sense, natural also to Men, determining them to be pleas’d with Actions, Characters, Affections’ (2008: 10).

Hutcheson delineates sense in terms of pleasure and pain, and states that in the case of the five external senses, one’s perception of pleasure and pain is independent of one’s will (2008: 8). The mind is passive and unable to vary the perception or idea (2008: 9). He distinguishes two types of perception: the ‘sensible perception’ of the external senses, where material objects operate upon the organs of sense; and the non-sensible perception of the internal sense, which is produced by the contemplation of ‘Uniformity, Order, Arrangement, Imitation’ (2008: 9). He defines the internal sense as a determination ‘to be pleas’d with certain complex Forms’ (2008: 8) and identifies ‘Uniformity amidst Variety’ as the basis of these forms (2008: 67). Hutcheson justifies classifying the internal sense as a sense because it is ‘very like, in many respects’ the external senses (2008: 24). Like the sensible ideas of the external senses, the ideas of beauty are necessarily and immediately pleasant—one cannot alter them by the exercise of one’s will or by consideration of one’s interest (2008: 24). The internal sense is similarly universal to human beings (2008: 63). Hutcheson devises several strategies for dealing with the apparent variations, which include: the association of ideas (2008: 63 & 67–68) the conflation of grandeur and novelty with beauty (2008: 69),¹⁰ and the influence of custom, education, and example (2008: 71–74). The perspective which emerges from Hutcheson’s defence is that the internal sense is universal, but not uniform, i.e. all human beings have a faculty for appreciating uniformity amidst variety, but there is little agreement on what constitutes uniformity amidst variety.

The moral sense is also non-sensible, and defined as the ‘Determination to approve Affections, Actions, or Characters of rational agents, which we call virtuous’ (2008: 8–9).¹¹ Like the external and internal senses, the moral sense is a passive power of receiving ideas: ‘The moral Ideas do arise also necessarily in our Minds’ (2002: 75). One can neither avoid approving (or disapproving) of the affections of others nor of one’s own affections (2002: 75–76). Like the internal sense, but not the external senses, the moral sense is not present at birth and appears to require a minimum degree of intelligence.¹² This does not, however, mean that the internal and moral senses are the result of education (2008: 10), nor of custom or example (2008: 99). Like Hutcheson’s internal sense theory, his MST has an empirical foundation. He maintains that where opinions are divided on natural good (such as health, wisdom, and wealth), moral good (virtue) is universally acknowledged to produce approbation towards those perceived to possess it (2008: 85). He appeals to the evidence of ‘universal Experience and History’, which reveal that there are some actions which produce approbation, and others which produce disapprobation, in all human beings (2002: 173). From these actions, he proceeds to the quality itself, which is ‘kind Affection, or Study of the Good of others’ (2002: 173–174). Hutcheson’s evidence for a universal conception of virtue from experience and history is not compelling, but the claim that all human beings have affection for others is—given the social nature of the species—worth taking seriously. Furthermore, if human beings have a universal affection for one another, then the three causes of diversity identified might very well account for the lack of uniformity in the way in which this affection is manifested.

Given that Hutcheson believes he has already demonstrated the existence of a sixth sense, he regards the empirical evidence for a seventh as conclusive. He nonetheless develops his position in his subsequent discussion of benevolence, posing the following question:

It is true indeed, that the actions we approve in others, are generally imagin’d to tend to the natural Good of Mankind, or that of some Parts of it. But whence this secret Chain between each Person and mankind’ (2008: 91)?

Hutcheson cannot explain the source of this chain beyond the existence of a moral sense—that will have to wait until 1871—but he identifies the chain itself as benevolence, which is necessarily opposed to self-interest (2008: 103) and ‘is the Foundation of all apprehended Excellence in social Virtues’ (2008: 118). Human beings both desire the happiness of others, and approve of this desire in themselves and others.

From the Whole it may appear, that there is in Human nature a disinterested ultimate Desire of Happiness of others; and that our moral Sense determines us to approve only such Actions as virtuous, which are apprehended to proceed partly at least from such Desire (2008: 104/225).

Benevolence is the sole ground of approbation (2008: 136), and the final section of An Inquiry into the Original of our Ideas of Beauty and Virtue explains the moral sense exclusively in terms of benevolence:

• (1) the true Original of moral Ideas, viz. This moral sense of Excellence in every Appearance, or Evidence of Benevolence’ (2008: 177).

• (2) there is something in Actions which is apprehended absolutely good; and this is Benevolence, or Desire of the publick natural happiness of rational Agents; and that our moral Sense perceives this Excellence’ (2008: 181).

Understood in these terms, the moral sense is a faculty for the approval of benevolence in oneself and others. The moral sense both approves of benevolence and disapproves of its absence, however, so it is more accurately characterised as a faculty for judging benevolence. The moral sense is furthermore a second-order faculty: human beings not only have a desire for the happiness of others, but approve of this desire, i.e. take pleasure in benevolence and believe that benevolence is itself desirable. The conception of the moral sense as a second-order affection is explicit, if undeveloped, in Shaftesbury’s original theory (2001: 28) and although Hutcheson does not employ the term there is evidence that he envisages the moral sense in this way throughout the treatises, particularly if one analyses the somewhat confusing notion of a “reflective sense” (which I have deliberately avoided).¹³ The reflective sense aside, it is suggestive that Hutcheson altered his initial view of the internal and moral senses as being a faculty for taking pleasure in beauty and virtue respectively to the internal sense taking pleasure in beauty and the moral sense approving of virtue (2008: 8–9). If the moral sense is understood as a second-order faculty for judging benevolence, then the perception it enables—the reception of moral ideas or the perceptual representation of moral properties—is paradigmatically evaluative: the perception (of, e.g., an action) and the judgement (approval or disapproval thereof) occur simultaneously and neither the l- nor h-properties are perceived in isolation. I shall return to this point in §5.

4 Evolution, Conscience & Moral Sense

In The Descent of Man, Darwin sets out to show that there is no fundamental difference or ‘impassable barrier’ between human and beast, and that human beings are therefore simply higher rather than lower animals (1871: 49). He claims that moral sense or conscience is the most significant difference between humans and animals (1871: 70), but that any animal possessed of both social instincts and well-developed intellectual powers would develop a moral sense (1871: 71–72). The social instincts cause an individual to take pleasure in the society of others, and experience a sympathy which manifests itself in assisting those with whom the individual associates (1871: 72). The development of memory facilitates dissatisfaction when past selfish actions are judged in terms of present sympathy. The development of language facilitates the expression and discussion of normative guidelines in terms of acting for the good of society, and the final factor strengthening the social (and other) instincts is habit (1871: 72). Darwin is uncertain as to precisely how sympathy developed in animals, but defines it as ‘an instinct, especially directed towards beloved objects’ (1871: 82). With the development of the intellect in primeval human beings or their ape-like progenitors, the individual would realise that the instinct to assist fellow humans was advantageous due to the reciprocal assistance they provided. This would result in the habitual performance of benevolent actions, which would in turn strengthen sympathy.¹⁴ Darwin maintained that the most powerful stimulus to the development of sympathy was the development of the love of the approbation, and fear of disapprobation, of fellow humans (1871: 164–163). Although he does not discuss this, it seems as if this desire for approval and fear of disapproval were—at least initially—self-serving, but subsequently strengthened the sympathy from which they arose.

The development of the moral sense in human beings appears to be along the following lines. Humans are subject to different desires, some of which are intense, but then disappear when satisfied (i.e. are strong, but leave weak impressions), and others which are less intense but more enduring (i.e. leave stronger impressions or are always active in the mind). In a case where myself and a colleague are both hungry, e.g., my intense desire for food might be stronger than my desire for my colleague’s well-being. As soon as my hunger is satisfied, however, my desire for food drops dramatically—below my desire for my colleague’s well-being—such that I am unhappy that I have caused her to go hungry. In order to prevent this unhappiness occurring in future, I may decide to share my food with her next time. The key factors here are not just sympathy, but the ability to remember, compare, and judge my desires and actions, which is why the moral sense requires a certain minimum intellect.

Darwin states that a ‘moral being is one who is capable of comparing past and future actions or motives and of approving or disapproving of them’ (1871: 88). He is clear that the social instinct is not stronger than selfish instincts, but that human beings cannot avoid reflecting on their past actions and experiences due to their intellectual development (1871: 89). Humans are therefore subject to the operation of the mechanism described above, and the comparatively weaker past impressions of acts of self-preservation with current feelings of sympathy cause disapproval of these actions and feelings of remorse (1871: 90). Darwin maintained that humans acquired the habit of self-command (which was probably inherited), that facilitated the following of the more persistent social instinct of sympathy (1871: 91–92). Thus:

The imperious word ought seems to merely to imply the consciousness of the existence of a persistent instinct, either innate or partly acquired, serving him as a guide, though liable to be disobeyed (1871: 92).

Darwin noted that even when a human being does not feel remorse for selfish actions, the individual is likely to feel dissatisfied with the social disapproval which would follow if the actions became public knowledge.¹⁵ There is evidence of the power of the effect of the judgement of the community in the development of ‘absurd rules of conduct’ and ‘absurd religious beliefs’ which have arisen throughout the world due to poor reasoning about what is actually in the interests of the communities in which they occur (1871: 99).¹⁶

Darwin maintains that anthropological evidence of both the moral sense and the social instinct being exclusive to the social community suggests that the former is derived from the latter (1871: 97). As civilisations developed and tribes were united into nations, reason led to the extension of human sympathy beyond the immediate community to the community of the nation, even though most of this community would be strangers.¹⁷

This point being once reached, there is only an artificial barrier to prevent his sympathies extending to the men of all nations and races. If, indeed, such men are separated from him by great differences in appearance or habits, experience unfortunately shews us how long it is before we look at them as our fellow-creatures. Sympathy beyond the confines of man, that is humanity to the lower animals, seems to be one of the latest moral acquisitions. It is apparently unfelt by savages, except towards their pets (1871: 100–101).

Darwin concludes by claiming that while there can be no doubt as to the immensity of the difference between the intellect of the highest animal and lowest man, the difference is one of degree rather than kind (1871: 104–105). It is the moral sense, however, ‘which perhaps affords the best and highest distinction between man and the lower animals’ (1871: 106).

Darwin employs moral sense and conscience as synonymous throughout his discussion, but distinguishes them as follows:

The above view of the first origin and nature of the moral sense, which tells us what we ought to do, and of the conscience which reproves us if we disobey it, accords well with what we see in the early undeveloped condition of this faculty in mankind (1871: 93).

The moral sense therefore motivates future action where the conscience is retrospective, and I take conscience to be one aspect of the moral sense (an aspect which is sometimes used synecdochically). The following description shows the similarity of Darwin’s conception to Hutcheson’s, which includes the use of several of the same terms:

Ultimately a highly complex sentiment, having its first origin in the social instincts, largely guided by the approbation of our fellow-men, ruled by reason, self-interest, and in later times by deep religious feelings, confirmed by instruction and habit, constitute our moral sense or conscience (1871: 165–166).

Thus understood, I take Darwin’s moral sense as compatible with Hutcheson’s MST. Moral sense is not merely social instinct for Darwin—the benevolent actions observed in numerous species—but a normative reflection on past or future behaviour, i.e. a faculty to judge the actions produced by the social instinct. Hutcheson refers to the sense as a passive power and Darwin maintains that human beings cannot avoid reflecting on their actions. Both agree that the sense is universal to human beings, even if conceptions of benevolence differ. The denotation I employed above, moral sense as a second-order faculty for judging benevolence, thus requires no amendment. Darwin’s contribution explains how the moral sense advanced in Hutcheson’s theory developed, determines the origin of the chain of benevolence which links human beings, and offers further evidence that the moral sense is independent of the will and universal to human beings.

The claim that human beings share a universal approval of benevolence may appear to admit of too many counter-examples to be of any philosophical interest. It is thus important to be clear precisely what would constitute a counter-example. The regrettable ubiquity of war and injustice throughout human history—and in the contemporary world—will not suffice. Darwin noted that the moral sense was extended from the tribe to the nation, most likely by the paradoxical means of the conquest of neighbours and the expansion of empires. One cannot say that the Greeks or Romans lacked a moral sense because their societies included slaves, or that the Aztecs and Incas lacked a moral sense because they routinely sacrificed prisoners of war, or that the Spanish colonists lacked a moral sense because they treated the South American population inhumanely. In each case the benevolence was simply, as both Hutcheson and Darwin noted, restricted to a particular group: citizens, not slaves; allies, not enemies; settlers, not natives. Much like the recent concerns about the relationship between humans and animals, slaves, enemies, and natives simply lacked moral status in the respective societies or, at best, enjoyed limited moral consideration. A counter-example to Darwin would be a society where benevolence was not restricted to a social elite, but absent, something akin to Hobbes’ Commonwealth, where human beings band together and subject themselves to the rule of a monarch for the sole reason of avoiding the anarchy of the state of nature (2008: 85). In such a society there would be no benevolence, only carefully calculated self-interest which sometimes resulted in apparently benevolent actions. I take it that no such society exists, and that there is no evidence for such a society having existed, which appears to be Darwin’s view. In addressing the three objections from §2, I shall now indicate how the moral sense grounds moral perception.

5 Moral Sense, Moral Perception

My concern with Cowan’s response to the Directness Objection is that dropping directness as a criterion for perception risks blurring the distinction between EMP and VMP, a distinction which is both tenuous and essential to the evaluative perception debate. The solution from MST is to retain the link between directness and sensory perception, but interpret directness in terms of Hutcheson’s definition of a sense as a faculty of receiving ideas (or the faculty of the perceptual representation of moral properties) independently of the will of the perceiver. If I am not visually impaired, I cannot help seeing, e.g., a postbox as red. The only way I can stop seeing the redness is to close my eyes or turn my head away. There is a parallel with Gilbert Harman’s example, which is as concise as it is compelling:

If you round a corner and see a group of young hoodlums pour gasoline on a cat and ignite it, you do not need to conclude that what they are doing is wrong; you do not need to figure anything out; you can see that it is wrong (1977: 3).¹⁸

If I am not morally impaired (like the hoodlums), then the only way I can stop perceiving the wrongness is to close my eyes or turn my head away. If I am looking at the postbox and the burning cat my perception is direct in both cases, i.e. necessary. I cannot will myself to see the postbox as being another colour, nor can I will myself to stop seeing the torture of the cat as wrong. It is also immediate in both cases: there is no time-delay between seeing the postbox and seeing its redness or between seeing the burning cat and seeing its wrongness.

Understanding directness in terms of necessity distinguishes moral perception from proton perception. Only a physicist seeing the vapour trail would know that he or she was seeing a proton; someone without that knowledge would merely see the vapour. In the case of the burning cat, all but the morally impaired see the wrongness. If proton perception operates by means of cognitive penetration, then the cognitive penetration appears to be both susceptible to the Directness Objection and distinct from the cognitive penetration involved in perceiving the wrongness of the burning cat. The difference means that I am not committed to the existence of a subatomic particle sense. I may be committed to the view that the moral sense requires a minimum level of both intellectual and emotional development. Indeed, the necessity of the perception of the wrongness of torturing the cat may, unlike the perception of the redness of the postbox, be explained by the emotional capacity of the perceiver. I do not wish to advance such a theory here, merely to acknowledge that if grounding moral perception in a moral sense commits me to a version of sentimentalism, then I do not regard that consequence as problematic. 

The objection to Audi was twofold: first, that moral and aesthetic perception are insufficiently alike for the latter to be a model for the former; second, that the perceptual model of moral knowledge is inadequate. The first of these appears telling against MST because Hutcheson employs what he regards as the plausibility of the internal sense to make his readers more amenable to the more controversial moral sense. Furthermore, the two senses share the characteristics of being non-sensible, independent of the will, immediate, and universal. Despite these similarities, however, there is one very important difference. The moral sense is a second-order faculty of judging benevolence whereas the internal sense is a first-order faculty of taking pleasure in complex forms. The internal sense may or may not require a minimum level of intellect or involve a degree of cognitive penetration, but it is nonetheless simply the recognition of beauty. In contrast, the moral sense involves not only the recognition of benevolence, but the judgement of this benevolence. I am uncertain as to whether or not Hutcheson believed that the internal sense was a first-order faculty. He employs the terms ‘contemplation’ (2008: 19) and ‘reflection’ (2008: 35) when describing the human reaction to uniformity amidst variety, but fails to provide evidence for a second-order faculty in the manner of the moral sense. Ultimately, his position on the internal sense is ambiguous, and there is good reason to hold that it is a first-order faculty if one considers the relevant l- and h-properties.

When I look at Guernica, I see both the l-properties (shapes and colours) and the h-properties (violence and dynamism). I can, however, focus on the l-properties alone, e.g. the way in which the shapes represent human and animal heads, or the variation of colour from black to brown to grey to blue to white. When I see the burning cat, I see both the l-properties (perceived by my senses of sight, sound, and smell) and the h-properties (perceived by my moral sense). I cannot, however, separate the sight, sound, and smell of the burning cat from its wrongness in the way I can with the l- and h-properties of the painting. This is why I maintain that sense-based moral perception is paradigmatic of evaluative perception. In the aesthetic case, there is—at the very least—an argument for separating the sensory perception from the aesthetic judgement. If they can be separated, then there may be two processes, perception plus evaluation, which seems closer to VMP than the EMP with which I am concerned. Alternately, where the perception and judgement cannot be separated, there is—again, at the very least—an argument that the perception and the evaluation are part of the same process, i.e. evaluative perception. MST does not therefore employ an aesthetic model of moral perception.

With regard to the second objection, MST does not entail that moral knowledge is predominantly perceptual. Hutcheson holds that there is no nation, club, or even single individual who is entirely indifferent to right and wrong (2008: 138), but his claim is unsubstantiated. Darwin’s view is much more sophisticated, as his notebooks reveal. Notebook M considers the issue of whether there is a moral sense which is universal to humanity. Darwin supports the view that all cultures perceive right and wrong, and maintains that the cross-cultural variation in ‘conscience’, i.e. as to what is right and what is wrong, is ‘no more wonderful than dogs should have different instincts’ (1838: 279). He picks this thought up again in Notebook N, where he writes that the result of the different instincts in solitary animals, bees, and dogs is that one would not expect a dog’s conscience to be the same as a man’s (1838–40: 330). This sentiment appears in The Descent of Man, where Darwin states that if other animals were to develop an intellect as powerful as man’s their moral sense would nonetheless differ:

In the same manner as various animals have some sense of beauty, though they admire widely different objects, so they might have a sense of right and wrong, though led by it to follow widely different codes of conduct (1871: 73).

He then states that if it could be proved that there was a variation in moral sense between nations, this would be evidence for rather than against the existence of such a sense (1838–40: 330), but does not return to this point. Darwin has a much more realistic view of the degree of variation than Hutcheson, and is able to accommodate the variety in his conception of a universal moral sense.¹⁹

I agree with Darwin: all human beings who are not morally impaired, i.e. are not suffering from antisocial personality disorder or an equivalent, have a moral sense, but the moral sense is universal rather than uniform. The second-order faculty of judging benevolence is diverse to the extent that different people will often perceive the same l-properties but different h-properties, and this diversity is cause for caution. As such, I also agree with Dancy that the perceptual model of moral knowledge is inadequate. Even if he overestimates the non-perceptual features of moral knowledge, the different ways in which this sense operates suggest not only that it cannot explain morality on its own, but that something more than moral sense is desirable when making moral decisions. That the moral sense fails to provide a comprehensive account of moral knowledge is not, however, an argument against its existence.

In this regard, MST may be most useful in conjunction with a two-tier explanation of moral judgement, such as R.M. Hare advances in Moral Thinking: Its Levels, Method, and Point (1981). One of Hare’s main concerns is moral conflicts within the individual and he accounts for their existence—and the means by which they should be resolved—in terms of a lack of complete compatibility between the intuitive and the critical levels of moral thinking (1981: 44–64). MST could enhance the understanding of the intuitive level while simultaneously showing the necessity of the critical level in moral judgement. Pekka Väyrynen’s²⁰ example of a utilitarian with moral justification for the burning of Harman’s cat illustrates the incompatibility nicely: despite knowing that the torture of the cat is right (at the critical level) and therefore failing to take action (because the critical level trumps the intuitive level), the utilitarian still perceives the same wrongness as the non-utilitarian. Recent work in empirical psychology, in particular Jonathan Haidt’s The Righteous Mind: Why Good People are Divided by Politics and Religion, appears to offer further evidence for the existence of a moral sense and also appeals to evolution (2012: 189–220).²¹ Hare and Haidt’s theories raise a further point of interest, the utilitarian perspective shared by Hutcheson and Darwin and the strength of the relation between MST and utilitarianism, which is regrettably beyond the scope of this paper. I shall therefore conclude by reiterating that MST could solve at least three problems in contemporary theories of moral perception and thus offers a satisfactory account of moral perception.²²