1. The Causal Challenge

Intuitions are widely assumed to play an important evidential role in ethical inquiry. Thus, an ethical theory is sometimes said to be justified either if, or only if, it achieves a good match with ethical intuitions. Read as a necessary condition, this claim explains the common practice of rejecting ethical theories when they fail to match common intuitions. So read, however, this claim is apt to encourage the dubious inference that ethical intuitions cannot derive justification from any other source. Read as a sufficient condition, the claim explains the common practice of appealing to intuitions in support of ethical theories. So read, however, the claim is apt to encourage the dubious inference that justification derived from a match with intuitions cannot be defeated by other considerations.

When and how appeals to intuitions are appropriate, and what their epistemological significance is, are questions often left unanswered by those who emphasise their evidential role in ethical inquiry. This claim is more often simply assumed than explicitly argued for. One recent exception to this rule is Brad Hooker, who has explicitly defended the claim that ethical intuitions are ‘beliefs that come with independent credibility’.Footnote ¹ Yet even this claim is open to more than one interpretation. First, we might think of independent credibility as indefeasibility. On this reading, to have an ethical intuition is to get things right. Some philosophers have arguably thought of ethical intuitions this way, and I shall return to this idea below. Second, we might think of independent credibility as defeasible. On this reading, intuitions are thought of as having some kind of presumptive credibility that can be defeated by other considerations brought to light in the course of inquiry. It is arguably this interpretation of ‘independent credibility’ that makes best sense of Hooker's claim. It is in any case this interpretation of the evidential role of intuitions that I shall take as my main focus in this paper.

The presumptive credibility of ethical intuitions can be challenged in a number of ways. One obvious challenge arises from the fact of disagreement between intuitions.Footnote ² Thus, when ethical theorists use thought experiments to draw out our ethical intuitions it is rare that everyone swings the same way, even within the same social context.Footnote ³ A second challenge arises from the apparent lack of coherence between intuitions within the same person. This problem is illustrated by the ease with which many of us are prone to get stuck when trying to ethically distinguish between variants of the same though experiment, such as killing one person in order to save five by, alternatively, pushing them onto a train, pushing the train onto them, wobbling a handrail on which they are leaning, flipping a switch that turns on their roller skates, or flipping a switch that changes a train's direction from one track to another.Footnote ⁴

Both of these challenges are closely related to a third, namely the challenge of showing that intuitions do not simply express some form of irrational prejudice. Thus, intuitive responses to thought experiments are sometimes alleged to track the presence of ethically irrelevant factors, such as novelty, excitement, disgust, surprise or arbitrary convention.Footnote ⁵ Thus, experimental psychologists have shown that people retain strongly negative reactions to imagined human incest behaviour even after being informed that the cases described are ones in which the ethically problematic features commonly associated with incest (such as risk of pregnancy, disability through inbreeding, familial trauma, and the like) are absent. The suggestion is that many people would disapprove of such behaviour even if it were not wrong. In a different kind of study, other experimental psychologists claim to have shown that some intuitive responses to thought experiments such as the infamous Trolley Problem are correlated with the presence of high levels of activity in parts of the brain causally responsible for emotional activity.Footnote ⁶ One response to these studies has been to reject all appeals to ethical intuition as unreliable on the grounds that the primacy of emotions in the genesis of intuitions undermines their credibility in contrast with beliefs that have their genesis in our reasoning or calculating faculties.Footnote ⁷

A fourth challenge is focused on causally more distant (social, historical or evolutionary) causes of ethical intuitions, as opposed to their more proximate (physical or psychological) causes. This challenge is closely related to the third (and thereby also to the second). At least some of the distant causes of ethical intuitions may figure in the causal explanation of at least some of the proximate causes of those intuitions. If so, the ethical rationale for these intuitions might be a function of the specific circumstances in which they originally appeared. It is then an open question whether these intuitions retain their ethical rationale in the current circumstances in which they later find their manifestation. First, the fact that these intuitions were somehow ‘adaptive’ in some distant ‘ancestral environment’ does not entail that they remain so adaptive. Second, even if the intuitions in question remain so adaptive it is possible to ethically question the kind of adaptive fitness they have historically served. Thus, the conception of our ethical sensibility as an adaptive mechanism for biological reproduction, for example, is one that a continues to be a subject of substantial ethical controversy.Footnote ⁸

In what follows I say nothing directly about the first two challenges. My primary aim is to evaluate the claim that the epistemological credentials of ethical intuitions are undermined by their causal pedigree and functional role. I refer to this claim as ‘the causal challenge’. This is not to suggest that I dismiss the first two challenges to the epistemological credentials of ethical intuitions. I do not. Addressing these challenges is a topic for another occasion.Footnote ⁹

There is more than one possible response to the causal challenge. Here I mention three. The first is one of rejection.Footnote ¹⁰ According to this response, ethical intuitions are in principle immune to the causal challenge. This response might be plausible if all ethical intuitions were indefeasibly directed at a domain of a priori knowable and necessary truths, our access to which were independent of any empirical evidence beyond that required in order to grasp the concepts out of which intuitions are said to be composed. It might also be plausible on a conception of ethical theory as a purely descriptive exercise of mapping the structure and content of some particular ethical sensibility. I shall not pursue either of these options here.Footnote ¹¹ First, even if it is plausible that some ethical intuitions can be fully accounted for in terms of the grasp of a priori knowable and necessary truths, it unlikely that all can. We still want to understand the epistemological credentials of ethical intuitions that do not fall into this category. Second, the idea of ethical theory as a purely descriptive exercise does not accurately describe actual practice in contemporary ethical theory. Much ethical theory consists in the critical evaluation of ethical intuitions in the context of more theoretical ethical claims, as opposed to a primarily descriptive mapping of their many and varied instantiations.

A second response to the causal challenge is one of scepticism. According to this response, the causal pedigree and functional role of ethical intuitions undermines their presumptive credibility. Even if we know that some of our ethical intuitions are plausible we also know that some of them are not, and no mere appeal to intuition is going to tell us which is which. Our only hope of avoiding ethical scepticism is therefore to base ethical inquiry on some objective foundation outside the domain of ethical intuition. One version of this response has recently been defended by Peter Singer, who suggests that we should abandon the reliance on ‘particular moral judgements we intuitively make’ and ‘start again from as near as we can get to self-evident moral axioms’.Footnote ¹²

In fact, Singer's response to the causal challenge is less radically different from the rejectionist challenge than these remarks may suggest. As I will show in the next section, an objective foundation of ethical theory in self-evident moral axioms just is a foundation of ethical theory on ethical intuition in at least one historically influential sense of that term. One of the things that a faculty of ethical intuition has traditionally been thought to do is provide access to self-evident principles of practical reason.Footnote ¹³ On a charitable reading of Singer's claim, what he is questioning is the presumptive credibility ethical intuitions that fail to meet this criterion. A more accurate way of thinking about his version of the sceptical response is therefore to think of it as a disjunctive claim that ethical intuitions have no presumptive credibility unless they can be shown to be at least consistent with self-evident principles of practical reason, via independently plausible empirical premises. This reading arguably also makes sense of Singer's contrast, implicit in the remark just quoted, between (general) moral axioms on the one hand, and what he refers to as ‘particular moral judgements’ on the other.

A third response to the causal challenge is one of integration. According to this response, some ethical intuitions are presumptively credible even though they may not be grounded in self-evident truths of practical reason. These intuitions are presumptively credible insofar as there is reason to believe that they would cohere with a causally informed ethical outlook. On this view, the rational response to the causal challenge is to ask whether any given set of intuitions targeted by that challenge would cohere with an ethical outlook informed by our best evidence about their proximate and distant causes. According to the version of the integrationist response I sketch in this paper, the answer to this question is sometimes positive. Moreover, for a significant range of ethical intuitions there is good evidence that this is so. The causal challenge therefore fails to undermine the presumptive credibility of ethical intuitions in the general way implied by the sceptical response.

In what follows I proceed in three steps. First, I clarify how the target of the causal challenge should be understood. I give an account of five different ways in which the term ‘intuition’ has been interpreted in modern moral philosophy and explain how ethical intuitions thus interpreted are potential targets for the causal challenge. Second, I give an account of the causal challenge by placing it in a wider theoretical context. I then explain why it is reasonable to believe that a significant range of intuitions would survive that challenge. In the last section of the paper I argue that the causal challenge should not only be understood as a narrowly epistemological challenge, but also as a substantially ethical one. In doing so, I briefly outline how this fact can illuminate the epistemology of ethical intuitions.

Before embarking on this task it is necessary to make two points of clarification. First, I distinguish the idea of an ethical intuition from the thesis of ethical intuitionism. Ethical intuitions as I shall understand them are individual attitudes (or the contents of attitudes) that in ethical theory have historically been invoked as justifiers, defeaters or otherwise evidential in the assessment of theoretical claims. Ethical intuitionism as I understand it is the thesis that some ethical claims (such as self-evident principles of practical reason) can play the role of non-inferentially grounded justifiers or defeaters in ethical theory.Footnote ¹⁴ Ethical intuitionism thus understood is consistent with, but does not entail, the thesis that the ethical claims that play this privileged role are ethical intuitions. As I will shortly argue, whether or not they are depends on how the term ‘ethical intuition’ is understood. Nor does the existence of ethical intuitions entail the thesis of ethical intuitionism under all its standard interpretations. First, if ethical intuitions have no presumptive credibility they are unsuitable for invocation as justifiers, defeaters, or as otherwise evidential in ethical theory. Second, if the presumptive credibility of ethical intuitions consists purely in their likelihood of fitting into a stable and coherent network of ethical and other beliefs their existence does not entail that some ethical claims are non-inferentially grounded. The case for the presumptive credibility of ethical intuitions is therefore in principle neutral with respect to the truth of ethical intuitionism in this sense.Footnote ¹⁵

Second, some versions of the causal challenge have a primarily metaphysical focus.Footnote ¹⁶ When the causal challenge has a metaphysical focus the main question is whether facts about the causal pedigree and functional role of ethical intuitions show that they are not (or cannot be) true, or that they should not be thought of as responsive to a mind independent ethical reality. These questions are orthogonal to the argument in this paper. The question at issue here is whether ethical intuitions are presumptively credible. This question could have a positive answer if there were no mind independent moral facts, if all ethical claims were false, or if ethical claims were not fundamentally in the business of aiming at the true representation of moral facts at all.Footnote ¹⁷ Any plausible view about the nature and status of ethical claims is committed to some conception of epistemological credibility for ethical claims, even if only to the minimally contrastive claim that some ethical claims are more credible than others. A commitment to the plausibility of this contrastive claim about ethical intuitions is all that is required for my purposes in what follows.

2. Five Conceptions of Ethical Intuition

Philosophical uses of the term ‘ethical intuition’ are many and varied. For present purposes it will be useful to distinguish five (the list is not exhaustive).

Sometimes when moral philosophers use the term ‘ethical intuition’ what they have in mind is nothing more than a common ethical judgement, widely shared between thinkers in a certain domain, and endorsed regardless of explicit engagement with, or commitment to, any particular ethical theory.Footnote ¹⁸ I shall refer to this as ‘the common judgement interpretation’. Thus understood, the idea of an ethical intuition is neutral with respect to its epistemological credentials. Common ethical judgements take a wide variety of forms in different circumstances, and are frequently based on prejudice or dubious authority, such as gender bias, racism or superstition. Intuitions as understood on the common judgement interpretation are therefore open to the causal challenge in most of its familiar forms.

On a second understanding ethical intuitions are thought of as ‘immediate’ or ‘non-inferential’ judgements. I shall refer to this as ‘the immediate judgement interpretation’.Footnote ¹⁹ In fact, there are two (in principle compatible) ways of understanding intuitions on this interpretation. On the first conception intuitions are defined as ‘strong, immediate reactions to the description of real or imaginary examples’, arrived at ‘intuitively’ and ‘spontaneously’.Footnote ²⁰ On the second conception, ‘a moral intuition is a spontaneous judgement, often concerning a particular act or agent, though an intuition may also have as its object a type of act or, less frequently, a more general moral rule or principle’.Footnote ²¹ Like the first interpretation, the immediate judgement interpretation is also neutral with respect to the epistemological credentials of ethical intuitions. It follows that the presumptive credibility of ethical intuitions is in principle defeasible due to ignorance of their causal pedigree and functional role, including the proximate psychological and largely sub-conscious mechanisms that throw up immediate and non-inferential judgements about cases, norms or principles. On this interpretation, therefore, intuitions are as open to the causal challenge as they are on the common judgement interpretation.

On a third interpretation it is a necessary condition of a subject having an intuition that he or she has achieved at least some critical understanding of its content.Footnote ²² On this interpretation an ethical intuition is a considered judgement the causal pedigree of which stops short of explicit engagement with ethical or other theorising. I shall refer to this as ‘the pre-theoretical judgement interpretation’. The contemporary philosopher who has arguably done most to articulate this interpretation is Robert Audi, who defines the idea of a ‘genuine’ intuition as follows: genuine intuitions are 1) cognitive (as opposed to purely affective or emotional); 2) ‘non-inferential (i.e. not consciously held on the basis of a premise); 3) firm (an intuition is ‘a moderately firm cognition’); 4) comprehended (i.e. ‘formed in the light of an adequate understanding of their propositional objects’); and 5) pre-theoretical (i.e. ‘neither evidentially dependent on theories nor themselves theoretical hypotheses’).Footnote ²³ On this interpretation intuitions are something of an epistemological mongrel. On the one hand, the idea of an ethical intuition is normatively loaded in virtue of the fact that there is a success-condition on having an intuition, namely an ‘adequate’ understanding of the nature of its object (although it may be subject to debate exactly what ‘adequate’ means in this context). On the other hand, retaining an intuition is compatible with massive ignorance both of its causes and of its justificatory relations to claims brought out in the course of theoretical inquiry. It follows that on the pre-theoretical judgement interpretation intuitions are open to the causal challenge.

A fourth interpretation ‘ethical intuition’ is that of an apparently self-evident truth.Footnote ²⁴ This ‘apparent self-evidence interpretation’ is also logically neutral with respect to the epistemological credentials of ethical intuitions. What initially appears as self-evident is defeasible, in part because a causal explanation of this appearance could show it to depend on criteria of ethical discrimination that would fail to earn their place in a causally informed ethical theory. The apparent self-evidence interpretation is therefore also open to the causal challenge.

On a fifth interpretation an ethical intuition just is a self-evident truth or principle. I shall refer to this as ‘the genuine self-evidence interpretation’. This interpretation is endorsed by Judith Jarvis Thomson in a recent book, where she describes ethical intuitions as ‘obvious truths’.Footnote ²⁵ The genuine self-evidence interpretation is obviously not neutral with respect to the epistemological credentials of ethical intuitions, correctness being one of their constitutive features. Even so, the definition of ‘ethical intuition’ as a genuinely self-evident truth only evades the causal challenge on the assumption that there actually is an interesting range of intuitions in the required sense. Furthermore, this interpretation is silent with respect to the epistemological credentials of the vast range of ethical claims that fail to have this status. The genuine self-evidence interpretation is therefore of limited interest when discussing the epistemological credentials of many of the allegedly ‘intuitive’ claims that moral philosophers have historically appealed to in the course of evaluating different ethical theories. Finally, it does not follow from the fact that some ethical claim is genuinely self-evident that every agent who entertains it is actually justified in believing that it is. One source of defeat for an agent's claim to be justified in believing that a genuinely self-evident ethical claim is genuinely self-evident is exactly the kind of causal factors targeted by the causal challenge, the apparent self-evidence of the claim in question being an illusory product of psychological or other causes of which the subject is either partly or wholly ignorant. It follows that the causal challenge is a possible epistemic debunker of ethical judgements the contents of which are, in fact, genuinely self-evident; although the subject in question is mistaken in thinking that they know that they are. Although there clearly is a justification for the claim in question, this is not a justification that the subject has.

How should we understand the idea of an ethical intuition? Given the variety of uses to which the term ‘ethical intuition’ has historically been put since its introduction into mainstream English speaking moral philosophy by Price and others in the Eighteenth Century, this is as much a terminological question as a matter of philosophical substance. In what follows I shall be using the term disjunctively, so as to include each of the four received interpretations of that term that fall short of genuine self-evidence. On this interpretation, the term ‘ethical intuition’ is taken to signify a pre-theoretical ethical belief or attitude, where this may include anything from a pre-cognitive ‘gut reaction’ on the one hand, to a considered ethical judgement on the other. I choose to define the term in this inclusive way for two reasons. First, these four interpretations together pick out a range of ethical responses that have each been the target of some version of the causal challenge. Second, the account I go on to sketch of how intuitions can acquire their epistemological credentials makes no fundamental distinction between any of these interpretations insofar as their epistemological credentials are concerned. However, understanding the idea of an ethical intuition in this inclusive way is also consistent with the possibility that the causal challenge is more serious when targeted at some of these intuitions rather than others, e.g. depending on the extent to which different intuitions result from more or less reflective awareness of their contents or causes. It is also consistent with the possibility that someone's capacity to respond to the causal challenge will vary, depending on the extent to which the intuitions in question are themselves revisable in response to inquiry (as some considered ethical judgements may be) or whether responding to that challenge depends on someone's ability to affect psychological, social or institutional factors external to those intuitions themselves (as might be the case with emotional ‘gut reactions’ that show a high degree of independence from rational or other agential control). I shall briefly return to this issue in the final section. For the rest of this paper I shall mainly set aside any further discussion of the genuine self-evidence interpretation. It is worth pointing out, however, that as this interpretation has had some currency in modern moral philosophy, the current debate about the place of intuitions in ethical theory cannot be simply projected backwards into the past without the danger of anachronism. To be an ‘intuition-basher’ at the start of the Twenty-first Century is not the same thing as it was to be an ‘intuition-basher’ at the start of the Twentieth or the Nineteenth.

3. Addressing the Causal Challenge

In this section I describe one way in which some ethical intuitions could stand up to the causal challenge. I argue that a significant range of actual ethical intuitions would, in fact, stand up to this challenge in this way. I proceed by first locating the challenge in a wider theoretical framework. I then address it directly as it affects the presumptive credibility of some common ethical intuitions.

I say that the causal explanation of an ethical intuition would epistemologically debunk that intuition if the following two conditions are met. First, the existing rationale for the intuition in question does not stand up to reflective ethical scrutiny. Such reflective failure will occur, for example, if the end states that the relevant causal mechanisms actually bring about are not ones that, on balance, we would endorse after causally informed ethical reflection. Second, there is no reflectively robust (and epistemologically accessible) alternative for the intuition in question. (This second condition is imposed in order to prevent the causal challenge from collapsing into an instance of the so-called ‘genetic fallacy’).Footnote ²⁶

On the picture of causally informed ethical inquiry that I sketch in this paper, addressing the causal challenge in practice is a matter of testing for reflective coherence between ethical intuitions and evidence about their causes and functional role. This process inevitably takes place against the background of an extensive network of existing ethical and non-ethical commitments. It will therefore make either explicit or implicit use of existing ethical commitments, at least some of which are themselves defeasible in light of further inquiry. These commitments will include judgements about what features of different situations are ethically relevant and which among those features are ethically more important than others in different situations. This reliance on prior ethical commitments does not amount to an embarrassing form of circularity. It is a necessary consequence of the fact that the debunking potential of causal explanations of ethical intuitions can only be comprehensively evaluated by confronting those explanations with a network of ethical and non-ethical commitments. Where coherence is found lacking the reasonable response is to make adjustments to the network of existing commitments, either by giving up the relevant intuitions or by revising some other ethical or non-ethical commitments that conflict with them, depending on the comparative degree of confidence it is reasonable to attach to each. The process of revising a body of ethical commitments in light of causally informed reflection is a holistic, fallible and open-ended process. It may nevertheless constitute a genuine epistemological improvement if the following three conditions are met. First, it is possible to access the relevant causal explanations. In other words, we are not stuck in an ‘Evil Demon’ type scenario with respect to the causal explanation of our ethical sensibility (rejecting this condition would remove much of the distinctiveness of the causal challenge as directed towards ethical intuitions in particular). To this extent, the epistemological credentials of ethical intuitions obtain relative to a given state of empirical knowledge of human nature and the social world. Second, it is possible to revise one's ethical judgements in light of inquiry if ethical intuitions fail to cohere with a better informed conception of their causal pedigree. In other words, we are not victims of a compulsive or fatalistic attachment to our ethical commitments. To this extent, the epistemological credentials of ethical intuitions may differ depending on the extent to which those intuitions themselves, as opposed to more considered ethical judgements, are revisable in response to reflection. As previously noted, intuitions that take the form of rationally or agentially independent ‘gut reactions’ could be more vulnerable in this respect. It is therefore crucial to bear in mind that our capacity for ethical judgement is neither completely exhausted, nor entirely controlled, by our capacity to have such reactions. What we think about some ethical issue is not always the first thing that springs to mind. Third, it is possible to apply a substantial and reflectively robust norm of coherence to evaluate the extent to which ethical intuitions are epistemologically credible in light of their causes and functional role. In other words, we have access to non-vacuous norms of practical and theoretical reasoning that are substantial enough to generate at least comparative evaluations of the epistemological credentials of a wide range of conflicting sets of ethical or non-ethical commitments. To this extent, the epistemological credentials of ethical intuitions are sensitive to epistemological considerations that extend far beyond the narrowly ethical. Thus, the considerations in question will include norms for how to estimate the comparative security of various ethical and non-ethical beliefs. (Given the broadly naturalistic explanatory framework assumed by standard formulations of the causal challenge, the basis for these comparisons is arguably not entirely symmetrical between the ethical and the non-ethical.)Footnote ²⁷

If each of these three conditions are met, the causal challenge can be confronted by testing for the reflective robustness of intuitions in response to evidence about their causes and functional role. Responding coherently to such evidence is one way of bringing ethical beliefs into what is known as ‘wide reflective equilibrium’.Footnote ²⁸ It is arguably true that as normally applied in contemporary ethical theory the idea of a reflective equilibrium is not always extended to include reflection on the wider network of non-ethical beliefs with which epistemologically credible intuitions ought to cohere. It does not follow that the underlying conception of a wide reflective equilibrium employed in this literature excludes the possibility of such an extension. On the contrary, it demonstrably requires it.Footnote ²⁹

Although some common ethical intuitions would not survive in a state of wide reflective equilibrium, other intuitions would. Thus, such widely shared ethical intuitions as that it is OK to care about the welfare of oneself and others, to seek the pleasures of genuine friendship and co-operative social relations, or to prevent the gratuitous suffering of other sentient beings are likely to survive the causal challenge in any plausible form it might take. It is a notable fact that ethical theories across a wide theoretical and historical spectrum, including any minimally plausible form of consequentialism, all claim to be consistent with a broad range of intuitions of this general form.Footnote ³⁰ There might be fewer pre-theoretical certainties about which among various more particular ethical intuitions would be reflectively robust when faced with the causal challenge. Even so, there is good evidence that a significant number of them (including a wide range of platitudes relating to practices of basic interpersonal care and reciprocity, or obvious prohibitions on gratuitous interpersonal harm and anti-social behaviour) are more credible than a vast range of absurd and preposterous claims with which they obviously conflict. There is such evidence in part (even if not only) because there is evidence that these intuitions cohere with beliefs that have proved themselves to be reflectively robust in response to causally informed reflection in a wide range of circumstances in the past. In light of this fact, it is a reasonable to assign presumptive credibility to a wide range of ethical intuitions, both general and particular, in the absence of evidence that these intuitions are guided by evaluative criteria that would fail to survive continued and causally informed reflection (as, indeed, contemporary critics of ethical intuitions have implicitly done when constructing experimental conditions that virtually everyone agrees have led their subjects astray). These intuitions are presumptively credible, not because they have some special property of ‘intuitiveness’ but because there is good evidence to believe that they would survive in a state of wide reflective equilibrium where their potential reliance on ethically irrelevant or otherwise discreditable criteria of evaluation would be exposed and washed out. Conversely, it is reasonable to assign comparatively less credibility to intuitions and other pre-theoretical beliefs to the extent that they depart from the domain of tried and tested ‘common knowledge’ by having as their object actual or counterfactual novelties or complexities about which our pre-theoretical sensibility cannot be antecedently assumed capable of reliably judging.Footnote ³¹ To the extent that recent critics of appeals to intuition in ethical theory have targeted their arguments at an overly complacent reliance on intuitions of this latter kind, their arguments are consistent with the argument of this paper. Where I depart from at least some of these critics is on the question whether there is a strong case for denying the presumptive epistemological credibility of ethical intuitions in general. I think there is not.

4. Three Objections

This cautious optimism on behalf of the presumptive credibility of some ethical intuitions is the potential target of several objections. Here I shall mention three. I claim that each objection is inconclusive.

The first objection has been made by Peter Singer, who writes that ‘the model of reflective equilibrium has always struck me as dubious’.Footnote ³² Singer's objection to the ‘model’ of reflective equilibrium can be formulated as a dilemma. On the first horn, the method of reflective equilibrium consists in producing a description of an existing ethical sensibility in terms of its explicit or implicit commitment to norms, principles or judgements about cases, mutually adjusted to exhibit explanatory coherence. Singer calls this the ‘narrow’ conception of reflective equilibrium. He objects that such a limited process of mutual adjustment between norms, principles and judgements about cases is too conservative, and therefore unable to critically vindicate ethical intuitions as presumptively credible in contrast to actual or possible judgements with which they conflict. On the second horn, the method of reflective equilibrium consists in producing an improved ethical sensibility by coherently integrating ethical intuitions within a maximally comprehensive network of ethical and non-ethical beliefs, including beliefs about the causes of ethical intuitions. Singer calls this the ‘wide’ conception of reflective equilibrium. He complains that the cost of interpreting the method of reflective equilibrium widely is explanatory emptiness. For understood this way there is nothing more to the so-called ‘method’ in question than the apparently trivial claim that a set of beliefs is epistemologically credible just in case it is reflectively robust. Yet this claim fails to specify any genuine procedure by means of which such a set can be generated in the course of inquiry. The idea of a distinctive ‘method’ of reflective equilibrium is therefore at best a misnomer and in any case quite useless for the purposes of substantial ethical inquiry.Footnote ³³

This is a false dilemma. With respect to the first horn, Singer may be right that ethical intuitions in a state of narrow reflective equilibrium are vulnerable to epistemic debunking in light of the causal challenge, and that the narrow conception is therefore unduly conservative. Yet nothing I have said on behalf of the presumptive credibility of some ethical intuitions implies the opposite. Furthermore, Singer's description of the narrow conception as unduly conservative underestimates the sense in which even reaching a state of narrow reflective equilibrium amounts to an epistemological achievement. Consider the case of two hypothetical agents, let's call them ‘Stubborn Believer’ and ‘Flaky Believer’. Stubborn Believer is someone whose intransigence in the face of reflection is such that he will never revise his ethical beliefs. Flaky Believer is someone whose flexibility in the face of reflection is such that he is disposed to abandon his ethical beliefs in the face of any potentially conflicting evidence whatever. Obviously, neither Stubborn Believer nor Flaky Believer present a serious philosophical challenge to the method of reflective equilibrium, even on the narrow conception. In order to decide whether an ethical intuition is reflectively robust it does not suffice to determine whether it would, in fact, remain in the light of inquiry. We also need to determine if retaining that belief in the light of inquiry is actually reasonable given the nature of the evidence and other beliefs. This normative condition applies regardless of whether this evidence and these other beliefs are broadly ethical (as on the narrow conception) or also non-ethical (as on the wide conception).

With respect to the second horn, Singer is mistaken if he assumes that the idea of a wide reflective equilibrium should be understood as a practice manual for ethical inquiry, as opposed to a general specification with a criterion of success. It is no objection to the ‘model’ of a wide reflective equilibrium that its acceptance fails to support one specific ‘method of doing normative ethics’ (e.g. consequentialism), as opposed to another (e.g. Kantian deontology). On the contrary, the choice between such ‘methods’ is itself partly a question of which of them (if any) would be favoured in a state of wide reflective equilibrium. Furthermore, the ‘method’ of reflective equilibrium as described in the previous section does include non-empty criteria for the evaluation of the reflective robustness of ethical intuitions in response to the causal challenge. This ‘method’ requires that ethical inquirers confront their existing network of ethical and non-ethical beliefs with empirical evidence regarding causally operative criteria of ethical discrimination in order to determine whether intuitive responses to norms and principles, or actual and hypothetical scenarios, are conditional on ill-informed or arbitrary prejudice, exposed as such by causally informed, but substantially ethical, reflection. Bracketing all-out ethical scepticism, the fact that a general specification of the ‘method’ of reflective equilibrium fails to provide an answer to this question a priori is no objection to that ‘method’ in the absence of independent evidence to believe that some alternative and equally reasonable short-cut is available. Finally, and pace Singer, the question at issue is not if the ‘method’ of wide reflective equilibrium would countenance the rejection of all our ‘ordinary beliefs’, but rather if this method provides an informative account of what it takes for ‘ordinary beliefs’ to be reflectively robust. Singer's dilemma fails to show that it does not.Footnote ³⁴

According to a second objection, the presumptive credibility of ethical intuitions is undermined by the fact that the causal explanation of their existence shows them to be epistemologically fickle.Footnote ³⁵ Although the point is not always made clear by proponents of this objection the fickleness in question can take at least two forms. The first kind of fickleness allegedly affects the credibility of ethical intuitions because we would still have these intuitions if they were mistaken. What explains why we would have these intuitions if they were mistaken is that we have them because they play certain functional roles in our social, psychological or biological economy, and not that we are epistemologically sensitive to their factual grounds. I shall refer to this kind of fickleness as ‘Flexibility Failure’. The second kind of fickleness allegedly affects the epistemological credentials of ethical intuitions because we would have had different intuitions if various non-ethical facts had been different, the ethically relevant facts remaining the same. What explains why intuitions would be different even with ethically relevant facts remaining the same is that our having those intuitions is contingent on their functional roles in our social, psychological or biological economy, regardless of our epistemological sensitivity to their factual grounds. I shall refer to this kind of fickleness as ‘Contingency Failure’. Both flexibility and contingency failure are forms of the same epistemological defect, namely the failure of judgements to bear a reliable connection to their factual grounds. I shall refer to this wider phenomenon of epistemological fickleness as ‘Tracking Failure’.

Both versions of this objection are inconclusive. With respect to inflexibility failure, it is not plausible that all ethical intuitions exhibit the kind of fickleness required to undermine their presumptive credibility. The reasons for this are familiar from the existing literature on moral explanations.Footnote ³⁶ Let us assume, for simplicity, that the tracking requirement is targeted at the relation between beliefs and their truth. First, suppose the truths in question are necessary (as at least some past philosophers thought the basic, or general, truths grasped by ethical intuitions are). If so, they could not have been different, so the counterfactual is empty, and the objection uninteresting. Second, suppose the truths in question are contingent (as at least some past philosophers thought the particular truths grasped by ethical intuitions are). If so, they could have been different. If they were, however, then so would at least some of the non-ethical facts upon which the truth of these beliefs depends. Consider a world in which the widely held belief that common theft is wrong is mistaken because the institution of property rights is socially non-adaptive in that world (as opposed to in the actual world, where I shall assume for the sake of argument that it is adaptive). Would we still believe that common theft is wrong in that world? Only if our belief that common theft is wrong in the actual world is counterfactually insensitive to the non-ethical truths on which its truth depends. Yet even if we can explain why we have the intuitions we do in terms of their functional roles in our social, psychological or biological economy, it does not follow that these intuitions are oblivious to the presence or absence of the non-ethical facts on which their truth depends. If they were, then reasonable ethical thought about the non-ethical grounds of ethical judgement would be impossible. But it is not. So the objector cannot assume that ethical intuitions are subject to inflexibility failure, even if the causal explanations of ethical intuitions targeted by the causal challenge are sound. True, these explanations may suggest that certain ethical intuitions are fallible in a particular way that is intimately related to their natural causes and functional role. Yet this fact (if it is a fact) does not thereby undermine their presumptive credibility. To think so is to confuse presumptive credibility with infallibility or truth.

With respect to contingency failure, the objection once more misses the target. First, the claim that we would have had very different beliefs if certain facts about our nature and circumstances had been different is true of all ethical and non-ethical beliefs alike. Thus, if we had been born blind we would have been unable to distinguish between the colours the way we do. And if we had been born with less sophisticated brains we would have been unable to grasp the truths of mathematics. Yet these facts do not constitute good evidence against the presumptive credibility of all intuitive judgements about colours, numbers or figures. Second, the fact that our ethical intuitions would have been different if our nature and circumstances had been different is actually an argument in favour of the presumptive credibility of those intuitions insofar as those differences in nature and circumstances would themselves make an ethical difference. Thus, a world in which there is no interesting causal connection between personal well-being and consistent parenting is one in which the absence of ethical intuitions in favour of existing family values would be neither surprising or disturbing, either ethically or epistemologically. In order for contingency failure to occur the envisaged counterfactual changes in ethical intuitions would have to bear no epistemologically sensitive connection to their factual grounds. Once more, however, the objector cannot assume that ethical intuitions are oblivious to the existence of the non-ethical facts on which their truth depends, even if the causal explanation of these intuitions show that they are fallible in away that is explicable with reference to their natural causes and functional role.

There is also a more general response to the claim that ethical intuitions are epistemologically fickle. This response applies equally to the possibilities of inflexibility and contingency failure. In subjecting ethical intuitions to reflective scrutiny we do so by starting with the intuitions we have, as given by our actual nature and circumstances. One way of testing the credibility of existing intuitions is to confront them with exactly the kind of counterfactual scenarios embodied in the arguments for contingency and flexibility failure just described. This way of testing the credibility of ethical intuitions is a natural extension of historical practice, in the long course of which moral inquirers in different times and places have acquired a significant degree of inductive support for a wide range of general and particular ethical intuitions. In testing intuitions this way, at least some ethical inquirers have to varying degrees been able to bring out some of the causal factors that explain their existence, and to reflectively evaluate the criteria of ethical significance they embody in light of their differential application to actual and counterfactual circumstances. Thus, reflective ethical agents have historically been able to extensively evaluate their intuitions about the ethics of property relations or familial obligations by considering how these intuitions either apply, or would have applied, to situations in which either they, or their circumstances, are significantly different in non-ethical respects.Footnote ³⁷ In being part of this process, these agents have implicitly been working towards a state of wide reflective equilibrium in which the intuitions with which they started have been abandoned, revised or retained in their original form. The fact that the intuitions that have gone into this process have a causal explanation in terms of their psychological, social or biological role does not preclude that the beliefs they have ended up with have been a genuine epistemological improvement on their predecessors. The fact that progress is uncertain and inconstant does not entail that it is impossible. Credibility does not entail infallibility, or even truth.

A third objection would maintain that the response I have sketched to the causal challenge is redundant because the challenge itself trades on confusing the explanatory causes of ethical intuitions with the justification of their content. The basic thought behind this objection is that the causes of coming to believe that so-and-so are, in most if not all interesting cases, irrelevant to the justification for believing that so-and-so. What matters for the justification for believing that so-and-so is the content of that belief and the evidence for and against it. Thus, even if the causal explanation of why I believe it is wrong to torture people for the sake of it is that I have been told that it is by a parent who is a compulsive liar, my justification for believing it is wrong to torture people for the sake of it is that gratuitous torture causes unbearable pain for no reason, and unbearable pain is intrinsically bad. As for ethical beliefs in general, so for ethical intuitions in particular. The question of their justification is not affected by the question of their causal origin. The idea that the presumptive credibility of ethical intuitions is undermined by the fact that they have causal explanations unrelated to their ethical content is therefore based on a basic confusion between reasons and causes.

This objection trades on an equivocation between two different interpretations of ‘ethical intuition’. True, the fact that an intuition was caused in a certain way need not imply anything about its objective justification, considered as a believed content. Thus, the causes of someone coming to believe a genuinely self-evident proposition do not affect the status of that intuition as genuinely self-evident and knowable as such to anyone who adequately understands it. Yet the fact that an intuition was caused in a certain way can have implications about a subject's justification in holding on to it, considered as a non-factive attitude with a believed content. Thus, the causes of my coming to believe a genuinely self-evident proposition may affect the epistemological credentials of that belief as held by me if the causal explanation of my coming to have it is evidence that I cannot have fully understood it. In general terms, facts about the causes of ethical intuitions, understood as non-factive attitudes with endorsed contents, can undermine the credentials of those intuitions in any case where a subject's epistemological state falls short of full, or otherwise adequate, information. Thus, unknown facts about the causes of someone's intuitions can be relevant to determine whether or not the evaluative criteria they embody are ones they would be able to coherently accept as ethically probative in light of further inquiry. Thus, the subjects described in Haidt's study described in Section 1, for example, might decide to modify their ethical beliefs about incest, for example, if causally informed reflection about the relevant intuitions is taken to suggest that the markers of wrongness in sexual intercourse are things like the risk of abuse, trauma or reproductive damage, as opposed to the mere fact of biological relatedness. (Given the statistical correlation between these markers in the actual world they might still, of course, disapprove of incest.) To simply deny this possibility as irrelevant to the epistemology of ethical intuitions is to claim for oneself the God-like status of self-conscious transparency and complete information on any ethical issue.

The causal pedigree of ethical beliefs can have positive as well as negative epistemological significance. First, the epistemological credentials of some widely shared ethical intuitions is supported by the fact that they have been passed down by individuals and communities that have repeatedly tried and tested them in moral practice and theoretical inquiry to the point that the platitudinous appearance of some of these intuitions may in fact serve to disguise rather than highlight their status as common ethical knowledge. Second, the fact that a belief was acquired by testimony from a trustworthy source is often described as a paradigmatic reason for thinking that this belief is presumptively credible, all else equal. This paper is not the place for a general discussion of the epistemology of testimony.Footnote ³⁸ It is worth noting, however, that one historical source of the idea that the causal challenge constitutes a sceptical problem for the epistemological credentials of ethical intuitions is the increasing lack of faith in the claim that they have their causal origin in the most trustworthy source of testimony imaginable, namely the mind of an all-knowing, all-powerful and benevolent God.Footnote ³⁹ One corollary of my argument in this paper is that this is an overly simplistic way of describing the epistemological alternatives.Footnote ⁴⁰

5. The Ethical Challenge

The challenge posed to ethical intuitions by their causal pedigree is not only epistemological. Facts about the causes of ethical intuitions can also be ethically significant in a more direct way. In this section I first give an account of what this challenge consists in. I then argue that getting clear about the nature of this challenge and the ways it can be coherently addressed serves to illuminate the epistemological challenge discussed in previous sections. The strategy is to construct an argument by analogy. The claim is not that either the epistemological challenge or the ethical challenge can somehow be reduced to, or be analyzed in terms of, the other. It is that the ethical aspects of the causal challenge can be shown to illuminate the epistemological aspects of that challenge.

Consider advertising. A persuasive advertiser aims to make prospective customers attracted to a product by associating its presentation with other objects of their conscious or unconscious attitudes, such as fear, admiration or desire. Thus, the famous ‘Marlboro Man’ (apparently one of the most successful advertising campaigns in history) associates a particular brand of tobacco with a certain lifestyle, involving a certain kind of rugged masculinity and closeness to nature. Neither association has an intimate connection with the intrinsic nature, effects, or comparative advantages of the product advertised (a brand of tobacco). Even so, exposure to these associations has obviously been thought to increase the likelihood of customers deciding to buy the product in part because of the way its advertising engages with attitudes that, although they have no intimate connection with the actual features of the product itself, are nevertheless strongly motivating for many people.

The existence of persuasive advertising raises at least three serious ethical challenges.Footnote ⁴¹ First, a potential customer's control of their own agency can be subverted by persuasive advertising that appeals to attitudes of which the potential customer is less than fully aware. In this way, persuasive advertising may negatively affect the extent to which the potential customer is able to exercise effective self-control. Second, the operative desires evoked by persuasive advertising can be ones that the potential customer either would not (or should not) on reflection be willing to endorse. In this way, persuasive advertising may negatively affect the quality of the potential customer's considered judgement as to whether the product advertised is one they should buy (in my example a highly addictive and toxic substance). Third, a potential customer exposed to persuasive advertising and unaware of its actual effects on their attitudes could struggle to make coherent sense of their own practical reasoning. In this way, persuasive advertising may negatively affect their practical rationality. In each of these ways persuasive advertising can undermine the self-knowledge and autonomy of anyone affected by it. To the extent that we think self-knowledge and autonomy are valuable features of persons, the existence of persuasive advertising is a potential obstacle to the pursuit or protection of those values. Thus, if our conception of a good life places a significant emphasis on some idea of self-constitution, for example, the mechanisms involved in persuasive advertising are relevant to our understanding of how, and to what extent, the activity of self-constitution is either possible, or ever actually realised, by creatures like us.Footnote ⁴² In a similar way, ignorance of these mechanisms will undermine the extent to which someone can be said to fully possess the virtue of self-knowledge. Because of this, causal facts about the mechanisms behind our purchasing decisions are a potential threat to the ethical and prudential credentials of those decisions, even if not in a narrowly epistemological way.

The mechanisms involved in persuasive advertising share a number of significant features with the mechanisms involved in the causation and retention of ethical intuitions. First, the effects of persuasive advertising are ubiquitous in a wide range of social transactions (such as public displays and product design). Second, persuasive advertising makes use of empirically knowable and partially controllable causal mechanisms (as targeted in controlled experiments and market research). Third, some of the effects of persuasive advertising can be controlled by means of discipline, regulation, and other forms of social and institutional design (such as restricting the range of legally advertised products or the range of potential customers exposed to it). Fourth, responding to persuasive advertising is compatible with informed and reflective ethical judgement (such as a considered belief that the product advertised is actually a bad thing). Fifth, the extent to which we judge the pactices of persuasive advertising as desirable overall depends on wider ethical considerations (such as our commitment to the values of self-knowledge and autonomy). It is clearly possible to be more or less optimistic regarding our general capacity to exercise these forms of informed and reflective judgement and control, either individually or collectively. There is no need to take any particular view on these issues here. What matters for present purposes is only the claim that people (collectively or individually) have some capacity to exercise the relevant kind of judgement and that the extent of this capacity is not trivial.

Each of the five features cited above has a close parallel in the case of the mechanisms responsible for the causation and retention of ethical intuitions. These mechanisms are also ubiquitous in a wide range of social as well as personal transactions. Yet they are also in principle knowable and partially controllable. Moreover, their effects can be resisted by means of regulation and other forms of individual, social or institutional design, such as legislation to combat widespread forms of ingrained and pernicious prejudice. Responding to the causes of ethical intuitions is therefore compatible with the application of informed and reflective ethical judgement. Finally, the extent to which we judge the effects of these mechanisms as desirable overall also depends on wider ethical considerations, which means that to judge the ethical credentials of ethical intuitions we must confront them with a causally informed and reflectively open network of ethical beliefs. Refusing to do so is ethically, as well as epistemologically, irresponsible. Once more, it is clearly possible to be more or less optimistic regarding our capacity to exercise these forms of informed and reflective judgement and control, either individually or collectively. Thus, in the case of some ethical intuitions (such as rationally or agentially independent gut-reactions) our ability to control their impact may require compensating mechanisms external to those intuitions themselves (whether intrapersonal or interpersonal) because these intuitions themselves are not readily malleable. In the case of other intuitions (such as considered judgements) it may be possible to control or change their impact through the process of ethical inquiry itself (whether intrapersonal or interpersonal). Once more, however, what matters for present purposes is the claim that people have some capacity to exercise the relevant kind of judgement and that the extent of this capacity is not trivial.

There are also obvious differences between persuasive advertising and the causes of ethical intuitions. First, the causes of ethical intuitions are not all the product of intentional agency in the way most aspects of advertising is. Second, a decision to purchase an item of merchandise does not necessarily involve making an ethical judgement in anything more than the broad sense of judging that something is ‘the thing to do’.Footnote ⁴³ It might be natural to think (although it is a subject of deep ethical controversy) that the transition from a personal decision whether to acquire an item of merchandise to an ethical judgement about the rightness or goodness of a practical option or state of affairs is a transition from a domain of presumptive permission to one of presumptive constraint. Third, it may be questioned whether the appearance of a conscious desire or intention to purchase a given product lends any kind of presumptive credibility to the claim that there is anything substantially evaluative to be said in favour of buying it. Each of these differences between persuasive advertising and the causal pedigree of ethical intuitions is irrelevant for the purposes of my argument here. What matters for present purposes is the following. First, the analogy with persuasive advertising brings out the way in which the causal challenge is as much an ethical problem as a narrowly epistemological one. It is ethically desirable to take an ethical view about the natural causes and functional roles of ethical intuitions and other commitments. Second, the example of persuasive advertising shows that there are established ways of getting both a theoretical and a practical handle on the ethical aspects of this challenge. Not everyone need fall helplessly into the arms of the Marlboro Man. Third, just as getting an ethical handle on the causal challenge can help to improve the quality of someone's purchasing decisions, so getting an ethical handle on this challenge can help to improve the epistemological credentials of someone's ethical commitments. The intellectual resources required to face the ethical aspects of the causal challenge substantially overlap with the intellectual resources required to face the epistemological aspects of that challenge. To this extent, thinking ethically about the causes of ethical intuitions can contribute to the project of better understanding their epistemology.Footnote ⁴⁴