Moral Constructivism: A Phenomenological Argument

Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 CARLA BAGNOLI MORAL CONSTRUCTIVISM: A PHENOMENOLOGICAL ARGUMENT1 Ethical judgments make claims about how we should live, and thus aim at guiding our actions and our attitudes, showing a specific kind of relation with our motivation. In this sense, they are normative and practical statements about what there is reason to do or to feel. There are standards of correctness for assessing whether these claims are legitimate or valid, and thus ethical judgments seem to aspire to objectivity. One might expect an adequate ethical theory to account for both the practicality and objectivity of ethical judgments. But many contemporary philosophers claim that these features to pull in opposite directions and thus they doubt that ethical theory should or could accommodate both. On this count, the very problem of contemporary ethics is the attempt at "reconciling" the apparent practicality and objectivity of ethical judgments.2 Why do practicality and objectivity appear to clash and resist reconciliation? The reason lies, I claim, in some tacit assumptions about the nature of ethical judgments and their aspirations. It is assumed that ethical judgments can legitimately aspire to objectivity only if there are values, moral properties and moral facts out there to be discovered. That is, it is assumed that ethical objectivity is vindicated only by ethical 1 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 realism. According to realism, ethical judgments are assertions about moral facts and moral properties independent of whether we could discover them. This thesis seems at odds with the claim that ethical judgments are practical: if values are independent of our capacity to arrive at them, it becomes mysterious how values could make claims on us, guide us, and motivate us to act. How is it that a moral view motivates us to act or shapes our beliefs and attitudes, if realism is true? In order to account for the apparent practicality and motivational force of ethical judgments, some philosophers have embraced non-cognitivism and maintained that ethical judgments are not descriptions of states of affairs or representations of properties, but expressions of non-cognitive mental states, such as decisions, desires, prescriptions, or attitudes. This view hardly makes sense of the claim that ethical judgments appear to be objective.3 The non-cognitivist strategy is to press the claim that ethical judgments do not represent anything because they are not assertions, and thus the question of objectivity does not arise for them. This debate shows that we reach an impasse when we try to accommodate the view that ethical judgments represent moral facts and that they express a motivational state. In order to overcome this meta-ethical deadlock, we are urged to give up either objectivity or practicality.4 But let us reconsider the two assumptions that are at the origin of this debate: objectivity is construed as the aspiration to correctly represent a moral reality, while practicality is construed as action-guidingness or the conative aspiration to motivate us to act. Once these assumptions are made explicit, it becomes apparent that the allegedly insoluble problem of contemporary ethics is less pervasive and more 2 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 modest than previously thought. The problem is that of reconciling a realist conception of objectivity with a non-cognitivist conception of practicality, and in this sense it affects only one specific philosophical project. My purpose in this paper is to challenge this account of the nature of ethical judgments, and to suggest that the reconciliation of their practicality and objectivity is not a genuine philosophical problem. The real issue is rather how to account for the fact that we consider our moral judgments important, authoritative, expressive of our moral personality and of our moral vision, and in this sense both objective and practical. The argument that supports this proposal is phenomenological. It starts with the phenomenological consideration that the agent experiences ethical judgments as normative: they are conceived and experienced as authoritative and as such that inform one's life in a special way. To take phenomenology at face value is to make the experience of the agent intelligible. My conclusion will be that neither realism nor non-cognitivism in their crudest versions engage in the search for the intelligibility of our moral experience, and for that reason they are not appealing alternatives. In their softer and sophisticated versions, both realism and non-cognitivism purport to explain moral phenomenology, but employ inadequate explanatory devices that seriously compromise their account. As I will argue, the phenomenological argument leads toward a constructivist account of ethical judgments. The conception of constructivism I advance is a form of cognitivist irrealism, thus opposed both to realism and non-cognitivism. 3 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 Here is how my argument will proceed. In section 1, I characterize the debate on ethical objectivity as exploiting the realist image of discovery and the non-cognitivist image of invention. With Wiggins I argue that these alternative metaphors fail to account for moral phenomenology. In section 2, I examine Simon Blackburn's non-cognitivist reply that utilizes the image of projection, and argue that while more sophisticated than the previous ones, this solution does not account for the authority of ethical judgments. In section 3, I take into consideration McDowell's analogy between values and secondary qualities and the metaphor of vision. My claim is that their argument succeeds in ruling out realism and non-cognitivism on phenomenological grounds, but it is insufficient as an account of values. My point is not simply that the analogy with perception is limited and incomplete, but that it is inadequate to account for the normative relations that the agent establishes with her surroundings. In section 4, I argue that that the appeal to phenomenology supports constructivism about normative relations and expose its features. Finally, in section 5, I show the merits of construction as an explanatory device. 1. Discovery or invention? A false dilemma. Having betrayed divine secrets to mortals, Sisyphus is condemned by the gods to roll a stone up the top of a hill; the stone will roll back down, and he will push it up again, forever. In "Truth, invention and the meaning of life", David Wiggins takes Sisyphus' myth as the image of a meaningless and absurd life, and considers the question of how meaning and significance can be restored to it. According to Wiggins, the realist says that life is (or is not) valuable quite independently of how Sisyphus experiences it. In fact, the 4 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 realist can say that Sisyphus's life has (or hasn't) value quite independently of whether it has meaning, that is, independently of whether it has a point or an end, and it makes sense for Sisyphus. But it sounds odd to say that the question of how Sisyphus perceives his life is irrelevant to the question whether his life is valuable. It seems that it should matter what Sisyphus thinks of his life, how he feels about it, and how he judges it. The agent's perception of whether his life is worth living should enter the account of whether it is. Otherwise, how could a completely external judgment be authoritative for Sisyphus? How is any external claim about value supposed to have a foothold in Sisyphus' character? According to Wiggins, the non-cognitivist may try and answer these questions by envisioning two modes of enrichment.5 First, Sisyphus might acquire a desire to roll up the stone. For example, he might learn to rejoice in assembling and arranging the stones on the top of the hill in a way that he finds beautiful or fascinating. By the adoption of the desire, the meaning of his life is restored: now his activity has a point. In a second scenario, the gods might take pity in Sisyphus and decide to implant in him an instinct to roll stones up to the top of a hill. Because Sisyphus' activity is intended to the satisfaction of his instinct, his life acquires meaning again. In fact, despite Wiggins' suggestion, these two strategies for restoring meaning and value to a valueless life are also available to a realist. For example, a realist naturalist might hold that our values are determined by our natural instincts and a valuable life coincides with a life in which all these instincts are properly adopted and satisfied. However, there is third strategy to restore value to life that is not available to the realist. 5 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 Sisyphus might embrace wholeheartedly his destiny and by a simple act of will his life will be made worthy.6 It is only this third alternative (that of producing or conferring value by an act of will) that the realists cannot undertake. That is because realists and non-cognitivists significantly disagree about the source of value. For the realist, values are a matter of discovery, for the anti-realist values are conferred by the agent's act of will or desire. Their disagreement does not necessarily concern the locus of value: they might concur that value is situated in the satisfaction of some desires. In holding that desires and instincts provide meaning to life in being the locus of value, both realists and non-cognitivists commit to an instrumentalist view of rationality. That is, since ends and values are given, set by nature or by gods' will, we only reason about the means that help us toward given ends. Further issues arise as to the extent to which we can reason about given ends, and whether we only discover which ends we have by nature or we also endorse them or specify them, or select them in some significant way. These issues require further determination and introduce important qualifications in the instrumentalist framework, but these qualifications will not substantially modify the central claim that ends are given.7 Although sharing the common instrumentalism bond, realism and noncognitivism grow apart in their account of the source of value. While on the realist view value originate in the recognition of a natural fact, on the non-cognitivist view values originate in some conative states of the agent (may those be passions, desires, or acts of will). The non-cognitivist conceives desires, passions, feelings and acts of will as states that arise independently of the objects to which they are directed. That is, they are neither 6 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 percipients nor determinants of perception. This claim is of importance because it amounts to denying that conative states are states of awareness that can be held as appropriate or inappropriate. As we will fully appreciate in section 3, the implication of this view is that to master an ethical concept is not to organize a mode of awareness.8 If for the realist value is a natural fact, for the non-cognitivist view the question arises as to "how value gets attached to the world of facts". The paradigmatic answers that non-cognitivism offers to the question of how a life can acquire value are in fact answers to the question as to the locus of value in a world of facts. The assumption is that value is injected into a neutral world by an act of will (e.g., the gods' decree) or by a desire (e.g., for beauty). While the facts are provided neutrally by intellect and experience, the ends are supplied by our feelings or will. Once we disengage our wills and our feelings, and that is "objectively", life is meaningless, the world emptied of all value. Meaning and value are attached to the world in this spurious way, in virtue of a mind with a tendency to spread onto it. While infectious, this tendency does not provide our subjective craving for objectivity with any ulterior reality.9 Subjective reasons are the only reasons available to us. Wiggins argues that the three claims I have just mentioned, instrumentalism, the conception of action-guidingness in terms of conative and subjective states, and the facts/value distinction, are combined in a way that undercuts the perspective of the agent. Because of its emphasis on conative and subjective states, the non-cognitivist account depends on the availability of an "inner view", that is, of the subjective perspective of how things strike us. Consequently, non-cognitivism has to admit of a distinction 7 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 between the outer and the inner: Things as they are differ from how they seem to be to us. There is a significant discrepancy between the deliverances of the inner and the outer views as to the status of values.10 Phenomenologically, that is, from the perspective of the agent's experience, as things appear to be in the inner view, there is a difference between a meaningful and a meaningless life and that bears on the question of whether such a life is valuable. The outer view insists that such a difference is invented, created by a need or a desire, decided by a decree, and thus fictitious. Since phenomenologically there are objective values, the non-cognitivist account of values as matter of creation or decision necessarily discounts the inner perspective as deceptive. In order to avoid the contradictory thought that life is and isn't meaningful, the inner view has to be unaware of the outer view. But this only means that the inner view enjoys an illusory kind of objectivity. 2. Earning truth: the metaphor of projection It might plausibly be objected that to claim that values are produced or invented does not necessarily imply that they are illusory. One may agree with Simon Blackburn that the values that are the outcome of an objectifying projection, which is sufficiently robust to underwrite an innocuous form of realism: quasi-realism. This echoes Hume's claim that while reason is inert, cool and disengaged, and intent at discovering what is already there, without adding or diminishing anything, our sensibility is "a productive faculty, and gilding and staining all natural objects, it raises in a manner of a new creation". 11 8 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 On this view, the image of projection encourages us to accept the right kind of realism, that is, a realism that does not presuppose the reality of values independently of the mind's tendency to spread itself onto the world, and has thus earned its right to speak of truth. Quasi-realism is meant as a reply to the charge that projectivism makes moral phenomenology unintelligible. In other words, quasi-realism (as a meta-ethical hypothesis) is designed to make projectivism (as an explanation of our practices) more attractive. Blackburn claims that projectivism succeeds in explaining how we can legitimately talk of moral truths, even if realism is false and there is no moral ontology. But the very issue is "whether quasi-realism is successful in explaining why we can permit ourselves the linguistic expressions, and the thoughts that enable us to express, if projectivism is true."12 Blackburn remarks that there are independent reasons for advocating projectivism: its ontological economy (as it allows for no ontological proliferation besides what is explanatorily necessary), its rejection of queer properties, and its congruence with a plausible conception of motivation.13 But I am only interested in the claim that quasi-realism is designed "to protect the appearance of morality: to urge that there is no error in our ordinary thought and our ordinary commitments and passions. This enterprise will interest a projectivist most, because it defends him against the most forceful attack he faces, which is that he cannot accommodate the rich phenomena of the moral life."14 Projectivism replaces objectivity with of objectification, but it insists that there is a significant difference among patterns of objectification. The relevant distinction is not between subjective and objective, but between appropriate and inappropriate patterns of 9 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 objectification. By revealing the mechanism of projection the outer view corrects a mistake of the inner view: the error of thinking of values as properties that some things have intrinsically or absolutely, independently of their relation to us, i.e. as properties of which our feelings or will constitute a kind of perception. That is, projectivism explains why moral phenomenology seems to encourage a straightforwardly realist metaphysics, and thus dispose of a powerful argument in the hands of realists.15 Moreover, quasirealism has the advantage of allowing for a reality of values that is not merely presupposed, but constructed out of a projection: Values are the children of our sentiments in the sense that the full explanation of what we do when we moralize cites only the natural properties of things and natural reactions to them. But they are not the children of our sentiments in the sense that were our sentiments to vanish, moral truths would alter as well. The way in which we gild or stain the world with the colours borrowed from internal sentiment gives our creation its own life, and its own dependence on facts.16 Ultimately, our evaluations are the upshot of our sensibilities, and yet they entitle us to make a legitimate use of the notion of truth. Despite their subjective origin, ethical judgments can be taken as expressing moral truths, which are neither merely subjective nor (always) mind-dependent.17 Because the creation of value starts as a subjective projection, it might seem rather dubious that quasi-realism is capable of gaining the kind of credibility that 10 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 straightforward realism has since it is not firmly anchored to anything outside the agent. Quasi-realism treats moral language as mimicking assertoric language, so as to exploit all the advantages of realism. "But on what ontological grounds?" the realist may rebut. Put in this way, the question bears on the kind of ontological support this model derives from the reference to "our sentiments", and thus it begs the question about realism. However, there is a way to object to Blackburn's projectivism that is not questionbegging. On Blackburn's view, we cannot say that were moral sensibilities to disappear, moral truths would vanish as well. But such moral truths originate in our sensibility, and acquire a further reality and autonomy. It is this autonomy that is hard to explain in terms of the metaphor of projection. That is, the issue at stake seems to me to be the authority of values created out of projected sentiments. Why should one recognize authority to the deliverances of a subjective projection, or even to the productive and creative power of sensibility? McDowell suggests that this problem arises because projectivism mistakes the order of explanation.18 By endorsing the metaphor of projection, quasi-realism assumes that since the world contains no such features, then our subjective responses to the world have to be explanatorily prior: values are the children of our sentiments. Alternatively, McDowell argues that "an adequate account of these matters will have to treat psychological states and their objects as equal and reciprocal partners".19 The elaboration of this account exploits the analogy with secondary qualities. In the next section, I will consider whether this is a significant improvement over Blackburn's proposal. 11 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 3. Values as secondary qualities: the metaphor of vision It is not obvious that the appeal to an analogy with secondary qualities is a promising strategy to restore moral phenomenology to its central place. This analogy has been traditionally used to support two competing conceptions of values. On a realist reading of the analogy, our moral sensibility is perceptive of a peculiar reality. We have the capacity to discern value in the world as well as the capacity for recognizing something as true or false. As McDowell recounts, this conception encountered a crisis when it became customary to distinguish between cognitive and conative elements of the evaluation.20 The alternative non-cognitivist view defended by J.L. Mackie endorses John Locke's thesis that secondary qualities enable the objects to produce some sensations in us, but do not literally belong to objects, and thus are illusory.21 Thus, contrary to the previous conception, on this one the analogy does not point at a perception of a peculiar reality. Rather, the analogy indicates a common mistake that we systematically commit when attributing ethical and perceptual concepts: that of presupposing that the experience of valuing or perceiving is like the conceptual operation of discerning some properties of the world, whereas values and secondary qualities are not genuine properties. The only genuine properties are primary properties constitutive of the fabric of reality. Value enters the world via the reflection of an attitude or the expression of a feeling, but these affective states are completely detachable from the true representation of reality. The experience of evaluating and perceiving is therefore illusory and deceptive, and it should be explained away by embracing error theory.22 Error theory reduces the apparent 12 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 objectivity of value properties to some patterns of objectification, discounts the deliverances of the inner view as not veridical, and thus fails to properly account for moral phenomenology.23 It appears that these two traditional readings of the analogy with secondary qualities misrepresent the experience of valuing.24 For McDowell this is because they rest on two false presumptions: that ethical concepts can be decomposed into cognitive and affective elements, and that the subjective is opposed to the objective. On both issues, McDowell's critique of this view echoes Iris Murdoch's complaints against the conception of concepts as rings laid onto portions of reality, and similarly suggests that we need an alternative view of moral concepts as deeper moral configurations.25 But contrary to Murdoch, McDowell's argument starts as a rehearsal of Wittgenstein's remarks about extending a numerical series.26 Mastering a concept, and formulating judgments in a way that counts as applying the same concepts to different objects, are conceptual operations that are internal to a practice. There is a strong temptation to describe this phenomenon as that of "following a rule", where rules are conceived as rails for us to follow, independently of our reactions and responses. The presumption is that ethical concepts correspond to classifications that are intelligible from outside the evaluative perspective to which they belong. Contrary to this claim, McDowell holds that it is not possible to properly understand and master an ethical concept unless we are acquainted with the practices to which the concept belongs and equipped with the relevant propensities and sensibilities.27 Moral knowledge is grounded on shared practices, shared judgments about particular 13 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 cases, shared practices of interpretation and application of concepts, shared propensities and sensibilities. In emphasizing the importance of background practices and sensibilities, this account propounds an alternative explanation of the nature of the expectations and reactions that we have when we use ethical concepts.28 Rather than being vindicated by the presumption of objective codified rules, these expectations and reactions are justified insofar as we are endowed with specific kinds of sensibilities shaped by common practices. Our sensibilities are not perceptive of a peculiar reality of sui generis qualities, and yet they deliver veridical and meaningful experiences. The alternative interpretation of the analogy is meant to underline, first, that ethical concepts cannot be adequately understood "otherwise than in terms of dispositions to give rise to subjective states", and yet they are genuine properties.29 Second, because ethical concepts are dependent on specific affective responses to a given situation, to master an ethical concept is to acquire a specific set of classifications and sensibility. The analogy shows how to overcome the false distinctions between the subjective and the objective, the affective and the cognitive. As judgments about colors state a certain response (that is, a visual experience) to some property in the object that we detect is appropriate, given a certain sensory equipment, somewhat similarly, we can learn to see the world in terms of some specific set of evaluative classifications, aesthetic or moral, only because our affective and attitudinative propensities are such that we can be brought to care in appropriate ways about 14 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 the things we learn to see as collected together by classifications.30 As McDowell acknowledges, the analogy breaks down at some point, and it is important to see at which point exactly.31 A morally relevant feature of an object does not merely induce us to value it, but also merits this appropriate response from us. The appropriateness of our moral vision depends on getting picture right, that is, on paying attention to what is salient and deserves our attention and concern. But how are the relations of salience established? How can we distinguish between traits that appear to be salient to us, and traits of the world that are salient and thus deserve our attention? The analogy cannot help answering this question. We gather that salience is not to be captured by following the rule, but presupposes a web of practices and certain sensibilities. Notice that the problem is not to have some standards of salience, rather the question is how to get salience aright. The responses of our sensibility are constrained by how the world is featured. The suggestion is that there are more or less refined ways of conceiving of and responding to the world. How are we supposed to trace this distinction? Absent this specification, the normative notions of merit and salience remain obscure. This is not problematic because we were expecting the analogy to support a full analysis of value concepts: we had no such expectation.32 Rather, the account is problematic because it points at a normative relation between the agent and her world that calls for an explanation.33 15 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 It could be suggested that the analogy must be developed in a rather realistic way so as to support the metaphor of vision.34 For Murdoch, the metaphor of vision is designed to overcome the distinction between the conative and the cognitive by redefining the scope of moral activity. Moral activity is a specific kind of cognition but is also characterized by a specific kind of motivation. Therefore, a moral vision is not something that can be put in purely propositional terms, and thus no further question arises as to whether the wise's affective sensibilities (or, alternatively, the wise's will) will conform to the world.35 To have practical knowledge is to see salience aright and to be appropriately motivated, or at least to be guaranteed that one is in principle suitably motivated (that is, if no disturbances interfere).36 The metaphor of vision succeeds in vindicating the claim that in moralizing we are cognizing the world, and in cognizing the world we are orienting ourselves toward it in a certain manner. For this reason, it represents an interesting complication in comparison with the simpler metaphors of invention and discovery. However, if it succeeds in overcoming the distinction between the conative and the cognitive, and the distinction between objective and subjective, this metaphor still leaves the question of salience rather mysterious. What exactly does the agent contribute to reality when she responds to salience? How does she arrive at considerations about moral salience? Why does she find these considerations binding? What does it constrain her findings? One might reply that McDowell's account of practical wisdom answers these questions by referring substantially to practices. After all, they are supposed to play a crucial role in accounting for how we learn moral distinctions and acquire moral 16 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 motivation. The constraints that bear on the appropriateness of some behavior are internal to specific upbringings and webs of practices. But exactly at this point the reference to practices becomes problematic. If we take these practices to be grounding moral claims, then we face two unwelcome options. According to one alternative, each practice legitimates its moral distinctions and specifies its own moral reasons. This is to embrace a form of relativism, which cannot explain how evaluators become aware, together with the acquisition of new moral concepts, that some moral conceptions are better than others, and not just different. According to the other alternative, some webs of practices are just "objectively" better than others, as if from an Archimedean standpoint. This is to embrace a form of absolutism that rests on no justification. Neither of these alternatives seems to me appealing, and I believe neither of them is the path chosen by those who support the response-dependence account.37 The reliance on webs of practices is not a threat to objectivity. The practices and the forms of life establish the very possibility of talking of values. Practices are the background against which questions of values arise: without such practices, questions about values would not arise. This is not to reduce the objectivity of values to the actuality of agreement. Agreement in practices does not have any foundational or justificatory role. Practices are the necessary pre-conditions for having values. But if we stay with McDowell and refer to practices merely as the background against which moral distinctions are comprehensible, rather than as grounding and determining their content, we do not gain much sense of the constraints on the appropriateness of our practical responses. 17 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 4. Moral constructivism: an outline I have argued that old-fashioned realism and non-cognitivism do not take moral phenomenology seriously, and that more recent and sophisticated theories, while attempting to make sense of moral phenomenology, fail to do so. I suspect that this is because they have attempted to resolve the wrong philosophical question. Moral phenomenology suggests that ethical judgments are both practical and objective, these philosophers say, and then they take at heart the problem of reconciling these two allegedly incompatible features. It seems to me however that phenomenology suggests us that we examine another phenomenon. Ethical judgments are practical normative claims, and that is to say that they state some reasons that are important, authoritative, and practical in that they direct us in leading our life by suggesting that some specific response is now appropriate. The phenomenon we have to explain is that we consider our moral judgments important, authoritative, indicative of our moral personality and of our moral vision, and in this sense both objective and practical. The action-guidingness of ethical judgments is a byproduct of their being claims that are so authoritative and important that inform the agent's conception of her possibilities and of her options. But it is misleading to treat practicality as simply the capacity to provide reasons for actions. Equally misleading is to treat action-guidingness as the feature of ethical judgments that make them "practical". Practical responses are not necessarily actions; they could be attitudes. Ethical judgments are practical not (only) because they guide our single actions but because they inform our conception of the kind of agents we are and of the life that we should live. For the very same reason, objectivity is not to be 18 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 understood as the realists do, in terms of truthfully representing a special sector of reality. Rather, ethical judgments are objective because they enjoy a special kind of authority: they importantly make claims on the kind of agents we are, and because of this they are inescapable. The philosophical issue is how to make this phenomenon intelligible. To take phenomenology at face value is not to argue that there are values because the agent feels or thinks so. Thus, the argument from phenomenology does not derive the reality of values from the experience of the agent. To the contrary, it aims at explaining the features of ethical judgments in a way that does not make them alien and unintelligible to the agent herself, that is, in a way that coheres with the ordinary experience of the agent. My present aim is to argue that in order to make sense of the phenomenology of the authority and importance, objectivity and normativity of ethical judgments we should pay attention to the ways in which the agent establishes normative connections with her surroundings. The question is not: how does value get attached to the world? The question is, rather, how the agent establishes normative relations with her surrounding and acquires moral reasons. And this is also the correct way to reconsider how Sisyphus could restore meaning and value to his life. My contention is that the metaphor of construction helps us to frame these questions correctly. Construction is an explanatory device that specifies three elements: the basis of construction, its object, and its methods and constraints.38 Since there may be significant disagreements concerning each of these elements, the metaphor of construction is likely to generate different kinds of constructivism. I am not interested in 19 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 investigating the differences among kinds of constructivism, but only in presenting an outline of a specific form of constructivism. First, constructivism specifies a basis of construction. The basis of construction is an assumption about the kinds of agents that we are. Kantian constructivism starts with the assumption that we are free and equal agents. Humean constructivism starts with the assumption that we are endowed with a specific kind of sensibility. My assumption is minimal: located in a finite world, the agent necessarily establishes normative relations with it. That is, she searches for reasons for believing, acting, and feeling. To be an agent is to engage in such a search. This is not the object of construction, but rather the starting point: constructing does not start from nothing. Second, the metaphor of construction specifies its object. I take the object of constructions to be ethical judgments about what there is reason to do or to feel, that is, about what kind of practical response is appropriate for us to give. These judgments are specific kinds of relations that I call normative insofar as they state reasons. It is my contention that these normative relations are constructed by the agent in the deliberative process through which she takes into account the constraints that bear on her situation. To say that ethical judgments are normative propositions is to say that they state that a certain practical response (an action or an attitude) is appropriate (demanded, required, prohibited, permitted, advisable, etc). I take the concept of reason to be a primitive normative concept: it is a consideration that counts in favor of something. I certainly do not mean to say that the contents of ethical reasons are taken as primitive. In fact, constructivism holds that the contents of reasons and their normative force are 20 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 determined by deliberation. What I mean is simply that I take the notion of reason as a consideration that counts in favor of something as a basic concept from which other kinds of normative relations and normative concepts can be derived.39 Reasons vary according to their normative status, that is, according to the position they hold and the relations they entertain with other reasons. For example, reasons might be prima facie, all things considered, overridden, overriding, non-overriding and non-overridden, canceled, silenced, etc. To defend constructivism about reasons is to suggest that these relations do not stand prior and independently of deliberation, but must be accounted for in terms of deliberation. Reasons also vary in their content, although there are limits to what can be regarded as a reason. The content of reasons and their normative strength depend on the deliberative context. They are object of construction. Ethical judgments invoke reasons of justification: by committing oneself to an ethical judgment one is committing oneself to presenting reasons for holding such a judgment. Reasons of justification are usually implicit in proffering a judgment. However, in contexts of perplexity, critical reflection or dialogue, such reasons of justification are made explicit and the evaluator might review the supporting reasons that were advanced in making evaluative judgments. The reasons for holding ethical judgments are also objects of deliberation. That is not to say that the content of reasons vary according to deliberative contexts. When we proffer an ethical judgment, we commit ourselves to hold reasons that justify such a judgment in relevantly similar contexts and we expect such reasons to be shared. The possibility of sharing reasons and thus the possibility and the expectation of convergence does not depend on having 21 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 adopted a realist conception of truth, but on having formulated a judgment based on reasons. Constructivism is thus a theory about the content of reasons, that is, the normative relations that we establish with the world, and about the way such normative relations are reciprocally related, organized, and supported. Finally, the construction is an activity that is regulated and governed by principles. There are methods for constructing and constraints that bear on construction. The methods of constructing are not themselves constructed, rather they are simply laid out. To say that there are methods for deliberating is not to say that there is an algorithm or that there is a complete decision procedure capable of determining for each deliberative situation what there is reason to do. For example, Kantian constructivism takes the categorical imperative to represent the method for moral reasoning, which is neither an algorithm nor simply a decision-procedure. Rather, the categorical imperative is a mode of reflection and representation of ourselves as moral agents that is articulated according to three complementary formulations, which are simply laid out.40 Different kinds of constructivism allow for different methods for moral deliberation; the individuation of the method of construction is clearly a significant source of disagreement among constructivists. Figuring out how to resolve the moral problem does not merely consist in removing obstacles along the way, but in considering and appreciating the nature of constraints. My contention is that an account of the methods available should start with an account of the constraints that bear on deliberation. This is motivated by the thought that normative determinacy (of the kind reached by the implementation of an algorithm 22 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 or a complete decision-procedure) is not a requirement of adequacy for ethical theory.41 I will simply outline a sketch of the constraints that bear on construction. In constructing normative relations, the agent has to deal with the significant traits of her surroundings. First of all, we should consider that constraints are not merely "limitations" to overcome, nor are they imputable to the agent as defects to correct.42 Recall that as the basis of construction, I take a consideration about human agency. Human agency is necessarily constrained and conditioned, finite and historical, and to this extent, such that it develops in time and is determined through time. Such a constrained agency is not a mere defective instance of practical agency: quite the contrary, it is what constitutes human agency. Constraints are either dependent or independent of the agent's deliberation. I want to emphasize that to hold that there are constraints that depend on the agent's deliberation is not to say that there are constraints that are "up to the agents". Constraints are never up to the agent. More generally, the method for deliberating is never up to the agent, but simply laid down. The distinction I am to draw concerns the source of constraints. Some constraints are independent of the agent's deliberation and arise because of the kind of agents we are, that is, finite, limited, and active in time. This kind of constraints is specified by what we take the basis of construction to be. The basis of construction is not only the starting point of the activity of construction, it also importantly frames and structures constructions. Some of these constraints are epistemic: the agent has limited information available, limited resources for acquiring and for processing such information. Another instance of this kind of constraint is that the agent has to act in time, she is a temporal agent, and her activity is diachronic. Finally, a 23 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 significant trait of the agent's context is that her world is inhabited by other interested, needy and concerned beings that make claims on her. The normative relations that the agent establishes with her world are partly determined by the relations she establishes with other agents. That is, other agents put some important constraints on what she can take as a reason, that is, on the normative relation she establishes with her surroundings. Because the activity of evaluating is so intimately connected with being an agent of the kind of we are, we should not imagine that we face isolated, disconnected acts of decision. Rather, deliberation is continuous and cumulative. We deliberate now partly on the basis of what we have deliberated about before. That is, our previous activities of deliberation importantly bear on the options that we now recognize. Qua agents, we establish normative relations with the world and we take some features of the world as significant. These normative relations are binding in any similar context. Thus, there must be constraints on how we deliberate about now that derive from and depend on past deliberations. In this specific sense, there are constraints that depend on our deliberation. Past deliberation constrains the options and informs the vision of the world that we have now. This is particularly important to make sense not only of the agent's sense of identity and unity over time, but also of her experience of ethical judgments as authoritative and inescapable claims, which is the phenomenon we sought to explain. 5. The Merits of Construction as an Explanatory Device The realist metaphor of discovery seems to vindicate the feature of objectivity, but it does not help us explain why ethical judgments are normative and authoritative; this is 24 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 because it discounts the activity of the agent. The non-cognitivist metaphors of invention and projection suffer from the opposite problem: they acknowledge no external constraints on the agent's activity through which values are produced. While these metaphors make sense of the feature of action-guidingness of ethical judgments in so far as they emphasize how values are the result of the agent's production, they cannot adequately make sense of their objectivity and also cannot justify why they are taken as authoritative by the agent. They also cannot make sense of the practicality of ethical judgments in a more complicated and broader sense than simple action-guidingness. The analogy with secondary qualities does not illuminate the normative relation between ethical properties and the agent's appropriate response. My contention is that the metaphor of construction is an improvement over all these other explanatory devices. In section 4, I have indicated that moral phenomenology shows ethical judgments to be authoritative, perceived as important to our identity and integrity. In this section, I argue that constructivism offer the best account of this phenomenon. Contrary to the realist image of discovery, the image of construction suggests that these normative relations are the products of the agent's activity. In envisioning and constructing a situation, we are active: we participate in making relevant connections, and in organizing the material in a certain manner. Most importantly, the construction is constrained and constraints are not up to the agent. The discussion of constraints on deliberation that derive from previous deliberation shows that constructivism can make sense of deliberative continuity. It can explain the notion of deliberative history, which identifies an agent that displays a specific kind of personality. Most importantly, these 25 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 kind of constraints show that the products of our deliberation have authority over us and impose themselves on our attention when we are about to deliberate further. That is true also when we cannot respond to their claims. This aspect of authority makes sense of an important yet disregarded meaning of ethical objectivity. When we say that some value is objective we often mean that it has a special kind of importance, that we cannot disregard it, or that by disregarding it, we subject ourselves to a special kind of sanction (a feeling of guilt, regret or remorse). To suggest that moral judgments make claims on us, and that we cannot easily turn away is to suggest that morality has authority over us. The authority of judgments depends on the manner of their justification. Insofar as an ethical judgment is adequately justified, one can expect it to be objective, that is, inescapable and such that it maintains binding force over time. An ethical judgment cannot be violated, without such a violation having important consequences for the integrity of the agent or for her moral outlook. To acknowledge constraints on deliberation that depend on our past deliberation is to acknowledge that the products of our deliberation remain important and authoritative over time, and influence future deliberations. They are objective in this sense. This is not to deny that in some cases deliberation shows that what we considered a reason in a previous deliberative context is overridden by another reason that was lacking in previous contexts. Relations of overridingness are normative relations that are established by deliberation. The new ordering and ranking of reasons, the way reasons are arranged and related is the product of deliberation. 26 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 The image of construction rules out a traditional conception of practical reasoning in which the agent confronts "the situation" or "the problem", she surveys the relevant facts, runs her deliberative procedure, and then draws her conclusion. Alternatively, the image of construction suggests that "the situation" is not merely given, standing before the agent, waiting for her to make up her mind or to simply discover the right solution to her problem. Rather, through the appreciation of multiple constraints, the agent constructs the situation as such, and figures the problem she has. The evaluative characterization of the situation and of the problem that the agent faces is also object of deliberation. The configuration of a situation is constructed together with the contents of reasons, their reciprocal relations, and their normative strength. Reasons are qualified, refined, articulated and revised through deliberation. This novel account of the evaluative characterization of the situation importantly bears on the issue of the fact/value distinction. To say that deliberation starts with the evaluative construction of the problem at hands is to say that it starts with producing an account of the relation between facts and values. Both gaps and relations between facts and values are determined by deliberation. The distinction between facts and values is not logical, but hermeneutic: it depends on the agent's activity of deliberation, and of how she relates to her surroundings. Similarly, the relations between facts and values (such as supervenience) are object of deliberation and established through the agent's activity. Thus, on my view the continuous activity of deliberation determines not only the nuances and the texture of the world, but also its bounds. 27 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 Questions of value arise in a social context, starting from specific needs, interests and concerns that are mediated, shaped, elaborated and defined across time in the dialogical exchange with other concerned, needy and interested beings. The image of constrained construction emphasizes the dialogical dimension of those normative relations that are called evaluations.43 That is another important and distinctive feature of constructivism. The image of a constrained construction highlights the diachronic dimension of our evaluations. Mastering ethical concepts is a capacity that we acquire and exercise in time. Evaluating is a practical capacity that requires education and reflection, appreciation of the salience of constraints that bear on the relations we establish with the world and all that it contains. Most meta-ethical accounts disregard the diachronic dimension of value, and focus exclusively on the perspective of a mature moral being. Constructivism is well equipped to address the question of how we develop into a moral being. More generally, the metaphor of construction well illustrates what it is to undergo a moral change. Because it encourages a complex view of how we develop into moral agents, the image of construction supports a more credible conception of moral rationality and moral knowledge. Moral rationality is the capacity of appreciating and interpreting multiple constraints that bearing on deliberation. Moral knowledge is not tantamount to the availability of a certain quantity of true propositions on what count as an appropriate ethical response, but it consists in adequately conceptualizing, comprehending, and structuring a situation from a moral point of view. 28 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 Finally, constructivism is capable of accounting for moral phenomenology without admitting any special ontology. Insofar as there is a correct way of applying moral concepts, and there is an adequate process of justification, there are regular moral truths. However, the availability of justification is within the bounds of morality: it pertains to the moral domain, and it is a moral activity, that is, deliberation. In this sense, the kind of constructivism that I have outlined is a form of cognitivist irrealism. On this view, ethical judgments establish some normative relations between the agent and the world. They are normative propositions about what there is reason for an agent to do or to feel. These normative propositions are truth-apt, and their conditions of truth and falsity are specified in terms of justification. To say that an ethical judgment is true is neither to say that there are some facts that make it true, nor that there is a portion of reality to which ethical judgments correspond. That is to say, ethical judgments have propositional content, but not factual content. For this reason, the claim about the truthaptness of ethical judgments does not commit one to realism. Most importantly, because this claim about truth-aptness is not the claim that there are facts that make moral judgments true, quasi-realism is beside the point. The defense of irrealist cognitivism is not the task I have undertaken here.44 Rather, I wanted to show that this meta-ethical view takes a specific form when it is combined with a certain hypothesis about the content of moral reasons. I defend constructivism as the best hypothesis about how the contents of such reasons are established. 29 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 It can be objected that the features I have just discussed as peculiar and distinctive to the metaphor of construction are in fact carried by other kinds of metaphors.45 For example, the metaphor of discovery might well be defended as a form of activity. Similarly, one could question that projection and invention are free from constraints. These criticisms are quite reasonable, but as I shall explain, they do not represent objections to my main claim. The metaphor of discovery is to be found in the realist tradition. There is nothing in the metaphor of discovery that prevents us from thinking that a discovering agent is, in some sense, an active agent. The question is how to understand the notion of activity. If the agent contributes nothing to the discovery of truth, if by discovering truths it is meant that there is an objective order of values that is independent of the presence and of the possibility of such agents, then this is not an interesting conception of activity. If, on the other hand, the discovery of values is not independent of the mind of the agents, in the way Wiggins and McDowell understand this claim, then the metaphor of discovery is brought to a non-realist camp. My claim is that it would be more suitable to call it construction, in order to emphasize the contribution of the agent. Something analogous can be said about the metaphor of invention. It is traditionally used by non-cognitivists, in order to stress that the agent's subjective states are the only source of value. But nothing in the metaphor of invention itself prevents us from claiming that any invention is subject to specific constraints, some of which might be independent of the agent. If we do read the metaphor of invention in this way, we have 30 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 abandoned the non-cognitivist camp. My claim is that it would be more appropriate to call it construction. I find comforting that despite his preference for the metaphor of projection, Blackburn helps himself to the metaphor of construction exactly when he argues about the right to use truth in ethics.46 My qualm about his conception is that he does not consider the constraints on the contents of moral reasons as originating from the appreciation of external traits of the world, but claims that they emerge as a creation of our sentiments. That is, our disagreement concerns the nature of the constraints that bear on the construction of truth. This shows that Blackburn's projectivism and other forms of constructivism differ only because of their respective claims about the nature of constraints they acknowledge. They do not differ as to the kind of representational device that they use. Finally, I would argue that the metaphor of construction is the necessary completion of the metaphor of vision. A moral vision is the result of construction, in that it is the way we organize cognitively the world and we orient ourselves toward it. Interestingly, Iris Murdoch holds that through the constructive moral activity we make reality normative.47 Consistently with the kind of constructivism I have defended, she argues that objectivity is a moral task for the agent to achieve in making reality normative, rather than an exercise in mirroring reality. I take these final considerations to support my claim that the metaphor of construction is the most helpful representational and explanatory device for elucidating the nature of our ethical judgments, coherently with moral phenomenology. At the same 31 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 time, these considerations also point at a continuity of intents in those contemporary philosophical projects, such as Wiggins's, McDowell's and Blackburn's that are preoccupied to vindicate moral phenomenology. My contention has been that constructivism provides the best theoretical framework for this vindication. BIBLIOGRAPHY Bagnoli C. "On the Objectivity of Practical Reason", Croatian Journal of Philosophy, 2001a, 3, forthcoming. Bagnoli C. "Realism as a Moral Achievement", in Ricciardi (ed.), The Philosophy of Iris Murdoch, 2001b, forthcoming. Bagnoli C. "Value in the Guise of Regret", Philosophical Explorations, 1, 2000, pp. 169-187. Bagnoli C. "Vision and Insight: The Intuitionist View of Normativity", American Philosophical Association, Pacific Division, 30 March 2001, conference paper Bagnoli C. Moral Dilemmas and the Limits of Ethical Theory, LED, Milano, 2000. Blackburn S. Essays in Quasi-Realism, Oxford, Oxford University Press, 1993. Blackburn S. Spreading the Word, Oxford, Clarendon Press, 1984. Brandom R. "Perception and Rational Constraints", in Villanueva1995, pp. 241-259. Crisp R. "Particularizing particularism", in Hooker and Little 2000, pp. 23-47. Cullity G. and B. Gaut (eds.), Morality and Practical Reasoning, Oxford University Press, 1997. Dancy J. "Two Conceptions of Moral Realism", in Rachels 1998, pp. 227-244. Dreyfus H. and S. Dreyfus "What Is Moral Maturity? Toward a Phenomenology of Moral Expertise", in Ogilivy J. (ed.) Revisioning Philosophy, Albany, SUNY Press, 1992. Gibbard A, Darwall and Railton (eds), Moral Discourse and Practice, Oxford University Press, 1997. Haldane J.and Wright C. (eds). Reality, Representation and Projection, New York, Oxford University Press 1993. Hare R.M. Freedom and Reason, Oxford, Clarendon Press, 1963. Holzman and Leich (eds.) Wittgenstein: to Follow a Rule, Routledge, Londra, 1981. Hooker B. and M. Little (eds.) Moral Particularism, Oxford University Press, 2000. Hume D. An Enquiry Concerning the Principles of Morals, edited by J.B. Schneewind, Hackett Publishing, 1983. Johnston M "Objectivity Refigured" in Haldane and Wright 1993, pp.85-130. Little M. "Virtue as Knowledge: Objections from the Philosophy of Mind", Nous, 1997, 31, pp. 59-79. Lovibond S. "Ethical Upbringing: from Connivance to Cognition", Lovibond and Williams (eds.) 1997, pp. 76-95. Lovibond S. and Williams G. (eds.) Identity, Truth and Value, Blackwell, Oxford, 1997. Mackie J.L. Ethics. Inventing Right and Wrong, Penguin Books, 1997. McDowell "Reply to Brandom", Villanueva (ed) 1995, pp. 290-300. McDowell J. "Non-Cognitivism and Rule-Following" in Holzman and Leich 1981. McDowell J. "Projection and Truth in Ethics", Lindsay Lecture, University of Kansas, 1987, in Gibbard, Darwall and Railton (eds) 1997, pp. 215-227. McDowell J. "Values and Secondary Qualities", 1985, in Rachels 1998, pp. 210-226. McDowell J. "Virtue and Reason", Monist, 1979, 62, pp.331-350. Pettit P. "Realism and Response-dependence", Mind, 100, pp.587-626. Rachels J. (ed) Ethical Objectivity, Oxford University Press, 1998. 32 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 Rawls J. "Kantian Constructivism in Moral Theory" (1980), in John Rawls: Collected Papers, Harvard University Press, 1999. Rawls J. Lectures on the History of Moral Philosophy, (B. Hermann ed.), Harvard University Press, 2000. Skorupski "Logical Grammar, Transcendentalism and Normativity", Philosophical Topics, 1999. Skorupski J. "Irrealist Cognitivism", Ratio, 1999, 12, pp. 437-438. Skorupski J. "Reason and Reasons", in Gaut and Cullity 1997, chapt. 8. Smith M. The Moral Problem, Blackwell, Oxford, 1994. Taylor R. Good and Evil, New York, Macmillan, 1970. Villanueva E. (ed). Perception, "Philosophical Issues", vol. 7, 1995. Wiggins D. "A Sensible Subjectivism", in Wiggins 1987, pp.185-214. Wiggins D. "Moral Cognitivism, Moral Relativism and Motivating Moral Beliefs", Proceedings of Aristotelian Society, 1990-91, pp.61-85. Wiggins D. "Truth and Truth as Predicate of Moral Judgments", in Wiggins 1987, pp. 139-184. Wiggins D. "Truth, Invention, and the Meaning of Life", 1976, in Wiggins 1987, pp. 87-138. Wiggins D. Needs, Values, Truth, Blackwell, London, 1987. Williams B. Making Sense of Humanity, Cambridge University Press,1995. Wright C. "Truth in ethics", Ratio, 1995, 8, pp. 209-26. Wright C., "Moral Values, Projections and Secondary Qualities", Proceedings of Aristotelian Society, supp.vol. 62, 1988, pp. 1-26. 1 A previous version of this paper titled "Truth, Invention, and Construction" has been published in Notizie di Politeia, 61, 2001, pp. 48-66. I have introduced many and substantial revisions, especially in sections 45. I should like to thank Luca Ferrero, Valerie Tiberius, Gabriele Usberti, Robert Schwartz, and Ralph Wedgwood, whose criticisms and suggestions guided me in rewriting this article. I am particularly indebted to Christine Korsgaard and John Skorupski. 2 As Michael Smith puts it: "The problem is that ordinary moral practice suggests that moral judgments have two features that pull in quite opposite directions from each other. The objectivity of moral judgment suggests that there are moral facts (…) but it leaves totally mysterious how or why having a moral view is supposed to have special links with what we are motivated to do. And the practicality of moral judgments suggests just the opposite that our moral judgments express our desires. While this enables us to make good sense of the link between having a moral view and being motivated, it leaves mysterious (…) the sense in which morality is supposed to be objective", Smith 1994, p. 11. See also Wiggins 1990. 3 Some of these non-cognitivists endorse a realist conception of objectivity, and deny that ethical judgments can legitimately aspire to it. Others, such as R.M. Hare and Blackburn, propose a conception of ethical objectivity that is alternative to realism. But Hare also argues that the objectivity (generated by universalizability) and action-guidingness (prescriptivity) are two distinct and separable features that depend of distinct logical functions of ethical judgments; see Hare 1963. Blackburn cannot easily be situated in either of the camps characterized as realist or non-cognitivist. This is because he defends the notion of objectivity and moral knowledge but argues that these notions are so thin and deflated that even a Humean can use them. My preliminary comments on the distinction realism/non-cognitivism are meant to uncover the assumptions underlying the claim that practicality and objectivity are necessarily at odds with each other. I will argue that the project of "reconciliation" is grounded on a very narrow account of the meta-ethical alternatives we do in fact have: Blackburn has attempted to broaden it, as I will argue in section 2. 4 For a map of these meta-ethical alternatives, see Smith 1994, pp. 3-13. 5 See Wiggins 1976, pp. 92-98; the example is drawn from Taylor 1970. 33 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 6 I add this third variation on the second scenario because it seems to me to better vindicate Wiggins' claim. However, the first non-cognitivist strategy could be read as equivalent to the third one: in this case, it is the endorsement of the desire, not merely having the desire, that makes Sisyphus's life meaningful. 7 Interestingly, this consequence of Wiggins' claim concurs with Rawls' claim that only on a constructivist account there is a vindication of the autonomy of practical reason. That is, only if one endorses constructivism and acknowledges the constructive and productive powers of reason one is capable of a non-instrumentalist conception of rationality. I consider this claim in Bagnoli 2001a. 8 This denial might sound plausible because to suggest that conative states are modes of awareness seems to encourage a form of intuitionism according to which our feelings are perceptions of moral properties. But as I will show in section 3, to accept the view that ethical judgments are modes of awareness does not commit one to intuitionism. 9 On a more sophisticated non-cognitivist view, the reality is provided in a manner of a new creation, see the discussion of Blackburn in the following section. 10 As Wiggins writes: "The extension of the concept objective is quite different on the inner view from the extension assigned to it by the outer view. And the rationale for determining is different also", Wiggins 1976, p. 99. 11 Hume, p. 88. 12 Blackburn 1984, p. 217. 13 Blackburn 1984, pp. 181-189, Blackburn 1993, pp. 149-181. 14 Blackburn 1993, p. 158. 15 Blackburn 1984, chapts. 5-6. Compare Blackburn 1993. 16 Blackburn 1984, p. 219. Italics is mine. 17 Blackburn 1984, pp. 217-218. 18 See McDowell 1987. 19 Wiggins 1976, McDowell 1987, p. 219-221. 20 McDowell 1979. 21 Mackie follows Locke on this claim. Mackie 1977, p. 98. 22 The claim that value judgments are all false because there are no such facts as values may be explained as the inability of mere conceptual analysis to have any ontological significance. The conceptual analysis of evaluations as assertions is misleading from an ontological point of view, since these assertions are about nothing. Therefore, conceptual analysis cannot explain fully the activity of valuing. This activity raises questions that Mackie calls 'factual', see Mackie 1977, p. 98. 23 Even though not revisionist, error theory unmasks a systematic mistake, and it is hard to understand how its general acceptance can leave the current practice as it is. The awareness of proffering a false judgment any time one proffers an evaluation must have an impact on the practice of the evaluation: it is likely to undercut our evaluations and also to undermine their authority and effectiveness. For this criticism, see Williams 1995. Mackie 1977, pp. 5-49. Consider that for Mackie "Values" stand loosely for rightness and wrongness, duty, obligations and other thick concepts, Mackie 1977, p. 15. Mackie 1977, pp. 322, 60-61, 73-74. 24 Neither Wiggins nor McDowell dwell on the first realist conception of the analogy between values and secondary qualities. This is, I suppose, because they share Mackie's view that the realist notion of moral perception is vulnerable to the objection of queerness. I have attacked the argument from queerness in Bagnoli 2001. My impression is that McDowell's reading of the analogy is closer to realism than he wants to admit. His insistence on falsity of distinction between cognitive and conative elements, which does not enter the realist conception of the analogy, is revealing. Moreover, the realist conception of the analogy invites the thesis that virtue is knowledge, which McDowell himself defends, see McDowell 1979. 25 Compare Murdoch 1957, Bagnoli 2001b. 26 Murdoch also questions the importance of rules, but she does not appeal to Wittgenstein: quite interestingly, she claims that Wittgenstein' argument about private language has encouraged the view that rules and universal practices are all important, and thereby encouraged a behavioristic ethics. I examine this argument in Bagnoli 2001b. 34 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 27 Even though I do not have here the opportunity to adequately support my contention, let me insist that I do think that the Wittgensteinian argument establishes something against codification, but I think that it establishes nothing about the possibility and the importance of moral principles. Uncodifiability is a thesis defended by most ethical theorists, myself included. Thus, the argument based on Wittgenstein's rulefollowing considerations has far less import for ethical theory, and especially for impartialist and universalist ethics, than many assume it has. Moreover, the argument has no implications whatsoever concerning the adequate ethical theory. In particular, contrary to what McDowell suggests, I believe that the Wittgensteinian argument does not support Aristotelian versus Kantian ethics. 28 This argument can be found in Wiggins 1976, McDowell 1979, pp. 336-343, McDowell 1981. 29 McDowell 1985, p. 170. "Secondary qualities are qualities not adequately conceivable except in terms of certain subjective states, and thus subjective in themselves in a sense that that characterization defines", McDowell 1985, p. 113. "To press the analogy is to stress that evaluative attitudes, or states of will are like (say) color experience in being unintelligible except as a modifications of sensibilities like ours", McDowell 1985, p. 175. 30 McDowell 1985, p. 142; see also McDowell 1979, pp. 332-333. 31 McDowell 1985, p. 175. 32 Dancy objects that the response-dependent account as formulated by Wiggins and McDowell is unstable and ultimately relies on a strong version of realism that denies response-dependence, see Dancy 1998. 33 I find that this is the most general way to formulate the objection against using the analogy with secondary qualities as an explanatory device. Blackburn more fastidiously gives us half a dozen reasons for rejecting the analogy: 1) moral properties supervene on their natural bases in quite a different way than colors; 2) the analysis of defective receptive mechanisms is not similar to the analysis of defects of character; 3) secondary qualities are mind-dependent in a way moral properties are not; 4) variations in moral practices differ dramatically from perceptual variations; 5) moral distinctions are intrinsic to the moral domain, in a way that perceptual distinctions are not; 6) evaluative practices are typically attributive, see Blackburn 1993, pp. 159-161. 34 Murdoch 1956. I have examined this thesis in Bagnoli 2001b. 35 This thesis has been objected because it seems to assume that moral claims have two directions of fit (like beliefs they conform to the world, and like desires they want the world to conform to them), see Smith 1994, p. 118. I agree with Little that these objections from the philosophy of mind do not succeed in undermining the view that virtue is a form of knowledge, see Little 1997. 36 These are two different claims. On the one interpretation, to know is to be motivated. On the other one, knowledge guarantees that one is appropriately motivated, but it is not equivalent to it. 37 McDowell 1979, p. 339. Compare Wiggins 1991, p. 79, Wiggins 1997, pp. 255-56. This account might seem to encourage the objection of radical conventionalism. Evaluators belonging to different communities do not share the same space of reasons. When coming from a different ethical community, reasons and criticisms are unintelligible, like noises. As Wiggins notices, such relativity pertains to the level of "sense." The contingency of human nature determines the "sense" of value language. Once this sense is fixed, "it is not human nature and responses that determine the reference or truth-value of putative answer to it", Wiggins 1991, p. 79. 38 Compare Rawls 1980. 39 I have attempted this reduction in Bagnoli 2000, chapter 3. 40 See Rawls 1980, Rawls 2000, pp. 212-214. 41 I have argued at length for this claim in Bagnoli 2000. 42 On the notion of practical constraints, see Bagnoli 2000a, pp. 169-187. 43 Despite his debts to Sellars, McDowell' account does not pay much attention to the social origin of norms. For this complaint (to my view quite adequate), see Brandom 1995; it is my opinion that McDowell's reply in McDowell 1995 fails to address Brandom's concerns. 44 On the formulation and defense of cognitivist irrealism, see Skorupski 1999, pp. 437-438, Skorupski 1997, chap. 8, and Skorupski's article in this volume. 45 I should like to thank Ralph Wedgwood and Valerie Tiberius for having brought this objection from different camps. 35 Carla Bagnoli “Moral Constructivism: a Phenomenological Argument” Topoi, 1-2 2002, pp. 125-138 46 47 See Blackburn 1984, pp. 196-202. Murdoch 1997, p. 201. I have proposed this interpretation in Bagnoli 2001b. 36