Review of "Impassioned Belief"

Essays in Philosophy, Jan 2015

Steven Ross

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

Review of "Impassioned Belief"

Essays in Philosophy Review of "Impassioned Belief " Steven Ross 0 0 Hunter College/Th e Graduate Center/CUNY - Im passioned Belief oBkReviw|s Steven Ross Hunter College/The Graduate Center/CUNY Impassioned Belief. Michael Ridge. Oxford: Oxford University Press, 2014; 288 pp.; $55 hc; ISBN 978-0-19-968266-9. t’s a funny thing. If you turn to virtually any issue in philosophy of mind, Cartesian dualism will loom large in the background, and rightly—but it does so in a particular way. Of course, it is a powerful and, in many ways, intuitive theory. Surely we have mental states, they are often clearly distinct from whatever intentional actions they go on to cause, and it is not obvious how we can identify consciousness, the experience of that fireplace with any purely material state or process. Fine. At the same time, everybody knows the theory cannot be right, for all sorts of reasons. But central is this: Cartesian dualism is just too stark, as it were; the framework within which we characterize what is and is not possible is just too meager. Too many features of our psychological and intentional life do not fit easily or naturally into “mind” or “body” as Descartes thought he was forced to understand these ideas. Mental states are clearly, must be, materially instantiated; they could not have their properties any other way. Certain behavior, like painting a landscape, is intrinsically intentional. And so we tend to begin our explorations into some topic in mind with Cartesian dualism in the background, but as a cautionary tale as much as anything else. It would be the odd philosopher, something of an ideologue, who would want to wind up at the end of their analysis with these categories, these rigid distinctions, wholly intact. To the contrary, we look for an analysis of mental life and intentional action that is more flexible, more subtle than what Descartes thought he was constrained by, an analysis not hostage to these old, unhelpful, crude categories. But in that branch of moral theory known as meta-ethics, it is very much otherwise. Here the dominant, and understandably dominant, dualism in the background is Hume’s. Hume’s dualism, as I will call it, is as brilliant, as seductive, and as straightforward as Descartes’. There is the world of facts, which in turn are divided into natural facts and “abstract ideas,” and we detect such facts with a distinct faculty, “reason.” This faculty represents such facts, and as such, it cannot be thought to cause action, or "supply a motive." Then there is the realm of feeling, or states, inside of us. Sometimes such states arise biologically (hunger, our love of children). And sometimes they are caused by how the external world impinges upon us (pain from being burned, finding smells disgusting). But this is the essential framework: facts out there, where what “facts” are is to be understood wholly within the constraints of naturalism, and feelings inside, where these feelings, or motives, or desires, are to be understood in terms of the state of affairs they seek to bring about, not in terms of representing anything, and so not capable of being termed true or false. When we turn to morality, two sorts of arguments, one from each part of this duality, support a certain picture of moral concepts and moral discourse. When we look out at the world of facts, we can find, initially, no place for moral notions. Goodness does not name a natural fact, it is not discovered through empirical investigation—just go look and see if you can find it, one might say!—nor is it plausibly thought of as an “abstract idea” (what expressivists today would call a discrete metaphysical entity). Arguing from the other half of our ontology, clearly moral qualities sometimes motivate. We sometimes act otherwise when we discover we are about to be rude, for example, or do something simply because we believe it is just; this cannot be denied. But motivated action cannot be caused by a mere representation of the world. So we have two sorts of arguments, one from each part of the world we have so divided, for the single truth: morality is essentially about feelings, responses, which then cause certain sorts of actions. I do not want to suggest this picture has survived whole, even for those who essentially adhere to it. No one I think would now characterize desires, or motivations, exactly as Hume did. Expressivists nowadays typically think of desires as having semantic content; they are certainly not “brute states” as Hume seemed to understand this idea. But these differences, or modifications, are I think quite modest in comparison to the degree to which the fundamental framework, the dualism, continues to be invoked to support some very powerful conclusions. First, in the seemingly undeniable fact that morality motivates, we have the beginnings of an argument for telling us what normative concepts name, what normative properties are. And in the world of facts, as we must understand that idea if we are to be sensible, we will find no moral properties. The world apart from our reactions, apart from our feelings, bears no moral content at all. It is not that these arguments strongly “influence” contemporary expressivism. Almost three hundred years since the publication of The Treatise, these arguments are being made more or less exactly as they were then. And what is invisible, what cannot be seen, when within Hume’s dualism, is invariably invisible when within contemporary expressivism too. The result, I shall argue, is a deep distortion of the subject matter from the start that is then typically elaborated upon with great imagination. Michael Ridge’s Impassioned Belief is I believe a very vivid case in point. Ridge has defended his version of expressivism, Ecumenical Expressivism, before; Impassioned Belief is not only a book length treatment of his view, it also gives Ridge the chance to take up at greater length several topics central to any version of expressivism: the connection between morality and feeling, the best way to understand the normative generally, why cognitivism cannot be right, how an expressivist can think of truth, and so forth. Those who know Ridge’s argument from previous articles might well find the presentation of the distinctive themes of Ecumenical Expressivism here actually a bit more opaque than in the past (I certainly did); to that end, in what follows, I allow myself to refer to language from one of these earlier essays too. Naturally, the account Ridge offers in Impassioned Belief is also far more comprehensive; this is especially so when it comes to Ridge’s commitments in language and mind. Readers particularly interested in Ridge’s elaborate arguments in philosophy of language will find a great deal here. But that said, Ridge’s ambitious, multifaceted argument almost everywhere rests on these unexamined Humean assumptions. And because this framework actually renders morality, and the normative generally, invisible, since it turns normative judgment and normative justification into something else (a reaction, a feeling, a decision to embrace a plan, whatever), the work is filled with fundamental errors and bad arguments. Not about language, of course; that would never happen. The level of sophistication about language is positively military. No; the casual nature of the arguments, the unexamined assumptions, the ease with which all sorts of questionable, sometimes truly ridiculous, assertions are made, the failure to take up at all seriously the complexity of the subject—these features of the book invariably arise when the author takes up morality. Ridge gets morality so wrong so pervasively, and with so little care, this book is bound to become a classic, beloved by meta-ethicists everywhere. Not since Mackie have I read a work allegedly about morality, or the normative, with so little interest in, or intuitive feel for, the subject. In what follows, I will take up some of the central themes. Some, like the particular picture of the relation between morality and feeling expressivists favor, are common to expressivists everywhere. Others, like the account of moral judgment as kind of hybrid state, with a representational belief component, are quite distinctive to Ridge. But before getting into any of this, let me say a bit more about the value of the book overall. If you are already an expressivist, this book will be of great interest to you. There is a lot of discussion of philosophy of language and to a lesser extent philosophy of mind, and an attempt to present expressivism within a “unified” view of these domains. Conversely, if you are hoping to see a version of expressivism that at last takes up the various features of morality that expressivism does not easily fit, this book is quite frustrating, largely an annoyance. The powerful intuitions to the contrary that fail to be taken up are not whatever is available so long as we stay within the Humean framework; it is not “realism” or “cognitivism” as such things would be within Humean dualism that fail to be examined. Naturally, there is adequate treatment of that. It is rather the absence of any attempt to construct a conception of the normative distinct from what is available when within the Humean framework that renders the central argument so unsatisfactory. The shadow of Humean dualism continues to fall on otherwise sophisticated philosophical treatments of normative judgment, and the deep, structural limitations of that framework are passed on wholesale. Descartes famously worried that if the foundation of his argument was not sound, all that was built upon that foundation would be subject to thoroughgoing challenge too. He very much got that right. SOME CENTRAL TOPICS: SO WHAT EXACTLY ARE WE TALKING ABOUT, TO BEGIN WITH? From the start, the idea of “the normative” is put forward in a deeply idiosyncratic way. It is essentially what grows out of our ability to deliberate rationally about our action, so long as there is some concern about the outcome. And so, from the start, “the normative” is tied essentially to what goes on when making a decision—any decision, it turns out. Ridge opens his first chapter, “Locating Normative Discourse” with the following allegedly paradigmatic illustration of the normative: If we are being especially careful and systematic [when reflecting on our options], we might even list what we take to be the “pros” and “cons” of the most salient of our options to help us figure out what to do. In doing so, we seem to presuppose that there are right and wrong answers to the question, “What should I do?” Indeed, it sometimes matters deeply to us that we answer this question correctly. For example, an agent might well wonder whether it makes sense to have a child. She realizes that this will involve large sacrifices of her other interests and projects. On the other hand, she also feels that it would significantly enrich her life, providing a kind of deep meaning and fulfillment. Clearly, people making such momentous decisions typically care deeply that they make decisions which are sensible. How, though, should such competing and very different values sensibly be weighed against one another? Even more deeply, how can one rationally determine whether either really even is a genuine value? It is no exaggeration to say that asking and answering questions about how we ought to live is an essential part of the human condition. We care deeply about getting the answers right, yet we are unclear about how to be sure we are getting them right. On reflection, we can become unsure just what could make a given answer right. We can even become unsure what the questions are even really about. This book is, very roughly, about these kinds of practical questions. More broadly, this book is about the kind of thought and discourse employed in asking and answering such questions. I call such thought and discourse “practically normative.” (15-16) A lot is going on in this passage; a careful reader is alerted to a great deal of the necessary stage setting. First, notice the easy-going equivocation between “the sensible” and “the right,” or, to put the point a different way, notice how “the right” is here put forward as just part of a list of terms we might understandably use in assessing personal decisions. Of course, there is a use for “right” in this way; I take that to be obvious. So I might say, “It was quite the right decision for me to have a child.” And here I mean: this turned out to make me happy. I submit this is a very different sort of use than what we see in, “When Truman integrated the armed forces, he was quite right to do so.” Before reading this passage of Ridge’s, I would have thought that the paradigmatic case of “the normative” might have been the second use, the use of the term that means something like “morally justified” or “admits of moral justification.” But ah no, it is not. We are in Jean-Paul Sartre country now; it is the personal decision, the weighing of options, by an earnest, free self, that gives us the central setting for, and so the central logic of, “the normative.” And so, the moral, from the start, is going to be just a species of this larger, more general category—assessment of options with some interest in the outcome, and to study “the normative,” (and so the moral), is just to study the means and mechanisms by which we come to make important personal decisions. The idea that the moral might traffic in properties or states of affairs that we either recognize or fail to acknowledge (as is illustrated in the Truman-integration example) is never given a chance from the start. This (to me fairly intuitive idea) is defined away, literally, on the very first page. To me, this woman’s dilemma is only with some strain thought of as raising “normative” matters in the first place. What she worries about is what course of action will make her happy, and I see nothing interesting about “the normative” in the undeniable platitude that we usually worry about our happiness when making important decisions in life. And because we know we cannot always anticipate our desires and proclivities in the future, in this case, there is a certain amount of speculation or uncertainty that no amount of “reflection” can wholly dissipate. At the same time, we know we are biological creatures, hard wired to love our children, and so we know that there is a good chance that we will love the child we have however much we give up in becoming parents. Perhaps I’m crazy, but I did not see anything “normative” in any of that. Why take this story as illustrating anything? So on my view, this story, and the sort of thing it is an instance of, very much mislocates the subject matter from the start. We are offered something far too wide as it were—it is something like “personal decision with some interest in the outcome.” Of course, when making such personal decisions, sometimes we do worry about (what I would call) the normative; of that there is no question. For example, we might worry whether we really do have a personal obligation to give to the desperately poor in disadvantaged countries, and if we do have such an obligation, how much personal sacrifice we ought, in taking up this obligation, to accept. And if we don’t happen to have this worry that solicitation from Oxfam will try to make sure that we do. There can be no doubt about that. But it hardly follows that the best way into that issue is by examining mental processes or the nature of deliberation as such. It begs the question to argue that, because we deliberate about normative matters, or where a normative predicate most lies (which action is the most just? which act is the more obligatory—the very stuff of tragedy) that “the normative” is to be found in deliberation. Consider the counterpart point made of normative concepts in the law. If you are on a jury, you must deliberate about whether the nightclub owner was negligent, whether he took, or failed to take, “reasonable” precautions against a “foreseeable” danger. Against the right fact pattern, this is tough stuff, and a lot of film agony has been made out of what goes on in the heads of jurors in a jury room. But clearly any talk of “deliberation” is derivative. “Negligence” will have a certain abstract definition in a certain jurisdiction, criteria for application, and jurors are charged with trying to apply that concept of “negligence” correctly. (That’s of course why it is so hard.) Just because we deliberate about negligence, “trying to get it right” one might say, shows absolutely nothing about how the concept of negligence is best understood. It is no accident that Ridge does not give us, as his example of “deliberation,” someone on a Hague tribunal wondering whether Nenad Banovic really is a war criminal, or someone on an arts committee wondering whether William Vollmann’s work really is original enough to deserve a grant. He does not like the deliberation to be about the application of (what I would call) some normative concept with more or less paradigmatic cases, more or less clear criteria, governing its use; that would then shift the attention to the concept. And this point leads us to another. Unsurprisingly, the central normative concepts as Ridge will understand “the normative” will be those that can stand in for “approval,” “endorsement” and the like—indicia of success, or satisfaction, in deliberation, abstractly conceived. This is what “good”—the alleged granddaddy of all normative language—and so the morally good—means: what we can approve or endorse. And when we construe “the good” or “the morally good” this way, there is now no logical or intrinsic conceptual connection between “the good” and the content laden normative language we employ in (what I would call) related domains, like art and the law. Ridge, like all expressivists, is superbly consistent in this regard. You will find no entrance in the index for “aesthetics” or “law” nor is there any mention of any particular theory of judgment in these domains—no entrance for Ronald Dworkin, or Clive Bell. (This is not hyperbole.) From the start, there is a resolute, unruffled disinterest in how normative language in developed, constructed domains actually works. Such things are not even on his radar. This mistake has been with us since Moore—the idea that “the good” can be understood, is best understood, in a way that has no connection whatsoever to related notions in related domains (e.g., “the just,” “the deserved,” “the rightly punished”). It seems very strange to me to have a theory of “the morally good” in which it is conceivable that we could, for example, have a complete conception of justice say—one on which we all agreed and all agreed to be as well justified as is imaginable—and this fact have no implications whatsoever for “the morally good” or the best theory for how to understand this concept. The idea that the criteria for judgment in other normative domains might have no implications for how to understand what we might mean by “good” as it operates in our moral life strikes me as all wrong. I take it as obvious that this thought, or thought experiment, would have struck Mill (or Kant) as just about incoherent—but we are now in a world where what Mill or Kant thought about “the morally good” is of no interest to anyone. This is the conceptual point, that “the morally good” is conceptually connected to other normative concepts in other domains, and to ignore this will generate certain serious limitations or artificialities in one’s treatment of morality. I make this point in a somewhat preliminary way now, because I will return to it in greater length below; it is crucial to assessing the famous argument Hare makes regarding “good” that Ridge repeats and relies on so significantly. Then, separate from this, is a point about practice. The way other normative domains actually work, how judgments are assessed or disputes structured, may have something to teach us about moral language and argument too. This is particularly so, and unsurprisingly so, with the law and legal life. I say it is “unsurprising” because after all, our moral life grew out of more primitive forms of social cooperation, tribal understandings and the like—just as our law did. Morality and law are, I submit, like man and ape in the Darwinian story, or like cosmology and literature—they each have exactly the same ancestor, the same form of life figuring in the causal story in the background. It would be very surprising if there were nothing to be learned about the nature of moral predication, moral disagreement, even moral motivation, by looking at the law. It would be very surprising because in fact, as I will show, it is very much not true. Ridge’s theory of Ecumenical Expressivism fails to take up what a concurrence is in the law, the agreement in outcome that stems from very different readings of the law (I do not think he even knows about this category), and his account of what it is for two moral agents to mean, or not to mean, the same, or different, things in their judgment is accordingly quite unsophisticated. (This is not to say his account of what is going on is semantically unsophisticated—hardly!) This, too, is a point to which I will return. Well, theories are theories, and if this is the best one, then so be it; the dislocations it would seem to require of us will in that case just have to be swallowed. After all, it is not obvious that time will slow down as we approach the speed of light either, but, since that is what the best theory requires we say, we better do our best to accommodate the idea. Let us now turn to some of the reasons Ridge, and expressivists generally, go this way, construe the normative along these lines. Let us look at what they understand their only opponents—“cognitivists”—to be, and why they feel that given that alternative, whatever difficulties there may be with expressivism, this is the theory to affirm and make better. And finally, let us look at the distinctive form of expressivism we have with Ridge, Ecumenical Expressivism. THE METAPHYSICAL ARGUMENT, THE ARGUMENT FROM MORAL DISAGREEMENT Expressivism, in this debate as it presently stands, is in contrast to descriptivism. A descriptivist account of a normative judgment takes its propositional content to “represent” “normative properties” and/or “normative facts.” But the prefix “normative” adds nothing to the ontology available to us without it. The properties and facts a descriptivist, within this framework, must say a normative judgment “represents” may be understood to be, or to reduce to, natural properties and facts. Or perhaps, in frustration with naturalism, yet determined to be factual about the good all the same (Moore’s argument in a nutshell), we turn instead, or simply posit, some irreducibly non-natural metaphysical entity. There is also the possibility that such facts are of a different kind altogether, constructed products of human intentionality. But expressivists only rarely take that third possibility seriously. Ridge very much embodies the more typical strand of expressivism when he writes: It is at this point [where cognitivism is judged trapped by the unsatisfactory consequences of naturalist reduction, whether such reduction is understood along analytic or synthetic lines—SR] that expressivism comes into its own. The expressivist gambit is to reject the cognitivist’s representationalist order of explanation. On the expressivist approach, we do not start with normative states of the world and explain normative judgment as cognizing them. Rather, we start with normative judgment and its distinctively practical functional role. Normative judgments function to settle “the thing to do,” and normative conversation allows us to deliberate together about the thing to do… (101-106) This pattern of argument [regarding truth] should seem familiar: it is precisely the pattern of argument used to motivate expressivism. In both cases, we find a meaningful predicate which persistently evades a plausible reductive analysis, but which cannot plausibly be understood as denoting a metaphysically primitive property. We therefore revisit our methodology. Instead of starting with metaphysical theorizing about the referent of the predicate in question, we begin with theorizing about the meaning of the predicate. Having concluded that meaning should be understood in terms of state of mind expressed, we next ask what sort of state of mind is expressed by declarative sentences in which this predicate is used. We then conclude that when used in assertoric discourse, at least, the predicate functions to express something other than a robustly representational belief. Instead, it functions to express some sort of pro-attitude… (201-202) Here we see the Humean framework in all its confident, exclusionary glory. Note the possibilities of “reference,” the range of what normative language might refer to, that, once exhausted, inexorably drive us towards an analysis that understands normative language as instead expressing a state of mind, a “pro-attitude.” We have natural facts, and combinations thereof, and metaphysically primitive entities. We are very much in Hume’s world now. It is perhaps not quite identical to “matters of fact” on one hand and “relations of ideas” on the other. The “ideas” that Hume speaks of, such as those of mathematics or geometry, could certainly be thought of as human constructions. Here, the expressivist argument refers to non-natural facts that must be thought of as “out there” in just the same way that natural ones are. But given that 1) the “meaningful predicate” “persistently evades” “a plausible reductive analysis” on one hand (consider the willful murder in all its facts and relations and the vice entirely escapes you, as Ridge’s more famous ancestor might have put it), and 2) the idea that good might refer to a nonnatural entity is rightly seen as not very plausible either, (if we took “good” to be an “abstract relation,” then it would have no intrinsic connection to human interests; Hume again), then 3) we must understand the predicate in terms of a “state of mind expressed” (‘tis only when you turn your attention towards your own breast…). Given what we have just ruled out, and that the term appears meaningful, appears stable, we must see moral language as non-factual, moral claims as non-representational, and moral judgments as simply expressive. I say Ridge embodies the strand of expressivist argument that does not take the third sort of fact very seriously for just this reason. There is no mention of any other sort of fact at all in this argument. Nor does this bother Ridge; he never even takes up a candidate predicate that appears to evade this framework, and to show that such a possibility is illusory after all. How can Ridge assert so blandly that this exhausts the range of how terms refer? I believe that this blandness, this confidence, is in fact just Hume’s framework speaking. Given that moral discourse cannot be said to track either Platonic entities or reduce to any naturalistic predicate, it must name a non-representational psychological state. But this argument completely truncates the alternatives out there, the actual possibilities. When we turn to accomplishment (and failure) concepts, we have the model we need. Consider: expressive, original, when said of art; witty, clever, insightful, when said of novels or philosophy (or persons); negligent, malicious, (or not), when said of defendants on trial; fair, unfair, when said of wages or policies… it is a long list, and it is not hard to keep going. Does anyone think for a moment such concepts name things in the world, where that is understood as some combination of natural predicates, with distinct causal powers, or, to the contrary, supersensible “entities” apart from the natural world (“Yes,” we might imagine G.E. Moore assuring us in law school, “negligent is negligent and there is nothing else to say about it.”) or simply refer to “pro-attitudes,” inner states that must be thought as having no representational content? That would be an interesting way to construe a verdict of, e.g., “insightfulness” regarding Proust’s treatment of sexual jealousy, as having no representational content! Such terms embody the essence of the normative, one might think—but they are invisible in Ridge’s picture of the possible, given his uncritical reliance on the Humean framework. Well, so what, the defender of expressivism might say. So Ridge left some things out you wish he hadn’t. And maybe as a general account of what is possible, how certain important parts of our language work, or refer, his argument suffers accordingly. But that does not mean he got good wrong. There are plenty of reasons to think he got “good” quite right. And one of the more powerful reasons to think that this is the right approach to take with respect to “good”—i.e., to see it as naming a non-representational state of “approval” or “endorsement”—is that this construal allows us to make the best sense of moral disagreement. Ridge, to his enduring credit, acknowledges his debt to R.M. Hare here, reproducing Hare’s example of moral disagreement and the argument that, if such disagreement is to be preserved, and to be thought as genuinely meaningful, then “good” cannot be construed any other way. First Hare (the passage is from The Language of Morals and is reproduced in its entirety in Ridge, p. 75): Let us suppose that a missionary, armed with a grammar book, lands on a cannibal island. The vocabulary of his grammar book gives him the equivalent, in the cannibal’s language, of the English word ‘good.’ Let us suppose that, by a strange coincidence, the word is ‘good.’ And let us suppose also, that it really is the equivalent—that it is, as the Oxford English Dictionary puts it, ‘the most general adjective of commendation’… If the missionary has mastered his vocabulary, he can, so long as he uses the word evaluatively and not descriptively, communicate with them about morals quite happily. They know that when he uses the word he is commending the person or object that he applies it to. The only thing they find odd is that he applies it to such unexpected people, people who are meek and gentle and do not collect large quantities of scalps; whereas they themselves are accustomed to commend people who are bold and burly and collect more scalps than the average. Now Ridge: Return again to Hare’s example of the cannibals. On the synthetic reductionist account, the cannibals’ use of ‘good’ refers to some descriptive property which scalping people has, while the missionary’s use of ‘good’ refers to some property which scalping people lacks… [I]t will follow that when the missionary says ‘scalping people is not good’ he says something true, and when the cannibals say the equivalent of ‘scalping people is good’ they say something true too. (91-92) In other words, if good referred to some property, then when people disagree, they must in fact be naming different properties. And voila! Now we cannot say they are “really disagreeing” after all, since each use tracks a different property. Therefore, “good” does not name a property; it names, instead, “commending” or “approval” or anything else along these lines. It is surprising this example, so (rightly) central to the expressivist argument, is not subjected to a bit more critical scrutiny. Suppose we were to substitute “deserved” in the right way and rephrase the dispute accordingly. Surely this is a fair move (whatever “fair” might mean). So now we have the missionary saying something like: “Your enemies do not deserve to be scalped and eaten.” And the cannibal says, “Yes, they do deserve to be scalped and eaten.” Now that the dispute between them has been rephrased in these terms, it seems to me pretty clear that the cannibal is simply wrong. Absent a further story, these enemy combatants do not deserve this treatment when they are captured in battle. And let’s be careful here. There is no reason to say the cannibal and the missionary are using a word (now the word would be “deserve”) with “different meaning,” and so “they do not really disagree.” This is a completely artificial and implausible interpretation of the dispute. They have different theories of desert (different conceptions, if you will); they categorize those who deserve and those who do not deserve a certain sort of respect differently. It is probably the case that the cannibal does not even employ the concept of “a person”—there is the tribe, there are enemies (and strangers, of course, to whom you probably have to show hospitality once they are shown not to be enemies) and that is that. For the missionary, there are persons, and as such, they deserve respect, even in battle, or when captured in battle. But when framed this way, it seems obvious to me that the cannibal is wrong. By employing a theory of desert that does not in turn follow the concept of the person, the cannibal makes a mistake, or, to use the language I think more appropriate for normative assessment, offers a theory that is substantially less justifiable than that of the missionary. (And I generally don’t like to agree with missionaries.) The concept of a person is a central concept in any plausible version of moral thought. I won’t defend this claim to any real degree here; I will content myself with the alarmist argument that, if you don’t think this, then you really cannot avoid the most corrosive forms of relativism. Anyone can get out of anything by subdividing persons into those who do and those who do not deserve respect along whatever contingent lines you like, and this subdivision cannot be challenged. Genocide is now not really genocide, you see, because it was not really persons who were targeted, just Jews, just Tutsis. No expressivist wants this result (whether they can avoid it is another matter) so I am going to assume the expressivist will swallow the point about persons. It is hard to avoid acknowledging that 1) morality takes up what we owe to persons—it surely is not about choosing wallpaper—and 2) that the persons/non-persons distinction cannot be understood to be governed by choice, or the intentions of the speaker. We may not want to blame the innocent cannibal for getting it wrong, of course. That is another matter, an easy one. But we cannot think his theory, his conception, is just as good as, is on a par with his missionary adversary. We cannot think it because it isn’t so. Enemies, it happens, really are persons too. This is something one can get right, or get wrong. The moral of the above argument is this. Sometimes what looks like a stopping point in moral analysis is in fact just the beginning. The seemingly intractable different uses of “good” mask different conceptions of desert (in this case), conceptions we can, in this case, easily assess for plausibility. “Good” may or may not operate as Hare and Ridge say it does. I really could not care. If we say they are right, then we must also say that “good,” the concept, turns out not to be so important when doing moral philosophy after all. And of course, we might equally say, they are not right: the concept, when used in a moral context, is best thought of as deeply, importantly linked to others that, once before us, must be thought as answerable to assessment, may not be thought of as governed by “intentions” or “pro-attitudes” at all. THE CONNECTION TO FEELING Crucial to any expressivist argument is the claim that, quite apart from the strengths or advantages of expressivism as a position in meta-ethics, moral life, moral judgment, just is deeply connected to feeling, and that the expressivist alone gets this fact right. On this issue, the expressivist argues that if we just look at the object discipline, just look at what actually goes on when people make and support their judgments, the expressivist orientation is vindicated. The expressivist meta-ethical analysis, that moral judgment expresses a pro attitude, rather than anything like a belief with representational content, best mirrors this deep fact about morality. Well, what sort of fact is this anyway? How are we to understand the role of feeling in moral judgment? That people often get excited about moral judgments is hardly terribly interesting. Many in New York City get far more excited about real estate than about anything else (and justly); no one would say statements about real estate prices essentially reflect a pro-attitude. No, the point is, that when you really press moral judgments, it turns out they can survive the disappearance of everything else, the disappearance that is of any fact, or the detection of any fact, in the world you care to name. They are, it turns out, essentially just feelings from the start. Here is Ridge: In one frequently cited study, American college students were asked to con sider a case in which a brother and sister have sex with one another. In the example, the siblings consent to have sex, use contraception, enjoy it, find that it strengthens their relationship, do it rarely… and keep it a secret. Eighty percent of the subjects judged this was morally wrong but they found it very hard to give any sort of cogent account of why it was wrong. Subjects typically cited concerns about deformed children, but were reminded that contraception was used. Other subjects cited worries about the impact on the community but were reminded that the sex was a well-kept secret. Subjects tended to admit that the counter arguments successfully refuted the reasons they gave for their judgment, but only 17 percent actually changed their judgment. The majority simply held onto their view and tended to just give up on principled argument. Instead, they tended simply to express their feelings, making remarks like “incest is nasty!”… [T]he best explanation of the ‘dumbfounding phenomena’ [where speakers can no longer explain why they continue to hold the view they do] is that speakers’ moral classifications reflect their immediate emotional responses. The reasons they give for that classification are post hoc rationalizations which do not really explain why they make the judgments they do. If the reasons given for the subjects’ normative classification were the real explanation of their judgment then one would expect it to be abandoned when those reasons were debunked by the subjects’ own lights… [I]t is plausible that affect plays a similar role in at least some of our non-moral normative judgments. For example, if we think someone is acting foolishly, we will tend to pity them… (59-60; last italics: SR) What is remarkable about this passage, this story, is that it never occurs to Ridge that it might, on reflection, support exactly the opposite picture of moral judgment than the one he thinks right. This particular example is so rigged in particular ways, it must be handled with tongs, but of that more below. The more general phenomenon is indeed a genuine one, and it deserves some thought. Sometimes indeed people affirm what seems to be a moral judgment and cannot, on reflection, really say why; cannot, on reflection, claim the reasons they had for the judgment in fact apply as they initially may have thought. Typically, as is the case here, such “judgments” have to do with sex, and draw on the deep, visceral role sex plays in our affective lives. I do not think this is at all irrelevant, and I am skeptical whether this case could be extended to cases where no such elemental force is in play. For example, I very much doubt a counterpart experiment structured around judgments about what is wrong with theft, say, or what is wrong with murder, could ever lead us to think there is nothing more than some inchoate, unjustified “feeling” behind such judgments (of course, Ridge is free to try and prove me wrong—survey away). And so: I do not think cases like this, judgments about incest (or bestiality, or sex with corpses), are at all central; they cannot be put forward as paradigmatic. But never mind; let’s say they are and pursue the point. Even as they stand, such cases do not show what Ridge here takes them to show. Sometimes these judgments remain largely dormant; we are not usually called upon to make judgments about bestiality say, and so the degree to which such judgments do not hook on to well supported reasons (if that is so) goes unnoticed. But sometimes the opposite happens. Social change forces a population to confront how well founded such previously unexamined judgments happen to be. And typically, if such judgments are not well founded, are in fact based on no more than some visceral response we cannot otherwise justify, then such judgments lose their authority, are abandoned. The place to see this point vindicated is in public life, where moral judgments deemed to be no longer supported by reasons are no longer enforced. Consider the judgment that homosexual behavior is just wrong. This judgment is still with us of course, but it is nowhere near as widespread as it once was. And the main point is, while some do feel it, very few have any appetite for enforcing whatever anti-sodomy laws still remain on the books, I take that as obvious. (These would be very easy arrests to secure.) Many might say “I still feel there is something wrong here, but I certainly don’t think these people should be punished for this behavior.” Well, if you feel that, how wrong do you really think it is after all? If there is no harm, no disutility, then the judgment tends to wither, tends to remain I would say only as a trace of a more robust judgment, one that was actually supported by reasons we once thought we could endorse. (Usually appeals to God, when no real reason can come to your aid). Along these lines, consider the “feelings” people may have continued to have about inter-racial marriage in a past generation even after the Loving case. Consider our confusion about prostitution, by which I mean, our uncertainty as to whether this really is wrong. Consider how “seduction” was once a crime, once actionable, now no longer. Ridge never looks to our public life and to the way in which some judgments are no longer enforced as showing anything. Au contraire, mon frere, the withdrawal of a judgment from the sphere of what is to be enforced shows everything. It shows there is just the feeling, the trace of a past habit. It is now no longer a judgment which we feel can be justified by reference to public reasons, and when that is the case, then these “judgments,” I submit, are but half way houses, incomplete judgments, not full judgments after all. What else does the disinclination to enforce show? Take away the reference to harm, or rights, or consent, or to whatever else might count as a reason, and Ridge is right—then there is nothing but the feeling. But now it is not clear we still have before us a moral judgment! To the contrary—this is no judgment at all. No justification story, you say, only mere feeling? Fine; then you just have the shell, the empty chrysalis. Ridge takes the inexplicable assertion “I just know it’s wrong” or “it’s just nasty, I can’t say why!” to give us the essence of moral judgment. But this simply, incredibly, begs the question in favor of his dogma. I don’t see this as the essence of moral judgment at all, and I would need a real argument (not a baptism ceremony) to be persuaded otherwise. If you look at our actual habits in public life, it would seem just the opposite is the case. “Moral judgment” is judgment that can point to considerations that can be plausibly put forward as having weight in a moral context. When reasons along those lines no longer apply, or are shown never to have applied, then the feeling that remains, if one does remain, has little authority, perhaps none at all. If you look at our actual habits in public life, it would seem this is how we regard feelings we cannot justify. They are but residues or shadows of judgment, and do not merit full blown endorsement. So what does this “experiment” show, after all? Not that moral life is pure feeling. Rather, that pure feeling is not yet moral life. And now a few words about the example, just because I promised. Typically, we make judgments about types or kinds of actions. Of course, how “types” are to be made out, and what description is warranted by the facts in play, can be a very vexed question. Is this racially charged talk free expression or incitement? Does it make a difference whether you take someone’s property using a false credit card or a gun? (Yes, it does.) But that is not our problem here. In this case, in the incest example above, we are asked to imagine the action-type, but then without the features that typically attend the type. What is before us then is quite suspicious. Among other things, we are asked to assume the parties “keep it a secret.” I thought it was well established after John Rawls’ “Two Concepts of Rules” (where we imagine an objectionable practice that is hidden from view, unknown to be operative, the benefits of which allegedly demonstrating what is wrong with utilitarianism) that practices are to be assessed publicly if they are to be truly assessed at all. Nothing follows about anything (certainly nothing follows about telishment and the plausibility of utilitarianism) should we imagine a practice and then imagine it also not meeting the publicity criterion. Why not ask these same students to “assess” a rigid caste system, but where it is also understood that all parties in the lower castes are hypnotized into believing they are in fact being treated well (the Brave New World case)? After all, having false beliefs about a practice is having false beliefs about a practice—it does not matter how such false beliefs come about. What would that “thought experiment” show us about caste systems, or about our judgments of them? I would say: absolutely nothing. Or consider this: someone deceptively alters another’s will, but in order to give all the proceeds to a charity; we of course also assume that the rightful heirs are well provided for from other sources (no bad consequences, see?), and of course, the initial deception is never discovered and so kept a secret. What would you say to that, huh? If I were asked this question, I would just laugh. Once you take away the results or features that are tied to an action type in the typical case, and then ask, well what do you think of that act or practice now, of course there will be uncertainty and confusion. Our judgments, to the contrary, are well founded, well justified, in so far as we can point to consequences of an action, or features of the action, that warrant praise or condemnation That is what we in fact do when justifying a judgment. Moral judgment is rooted in considerations. Take away those considerations, ask for a response all the same, and the results, whatever they are, will show you nothing about the nature of that practice or about the nature of moral judgment. And to the contrary, to go back to point made above, if we come to think that the considerations we once thought did attach to an action type in fact do not (the condemnation of homosexual behavior case, the seduction case), then, whatever “feelings” do persist are not going to be given very much weight or moral authority. And finally, just where is the evidence for the remarkable claim that the normative judgment of, e.g., foolishness, goes hand in hand with some sort of affective response, in this case, pity? When I say, “Stalin was very foolish to think Hitler would not attack the Soviet Union” or “Hitler was very foolish to extend his armies as far as Stalingrad” where in heaven’s name is the pity? The claim is simply ridiculous. And here is the central thing: if any of these claims just taken up were the counterpart claims in the philosophy of language, you can be sure, Ridge would be exacting in his examination. Different views would be canvassed, different arguments meticulously summarized and assessed. The bibliography would positively swell with references to important articles by important philosophers. But if it is a claim about our moral life, a breezy unexamined assertion will do. No mention of the contrary position, let alone rehearsal of the contrary argument, need be made, even if the contrary positions are held by some of the most formidable thinkers in the history of moral philosophy. Why is this? I think we must say it is simply because these misrepresentations are the ones that fit the expressivist program. The whole argument in Impassioned Beliefs is like an exquisitely made mechanical model of the Ptolemaic universe. The underlying assumptions are never examined critically, just the internal workings of what follows if those assumptions are granted. But the assumptions are almost everywhere implausible. MOTIVATION Expressivists are internalists. That is, expressivists hold that there is an internal relation between a judgment that X is good and being motivated to do X (or, between judging that X is good and having some reason among others to do X, or however else you wish to put the point). Indeed, this feature of the view is typically put forward by expressivists as a near decisive reason in favor of the expressivist argument. Here again, the original Humean argument is more or less wholly intact: it is the same argument, essentially, as Hume’s original claim that “reason can never supply a motive.” Facts, or detections of facts, cannot on their own motivate (obviously) and since morality does motivate, moral properties cannot then name facts in the world. On a more modern note, ever since John Mackie taunted what he imagined moral realists to be with the “queerness” of the idea that “to be done-ness” could be “woven into” a fact (and truly, how silly is that?), it has seemed to many that we really should have an account of moral predication in which judgments of goodness are logically, conceptually tied to motivation from the start. If claims of “goodness” express a pro-attitude, the approval or endorsement of some plan, then it would seem at least that difficulty has been met more or less perfectly. Ridge could not put his commitment to internalism more strongly: Belief qua representational state cannot motivate all on its own. It needs the help of desire. The belief that grass is green represents the world as being a certain way, but this does not in itself motivate someone who believes it to act one way rather than another. Even the belief that something will be painful will not motivate someone who is depressed and thereby rendered listless, or thinks he deserves pain… Moreover, our implicit conception of normative judgments as distinctively motivating is not best understood as a merely contingent fact about those judgments. The thesis is rather that simply qua their particular character and content, first person normative judgments are necessarily capable of motivating without the help of any independent desire. Call this doctrine “Capacity Judgment Internalism,” or “CJI” for short. CJI is a necessary truth about the nature of normative judgment. (49-50) Ridge may simply not have noticed that his actual language here is fatally imprecise. To say a normative judgment is necessarily capable of motivating says very little. For, after all, this is consistent with the easy-going counterfactual “if the agent had had the right character, then his judgment would have caused him to act.” (The judgment as such remains credited with this “capacity.”) And now, it is hard to know who could disagree. (No wonder it is a necessary truth.) Later on, Ridge makes the claim more clearly. Internalism is committed to denying the possibility of a moral agent making “full-blooded moral judgments” while at the same time being “entirely unmoved by them.” (51) And this then would seem to rule out the very idea of “the amoralist,” a figure made famous by Dostoevsky, of great use to adjuncts in intro courses, and understood to embody just this possibility. (51-52) Ridge spends a great deal of energy on this particular challenge to internalism. Once again, surveys are cited: here students are asked whether they think a psychopathic criminal really understands the judgments he never abides by, and while many (naively) do say “yes,” questions are raised, naturally, as to whether “understanding that” and “belief in” are adequately distinguished in these thought experiments. So, I understand unicorns have horns, but I do not believe in unicorns. Might something along those lines be the right way to understand the psychopath? (52—54) Armed with considerations like these, Ridge thinks he can battle his opponent to a draw, and perhaps he is right when we take a case as extreme as this. (“The appeal to folk intuition about the possibility of amoralism leads, at best, to a stalemate.” 54) But suppose we leave such high drama aside and ask if there could just be, in everyday life, by ordinary people, what I would call normative detection without motivation. And here the answer seems to be, “sure; why not?” And here again, it will be helpful to move away from the abstract term “good” and take up instead normative concepts more or less criteria governed. Consider “deserves.” It seems very obvious that one could say “yes, I can see that Smith deserves this raise, but I am just not interested in giving it to him.” This certainly does not seem very hard to imagine to me. Or: “It is very compassionate of Peter S. to give to the poor as he does, no doubt about that, but I just do not have that interest.” Consider also the application of normative terms in art, or the law. So, when we see that (or come, as a result of our deliberations given the testimony and the facts, to judge that) the nightclub owner was negligent, what follows motivationally? Absolutely nothing. When a musician says, yes, Debussy’s late piano work is very expressive, but I myself do not respond very deeply to that genre—do we feel he cannot possibly be using the word correctly, cannot know what he is talking about, cannot really understand the concept in play? I do not think so. In the interests of full disclose, I should say that not only do I not think tying moral judgment to motivation in a “logical” (or any excessively strong) way is right, I do not frankly see much point. It seems obvious we can attribute normative predicates all the time without this having any connection whatsoever to motivation or response. And it seems obvious that those who take this line can easily make sense of how we come to be moved by moral argument by a different sort of story. This was Mill’s view, famously—whatever your view of “the good” is, he argued, you will need to tell some further story about how we come, through upbringing and education, to be motivated to do it (“Of the Ultimate Sanction of the Principle of Utility”). That has always seemed right to me. (The only reason you might not think it is so is if you thought “the good” had no specific content, named mere approval; then the argument’s initial premise does not really apply.) But more generally, this dispute is, I think, a largely fruitless one. What happens in these arguments is that inevitably, for the internalist, being motivated is simply made the criterion for “really understanding” that N is the case, by definition. It is not a definition I myself recommend but, like most analytic claims, it is not easily engaged by experience. ECUMENICAL EXPRESSIVISM And now to the heart of Ridge’s distinctive contribution to the expressivist argument, what he calls Ecumenical Expressivism. Ecumenical Expressivism is the view that normative judgments are to be thought of as a kind of hybrid, as being constituted by a desire (roughly) and a belief. But it remains expressivism, and not cognitivism, which of course could also admit of a hybrid form, because: 1) the content of the moral claim is driven by the desire (or “approval” or “proattitude”), and 2) the normative content of the moral claim does not have truth conditions; it is not to be thought of as making any representations about the world that in turn could be true or false. By contrast, the belief in question is a real belief—it has representational content, it is to be thought of in terms of the world to mind fit. And, most gratifyingly, it can be true, not simply in the somewhat evasive sense that is available to us when within the deflationist theory of truth, but under any conception of “truth” you like. This is a substantive motivation behind Ecumenical Expressivism—we have, in it, a form of expressivism that is not hostage to the deflationary theory of truth being deemed right in time (our theory is not, as Ridge would say, “hostage to fortune” here). Rather, for once, it would seem, an expressivist can say moral discourse is “truth apt,” as such; or, regardless of whatever theory of truth you think will be vindicated. Unfortunately, this “advantage” turns out to be trivial. And that is because the “belief” in question is trivial—or more accurately, what this belief represents, what bits of the world it picks out, is trivial, attesting only to what one might call the minimal competence of a speaker’s understanding what his “normative commitments” actually in fact entail. But this is not the most serious difficulty with Ecumenical Expressivism; not at all. The most serious difficulty with Ecumenical Expressivism is that it redescribes normative judgment in a way that makes moral justification, and moral theory, invisible. Justification, supporting one’s judgment by reference to considerations taken to be relevant in the context at hand, disappears. Moral theory, which is simply a comprehensive or wide ranging justification story regarding general principles as well as particular actions, disappears. Instead, what we have is “approval”—and we have it twice over. We approve of a “normative perspective,” and a “normative perspective” is simply an approval function, a brute, not itself to be justified preference for action types of a certain kind, where the kind, hopefully, can always be specified in a suitably nonmoral vocabulary. (It often can’t, but we will get to that difficulty separately.) Here is Ridge: An agent’s normative perspective just is a set of relatively stable self-governing policies about which standards to reject and accept. More specifically, a normative perspective in my sense is a set of relatively stable policies against accepting certain kinds of standards of deliberation. (115) …I propose that to say the judgments of the form ‘X is good as an end’ can be understood as equivalent to judgments of the following form: ‘X would be highly ranked as an end by any acceptable ultimate standard of practical reasoning.’ …However, any token of this [non-representational] belief type is on my view itself necessarily a hybrid state. More specifically, any such token will be constituted by a normative perspective and a representational belief whose contents are linked in the right way. So any token of (1) will be constituted by: 1a) A normative perspective. 1b) The [representational] belief that X would be highly ranked as an end by any admissible ultimate standard of practical reasoning. [W]e then understand the judgment so paraphrased as a relational state, such that any token of that state type will necessarily be a normative perspective/representational belief pair. The content of the representational belief…is fixed by replacing all uses of ‘acceptable’…with ‘admissible,’ where this in turn is defined relative to speaker’s normative perspective. (119-120; additions SR; italics Ridge) Even by the very forgiving standards of current meta-ethics, this last passage is not as clear as it could be. Ridge has changed his account of what is distinctive about Ecumenical Expressivism from earlier versions, but, as the reader will see, not by very much, and for purposes of clarification, it is worth reproducing one of those earlier descriptions here. On the version of Ecumenical Expressivism I favor, normative utterances express: 1. A suitable state of approval to actions insofar as they would garner the approval of a certain sort of advisor. And 2. A belief which makes suitable anaphoric reference back to that advisor. The basic idea is best illustrated through examples. Suppose I am a utilitarian. In that case, my claim that charity is right expresses my perfectly general attitude of approval to actions insofar as they would garner the approval of a certain sort of advisor, which in this case is an advisor who approves of actions just insofar as they promote happiness. My claim also expresses a belief which makes anaphoric reference back to that advisor (the one which figures in the content of the attitude I expressed). In this case, the belief is that such an advisor would approve of charity. (Ridge, “The Truth in Ecumenical Expressivism,” 2009) Parenthetically: Ridge moves from his earlier account, where an agent approves of what some ideal observer would approve of and also (representationally) believes that such approval extends to this case here, to the language of Impassioned Beliefs, with its talk of normative perspectives and what such perspectives do and do not exclude (what would and would not be “admissible” under that perspective) because he thinks such talk better fits first order experience. After all, people usually have not embraced some pre-existing theory wholesale when making a judgment in a particular case, and to have a stable normative standpoint just is best thought of in terms of what is ruled out, what you won’t do—what you will do, or might do, is often yet to be determined, is still “on the table” (115). These changes are not important to us now. What is fundamental to Ridge’s Ecumenical Expressivism—the “hybrid” story, what is and what is not representational—remains exactly the same. Now that the reader has the basic idea, I reproduce here a few more passages that develop it, and draw out some of the more important implications. Because what is expressed is a deflationist belief type and the type in question is multiply realizable, there is no particular normative perspective / representational belief pair expressed by a given token claim. The [non-representational] belief that the wealthy are morally required to give to charity can be partly constituted by a representational belief about consequentialist standards, or about standards fixed by what God commands, or about standards fixed by what treats people with respect… and so on ad infinitum. (196) Normative judgments are therefore massively multiply realizable. There are indefinitely many possible normative perspectives, and so there are indefinitely many relational states [where the particular normative judgment follows from a normative perspective/representational belief pair—SR] which can constitute a given normative judgment. (121) There are two distinct, equally important things going on. One is a set of claims about how representational beliefs figure in normative judgments; they are said to “partly constitute” a normative judgment, and this is what is going to make such judgments “truth-apt,” quite regardless of your particular theory of truth. The other is a claim about “multiple realizability,” that normative judgments are multiply realizable (actually, “massively” multiply realizable) because they can easily be conjoined with a variety of normative perspectives (and so, of course, with a variety of normative perspective / representational belief hybrid pairs). So, to illustrate, the normative judgment token, “giving to charity is good,” can easily be combined with a Kantian normative perspective, a utilitarian normative perspective, and so forth. And of course, each of these will be conjoined with the representational belief that from each such perspective, one would approve of charity. It would be true that one with this perspective would approve of this act, or action type. But the claim that the representational belief “partly constitutes the normative judgment” is, I submit, extremely misleading. The representational belief Ridge refers to here is “about consequentialist standards” (to stay with the earlier example) only in the most minimal and, I would say, trivial way. Let us first make clear what it is not, what it cannot be. It is not a belief that consequentialist standards are well founded, or are justified. That belief could not possibly be representational, on the expressivist view, obviously. No; this “belief” is the belief that, if you were a consequentialist, you would then find this act, whatever it is, implicated by your view. Or: consequentialists, as a matter of fact, do approve of this act. That is the belief. It is the more or less B-or-better answer a student might give on a philosophy mid-term exam to the question “what do consequentialists think?” Because that is the only thing we can say about consequentialism that admits of “representational” character, that can be characterized in terms of that almighty God, the world to mind fit. And that is the degree to which this “belief” is “about consequentialism.” Forgive me, I find this almost ridiculous. It is nothing more than an expression of semantic competence; in holding this “belief,” the agent shows he truly does understand what consequentialism is, that’s all. (The point is simply rephrased when we speak, in the later work, of what some normative perspective allows or excludes.) This amounts to a meager, to me, virtually pointless, difference from those forms of expressivism that simply leave such “representational beliefs” out altogether. Since such beliefs in no way speak to the plausibility or acceptability of the content of the normative perspective, what does it add to expressivism to include them? Why not equally say: “and in addition I know it is I who am making this judgment,” and make that the “constitutive representational belief” that will always be truth apt regardless of your truth theory? It is a novel addition to expressivism, and it has received a lot of attention. But it is also a strange and silly one. It shows the degree to which expressivism is anxious to be able to claim that normative judgments really can be truth apt. And in Ridge’s case, it shows an admirable sensitivity to the danger, not at all noticed by Blackburn or Gibbard, that in relying, for success at this ambition, on a deflationist theory of truth, one may be relying on a theory that in time may become discredited. But it is a mere gimmick; it is no solution at all. This “representational belief” simply takes the implications of the content of the normative perspective as its object. Whatever connection there is between this view and the world, however plausible or implausible the view itself may be, (for this will depend, of course, on the view itself), never even arises. As Ridge himself concedes, the overall “judgment,” with this artificial addition sown onto it, can now be said to be “truth apt,” but nothing about the normative perspective understood as such can be said to be so. Utilitarian principles, utilitarian reasons (or those of the chosen normative perspective, to use the later language) cannot for a moment be thought of as true, and cannot be assessed in this way at all. One’s normative (non-representational) beliefs are no more truth apt than one’s inner desire for an ice cream cone. I should say: strictly speaking, this is fine with me. I do not think, as I will argue below, moral theories should be assessed against “the language of truth,” I don’t think we “unhesitatingly” apply the truth predicate to our normative judgments. The language of justification is the one to use. But the expressivist must pursue this issue against a particular, forbidding framework: there are representational states that mirror the world and so can be true on one hand, nonrepresentational states on the other, and he wants to say moral claims, which cannot in any sense be about the world, nevertheless can be truth apt all the same. It’s a hard square to circle, especially if you don’t hide behind the Unitarianism of the day, a deflationist theory of truth. Now, normally, when we worry about truth in moral discourse, we wonder if the application of a normative judgment-predicate to a state of affairs is right. Is capital punishment (really) wrong; is charity (really) obligatory? Are these claims “true”? This, I submit, is the natural way in which the issue arises. But Ridge has us speaking instead of the semantic entailments of a doctrine. Is it true that the doctrines of utilitarianism require some acts of charity? Is it true that my normative perspective allows me here to lie? I leave it to the reader to decide what sort of advance this represents. What about the claim regarding multiple realizability? One might think this is innocuous enough, hardly worth arguing about. Isn’t it obvious that, as a matter of fact, different (what I would call) moral theories will often reach similar verdicts about similar action types? Sure; but Ridge’s language here is also very misleading. The reality, where moral judgment is concerned, is far more interesting and more complicated. And as a result, the relation between judgment and “perspective” (again, what I would call theory) is no type-token relation, that is for sure. There is in fact already with us an ongoing, stable, normative practice where something more or less identical to what Ridge mentions here is an everyday affair. In legal decisions reached by a panel of judges, there is a category known as the concurrence. One can join the holding; one can dissent; one can write a concurrence. What is a concurrence? It is agreement with the majority that one or the other party does in fact win, but for very different reasons. Justice Kennedy agrees with Justice Scalia and the plurality that Seattle’s racial redistribution plan is unconstitutional, (and so he votes, with Scalia, for the parents who brought suit) but not because he thinks Seattle has no legitimate purpose, as The Court’s decision states. Kennedy thinks Seattle’s purpose is just fine (he joins the dissent in this), but that Seattle failed to show it could not achieve it by race neutral means. So, to use Ridge like language, he makes the same judgment, but from a different (let us call it this, just to keep things simple) “normative perspective.” But is it the same judgment? Is it really a token of the same type? I would not say so. Nor would Kennedy, nor would anyone wondering about the future of such litigation. And here is why: obviously, the two claims admit of very different counterfactuals, and that is not something we should ever be able to say about two tokens of the same type. So, if another city did show race neutral means could not work to achieve this end, Scalia would still find it unconstitutional, Kennedy would not (each makes this point perfectly clear). That is why there is the category of the concurrence to begin with. It is important, not just what verdicts we reach, but our reasons for reaching them. Because the reasons make clear what the verdict really is, what you really hold, when you say “so and so is right.” And lest this seem a fussy artifact of mere legal reasoning, consider the following. Suppose you are a libertarian along Ayn Rand like lines, and I am a utilitarian. We both say “private property rights should be respected.” Fine. Do we mean the same thing? Is this a “judgment token” that in turn is being “multiply realized,” realized alongside different normative perspective/ representational belief pairs? Well, sure, maybe, in one sense. We do assert the same sentence, and we certainly disagree with the imagined Marxist who asserts the contrary. But, because of the different justification stories we hold, very different counterfactuals will hold too, and this, the difference in counterfactuals, is crucial in normative life. I am utilitarian; I am quite at ease with understanding these property rights in incremental ways. I think inheritance tax is fine, adverse possession is fine, various degrees of state interference with property, like use regulations or landmark regulations, are fine. You are a pure libertarian; you do not think any of these things are fine at all. So we won’t together assent as soon as we imagine any of these things arising, (for example, a tax on property inheritance) and asking if they are justified. Do we mean the same thing when we say private property rights must be respected? Do we mean the same thing when we use the word property? If we say yes, it is sort of surprising how quickly we are going to disagree as we continue to talk about property and property rights. And I certainly don’t want to say “no, we do not mean the same thing” in the tired old sense in which “it would then follow” that “we are not in fact really disagreeing after all, because we do not mean the same thing.” How silly is that! We disagree, because we hold different theories of property. This is obvious. It was certainly obvious to Marx. But it is obscured if you focus on a sentence. And to return to the earlier point, the one raised by the very category of the concurrence, why feel we have to say yes or no here to begin with? Why have to answer: “does the sentence mean the exact same thing in both cases, or not?” Why not just say, what I have just said—we assert the same sentence, but because we offer different justifications, hold different theories, the meaning of the sentence is not quite the same for the two speakers. I realize it is a long time since the pieties of Word and Object were on everyone’s lips, but whatever happened to the old saw that sentences face the world not singly but holistically, as part of a web of theory? Change the web, and surely you change the sentence. (Consider: when a Muslim, Evangelical Protestant, and Hindu say “there is a God”—and surely they all do—do they utter the same sentence token and mean the same thing? Let’s see that one worked out then.) In law, the importance of this distinct category is explicitly affirmed from the start; it is woven into how we understand the practice. If you were to ask a lawyer “Do Scalia and Kennedy mean the same thing when they say the Seattle plan is unconstitutional?” he would look at you funny, and rightly. The answer is clear: they reach the same verdict for different reasons. That’s why we have the category of a concurrence to begin with, and why Kennedy writes one here. They “mean the same thing” in one sense, not in another. It is not so hard to keep this more complicated answer in mind. Throughout the past few paragraphs, I have, as the reader will have surely noticed I am sure, used “justification” and “theory” talk where Ridge would speak instead of “approval” and “normative perspective.” And this difference—which speaks to my inability even to describe what goes on in moral life when using only the language expressivism prefers—brings me to what is I think the deepest flaw in Ecumenical Expressivism. Ecumenical Expressivism is unable to make sense of the distinctive nature of justification; indeed, justification, as I have said before, is actually quite invisible in this framework. Instead, expressivism seeks to characterize what goes on in justification using the much weaker notion, “approval.” In this respect, I want to suggest an intentionally unflattering parallel between the expressivist project and reductionism in philosophy of mind, the reductionist characterization of intentional acts solely in terms of physical causation stories for example. Of course, every intentional act is also a physical event, and so admits of that description too. But so what? That hardly means we capture what is central about intentional action in the language of physical causation. We do not. And of course any policy or action I think justified is one I will, obviously, also “approve of.” That follows almost grammatically. But approval is a far weaker notion, and does not capture the nature of justification at all. An illustration here may help. Let us consider the utilitarian argument. Mill does not simply “approve of” maximizing utility. What a caricature that is! Mill thinks this policy is warranted, given certain plausible assumptions. What are these assumptions? Here is a sketch of some of them: we assume that “utility” is a reasonable metric to capture otherwise diverse objects of desire and/or the value agents place on success in intentional action. We assume persons seek a rule to manage their public cooperation and occasions of conflict, and we assume persons will see themselves as entitled to be counted equally in any such rule. Given these considerations, it follows, Mill thinks, that the rule, “have those policies that promote the greatest net utility for the relevant population,” is justified. Certain considerations are pointed to as relevant to the project, certain ways of managing these considerations are put forward as more satisfactory than others. The reasons for such claims of relevance, or satisfactory management, in turn will be tied to how we understand what it is that morality seeks to do. So for example, the emphasis on equality, and the fact that in some important respects, utilitarianism is quite egalitarian—these are considerations that count towards justifying the theory. And the thought is: surely whatever else may be wrong with this theory, paying attention to equality counts in favor of it. And this idea of some consideration “counting in favor” of a theory cannot possibly be rephrased as a fact about approval or pro-attitude. To the contrary, it is put forward as a fact about the theory. But the idea that certain facts, or considerations, might support a theory, that this might be a fact about some theories, and not others, can get no foothold here. We can speak of causation—some facts might somehow “cause goodness.” We can speak of detection, and so, if intuitionists, of “detecting” goodness. If goodness were indeed a natural or non-natural fact, we could have representational beliefs about it. And by contrast, we can speak of “approval,” a non-representational state (of course), in which we seek to settle “the thing to do” (somehow). But there is no place for the central phenomenon, for what writers like Mill and Kant for example actually undertook—pointing to certain considerations in support of a certain conception, or theory, of moral goodness. And this is also what we do, more or less, when elaborating upon our particular moral judgments. We go back and forth between those facts we think relevant and the more general or theoretical considerations that make sense of such importance. No one “just approves of” promise keeping. (Well, a child might be taught to speak this way.) We think it is right because of the respect it shows the one to whom the promise has been given, or because of the stability of expectations such a practice makes possible. No one “just approves” of giving modest amounts of money to the desperately poor. One thinks it is right because of the great benefit given to other persons (who will be thought of as creatures that must count) for little sacrifice. The phenomenon here, generally understood, is not so different from justification elsewhere. I support, I try to justify, my view that Lawrence is a better writer than Forster by pointing to certain facts about each, and by pointing to what I hope are fairly persuasive ways of understanding literature. I don’t “just approve” of Smith over Hoskins; I try to justify my view that this candidate rather than that one deserves an offer of admission in light of certain facts about the candidate and how I understand the point of philosophy. We offer more or less general considerations that we hope take account of the central cases well, and support one particular judgment over another elsewhere. We offer these considerations in light of a theory we are prepared to defend of what is important here, and what is not, and all of this will be subject to criticism. So, when considering utilitarian theory, the consequences of being willing to treat welfare in an aggregative manner is rightly subject to criticism, however much some die hard utilitarian might insist he “just approves.” But the expressivist can make no sense of these central phenomena, justification and criticism, the very essence of normative life. We have only the tools of naturalism—causation, detection, representation, non-representational approval. The background framework makes it impossible to capture the distinctive nature of justification and critical assessment from the start. Under Ridge’s Ecumenical Expressivism, all we get, as I have said, is ungrounded approval twice over. Moral theories, rebranded “normative perspectives,” are understood as mere approval machines. The utilitarian is someone who “just approves” of maximizing pleasure, the Kantian of respecting persons, and so on. Reasons for such conclusions, justification stories, are left out. And agents for their part simply do or do not approve of such schemes of approval. Moral theory is truncated, moral justification disappears. BUT WHAT ABOUT THOSE STANDARDS? When Ridge speaks of “admissible” standards, does this make a difference? After all, justification invariably proceeds against some standard or other. Perhaps, armed with a notion of admissible standards, the accusations above regarding justification can be met. But in Ecumenical Expressivism, to say a standard is “admissible” is not to attribute any intrinsic property to it. A standard is “admissible” just in case if it is “not excluded” by whatever “normative perspective” one has reached by deliberation. (“The content of the representational belief…is fixed by replacing all uses of ‘acceptable’…with ‘admissible,’ where this in turn is defined relative to speaker’s normative perspective.” 119-120) As it happens, it is extremely unclear in Impassioned Beliefs what it means for a standard to be “admissible” under a normative perspective, or conversely, what it might mean for a standard to be excluded by one. From my point of view, this is not that important a criticism, as I do not think the framework by which normative perspectives are generated in the first place is at all satisfactory. But for those attracted to developing Ridge’s project, as things stand, this is a serious instability in the argument and someone should take it up. After all, “exclusion” cannot be the same thing as mere difference. Presumably, my love of things Italian does not exclude an interest in pursuing things Korean, a sensibility it happens I am so far utterly indifferent towards. Does the idea require overt enmity then? Belief in revenge-for-an-insult certainly excludes turn the cheek pacifism. But the line between enmity and mere difference is elusive. Suppose my standards come from a perspective that values highly earning money for my loved ones, yours from a perspective that values solitary religious contemplation. It seems the perspective of one excludes the standards of the other, but maybe not. Could I not have the perspective that endorses doing one, then the other, sequentially? (This is in fact a bit of folk wisdom in the Hindu tradition.) If you are an abstemious libertarian, finding taking drugs and buying sex quite beneath you, but believing in maximal personal freedom, what exactly does “your” normative perspective “exclude”? I myself am at a loss to say. How “exclusion” within Ridge’s standard talk is to be understood, is, I submit, actually a very tricky matter. I am skeptical there is a clear way to fix this idea. But Ridge has little interest in it. An unexamined quasilogical operator (“exclude”) is just going to have to be enough. Obviously, the real issue here is: can normative perspectives, or, the content of the standards they fail to rule out (for they are the same thing), themselves be assessed? And the answer is, yes, sort of, certainly in so far as we draw upon noncontroversial notions of psychology. For a standard to function as a standard of practical reasoning is for it to guide an agent in deliberating about what to intend and do…. Clearly, this construal places no substantive constraints on the contents of standards of practical reasoning. A standard which recommends always maximizing one’s own agony can, for a given agent, be a standard of practical reasoning just as well as a more sensible standard. This is as it should be, since people can adopt absurd standards of practical reasoning…. Equally clearly though we want some way to privilege standards of practical reasoning which we think deserve to be taken seriously from those which do not. (40) It is typical for the expressivist to “concede” a case like this. Sure, crazy standards like these are logically possible, the passage tells us; my theory not only has no ability to rule them out, it would be a mistake even to try. But that (rare, peculiar) fact about rare and peculiar people need not distract us; it is hardly what we usually find. And here Ridge is right—these cases don’t pose any interesting problems. The case a reader assessing expressivism wants to see taken up and satisfactorily handled will be the less exotic, all too familiar one where reasonable self-interest and the morally awful intersect nicely. Consider a man with the standard that seeks to preserve the second class status of women. If this is not a “sensible” standard, I need an argument to show me why. It has made many far happier over the years than the pursuit of agony; that is for sure. It might seem that I am now very far from the criticism I started with, that expressivism is driven by the unsatisfactory framework it uncritically inherits from Hume. After all, for Hume, our values (our standards) would have been thought as largely fixed by nature. Deliberation would take up the best means to ends that were themselves (almost always) generated by causation and not subject to assessment; in the expressivist argument, deliberation takes as its object final ends themselves. But the Humean framework is driving expressivism here all the same. Because normative judgment cannot be representational—and this is quite right, given what “representational” can mean when within this framework—if there is to be deliberation about ends, about standards, such “thought” cannot be criteria governed. There is a lot of evasion before this point, much of I think simply instinctive (not calculating); it is just not an appealing part of the view that it has no real tools at its disposal to handle what we might call the wrong sort of relativism. By that I mean, given how “good” or “ought” or “acceptability” are understood when within expressivism, the theory cannot characterize a plan or standard as “wrong” except as relative to whatever other standard that one’s deliberation has led one to. That means the accusations of moral error moral antagonists will make of one another must always be logically on a par. Now sometimes this is just what we want. It is one thing to say “on my theory of how “morally right” or “the right thing to do” is to be understood, there will be acceptable stories both defending and condemning capital punishment.” That seems quite right. But it is quite another to say, “on my theory of how “morally right” or “the right thing to do” is to be understood, there will be acceptable stories both defending and condemning second class status for women.” That does not seem right at all. But the expressivist must resist any talk of the nature of persons, or the nature of morality (or anything else), generating criteria for satisfactory (or unsatisfactory) norms. For in that case, normative deliberation would be (or should be) about something, and we might as well just turn to what that something is. The deliberation would be “derivative,” and the expressivist project would be doomed. Consistent with the expressivist project, deliberation as such is all we can offer. Value cannot be out there, cannot be a natural fact or a metaphysical entity (this, happily, is true). Since value cannot figure in a world to mind representational belief, there is nothing “out there” that can govern what goes on inside our heads when we wonder about “what is best.” The result, again, is that deliberation itself, and what it aims at, is all we can point to. Normative judgments function to settle “the thing to do” and normative conversation allows us to deliberate together about the thing to do and enhance our individual deliberation while also better coordinating our joint efforts toward common goals. (101-102) This ghastly, bloodless empty talk; always abundant in the writings of expressivists. Whatever can it mean? What, more pointedly, if anything does it rule out? One thinks of the young Wehrmacht officers, getting together, in a clubby, congenial fashion on chilly autumnal evening in Poland, to think about all the things they had to do, and to enhance their individual deliberations while exploring their common goals. Across the border, NKVD apparatchiks worried how best to advance the starvation in the Ukraine, meeting far into the night to discuss techniques for disgorging hidden grain from those uncooperative peasants. Readers may well find these remarks uncharitably harsh; I ask the reader however to remember they describe actual events, not made up fantasy. And truly, they are quite continuous with Ridge’s account of “locating the normative” in simply worrying about having a child with which we began. AND FINALLY, A FEW WORDS ABOUT TRUTH One of the interesting ways in which current expressivists truly do differ from the original Humean inheritance is on truth; specifically, on the capacity of moral judgments to be thought of as truth apt. Expressivists may share Hume’s view about how the world is to be divided—various natural and metaphysical facts out there, reactions and attitudes in here—but Hume thought of our moral responses not simply as non-cognitive, but as states with no semantic content, and so as states that could not be thought as true or false. The emotivists, Ayer and Stevenson, followed Hume in this way, famously comparing moral judgments to interjections or exhortations. And this leads to easy dismissal, for when understood this way, moral judgments cannot even figure straightforwardly (as they surely do) as objects of quantification, in embedded inference or implication claims. (And this of course is the famous Frege-Geach problem, about which I will on principle have absolutely nothing further to say.) Current expressivists (Blackburn, Gibbard), usually affirming what has come to be called “quasi-realism,” struggle strenuously to avoid this result. Moral claims can, it is said (by the quasi-realist at any rate), be both non-representational and truth apt. When you put it that way, it is pretty clear that such views will be happiest if “truth” turns out to be something not signifying very much, which is to say, understood along deflationist lines. Ridge sums up the quasi-realist position here nicely: It is easy to see how this sort of deflationism [Blackburn’s, where calling a sentence “true” is understood dispositionally, in terms of what inferences are and are not warranted - SR] might seem like a powerful tool for quasi realists. Suppose there really is nothing more to saying that it is true that charity is good than there is to saying charity is good. This suggests that the expressivist can allow that normative sentences are truth-apt. The point will simply be that in saying that it is true that charity is good one is expressing one’s attitude in favor of charity in just the same way that one does when one says that charity is good. (202) It is to Ridge’s credit that this is not very satisfying for him—though this is largely because he thinks, rightly, there are many general difficulties with deflationism as a theory quite apart from what it says about normative discourse, and it would be a pity if a theory as near-flawless as expressivism were to be so closely tied to deflationism’s uncertain fortunes. I have already rehearsed what it is about normative judgments that would be truth apt, according to Ridge, and why on my view this account adds so very little. There is no need to go into that again. I want instead to conclude with a few reflections on a remark Ridge makes when ruminating on the puzzles that attend thinking about truth in normative discourse generally, before he turns to his own view. Ridge considers a possibility rarely entertained in these discussions, and then unsurprisingly, quickly sets it aside. But on my view, this (to Ridge) almost crazy idea may contain the seeds of what is exactly the right approach. Mark Richard has argued at some length that expressivists should not be so afraid of denying truth-aptness… [H]e usefully reminds us that ordinary speakers do sometimes balk at the inference from ‘p’ to ‘p is true’. The fact that such speakers sometimes speak loosely and use ‘true’ as a grammatical device to indicate assent to a whole class of propositions does not show that they think those propositions are really true. It is a platitude that people often speak loosely if it is a more efficient way to get their ideas across, after all. Richard himself is agnostic as to whether the logical space at which he adverts is plausible in the case of moral thought and discourse, and for good reason. In the moral context, I doubt very much that ordinary speakers tend to use ‘true’ only loosely and as a grammatical device, where they would (under Socratic pressure, say) distinguish this in some way from “real truth.” If “is true” is always a mere grammatical device, then it will be a grammatical device in these contexts too. What is implausible is that some more robust conception of truth is correct more generally and available to speakers, but that competent speakers do not intend to invoke that robust notion in normative contexts. (197) Before taking up what Ridge terms implausible and what I think is in fact a genuinely promising framework, I cannot resist noting the following. Has anyone ever noticed that expressivists are always trying to convince us that, on one hand, it turns out that “ordinary people” simply have these brute unjustifiable feelings driving their moral judgments (“I don’t know why I think this; incest is just nasty!”) and at the same time, “unhesitatingly believe” their normative judgments are “truth apt”? Of course there is only one way for both of these to be true at the same time—all of us, it turns out, just are, instinctually, expressivists who hold to a deflationist theory of truth. Like Socrates exposing the Platonic knowledge even a slave has of the Forms, expressivists have revealed the deep, true, philosophical commitments that just happen to be instinct in us all. (As if.) Well, I am skeptical. The expressivist always insists ordinary people “unhesitatingly” use the language of truth before their normative discourse—this is why expressivists always quickly distance themselves, patronizingly, from A. J. Ayer. But I think this claim needs to be examined, not repeated. Let us leave the world of surveys of college students and look instead at some ongoing, stable, structured, normative practices, say literary criticism and legal interpretation. We can, and will, return to the students afterwards. I am very much not sure that those within these disciplines would say they use “truth” in the “ordinary” sense at all. Of course, Justice Kennedy would say, I think it is true that my reading of Concerned Parents v. Seattle Schools is better than the Court’s, my understanding of “legitimate Constitutional purpose” is more in line with what the post Brown v. Board of Topeka cases require, but to say my reading is right (more awkwardly, “is true”—an expression I venture to say he would never use, much less “unhesitatingly”), is to say my reading is more justified. It is not to say there is some fact I am “detecting,” for there is not. It is certainly not to say Scalia’s reading is false, in the sense that it has no merit whatsoever. And this brings us to an interesting, important, and rarely acknowledged asymmetry in “ordinary usage” between truth talk and justification talk, and one of the reasons why people often in fact do balk at using truth talk in normative life, particularly in its “more robust” sense. When I say it is true that the earth revolves around the sun, I am also saying it is false that it doesn’t, and so it is false that the sun moves around the earth. The fact, presumably, just is one way, and we represent it, correctly, or not. Here is the familiar framework of the world to mind fit, and for every unit of representation that seeks to picture the world a certain way there can be no “degrees” of truth and falsity. But before complex intentional structures—the law, the novel—there often will be a range of legitimate readings available. And so the real central term of assessment here—“is justified”—will admit of degrees. This is a very important part of how we understand normative judgment. In saying Scalia is wrong, Kennedy is saying that his reading is more justified than Scalia’s. He is not saying his reading is “true” and so Scalia’s must be “false.” This is obvious. He is certainly not saying that Scalia’s reading cannot be justified at all, or even that it does not admit of a relatively good justification story. He is saying simply that his justification story is at least a little bit better. And the result is that ordinary people, knowing this fact about justification and normative judgment, often do hesitate to use the language of truth “in its more robust sense” before the normative judgments they put forward as right. They do not wish to imply that the opposing, or contrary, or rival judgments are necessarily without merit. (My students at least talk this way all the time.) It would be very interesting to see how the expressivist would characterize grading, that very ordinary business of assessing the written work of young philosophers. (And surely this is a normative practice if any is, though I am not sure anyone is “settling the thing to do.”) Is it that each sentence is either true or false (this is the language of assessment we “unhesitatingly apply,” remember?) and we add up the net average? No; this does not sound like a very plausible account of grading to me. Rather we look at the degree to which the claims are justified. Yes, that sounds like a better description to me. I submit, before the expressivist can tell me with confidence what in fact goes on in moral life, let me hear his account of grading and find that plausible. I am quite serious. Now, of course, there are some deep, important asymmetries between moral judgment and grading, or the law. Arguably, the standards in moral life are less settled (but one should not exaggerate the degree to which they are fixed in grading or legal interpretation either—obviously, or it would not matter so much which President appoints which Justices). And sometimes, in any normative domain, there will be cases where we feel there is no rival to a particular justified judgment. For many, this is when talk of “truth” seems most natural, i.e. when no plausible rival justification story is on hand. And these are the examples beloved in meta-ethical debate, particularly, now, by so-called “realists”: it is true that “Rape is wrong!” it will be said, and with a particular tone of voice too. I do not dispute it, I do dispute that this shows what it is typically taken to show. It does not show we are detecting a mind independent fact and that our judgments are truth apt because of that fact. It is instead more plausible to say that here, no good justification story can challenge the justification (or condemnation) story on offer. We rarely see as an example of “truth-apt” normative discourse a sentence like the following: it is true that “Capital punishment cannot be justified!” One wants to say to that last remark: Really? Not at all? That can’t be right; of course it can, to some degree. But that sort of talk is exactly what is not available to us when within the more robust conception of truth that appropriately operates in naturalistic discourse. If p is true, then not-p must be false. But if Kennedy’s view is well justified, well supported it does not follow that Scalia’s cannot be justified at all. And I think in fact, far from “unhesitatingly” reaching for the truth predicate when complimenting their moral judgments, most moral agents are instinctively attuned to the scalar and non-exclusionary nature of justification, and so instinctively hesitate, instinctively feel squeamish, using ordinary notions of truth or falsity to modify the many normative judgments they make, for they also know the rivals to these judgments are usually not without some plausibility too. OK; let’s do what expressivists do all the time. Let’s just assume that the account of normative discourse I just gave is right. What, from the standpoint of truth, follows? Well, very little actually. One could be deflationist, of course. Being a deflationist about truth as a predicate leaves wide open how you think assessment operates within particular domains. Deflationism simply tells you: don’t look for a “theory of truth” to tell you how “is true” operates in science, in law, in etiquette, in games, and so forth and so on. As Wittgenstein would have said, don’t theorize here; just look and see. And that would be fine with me; on my view of normative discourse, the central term of assessment here is “is justified,” and any truth talk that made any sense, that in any way seemed natural, would always refer to that concept, would always have that term embedded in it. So we could certainly say “Kennedy’s reading is more justified than Scalia’s” is true. Or, in the moral case: “liberal egalitarianism is more justified than the caste system” is true. And these sentences would mean no more and no less than Kennedy’s reading is more justified than Scalia’s, and that liberal equalitarianism is more justified than the caste system. That would be fine (that would be, in fact, how it is). But one could equally well say: “When I talk of truth, I want a more robust notion, something about “matching the world” where it is understood that sentences either do or do not do this. Normative discourse, with its commitment to justification, and degrees of justification, can’t be thought of this way.” That would also be fine with me. I cannot see why it matters from the standpoint of normative theory. So long as we do not disfigure what normative life is in fact like to fit our theory of truth, so long as we are faithful to its distinctive structure, we should be easy going about what conception of truth in the end is deemed the right one. If it turns out that there is a general theory of truth that actually fits what goes on in each domain, that each domain can be plausibly thought of as an instance of this general account, fine. If it turns out that the best theory of truth is one that can be said to apply only in some domains, and in others, we are better off speaking of other notions that are not (on this account) quite the same as truth, like “justification,” that would also be fine. What matters is simply being faithful to the underlying domain. Where moral life is concerned, this means that what counts (above all) is doing justice to the distinctive structure of justification, not turning it into the detection of some fact, natural or otherwise, or the act of approval. Get the domain right; get the domain right; get the domain right. But the expressivists never do.

This is a preview of a remote PDF:

Steven Ross. Review of "Impassioned Belief", Essays in Philosophy, 2015,