7. IS GIBBARD'S VIEW PSEUDO-GALILEAN?

     This question, if I've done my job, should seem superfluous. Gibbard holds his view to be an exercise in, and an advancement of, Galilean reduction. Yet I have presented a number of reasons for the reader to believe that Gibbard has succumbed to some very ordinary anti-Galilean influences.

     I will need to argue against the subjectivist perspective, but most of the other arguments have already been presented. I should provide a synopsis of these, (and expand on them briefly where additional material will be valuable), and present additional evidence from Gibbard's book revealing a pseudo-Galilean perspective.

7.1 Signs of Anthropocentrism and Mistaken Essentialism

     As I've already discussed, Dennett (1995) argues persuasively that mistaken essentialism (of which I identified anthropocentrism as a special case) is a prime influence against adopting a fully scientific stance with respect to the origin and nature of human qualities. Here is additional evidence that Gibbard has fallen victim to these.

     When discussing inclusive fitness³² and telos (the apparent aim of Life to maximize inclusive fitness), Gibbard goes so far as to call telos a "surrogate of purpose" (28 note 5). But a surrogate is one who stands in for, and performs some function of, another. Telos is simply the largest expression of Life's "ambition" to be all that it can be. Life cannot motivate us directly, but requires (and developed) a surrogate to do the job: an internal motivational state we call 'purpose'. Indeed, our only scientific account of human purpose is as part of a repertoire of desires, appetites, and other impulses that developed through the evolutionary process -- leading us to behave in ways that served our telos. It would appear that human purpose is a surrogate of telos instead!

     The anthropocentrism is almost palpable when Gibbard pointedly separates humans from beasts in his discussion of conflicting "animal" and "human" control systems (56). Investigation of this conflict leads him to examine not the fundamental functional (neuropsychological) differences of control systems, but rather the differences between those control systems that directly involve, and respond to, human language, and those that don't. That, it seems likely, is why he restricts 'normative governance' to language-dependent processes such as normative discussion.

     I mentioned at the outset that Gibbard illuminates some proximate causes for normative behavior, but not the ultimate cause. By now, my meaning should be clear: Gibbard was concerned with why those practices were adaptive, and not with why adaptivity matters. As a result, he decided that an adaptive pay-off, coordination, was at the heart of things. Further investigation, I claim, yields the reason adaptivity matters: there is objective, intrinsic value in the world (in the form of reflexive functionality -- Life and living pursuing their own ends. Life's search for adaptive means to fulfill its intrinsic purpose is the ultimate cause of human normative behavior.

     Coordination is naturally, but mistakenly, viewed as a feature of the dynamics of the current environment rather than as a product of the evolutionary design process. Since we don't attribute morality to mere animals, the argument might go, we need to find something peculiar to humans to account for morality -- say, language. And if language, as a quintessentially human acquisition, allows us to engage in normative discussion, then normative governance is a matter of linguistic endeavor.

     But the benefits of internalized norms for our primate cousins (and ancestors) probably consisted largely in improved coordination. Is the acceptance of human norms supposed to provide a quantum leap in benefits from coordination? I doubt it. In fact, internalized norms are by nature much more powerful than accepted ones, for they can govern us without our acceptance.

     "Now if we had one master source of moral concern -- benevolence, say, or respect -- things might be straightforward. We might find a clear rationale in that source of concern, and just adopt the rationale." Here again (254) Gibbard is looking to current circumstances for his "master source" -- not to evolutionary design. He sees an array of competing impulses (benevolence, reciprocity, respect, etc.), of which none has clear title to pre-eminence, and concludes the absence of a chief impulse equates to the absence of a master source of moral concern. He doesn't consider that all of the competing impulses may have been produced over time as responses to a master source of moral concern in nature -- say, the universal importance for organisms of living and reproducing (intrinsic values, I claim, whose pursuit accounts functionally for all the impulses, desires, purposes, etc., experienced by living organisms).

     This is tricky, for Gibbard certainly does recognize a role for evolution in producing such impulses. He just doesn't see that role as being the master source he seeks. Instead, he sees evolution as the progenitor of our highly social, lingual species, accounting for our shared propensities (including some for normative thought and discussion) ... period. The interesting answers to normative behavior he seeks in the immediate activity generated by human psychology and social interaction -- in particular, in normative discussion.

     This attitude is a textbook example of mistaken essentialism.

7.2 Evidence that Value is Subjective for Gibbard

     For some readers, I might as well be waving at evidence that Gibbard is bipedal. It seems obvious that intrinsic value inheres in human experience, and it would be rash to suggest that any eminent philosopher should think otherwise.

     On the one hand this standard view makes my immediate task quite easy: demonstrate that Gibbard has an attitude that should be expected of him. On the other hand, this is probably the greatest obstacle to the acceptance of my own view, for I claim that what seems obvious about subjective value is actually an evolutionarily useful (and perhaps functionally necessary) mistake. That will be our next topic to consider ... in a moment.

     Gibbard has little to say directly about value, but several examples should serve my purpose in proving a point that Gibbard would probably concede anyway.

     In considering how a plausible story for feelings might go (279), Gibbard begins with an implicitly subjectivist option: "First, we might try a pragmatic story: a story of how norms for feelings contribute to a life worth leading. ... We experience life through our feelings, and so norms that endorse our feelings enhance our experiences." This makes it clear that, for Gibbard, what makes a life worth leading isn't primarily, say, the contribution it makes to the flourishing of its parent process, Life, but rather the quality of one's experience of that life.

     This is underscored in the second (and last) example I'll offer here. While worrying about potentially destructive requirements (such as consistency) of moral system (324), Gibbard asks,

Suppose we found a system that stood up rigorously to all challenge. Might that drive us away from a morality we can thrive on? Are the judgments we would attain ones that could sustain us in the flux of life? Could they give our lives the sense of meaning our old, messy prejudices gave it? Perhaps rigor in moral thinking must lead us down the paths of Sidgwick and Harsanyi (or perhaps, instead, of Kant), and those are not systems anyone could fully accept and still live a life most worth leading.

The issue, then, is partly what is needed to live with felt meaning and enriching commitments.

     The value Gibbard cares about here lies in the feeling and the enriching rather than in the meaning and the commitments -- an unsurprising subjectivist perspective.

7.3 A Case Against Subjective Value as Intrinsic Value

     It's high time I explained just what is wrong with the subjective perspective, the traditional presumption that intrinsic value inheres in experience. My argument has three prongs. First, both the evidence and our body of Galilean theory support the view that subjective valuing has played an exclusively instrumental role in the evolution of creatures like ourselves; furthermore, the Galilean body of theory cannot in principle account for any evolutionary mechanism by which a new purely subjective variety of intrinsic value could be introduced into (or spread through) a population, or for the evolution of any faculty other than full-blown reason that could recognize or appreciate non-biological intrinsic value. Second, the claim that being the object of a subjective, personal preference or judgment of importance is the defining characteristic of value fails because such defining characteristics must be both necessary and sufficient for their possession to point out, unambiguously, the defined trait -- and subjective valuing is neither a necessary nor a sufficient characteristic of value. Third, the alternative subjectivist view that subjective states such as happiness are intrinsically valuable (and perhaps the source of all value) defies Galilean explanation -- for the emergence of new intrinsic value in living organisms can't, in principle, come from an evolutionary (selective) process.

     I think our best account of nature and the evolution of both life and human experience provides two very important things: (i) evidence of something that we can (and should) call 'value', inhering in a matter of natural, material fact: the essential nature of Life; and (ii) evidence that "values" as traditionally conceived were instrumentally (but not intrinsically) valuable during their development.

     I discuss the nature of Life and its claim to intrinsic value in the next section. As for subjective values, there is little disagreement as to how they evolved for our pre-human ancestors. Valuing those things that contributed to an individual's survival and reproduction was adaptive, and propensities for apt valuing were therefore favored through differential reproduction. In other words, we have a capacity for subjective valuing because it was instrumentally valuable to our ancestors. This theory is supported by much evidence from neuroanatomy, ethology, and experimental psychology (parts of which are scattered throughout this paper). More to the point, as far as I know it's the only surviving scientific theory we have on the subject.

     In light of this evidence, if we accept it, we must ask whether such experiential "values" continue to have purely instrumental value, whether some of them have become "empty" (having lost their instrumental value but acquired no intrinsic value), and whether any of them (especially modern "human values", which will be primarily at issue) have justifiable claim to intrinsic value independent of their biological origins.

     These questions fly in the face of deeply ingrained subjectivist attitudes. For virtually everyone, it is obvious that the beauty of a sunset and the tart flavor of a newly ripe peach have value. We value them, and that's the end of the story! But I think this obviousness is on a par with the way the flat Earth, the Earth-centered universe, and the direct perception of external objects have been obvious -- yet mistaken.

     I'm aware of four defenses for subjective value, each seriously flawed. The first is self-validating intuition -- reflexive obviousness or self-evidence. The simplest rebuttal is to reject this as a gratuitous assertion. Though logically effective, that may not satisfy, however. So, let us consider an epistemic argument.

     It is impossible to distinguish, simply by the quality of our experience, purportedly intrinsic subjective value from the ordinary intermediate kind ... or, for that matter, from pathologies -- psychological malfunctions that generate illusions of intrinsic value. This point bears reiteration. Without reference to objective standards, there is no way to tell whether seeming intrinsically valuable is the result of true apprehension of intrinsic value or of misapprehension, pathology, or hallucination. A variation on Descartes' radical skepticism, applied to the experience of subjective value rather than to sensory experience, is called for -- and Galileans will not be able to resolve the problem by an appeal to God's goodness, which itself constitutes the second defense: God would not deceive us, so subjective value is genuine and unproblematic. This requires proof of God's existence, for starters -- something we still await.

     The third defense is the converse of the naturalistic fallacy: our lives would be impoverished if we were wrong about subjective value, and that would be unacceptable -- so we must be right (ought implies is). The fourth is democratically vindicated intuition: subjective value is obvious to most people, so it must not be a mistaken notion. Like other rhetorical bandwagons, this one lacks wheels. It rejects the possibility (the actuality, in this case, as we shall see) of systematic error.

     The near-universal belief in subjective value as intrinsic value, however, is more than understandable. The survival of our mammalian ancestors depended on a psychology in which the kinds of value were systematically confounded. This really shouldn't come as much of a surprise, for we already know that our common-sense conception of most (if not all) of our phenomenal experience is simply wrong -- as Gibbard notes (277) -- but in a useful way. This common-sense view is called 'naive realism' by D.L.C. Maclachlan (1989, pp. 43-48). We naively think the material objects of our experience have colors and flavors and fragrances, but they really have only characteristic molecular structures to which our senses respond (indirectly where there is an intervening medium) in highly reliable ways. We naturally imagine that we experience a Euclidean world of three spatial dimensions and a temporal one, but we really experience patterns of neurochemical activity (or, some argue, supervenient mental states) quite different in kind from the purported objects of our imaginings.

     The Galilean core explains both the nature of these errors and their utility: by responding appropriately to representations of important features of our environment, we tend to act appropriately with respect to that environment itself. This works only because we mistakenly (in an epistemic sense) treat our representations of the environment as the environment itself.

     There is a great irony here. We have come to accept the Galilean account of how our exquisitely vivid sensory experience of the external world is fundamentally mistaken; yet we cling to the notion that our experience of subjective value -- as controversial, indistinct, and mysterious as it is -- somehow directly apprehends a metaphysical truth! We are long overdue for a reassessment of this view.

     How is it that we systematically confound the kinds of value as we respond to our environment? We might call subjective value (which our early, pre-human ancestors experienced as appetites, desires, etc.) a variety of second order intermediate value. The benefit from satisfying an appetite (such as hunger) is that in doing so we acquire a first order instrumental value (food, say) that directly contributes to some intrinsic value (in this case, staying alive).

     We pursue subjective value for its own sake, as though it had intrinsic value; the pursuit leads us to an associated (but qualitatively different) instrumental value, which in turn contributes to acquiring or maintaining an actual intrinsic value. I call this a 'displaced response strategy'. It is how subjective value worked for our pre-human ancestors, and how (I think we must acknowledge) it clearly works to satisfy our own primitive needs. Can we justify the claim that it works differently with respect to more sophisticated modern human interests? I think this turns out to be an epistemic question we will need to examine in more detail later.

     If we are naturally mistaken about subjective value, why might we have persisted in our error for so long? The answer has to do with the notions of semantic transparency and opacity discussed in section 5.2. It is comparatively simple to recover from a mistaken belief in something like the direct perception of external objects; one must only realize that the direct objects of our experience are phenomenal representations of material objects rather than those objects themselves, and that the error arises because the phenomenal representations are semantically quite transparent -- it is almost as though we can "see through" those representations to the objects themselves. Because the representation resembles the object, a response to the representation functions well as a response to the object.

     Recovering from a mistaken belief in subjective value is much more difficult, however, for the representations involved are highly opaque. We respond to hunger -- not to a depletion of nutrients; the experience of hunger in no way resembles the material state of affairs in which cells are running out of fuel. Subjective value, when it functions as designed by evolution, naturally represents objective value: its existence reflects the existence of objective value.

     It will be tempting to object that these concerns may have little or nothing to do with modern values not directly related to survival or reproduction. My desire to compose a symphony, or the value of doing so, represent absolutely nothing in the world; they simply are. If they did represent anything, it could only be themselves. What they represent, however, is the existence in our evolutionary past of kinds of things to which we had adaptive responses -- things which happen to resemble certain ingredients of a symphony: sounds similar to the human voice (particularly protracted cries evocative of important circumstances), rhythms of a mother's heartbeat or of one's natural stride, or the impression of consonance or dissonance of a harmonic interval that is imposed in large part by our perceptual apparatus.

     For now, I want to engender some healthful skepticism about subjective value based on these evolutionary grounds: (i) Its functional role has been as a second order instrumental value masquerading as intrinsic value; its success has depended on deceiving us that it is intrinsic value -- so why accept it at face value now? What reason could we possibly have for believing that our valuing mechanisms function differently for our idiosyncratic values than they evolved to function for our biological ones? (ii) Though subjective value might have taken on a new function other than the one it evolved to perform (as various traits of organisms occasionally do), we know of no such accidental function that has emerged because of its intrinsic value; the known examples have proved instrumentally valuable to the organism's survival or reproduction.

     Let's consider the latter point briefly. How might a new variety of intrinsic value spread through a population of human or pre-human creatures? Not by natural selection for its intrinsic value, because intrinsic value is not responsive to natural selection -- which can only respond to value that is somehow instrumental in promoting the intrinsic values of survival or reproduction. Certainly it is possible that an intrinsic value happens to have an instrumentally valuable aspect, or otherwise to be linked to an instrumental value -- thereby enjoying a selective advantage and spreading accordingly through a population. But the instrumental value (or component of value) does all the evolutionary work here; the intrinsic value is just along for the ride. Any argument for an evolved intrinsic value must treat that value either as an accidental aspect of something that is instrumentally valuable, or as an accidental companion to some instrumental value. If we concur with the venerable judgment of philosophy that accident explains nothing, then the notion of newly evolved purely intrinsic value seems to be in trouble.

     There also appears to be no available evolutionary path for developing the recognition or exploitation of intrinsic value (save the sort of full-blown rational assessment in which we're now engaged). Nature rewards adaptations which are instrumentally valuable for living and reproduction. Some new method of direct apprehension of the intrinsic value of life, say, would be a good candidate to spread through a population once introduced -- but where could it possibly come from? It would be astronomically improbable for such a thing to appear as a single, large-scale variation in a species. That leaves the standard mode of evolutionary development, whereby a new trait is acquired by a small departure from previously useful attributes -- but what could these possibly be, if not the attributes of reason itself?

     What could be not quite capable of directly apprehending intrinsic value, yet useful in some other respect? We can rule out a sensory apparatus, unless there is something similar to (but not quite) intrinsic value to which it could have responded. Intuition is another possibility, but Galileans require an account in terms of brain function -- say, the pattern-recognition capabilities of the right cerebral hemisphere. Inference from recognized patterns, though, is not direct apprehension, but indirect conjecture. And if we're resigned to indirect apprehension, perhaps again reason is the most reliable alternative; I think it is.

     In the interest of full disclosure, I should mention here that I think there is a class of subjective value that is indeed intrinsic in nature -- but it is intrinsically valuable to itself rather than to any living organism that might hold it, though it may also be of instrumental (objective) value to the organism. Indeed, having instrumental biological value was how it developed. What Richard Dawkins (1976) calls 'memes' -- the psychological analogs of genes, which actively work to reproduce themselves culturally in the same way that genes actively work to reproduce themselves biologically -- include what we ordinarily think of as subjective values and the norms associated with them. Readers unfamiliar with memes may find the notion bizarre or even nonsensical; perhaps my functional analysis of value in the next section will make it seem more plausible. At any rate, because the ordinary meaning of subjective value concerns value to the organism experiencing it rather than value in and for the subjective value itself, my foregoing objections to ordinary claims of intrinsic subjective value are not undermined by this exception.

     It's clear that many things we value -- basic needs like food and social belonging -- still have instrumental value for us. That value is objective, unbeholden to any confirming judgment: food keeps us alive, and social belonging offers many benefits -- mating opportunities, division of labor, common defense, education, etc. It's also clear that we normally treat many such things as having intrinsic value, though upon consideration from a Galilean perspective we acknowledge that their contribution to values such as survival and reproduction accounts for their existence. Their continuing role as intermediate values (at least in many familiar situations) is confirmed.

     Let's move now to the second prong of my argument. If value were purely a subjective matter, then subjective valuing would be both necessary and sufficient for a thing to have value. Let's consider these criteria.

     Is the act of valuing necessary for something to have value? Antibiotics, we know, are of value in keeping an infected child alive (a goal few would dispute as valuable), whether or not the child knows what antibiotics are or cares a whit about them. This would still be the case whether or not any other person knew or cared about antibiotics. An appropriate dose of the stuff introduced into the child's bloodstream (by unknown means philosophers are allowed to gloss over) would benefit the child despite universal ignorance of what was happening. We must conclude that valuing is not necessary for value.

     Is the act of valuing sufficient for something to have value? Probably not. There are healthy, able-bodied individuals in our society who value the state of intoxication enough to wish it could be perpetual for them. Most of the rest of us tend to regard habitual severe intoxication as an unqualified liability -- as having unquestionably negative value. If we are subjectivists, we might say that drug abuse is a value for addicts, but the value is somehow "empty". Taking a different view, we might say that the addict's valuing is mistaken in one of two ways: either the addict's various values are inconsistent (and presumably the inconsistency could be satisfactorily resolved by forcing the rest of the addict's values to conform to his high regard for intoxication), or the addict is not responding appropriately to the concept of intoxication. This exposes an equivocation in common usage concerning subjective value. Is the value the object itself -- the pleasure we experience or the ice cream we crave -- or is the value a response to the object (or to some property or properties of the object)?

     Each of these three alternatives undermines the hypothesis of subjective value. (i) If value can be empty, we seem to have the oxymoronic notion of value lacking value. (ii) If mistaken value is mere inconsistency, then we legitimize any coherent set of "values", including perhaps the logical complement of our own; there would be no grounds for preferring one coherent set over another -- but, as Gibbard points out, people view their normative judgments as objective, as interpersonally valid (155). We clearly do claim that some sets of values are better than others. (iii) If mistaken value is a misapprehension of the (objective) value-defining properties of the object we value, then value can hardly be considered to be subjective in nature.

     Subjective value is vulnerable to a collection of pitfalls reminiscent of moral relativism. There are no grounds for preferring one set of values over another; there are no grounds for recommending any particular value to other people; value is personal, not interpersonal (and therefore hardly a useful topic for the normative discussion that is central to Gibbard's account). And why should behavior conform to values, when values can so easily conform to behavior?

     We could certainly decree that 'value' is subjective, and assign some other term to represent what I treat as objective value -- 'benefit', say, or 'utility' (though it would be difficult to make room for intrinsic objective value, since these terms are so transparently concerned with instrumental value). Semantics is after all a matter of definition. But we want our concepts, and our terms for them, to be theoretically useful in developing and maintaining a metaphysical framework that explains our experience. It is also helpful to use definitions that correspond to the most important meanings already in use for our terms -- and, of the meanings associated with 'value', I think objective value is not only the most important, it is fundamental and indispensable. I have explained how it is the source of subjective value.

     It is also the standard by which subjective value is naturally measured. In matters of purely personal concern, subjective values can be ranked according to how close they come to the importance of continued living -- and if someone ranks some such value higher than his own survival, we might question his sanity (unless that value concerns Life in the abstract). Gibbard uses the example (165-166) of the anorexic who would truly accept starving to death as the price of a very slender figure. Similarly, in matters transcending the personal, whoever values something more than the continued survival of Life in the universe might be likewise psychologically suspect.

     Let me summarize the case that subjective value has no legitimate claim to primacy for humans.

     The act of valuing is not necessary for an object to have value, for there are valuable objects of which we are unaware and therefore fail to value. The act of valuing is not sufficient to confer value on an object, for we can be mistaken in our valuing. Subjective valuing, then, can't be all there is to value. Worst of all, evolutionary theory (thoroughly confirmed so far by the empirical evidence) shows that we are designed to respond to objective values in the world around us by constructing internal motivational states that we experience as subjective values cognitively associated with those objective values. Experientially, we treat these subjective values (mistakenly and systematically) as fundamental -- as intrinsic -- which is the way subjective value functions as an instrumental means of acquiring objective value. Though some modern subjective values might have an intrinsic component, evolution can account for that in only two ways: either arrival by accident, or linkage to some instrumental value.

     For claims of intrinsic subjective value, then, a substantial (and I think prohibitive) burden of explanation lies with the subjectivist.

     As for the implications of the foregoing for Gibbard, I intend it to undermine his notion of "what makes life worth living". If the reader believes life has value, and I have cast doubt on the claim that life's value inheres in the experience of it, then my radical claim about objective value may be better received because it affirms the value of life.