Norms as Hypothetical Imperatives
Note: This may be a tad technical for some readers. In my postings here at SO I recognize that many readers are not academic philosophers, so I try to walk a line that balances philosophical accuracy with accessibility. Here I go into somewhat greater detail and depth on an issue that non-philosophers may regard as esoteric. The issue is actually quite important, dealing with the sources and justification of ethical and moral norms. My reflections here spring from recently going over Hilary Kornblith’s Knowledge and its Place in Nature with the students in this term’s epistemology class. I am an admirer of this book and I here am offering some of my own reflections inspired by that reading.
A common charge against epistemological or ethical naturalism is that it is hard to see how norms derive from such views. Naturalism characteristically responds to intractable philosophical issues by invoking physical science and recommending that philosophical methods be replaced by scientific ones (Quine), or that philosophical methods be construed as continuous with scientific methods (Kornblith). I favor the latter view. I hold that philosophy should be conceived as a kind of inquiry not fundamentally different in methods or aims from natural science. Such a commitment carries both negative and positive implications. Negatively, it implies that philosophy is not to be done by appeal to allegedly a priori intuitions. Rather, intuitions are to be seen as a posteriori and pre-theoretical insights that have some purchase on reality, but are to be modified or rejected as well-confirmed theoretical knowledge accumulates. Further, philosophy can no longer be viewed as the tribunal of pure reason, the court of last appeal in matters of rationality. Among other consequences, this means that philosophers have no right to prescribe that scientists toe a doctrinal line (Christian, communist, feminist, or whatever), or insist that scientists steer clear of what philosophers have traditionally marked as their own turf (e.g., ethics, epistemology, aesthetics). When philosophers do make such pronouncements, they only succeed in debasing their practice into a kind of obscurantism.
Positively, a commitment to a naturalized approach to philosophy permits a close and cooperative relationship between philosophers and empirical investigators. Too often in the past philosophers and scientists have gone their separate ways. For instance, the standard approach of epistemologists has been to offer propositions of a form such as this: “S knows that P (or is justified/warranted in believing that P) if and only if conditions C1, C2, C3,…Cn hold.” Such proposals are generally offered as a priori, comprehensive, and one-size-fits-all definitions. Critics are then invited to propose counterexamples to show that the stated conditions are insufficient and/or not necessary for knowledge, justification, or whatever. Such counterexamples are generally highly imaginative, invoking brains in vats, regions where the locals erect fake barn façades, clairvoyants who are ignorant of their paranormal powers, and people who are always right but have no idea why they are. These debates appeal to our epistemic intuitions about knowledge or justification, and these intuitions are taken as decisive and authoritative, as though no real empirical inquiry were required to settle such issues. Meanwhile, scientific approaches are much more piecemeal, less interested in global pronouncements, and closely tied to empirical results (and, BTW, productive of such results). Philosophical naturalists ally themselves with this latter approach.
But if we seek to justify norms, as we must in ethics and epistemology, can we shirk appeals to the a priori? Don’t genuine norms have to be categorical, and, as such cannot be based upon contingencies of any sort? Kant, of course, was very clear that ethical norms must be categorical, that is, absolute. A genuine ethical command must be a pure imperative, an unqualified “thou shalt” or “thou shalt not,” dependent upon no “ifs,” “ands,” or “buts.” For Kant it was a cardinal mistake to think that ethical imperatives could be grounded in any matter of fact, such as human psychology or even the command of God. Any such attempt corrupts the fundamental ethical motivation, which is to do one’s duty simply because it is one’s duty and for no other reason. Therefore, Kant proposes to found his Categorical Imperative on a purely a priori basis—the form of the Moral Law. The Categorical Imperative is thus a purely formal requirement, namely, that the maxims that motivate our individual behavior must be such that they can rule as universal laws. Hence, for Kant, the job of philosophy is to produce an a priori metaphysics of morals, and this is very different from the job of any natural science.
Hume, though he certainly did not recommend an a priori grounding for norms, did a great deal to reinforce the conviction, now knee-jerk, that facts and values occupy separate realms. Did he not eloquently argue that you cannot derive an “ought” from an “is?” Science is the realm of “is;” it establishes natural facts, but, Hume argues, values do not derive from any statement of fact. You may look at a gruesome murder in as many ways as you please, cataloguing all of the stomach-churning facts that go into the CSI reports. Yet, nowhere in that description will you find the information that the murder was morally wrong. For Hume, that judgment is supplied by feeling, not fact. Though G.E. Moore was an intuitionist, not a subjectivist like Hume, his argument concerning the alleged “naturalistic fallacy” also inculcated the conviction that moral value cannot be defined in terms of any natural state, such as happiness (since it is always possible to ask further of any such state “Is it good?”). So successful have these arguments been that a fact/value dichotomy is now just taken for granted by many intellectuals. Stephen Jay Gould’s proposal of “nonoverlapping magisteria” as a solution to the science and religion issue simply assumes that religion deals with values and science with fact, where the two are wholly distinct (“nonoverlapping”) spheres.
Traditional epistemology has also sought categorical norms. For foundationalists and coherentists, actual scientific results play no part in specifying the conditions of justification. The conditions they lay down are abstract, formal, a priori, universal, and categorical. For the foundationalist, justification is a matter of identifying a set of privileged basic beliefs for which it is illegitimate to ask for any further justification. Properly basic beliefs are self-justifying. Candidates for such privileged beliefs include those that are allegedly self-evident, incorrigible, or evident to the senses. Non-basic beliefs, those requiring justification, are then justified by being put into a proper inferential relationship with basic beliefs. Thus, for instance, non-basic belief A is justified by being correctly inferred from non-basic belief B, which is justified by being correctly inferred from basic belief C, which needs no further justification. Coherentists hold that our beliefs are justified by being in a relationship of coherence with the entire body of our other beliefs. For coherentists, justification is primarily a feature of the entire system of beliefs and individual beliefs are justified by being integrated into that system. Despite their differences, foundationalism and coherentism concur in asserting categorical imperatives for justifying our beliefs: Justify your beliefs by putting them into the appropriate structure with other beliefs, either with beliefs that are closer to the foundational base, or with the body of your beliefs as a whole. The structure of those allegedly justifying relationships is given a priori and unconditionally.
But, even conceding that categorical imperatives may be coherently specified, what if they are useless or empty? As Alasdair MacIntyre points out, Kant’s Categorical Imperative can be made consistent with all sorts of maxims, even those that endorse moral atrocity. ISIS could, and probably does, state its maxims in a consistently universalizable way. ALL who deny the truth of their version of Islamist claptrap must die. As for categorical epistemic norms, they have the opposite problem: Instead of proposing far too weak a requirement, they pose one that is far too strong. For foundationalists, the class of kinds of properly basic beliefs is extremely exiguous, and the class of non-basic beliefs needing their grounding is tremendous. For instance, can the entire fantastic corpus of our scientific beliefs, including those about highly theoretical entities, processes, and relations, be reduced to an individual’s set of incorrigible sensations? Well, Carnap gave it the best possible try in The Logical Structure of the World (1928), but even he wound up admitting that it was impossible. As for coherentism, only God or some other omniscient being could know that its body of beliefs was coherent. For finite minds, the task vastly exceeds our capacity. For the coherence of our beliefs to be cognitively accessible—for us to be capable of knowing that our beliefs are coherent (and otherwise, what use is a coherentist criterion?)—we must be able to access the entire body of our beliefs. Hilary Kornblith comments:
Now I think it is safe to say that no human being ever does this, and, indeed, that no human being is even capable of it. While I have frequently reflected on my beliefs and self-consciously raised questions about their coherence, the number of beliefs I am capable of entertaining at any one time is quite small [!] in comparison with the total body of my beliefs…Grasping one’s total body of beliefs is something that cannot be done (Knowledge and its Place in Nature, p. 124).
Perhaps, then we should seriously consider the possibility that both moral and epistemic norms could be hypothetical rather than categorical. Norms could aim at the actualization of values—ethical or epistemic. A norm could have the form of a practical syllogism: “If we value V, and if action A tends to actualize V, then perform action A.” Ethical norms might therefore specify actions that tend to promote the well-being of sentient creatures, if such well-being is our value. Epistemic norms might specify the actions we need to take to maximize our chances of arriving at truth. For instance, we might say that to put yourself in a position to improve your chances of knowing the truth about climate change, read the findings of the qualified climate scientists and do not listen to the effusions of an AM radio bloviator. Which actions tend to actualize which values will be a matter of empirical fact. For instance, it is a matter of demonstrable empirical fact that countries with single-payer universal healthcare systems achieve better medical results for a lower per patient cost than in the U.S. Similarly, it is demonstrable that, for instance, double-blind experimental designs are more effective at assessing the effectiveness of pharmaceuticals than single-blind tests.
The most likely objection to recognizing that norms, particularly ethical norms, are hypothetical imperatives, is that such imperatives have to assume a set a values. They do not tell us what we should value. Assuming that I value the well-being of sentient creatures, I can, if practically wise, generate imperatives telling me how to pursue this in a given situation. But why should I value the well-being of sentient creatures? But how do we justify values? On what basis do we say that something is not just desired, but is desirable, namely that it merits or is eminently worthy of being desired? The only way seems to be to point out that it is desirable because it either is conducive to something of intrinsic worth or itself possesses intrinsic worth. When we reach those things that are intrinsically worthy, the things we value for their own sake, then the process of justifying comes to an end. In other words, we do not and cannot justify our basic values; we can only point them out. We discover inherent worth as an explorer discovers a river or mountain range. When we do discover what is really valuable, we cannot prove it, but, like Cortez silent upon a peak in Darien, we can only stand and admire.