Epistemology and Ethics, Parallel Between

views updated


Usually, actions are taken and policies adopted to realize envisaged goals, and they are undertaken because of belief that they will probably realize the goals. Actions and policies may be criticized, then, on one of two grounds: that the goals are ill-chosen or that the belief that the actions or policies will probably achieve the goals is ill-founded. It is interesting, and perhaps indicative of the facts to be examined below, that many words of appraisalsuch as "justified," "warranted," "reasonable," "right"are used, although possibly in slightly different senses, to indicate the acceptability or unacceptability of both goals and beliefs. Moreover, such appraisals obviously have two features in common. First, the appraisal of a particular goal or belief can be made only in view of some general principle or standard; second, the standards and principles in question are not self-certifying, and their rational justification must be a serious question for a thoughtful person. The appraisal of actions and policies thus raises two questions about both goals and beliefs: What are the proper principles or standards to be used in appraisals, and what is the rational basis for regarding any principle or standard as proper?

Historically, ethics has been the philosophical discipline concerned with these two questions about goals. (However, "goals" must be taken broadly to include not only the question of ultimate values but also the question of moral obligations and moral rights.) And historically, epistemology has been the discipline concerned with the same questions about beliefs. (Again, "beliefs" must be taken broadly to go beyond mere predictions of consequences of the use of certain means to include theories, explanations, and systems of mathematical thought.)

In order to develop the parallel between ethics and epistemology, it is convenient to identify ethical and epistemological statements. Ethical statements are those that imply a statement that could be expressed by some English sentence containing essentially "is a good thing that," "is a better thing," or "is morally obligatory that" (on the assumption that "morally obligatory" does not introduce special concepts different from a phrase such as "it is morally wrong not to"). The class of statements thus specified will be identical with the class of statements that moral philosophers have traditionally been concerned with. Similarly, epistemological statements are those that imply a statement which could be expressed by some English sentence containing essentially "It is reasonable (or warranted) for a person S to place more confidence in h than in i," in which it is understood that for h and i can be substituted expressions of the form "its being true that. " The class of statements thus specified is identical with the class of statements that epistemologists have been concerned with. It is useful to identify as ethical terms those phrases whose occurrence in a sentence distinguish it as an ethical statement and to identify as epistemic terms those phrases whose occurrence in a sentence characterize it as an epistemological statement.

Problem of Ethics and Epistemology

Moral philosophers have not, at least qua moral philosophers, been concerned with the acceptability of particular ethical statements such as "It would be a good thing if Jones learned to play the piano." Rather, moral philosophers have attempted to arrive at acceptable universal ethical statements which could serve as standards for the appraisal of particular situations. Thus, moral philosophers have defended or criticized such statements as "Enjoyment is always a good thing, abstracted from all consideration of consequences; nothing else is so" and "An action is morally right if and only if it will produce consequences as good as those of any other action the agent could have performed instead." The formulation of such generalizations, together with the proposal of reasons in support of them, is generally called normative ethics. Like moral philosophers, epistemologists are not concerned with the acceptability of particular epistemological statements such as "It is now highly probable that viruses are the cause of some forms of cancer." Instead, they have attempted to arrive at acceptable universal epistemological statements to be used as standards in appraising particular statements. Thus, they have examined the acceptability of such statements as "If at a time t person S seems to remember that he had the experience E at an earlier time, then he is initially warranted in believing that he did have the experience E " or "If at time t person S believes the statement h about his own experience at t, then it is reasonable for S to place at least as much confidence in h as in any other statement." The formulation of statements like these and the proposal of reasons in support of them have traditionally been the main occupation of epistemologists. In order to distinguish this task from other concerns of epistemologists, we may call it epistemology proper.

some minor points of similarity

If it is morally wrong for a person to take action A, but he takes that action, then in the absence of any excuse we attribute to him a fault of character and say he is morally blameworthy. Similarly, if a person has good reason for believing a certain statement but he does not, then in the absence of any good excuse we attribute to him an intellectual fault and characterize him as intellectually open to criticism. (The excuse in either case might be much the same; for instance, a person might plead that he was very upset, not "himself.") It is sometimes said that the parallel extends further in another direction, that just as there are several senses of "morally obligated," so there are corresponding senses of "reasonable to believe." For instance, it is widely believed that "morally obligatory" is sometimes used to mean the act which a being omniscient about the facts of the case and about moral principles would be morally blameworthy for not doing if he were in the place of the agent; this is said to be a sense of "morally obligatory" different from that employed when a person with possibly faulty information about the facts and imperfect clarity about moral principles is said to be morally obligated to do something. Correspondingly, it is sometimes suggested that there is a sense of "reasonable to believe" in which we may say that it is reasonable for a person to believe that any statement is true ; this sense is contrasted with the sense in which we say it is reasonable for a man to believe what is supported by the evidence he has. Whether there are such different senses in either case we must leave an open question here.

Epistemology Reducible to Ethics?

Some philosophers (R. Chisholm, for example) have thought that there is more than just a parallel between epistemology and ethics. They have thought that epistemic terms are properly defined by means of ethical terms. If this is correct, epistemological statements are complicated ethical statements, and, presumably, epistemology is a branchdoubtless a somewhat special oneof ethics. In accordance with this view, for example, the statement "It is warranted for S to place more confidence in h than i " might be taken to mean "For any good thing G, if S had to choose between risking it by a wager on the truth of h or risking it by a wager on the truth of i, he would be obligated to wager on h." If sound, this definition has the advantage of reducing the number of undefined terms in one's total system of concepts. The disadvantages of the definition are (1) that it is doubtful whether there is any useful sense of "obligated" in which the implied equivalence is true, (2) that the definition seems to be more obscure than the definiendum, and (3) that it does not seem that the meaning of "warranted belief" involves the notion of moral obligation but, conversely, that a person's being obligated to do something, in one ordinary sense, can be explained only by reference to the propositions it is reasonable for him to believe about his situation.

Theories of Meaning and Verification

If epistemology is not reducible to ethics, there is still a parallel between the higher-order questions and theories to which epistemology proper leads and those to which normative ethics leads. The discipline dealing with these epistemological questions and corresponding to the discipline of metaethics we may call metaepistemology. The central question of these disciplines, roughly, is how the statements of normative ethics and epistemology proper, respectively, can be supported, what is their "logic." The task of showing this is different, of course, from the task of producing a specific line of reasoning in support of specific ethical or epistemological principles. Can such principles be supported in the same way that propositions in the empirical and mathematical sciences can be? Whether they can be depends in part on what the meanings of the special epistemological or ethical terms are. Moral philosophers recognize three main views about the meaning of ethical terms and, correspondingly, about the ways in which ethical principles may be justified. Three very similar views have been given by epistemologists for the meaning and support of epistemological principles.


The first view, which we may call nonnaturalism in epistemology and ethics alike, affirms two things. It affirms (1) that epistemic and ethical terms are meaningful and that epistemological and ethical statements are true or false and (2) that epistemic and ethical terms do not name observable qualities (such as color or shape) and that their meanings cannot be defined, even partially, by citing a relation between them and names of observable qualities. Epistemic and ethical terms can be explained only by way of other epistemic and ethical terms. Hence, neither epistemic nor ethical principles can be confirmed by observation in the way that principles in the empirical sciences can be. This means that when we know ethical and epistemological statements that are not analytically true (as contrasted with ones like "A person ought to do his duty" or "One cannot know something unless it is reasonable for him to believe it"), our knowledge is synthetic a priori knowledge. A clear example of this view is the theory of probability held by J. M. Keynes, who thought that "probable" is an indefinable concept and that the axioms of probability theory are a priori synthetic knowledge.


The second view can be called naturalism or "definism." It agrees with the first affirmation of nonnaturalism but denies the second. It holds that epistemic and ethical terms can be explained without the use of other epistemic or ethical terms, that they can be explained exclusively by use of empirical and logical concepts. As a result it holds that nonanalytic epistemological and ethical principles have the same logical status as the principles of the empirical sciences and can be appraised ultimately by reference to the data of observation or introspection.

For example, according to one such definition of an epistemic term, the statement "It is reasonable for S to believe h " means just "S believes h." A more plausible theory defines "know" as follows: "S knows that h at time t " means that h follows logically from the propositions S believes about his own experience at t (including what he seems to remember), plus the following (enumerated) principles of inductive logic and principles about the truth or probability of memory beliefs. Examples of parallel definitions in ethical naturalism are familiar. If the second definition given were accepted, it would be an analytic proposition that the principles of deductive and inductive logic are known by everyone, just as, given a utilitarian definition of "right," the principle of utilitarianism is analytic.

Parallel to the claim of the ethical relativist that conflicting basic ethical principles may be affirmed with equal warrant by different persons or social groups, is it possible that conflicting basic epistemological principles may also be affirmed with equal warrant by different persons, even if a naturalistic analysis of epistemological terms is adopted? For instance, just as the ethical relativist may affirm that different assessments of the ethical obligation of making a promise may be made with equal warrant by different persons, may someone claim with reason that different assessments, say, of the weight of an additional observation in support of some general law may be made with equal warrant by different persons? Such questions must be left unanswered here. It is obvious, of course, that given different evidence, it may be reasonable for a person S to believe some propositions which it would be unreasonable for person S to believe.


The third view, which may be called noncognitivism, denies the thesis common to naturalism and nonnaturalism that epistemological and ethical statements are true or false but agrees with nonnaturalism that epistemic and ethical terms cannot be defined by means of empirical and logical concepts. Noncognitivism holds, however, that epistemic and ethical terms have a function and, in a sense, ideas in meaning. Ethical terms have been assigned various functions by different writers (functions like expressing the speaker's attitudes, changing the audience's attitudes, issuing prescriptions, declaring one's principles, giving advice, entreating, urging, exhorting, and so on).

Somewhat similar proposals have been made for epistemic terms. "It is probable that h," for example, is sometimes said to be a guarded way of affirming h or a cautious way of encouraging others to believe h. Again, "it is probable that h " is suggested not to assert that the speaker believes h but to express his belief in h. A more complex suggestion is to say that "it is reasonable to believe h " declares or expresses the speaker's own somewhat guarded inclination to believe and usually, at the same time, as a result of people's conditioning in the use of the language, strengthens the beliefs of others in h.

Further, parallel to ideas in C. L. Stevenson's ethical theory, one might say that epistemic terms also have some rather indefinite descriptive meaning, perhaps to the effect that acceptance of the proposition in question would conform to generally recognized standardsthe standards, perhaps, of scientists. One could say, as P. H. Nowell-Smith does in his ethical theory, that epistemic terms have various functions in various contexts and that it is a mistake to look for some single function performed by them on every occasion. It could be added that the use of epistemic words (similar to the suggested parallel to Stevenson's theory) carries special contextual implications which distinguish them from nonepistemic terms. Since the noncognitivist does not think that epistemic and ethical statements are either true or false, his view does not contain any theory about how the truth of such statements may be established, although it often contains a descriptive account of various ways in which persons do or could try sensibly, in view of the kind of statement in question, to remove their disagreements. Defenders of the noncognitive view in epistemology include R. Chisholm, Stephen Toulmin, and J. N. Findlay. Although they have been popular, noncognitive views, in epistemology and ethics alike, appear to face some difficulties. (For instance, in conditional clauses, such as "If it were known that , then " or "If it were a good thing to , then ," the ethical and epistemological terms seem to be used in their normal sense, but obviously nothing is being urged or expressed. Again, whatever specific function one assigns to terms of either type, it seems possible to find affirmations which employ the terms and cannot plausibly be said to be performing that function.) Moreover, it is doubtful whether there are conclusive reasons for rejecting "definism" (naturalism) in all its possible forms.

Some Broader Perspectives

One feature of both ethical and epistemic terms is that even very educated people do not have any clear idea of their meaning or of how to support their applicability in, for example, the way they are able to support the assertion that someone is a bachelor. Nor is it clear how an appeal to whatever vague meaning there is could be used reasonably to require anyone, on pain of inconsistency, to accept a corresponding ethical or epistemological principle; the fact that a person rejected such a principle would always be good reason for saying that his use of the epistemic or ethical term does not correspond to the sense needed in order to require the principle. It would appear that one of the functions of moral philosophy and epistemology is, rather, to propose helpful and clarifying uses for these terms.

What would be a helpful and clarifying use for these terms must presumably be decided by a broad view of human nature and society, a view of action and the requirements of reliable prediction for the purpose of action, the need of informal social controls in a complex society, and the necessity for impartial and general rules for such control, given that human beings are for the most part intelligent, self-interested creatures. However this may be, it is likely that reflection on the functions of science (reasonable beliefs) and conscience in society may provide, in the case of science at least as well as in the case of ethics, a guide for the philosopher's reconstruction of the meaning of ethical and epistemic terms and a basis for the appraisal of epistemic and ethical principles.

It should be noted that metaethics and metaepistemology both have another part. That part is the explanation of various concepts necessary for the understanding of terms and statements of both epistemology and ethics and for an understanding of theories about how epistemological and ethical statements may be supported. Among the concepts that are epistemologically important are "meaning," "truth," "reference," "analytic," and "a priori." Some of these concepts are also important for ethics; other concepts important for ethics are "action," "choice," "free choice," "voluntary," "intention," "motive," and so on. Both metaethics and metaepistemology, then, have branches devoted to these auxiliary concepts.

If there is a parallel between ethics and epistemology, should we go on to say that there is a parallel between ethics and science since, after all, to assert any statement in science is at least to imply that there is evidence for what is stated, that it is generally known, and so on? Is not every scientific statement, at least covertly, an epistemological statement? It seems clarifying not to go this far. Epistemological statements are of the form "It is warranted for S to believe that ," and we should note that what comes after the "that" is the sort of statement that occurs in the textbooks of science. If the meaning of this statement also has to be construed as somehow epistemological, difficulties arise.

See also Analytic and Synthetic Statements; Chisholm, Roderick; Epistemology, History of; Ethical Relativism; Ethics, History of; Keynes, John Maynard; Metaethics; Naturalism; Noncognitivism; Stevenson, Charles L.


Adler, Jonathan E. Belief's Own Ethics. Cambridge, MA: Bradford Book MIT Press, 2002.

Alston, William. Epistemic Justification: Essays in the Theory of Knowledge. Ithaca, NY: Cornell University Press, 1989.

Atkinson, R. F. "'Good' and 'Right' and 'Probable' in Language, Truth and Logic." Mind 64 (1955): 242246.

Audi, Robert. Belief, Justification, and Knowledge. Belmont, CA: Wadsworth, 1988.

Austin, J. L. Philosophical Papers. Oxford: Clarendon Press, 1961. Pp. 6771.

BonJour, Laurence, and Ernest Sosa. Epistemic Justification: Internalism vs. Externalism, Foundations vs. Virtues. Malden, MA: Blackwell, 2003.

Chisholm, R. "Firth and the Ethics of Belief." Philosophy and Phenomenological Research 51 (1) (1991): 199128.

Chisholm, R. Perceiving. Ithaca, NY: Cornell University Press, 1957.

Chisholm, R. "Evidence as Justification." Journal of Philosophy 68 (1961): 739748.

DePaul, Michael. Balance and Refinement: Beyond Coherence Methods of Moral Inquiry. New York: Routledge, 1993.

DePaul, Michael, and Linda Zagzebski, eds. Intellectual Virtue: Perspectives from Ethics and Epistemology. Oxford: Clarendon Press, 2003.

Fairweather, Abrol, and Linda Zagzebski, eds. Virtue Epistemology. New York: Oxford University Press, 2001.

Feldman, Richard. "Epistemic Obligations." In Philosophical Perspectives 2, edited by J. E. Tomberlin. Atascadero, CA: Ridgeview Press, 1988.

Firth, R. "Chisholm and the Ethics of Belief." Philosophical Review 68 (4) (1959): 493506.

Firth, R., and R. Chisholm. "Ultimate Evidence." Journal of Philosophy 53 (1956): 722739. A symposium.

Greco, John. "Virtues and Vices in Virtue Epistemology." Canadian Journal of Philosophy 23 (1993): 413432.

Lewis, C. I. The Ground and Nature of the Right. New York: Columbia University Press, 1955. Chs. 23.

Sinnot-Armstrong, Walter, and Mark Timmons, eds. Moral Knowledge? New York: Oxford University Press, 1996.

Sosa, Ernest. Knowledge in Perspective. Cambridge, U.K.: Cambridge University Press, 1991.

Steup, Matthias, ed. Knowledge, Truth, and Duty. New York: Oxford University Press, 2001.

Toulmin, Stephen. "Probability." In Essays in Conceptual Analysis, edited by A. G. N. Flew. London: Macmillan, 1956.

Urmson, J. O. "Some Questions concerning Validity." ibid.

Williams, Bernard. Ethics and the Limits of Philosophy. London: Fontana Press, 1985.

Zagzebski, Linda. Virtues of the Mind: An Inquiry into the Nature of Virtue and the Ethical Foundations of Knowledge. Cambridge, U.K.: Cambridge University Press, 1996.

Richard B. Brandt (1967)

Bibliography updated by Benjamin Fiedor (2005)