SOURCE: Rubidge, Bradley. “Psychological Atomism, Amour-propre, and the Language of Generosity.” In La Rochefoucauld, Mithridate, Frères et sœurs, Les Muses sœurs: Actes du 29e congrè annuel de la North American Society for Seventeenth-Century French Literature, edited by Claire Carlin, pp. 43-52. Tübingen, Germany: Gunter Narr Verlag, 1998.

[In the essay below, Rubidge argues that elements of La Rochefoucauld's views, including his criticism of generosity, have similarities with the philosophical position called eliminative materialism.]

La Rochefoucauld's eighty-third maxim exhibits some argumentative moves that are typical of the Maximes:

Ce que les hommes ont nommé amitié n'est qu'une société, qu'un ménagement réciproque d'intérêts, et qu'un échange de bons offices; ce n'est enfin qu'un commerce où l'amour-propre se propose toujours quelque chose à ganger.¹

Here we see the characteristic move of reduction: the statement that one thing is nothing but another thing. We also discover, in friendship, a typical target of such reduction: a kind of behavior, relationship, or virtue that seems admirable. This ideal is denounced as a specious manifestation of amour-propre or intérêt—which we can, for the moment, treat as synonyms. The maxim is striking not just because it claims that friendship is something other than friendship, but because it asserts that friendship is, in fact, the thing most opposed to friendship. The word friendship, the maxim affirms, denotes behaviors and relationships that are in truth un-friendly. We are invited to conclude that, if friendship is nothing but these behaviors, then it does not really exist and we delude ourselves by talking about it.

Such paradoxical reduction is practiced throughout the Maximes. The target of the reduction is usually the ethic of generosity—a Neo-Stoic morality of heroic, self-denying virtue that seems diametrically opposed to the impulses of amour-propre. This ethic depends on a set of assumptions about human nature, especially the belief that people can sacrifice their own advantage for duty or for a superior good. The ethic also relies on a set of terms that articulate those beliefs and constitute a moral discourse—the language of generosity. La Rochefoucauld's reductive attack challenges both the reality of disinterested behavior and the meaningfulness of the language that describes altruistic conduct.

Other moralists of the seventeenth century, especially Pierre Nicole, suggest that all human behavior, even apparently charitable conduct, can be explained as the result of egoism. Nicole proposes an analogy from Cartesian physics to explain how egoism alone can form an orderly society that seems to display charitable conduct, but in fact survives without anyone's being truly disinterested:

Rien n'est plus propre, pour représenter ce monde spirituel formé par la concupiscence, que le monde matériel formé par la nature, c'est-à-dire cet assemblage de corps qui composent l'univers; …²

Though Descartes was not an atomist, Nicole's reference to small corps and his invocation of only physical causes recalls the doctrine of atomism. Nicole and La Rochefoucauld's psychological reduction implies that self-love and self-interest are a kind of atom, a building block from which other phenomena are constructed and into which these higher order phenomena can be analyzed. La Rochefoucauld sometimes seems to go even further. Though Nicole admitted the possible existence of charitable motives, La Rochefoucauld's psychological atomism appears also to be a sort of monism, for he suggests that self-interest is the only motive that really exists and influences our behavior, while altruism is a mere illusion³.

In this paper I want to pursue Nicole's suggestion that we refer to physics and metaphysics when we think about the relationship between altruism and egoism. The psychological reduction proposed by moralists like Nicole and La Rochefoucauld bears a striking resemblance to materialist responses to what philosophers call the mind-body problem. The suggestion that altruistic motives and behavior are illusions and that all human behavior is the result of amour-propre resembles the claim that people do not really have minds and that what we believe to be our minds are just deceptive effects produced by physical causes. The similarity of the reductive move is heightened by the sharp opposition drawn between altruism and egoism, which recalls the radical distinction between the mental and the physical. Even though the parallel is not strict, exploring it can help us better understand the nature and power of psychological atomism.

Though the mind-body problem became a prominent issue in the seventeenth century, the comparison I wish to draw is based on recent work on the mind-body problem by analytic philosophers. I focus on a position called eliminative materialism. Eliminative materialism is especially interesting because, in addition to claiming that the mental is really just the physical, it also challenges our usual way of talking about mental entities and experiences. Eliminative materialism's challenge to this language resembles La Rochefoucauld's attack on the discourse used to describe generous, disinterested behavior. Since late twentieth-century readers often accept La Rochefoucauld's claim that all behavior can be accurately explained by reference to amour-propre, I am particularly interested in exploring whether such a complete elimination of the language of generosity is possible.

Philosophers working on the mind-body problem often distinguish between ontological and semantic problems. We raise ontological questions when we ask whether mental entities, such as feelings and beliefs, exist and what their true nature is. We raise semantic questions when we try to explain how ordinary language statements about mental entities are meaningful. The distinction is important because materialists, who deny that mental entities exist, have to explain how our ordinary language, which assumes mental entities, can be meaningful. Eliminative materialism adopts a particularly radical thesis about the meaningfulness of ordinary language statements about mental entities. This position was developed in response to other materialist proposals' failure to resolve the semantic problem.

Logical or philosophical behaviorism was an influential theory of mind from the 1930s through the 1950s. It was originally formulated more as a semantic claim than as an ontological claim. Gilbert Ryle's presentation makes the position clearest. Ryle argued that there is no “ghost in the machine.” We do not really have minds, and ordinary talk about mental states is merely a shorthand for statements about dispositions. On this account, saying that someone has a belief or feeling is really a way of saying that he or she has a disposition to a certain behavior that will be fulfilled in certain circumstances, circumstances that are implied in the “mind talk” statement⁴. This view implies that ordinary mind talk should be abandoned in favor of statements about dispositions and observable behavior.

The first logical behaviorists, who were members of the Vienna Circle, made a more definite claim about mentalistic statements. Taking as their starting point the verification principle (that a statement's meaning is established by the conditions for its verification), they concluded that the only meaningful mentalistic statements are those which can be translated into statements about observable physical properties or behavior (and thus be verified). Meaningful ordinary language statements about the mental, then, are really statements whose sentences can be replaced with sentences about the physical⁵. These logical behaviorists held out the prospect of a physicalist psychology based on such translations. The translations, however, were not forthcoming, and it soon became evident that they could not be expected. The translations logical behaviorists proposed for statements about the mental were too unwieldy. Too many conditions and circumstances had to be enumerated, and it was difficult to eliminate reference to mental entities such as beliefs or desires, especially from statements about the pre-existing conditions “in” the subject.

The Identity Theory was initially developed to fill in some gaps in Rylean logical behaviorism, but was soon presented as a complete replacement. Unlike logical behaviorism, the identity theory did not begin as a semantic claim. Eventually, however, the identity theory ran into problems that centered on the status of ordinary language statements about mental entities.

Identity theorists did not, like the early logical positivists, claim that most ordinary language statements about mental entities are meaningless. On the contrary, they argued that they are meaningful and refer to entities that really exist. Their innovation was to identify the referents of mentalistic statements with the referents of materialist statements about the brain. Using Frege's distinction between sense (Sinn) and reference (Bedeutung), identity theorists claimed that words for mental entities or phenomena have a different meaning from words for brain states, but that the mentalistic terms denote the same thing as the neurophysiological terms⁶. Mental states, then, are nothing but brain states. Identity theorists argued that this identity is not, as the logical behaviorists had suggested, as a priori, analytic identity based on the definition of terms. On the contrary, they insisted that the identity of mind and body is an a posteriori, synthetic, contingent identity that can be discovered by empirical research, as it was discovered that water is H₂O.

The identity theory does not demand translation of mentalistic statements into materialist terms, since it recognizes the meaningfulness of mentalistic language. Still, it suggests that such a translation is desirable and that neurophysiological terminology is preferable to mentalistic language, since mental entities really are states of the brain. But the formulation of the identity theory did not allow for a persuasive argument that the meaningfulness and value of mentalistic terms can be respected at the same time as the superiority of neurophysiological terms can be affirmed. Various technical efforts to overcome these problems have been proposed without great success. But early on a radical solution was offered, a position that has recently been revived.

Developing a line of argument introduced by Paul Feyerabend, Richard Rorty argued for eliminative materialism. Rorty claimed that advocates of the identity theory were mistaken when they tried to respect the value of ordinary language claims about mental entities and experiences⁷. A thorough-going materialist monism must accept that mental entities, as described in ordinary mentalistic statements, do not exist. Rather than saying that mental experiences are identical with brain states, it is more accurate to say that what we call mental experiences are brain states. The reformulation suggests that we may eventually learn not to use mentalistic language, abandoning the belief that its terms refer to anything that really exists. Eventually we will refer only to brain states. Rather than suggesting that ordinary language statements about the mental be translated into physicalist terms, Rorty argues that mentalistic language should, and will, disappear. As a better language arises for describing the phenomena in question, we will accept that the entities posited by the old language do not really exist. Rorty argues that similar changes in our language have resulted from scientific progress. People used to believe in many entities, such as witches, about which they could make meaningful statements. Even though such statements remain comprehensible, we no longer believe in those entities and do not accept statements about them as true descriptions of reality.

Eliminative materialism was revived and extended with new arguments in the 1980s, in ways that make its similarity to La Rochefoucauld's project especially clear. Paul Churchland and Stephen Stich argue for eliminative materialism by carrying out attacks on what they call folk psychology. Folk psychology is our ordinary way of talking about mental entities. Recent eliminative materialist arguments depend on describing folk psychology as a theory. Our commonsense accounts of mental experience and behavior can be counted as a theory because we do not merely use certain terms. We also assume the existence of certain phenomena and entities and we posit laws about the relationships between them so that we can predict and explain them. Since the discourse used to express the ethic of generosity similarly implies belief in certain entities and assumes laws or regularities that allow explanation and prediction, the attack on folk psychology resembles La Rochefoucauld's attack on the language of generosity and gratitude.

Churchland's argument is particularly dependent on identifying Folk Psychology as a theory. Having argued that it is a theory, he tries to show that it is such a flawed, inadequate theory that it must be abandoned. He argues for the elimination of mentalistic terminology and ontology on the grounds that what philosophers of science call a smooth intertheoretic reduction of folk psychology to modern neuroscience is impossible. Some sciences, such as chemistry, can be reduced to physics by the introduction of bridge laws that translate the terms of chemistry into the terms of physics. Churchland insists that the imperfections of folk psychology make it impossible to find such bridge laws, and that rather than being reduced or translated, folk psychology will have to be abandoned⁸.

Though he is more cautious about advocating eliminative materialism than Churchland, Stich's arguments make the connection to a position like La Rochefoucauld's clearer, for he offers concrete evidence of folk psychology's inadequacy and modern cognitive science's superiority. Experiments in social psychology concerning attribution and cognitive dissonance show that people often make mistaken claims about their behavior and their mental experiences. By falsely describing the effect of placebos, for example, experimenters can induce their subjects to give erroneous accounts of their experiences and the causes for their behavior. People can thus be mistaken about beliefs and desires, which would seem to be states to which they have privileged access, and which provide causes or conditions that we usually think affect behavior⁹. These experiments suggest that we do not have access to the real causes of our conduct; the language we think describes those causes is deceptive and should be abandoned. The claim recalls La Rochefoucauld's observation that we frequently make errors in explaining our mental states and the causes of our behavior.

Both Churchland and Stich argue against retaining our ordinary language about mental experience and for replacing folk psychology by modern neuroscience. This project appears hard to imagine, for it seems that our ordinary discourse about mental experiences is meaningful and that much would be lost of we stopped talking that way. In his work on eliminative materialism, however, Rorty explained how such a change might occur, and Churchland makes much the same kind of argument. Rorty argues that there used to be coherent, meaningful, well-developed discourses about phenomena and entities in which we no longer believe. As examples he mentions witchcraft, demonic possession, and ghosts. Statements about such entities have not been translated into modern scientific language, but have instead entirely disappeared from respectable scientific discourse, because such statements, although they may be meaningful, do not refer to anything that really exists. This is why Rorty argues for what he calls a “disappearance” version of the mind-body identity theory. In normal, serious discussions of events in the real world, mind talk will eventually be displaced by materialist statements.

The development of the social sciences makes the analogy between La Rochefoucauld's observations and the project of eliminative materialism plausible. Much of the progress in the social sciences has been achieved by adopting economic models to describe human behavior. People are assumed to behave egoistically, and behavior is explained by showing how it is advantageous. Though this is only a methodological assumption, and though careful social scientists admit its limitations, the assumption of universal egoism is widely accepted, so much so that we are skeptical, even in ordinary life, when a person claims to act disinterestedly. In addition to asking “What's in it for me?” we often probe suspicious offers by asking “What's in it for you?” Perhaps the language of generosity, like claims about witches, is on its way to elimination. In the future we may only talk about rational egoists allocating scarce resources to secure the maximum return.

But is it really possible that reference to altruistic motives and behavior will disappear? The limitations on comparing La Rochefoucauld's psychological atomism to materialist solutions to the mind-body problem indicate that the language of generosity cannot be displaced.

As I indicated earlier, most philosophy of mind makes two sorts of claims. There are ontological claims about what really exists and its nature, and there are semantic claims about the meaningfulness of our statements about what exists. The first response one can make to La Rochefoucauld's reductive claim is ontological. Though La Rochefoucauld regularly uses the reductive formula n'est que, and although he contrasts egoism and altruism very strictly, the reduction he proposes cannot be as rigorous and eliminative as those proposed in philosophy of mind. The contrast between mind and body can be coherently formulated as a substance dualism, the claim that there are two completely different kinds of substance. The contrast between amour-propre and charity is strict, but it is not a substance dualism because egoism and altruism are supposed to be different kinds of the same thing, namely a motive or impulse. Moreover, we cannot even construe the opposition between egoism and altruism as a property dualism, according to which the same cause produces radically different properties. After all, amour-propre and generosity are supposed to produce phenomena that are similar, indeed, in some cases, indistinguishable. Though egoism and altruism are different, they share most of their properties.

If egoism and altruism are similar entities and produce phenomena of the same basic type, then there is no great urgency to reduce one to another, because the problems of dualism do not arise. In a dualist system we have to explain how the two kinds of thing can interact, despite their radical differences, and how two radically different kinds of phenomena can occur. Here there is no such difficulty. Indeed, it is easy to say that egoism and altruism interact and mingle. But because of that connection, it also seems easy to reduce these two motives to a single type and hard to argue that such a reduction is impossible. Dualist philosophers of mind can argue that the mind must exist as a substrate for mental properties by claiming that mental properties are irreducible to physical properties because they are of different kinds. Once we admit that the phenomena caused by generosity are similar to those produced by egoism, then we cannot rule out the claim that they are, in fact, produced by self-love alone.

But the claim of a thoroughgoing psychological egoist, who affirms that self-interest is the fundamental motive underlying all behavior, goes further than saying that we can explain apparently altruistic conduct by reference to self-interest. He argues that all behavior really is caused by self-interest. He thus makes a claim about the nature of self-interest and its relation to altruism, affirming that amour-propre is more fundamental than love-for-another or generosity. Such a claim cannot, however, be advanced a priori. On the contrary, there are better grounds for suggesting that interested and disinterested motives are of equal status. The very words self-interest and amour-propre suggest that egoism is not a simple, atomistic, fundamental motive. The term amour-propre indicates, by its reflexive form, that self-love is a subspecies of love. Thus, self-love is no more fundamental than love-for-another; both are forms of love distinguished by their object. If any impulse or motive is irreducible, it is love itself¹⁰.

Ontological issues, however, are not my main concern. In most of this essay I have focused on the semantic claims of philosophers of mind. I compared the language of generosity to folk psychology, and suggested that La Rochefoucauld's critique of explanations of behavior that refer to disinterested motives resembles eliminative materialists' challenge to folk psychology. In fact, La Rochefoucauld's critique is less extreme, and we are now in a better position to understand why he does not go so far.

Maxim 83, which I read at the beginning of this talk, is somewhat atypical. Most of the maxims include attenuating or mitigating terms, like souvent and d'ordinaire. These words temper the maxims' reductive affirmation and admit that disinterested motives and behavior could indeed exist. This mitigation indicates not only that a “disappearance theory” of the discourse of generosity is impossible, but even that a “translation theory” is impracticable, because there are some cases in which reduction or translation is false.

Despite its reductive claims, La Rochefoucauld's discourse of egoism cannot avoid referring to the motives it denounces. That is because La Rochefoucauld's discourse is not a new theory. On the contrary, the account of motivation and behavior he proposes is part of the conventional language of generosity and gratitude. The traditional moral discourse that includes reference to disinterested motives like charity, generosity, and gratitude recognizes that not all persons are altruistic, and that not all actions are self-sacrificing. Indeed, this moral discourse is largely constructed so as to distinguish actions and agents according to the degree of generosity they display. This discourse does not rule out the claim that a given act or a given person is self-serving—on the contrary, it requires that such claims sometimes be made, and be accepted as true.

But, you may argue, La Rochefoucauld tends to go further. He doesn't just say that people are usually self-interested, a position the language of generosity would admit. He also claims that actions and motives that appear to be self-seeking are in fact egoistical, and that we are deceived when we think they are altruistic.

That move, too, is contained in the language of generosity. The traditional moral discourse recognizes that external observers may misconstrue the motives of others, that people may mislead others about their motives, and that people may even be mistaken about their own motives. The reductive claim made by La Rochefoucauld is part of the language of generosity. All La Rochefoucauld does is claim that we are more frequently deluded about our own and others' disinterestedness than the language of generosity usually admits. But he cannot logically deny the operation of all generosity, because, as I indicated above, self-interest is not, logically or ontologically, more fundamental than disinterestedness. At most he could make an empirical claim which is fairly easily challenged.

La Rochefoucauld's insertion of attenuating terms indicates he was aware of the limitations on his reductive project. Having compared his reductions to those proposed in philosophy of mind, we are in a better position to explain the limitations on his claims. Like the traditional moralists, La Rochefoucauld admits the possibility (though he is skeptical about the reality) that two contrasting motives can cause behavior, amour-propre and disinterest. These motives have equal status; they are both either basic drives, or secondary drives growing out of a more fundamental impulse, love. Self-interest, however, is a special sort of motive because it can produce behavior, and even internal experiences, that resemble the effects of self-sacrifice. There does exist, therefore, a specious kind of generosity or charity that can, in fact, be reduced to amour-propre. But the fact that we can carry out a reduction of this false altruism does not entail the nonexistence of true disinterestedness, or its ineffectiveness. The existence of false appearances of generosity merely means that we may often be mistaken in thinking we or others act disinterestedly. That is as far as La Rochefoucauld's claim goes, or can go. The consequence is that, although we may be skeptical about claims of disinterestedness, we cannot do without a language and a moral philosophy that include descriptions of altruistic behavior and motives.

Notes

1. François de La Rochefoucauld, Maximes 3rd., rev. & aug. ed., ed. Jacques Truchet (Paris: Classiques Garnier, n. d. [1982]).

2. Pierre Nicole, Œuvres philosophiques et morales (1845; reprint, Hildesheim: Georg Olms, 1970), p. 182. The passage continues to mention Descartes' tourbillons.

3. For the sake of brevity I disregard the epiphenomenalist position. This essay focuses on contemporary positions in the mind-body debate, and epiphenomenalism is nowadays almost entirely abandoned. Perhaps I should also mention that the term psychological atomism alludes to Bertrand Russell's doctrine of logical atomism.

4. Gilbert Ryle, The Concept of Mind (New York: Barnes & Noble, 1949).

5. See Carl G. Hempel, “The Logical Analysis of Psychology,” in Readings in Philosophy of Psychology, ed. Ned Block (Cambridge, MA: Harvard UP, 1980), vol. 1, pp. 13-23 (originally published in Revue de Synthèse [1935]) and Rudolf Carnap, “Psychology in Physical Language”, in Logical Positivism, ed. A. J. Ayer (New York: Free Press, 1959), pp. 165-98 (originally published in Erkenntnis 3 [1932-33]).

6. Herbert Feigl, “Mind-body, not a pseudo-problem”, in The Mind-Brain Identity Theory, ed. C. V. Borst (N. p.: Macmillan/St Martin's, 1970), pp. 33-41 (originally published in Dimensions of Mind, ed. Sidney Hook [New York: NYU Press, 1960]) and J. J. C. Smart, “Sensations and Brain Processes”, in Materialism and the Mind-Body Problem, ed. David M. Rosenthal (Englewood Cliffs, NJ: Prentice-Hall, 1971), pp. 53-66 (originally published in The Philosophical Review 68 [1959]).

7. Paul Feyerabend, “Mental Events and the Brain”, in Materialism and the Mind-Body Problem, ed. Rosenthal, pp. 172-73 (originally published in The Journal of Philosophy 60 [1963]); Richard Rorty, “Mind-Body Identity, Privacy, and Categories,” in Materialism and the Mind-Body Problem, ed. Rosenthal, pp. 174-99 (originally published in The Review of Metaphysics 19 [1965]); Richard Rorty, “In Defense of Eliminative Materialism”, in Materialism and the Mind-Body Problem, ed. Rosenthal, pp. 223-31 (originally published in The Review of Metaphysics 24 [1970]).

8. Paul M. Churchland, “Eliminative Materialism and Propositional Attitudes”, Journal of Philosophy 78 (1981), pp. 67-90. For Churchland's account of how his theory is related to other views, see Paul Churchland, Matter and Consciousness, rev. ed. (Cambridge, MA: MIT UP, 1988).

9. Stephen P. Stich, From Folk Psychology to Cognitive Science: The Case Against Belief (Cambridge, MA: MIT Press, 1983), especially pp. 230-37.

10. Philippe Sellier provides a good explanation of Augustine's formulation of the view that love is the basic human motive in Pascal et Saint Augustin (Paris: Armand Colin, 1970).