obert C. ROBERTS
An important aspect of Ellis’s theory and practice of therapy is the promotion of self- acceptance in the client. This is accomplished by getting the client to give up all self-evaluation. Ellis appears to give three different rationales for this practice/theory: 1) the selves are functions, 2) that selves are ontologically incapable of taking value-predicates, and 3) that it is therapeutically necessary to forswear self-evaluations. It is argued that Ellis needs a rationale for rejecting all self-evaluation, but on his own principles none of the three available to him works. Thus his psychotherapy fails a crucial test of theory adequacy: consistency. If RET is to became a rational system, a way must be found to eschew self-evaluation in terms of traits and behavior without forswearing all self-evaluation. For this purpose the theory needs to be supplemented with some plausible account of human dignity or value, in which clients can be rationally convinced that they share, independently of the value of their traits and behavior.
Keywords: RET, self, evaluating selves
Republished from Psychotherapy: Theory/Practice/Research/Training, 1987 with the kind permission of the APA
Correspondence concerning this article should be addressed to:
Robert C. Roberts Department of Philosophy Baylor University PO Box 97273 Waco, TX 76798-7273, USA
Rating people is irrational, says Albert Ellis (1977b): “Value is a meaningless term when applied to man’s being… it is invalid to call him either ‘good’ or ‘bad’” (p. 101). “ ‘Although I can rate my various acts of behaviors in terms of whether they help me survive and enjoy myself, I cannot legitimately assess or evaluate my total self, or my humanity’” (Ellis, 1977a, p. 31). Why is it meaningless or illegitimate to evaluate a self? At different points Ellis offers three different explanations.
He sometimes suggests that there is no such thing as a “self,” that a human is nothing but a loosely tied bundle of acts, beliefs, experiences, short-term abilities, and dispositions of widely disparate sorts. Thus he occasionally sounds like the empiricist philosopher David Hume, who says that human beings “are nothing but a bundle or collection of different perceptions, which succeed each other with an inconceivable rapidity, and are in a perpetual flux and movement” (Hume, 1739, p. 252) On this view, to rate a “self” would be like rating a bagful which consisted of a juicy lemon, an old carburetor, some clothespins, a new size 8 running shoe, and half a bottle of insecticide. How shall we rate the bag? It would be meaningless to try, because the contents do not constitute anything integral enough to make that activity rationale.
All of man’s traits are different – as apples and pears are different. Just as one cannot legitimately add and divide apples and pears and thereby get a single, accurate, global rating of an entire basket of fruit, so one cannot truly add and divide different human traits and thereby obtain a single, meaningful, global rating of a human individual (Ellis, 1977b, p. 109).
If selves are fictions like basketfuls, then evaluating them is on all fours with avoiding ghosts; neither activity makes any sense, because of the object acted on does not exists.
At other junctures Ellis seems to say that while selves exist, they are a kind of being which cannot be overall good or bad. “Value is a meaningless term when applied to man’s being… it is invalid to call him either ‘good’ or ‘bad’” (Ellis, 1977b, p. 101, italics added). On this interpretation ascribing goodness to a self is like ascribing redness to the number two: just as numbers ontologically resist color-predicates, so selves ontologically resist value-predicates. If we assume that a self exists as a composite of dispositions, abilities, experiences, and so on, then a self on this view would differ essentially from other integral but composite kinds of thing, for example, a house. A house is composed of a number of elements or aspects: a foundation, sides, windows, carpets, roof, heating system, layout, and so on. But with houses it makes good sense to evaluate both the parts or aspects that go to make it up, and the house itself, overall. And these two evaluations are logically connected. Roughly, the principle would be: a) the greater the number of things that are bad about the house, the more likely it is to be overall bad; b) the more basic a defect is, the more likely it is to make the house bad overall (a bad foundation is more infecting than a bad window); and c) the more difficult to rectify a defect is, the more likely it is to make the house bad overall (having toxic insulation between the walls is worse than being infested with plague-ridden rats, because the latter is more remediable). While selves, like houses and some other composite beings, have enough integrality to exist, they differ from houses in a basic way: they have the unusual property of being incapable of having overall value. They are what we might call ontologically nonvalue-predicable.
Yet a third position can be discerned in Ellis’s writings: while selves may exist and have value, it is necessary for purely practical reasons not to evaluate them. Evaluating them negatively is a major ground of mental disturbance, and evaluating them positively opens them to the possibility of mental disturbance should circumstances warrant a change of evaluation. Thus it is better, in the interest of mental health, not to get into the business of evaluating selves at all. In this case the situation is like that in a school, where it is clear enough that children do, objectively speaking, perform at different levels, but were this fact is concealed from the children because it is thought that their rating their performances and one another’s would create an atmosphere less conducive to education. The difference is that Ellis proposes not just that a certain class of persons (pupils, therapists, disturbed people) refrain from evaluating selves, but that such evaluations be generally ruled out.
Upon reading Ellis one is left with the impression that he holds all three of these positions at different points in his writings. Perhaps so. But notice that they are mutually contradictory: The first contradicts the second and third by denying that selves exist; and the second contradicts the third by denying that selves have value. Clearly, Ellis cannot rationally hold all three of these positions.
It is not entirely clear which of these rationales Ellis advocates. The most efficient position for him to take would be 1), since the nonexistence of selves would entail what is most attractive for him about 3) (namely, that selves, should not be evaluated in practice) and 2) (namely, that selves, do not have value). His contributor Arnold Lazarus states the position pretty clearly and he calls into question the existence of selves:
Is there such an entity as a ‘self’? What does it mean to be ‘liked for one’s self’? I suppose that there are certain continuities in behavior that let us make predictions about our responses and other people’s reactions in different situations, but we are still talking about behavior. If people like me for ‘myself’ it means that they value all respect the way I behave at various times and across different situations. (Lazarus, 1977, p. 115)
Ellis does not consistently deny the existence of persons; indeed most of the time he seems to assume, quite commonsensically, that human selves – and not just their traits and behaviors – exist. At one point, he even defines the therapeutic process as the replacement of ‘egos’ with ‘persons’:
According to… rational emotive therapy, man attains maximum understanding of himself and others and minimum anxiety and hostility when he follows the rule ‘Where ego was, there shall the person be.’ By ego, of course, I mean man’s self-rating and self-justifying tendencies. (Ellis, 1977b, p. 110)
And by implication, by ‘person’ Ellis must mean a being, a human entity, whose emotional dispositions are no longer characterized by reflexive global evaluations. In other words, a person is just a self that has learned to be quite consistent in refusing to evaluate itself. According to this quote, to rule out talk about selves is not really to claim they do not exist but to prescribe that we not rate or justify them. Ellis also seems to endorse the concept of a self in statements like, “Positive or negative evaluation of a person… may well encourage him to be less of a self or of a self-directed individual than he would enjoy being” (Ellis, 1977b, p. 105).
So if Ellis does not deny the existence of selves, maybe he is claming that they have the unusual property of being nonvalue–predicable – that is, by nature not capable of being either good or bad. This would seem to be a possible interpretation of Ellis’s repeated assertion that it is fallacious to move from a claim about the value of a person’s acts or traits to a claim about the value of the person. “Blaming or praising the whole individual for a few of his acts is an unscientific overgeneralization” (Ellis, 1977b, p. 103). But the invalidity of the inference, or the meaninglessness of the ascription of value to man’s “being,” would have to be a function of something peculiar about human beings. It is certainly not in general a fallacy to give something an overall rating on the basis of rating some of its aspects or parts. If a house has a bad roof, a bad foundation, leaky windows, smells bad inside, and a few other things, it is legitimate to call it a crumby house, a louse of a house, an overall loser or turkey of a house. It would, of course, be silly to say that a house was bad overall because it had a couple of minor defects, just as it would be overgeneralizing to say that somebody is an overall jerk because she makes a few social blunders at cocktail parties. But if the analogy with houses holds, it would not be overgeneralizing to call a person bad who consistently displayed a wide range of importantly bad traits. Ellis tells us that “if a ‘rotten person’ truly existed, that individual would have to inevitably, for all time and all times, act rottenly”(Ellis,1977a, p. 29). This is an odd thing to say, and suggests that Ellis knows something about the ontology of persons that the rest of us do not know. He would surely not say that if a rotten house existed, it would have to be rotten in every respect. So he must believe in a disanalogy between persons and other things when it comes to inferring their overall value from the value of their parts or aspects.
If humans are nonvalue-predicable, we would like to have an account of why this is so. That numbers do not take color-predicates does not cry out for an explanation, since people do not tend to think that numbers have color. By contrast virtually the whole human race throughout recorded history has thought that people, and not just their traits and performances, have value. Ellis himself thinks the tendency to evaluate our selves is one that humans are universally born with (Ellis, 1976). If Ellis’s opinion is that selves, though existent, are nonvalue-predicable, he is offering an interesting and counterintuitive hypothesis for which he needs some argument.
Despite statements suggesting that selves are fictions because of the mutual disparity of their traits, and claim the value of the fallaciousness of inferring the value of the whole form the value of parts or aspects, the bottom line for Ellis is probably that self-evaluation has bad behavioral and emotional consequences. Ultimately, he rejects self-evaluation not on metaphysical but on pragmatic grounds.
Both positive and negative self-evaluation are inefficient and often seriously interfere with problem-solving. If one elevates or defames himself because of his performances, he will tend to be self-centered rather than problem-centered, and these performances will, consequently, tend to suffer. (Ellis, 1977, p. 102 ff., italics added)
There are many pragmatic reasons for forswearing self-evaluations: they lead to one-upmanship and one-downmanship, they tend to narrow a person’s range of interests and enjoyments, they promote bigotry and block social change, they sabotage empathic listening and free will, and foster musturbating (see Ellis, 1977, pp. 103-106).
But if these remarks about the firs two of Ellis’s rationales are correct, and yet RET commits us, on pragmatic grounds, to give up evaluating selves, RET must abandon the rational ideal. For surely, if people have a value it is inconsistent to have a practical policy of never admitting this. The situation would be like that of a prisoner serving a life sentence who cannot be free of depression and anxiety unless he believes that he is soon to be paroled. So he surrounds himself with symbols of his release: a calendar with a day one month hence circled in red; a letter to himself that he has written in another’s handwriting inviting him on a sailing trip when he is released; and so on. Perhaps it is implausible to think any individual is so rigidly constitute that the only way to equanimity is to deny the truth about his future. But if someone were, this person would be forced to choose between equanimity and truth. His ritual denials might even succeed in getting him peace of mind, just as Ellis’s statements about the “fallacy” of overgeneralization may be rhetorically effective in getting people to give up self-evaluation. But such practice falls far short of the rational ideal.
The inconsistency of RET as Ellis presents it, like that of our imagined prisoner, is forced on us by the nature of things; it is not one of which Ellis is guilty. If mental health is inconsistent with acknowledging the truth, then we must choose between the truth and mental health; and presumably Ellis has resolved this dilemma by choosing mental health at the cost of truth, even if his arguments have made him less than fully aware of having lost the truth about selves. For the falsity of the conclusions of these arguments – that selves do not exist, and that selves are nonvalue-predicable – is patent enough unless very strong arguments to the contrary are forthcoming. As we have seen, they are not forthcoming.
The Form of a Remedy
But we are really in the dilemma that Ellis has (inadvertently) cast us into? If we are committed to mental health, must we systematically refuse to acknowledge the existence or value of selves? Is there no adjustment by which RET might be made consistent with the nature of selves?
Ellis assumes that there is only one way to circumvent negative self-evaluations. Since all evaluation of selves must be on the basis of performances and traits, and even the best persons in terms of performances and traits may turn into the opposite, we should refuse all evaluations whatsoever. But maybe there is another way of ascribing value to selves that is not “dangerous” in the way that ordinary value-ascriptions are. Some have thought that humans, just by their nature, independently of particular traits or achievements, have a dignity and worthiness of respect. Immanuel Kant, for example, wrote:
Humanity itself is a dignity; for man cannot be used merely as a means by any man (either by others or even by himself) but must always be treated at the same time as an end. And it is just this that comprises his dignity (personality), by virtue of which he assumes superiority over all the other beings in the world which are not men and can be used – hence over all things. (Kant, 1964, p. 132; see also Hill, p. 1973)
For Kant, the dignity or “absolute value” of a human being is consequent upon his freedom of will (Kant, 1948, p. 101), and freedom of will is construed in terms of the possession of “practical reason,” which is the faculty which knows moral right from wrong and indeed is the source of that distinction (Kant, 1964, p. 102). The distinction between right and wrong is captured most fundamentally in the rule that each person “should treat himself and all others, never merely as a means, but always at the same time as an end in himself” (Kant, 1964, p. 101). This is why humans, simply in view their nature, call for the attitude Kant calls “reverence.” Paton (1947) points out that “dignity” is a technical term borrowed from stoicism and is meant to be contrasted with “price.” (Paton, 1947, p. 189). Price is the value something has with respect to some context or market; dignity is an absolute value, determined not by any context but merely by the nature of the thing in question. Since dignity is built into human nature it cannot be lost.
This idea would, of course, need explication and defense to be made plausible enough for use in helping people to positive self-regard despite their poor traits and performances. Whether it could be made plausible would be found out only by trying to make it so. (Ellis sometimes says that beliefs need to be “empirically proven” to be rationally acceptable; but this is an impossibly stringent standard. Can it be empirically proven that beliefs must be empirically proven to be rationally acceptable? Kant’s view of human value would be rationally acceptable if some strong reasons could be given for it and most of the important objections against it answered.) But if a therapist believed that people have an inalienable dignity in virtue of belonging to a race of creatures capable of making moral distinctions, and if she could convey this compellingly to her clients, this doctrine could play the role in RET that I am arguing needs to be played by some doctrine. I do not know whether this would work. Probably it would work with some therapists and some clients (namely, some rather abstractly intellectual ones with a strong sense of objective moral obligations), but not with all.
Ellis half endorses in thinned-out version of this Kantian rationale for positive evaluation of selves:
Perhaps the only sensible way of making a global rating of an individual is on the basis of his aliveness: that is, assuming that he is intrinsically good just because he is human and alive (and that he will be nongood or nonexistent when he is dead). (Ellis, 1977b, p. 108)
Telling the client that she has worth because she is alive and human may help her accept herself. But if she is at all critical, the question will leap to mind, “How does being human and alive give me any goodness? Maybe human beings are just fancy meat, or highly evolved rats, or something.” Or maybe the client will be helped because she has something like the Kantian theory of persons in terms of which to interpret Ellis’s words about being human and alive. But on its own terms this thin version of the Kantian theory goes no distance at all toward making RET more rational, for two reasons. First, it is not in any way integrated into the theory and practice of RET. The second, the words remain merely rhetorical unless they are filled out with an actual account of how being human involves having some kid of global self-worth – in other words, something like a theory of the value of human selfhood. Ellis himself makes this point when he criticizes another rational therapist for bad strategy in trying to get a depressed client to convince herself, “I happen to be a rotten mother – that’s true – but I am a wonderful human being.”
For practical purposes, this solution is all right since she can hold that she is a wonderful person merely because she exists, in spite of her failings. But philosophically, she has no more reason for calling herself a “wonderful person” than she has for calling herself a “bad person.” (Ellis, 1971a, p. 131; see also Ellis, 1971b, pp. 19-20)
Another rationale, and one which for many clients and some therapists is more emotionally powerful than the Kantian one for engendering positive self-rating, is a theological one, according to which humans have value because of their being made in the image of God, or loved by Him, or destined by Him for a noble life beyond this world. Ellis is wrong when he claims that this religious rationale for positive global self-evaluation reduces to the claim “that a human is globally ‘good’ just because he is human and alive” (Ellis, 1977b, p. 108). Typically of those in the verificationist tradition, Ellis ignores the enormous difference in theoretical content between the theological rationale and the empty rhetorical one. He thinks that in cases were are two utterances have the same observable effect (namely, reduction of emotional disturbance), they have the same meaning. The three rationales (the rhetorical one, the Kantian one, and the Judeo-Christian one) can be ranked on a scale of theoretical richness. Just as the idea I quoted from Ellis is a radically thinned-out version of Kant, it could be argued that Kant’s theory is a thinned-out rationalist version of a Judeo-Christian picture of the self. The Kantian theory trades on the concepts of free will and moral obligation; the Christian picture trades on these plus the concepts of a loving and righteous God and His kingdom; by contrast, Ellis’s references to the value of being human and alive have no conceptual background at all accept what is accidentally supplied – probably by vague association – in the mind of the client.
- Ellis, A. (1971a). Growth Through Reason. North Hollywood, Calif.: Wilshire Book Company.
- Ellis, A. (1971b). Humanistic Psychotherapy. New York: McGraw-Hill.
- Ellis, A. (1976). The biological basis of human irrationality. Journal of Individual Psychology. 32(2), 145-168.
- Ellis, A. (1977a). The basic clinical theory of rational emotive therapy. In A. Ellis and R. Grieger (Eds.), Handbook of Rational Emotive Therapy. New-York: Springer.
- Ellis, A. (1977b). Psychotherapy and the value of human being. In A. Ellis and R. Grieger (Eds.), Handbook of Rational Emotive Therapy. New-York: Springer.
- Hill, T. E. (1973). Servility and self-respect. Monist, 57, 87-104.
- Hume, D. (1739). A Treatise of Human Nature.(Edited by L. A. Shelby-Bigge, 1888) Oxford: Oxford University Press.
- Kant, I. (1948). Groundwork of the Metaphysic of Morals (trans. H. J. Paton). London: Hutchinson. (Reprinted 1959, New-York: Harper&Row.)
- Kant, I. (1964). The Doctrine of Virtue (trans. Mary J. Gregor). Philadelphia: University of Pennsylvania Press.
- Lazarus, A. (1977). Toward an egoless state of being. In A. Ellis and R. Grieger (Eds.), Handbook of Rational Emotive Therapy. New-York: Springer.
- Paton, H. J. (1947). The Categorical Imperative. London: Hutchinson.
While the article –printed here with the permission of Prof. Roberts–was originally published in Psychotherapy: Theory/Practice/Research/Training (1987: 24, 4, 821-825), it continues to have great significance today. Ellis considers unconditional acceptance of self to be the distinguishing factor between his own and other cognitive-behavioral theories. For example, Beck, Meichenbaum, and Freeman variously offer interventions for symptom relief and personality reorganization. However, Ellis has claimed unconditional self acceptance (USA) synonymous with elegant change that other CBT interventions do not achieve. He considers his own ABC and other CBT changes to be secondary. Not only has Prof. Roberts, now Distinguished Professor of Philosophy at Baylor University, commented on shifting arguments-pragmatics by Ellis concerning self, he has suggested other possibilities that might work well within REBT. McMahon, in a chapter in Dryden’s text on new directions in REBT (2004), commented that the various definitions by Ellis might also be viewed as a tribute to his diversity. The reader might consider both points of view, as well as the lead, overview article in this issue in helping him or her to make a judgment about whether the concept that Ellis claims to be cornerstone to his own theory has been stated coherently by him either before or after 1987. Ed.
Republished with the kind permission of the American Psychological Association