Self-Esteem and the Death of the Subject

I have written here repeatedly about the problems with person-oriented reactive attitudes and character skepticism. But recently I came across the work of the psychologist Albert Ellis, whose work is at the intersection of therapeutic psychology and philosophy. His work on self-esteem and person-oriented assessment suggests an interesting new direction for the general insight that we are in error when we attribute actions, habits, and tendencies to a self or a subject.

Ellis calls this “unconditional self-acceptance.” Where the psychology of self-esteem encourages us to continually affirm (perhaps daily) propositions about how lovable and capable we and others are, Ellis’s unconditional self-acceptance instead suggests that we forgo these exercises and the global evaluations they require for more careful assessments of acts and behaviors. The same applies to our assessments of others, and thus he offers a good case study of the attempt to operationalize a rejection of person-oriented reactive attitudes through “unconditional other-acceptance.”

Ellis’s student David Mills summarizes the argument like this:

  1. Most people unfortunately believe that self-esteem must, in some way, be earned through accomplishments.
  2. When self-esteem is based on accomplishments, it must be earned repeatedly. It is never permanent.
  3. The concept of self-esteem leads intermittently to self-damnation.
  4. The concept of self-esteem usually promotes social and behavioral inhibition.
  5. A compulsive drive for self-esteem leads to frequent anxiety. And self-esteem-related anxiety is an obstacle to achieving those goals essential to our self-esteem!

Now, I think there’s a lot of truth in Ellis’s diagnosis. We have good reasons to believe that our acceptance within the community is predicated on the judgments of our peers. So we are right to self-monitor the likely assessments of others, to avoid transgressing crucial communal norms, free-riding on the efforts of our collaborators, or running afoul of the unwritten standards of behavior and comportment. There’s some reason to believe that this monitoring is the basis for person-oriented status judgments: we assess others and ourselves in order to determine the standards for preserving our group membership, and the continued existence of social exclusion and individual choice proves that we’re not living under conditions of unconditional acceptance.

Yet at the same time, we also know that our assessments and attributions suffer from serious errors and biases. Psychology has begun to catalog these biases and give them catchy names like the spotlight effect and fundamental attribution bias, but the basic insight is just that we’re often very deeply wrong about these assessments.

As a result, Mills (following Ellis) recommends an elegant solution:

  • To overcome self-esteem-related anxiety and inhibition, recognize that your choice is not between self-esteem and self-condemnation. Your choice, rather, is between establishing an overall self-image and establishing no self-image. That is, you can choose to view your external actions and traits as desirable or undesirable, but abstain from esteeming or damning yourself as a whole.

This is a philosophically dense proposal, one that assumes that by changing our metaphysical orientation to persons, we can overcome the pernicious (and importantly false!) habits of anxiety, self-blame, and self-destruction. In so doing, we can also develop a more sensitive and sophisticated attitude towards our neighbors and fellow citizens.

Of course, the practical efficacy of these attitudes are difficult to measure; apparently there’s been little empirical work on the topic, but to assess the model it helps to think through the best case scenario. Let’s assume that forgoing global evaluations of self and other has the effects promised: less anxiety, fewer fundamental attribution errors, improved mental health outcomes, etc.

Yet as we think about these themes, and especially about prescriptive metaphysics required for this to function, I wonder if we can preserve the sense of accuracy. Is this merely an exercise or is it meant to actually be supplying more accurate claims about the world? Is it convenient or true?

Academics of a certain stripe have been rehashing the “death of the subject” for a while now. The best reasons for rejecting person-oriented reactive attitudes seem to follow in this mold: one cannot judge a person without judging her acts, yet single acts are insufficient for a whole judgment of her person. Her acts are multifarious and varied, yet domain-specific judgments are subject to contextual factors. She is the agent of her acts, yet agency is empirically undermined by context.

Ellis himself claims the mantle of truth for this rejection of global judgments, but since his primary work is with patients who aren’t all willing to accept the full set of metaphysical presumptions here, he also suggests a “pragmatic” and “inelegant” alternative:

“If, however, you have difficulty refusing to rate your self, your being, you can arbitrarily convince yourself, ‘I am “good” or “okay” because I exist, because I am alive, because I am human.’ This is not an elegant solution to a problem of self-worth, because I (or anyone else) could reply, ‘But I think you are “bad” or “worthless” because you are human and alive.’ Which of us is correct? Neither of us: because we are both arbitrarily defining you as ‘good’ or ‘bad,’ and our definitions are not really provable nor falsifiable. They are just that: definitions.

Defining yourself as ‘good,’ however, will give you much better results than believing that you are ‘bad’ or ‘rotten.’ Therefore, this inelegant conclusion works and is a fairly good practical or pragmatic solution to the problem of human ‘worth.’ So if you want to rate your self or your being, you can definitionally, tautologically, or axiomatically use this ‘solution’ to self-rating.”

This was always the real problem with the self-esteem movement and with the two kinds of respect Stephen Darwall identified; it’s very difficult to preserve recognition respect, a sense of respect-for-persons that rates them higher than chairs, concepts, or other animals while simultaneously pretending that there are no further forms of appraisal like their skills, competences, and morally salient decisions.

We sometimes pretend that maximal attention to the norms of recognition respect eliminate the room for appraisal respect. Thus, because humans all have this recognition respect in the form of what Kant called “dignity” there’s no room for social status differentiation. But we play favorites. I have favorite people (friends), favorite scholars (idols?), favorite religious groups (Quakers!), and even favorite politicians (Elizabeth Warren, who was once a favorite scholar!) What’s more, I have good days and bad days, days where I’m proud of my teaching and writing, and other days where I feel like I failed to live up to my own expectations.

Ellis claims that we should actively resist any effort to assemble all these appraisals into a complete picture of the person. That we can assess the actions without making all the troublesome metaphysical assumptions required to attribute those actions to a person. Indeed, perhaps I shouldn’t give Elizabeth Warren the Senator so much credit for the work of Warren the Law Professor.

But I’m still giving Warren credit. And that’s the problem. I’m starting to think we can’t duck person-oriented reactive attitudes by merely reducing them to action-oriented reactive attitudes. Going back to the original Strawson paper, we don’t get angry at the painful blow, or fall in love with the witty reply. We get angry at the person who lands the painful blow; we fall in love with the person who offers the witty reply.

So how can we avoid the Nietzschean invention of a doer for every deed? Can we stop ourselves from filling in the back story of the driver who cuts us off in traffic to show that he is a terrible human being? And if we can, should we? Or should we continue to pretend?

We still might want to say that global judgments are a mistake. The person who offers you witty replies on a first date may also be kind of boring sometimes. The person who assaults you may also be a loving father or an honor roll student. It may well be that we learn remarkably little about most people from what we see of them, and that we fill in this ignorance with heuristics and biases that are more rough than ready.

It’s hard not to equate Nietzsche and Ellis here with Buddhist reflections on the illusory nature of selfhood. And it’s hard, too, not to think that this demand that we amend our syntax and our ethics begs the question.

Are we merely doing this to get off the treadmill of anxiety, to overcome maladaptive perfectionism? Is all this elaborate metaethical reflection really just therapeutic? Is it the philosopher’s obsessive #actually that demands we reassess the common sense for no other reason than to avoid imprecision? Is there a pragmatic upshot? What’s the cost of self-esteem? And what are its benefits?