Category Archives: philosophy

defining “games”

Leave a reply

I am reading Josh Lerner’s Making Democracy Fun: How Game Design Can Empower Citizens and Transform Politics because it makes an important argument. Games are fun for specific reasons; most political processes fail to be fun because they lack those elements; and we could make politics more fun without sacrificing serious purposes if we learned from game design. That’s the great value of the book, but here is a philosopher’s digression ….

Lerner (p. 29) defines games as “systems where players engage in artificial conflict, defined by rules, that results in measurable outcomes.” My ears perk up at any definition of “games” because Ludwig Wittgenstein famously avoids defining that word in his Philosophical Investigations. There he observes that games come in many different forms and asserts that no single feature defines them all. Games constitute a family of cases, each of which resembles several others even though they are not all alike in any particular respect. We know how to use (and teach) the word “game” even though we cannot define it in terms of necessary and sufficient conditions. This observation is important for Wittgenstein because he believes that language is a heterogeneous set of games. And we think in language. Thus our thought is a set of practices that lack a common feature, yet we can learn to think and communicate.

Lerner offers a definition. He emphasizes relevant and important features of many practices that we call “games”–features that we should heed when we design political processes, which is Lerner’s interest. One wouldn’t need his definition to understand the word “game”: I have been playing games for almost half a century without thinking in Lerner’s terms. His doesn’t exactly work as a literal definition, because, for instance, a business competition could easily be an “artificial conflict, defined by rules, that results in measurable outcomes” such as profit and loss. If that competition is devoid of fun, we wouldn’t call it a “game,” except metaphorically. Also, if you showed Lerner’s definition to someone who had never played a range of games, it wouldn’t communicate what he has in mind. This person might think of standardized tests, duels, court cases, and other artificial conflicts that we don’t usually call “games.”

This is not a criticism of Lerner. I think his definition plays its intended role in his book. He presumes some real world experience with games and provides many vivid examples to expand one’s store of cases. His definition points to general tendencies in those examples that are important in a different context, politics. That is a typical and appropriate way to advance an argument. But I am left thinking that Wittgenstein was right about the indefinability of the word “game.”

(As a digression on this digression: Wittgenstein wrote in German, and the word “Spiel” means both “game” and “play.” For Lerner, the differences between the English words “game” and “play” are important; to make politics more game-like is different from making it more playful. Does Wittgenstein fail to see a common denominator to all “Spiele” because that word encompasses play as well as games? I don’t think so: all of his examples are actually “games” in the English sense. His argument works perfectly well when translated.)

voting and punishment: Foucault, biopower, and modern elections

Leave a reply

Michel Foucault wrote a great deal about punishment as a tool that governors use to discipline the governed. Voting seems like the opposite: a device for the governed to discipline the governing. But Foucault’s concept of bio-politics can be illuminatingly applied as a critique of modern voting.

Foucault begins “Security, Territory, Population” (his 1977-8 lectures at the Collège de France) with a “very simple, very childish example” of punishment in three forms.

  • Juridico-Legal: The law defines a category of actions as a crime (e.g., theft), and sets a certain punishment to follow it in order to restore justice. This punishment is usually conducted in public and on the body of the accused.
  • Disciplinary: Punishment is used to influence behavior, both of the person being punished and of others who may be deterred. Punishments are now designed to have results; for instance, prisons become “houses of correction.” If a given punishment lacks beneficial consequences (as Cesare Beccaria argued of torture), it should be repealed. But in Discipline and Punish, Foucault interprets this apparent humanity or leniency as a reflection of an ominous improvement in the efficiency of discipline, whose purpose is “not to punish less, but to punish better.'”
  • Security: The objective becomes to influence the frequency of undesirable actions (such as theft) in the population as a whole. Outcomes are measured statistically, for instance, in terms of crimes/capita or probabilities of recidivism. A given punishment, such as imprisonment, is now a mere tool for security, to be assessed by its aggregate costs and benefits and compared against other tools, such as paying or training people to behave as desired or subjecting them to surveillance and monitoring.

Foucault emphasizes that these three “modulations” of punishment have not simply replaced one another in a historical sequence. Even medieval law sometimes aimed at security; juridico-legal thinking remains alive today. But security has become far more prominent in the current era than it was before.

Like punishment, voting has adopted relatively durable forms but has changed its purposes and rationales in profound ways. Drawing on Michael Schudson’s accessible history, I would identify the following three stages in the history of US voting:

  • Nineteenth Century: Voting is mostly a public expression of full membership in a group. By voting at all, a man shows that he is a full and free US citizen. By voting for a party, he shows his loyalty to a sub-population, e.g., Southern white Protestant farmers vote for Democrats. Voting is conducted in public (ballots are not secret) along with torchlight parades and other public rituals. Generally, everyone in a given community votes alike and reinforces each other. Voting is an obligation.
  • Progressive Era: Voting is a private choice among independent candidates and ballot questions. Voting maximizes the degree to which the government represents the voter’s interests and values. Elections also punish corrupt or incompetent incumbents by rotating them out of office. To enable a free and precise choice, the ballot is now secret; candidates are distinguished from parties; numerous offices are made elective; and important questions are put to referenda. Reporters, experts, and civic educators purport to assist voters in making up their own minds. Voting is a source of power that should be employed responsibly.
  • Post-Watergate: For individuals, voting is one means of influencing the government (at a time when other means have proliferated) and is one optional way to spend time and energy. A prospective voter is assumed to weigh the costs of voting–including the costs of becoming informed–against its benefits. The population is assumed to vote as a function of large external factors, such as the billions of dollars spent on campaign advertising and the constantly shifting procedures for registering and voting. Candidates are entrepreneurs who make heavy use of Big Data to target and influence citizens. Some prominent political scientists and jurists defend private campaign finance on the basis that the various campaign donors cancel each other out in a competitive market. Voting, running for office, and giving money are choices; aggregate results can be predicted.

The three stages of voting resemble those of punishment. In each case, we see a move from 1) symbolic to 2) deliberately manipulative to 3) scientific and statistical. We also see a move from 1) automatic to 2) individually tailored to 3) designed at a social scale. And a sequence of 1) physical impact on bodies, to 2) influence over individual minds, to 3) tweaking the milieux that shape mass behavior. Foucault calls scientific control over the contexts that shape human behavior “bio-politics,” which is the ascendant norm.

In the case of punishment, the tool’s effectiveness has increased, but control is increasingly dispersed. The medieval king was fully in charge of the gallows, but he couldn’t influence much of his realm with it. The modern regime of schools, prisons, and police is much more effective and pervasive, but there is no single king. Power strengthens but also multiplies.

In the case of voting, the tool may possibly have become more powerful, but the individual voter pretty clearly has less influence today, for other political acts (from drawing district lines to allocating campaign dollars) have become highly sophisticated and effective. Voting looks more like a dependent variable than the cause of anything.

If this portrait of the current situation is accurate, we need both an assessment and a strategy for improvement. Foucault proposes some theses about assessment and strategy at the outset of “Security, Territory, Population”:

I do not think there is any theoretical or analytical discourse which is not permeated or underpinned in one way or another by something like an imperative discourse. However, in the  theoretical domain, the imperative discourse that consists in saying “love this, hate that, this is good, that is bad, be for this, beware of that,” seems to me, at present at any rate, to be no more than an aesthetic discourse that can only be based on choices of an aesthetic order. And the imperative discourse that consists in saying “strike against this and do so in this way,” seems to me to be very flimsy when delivered from a teaching institution or even just on a piece of paper. … So, since there has to be an imperative, I would like the one underpinning the theoretical analysis we are attempting to be quite simply a conditional imperative of the kind: If you want to struggle, here are some key points, here are some lines of force, here are some constrictions and blockages. In other words, I would like these imperatives to be no more than tactical pointers. … So in all of this I will therefore propose only one imperative, but it will be categorical and unconditional: Never engage in polemics.

Contra Foucault, I would like to assert that the current system of elections (and much worse, of prisons) in the US is bad; that this is not a merely aesthetic judgment; that making such judgments is worthwhile if you defend them; and that effective polemics are badly needed. But I take Foucault’s point that a paper argument against the status quo can be valueless or arbitrary. As always, the question “What should we do?” requires tough-minded analysis that is about strategy as well as facts and values. Specifically, if we want to defend the Progressive Era ideal of voting, we must take seriously the deep shift toward what Foucault called “bio-power” in the society as a whole.

See also:when society becomes fully transparent to the state; qualms about Behavioral Economics; citizenship in the modern American republic: change or decline?

assessing a discussion

Leave a reply

We discuss in order to address public problems together. We also develop morally through discussion–which, by the way, I would define very broadly to encompass a conversation with your neighbor over the backyard fence, with Leopold Bloom in the pages of Ulysses, with Angela Merkel through the New York Times, with Jesus in prayer, or with your late parent through memories and imagination.

I posit that the quality of discussion is a function of the skills, attitudes, and beliefs of the participants; the nature of the question under consideration; and the format. An individual’s contribution to a good discussion must be understood in context, because a given discursive act (such as making a concession or repeating a claim) can either be helpful or harmful, depending on the situation.

The tool I would use to assess discussion is a network map, where the nodes are the assertions made by the participants, and the links are explicitly asserted connections, such as “P implies Q” or “P is an example of Q” or “P is just like Q.” The network grows as the conversation proceeds–except when people stop adding new ideas and links–and each contribution can be assessed in terms of how it changes the network. A person’s statement can (for example) make a network larger, richer, denser, or more coherent.

As an illustration, I’ve mapped a 2005 Pew Research Center debate on the right to die (prompted by the then-recent Terry Schiavo case) that involved Daniel W. Brock (a medical ethicist), R. Alta Charo (a law professor), Robert P. George (a political theoriss), and Carlos Gómez (a hospice physician). The transcript is here and my map can be explored here:

This topic (end-of-life decisions) has certain features: it raises fundamental metaphysical questions rather than empirical questions that could be settled with data. It poses absolute and irrevocable decisions, unlike questions about the distribution of scare resources, which can be negotiated. As for the format, it involved relatively long prepared statements by just four experts, in contrast to a free-for-all among a larger group, which would have a different structure. And the speakers, although diverse in perspectives, were all accustomed to a certain style of argument (relatively abstract and organized). It would be interesting to contrast this transcript to, for instance, a New England town meeting about a budget.

Dan Brock goes first and has a chance to lay out a position in favor of allowing a patient or her surrogate to end life support. His position is neatly organized, with the principle of autonomy at the center. He names that principle as the underlying rationale for a series of professional reports and court decisions that represent what he calls the current consensus. He connects autonomy to several related concepts: bodily integrity, privacy, self-determination, and choice. He draws the explicit implication that an autonomous patient must be able to choose or refuse any treatment. He adds the idea that when a person is incapable of exercising autonomy and has not made an advance directive, the best course is to empower a surrogate to choose. And he denies that the patient’s or surrogate’s choice should be constrained by supposed distinctions between starting versus stopping care, hydration/nutrition versus medical treatment, or a terminal versus a stable condition. Below is his position, isolated from the rest of the network.

Screen Shot 2015-05-09 at 5.56.20 PM

Brock’s position is consistent (no nodes contradict each other), coherent (all nodes are connected), and centralized around the concept of autonomy. I would attribute those features of his position to: 1) the format (he gives prepared remarks that come first in a debate), 2) the professional style of the speaker (a professional philosopher), 3) the nature of the topic (bioethics), and 4) Brock’s position as a liberal who strongly favors autonomy. Indeed, Robert P George, the conservative theorist, says later in the debate: “liberals have to come up with a justification for placing autonomy in the central position in the first place, and that requires the defense of a moral proposition.” Note George’s use of a network metaphor to characterize Brock’s view.

Dr. Carlos Gomez speaks second. Unlike Dan Brock, he doesn’t produce a single, organized argument with explicit connections than link all of his ideas. I count nine different clusters of points in his remarks. Gomez’ points–isolated from the rest of the network–are shown below.

Screen Shot 2015-05-10 at 1.11.34 PM

An important claim for Gomez only becomes evident to me (although this might be my own limitation as an interpreter) during the following exchange from the Q&A:

MODERATOR: Actually, before we go to the next question, when you said autonomy misses something essential in this sort of doctor-patient relationship, would you elaborate a little bit more on what that means in the real world?

MR. GOMEZ: Yeah, I’ve never had a patient knock on my office door, come in, sit down, and say, “I’m here to exercise my autonomy.” Now I may be a little too glib there, but what I’m suggesting is that one of the reasons that they are coming to me is precisely by nature of what I profess as a physician, by nature of what I know in terms of my skills, and also by nature – and on this I think Robby is dead on – by nature of the fact that there is a moral construct to what it means to be a physician or a nurse, or any other professionals that professes publicly what they’re going to do.

I think what Carlos Gomez has been implying all along is that nurses and doctors are required to show care for a patient, and an ethic of care is inconsistent with ending the patient’s life. Further, caregivers should have a strong voice in the debate about bioethics. Unlike Dan Brock, however, Gomez does not present that position as an organized argument but alludes to it with relatively scattered claims about how, for instance, there is actually no consensus about end-of-life treatment and the press is uninformed about hospice care. If I were to evaluate Gomez’ participation, I would say that he is less rhetorically effective than he might have been because he never states a claim that actually is central for him. The moderator assists not only Gomez but also the group by drawing out one central node that had not been clear before. On the other hand, Gomez clearly contributes ideas to the conversation and connects many of them to points already introduced by Dan Brock; so he broadens and enriches the discussion.

Alta Charo, a law professor, speaks third. She makes a cluster of points about how people mistake biological patterns for moral imperatives, and a related cluster of points about how sometimes the law appropriately creates “fictions” that are not based on biology, such as the idea of adoptive parenthood. She also makes at least nine other points that don’t explicitly connect to these two clusters. Her view is about as coherent as Carlos Gomez’. However, she is in a different position from him. She generally holds the same liberal position as Dan Brock, who has already spoken. It would not contribute to the conversation for her to repeat Brock’s argument for the centrality of autonomy, although she does state that choices about life must be personal and free. Instead, she builds ideas around the structure than Dan Brock has already laid out.

George follows Charo, and he lays out an alternative view to Brock’s, in which autonomy is explicitly not the central idea. Instead, “human life, even in developing or severely mentally disabled conditions, [is] inherently and unconditionally valuable.” His structure is about as consistent, coherent, and centralized as Brock’s, but it has a different center. Below is shown a network consisting only of the ideas proposed by Brock and George. “Human life is unconditionally valuable” is a central node in the top third of the picture; autonomy is a different center about two-thirds down.

Screen Shot 2015-05-10 at 1.30.00 PM

The two networks touch at multiple points, either because George contradicts Brock (I show explicit disagreements with darker lines) or because he acknowledges specific areas of agreement.

Later, in the Q&A, George makes a discursive move that can sometimes be helpful to a group. He says, “As much as I love disagreement and dissent, I think that on one point on which Carlos and Dan thought they were arguing, there’s not actually a disagreement.” This is an example of tying together two points that have already been made in order to increase the coherence of the network. It is a helpful move–unless the two points are not actually alike.

By the time the session ends, the whole network is fairly connected. But certainly, no agreement has been reached, and two nodes remain central for different people but mutually inconsistent. That may be an inevitable feature of debates about the ends of life, or it may be a function of the way these speakers reason about such questions. Although they are speaking lightly at this juncture, Brock and Gomez imply a serious point about the impasse between them:

Dan Brock: Well Carlos and I first met on a PBS show about assisted suicide I guess 15 years ago, was it, Carlos? And we disagreed then roughly the way we do now, so –

MR. GOMEZ: I’m unteachable.

MR. BROCK: So am I.

I would hope that more mutual learning can occur when issues are either more empirical or more negotiable than this one is.

it’s not just what you think, but how your thoughts are organized

We come into the world with no moral ideas at all and must learn them from others. We learn not just from arguments and explicit principles, but also by observing practices and experiencing emotional reactions.1 We must make judgments about complex, evolved, historically contingent phenomena (such as, among many others, marriage, democracy, and art) that we cannot apprehend as wholes but must learn to assess from accumulated and vicarious human experience.2 In Habermas’ terms, we begin with a “Lifeworld” formed of our shared experiences and improve it through explicit deliberation with diverse people in civil society.3

Some people are much better at this process than others are, and we can explain why by understanding their moral worldviews as networks of ideas and connections and considering how their whole networks are organized. Consider these hypothetical discussion partners:

  • Aaron constantly returns from any situation or moral consideration to the same value. He considers that value immediately relevant to all others and nonnegotiable. It defines his moral identity and appears to him manifestly true. Deliberating with Aaron is impossible, but not because his network contains a foundational belief in the sense of one that is “infallible, or indubitable, or incorrigible, or certain.”4 What makes him a poor deliberator is rather the over-centralization of his network of moral ideas. One cannot find a route around his core principle.
  • Bao endorses a lot of moral ideas, examples, and principles. But he cannot connect one to another. Asked why he believes P or Q, he has nothing to say about his reasons, let alone can he offer a chain of reasons that connects P to Q. It is hard to talk to Bao because his network is disconnected.
  • Carlos simply has nothing to say about many choices, dilemmas, and cases that arise in conversation and practice. He can discuss some topics cogently, but many others seem not to interest or concern them. The problem with Carlos’ network is that it is too small (having too few nodes) or has too restricted a scope.
  • Dominique cheerfully holds both P and not P, depending on her mood or perhaps her self-interest or convenience. Dominique frustrates deliberation because her network harbors blatant inconsistencies that she does not attempt to resolve.
  • Eduardo is committed to one idea, like personal liberty or economic equality, and he will not recognize the legitimate pull of other values that conflict with his summum bonum, e.g., order and security, solidarity and community, or democracy. Eduardo’s network is consistent but impossible to connect to if one holds other values.
  • Fiona holds many ideas and can thoughtfully connect them to each other. But asked whether she has tried to apply any of his ideas in practice or observed them in application, she demurs. Fiona’s network is well structured for talk but disconnected from experience.

This list can be extended. The point is that the structure of a moral network is important. That follows from the premise that we each begin with whatever ideas and connections we happen to hold, and our responsibility is to refine the whole set in discussion and collaboration with others. In that case, we should be concerned not only about the various values that we endorse, but also with how they are configured. The best networks for discussion are likely rich, complex, connected, not overly centralized, and not necessarily fully consistent.

Notes

  1.  Cf. Owen Flanagan, “Ethics Naturalized: Ethics as Human Ecology,” in Larry May, Andy Clark , and Marilyn Friedman (eds.) Mind and Morals: Essays on Ethics and Cognitive Science (Cambridge, MA: MIT, 1998) p. 30: “The community itself is a network providing constant feedback to the human agent.”
  2. See Richard N. Boyd, “How to be a Moral Realist,” in Geoffrey Saye-McCord, Essays on Moral Realism (Ithaca, NY: Cornell University Press, 1988), p. 205: ““Much [moral] knowledge is genuinely experimental knowledge and the relevant experiments are (“naturally” occurring) political and social experiments whose occurrence and whose interpretation depends both on “external” factors and upon the current state of our moral understanding.” Cf. Friedrich A. Hayek, The Constitution of Liberty (Chicago: University of Chicago Press, 1960), pp. 59-65
  3. Habermas’ preferred metaphor is a horizon, but he explicitly mentions networks in Jürgen Habermas, Between Facts and Norms: Contributions to a Discourse Theory of Law and Democracy, trans. by William Rehg (Cambridge: MIT Press, 1996), p. 18.
  4. Geoffrey Sayre-McCord, “Coherentist Epistemology and Moral Theory,” in Saye-McCord, p. 154.

empathy: good or bad?

1 Reply

I am speaking next week on a panel about empathy:

“Generative Empathies” (Rabb Room, Lincoln Filene Hall, Tufts University, March 30, 12 pm) with …

  • Amahl Bishara, Assistant Professor, Department of Anthropology, Tufts
  • Doris Sommer, Ira Jewell Williams, Jr. Professor of Romance Languages and Literature and Director, Cultural Agents Initiative, Harvard University
  • Peter Levine, Associate Dean for Research, Tisch College

I don’t know quite what to say yet, but I am inclined to raise the following points.

First, for a very long time, writers have argued that sad stories generate empathy and improve the character. From his dismal exile on the shore of the Black Sea, the poet Ovid addresses a soldier friend in these lines:

Is it true? When you heard of my misfortune
From a distant land, was your heart sad?
You can hide and shrink to say it, Graecinus,
But if I know you well, it was sad.
Revolting cruelty does not fit your type,
And even less your avocation. For
The liberal arts, your highest concern,
Soften the chest so that harshness escapes.
— Ex Ponto, 1.6 (my trans.)

Ovid presumes that his story will soften the gruff Roman’s heart, especially because it comes in the form of a poem and the soldier is a devotee of the artes ingenuae: the liberal arts, or literally, the freeborn arts. The poem will work because the reader has been habituated by many previous poems to dislike cruelty. Apparently, “ingenuae” has aristocratic connotations, and so Ovid’s phrase for the “liberal arts” implies a higher class of people who have been civilized or humanized by the arts.

Here is another classic source for the idea that writing generates empathy:

  1. And early in the morning, he came again into the temple, and all the people came unto him; and he sat down, and taught them.
  2. And the scribes and Pharisees brought unto him a woman taken in adultery; and when they had set her in the midst,
  3. They say [sic] unto him, Master, this woman was taken in adultery, in the very act.
  4. Now Moses in the law commanded us, that such should be stoned: but what sayest thou?
  5. This they said, tempting him, that they might have to accuse him. But Jesus stooped down, and with his finger wrote [or drew] on the ground.
  6. So when they continued asking him, he lifted himself up, and said unto them, He that is without sin among you, let him first cast a stone at her.
  7. And again he stooped down, and wrote [not drew] on the ground.
  8. And they which heard it, being convicted by their own conscience, went out one by one, beginning at the eldest, even unto the last: and Jesus was left alone, and the woman standing in the midst.
  9. When Jesus had lifted up himself, and saw none but the woman, he said unto her, Woman, where are those thine accusers? Hath no man condemned thee?
  10. She said, No man, Lord. And Jesus said unto her, Neither do I condemn thee: go, and sin no more (John, 8:2?11)

What was Jesus writing? One answer: something concrete about the specific Woman, which made the scribes and Pharisees think about her (and about themselves) instead of applying the abstract law.

For centuries in the English-speaking world, to enter the ranks of the civilized and humane meant reading Shakespeare. One possible reason: Shakespeare’s special capacity for empathy, which is related to his refusal to push arguments of his own. Keats found in Shakespeare the quality that he called “Negative Capability, that is when man is capable of being in uncertainties, Mysteries, doubts, without any irritable reaching after fact and reason.” Other critics have noted Shakespeare’s remarkable ability not to speak on his own behalf, from his own perspective, or in support of his own positions. Coleridge called this skill “myriad-mindedness,” and Matthew Arnold said that Shakespeare was “free from our questions.” Hazlitt said that the “striking peculiarity of [Shakespeare’s] mind was its generic quality, its power of communication with all other minds–so that it contained a universe of feeling within itself, and had no one peculiar bias, or exclusive excellence more than another. He was just like any other man, but that he was like all other men.”

So we have a model of the humane and sensitive educated person as one who has been habituated by the reading of moving stories to be empathetic and thus to show mercy or otherwise depart from harsh decisions.

This model conflicts with the idea that a just person knows the truth and obeys the consequences.  St. Augustine recalls his sinful younger self enjoying the theater, where he was “forced to learn I don’t know what wanderings of Aeneas, oblivious to my own, and to lament the dead Dido, because she killed herself for love, while meanwhile with dry eyes I endured my miserable self dying among these things before you, God, my life. … In the theaters I took pleasure along with the lovers when they used each other for vice, even though their behavior was just the imaginary sport of a play, and when they parted I was sad along with them, as if I were really compassionate; yet I enjoyed both parts.”  At the moment of his conversion, Augustine hears a voice saying, “take up and read, take up and read.” He understands this as a command to open the Bible at random. The first words he finds are those of Paul: “But put ye on the Lord Jesus Christ, and make not provision for the flesh, to fulfil the lusts thereof” [Rom. 13:13?14]. Augustine recalls: “I wanted to read no further, nor was there need” (Conf., 1.13.20; Conf. 3.2.3.; Conf., 8.12.29).

This is a model of the just person as one who is unmoved by inappropriate empathy and who avoids reading texts that might make him sympathize with sin. Although he is a very different kind of person from Augustine, Judge Richard A. Posner writes in “Against Ethical Criticism” that “empathy is amoral.”

Imaginative literature can engender in its readers emotional responses to experiences that they have not had. We read King Lear and feel how–or some approximation to how–a failing king feels, the wicked bastard feels, the evil daughters, the good daughter, the blinded earl, the faithful retainer, the corrupt retainer, the fool, all feel. We experience simulacra of the agony of madness and the pang of early death in Hamlet, the depths of mutual misunderstanding in The Secret Agent, the loneliness of command in Billy Budd, the triumph of the will in Yeats’s late poetry. This is the empathy-inducing role of literature of which [Hilary] Putnam and [Martha] Nussbaum speak. But empathy is amoral. The mind that you work your way into, learning to see the world from its perspective, may be the mind of a Meursault [from The Stranger], an Edmund [from Lear], a Lafcadio [the lion?], a Macbeth, a Tamerlane, a torturer, a sadist, even a Hitler (Richard Hughes’s The Fox in the Attic).

Empathy can even undermine justice. It can make the empathetic person feel more virtuous without doing anything, and it can even strengthen his position in a conflict by making him look better to third parties. This can be true of sincere empathy. I believe, for instance, that the median Israeli voter has achieved some empathy for Palestinians, and that feeling both blunts the urgency of justice and makes Israel look better than it should in the eyes of the world. Note the applause in this speech by Barack Obama in Jerusalem on March 21, 2013:

I — I’m going off script here for a second, but before I — before I came here, I — I met with a — a group of young Palestinians from the age of 15 to 22. And talking to them, they weren’t that different from my daughters. They weren’t that different from your daughters or sons.

I honestly believe that if — if any Israeli parent sat down with those kids, they’d say, I want these kids to succeed. (Applause.) I want them to prosper. I want them to have opportunities just like my kids do. (Applause.) I believe that’s what Israeli parents would want for these kids if they had a chance to listen to them and talk to them. (Cheers, applause.) I believe that. (Cheers, applause.)

In sum: I don’t think empathy will suffice on its own. It must be connected somehow with justice and with actually taking just action. If you favor systematic moral theories, than you may recommend using one or more general moral premises that distinguish good empathy from bad empathy. A feeling of empathy will not be a reliable guide to right action, only an urge that you must critically assess in other terms.

If, like me, you are skeptical about organized moral theories and believe that empathetic responses can convey truths about the world, then you will view an empathetic response as a valid source of guidance. But not as the only kind of valid input: relatively abstract and impersonal considerations must also apply.