Category Archives: science, technology and society

Tamsin Shaw’s critique of moral psychology

Leave a reply

I think that Tamsin Shaw’s article “The Psychologists Take Power” (New York Review of Books, February 25, 2016) is very important. I enjoyed an informal seminar discussion of it on Friday, but that conversation made me realize that the article is rather compressed and allusive, and its argument may not convey to readers who are unfamiliar with the research under review or with important currents in moral philosophy.

This is how I would reconstruct Shaw’s argument:

First, the psychological study of morality presents itself as a science; it claims to be value-neutral and strictly empirical. The phenomena under study are called “moral,” but the researchers purport or at least strive to be value-free.

Given that self-understanding, psychologists are attracted to three research programs: evolutionary biology, neuroscience, and game theory. Each presents itself as value-neutral. The three programs can be made highly consistent if one focuses on rapid human reactions to very basic stimuli, such as sexual desire or perceived threat. These reactions presumably arose well before cultural differentiation, they have Darwinian explanations, they would serve individuals or groups in competitive situations (e.g., while struggling for food or mates), and they light up specific parts of the brain. Findings that seem consistent with all three streams of research have special prestige because they seem particularly hard-headed and empirical. (A perfect example is the Times’ article yesterday: “What’s the Point of Moral Outrage? It may seem noble and selfless, but it’s also about improving your reputation.”)

People who think this way about morality are basically amoral. They have no independent moral compass. Yet they learn techniques that are useful for manipulating subjects, particularly in extreme situations where instinctive human impulses are most pertinent. Therefore, it is no surprise (Shaw writes) that some of them became professional advisers on torture during the first years of the Iraq occupation. Any argument against torture will seem to them arbitrary and subjective.

The last point may be a bit of an ad hominem, although it is certainly worth taking seriously as a warning. But even if all psychologists use good professional ethics, the agenda of making moral psychology strictly empirical needs to be challenged.

For one thing, you can’t study phenomena categorized as “moral” without independently deciding what constitutes morality. We have many deep, instinctive impulses. For instance, we are capable of altruism and even self-sacrificing love, but also of violence and greed. It’s plausible that many of these impulses have evolutionary roots and can be explained in game-theoretic terms. But only some of them are moral. Imagine, for instance, that I said, “Greed is a moral virtue that we developed early in our evolution as a species to motivate individuals to maximize resources.” This would not be a scientifically false statement. It would be morally false. The mistake is to call greed a “virtue.”

Jonathan Haidt likes to provoke liberals by describing “authority” and “sanctity” as moral values. They may be, but that requires a moral argument against the position that only care, fairness, liberty, and loyalty count as moral. The fact that some people see authority and sanctity as virtues does not make that opinion right. Hitler thought that racial purity was moral, and he was wrong. So moral reasoning is indispensable.

Further, when we reason morally, we are usually thinking about very complex, socially constructed phenomena that we don’t directly perceive. We certainly don’t experience them as immediate sense-data. I wrestle with my feelings about democracy, the United States, academia, capitalism modernity, etc. These things don’t appear in my visual field like violent threats or piles of yummy food. I experience such institutions through speech and text, through vicarious reports, and by accumulating experience and arguments over decades. Possibly the impulses that homo sapiens developed early in our evolution influence my judgments. For instance, I may have a deep, unconscious tendency to separate people into in-groups and out-groups, and that may affect my tendency to see the USA as my group. But I could treat another unit as my main group, I could be uninterested in (or even unaware of) the USA as an entity, or the country might not even exist. A nation is a social construction, built by people for complex reasons, that we understand in a mediated way. It would be a contentious assumption, not a hard-nosed scientific premise, that our most primitive impulses have much to say about institutions or our attitudes toward them.

See also: Jonathan Haidt’s six foundations of morality; neuroscience and morality; morality in psychotherapy; on philosophy as a way of life; is all truth scientific truth?; and right and true are deeply connected.

why social scientists should pay attention to metaphysics

Leave a reply

Yesterday, I introduced the substance of Brian Epstein’s book The Ant Trap. Epstein analyzes the metaphysics of social phenomena, such as groups. Here I want to argue that social scientists should be more attuned to metaphysical issues in general.

In social science, we think naturally of certain relationships, such as correlation and causation, and of certain kinds of objects, such as individuals and groups. But other relationships are present although less explicit in our work. For instance, the members of the US Congress do not cause the Congress; they compose it. Composition is a relationship that is named (but rarely explored) in standard social science.

One can ask, more generally, what kinds of relationships exist and what kinds of things are related to each other. Constitution and causality are two different relationships. Groups, moments in time, and ethical qualities are three different kinds of things. These types and relationships can go together in many ways. We can ask about their logic or their epistemology, but when we ask specifically, “What kinds of things are there and how do they go together?” we are putting the question in terms of metaphysics.

Social scientists should be concerned with metaphysics for two big reasons. First, in our actual writing and modeling, we often use some metaphysical terms (e.g., object, composition, causation), but only a few of those get explicit critical attention. In my experience, most of the meta-discussion is about what constitutes causality and how you can prove it—but there are equally important questions about the other relationships used in social science.

Second, professional philosophers have developed a whole set of other types and relationships that are typically not mentioned in social science but that can be powerful analytical tools if one is aware of them: supervenience, grounding, and anchoring being three that play important roles in The Ant Trap.

Since metaphysics is a subfield of philosophy, and since philosophers are probably outnumbered 50-to-one by social and behavioral scientists, it’s easy for the latter to overlook metaphysics. In fact, I suspect that the word “metaphysics” (as modern academic philosophers use it) is not well known. If you Google “metaphysical relationships,” you will see New Age dating tips. But all scientific programs involve metaphysics, and it is important to understand that discourse–not only to be more critical of the science but also to develop more powerful models.

is social science too anthropocentric?

Leave a reply

Consider these statements: “A group just is the people who make it up.” “If a group can be said to have intentions at all, its intentions must somehow be the intentions of its members.” Or: “When a convention arises, such as the convention that a dollar has value, it must exist because the people who use dollars have imposed some meaning on material reality.”

In The Ant Trap: Rebuilding the Foundations of the Social Sciences, Brian Epstein criticizes an assumption that is implicit in these statements (which are mine, not his): that social phenomena can be fully explained by talking about people. It’s obvious that non-human phenomena–from evolution to climate change–influence or shape human beings. But the thesis that people fully determine social phenomena is worth critical scrutiny.

Epstein’s book is methodical and not subject to a short paraphrase, but some examples may give a flavor of the argument. For instance, is Starbucks composed of the people who work for it? Clearly not, because the coffee beans and water, the physical buildings, the company’s stock value, the customers and vendors, the rival coffee shops in the same markets, and many other factors make it the company that we know, just as much as its own people do. Indeed, its personnel could all turn over through an orderly process and it would still be Starbucks.

Likewise, if the Supreme Court intended to overturn the ban on corporate campaign contributions, was its intention a function of the preferences of the nine individual justices? No, because in order for them to intend to overturn the ban, they had to be legitimate Supreme Court justices within a legal system that presented them with this decision at a given moment. I could form an opinion of the Citizens United case, but I could not “intend” to rule for the government in that case, because I am not a justice. And what makes someone a justice at the moment when the Citizens United case comes before the court is a whole series of decisions by people not on the court, going back to founding era.

In general, Epstein writes, “facts about a group are not determined just by facts about its members.” And it’s not just other people who get involved. Non-human phenomena can be implicated in complicated ways. For instance, the Supreme Court is in session on certain days, and on all other days, a “vote” by a justice would not really be a vote. What makes us say that a certain day has arrived is the movement of the earth around the sun. So the motion of a heavenly body is implicated in the existence and the intentions of the Supreme Court. That is an apt example, because Epstein calls for a Copernican Revolution in which we stop seeing the social world as “anthropocentric.”

Note that we are talking here about grounding relations, not causation. Public opinion may influence the composition of the Supreme Court and its decisions. The movement of the earth does not influence or affect the Court, and you wouldn’t model it that way (with the earth as an independent variable). Rather, the court is in session on certain dates, and the calendar is grounded in facts about the solar system. Likewise, a president can influence the court, and you could model the president’s ideology as an independent variable. But the composition of the court is grounded in decisions by presidents and senates in a more fundamental way than causation. To be a justice is (in part) to have been nominated and confirmed.

When people criticize anthropocentrism, usually they mean to take human beings down a peg. But in this case, the critique is a testament to our creativity and agency. Human beings can create groups in limitless ways. We can intentionally ground facts about groups in circumstances beyond the control of their members, or indeed in facts that are under no human’s control (like the motion of the earth). It can be wise to limit the power of group members in just these ways. Epstein writes, “Our ability to anchor social facts to have nearly arbitrary grounds is the very thing that makes the social world so flexible and powerful. Why would we deprive ourselves of that flexibility?” But the same flexibility that empowers the human beings who design and operate groups also creates headaches for the analysts who try to model their work. “Compared to the social sciences, the ontology of natural science is a walk in the park.”

The Ant Trap does not offer one model as an alternative to the standard anthropocentric ones, because social phenomena are diverse as well as complex. But if we narrow the focus a bit from the whole social world and look at groups, they tend to require (in Epstein’s analysis), a two-level model. Various facts about each group are grounded in other facts. For instance, the fact that the Supreme Court is in session is grounded in facts about the calendar (as well as many other kinds of facts). In turn, these grounding relationships are anchored in different facts–for instance, facts about how US Constitution organized the judiciary system.

My day job involves very conventional social science. We study various groups, from Millennials and voters to Members of Congress. After reading The Ant Trap, I won’t think of groups in the same way again. I am not yet sure what specific methodological implications follow, but that seems an important question to pursue.

See also Brian Epstein’s TedX Standford talk, which captures some of the book.

 

Bent Flyvbjerg and social science as phronesis

1 Reply

I am co-teaching the Summer Institute of Civic Studies and blogging about roughly half of the 18 topics on our syllabus. Last Friday morning, we discussed Bent Flyvbjerg and social science as phronesis. The readings were:

  • Bent Flyvbjerg, “Social Science that Matters” (2006)
  • Bent Flyvbjerg, “Making Organization Research Matter: Power, Values and Phronesis” (2006)
  • Bent Flyvbjerg, Making Social Science Matter, Chapter 10, pp. 141-65
  • David Garvin, “Making the Case,” Harvard Magazine, September-October 2003

Flyvbjerg started as a planning professor in Denmark who uncovered and successfully addressed corruption in the city of Aalborg by finding specific information and using it to challenge power. The key moment was when he discovered five lines of a specific memo which guaranteed that “The Chamber of Industry and Commerce” (a private interest) would be included in all official meetings about redesigning the city center. He asked why they would be there, and that revealed a whole scandal. The Chamber was actually pro-car and had been the most powerful force in city planning.

Flyvbjerg has since developed a broader theory of social science as “phronesis,” the Greek word for practical wisdom. His theory has been highly influential; for example, it provided one inspiration for the Perestroika movement in political science.

In Latin, the word phronesis was translated as prudentia. The English derivative word “prudence” is too narrow, but it reminds us that phronesis is a virtue, not just a cognitive skill. Aristotle distinguishes phronesis from:

  • Sophia (wisdom about abstractions, often spiritual in nature)
  • Techne (art), which is knowledge about how to make things or make things work
  • Episteme (knowledge that consists of general propositions, connected logically, and therefore teachable)

Flyvbjerg asserts that social science tries to be an episteme, but as such, it does not work. “No predictive theories have been arrived at in social science, despite centuries of trying. This approach is a wasteful dead-end.”

Continue reading →