Less Wrong is a community blog devoted to refining the art of human rationality. Please visit our About page for more information.

Modularity, signaling, and belief in belief

18 Kaj_Sotala 13 November 2011 11:54AM

This is the fourth part in a mini-sequence presenting material from Robert Kurzban's excellent book Why Everyone (Else) Is a Hypocrite: Evolution and the Modular Mind.

In the previous post, Strategic ignorance and plausible deniability, we discussed some ways by which people might have modules designed to keep them away from certain kinds of information. These arguments were relatively straightforward.

The next step up is the hypothesis that our "press secretary module" might be designed to contain information that is useful for certain purposes, even if other modules have information that not only conflicts with this information, but is also more likely to be accurate. That is, some modules are designed to acquire systematically biased - i.e. false - information, including information that other modules "know" is wrong.

continue reading »

Strategic ignorance and plausible deniability

36 Kaj_Sotala 10 August 2011 09:30AM

This is the third part in a mini-sequence presenting material from Robert Kurzban's excellent book Why Everyone (Else) Is a Hypocrite: Evolution and the Modular Mind.

The press secretary of an organization is tasked with presenting outsiders with the best possible image of the organization. While they're not supposed to outright lie, they do use euphemisms and try to only mention the positive sides of things.

A plot point in the TV series West Wing is that the President of the United States has a disease which he wants to hide from the public. The White House Press Secretary is careful to ask whether there's anything she needs to know about the President's health, instead of whether there's anything she should know. As the President's disease is technically something she should know but not something she needs to know, this allows the President to hide the disease from her without lying to her (and by extension, to the American public). As she then doesn't need to lie either, she can do her job better.

If our minds are modular, critical information can be kept away from the modules that are associated with consciousness and speech production. It can often be better if the parts of the system that exist to deal with others are blissfully ignorant, or even actively mistaken, about information that exists in other parts of the system.

In one experiment, people could choose between two options. Choosing option A meant they got $5, and someone else also got $5. Option B meant that they got $6 and the other person got $1. About two thirds were generous and chose option A.

A different group of people played a slightly different game. As before, they could choose between $5 or $6 for themselves, but they didn't know how their choice would affect the other person's payoff. They could find out, however – if they just clicked a button, they'd be told whether the choice was between $5/$5 and $6/$1, or $5/$1 and $6/$5. From a subject's point of view, clicking a button might tell them that picking the option they actually preferred meant they were costing the other person $4. Not clicking meant that they could honestly say that they didn't know what their choice cost the other person. It turned out that about half of the people refused to look at the other player's payoffs, and that many more subjects chose $6/? than $5/?.

There are many situations where not knowing something means you can avoid a lose-lose situation. If know your friend is guilty of a serious crime and you are called to testify in court, you may either betray your friend or commit perjury. If you see a building on fire, and a small boy comes to tell you that a cat is caught in the window, your options are to either risk yourself to save the cat, or take the reputational hit of neglecting a socially perceived duty to rescue the cat. (Footnote in the book: ”You could kill the boy, but then you've got other problems.”) In the trolley problem, many people will consider both options wrong. In one setup, 87% of the people who were asked thought that pushing a man to the tracks to save five was wrong, and 62% said that not pushing him was wrong. Better to never see the people on the tracks. In addition to having your reputation besmirched by not trying to save someone, many nations have actual ”duty to rescue” laws which require you to act if you see someone in serious trouble.

In general, people (and societies) often believe that if you know about something bad, you have a duty to stop it. If you don't know about something, then obviously you can't be blamed for not stopping it. So we should expect that part of our behavior is designed to avoid finding out information that would impose an unpleasant duty on us.

I personally tend to notice this conflict if I see people in public places who look like they might be sleeping or passed out. Most likely, they're just sleeping and don't want to be bothered. If they're drunk or on drugs, they could even be aggressive. But then there's always the chance that they have some kind of a condition and need medical assistance. Should I go poke them to make sure? You can't be blamed if you act like you didn't notice them, some part of me whispers. Remember the suggestion that you can fight the bystander effect by singling out a person and asking them directly for help? You can't pretend you haven't noticed a duty if the duty is pointed out to you directly. As for the bystander effect in general, there's less of a perceived duty to help if everyone else ignores the person, too. (But then this can't be the sole explanation, because people are most likely to act when they're alone and there's nobody else around to know about your duty. The bystander effect isn't actually discussed in the book, this paragraph is my own speculation.)

The police may also prefer not to know about some minor crime that is being committed. If it's known that they're ignoring drug use (say), they lose some of their authority and may end up punished by their superiors. If they don't ignore it, they may spend all of their time doing minor busts instead of concentrating on more serious crime. Parents may also pretend that they don't notice their kids engaging in some minor misbehavior, if they don't want to lose their authority but don't feel like interfering either.

In effect, the value of ignorance comes from the costs of others seeing you know something that puts you in a position in which you are perceived to have a duty and must choose to do one of two costly acts – punish, or ignore. In may own lab, we have found that people know this. When our subjects are given the opportunity to punish someone who has been unkind in an economic game, they do so much less when their punishment won't be known by anyone. That is, they decline to punish when the cloak of anonymity protects them.

The (soon-to-expire) ”don't ask, don't tell” policy of the United States military can be seen as an institutionalization of this rule. Soldiers are forbidden from revealing information about their sexuality, which would force their commanders to discharge them. On the other hand, commanders are also forbidden from inquiring into the matter and finding out.

A related factor is the desire for plausible deniability. A person who wants to have multiple sexual partners may resist getting himself tested for sexual disease. If he was tested, he might find out he had a disease, and then he'd be accused of knowingly endangering others if he didn't tell them about his disease. If he isn't tested, he'll only be accused of not finding out that information, which is often considered less serious.

These are examples of situations where it's advantageous to be ignorant of something. But there are also situations where it is good to be actively mistaken. More about them in the next post.

Consistently Inconsistent

60 Kaj_Sotala 04 August 2011 10:33PM

Robert Kurzban's Why Everyone (Else) Is a Hypocrite: Evolution and the Modular Mind is a book about how our brains are composed of a variety of different, interacting systems. While that premise is hardly new, many of our intuitions are still grounded in the idea of a unified, non-compartmental self. Why Everyone (Else) Is a Hypocrite takes the modular view and systematically attacks a number of ideas based on the unified view, replacing them with a theory based on the modular view. It clarifies a number of issues previously discussed on Overcoming Bias and Less Wrong, and even debunks some outright fallacious theories that we on Less Wrong have implicitly accepted. It is quite possibly the best single book on psychology that I've read. In this posts and posts that follow, I will be summarizing some of its most important contributions.

Chapter 1: Consistently Inconsistent (available for free here) presents evidence of our brains being modular, and points out some implications of this.

As previously discussed, severing the connection between the two hemispheres of a person's brain causes some odd effects. Present the left hemisphere with a picture of a chicken claw, and the right with a picture of a wintry scene. Now show the patient an array of cards with pictures of objects on them, and ask them to point (with each hand) something related to what they saw. The hand controlled by the left hemisphere points to a chicken, the hand controlled by the right hemisphere points to a snow shovel. Fine so far.

But what happens when you ask the patient to explain why they pointed to those objects in particular? The left hemisphere is in control of the verbal apparatus. It knows that it saw a chicken claw, and it knows that it pointed at the picture of the chicken, and that the hand controlled by the other hemisphere pointed at the picture of a shovel. Asked to explain this, it comes up with the explanation that the shovel is for cleaning up after the chicken. While the right hemisphere knows about the snowy scene, it doesn't control the verbal apparatus and can't communicate directly with the left hemisphere, so this doesn't affect the reply.

Now one asks, what did ”the patient” think was going on? A crucial point of the book is that there's no such thing as the patient. ”The patient” is just two different hemispheres, to some extent disconnected. You can either ask what the left hemisphere thinks, or what the right hemisphere thinks. But asking about ”the patient's beliefs” is a wrong question. If you know what the left hemisphere believes, what the right hemisphere believes, and how this influences the overall behavior, then you know all that there is to know.

continue reading »

Modularity and Buzzy

24 Kaj_Sotala 04 August 2011 11:35AM

This is the second part in a mini-sequence presenting material from Robert Kurzban's excellent book Why Everyone (Else) Is a Hypocrite: Evolution and the Modular Mind.

Chapter 2: Evolution and the Fragmented Brain. Braitenberg's Vehicles are thought experiments that use Matchbox car-like vehicles. A simple one might have a sensor that made the car drive away from heat. A more complex one has four sensors: one for light, one for temperature, one for organic material, and one for oxygen. This can already cause some complex behaviors: ”It dislikes high temperature, turns away from hot places, and at the same time seems to dislike light bulbs with even greater passion, since it turns toward them and destroys them.” Adding simple modules specialized for different tasks, such as avoiding high temperatures, can make the overall behavior increasingly complex as the modules' influences interact.

A ”module”, in the context of the book, is an information-processing mechanism specialized for some function. It's comparable to subroutine in a computer program, operating relatively independently of other parts of the code. There's a strong reason to believe that human brains are composed of a large number of modules, for specialization yields efficiency.

Consider a hammer or screwdriver. Both tools have very specific shapes, for they've been designed to manipulate objects of a certain shape in a specific way. If they were of a different shape, they'd work worse for the purpose they were intended for. Workers will do better if they have both hammers and screwdrivers in their toolbox, instead of one ”general” tool meant to perform both functions. Likewise, a toaster is specialized for toasting bread, with slots just large enough for the bread to fit in, but small enough to efficiently deliver the heat to both sides of the bread. You could toast bread with a butane torch, but it would be hard to toast it evenly – assuming you didn't just immolate the bread. The toaster ”assumes” many things about the problem it has to solve – the shape of the bread, the amount of time the toast needs to be heated, that the socket it's plugged into will deliver the right kind of power, and so on. You could use the toaster as a paperweight or a weapon, but not being specialized for those tasks, it would do poorly at it.

To the extent that there is a problem with regularities, an efficient solution to the problem will embody those regularities. This is true for both physical objects and computational ones. Microsoft Word is worse for writing code than a dedicated programming environment, which has all kinds of specialized tools for the task of writing, running and debugging code.

continue reading »