Incognito (17 page)

Abraham Foxman, head of the
Anti-Defamation League, expressed outrage that there was no reference in the apology to the anti-Semitic slurs. In response, Gibson extended a longer note of contrition specifically toward the Jewish community:

Gibson offered to meet one-on-one with leaders of the Jewish community to “discern the appropriate path for healing.” He seemed genuinely contrite, and
Abraham Foxman accepted his apology on behalf of the Anti-Defamation League.

Are Gibson’s true colors that of an anti-Semite? Or are his true colors those he showed afterward, in his eloquent and apparently heartfelt apologies?

In a
Washington Post
article entitled “Mel Gibson: It Wasn’t Just the Tequila Talking,”
Eugene Robinson wrote, “Well, I’m sorry
about his relapse, but I just don’t buy the idea that a little tequila, or even a lot of tequila, can somehow turn an unbiased person into a raging anti-Semite—or a racist, or a homophobe, or a bigot of any kind, for that matter. Alcohol removes inhibitions, allowing all kinds of opinions to escape uncensored. But you can’t blame alcohol for forming and nurturing those opinions in the first place.”

Lending support to that outlook,
Mike Yarvitz, the television producer of
Scarborough Country
, drank alcohol on the show until he raised his blood alcohol level to 0.12 percent, Gibson’s level that night. Yarvitz reported “not feeling anti-Semitic” after drinking.

Robinson and Yarvitz, like many others, suspected that the alcohol had loosened Gibson’s inhibitions and revealed his true self. And the nature of their suspicion has a long history: the Greek poet
Alcaeus of Mytilene coined a popular phrase
En oino álétheia
(In wine there is the
truth), which was repeated by the Roman
Pliny the Elder as
In vino veritas
. The
Babylonian Talmud contains a passage in the same spirit: “In came wine, out went a secret.” It later advises, “In three things is a man revealed: in his wine goblet, in his purse, and in his wrath.” The Roman historian
Tacitus claimed that the Germanic peoples always drank alcohol while holding councils to prevent anyone from lying.

But not everyone agreed with the hypothesis that alcohol revealed the true Mel Gibson. The
National Review
writer
John Derbyshire argued, “The guy was drunk, for heaven’s sake. We all say and do dumb things when we are drunk. If I were to be judged on my drunken escapades and follies, I should be utterly excluded from polite society, and so would you, unless you are some kind of saint.” The Jewish conservative activist
David Horowitz commented on Fox News, “People deserve compassion when they’re in this kind of trouble. I think it would be very ungracious for people to deny it to him.” Addiction psychologist
G. Alan Marlatt wrote in
USA Today
, “Alcohol is not a truth serum.… It may or may not indicate his true feelings.”

In fact, Gibson had spent the afternoon before the arrest at the house of a friend, Jewish film producer
Dean Devlin. Devlin stated,
“I have been with Mel when he has fallen off, and he becomes a completely different person. It is pretty horrifying.” He also stated, “If Mel is an anti-Semite, then he spends a lot of time with us [Devlin and his wife, who is also Jewish], which makes no sense.”

So which are Gibson’s “true” colors? Those in which he snarls anti-Semitic comments? Or those in which he feels remorse and shame and publicly says, “I am reaching out to the Jewish community for its help”?

Many people prefer a view of human nature that includes a true side and a false side—in other words, humans have a single genuine aim and the rest is decoration, evasion, or cover-up. That’s intuitive, but it’s incomplete. A study of the brain necessitates a more nuanced view of human nature. As we will see in this chapter, we are made of many neural subpopulations; as Whitman put it, we “contain multitudes.” Even though Gibson’s detractors will continue to insist that he is truly an anti-Semite, and his defenders will insist that he is not, both may be defending an incomplete story to support their own biases. Is there any reason to believe that it’s not possible to have both racist and nonracist parts of the brain?

Throughout the 1960s, artificial intelligence pioneers worked late nights to try to build simple robotic programs that could manipulate small blocks of wood: find them, fetch them, stack them in patterns. This was one of those apparently simple problems that turn out to be exceptionally difficult. After all, finding a block of wood requires figuring out which camera pixels correspond to the block and which do not. Recognition of the block shape must be accomplished regardless of the angle and distance of the block. Grabbing it requires visual guidance of graspers that must clench at the correct time, from the correct direction, and with the correct force. Stacking requires an analysis of the rest of the blocks and adjustment to those details. And all these programs need to be
coordinated so that they happen at the correct times in the correct sequence. As we have seen in the previous chapters, tasks that appear simple can require great computational complexity.

Confronting this difficult robotics problem a few decades ago, the computer scientist
Marvin Minsky and his colleagues introduced a progressive idea: perhaps the robot could solve the problem by distributing the labor among specialized subagents—small computer programs that each bite off a small piece of the problem. One computer program could be in charge of the job
find
. Another could solve the
fetch
problem, and yet another program could take care of
stack block
. These mindless subagents could be connected in a hierarchy, just like a company, and they could report to one another and to their bosses. Because of the hierarchy,
stack block
would not try to start its job until
find
and
fetch
had finished theirs.

This idea of subagents did not solve the problem entirely—but it helped quite a bit. More importantly, it brought into focus a new idea about the working of biological brains. Minsky suggested that human minds may be collections of enormous numbers of machinelike, connected subagents that are themselves mindless.
¹
The key idea is that a great number of small, specialized workers can give rise to something like a society, with all its rich properties that no single subagent, alone, possesses. Minsky wrote, “Each mental agent by itself can only do some simple thing that needs no mind or thought at all. Yet when we join these agents in societies—in certain very special ways—this leads to intelligence.” In this framework, thousands of little minds are better than one large one.

To appreciate this approach, just consider how factories work: each person on the assembly line is specialized in a single aspect of production. No one knows how to do everything; nor would that equate to efficient production if they did. This is also how government ministries operate: each bureaucrat has one task or a few very specific tasks, and the government succeeds on its ability to distribute the work appropriately. On larger scales, civilizations operate in the same manner: they reach the next level of sophistication when they
learn to divide labor, committing some experts to agriculture, some to art, some to warfare, and so on.
²
The division of labor allows specialization and a deeper level of expertise.

The idea of dividing up problems into subroutines ignited the young field of artificial intelligence. Instead of trying to develop a single, all-purpose computer program or robot, computer scientists shifted their goal to equipping the system with smaller “local expert” networks that know how to do one thing, and how to do it well.
³
In such a framework, the larger system needs only to switch which of the experts has control at any given time. The learning challenge now involves not so much how to do each little task but, instead, how to distribute who’s doing what when.
⁴

As Minsky suggests in his book
The Society of Mind
, perhaps that’s all the human brain has to do as well. Echoing
William James’ concept of instincts, Minsky notes that if brains indeed work this way—as collections of subagents—we would not have any reason to be aware of the specialized processes:

When scientists began to look into the brains of animals, this society-of-mind idea opened up new ways of looking at things. In the early 1970s, researchers realized that the frog, for example, has at least two separate mechanisms for detecting motion: one system directs the snapping of the frog’s tongue to small, darting objects, such as flies, while a second system commands the legs to jump in response to large, looming objects.
⁶
Presumably, neither of these systems is conscious—instead, they are simple, automated programs burned down into the circuitry.

The society-of-mind framework was an important step forward.
But despite the initial excitement about it, a collection of experts with divided labor has never proven sufficient to yield the properties of the human brain. It is still the case that our smartest robots are less intelligent than a three-year-old child.

So what went wrong? I suggest that a critical factor has been missing from the division-of-labor models, and we turn to that now.

The missing factor in Minsky’s theory was
competition
among experts who all believe they know the right way to solve the problem. Just like a good drama, the human brain runs on conflict.

In an assembly line or government ministry, each worker is an expert in a small task. In contrast, parties in a democracy hold differing opinions
about the same issues
—and the important part of the process is the battle for steering the ship of state. Brains are like representative democracies.
⁷
They are built of multiple, overlapping experts who weigh in and compete over different choices. As
Walt Whitman correctly surmised, we are large and we harbor multitudes within us. And those multitudes are locked in chronic battle.

There is an ongoing conversation among the different factions in your brain, each competing to control the single output channel of your behavior. As a result, you can accomplish the strange feats of arguing with yourself, cursing at yourself, and cajoling yourself to do something—feats that modern computers simply do not do. When the hostess at a party offers chocolate cake, you find yourself on the horns of a dilemma: some parts of your brain have evolved to crave the rich energy source of sugar, and other parts care about the negative consequences, such as the health of your heart or the bulge of your love handles. Part of you wants the cake and part of you tries to muster the fortitude to forgo it. The final vote of the parliament determines which party controls your
action—that is, whether you put your hand out or up. In the end, you either eat the chocolate cake or you do not, but you cannot do both.

Because of these internal multitudes, biological creatures can be conflicted. The term
conflicted
could not be sensibly applied to an entity that has a single program. Your car cannot be conflicted about which way to turn: it has one steering wheel commanded by only one driver, and it follows directions without complaint. Brains, on the other hand, can be of two minds, and often many more. We don’t know whether to turn toward the cake or away from it, because there are several little sets of hands on the steering wheel of our behavior.

Consider this simple experiment with a laboratory rat: if you put both food
and
an electrical shock at the end of an alley, the rat finds himself stuck at a certain distance from the end. He begins to approach but withdraws; he begins to withdraw but finds the courage to approach again. He oscillates, conflicted.
⁸
If you outfit the rat with a little harness to measure the force with which he pulls toward food alone and, separately, you measure the force with which he pulls away from an electric shock alone, you find that the rat gets stuck at the point where the two forces are equal and cancel out. The pull matches the push. The perplexed rat has two pair of paws on his steering wheel, each pulling in opposite directions—and as a result he cannot get anywhere.

Brains—whether rat or human—are machines made of conflicting parts. If building a contraption with internal division seems strange, just consider that we already build social machines of this type: think of a jury of peers in a courtroom trial. Twelve strangers with differing opinions are tasked with the single mission of coming to a consensus. The jurors debate, coax, influence, relent—and eventually the jury coheres to reach a single decision. Having differing opinions is not a drawback to the jury system, it is a central feature.