AS WE HAVE SEEN, MUCH of the individual differences in human IQ seem to relate to differences in working-memory capacity. The comparative psychologist Tetsuro Matsuzawa and colleagues have conducted innovative studies on chimpanzees at Kyoto University that suggest surprising memory capacities. In these experiments the chimpanzee Ai faced a computer touch screen and was trained to press numbers in ascending order that appeared in random places on the screen. In one version, when the first number was pressed, the other numbers were masked with white squares, and Ai had to press them in order of the numbers that were no longer visible. She performed about 65 percent correctly when there was a total of five numbers on the screen. It has therefore been suggested that Ai had a working-memory capacity of five. Though impressive, this performance may only require recall of three, not five, numbers. The first number press need not involve memory (one can point straight away), and the last number is always whichever square is left.
A more recent study reported that some chimpanzees can do this with up to nine numbers (suggesting a memory of seven chunks). Ai’s son Ayumu could even beat humans on a version of the task when five numbers were presented for merely a fifth of a second on the screen—too quick to explore the screen through eye movement. This is a very impressive performance when you see it in action. The numbers appear and disappear, and then Ayumu quickly presses the five locations in ascending order. At a conference we were shown a video of a chimpanzee being interrupted on a trial, looking away, and then returning to complete the sequence in a flash. The researchers now argue that the chimpanzees solve the task with something akin to photographic memory. The findings that the chimpanzee did better than adult humans caused significant attention in scientific and nonscientific circles alike. The comparison was perhaps not entirely fair: humans did not receive anywhere near the practice on the task that the chimpanzee did. In follow-up research humans were given practice and subsequently outperformed the chimpanzee.
What, then, is the working-memory span of chimpanzees? No conclusive tests have yet been conducted that are directly comparable to human working-memory measures (which typically involve distractor tasks). Analysis of various task performances in the wild and in the laboratory led anthropologist Dwight Read to suggest that chimpanzee working-memory capacity is actually limited to between two to three concepts. He examined, for instance, the number of intelligible word combinations produced by Kanzi and Nim Chimpsky and the number of object combinations in natural tool use. Such a working-memory capacity might account for a lack of embedded thought and could represent a fundamental limit. A gradual increase of working-memory capacity in human evolution could explain a lot about the qualitative changes that characterize the human mind. This is an intriguing hypothesis. But since there is no established nonverbal measure of working memory in nonhuman animals, we will need to wait for further research before drawing a firm conclusion.
I NOTED PREVIOUSLY THAT ANIMALS can learn some things better than others. For instance, rats can learn that a taste, but not a sound, predicts later nausea. Evolution may have shaped what learning matters for a species. Given that rats are generalists that frequently explore novel food sources, learning to link taste with sickness is important. Many species show clever behavior in specific contexts but not in others. According to the primatologists Dorothy Cheney and Robert Seyfarth, animals show “laser-beam intelligence.” David Premack concurs and points to teaching in cats as an example of a clever but restricted ability that serves one goal: teaching hunting. Human teaching, by contrast, is domain general and can serve many goals, as I will discuss in the next chapter. Premack argues that what sets human intelligence apart is our flexibility. The abilities we have discussed thus far—language, foresight, mind reading, reasoning—are not fixed to a particular domain but can be employed to virtually endless varieties of goals.
Species differ in the way they can interact with the world. Some species have few means of responding to environmental challenges (e.g., a snail’s defense is to hide in its shell), and others have a diversity of options at their disposal (e.g., a monkey may threaten, hide, recruit support from its group, or climb to safety). The philosopher Kim Sterelny refers to this as “response breadth.” Humans can react to situations in flexible, diverse ways. We innovate new ways of responding to situations. We are also correspondingly curious. We seek out novel information and prefer situations that are likely to yield new insights. The German word for curious is Neugierig, which literally means “news greedy.” Indeed, we generally crave new information, and endorphins are released as we comprehend—they give us pleasure by activating the same opioid receptors that are activated by certain drugs. We all know that reading a good book can be rewarding. Irving Biederman has called humans “infovores” to highlight our innate hunger for novel, interpretable information. (He does not, however, restrict the term to humans.)
A classic study on zoo animals presented sets of wooden blocks, dowels, chains, and rubber tubing to over one hundred species. Primates and carnivores were found to be more than twice as curious as rodents and other mammals. Furthermore, great apes spend twice as much time observing the objects than the other primates tested. Our closest relatives are not only curious but also quite innovative. Andrew Whiten and I noted that chimpanzees who were unable to poke out a bolt necessary to obtain a treat from inside a puzzle box tried thirty-eight different ways to solve the problem. They employed one hand and two hands; they used their lips, feet, and a tool. They pushed, pulled, and poked. They grasped, gripped, and hit. They did not give up easily. Baboons used far fewer ways of exploring the same apparatus, in spite of the fact that their hands, including the thumb to finger ratios, are more like ours than those of chimpanzees. One way to examine response breadth, then, is to offer animals objects and record the diversity of responses they produce.
One study recorded the diversity of actions primates directed at a knotted rope that had one end secured outside the cage. There was no food reward to be gained; the researcher examined playful rather than functional manipulations. Great apes employed significantly more combinations of body parts and actions than any other primates. These results have been replicated: great apes appear particularly inventive. To put it a different way, their behavior is the least predictable. This result is in line with field observations of innovation and creativity with objects and social others,14 as well as with the diversity of socially maintained traditions we will discuss in the next chapter. Thus there appear to be some signs of gradual change in the evolution of intelligence and creativity. Our closest relatives are particularly adept at flexibly interacting with their environment.
Nonetheless, human response diversity is exponentially larger. Our inventiveness appears to know no bounds. We can generate virtually infinite combinations of elements, creating novelty in behavior, tools, and sentences. With language we can learn from the responses of others, even if we did not witness them ourselves. With mental time travel we can test consequences of potential actions, even if we do not physically try them out. We can therefore overcome obstacles and discover opportunities in our mind. We can treat scenarios as chunks of information and use placeholders to construct higher-order relations. We can decontextualize these relations and reason about entirely abstract concepts. We can construct elaborate theories about the forces that govern this world and systematically test whether they are correct. Only humans do science.
In a sense Cicero may have been right in asserting that “before all other things, man is distinguished by his pursuit and investigation of TRUTH.” The acquisition of knowledge is a goal driving many human endeavors, and we derive happiness from gaining understanding. We can build on others’ insights and observations to accumulate knowledge. We have established cultures that permeate virtually everything we do and help us act smart in our environment. To culture, then, we turn next.
1Individual scores are converted to a scale on which 100 still indicates the population mean. The distribution has a standa
rd deviation of 15, and this entails that just over two-thirds of people’s IQs fall between 85 and 115. If your IQ is 115, you score higher than 84 percent of the population; if it is 130, only 2 percent of people have a higher score.
2Especially in the United States, intelligence testing has been of immense influence. People who score low on IQ tests are more likely to be incarcerated, bear children outside of marriage, and be welfare recipients than those who score highly. These data have led to some debate about whether American society is increasingly becoming stratified according to intelligence. Note, however, that testing itself affects this result, as only people with certain scores will be given certain opportunities (e.g., to study).
3I once made a large version of a test called Raven’s progressive matrices (a series of pattern completion tasks, highly correlated with IQ) and tried to test the chimpanzees Cassie and Ockie. Alas, they did not understand the basic premise, and I quickly abandoned the attempt.
4In this light, the term “artificial intelligence” may be a misnomer. Computers have better memory than we do and may calculate numbers more quickly and accurately, but as long as they do not want to achieve anything, they may not be regarded as intelligent. This “want” refers not just to a goal (which is easily programmed) but to what William James called “having an interest.” Computers do not care if you turn them off—as far as I can tell.
5The psychologist Alan Baddeley proposed the term “working memory” and suggested that it comprises distinct components: a phonological loop and a visual-spatial sketch pad that operate independently from each other. Furthermore, he proposed a “central executive” that controls how these components are used. In a later version Baddeley added an “episodic buffer,” a limited capacity store in which information can be integrated, combined, and manipulated.
6Recent research suggests that capacity for integration of relations and for storage and processing of chunks are related but distinguishable concepts.
7Children who failed theory of mind tasks produced few correct solutions. They often searched the test room for answers, and a salient idea seemed to stay in their focus of attention. When asked to name things that are red, for instance, they might say, “fire engine,” and then keep listing related items, rather than to note that books, balls, or almost anything else could also be red. Flexible scanning of one’s own knowledge for potential solutions may require executive capacities to disengage from an answer and meta-representational skills to generate and evaluate potential solutions.
8It should not be surprising if some animals monitored uncertainty. Many species have to track indicators of, say, whether to attack or to flee, and it could be highly beneficial to recognize uncertainty about what one can or cannot handle. A monkey traversing the canopy needs to take into account how far it can jump.
9While big consequences can lead even to one trial learning (you do not need to burn your hand on a stove repeatedly), minor ones may not have much power in shaping behavior. Consider, for example, whether you turn your key left or right when opening a car. This is an action that you might have done thousands of times, and you get either rewarded when moving the right way (the door opens) or punished when moving the wrong way (the door is still shut). I still keep getting it wrong.
10Another recent study suggests that these crows can consider hidden humans as causal agents. When the crows saw a human go into a hide, then saw a stick moving through a hole in the hide’s wall, and finally saw the human leave again, they were more willing to approach the hide than when they just saw the moving stick. This suggests that the birds attributed the movement in the former condition to the hidden human.
11Paula Irving and I once tested dolphins on a similar task at SeaWorld. At first, we rewarded them for touching with their nose one of two boards that displayed a symbol they were shown beforehand on the other side of a pontoon. We then wanted them to match items that were perceptually different but represented an analog relationship. Alas, the dolphins were poor at this task. When they got it wrong, instead of examining the problem, they would try to hit the board harder or do a flip first or some other fancy trick. As we have seen, negative results are difficult to interpret. One possible explanation for our findings is simply that their interactions with humans typically involve fish rewards for acrobatics rather than for correct choices.
12It had been argued that Call’s tape recorder condition does not address the possibility that an association between the compound of shaking and sound was learned. To examine this, Andrew indicated the location of the treat through the sound of shaking a duplicate cup. The apes performed at chance in this condition, but clearly above chance when the target cup itself was shaken.
13For instance, when we recently gave New Caledonian crows the choice between two boxes with a food reward, one with a stick poking out with the meat skewered at the end and another where the food was either not attached to the stick, attached to a broken stick, or otherwise not functional, the birds acted randomly. When given the option between functional and broken tools to be used to rake in food, they immediately made the correct choice. Though avid stick tool users, it took them many trials to learn what to do when the stick already had food on it. Some things are easier to learn than others.
14Rates of behavioral innovation in animals are associated with brain size measures as well as with frequency of tool use and social learning.
EIGHT
A New Heritage
The primary difference between our species and all others is our reliance on cultural transmission of information and hence on cultural evolution.
—DANIEL DENNETT
WE ARE DEEPLY CULTURAL BEINGS. That does not mean that we are all connoisseurs of classical music, literature, and fine arts. Culture in a broad sense comprises everything enduring that we learn from others; it includes the commonplace—even banal—customs, values, knowledge, and objects our societies have invented and propagated. Shoes, for instance, are deeply cultural. Someone realized that adding a sole to one’s foot is a good thing, and ever since, people around the world have created new versions. You and I benefit from this knowledge, even if we did not conceive of the idea, obtain the raw material, design shoes, nor produce shoes—all we had to do was buy them. This cooperation is extraordinary. No monkey has any shoes—at least not any made and sold by other monkeys.
What is most powerful here is the accumulation of knowledge, skills, and artifacts over time. We benefit from what others have done long ago. We do not have to reinvent the wheel, as the saying goes. Someone invented the wheel some six thousand years ago or so, and the idea spread rapidly. From the original uses as potters’ wheels, such as in the Mesopotamian city of Ur, to chariots, mechanical clocks, pulleys, and the hula hoop—there are many thousands of uses to which this basic idea has been applied. We build upon the cultural achievements of others.
Daniel Dennett argues that cultural objects make us smarter. They allow us to do things we could not do before and hence enable us to explore new ways of intelligently interacting with the world. The day someone invented the boat, oceans of possibilities opened up before all humankind. Such broadening of horizons occurs for nonmaterial aspects of culture as well. Culture provides us with mind tools. Words, for instance, are tools not only for communication but for categorizing, thinking, and reasoning. We do not have to reinvent concepts and symbols anew but can acquire those of our group. The word “shoe” itself is a part of your cultural heritage. As we saw earlier, you may know tens of thousands of words, but few, if any, are words you invented yourself. Most of the concepts you have are imported from others. The ideas of “software” and “evolution” are recent cultural inventions, and you can use them without having had to lay the theoretical groundwork. A mind with words is profoundly different from one without, and different words may influence your thinking in different ways.1
Although our individual understanding is often flawed and our foresight misguided, by linking our minds to those of ot
hers we have enormously increased our predictive capacities and powers of control. With theory of mind and language we are able to wire our scenario-building minds into much larger networks. We teach each other and copy each other, allowing us to pass on what we have experienced, abstracted, innovated, or learned from another. Thus are populations able to socially maintain and accumulate knowledge, customs, and survival strategies.
Culture penetrates most of what we do. We are part of a larger matrix that links us to the cultural achievements of our forebears and contemporaries. Our minds are shaped by the cultural heritage of our group. Alone we may be weak, but together we are strong. Human culture has led to civilizations that, for better or worse, have changed much of the planet. This system is built on extraordinary levels of cooperation. In this chapter I will discuss culture first in terms of the fundamental problems of widespread cooperation that had to be overcome and then in terms of the key mechanisms through which culture is transmitted and changed.
WHATEVER YOU MIGHT THINK ABOUT humanity, we are a remarkably cooperative lot. People habitually collaborate with friends and family members, communities, teams, clubs, companies, societies, associations, and national or international institutions. With the ability to read minds and tell each other what we are thinking, we can coordinate our actions in an unprecedentedly flexible manner. With foresight we can construct and pursue long-term collaborative plans. We even cooperate with people we do not know. I banked on the truth of this claim when I hitchhiked through various parts of Europe, Asia, and America. Wherever I travelled, I benefited from the kindness and knowledge of strangers. You can cram us in a bus or a football stadium, and only rarely does mayhem ensue.
We also cooperate economically. For the right price most people will trade goods and services with almost anyone. Most of your possessions were likely made by others: your clothes, furniture, music, spices, art—certainly, the book you are reading right now. Though I am writing this down under, you may be reading it on the other side of the world. You are benefiting from my labor (whether or not you agree with what I write).
The Gap Page 19
