The Making of the Mind: The Neuroscience of Human Nature (6 page)

Finally, the output of the face module is shallow and limited to only seeing the perceptual structure of a face. The Italian artist Arcimboldo painted a series of composite faces in which common objects made up the facial parts. In viewing such paintings, we become aware of both the configuration of the face and the nonfacial component objects. This kind of double awareness
occurs because the output of the face recognition module and the output of the system responsible for object recognition both gain access to consciousness. In patients suffering from selective damage to the module that performs the recognition of common objects other than faces, a different outcome takes place. When shown an Arcimboldo painting, such a patient is fully aware that it looks like a face, because the intact face recognition module automatically provides its output to consciousness. However, the face module cannot see any deeper to see the identity of the objects used as facial parts. Without an intact object recognition module working in tandem with the face recognition module, the Arcimboldo painting loses its intriguing effect.
⁸

Modules are thought to be biological adaptations that were selected for during human evolution and are now hardwired in the human genome. It is easy to see why face recognition would be important enough to warrant a dedicated perceptual network. Similarly, there is strong evidence for a speech recognition module that rapidly and accurately identifies the short bits of sound (i.e., the phonological segments or phonemes) that are the building blocks of words in spoken language. These perceptual modules have been theorized to feed into more abstract cognitive modules that further organize knowledge within long-term memory.
⁹
Knowledge about people includes cognitive modules for processing facial expression, language, nonverbal behaviors, and the mental states of others. It also includes modules for organizing knowledge about the self. Another important set of cognitive modules in the domain of social information concerns groups of people. For example, recognition of kin and appropriate ways of behaving toward those biologically related to us is one such cognitive module. Other cognitive modules are dedicated to knowledge about the in-group, the out-group, and other dimensions of group identity such as ideology. Cognitive scientists are still debating how many modules exist within the system of long-term memory. The cognitive modules that represent knowledge about people are certainly only a subset. Biological knowledge about plants and animals and physical knowledge about climate, landscapes, and other ecological features are also important for obtaining the resources needed to survive and reproduce.

Whereas modules are narrowly tailored to the needs of a specific domain, working memory is the
general-purpose brain system. It can combine knowledge stored in separate domains of long-term memory on a temporary basis. Consider again the Paleolithic artwork of the Hohlenstein-Stadel figurine. The module representing knowledge about the shape of the human body is stored separately in long-term memory from knowledge about animals, such as lions. The knowledge modules of Neanderthals were possibly as detailed as those of modern humans when it came to such fundamentally important domains as people and lions. But Neanderthals apparently never thought to combine the body of a man with the head of a lion in creating a work of art. This combination required a degree of executive attention and possibly visual and spatial storage capacity in working memory unavailable to them. The advanced form of working memory associated with modern human beings provided a general workspace where the modules of diverse domains could be brought together in innovative ways.
¹⁰
The artist was capable of combining knowledge from social modules representing knowledge about people with knowledge about animals as represented in biological modules. The man-lion can be seen as a symbolic representation of an abstract idea, such as the idea of a lion that has taken on human features or vice versa. Either way, the Hohlenstein-Stadel figurine illustrates that modern humans were able to combine in a common workspace multiple domains of knowledge to create artifacts with symbolic meaning.

Similarly, Neanderthals were able to master the procedural knowledge of how to make a stone scraper versus a spear point. A cognitive module for tool making was apparently well established in the Neanderthal mind as part of their physical knowledge of the environment. They could acquire expertise through learning and apprenticeship and store these procedural skills in long-term memory. However, Neanderthals seemed to lack what modern humans possessed: the advanced working memory—particularly executive attention—needed for inventing and planning new ways of doing things. Without innovation, there was no possibility of improving the design of tools from one generation to the next in Neanderthal culture, and this could account for the static archeological record regarding tool design.
¹¹

EVIDENCE FOR CONTINUITY

As will be described in the last section of the chapter, there are abundant reasons for the claim that the prefrontal cortex of the human brain endows us with a remarkable network of executive attention. Through it, modern humans can, with some effort, inhibit impulses that run counter to our goals for staying healthy or succeeding at school or work. We can delay the need for immediate gratification in order to obtain a superior reward at a later time. We can plan solutions to novel problems. Yet, according to the ensemble hypothesis, these enhancements of working memory were only part of the total story of the human revolution during the Upper Paleolithic. So before turning to characterizations of the power of human executive attention, the evidence for continuity in working memory with other primates will be outlined.

The prefrontal cortex refers to the portions of the frontal lobe that lie in front of the premotor and motor cortex. These motor regions, as well as perceptual regions of the occipital lobe, temporal lobe, and parietal lobe, all provide inputs to the prefrontal cortex. So, too, do the limbic regions and other subcortical structures. Because location and connections matter in the brain, just as they do in the human social world, the prefrontal cortex is positioned “to coordinate processing across wide regions of the central nervous system.”
¹²
The size of the frontal lobe and in particular the prefrontal cortex is quite small in other species compared with human beings. In cats and dogs, for example, the prefrontal cortex is but a small fraction of the total frontal lobe. This disparity in the proportion of the frontal lobe dedicated to prefrontal cortex is also seen in comparing the brain of a rhesus monkey with a human brain. Strikingly, the prefrontal cortex makes up about half of the human frontal lobe; this lobe in turn constitutes about a third of the cerebral cortex in humans.
¹³

It is well-established that the human frontal cortex is large and complex. Yet it is not obviously discontinuous with the frontal cortex of the species most closely related to human beings at a genetic level of comparison. Although a reorganization of the brain is apparent when comparing human beings to non-primate mammals, such as dogs and cats, and to primates, as represented by monkeys, the great apes present a different picture. When magnetic resonance
images are taken that allow a precise measurement of the size of the frontal lobe relative to the overall size of the brain, continuity is apparent in comparisons of the brain of a human (36.4–39.3 percent) and the brain of a chimpanzee (35–36.9 percent), orangutan (36.6–38.7 percent), and gorilla (35–36.7 percent). All of these species differed from the lower percentage of total cortex allocated to the frontal lobe in the gibbon, a lesser ape (27.5–31.4 percent), and in monkeys (29.4–32.3 percent).
¹⁴
These findings contradict older studies that may have suffered from looking at too few or too limited a range of brain sizes in chimpanzees and humans. Moreover, the research tentatively suggested that the prefrontal region “is as large as expected for an ape brain of human size, and that individual humans and individual great apes (but not lesser apes or monkeys) overlapped.”
¹⁵
Thus, it is harder than was once thought to pinpoint a reorganization of the frontal lobe in general and even the prefrontal cortex in particular that resulted in the superior executive function in human beings.

Continuity between human beings and the great apes can also be seen in a comparison of the storage capacity of working memory. It has been firmly established that the capacity of working memory in modern humans is strictly limited to four items or chunks of information. It had once been thought that human beings could store about seven chunks of information—the length of a telephone number—in the phonological loop. Yet careful investigations revealed that this higher estimate reflected the benefit of rehearsing some of the items and of retrieval from long-term memory. Once a pure measure of working memory capacity is taken, without contamination from long-term memory, a surprisingly small and strict limit of only four chunks is observed.
¹⁶
For example, suppose you are trying to keep track of the names of new acquaintances just introduced to you at a party. At the very most, you will be able to retain four such first names within the capacity of the verbal store of working memory. The only way to do better than that is to rehearse some of the names silently in an effort to learn and store them in long-term memory or to use some other mnemonic technique. Obviously, such a verbal test cannot be done with great apes because they lack language and a specialized store for maintaining names. However, they share with us the visuo-spatial sketchpad for maintaining images of objects and scenes. Because the limit of four chunks
also applies to the visuo-spatial sketchpad, it is possible to devise a test that directly compares human and nonhuman species.

The Corsi Tapping Test is a common way of assessing the ability to retain nonverbal information for a short period of time. A set of identical icons appears on a touch screen of a computer. The icons then flash or change color one at a time in a sequence. The test taker must then reproduce the sequence by touching the icons on the screen in the correct serial order. One study adapted this test for baboons by rewarding the animal with food when it correctly recalled the sequence by touching the items on the screen in the right order.
¹⁷
The number of items in the sequence was varied. Two baboons were tested and both were about 70 percent accurate for three items and only about 20 to 30 percent accurate for four items. A single item more, using a five-item display, resulted in near zero accuracy. For human beings, by comparison, the accuracy rate was 97 percent, 92 percent, and 78 percent. Whereas human beings were clearly better at the task overall, they made errors relatively often when more than four items needed to be retained. For the baboons, this limit was three items instead of four. So the capacity of the visuo-spatial sketchpad in baboons is a bit more limited than in human beings, but the difference is by no means massive.

Finally, the studies that initially pointed to the prefrontal cortex as the site of working memory were performed with monkeys using a different nonverbal test.
¹⁸
In a spatial delayed response task, the animal must remember which of two locations contains a food reward. The experimenter baits one of the two locations by slipping food into a covered tray and then blocks the view of the two trays during a short delay period. The monkey gets the food if it can remember where the food is located after it has been out of sight for a while. Electrodes are surgically implanted into the brain of the animal to allow the recording of neural firings while the working memory task is being performed. Such work revealed that neural activity in the dorsolateral prefrontal cortex is central to accurate performance in the task. The important point is that the analogous brain region in human beings has been implicated in spatial working memory. In other words, primates of all kinds rely on the same prefrontal cortical regions to retain the spatial locations of objects over short periods of time.

THE PHONOLOGICAL LOOP

Although the human frontal lobe is not disproportionately large, given the overall size of the brain, it is indeed massive. On average, its volume measures five to six times the volume of the frontal cortex of a chimpanzee; the same overwhelming size advantage applies in comparing the human prefrontal cortex to that of chimpanzees.
¹⁹
What use do we humans make of all this cortical space? If we use the metaphor of a house with, say, four times the square footage of another smaller house, then how has the added space been allocated? One option would be to add a room with all the extra space. Perhaps what had been a small entry space just inside the front door could be expanded into a separate parlor, for instance. In other words, the new extra space could be used to add a new room altogether rather than just making the entry way and the living room proportionally larger. If a new room is thought of as a new cognitive function, then it is instructive to use this metaphor, for the human brain is unique in dedicating cortical areas for a verbal store of working memory specialized for learning and using language.

As adults we rely on the phonological loop for comprehending complete spoken sentences. It is also used as you silently read these words, providing an inner voice that accompanies the eye movements of reading. The phonological loop is also important in the production of speech and writing. However, at the most fundamental level the phonological loop is arguably an adaptation that allows very young children to learn words, particularly during the first few years of their lives.
²⁰
By allowing the phonological representation of a novel word to linger in the mind's ear, so to speak, the child can more readily learn to produce the new word and to link the word with its meaning.

Vocabulary acquisition is an enormously important task for the developing child—it is in fact essential for acquiring the culture of one's parents, family, and community. Learning to communicate through language of some sort—spoken or signed—may be the single most important factor in a child's survival and success in life. For children with normal hearing who are immersed in a culture of the spoken word, the acquisition of their native tongue comes easily and at a remarkably fast pace in terms of vocabulary as well as grammar. For word learning, the “rate of acquisition increases during
infancy, so that by the age of 5 years, mean vocabulary is in excess of 2000 words. Peak rates of vocabulary growth occur during the school years…typically…on average 3000 words every year.”
²¹
When children are assessed for their capacity to store information in the phonological loop, they are asked to remember a series of random digits or to try to repeat back nonwords. The digit span test is like trying to remember a new phone number, for example. The repetition test uses nonwords to prevent the reliance on meaning or any other representation of an actual word stored in long-term memory. Instead one must remember phonological segments or syllables that are similar to words but do not fit the pattern of an actual word (e.g.,
woogalamic
). From subjects ages three to eight years of age, researchers have found statistically significant correlations between a child's vocabulary size and the ability to perform the digit span and nonword repetition test.
²²
Also, when adults try to learn the words of a foreign language in a laboratory setting, their vocabulary learning is impeded when the unfamiliar words sound very similar and are easily confused within the phonological loop. Other experiments injected irrelevant sounds into the phonological loop by having the learner silently repeat the word “the, the, the” over and over again—this suppresses silent articulation and is a good way to distract the loop from accurately storing the sounds of a new foreign word. Learning foreign vocabulary is much impaired when adults are prevented from silently rehearsing each new word in the phonological loop.
²³
The evidence from both children learning their first language and adults learning a foreign language shows that the phonological loop has an important cognitive function as an aid to language learning.