Robin Dunbar
The Human Story
A new history of mankind's evolutionb
faber and faber 2004

Dunbar pg 65

Deep Thought

One of the very conspicuous features of our mental world is the way we rehearse what we are going to do. This often entails explicitly considering alternative options, evaluating their likely outcomes and, having chosen one, rehearsing how we can best approach its execution. This process is so much a part of our mental life that we barely give it a moment's thought. But perhaps it provides us with the clue we have been looking for.

This kind of mental rehearsal is actually quite a complex task and involves bringing into play a number of quite different cognitive abilities. At the very least, these include the ability to reason causally (to follow the sequence through from cause to likely effect), to reason analogically (to recognise that A is to B as X is to Y), to run several alternative scenarios in parallel and, finally, to do so on an extended time frame into the future.

Analogical reasoning may be the dark horse here. Lera Boroditsky has in fact recently suggested that we use it to develop a sense of time off the back of a sense of space. Time, after all, is something that we can only imagine: we cannot touch or sense it directly. Instead, we reflect on memories of past circumstances, compare them with the corrent state of the world, and infer that time has passed. In contrast, we have direct sensory experience of the world of physical space through both vision and touch. Boroditsky showed that, when primed with statements about spatially arranged objects ('The flower is in front of me'), subjects were more likely to interpret an ambiguous statement about the temporal arrangement of an event ('Next Wednesday's meeting has been moved forward two days: which day is it now on?') in ways consistent with the spatial arrangement that they had been shown; however, a statement about a temporal event did not have the same effect on how they interpreted an ambiguous statement about a spatial arrangement. She argues that we develop a sense of time analogically from our sense of space, and this explains why we use so many spatial metaphors when referring to time. Things happen before other things; we look ahead or forward to the future; we fall behind schedule.

Analogical reasoning may play an unexpectedly crucial role in the story of the human mind because it provides us with a platform for understanding other minds. I use my experience of my own mental processes to imagine how someone else's mind might work. I use it to empathise with your predicament when the ATM swallows your cash card, as well as with your feelings when you stub your toe. This phenomenon may be even more important when it comes to dealing with the world of relationships. We can experience directly the interactions that pairs of animals engage in: we see these happening before our eyes, we may even experience them directly through touch or hearing.
But when it comes to understanding the nature of the relationship between two individuals, we have to deal with something that we cannot experience directly. As the ethologist Robert Hinde pointed out more than three decades ago, we abstract relationships out of the observations we make of individuals interacting. Relationships are things that happen only in a virtual world, and we have to be able to move backwards and forwards between the physical world of interactions (real events) and the virtual world in which these events are constituted into relationships in order to be able to understand what the significance of specific actions is or might be, or how two relationships impinge on each other.

Testing animals' abilities to engage in analogical reasoning is not exactly easy. Most attempts to explore this phenomenon have focused on the rather simpler task of whether they understand analogies between categories ('Tap is to water as key is to...[lock]?'). This is not quite the same thing as being able to exploit the analogy between how, say, one mind works and use it to model the behaviour of another mind. Or to take a social process and use it to model a physical process, or vice versa. All the studies of this phenomenon so far have concentrated on perceptual similarity rather than conceptual similarity. Conceptual similarity may be more important in dealing with the social world because relationships cannot be experienced directly. Attempts have been made, however, to test great apes' abilities on at least two of the other processes, namely causal reasoning and mental rehearsal.

The test we used to assess an understanding of causality is one that has been used extensively on very young children - even babies as yonng as six months. The design is relatively simple: a subject (human or otherwise) is shown a video clip of, say, one object hitting another, thereby causing it to move. The sequence is repeated over and over again until the child habituates (ceases to pay attention); it is then shown a clip in which movement occurs despite the fact that the two objects did not touch (or 'collide'). If it suddenly pays attention, this is taken as evidence that it appreciates that something odd or out of the ordinary has happened (at least, providing it can be shown that it does not respond as strongly to a simple change of scene). This is taken as evidence of an intuitive understanding of causality. Children as young as six months pass this test, and so it seems do chimpanzees - but not monkeys. Rather similar results using different kinds of causal reasoning tests have been obtained by the Italian psychologist Elisabeta Visalberghi and her collaborators in Rome.

We tested apes' ability to engage in mental rehearsal by comparing the time taken to open a puzzle box after the animals had had an opportunity to look at (but not touch) various puzzle boxes for a day with that when they were given the puzzle box without any rehearsal time. Chimpanzees, orang-utans and young children (aged five-seven years) all did much better after an opportunity to think about a box than they did when given the box cold. But, significantly, even such young children were a great deal faster on the task than either of the great apes.


Full-blown human-style social cognition

These results suggest that such basic abilities may be quite widespread, at least among the apes. Nonetheless, it is clear that however good the apes are on these kinds of tasks, human children are simply orders-of-magnitude better, even at very young ages. Two things are likely to be important here. One is that, for full-blown human-style social cognition, all four abilities have to be brought into play together: having only some of them is useful, but it does not allow you to engage in the kinds of complex thinking that goes into fourth- and fifth-order intentionality.
Second, the extent to which these abilities can be deployed probably depends directly on the size of your computer (i.e., brain): how many video tapes you can run side by side, and how far into the future you can run them, may be a simple function of how much neural circuitry you can afford to devote to the task (and still have enongh left over to keep the rest of the body going). Planning depth may prove to be especially important in this context: it is what allows us to play chess, as wel1 as to scheme our way into other people's good books. My colleague Louise Barrett, who studies baboons in the wild, always insists that the trouble with baboons is that they live wholly in the here-and-now and do not seem to be able to figure out that sometimes holding back on the emotions of the moment might allow them to do better in the long run.

The ability to step back from the immediacy of the world may be crucial in allowing us to assess the consequences of alternative courses of action. This, after all, is in effect what theory of mind is all about - the ability to step back from one's personal experience and imagine that the world could be other than it is, to imagine that someone else could have a false belief about the world.
Differences between species in these very basic abilities may explain why baboons and great apes are very competent socially but autistic children are not, even though none of them can pass a false-belief task.

Brain Story

Phineas Gage has the rare distinction of having achieved immortality. Sadly, it is probably not the kind of immortality he might have had in mind had he given a moment's thought to it. So far from being physically still with us or remembered for the fine symphonies he composed or the exquisite paintings he produced, he lives on as one of the most celebrated cases in neuropsychology, familiar to generations of psychology stu­dents who continue to learn the bare bones of his life story a century and a half after he himself died.

Phineas had been the foreman in charge of a road gang laying tracks for the new railway near Cavendish, Vermont, in north-eastern USA. He ran a good crew, among whom he maintained discipline and one of the best work rates on the line by dint of a forceful personality and an ability to cajole and persuade - no mean feat, given that road gangs were mostly made up from a hardbitten and fractious bunch of social misfits. Then, one fateful day in September 1848 while he was preparing a charge of explosives to blast through a rocky cut­ting, the gunpowder accidentally ignited as he was tamping it down in the hole in the rock with a three-foot-long metal rod. The force of the explosion drove the tamping iron straight up through the front of his skull, destroying a large chunk of frontal cortex. Since his skull was preserved afier his death, it has been possible to use modern computer methods to model the passage of the tamping iron throngh his brain, allowing us to determine exactly which bits of his frontal cortex were destroyed and which spared.
Miraculously, Gage survived. But, as he recovered, his per­sonality changed completely. The once masterful manager of a crew of tough, hard-drinking labourers was no longer able to maintain effective social relationships and became chronically incapable of getting a job done. He was unpredictable and inconsiderate, constantly blaspheming and behaving in an impatient, obstinate and capricious manner. According to some (albeit possibly apocryphal) acconnts, he died penniless in an institution, having drunk himself to death twelve years after his accident.

Gage's importance in the history of neuropsychology lies in what his unfortunate accident has to tell us about the functions of the frontal cortex, that bit of the outer layer of the brain that lies above the eyes and, ronghly speaking, forward of the ears. We have long appreciated that herein lies the seat of conscious mental activity, the part of the brain that is intimately bonnd up in all those smart activities that we especially associate with humans. Yet, Phineas Gage's experience tells us that we can sur­vive pretty well without substantial chunks of this bit of the brain. We just do not need it for day-to-day survival. Gage, after all, lived for a dozen years after his accident and, while the later part of his life may have been less fnlfilling than his early history might have promised, it was by no means unhappy. He, at least, seems to have been pretty contented with his life in his later years, even if others were not quite so enthusiastic about the way he treated them. What it does suggest, however, is that there is something special about this chunk of the brain that plays a crucial role in smoothing our way through the bumpy vagaries of our social world.

Phineas Gage's sad story reminds us that much of what we do in the social domain is a fine-tuned balancing act teetering precariously on the brink of social disaster. That most of us humans manage to keep to the side of social cohesion is largely thanks to the frontal cortex of our brains. Whatever the psychological processes involved may be, they must in the end be consequences of brain activity. Consciousness, as we experience it, is nothing more than the emergent property of electrical activity in the brain as interconnected neurones exchange electro-chemical messages. We can reflect on these events (the phenomenon we call 'self-consciousness') because we have theory of mind and can stand back from our immediate experiences and ask how it feels to think something. In other words, we can ask, how do I know that I know something is the case?
But how is it that only we humans can do this? A hint as to where the answer to this question might lie comes from the observation that social group size in primates correlates with the relative size of the neocortex of the species concerned. The neocortex is the relatively thin sheet (about 6 millimetres deep) that is wrapped around the inner core of the ancestral reptilian brain that all vertebrates share. In mammals, this sheet typically accounts for between 10 per cent and 40 per cent of total brain volume, but in primates it starts at 50 per cent (in the prosimians) and rises to as much as 80 per cent of , total brain volume in humans. Put somewhat simplistically, the neocortex is the thinking part of the brain and big neocortices are the primate speciality.

It is important to notice that, during the course of primate evolution, the brain has expanded forwards from back to front, so that the bit that has increased out of all proportion in modern humans is the frontal lobe. The bits at the back and sides of the brain are mainly devoted to vision and other aspects of sensory perception, sensory integration and memory. It is the increased size of the frontal lobes that is largely responsible for the much greater intelligence of species like apes and humans. Of course, this is not the whole story, since in reality the brain is a highly integrated organ, with complex interconnections between different parts of the neocortex as well as between the neocortex and some of the more primitive parts of the brain (notably some parts of the limbic system, which deals with emotions and responses to emotional cues). However, this simplified picture provides us with a good enough basis for understanding some of the key cognitive differences between humans and other primates.

There is a correlation between social group size and the volume of the neocortex in primates which soggests that it has been the need to manage the complex social world in which primates live that has driven the evolution of ever-larger brains. The important point for the present story is that we humans fit neatly onto the same scale as the other primates. Group size in humans is about 150: this is the number of people that you know personally and have some kind of meaningfol relationship with - as opposed to the people you know by sight or those with whom you have a strictly business relationship. Chimpanzees live in communities that have an average size of about 50-55, and their neocortex is proportionately smaller.

However, it turns out that, as brain size has increased during the course of primate evolution, the various parts of the neo­cortex have not expanded in the same proportions. The sensory processing areas of the neocortex seem to increase in size less fast than the non-sensory components in the frontal lobe. This is mainly because there is no advantage in having a computer to analyse sensory input (the information tbat comes in throngh your eyes, ears, nose and so on) that is bigger than that minimally necessary to make sense of the signals from the relevant organs. Given that the neocortex as a whole is increasing at a much faster rate, more spare capacity becomes available for the smart stuff that goes on up at the front end - in other words, social skills such as theory of mind - as brain size increases across the range of sizes found in the monkeys, apes and humans. The amount of spare capacity in the frontal lobe over that available to monkeys first begins to show a significant increase at around the brain size of great apes (which may explain why they, but not monkeys, can just about manage theory-of-mind tasks) but it is more than four times greater in modern humans than in great apes and the rate of increase seems to be exponential.

There is now quite a lot of dinical evidence to support the suggestion that the frontal lobe of the brain might play the crucial role in mind-reading. Patients who have suffered lesions in the frontal cortex as a result of accidents or strokes, for example, invariably lose their social skills. In some cases, they merely lack the ability to execute the usual social graces and behave like people with Asperger's syndrome, trampling unwittingly on others' social sensibilities without embarrassment; in other cases, like the unfortunate Phineas Gage, their entire personali­ty may change and they become more aggressive and less considerate of others' interests.

Recently, new technology has allowed us to peep inside the brain while it is actually working. The technology depends on the reasonable assumption that when bits of the brain are actively working on a problem, they consume more oxygen than the resting brain and so blood-flow to those particular points increases. Blood-flow throngh small segments of brain can be measured indirectly using the changes such activity cre­ates in the electromagnetic fields that surround the brain, or in the frequency with which electrons are emitted, both of which can be picked up by powerful recording devices. Studies of the active brain suggest tbat various areas up in the frontal cortex are particularly active when we are engaged in thinking about social cognition tasks like the Sally-Ann task, but not when we are thinking of simpler tasks like recognising shapes or reading words.

Taken together, these results suggest that, as the ape and human brain has evolved in size, the extra capacity has largely been added on at the front where it can be put to use developing more powerfol social cognitive abilities. Eventually, at some point in hominid evolution, sufficient extra computing power was available to make that crucial transition into the kind of cognitive reflexivity that allowed us to engage in second- and third-order intentional analyses of the world we lived in.

Inevitably. the obvious question at this point is: When did our ancestors pass through the critical Rubicon at which theory of mind and higher orders of intentionality became possible? The short answer, of course, is that it is rather hard to say because neither brain nor behaviour, let alone mental states, fossilise terribly well. However, we can gain some idea by relating the kinds of findings I've discussed above to the changes in brain size in the hominid lineage. We can do this because scaling relationships within the brain mean that overall brain volume gives us a reasonable idea of the relative sizes of the constituent bits.

If we map the intentionality levels of monkeys (at first­order), apes (at second-order, just) and modern humans (at fifth-order) onto the relative size of their frontal lobes, we get a surprisingly good straight line relationship. Using this relationship, we can discover what neocortex size would equate with third-order intentionality, and then find out when the equivalent size of brain appeared in the human fossil record. Since one of the things that can be determined reasonably well from fossil specimens is brain volume (the skull, being especially hard, tends to be preserved better than most bones), it should be possible to map onto the history of hominid evolution the pattern of change in these crucial mentalising abilities.

Mapping this relationship onto the graph of changes in brain size in the hominid lineage, making the necessary adjustment for a very straightforward relationship between total brain volume and the volume of the frontal lobe, gives us the results shown in Figure 6 (p. 19l). These suggest that third-order intentionality would have appeared for the first time with Homo erectus, around two million years ago. Fourth­order intentionality, however, would not have made its appearance until sometime around 500'000 years ago when archaic Homo supiens (our own species) came on the scene. Because brain size continues to increase dramatically in the human lineage, fifth-order would have followed fairly quickly on its heels. It is worth noting that both the Neanderthals and the Cro­Magnons, like contemporary humans, had brain sizes large enongh to accommodate fifth-order intentionality. It seems that the Neanderthals might not have been the intellectual slouches of common myth.

It seems that although the critical first step into higher levels of mentalising was made quite early on, the critical ones that radically distinguish us from our ape cousins - the higher orders of intentionality - probably came in very late: at the earliest, with the appearance of Homo supiens. Whether or not the Neanderthals shared these capacities with us really depends on whether their brains were organised in exactly the same way as ours. The famous Neanderthal 'bun' (the enlarged back part of their skull) suggests that they might have had a much bigger visual area than we do (something that is confirmed by the relatively much larger size of their eyes); if so, then it is possible that they would have had less neocortex volume in their frontal lobe - which, if true, could have limited their social cognitive abilities to fourth-order intentionality (the level they would have inherited from the archaic humans that we and the Neanderthals share as a common ancestor). If this is so, then fifth­order intentionality, and all the complex social phenomena that depend on this, would not have appeared until anatomically modern humans (our own sub-species) came on the scene a mere 200,000 years ago.

The conclusion we have been drawn to in this chapter is that, although apes and humans share a number of important advanced cognitive abilities, they differ in one key respect: the extent to which humans can detach themselves from the world as they experience it. This allows humans to reflect on the world as they find it, to wonder whether it could have been otherwise. In contrast, apes (and certainly all other animals) have a much more direct, straightforward experience of the world. Their noses are thrust firmly up against reality. In the following chap­ters, we shall see that this has very important implications for some of the more explicitly human aspects of our behaviour.