3Animals may be trained to paint in specific ways that resemble human art. I have heard that some elephants in Thailand, for instance, paint what looks like flowers for tourists. But these actions are shaped by the trainer, so this does not suggest that the animals intend the picture to represent something else.
4As it turned out, the children’s problem in the previous studies was not with understanding the relationship between the picture and the room, but with persevering over repeated trials. They typically did well on their first trial, but on subsequent trials they had a tendency to return to a previously successful location. When I presented children not with the typical four trials in one room, but with one trial each in four different rooms, even twenty-four-month-olds could solve the task.
5Using a task similar to the game Pictionary, researchers are beginning to examine how graphic pictures can become conventionalized into abstract symbols.
6Though they usually do understand high German too. Being multilingual is another solution that allows people to interact across cultural divides. Traders who interact with many different peoples, such as the Dutch traditionally did, benefit from being able to speak a variety of languages. It is also not uncommon for people whose language has relatively few speakers, such as tribal groups of Australians and New Guineans, to speak numerous other languages spoken by surrounding tribes. Incidentally, people who speak more than one language, it has recently been documented, are less likely to suffer from Alzheimer’s dementia in old age—so if you are worried about that, go on, learn another language.
7Some languages can make compounds virtually without limits. Consider this extraordinarily long Maori place name: “Taumatawhakatangihangakoauauotamateaturipukakapikimaungahoronukupokaiwhenuakitanatahu”—which translates to “The hilltop where Tamatea with big knees, conqueror of mountains, eater of land, traveller over land and sea, played his flute to his beloved.”
8To get this going, one first needs to assume F1 = F2 = 1.
9Languages, Chomsky argues, differ in terms of their surface expressions rather than their deep structure. For instance, the simple sentence “The computer copied over my files” can be considered as a noun phrase (“The computer”) and a verb phrase (“copied over my files”). The latter comprises a verb (“copied”) and a prepositional phrase (“over”) and another noun phrase (“my files”). While English tends to structure these elements such that subjects precede verbs, which in turn precede objects, Japanese has a subject-object-verb surface structure.
10Skinner had been enticed to provide a behaviorist account of human language following an encounter with the philosopher Alfred North Whitehead. In 1934 Skinner found himself seated next to Whitehead at a dinner and boasted about the power of his learning theory. Whitehead challenged: “Let me see you account for my behavior as I sit here saying, ‘No black scorpion is falling upon this table.’” It took Skinner twenty-three years before he published a book attempting to explain language in terms of associative learning. In an appendix he responded to the challenge, surprisingly citing Freud, by suggesting that Whitehead said what he did because he was afraid behaviorism, symbolized by the black scorpion, would take over. Yet something else was about to take over. In the same year, Noam Chomsky published his entirely different, and ultimately far more influential, theory of the nature of language.
11I did this even though I actually like the word “autapomorphy.” The term refers to a derived trait that is unique to one member of a clade, and not found in any others, not even those most closely related or a common ancestor.
12Speech involves complicated deliberate motor acts. Great apes have much more voluntary control of their hands, which they use, for example, to extract foods, than over their articulations. If our common ancestor had such control, it would appear easier for natural selection to tinker with such control than for it to invent vocal control de novo. Gestural communication also has the advantage of being more iconic, where words are almost entirely arbitrary. Indeed, some things are much more intuitively communicated by gestures: try to explain to someone what a spiral is. So perhaps language evolved first in the gestural domain, as Michael Corballis argues, and only later was complemented and largely displaced by articulations.
13The revealing 2011 documentary Project Nim tells the sad story of Nim and his fate following Terrace’s research effort.
14Chimpanzees tend to have severe problems with some basic elements of symbol-referent links. For example, after intensive training they may learn that the color red is to be matched to a symbol for red. However, as soon as the relation is reversed, they struggle. That is, they find it difficult to select between red and blue colors when presented with the symbol for red. The bidirectional correspondence seems to look obvious only to us.
15Tamarins have been found to be able to learn a nonrecursive acoustic sequence but not a recursive one. Subsequently research has used this test to suggest that starlings can learn recursive rules. The test itself, however, has been criticized by Michael Corballis because it can actually be solved in nonrecursive ways. In this case the killjoys appear to have the upper hand over romantic explanations.
FIVE
Time Travelers
Forethought is the most important of all the causes that make human life different from that of animals.
—BERTRAND RUSSELL
IMAGINE SOMEONE HAS FINALLY INVENTED a time machine. Where in time would you go? Would you want to travel to the distant future, or witness a special event in the past? Perhaps you are the type of person who is content to stay where you are. Regardless of your personal preferences, the idea of a time machine has long had a curious appeal—its endless possibilities tickle our fancy. Alas, modern physics suggests time travel will never become a reality. We must make do with time traveling in our minds. We can recall past episodes and imagine future events, including entirely fictional scenarios (such as the invention of an actual time machine). Much of my research has centered on this fundamental human capacity. It therefore seems fitting to begin with some reminiscing of my own.
As a child, I struggled to come to terms with the most unwelcome of all realizations: the fact that, one day, I would die. I lay in bed, staring at the ceiling, and attempted to imagine “not being.” I thought I might be able to do this, as the state seemed little different from dreamless sleep. However, I couldn’t get my head around the idea of never existing again—never to wake up; to be gone forever. Even now the thought makes me feel queasy, and I appreciate why people seek comfort in the idea of an afterlife. I think I realized even then that this dilemma was entirely due to our ability to project ourselves forward in time. I am not sure if I considered whether this was a uniquely human problem, but this was to become an important question in my studies.
In my first year of undergraduate psychology at a German university, we read a book by Norbert Bischof, a student of Konrad Lorenz, on the nature of incest avoidance. Buried a few pages from the end was his claim that only humans have “time representation.” This notion resonated with me, and I began to investigate the questions it raised. What is the nature of the human ability to think beyond the here and now? How does it develop in children? What are the temporal capacities of animals? Don’t they ponder the good old days? Don’t they imagine what the future might hold? In a radical shift in lifestyle, I ended up moving to New Zealand to research these questions in my master’s thesis.
I eventually found an idyllic houseboat in the mangrove swamps of an offshore island and a wonderful mentor, Michael Corballis, at the University of Auckland. He gave me an early laptop that I powered with a couple of truck batteries that in turn were charged through a solar panel. I had time to focus—there were few distractions other than the mosquitos. One day, after I had almost completed my thesis, WordPerfect was in the process of saving my files when the power supply seized up. I lost much of my work. Michael consoled me with the wise counsel that rewriting would probably improve the whole flow immensely. I dutifully rewr
ote what I had done—only to lose it again, this time on a computer at the university.
I learned a thing or two from this experience about backing up my work, something we do thanks to our capacity to think about the past and the future (the backup software on my current computer is in fact called Time Machine). Finally we turned my thesis into a monograph in which we proposed that mental time travel into the past and mental time travel into the future are two aspects of the same faculty. We could not find compelling evidence for anything like this faculty in other animals and argued that its emergence must have been a prime mover in human evolution. Mental time travel, as it turns out, explains many of the most peculiar human traits, from celibacy to suicide and from diverse expertise to greed.
“It’s a poor sort of memory that only works backwards,” the Queen remarked.
—LEWIS CARROLL
WE HAVE SEVERAL DISTINCT MEMORY systems. We can experience problems with one system without impairment of the others. The English musician Clive Wearing, for example, developed amnesia after an infection of the herpes simplex virus destroyed a part of his brain known as the hippocampus, which, as you may recall, plays a role in mental maps. He has retained many skills (such as how to play the piano) and knows facts about the world (such as what a piano is), yet he cannot remember a single event that has ever happened to him (such as his having given a concert). He knows he is married but cannot remember getting married. Because of such dissociations, researchers commonly distinguish the following memory systems: memory for how to do things (procedural memory), memory for facts (semantic memory), and memory for events (episodic memory).
Episodic memory is probably closest to what we typically mean when we use the word “remember” rather than “know.” The psychologist Endel Tulving first proposed this concept to refer to memory of one’s own past experiences. When you retrieve episodic memories, you travel mentally back in time and reexperience perceptions, actions, emotions, or thoughts of a past episode of your life. You may revel in reliving your past successes or mourn your failures. Clive Wearing, on the other hand, continually reports that he has just woken up from unconsciousness and is seeing and experiencing things for the first time—only to forget about them and then experience yet another resurrection.1 Without episodic memory, he is stuck in a time bubble.
Although you may treasure many of your episodic memories, it is not entirely clear what the ultimate function of this capacity is. At first glance, you may think that its job must be to provide a faithful record of our past. Indeed, memory research in the past 150 years has focused almost exclusively on factors affecting memory accuracy. What this research has found, however, is that our episodic memory system is not particularly comprehensive or reliable. Perhaps this comes as a bit of a relief—it certainly makes me feel better about my own apparent memory shortcomings. We tend to struggle with simple requests. Try to recall what happened on your birthday three years ago. Now try to recall what happened at lunch two days before or after that birthday. Our memory for events is nothing like a recording; we cannot simply rewind and press play. In fact, it seems we forget most events. Think of all the hours at school you sat on a chair in class. You may remember your classmates, the teachers, and some special interactions, but for the most part your memories blend together. I suspect you would struggle to recount the exact events of any single day or even the details of a specific hour.
We do, of course, remember some events in great detail. Perhaps you recall the moment when you found out about the 9/11 attacks on New York. Yet even here there is reason for pause. Research shows that we can be wrong about events we think we recall clearly. Naturally this poses a serious problem for all those situations in which someone’s memory is all we have to go by—as often happens in the courtroom. Jennifer Thompson thought that Ronald Cotton had raped her at knifepoint in 1984. Her identification led to his conviction in 1986, and he served eleven years in jail before his innocence was revealed. After a DNA test, another man, Bobby Poole, whom the victim did not identify when he originally appeared in court, admitted to the rape. Thompson’s memory had been wrong. There has been extensive work on the reliability of eyewitness testimony, and the results are not exactly reassuring. Untrue information about crucial details, be it the presence of a stop sign or a mustache, is sometimes falsely recalled. People’s accounts can be influenced by information introduced after the event. Suggestions by other witnesses, police, or lawyers, for instance, may be incorporated into eyewitness reports. What is more, confidence in the accuracy of one’s memory is a surprisingly poor predictor of which detail in the eyewitness report is false and which is correct. Even when you are positive you remember something correctly, you might be wrong.
Remembering episodes is a reconstructive process that draws on some stored gist that is then actively expanded as we rebuild the scenario of the past.2 We may embellish to create a better story or adjust our reconstructions to make them consistent with our prevailing attitudes. Repeated acts of reminiscing and retelling can lead to increasing distortions. So we bolster our faulty memories with external storage systems, such as diaries, photos, drawings, books, and data sticks—in acknowledgment that our memory of past events is not a terribly reliable record keeper.
How did such a flawed system evolve? Unlike a memory researcher using a video recording, natural selection cannot go back in time and check your memory matches accurately with the original event. If a false memory or a memory bias improves fitness (the capacity to survive and to propagate one’s genes), the memory system that produced it has a selective advantage no matter how inaccurate. For example, you typically recall your own good behavior better than your bad behavior, but show no such bias when recalling the behavior of others. This partiality leads to a flawed reflection of the past, but if it improves, say, the chances of impressing potential mates, then those with the distorted view may on average have more children in the next generation than those with an accurate view. And so the bias spreads. Evolution works only on how memory influences fitness, not for how accurately memory reflects the past per se.
In this sense, accurate prediction is more important than accurate recall. Indeed, all memory systems are inherently future-directed, rather than oriented toward the past, as one might assume. Take the classical conditioning first described by the physiologist Ivan Pavlov. He found that a dog learns to associate the sound of a bell with the arrival of food. The next time the dog hears the bell, it starts salivating before the food arrives. Although people typically think of this as memory, the dog is not really salivating at the memory of food, but in anticipation of, and preparation for, food. The sound of the bell makes the dog predict that food is about to arrive. Similarly, the memory system for facts must have evolved for its future-directed benefits. For instance, knowing where one’s hideout is can quickly become fitness relevant. A warthog that knows this location has a better chance of surviving a lion attack than one that does not. Memory matters because of what it can do for you now and in the future. How might this work for episodic memory? The main benefit of memory for past events may be that it allows us to imagine future events.
Perhaps the Red Queen was right. It would be a poor sort of memory that only works backwards. Are our capacities for mental time travel into the past and into the future two sides of the same coin? Amnesic patients such as Clive Wearing who have lost their episodic memory have similar problems imagining future events; they draw a blank when asked about what they will do tomorrow. We have found that young children’s capacity to answer such questions is linked with their ability to report what they did the day before. Introspectively, there are some similarities. For example, we tend to conceive of fewer details of events the further they are removed from the present—whether into the past or into the future. In old age we tend to report fewer details of both past and future events. Suicidally depressed and schizophrenic patients have difficulties recalling specifics about past events and show similar problems imagining the future
. Brain imaging studies have found that when participants are asked to recall past events and imagine future situations, the same areas of the brain (including the hippocampus and regions in the prefrontal, parietal, and temporal cortex) are involved. Although there are some important differences between mental time travel into the past and mental time travel into the future—after all, one has happened and the other hasn’t—many commonalities have been documented in recent years. Substantial evidence supports the proposal that episodic memory and episodic foresight are fundamentally linked in mind and brain.3
OUR EPISODIC MEMORY IS NOT limited to working backwards. One way for it to go forward is simply for it to project into the future. After all, the best predictor of future behavior, generally, is past behavior. Your dog’s reaction the last time you attempted to steal his bone provides a reasonable guide about what to expect should you try to do so again next week.
However, you can do much more than predict reoccurrences of events. You can imagine situations you have never experienced before. For example, you might mentally simulate how you could distract the dog before going for his bone. You can imagine a virtually infinite number of future scenarios. You may consider a few options before pursuing the most promising path—no one likes being bitten. You do not need to suffer all the real-life consequences of trying out each possibility. We can test most things in our mind and assess how likely or pleasant they might be. For instance, I suspect you can imagine whether you would enjoy hot mustard mixed with your vanilla ice cream without having tried the combination before. To imagine new events you need an open-ended system capable of combining old information into new scenarios. If mental time travel evolved for this purpose, then the price of this flexibility is that we may at times reconstruct past events creatively rather than faithfully—which explains some of the typical errors of episodic memory.
The Gap Page 11