Reliving the Past
I have been through some terrible things in my life, some of which actually happened.
—Mark Twain
In Hidden Lives, a memoir of her family, the British novelist Margaret Forster describes herself in the first five years of her life in the third person. From the age of five, however, she switches to the first person. She explains the switch as follows:
It was at this time, in 1943, when I was five, that my own real memory begins, real in the sense that I can not only recall actual events but can propel myself back into them, be there again in my Aunt Jean’s roomand-kitchen, standing by the window at the back of the buildings, staring out at the outside staircase and the tops of the wash-houses, while behind me Jean asks what is the matter.… That is what I call real memory, not at all the same as “remembering” being taken to Ashley Street School to demonstrate my boasted ability to read. Though, because my mother later told me about it so often, I often claimed to remember it and could easily convince myself that I did.1
Up to that point in the book, she describes the life of a rather difficult little girl called Margaret as just another member of the family, but from then on everything is viewed through the lens of personal experience. The transition, as biography switches to autobiography, may even herald the arrival of the concept of self. It is the beginning of memory as a recursive phenomenon, with previous experience inserted into present consciousness.
My own conscious memory also begins at around five, when I trudged about a mile up a country road to my first school, accompanied by older boys from down the road whom my mother entrusted to look after me, but in fact they tormented me for wearing glasses. There are a few earlier episodes that I imagine I remember, but they may well be fabrications from things that were told to me. I often ask my undergraduate classes to declare their earliest memories. There are always a few souls who claim memories going back to infancy, even to birth, but most of the memories come up from age four or five. This pattern is corroborated in adult surveys of remembered events before the age of eight. Virtually no memories were recorded from the first three years of life, rising steadily to all adults remembering events from the first eight years.2
In the nomenclature of modern cognitive psychology, Margaret Forster’s distinction between “real memory” and “remembering” is out of line. The term “remembering” is usually taken to refer to memory for actual events, located in time and space. This is also known as episodic memory. It involves consciously projecting oneself back in time, just as Margaret Forster did. The term “knowing” refers to the other kind of memory, also known as semantic memory, which is the storehouse of knowledge that we possess, but that does not involve any sense of conscious recollection.3 I know that Paris is in France, but have no conscious memory of when I learned this. On the other hand I can vividly remember being in class as an eleven-year-old when I discovered that (2n + 1), where n is any integer, was a general expression for an odd number. My memory is tinged with triumph, because we were asked to generate this expression ourselves, and I was the only one who did so.4
Episodic memory may combine with some aspects of semantic memory to make up what is known as autobiographical memory.5 As the passage from Margaret Forster illustrates, it is often difficult to distinguish which parts of autobiographical memory are based on remembered episodes, and which on knowledge. In the study referred to above, the adults were not only asked to recall memories, but were also asked to describe things that they knew had happened to them, but could not actually remember. Knowing but not remembering early events depends largely on family folklore—stories, often repeatedly told, of childhood events, and often embellished through imagination to the point that one has the sense of re-experiencing them. Nevertheless “known” events declined steadily from birth to age eight, while “remembered” events increased from age two to age eight.
Endel Tulving has described remembering as autonoetic, or selfknowing, in that one has projected one’s self into the past to re-experience some earlier episode.6 Simply knowing something, like the boiling point of water, is noetic, and implies no shift of consciousness. Autonoetic awareness, then, is recursive, in that one can insert previous personal experience into present awareness. This is analogous to the embedding of phrases within phrases, or sentences within sentences. Deeper levels of embedding are also possible, as when I remember yesterday that I had remembered an event that occurred at some earlier time. Chunks of episodic awareness can thus be inserted into each other in recursive fashion. Having coffee at a conference recently, I was reminded of an earlier conference where I managed to spill coffee on a distinguished philosopher. This is memory of a memory of an event. I shall suggest later that this kind of embedding may have set the stage for the recursive structure of language itself.
Amnesias
Brain imaging suggests that there may be considerable overlap in the areas activated when people retrieve episodic and semantic memories, but each kind of retrieval also activates unique regions.7 In cases of amnesia, which is the loss of memory caused by brain injury, it is episodic memory that usually carries the brunt of the loss. Semantic memory is relatively unimpaired. One well-known case is the musician Clive Wearing, whose plight has been aired in several television documentaries and in a book written by his wife Deborah.8 Wearing was an acknowledged expert on early music and had built up a musical career with the BBC, when at the age of 46 he became ill with the herpes simplex virus. This effectively destroyed his hippocampus, a structure in the inferior temporal lobe that is critical to memory. Wearing’s condition is well described by the title of a 2005 ITV documentary, The Man with the 7-Second Memory. That is, he has sufficient short-term memory to be able to respond to questions, and even converse, although he quickly forgets topics that he spoke about moments earlier.9 He does remember some aspects of his life before the illness. For example he recognizes his children, but does not recall their names. Much of his semantic memory, though, remains intact. His vocabulary is largely unaffected, and he still knows how to play the piano or conduct a choir.
The most extensively studied case of amnesia for events is that of H. M., who underwent surgery for intractable epilepsy at the age of 27.10 The surgery destroyed most of his hippocampus, along with surrounding regions, and left him in much the same state as Clive Wearing. He remembered some of his pre-illness life, but suffers memory loss reaching back in time from the operation, with better recall of earlier than of later events. This has led to the view that the hippocampus is responsible for laying down memories, and somehow holds them temporarily while consolidation takes place elsewhere in the brain. Loss of hippocampal function therefore not only prevents new memories from being formed, but also destroys past memories that are temporarily held there, before being distributed to other brain regions. The holding operation may last several years, with diminishing effectiveness, as the hippocampus relinquishes its grip.11
There has been some dispute about whether semantic memory and episodic memory are truly distinct. One might suppose, for example, that episodic memory is fragile simply because it is based on single events, whereas semantic memory is based on information that is repeated, and may be present in multiple episodes. Endel Tulving has undertaken careful studies of a particular patient, known as K. C., and shown that his difficulty with episodic memory relative to semantic memory cannot be attributed to the frequency with which the recorded information is presented. K. C. can recall factual information that he is unlikely to have rehearsed repeatedly, but cannot bring to mind events that lasted several days, such as being evacuated from home, along with tens of thousands of others, when a nearby derailment released toxic chemicals. This suggests that episodic memory can fail even when there is likely to have been repeated rehearsal. Tulving refers also to controlled experiments leading to the same conclusion.12
Tulving nevertheless argues that the storage of episodic memories depends on semantic memories that are already in place—one can scarcely record a visit to a restaurant without already knowing what a restaurant is, and what happens there—but are then related to the self in subjectively sensed time. This allows the actual experience of the event to be stored separately from the semantic system.13 In this view, episodic memories could not be stored in the absence of semantic memory, which is perhaps why our childhood episodic memories do not begin until the semantic system is well established, by around age four or five. Nevertheless there is some evidence that episodic and semantic memories are more broadly dissociated. People with semantic dementia, a degenerative neurological disorder that afflicts some people in late adulthood, show severe decline in semantic memory, but their episodic memories remain remarkably and surprisingly intact.14
Unconscious Memory
There is another category of memory that appears to remain intact in cases of profound amnesia. This illustrated by the finding that H. M. could learn new skills, such as mirror drawing, where the task is to draw or trace a shape while viewing the hand and shape only in a mirror. This is a difficult task, because the tracing movements you see in the mirror are front-back reversed relative to the actual movements. H. M. improved with practice, just as people without memory deficits do, yet never remembered having done it before. He even showed consistent improvement on tasks that require mental operations, rather than learned movements. One example is the Tower of Hanoi task, in which rings are stacked in order of size on one of three pegs, with the largest ring at the bottom. The task is to move the entire stack to another peg one at a time, using all three pegs, but without ever placing a larger ring on a smaller one. Again, H. M. showed systematic improvement, although he never consciously recalled seeing the problem on previous attempts.
The unconscious memory that underlies these skills is called implicit memory, which enables us to learn without any awareness that we are doing so. It is presumably more primitive in an evolutionary sense than is explicit memory, which is made up of semantic and episodic memory. Explicit memory is sometimes called declarative memory because it is the kind of memory we can talk about or declare. Implicit memory does not depend on the hippocampus, so amnesia resulting from hippocampal damage does not entirely prevent adaptation to new environments or condition, but such adaptation does not enter consciousness. I am a poor typist, and could not consciously tell where the keys are placed on the keyboard, except perhaps for the rather touching [sic] fact that U and I are side by side, but my fingers find their way very quickly to the appropriate keys. This is implicit memory. I should add that I only use two fingers. The others know nothing.
Implicit memory is sometimes measured through a technique called priming. For example, if people are asked to recall a series of words they have been shown in the laboratory, their recall of individual words may be enhanced if they are preceded by some related word, or prime. For example, the word animal may prime recall of the word aardvark. Priming is remarkably resilient. In one study, for example, fragments of pictures were used to prime recognition of whole pictures of objects. When the same fragments were shown 17 years later to people who had taken part in the original experiment, they were able to write the name of the object associated with each fragment much more accurately than a control group who had not previously seen the fragments. Some of the people involved had no conscious recollection at all of the laboratory experiment 17 years earlier, and these people actually scored slightly higher on the priming measure than those who did remember the previous occasion.15
The Fragility of Memory
Implicit memory, then, is unconscious and seemingly invulnerable, whereas our episodic memories are fragile, even in the absence of brain injury. Indeed, we are not nearly as good at remembering events as we think we are. At one time it was thought that we store everything that we were ever conscious of, and that the reason that we seem to forget so much has to do with retrieval failure. That is, there are lots of things stored in memory that we can’t find. It is true that retrieval is fickle—we all have the experience of being embarrassingly unable to remember someone’s name when we need to introduce that person to a group of friends, only to remember it later when it no longer matters. More recent research suggests that memory failure is not just a matter of retrieval failure, which of course does occur, but also reflects failures of storage.16 For this reason, a school reunion is a sobering experience.17 At one such event, my school friends frequently recounted some apparently salient event that seemed to have completely escaped my own recollection, even though I was supposedly present when it happened. Symmetry was restored when I told of events that others seem to have forgotten. And of course some stories were embellished beyond recognition anyway. Some of my old mates, I do remember, were compulsive liars.
We now know that our memories are sieve-like; in fact, such is the length and complexity of our conscious lives that we retain probably only a tiny fraction. The émigré Czech writer Milan Kundera, in his novel Ignorance, put it like this:
The fundamental given is the ratio between the amount of time in the lived life and the amount of time from that life that is stored in memory. No one has ever tried to calculate this ratio, and in fact there exists no technique for doing so; yet without much risk of error I could assume that the memory retains no more than a millionth, a hundredmillionth, in short an utterly infinitesimal bit of the lived life. That fact too is part of the essence of man. If someone could retain in his memory everything that he had experienced, if he could at any time call up any fragment of his past, he would be nothing like human beings: neither his loves nor his friendships nor his capacity to forgive or avenge would resemble ours.18
Well, perhaps that’s an exaggeration—a hundred-millionth would amount to about 15 minutes’ worth. Still, most of our conscious lives slip by without lasting trace.
The paucity of our memory for events may nevertheless be adaptive, since to remember everything that happens to us could be debilitating, clogging our minds with unwanted junk—like old clothes. Some individuals, in fact seem to suffer precisely because their memories are too good.
Can You Have Too Much Memory?
New Zealand children of my generation were brought up with a substance known as Marmite, a spread that is roughly the cultural equivalent of peanut butter in the United States. The jar contained the message “Too much spoils the flavor.” To those not raised with this substance, the flavor is repugnant anyway, no matter how little you smear onto your slice of bread. The idea that too much might spoil the flavor must have been a poor marketing strategy, and the message no longer appears. Memory, perhaps, is a bit like marmite. Too much may spoil the mind.
There are indeed reasons to suppose that a good memory may be an impediment. A condition known as savant syndrome can result in prodigious powers of memory, but deficiencies in other aspects of intelligence. The most extraordinary case so far described is that of Kim Peek, who was the inspiration for the movie Rain Man.19 Peek, known to his friends as “Kim-puter,” began memorizing books at the age of 18 months and in his midfifties has now memorized 9,000 books. He has a vast storehouse of knowledge in history, sports, movies, space programs, literature, and Shakespeare, among other things. He has an extensive knowledge of classical music, and in middle life has even begun to play it. In common with other savants, he is an expert at calendar calculation, and is able to instantly tell you the day of the week for any given date. This is known to depend on massive memory.
Yet on a standard test of intelligence, Peek scored only 87 (the population average is 100). There was great variation, though, in the subscales that go to make up the intelligence score, with some in the superior range and others in the range of the mentally retarded. His brain is abnormal, being unusually large and lacking the corpus callosum, which is the main fiber tract that connects the left and right sides of the brain. Two other tracts that normally link the sides of the brain are also missing. He has an unusual sidelong gait, cannot button his clothes, and cannot handle the chores of daily life. He also has difficulty with abstract ideas. What this profile suggests is that a large and detailed memory can work to the disadvantage of other mental skills, and a memory that is too particular can impair ability to see relations and form abstractions. Too many trees, and it’s hard to see the wood.
Another case in point was Solomon Shereshevskii, better known as “S,” whose prodigious feats of memory were described by the Russian neuropsychologist Aleksandr Romanov Luria in his 1968 book The Mind of a Mnemonist. According to Luria, S’s memory capacity was seemingly without limit, and he remembered trivial things for extremely long periods of time. For instance, he could accurately recall lists of words that Luria had presented 16 years earlier. His memory was largely visual, and when given verbal items he could transform them mentally, either by arranging them spatially, or using the “method of loci” whereby he would imagine them in familiar locations. The particularity of his memory was actually an impediment, because it prevented him from forming general concepts. He was unable to make sense of novels, since he would imagine scenes in precise detail, only to find his images contradicted at later stages. S was also a synaesthete—a condition in which inputs in one sensory modality give rise to sensations in another—so that spoken words were accompanied by visual sensations, such as “puffs” or “splashes,” and a tone at 30 cycles per second and 100 decibels gave rise to “a strip 12–15 cm in width the color of old tarnished silver.”20
Another case that has recently come to light is that of a woman known in the literature as A. J.21 She would not be classed as a savant, but nonetheless has an unusual and intrusive autobiographical memory. She was born in 1965, but claims that if you give her any date between 1974 and the present she can tell you what day of the week it was, what she was doing on that day, and whether anything of great importance (such as the Challenger explosion of January 28, 1986) occurred. She kept a diary from age 10 to age 34, but her memory for events evidently does not depend on her rereading the diary, which she hardly ever does, nor does she consult a calendar before being asked questions about her life. She spends an excessive amount of time recalling her past, and tests have shown that what she does recall is extraordinarily accurate.
Paradoxically, though, when the researchers made a videotape of her and showed it to her a month later, she remembered very little of the video session itself, suggesting that her autobiographical memory is in fact very selective. Although she seems to remember something from every date, the actual amount remembered may be only a fraction of what actually happened. She claims to have had trouble memorizing historical dates, learning foreign languages or science, and got Ds in geometry. Her full scale intelligence quotient (IQ) is 93, a little below average, although she scored 122 on a standard memory test. Despite her academic shortcomings, she completed a bachelor’s degree in social science, graduating at the age of 23.
Memory, whether exceptional or not, requires huge storage space. As mentioned earlier (in case you have forgotten), the average literate adult probably knows about 50,000 words, which implies that she knows about the same number of concepts, including objects, actions, qualities, professions, emotions—and on and on.22 Episodic memories are generally made up of different combinations of otherwise familiar concepts, as in who did what to whom, when where and why. To accommodate all of this memory, it is not surprising that the human brain is large, even by primate standards. Something like 10 billion neurons exist in the human brain, each connected to other neurons—in fact, some neurons have as many as 30,000 connections. It’s hard to get one’s head around such numbers, even though they are in fact contained within the head. Given that memories are probably stored in terms of interconnections between neurons, one may be tempted to believe that the brain has infinite capacity. But the brain has lots of other things to do besides store memories. I suspect therefore that there is something of a trade-off between different mental pursuits, and obsession with some pursuits may prevent others from developing. Sometimes, as in the case of people with savant syndrome, the reason for the imbalance may be pathological—some defect of growth or hormone balance that allowed some faculties to develop at the expense of others. In other cases, perhaps, obsessions may arise in response to cataclysmic life events. And there is no doubt normal variation. Life would be dull if this were not so.
False Memories
When I was younger I could remember anything, whether it had happened or not, but my faculties are decaying now, and soon I shall be so I cannot remember any but things that never happened.
—Mark Twain
Although our memories of the past can be vivid, to the point that we seem to re-experience them, they can fail to be true to the original experience. Our memories of earlier exploits are often embellished or distorted, and often downright false—we “remember” events that simply never happened. The pioneer of research in this field, Elizabeth Loftus, recounts a false memory of her own mother’s death:
I remember a summer many years ago. I was fourteen years old. My mother, my aunt Pearl, and I were on vacation, visiting my uncle Joe in Pennsylvania. One bright sunny morning I woke up and my mother was dead, drowned in a swimming pool.23
The scene to her is vivid. In her mind she sees and smells cool pine trees, tastes iced tea, and sees her mother in her nightgown, floating face down. She cries out in terror, starts screaming, sees the police cars, lights flashing, and the stretcher carrying her mother’s body. But the memory is false. She was in fact asleep when the body was discovered, not by her but by her Aunt Pearl. The memory is a construction, built partly of knowledge of what happened, and partly of extra details supplied by imagination.
Loftus studied false memory experimentally by giving people short narratives, constructed by family members, of events in their childhood. She then asked them for their memories of these events. One of the narratives described being lost in a mall, but was entirely false. About a quarter of her sample claimed to remember the event, and some even supplied details that had not been mentioned in the narrative. This study took place in 1991, and has been repeated many times with other false narratives that were nevertheless “remembered,” such as being taken for a ride in a hot-air balloon, being attacked by a vicious animal, or being nearly drowned and rescued by a lifeguard. False memory is also studied using more prosaic methods, such as giving people lists of words to remember that omit one very obvious word; when asked to recall the list, participants will often insist that the missing word had actually been presented. For example, the list may include words like bed, dream, rest, awake, and so on, omitting the word sleep, but around half of the participants later remember the word sleep as having been in the list.24 This experiment has become something of a cottage industry in experimental psychology, and is a regular feature of undergraduate laboratories.
Actually, false memories were widely discussed in the late nineteenth century, and were known as paramnesias.25 An especially horrific example is described by the hypnotherapist Hippolyte Bernheim, who recounts how he suggested to a patient that she had witnessed through a keyhole an old man raping a little girl, who struggled, was bleeding, and was then gagged. His suggestion concludes “When you wake up you will think no more about it. I have not told the story to you; it is not a dream; it is not a vision I have given you during your hypnotic sleep; it is truth itself.” Three days later Bernheim asked a friend and distinguished lawyer to question the patient as though he were an examining judge. The patient recounted the events in detail as suggested to her, and even when encouraged to doubt them maintained the truth of the testimony “with immovable conviction.”26 Today, fortunately, ethical regulations would forbid such a study.
In recent times, the question of false memory has been revived and has assumed special importance. In the 1980s and 1990s, many therapists adopted the view that psychological problems in adulthood could be traced to sexual abuse during childhood. Because of their traumatic nature, such memories were often repressed, and the main purpose of therapy was to recover these memories, so that patients could then face the real causes of their problems and deal with them—presumably with the therapist’s help. The most extreme expression of this view was a book by Ellen Bass and Laura Davis entitled The Courage to Heal, which was first published in 1988 but has since gone through several editions. Bass and Davis, who had no formal training in psychology or psychiatry, were nevertheless bold enough to tell their readers:
If you don’t remember your abuse, you are not alone. Many women don’t have memories, and some never get memories. This doesn’t mean they weren’t abused.
In another notorious quote from the book, Bass and Davis write, “If you think you were abused, and your life shows the symptoms, then you were.” This statement commits the logical fallacy of affirming the consequent. It is undoubtedly true that childhood abuse can cause later symptoms of psychological distress, but this does not mean that psychological stress must have resulted from childhood abuse. Murder results in death, but this does not mean that death is always due to murder. Unfortunately, widespread acceptance of Bass and Davis’s edict too often led to aggressive therapy designed to help distressed people recover the memories of the abuse that led to their distress.27 This led to the recovery of “memory” for events that did not happen, just as Loftus’s experimental subjects were never in fact lost in a mall. The whole issue of recovered memory versus false memory became intensely political in the 1990s, with many cases of men being given jail sentences for abuse despite being entirely innocent;28 one such case in my own country is described in detail in a prize-winning book.29
An Evolutionary Perspective
By now it will be apparent that memory is not quite what it seems. It is far from a faithful record of past events, and is also complex, made up of several systems. This is an appropriate point, then, to consider the evolution of memory systems, and how they might contribute to the adaptive fitness of organisms. I shall suggest that there is a hierarchy of memory systems, each contributing in a special way. I will here take “memory” to mean any adaptation involving some change in an animal’s behavior that results in increased fitness.
A good proportion of behavioral adaptations are instincts, which are “memories” only in the sense that they refer to adaptive changes that took place in earlier generations, and were transmitted genetically. Genetic endowment is not sufficient, though, and instinctive behaviors will often not develop unless the environmental conditions are adequate. Instincts can be quite complex, and include activities such as seasonal migrations of birds and animals, hoarding and caching, dam-building in beavers, and so forth. These are effective adaptations to predictable seasonal events, and are effective so long as seasonal changes are indeed regular. Such instincts can be disrupted by climate change, or by the predations of other species, especially humans, with our unparalleled capacity to wreak destruction, not only on the geography of the planet, but also on seasonal change itself.
Of course we humans also have lots of instincts, such as emotions, sexual drive, and, one may think, a predisposition to gossip. Steven Pinker has maintained that language itself is an instinct;30 given normal development all humans talk or sign, some of them too much. Again though it does depend on the environment, and a child brought up without human contact will not develop language, as illustrated by the famous case of Genie, a girl who was isolated until the age of 13 and never learned proper language.31 Of course the particular languages we learn also depend on the particular linguistic environment we are brought up in; I fear I will never learn Chinese. The development of birdsong involves similar principles. Although fundamentally instinctive, the song of the chaffinch requires that the young bird be reared where it can hear the song of other chaffinches.
Learning offers a more flexible means of adaptation, since it allows adjustment to environmental events or contingencies that occur in the individual’s own lifetime. One form of learning is classical conditioning, made famous by Ivan P. Pavlov’s demonstration that if a bell is sounded prior to the presentation of food, animals soon salivate to the bell.32 More generally, classical conditioning may serve to generate appropriate emotional or anticipatory responses to situations, animals, or objects that pose threats or promise rewards. Classical conditioning depends on the involuntary elicitation of bodily events, such as salivation, blushing, or emotional responses like fear and anger, which then become attached to previously neutral events.
Pavlov’s experiments were carried out with dogs. The most famous (not to say notorious) study of classical conditioning in humans was carried out by the founder of behaviorism, John B. Watson. The subject 33 was a nine-month-old orphan boy known as Little Albert.34 Watson first determined that Little Albert was not afraid of objects, such as a rat, rabbit, monkey, dog, or mask, but that he was afraid of a loud noise, created by banging a metal bar with a hammer. By presenting the rat and the noise at the same time, Watson was able to condition Little Albert to be afraid of the rat alone—and incidentally afraid of all the other objects as well. Little Albert was adopted shortly after these unfortunate experiences, and Watson did not have the opportunity to recondition him not to fear these objects. One must feel sympathy for both little Albert and his adoptive parents, and as in the earlier study by Bernheim, no ethics committee nowadays would permit such an experiment to be repeated.
Another form of conditioning, known as operant conditioning, involves the shaping of behaviors that are emitted rather than elicited, and might therefore be considered voluntary rather than involuntary. Such behaviors might include a rat pressing a bar, a pigeon pecking a disk, or a human operating a gambling machine in a casino. In operant conditioning, such behaviors are progressively controlled, or “shaped,” by the delivery of rewards and punishments. They can also be brought under what operant psychologists call stimulus control; if otherwise neutral stimuli or events are associated with the rewards and punishments, they too can shape behavior. The dinner bell, then, may not only induce salivation, but may also serve as the signal to enter the dining hall. The behaviorist B. F. Skinner sought to explain all human behavior, including language, through the principles of operant conditioning,35 leading to the somewhat robotic view of human society and behavior described in his utopian novel, Walden Two.36
These different forms of conditioning may be said to fall into the category of implicit memory. As we have seen, implicit memory may also include the learning of skills and even mental strategies to cope with environmental challenges. Implicit memories are elicited by the immediate environment, and do not involve consciousness or volition. Of course one may remember the experience of learning to ride a bicycle, but that is distinct from the learning itself. Many of my own childhood memories involve bicycles, including being launched by one parent in the direction of the other, turned around and launched again, thus riding in wobbly fashion back and forth. I still bear the scar from a later fall, which I remember vividly. I was nine years old, and I can remember the doctor’s amusement, as he stitched up the gash in my knee, when I asked him if the wound would be fatal.37 These are episodic memories, independent of the process of actually learning (more or less) to ride the bike.
Compared to implicit memory, explicit memory provides yet more adaptive flexibility, because it does not depend on immediate evocation from the environment. That is, we can consciously bring to mind such memories at any time, as when we muse about what happened yesterday, or try to bring to mind specific places or people that we know. This is not to say that the environment is unimportant, since we are constantly reminded of events by happenings in the world around us, or by probing questions, such as What on earth were you up to last night?
As we have seen, explicit memory includes both semantic and episodic memory. Our semantic memories are vast storehouses of facts about the world, including physical knowledge of the world and the particular environments that we inhabit, knowledge of our friends and how they behave, and even some knowledge of how language works, including vocabulary, some rules of grammar, and pragmatics—the different ways we can use language to achieve social goals. Language, though, is complex, and grammar for the most part depends on implicit rules—even linguists haven’t figured them all out explicitly. By and large, semantic memory is the stuff of education, and is continually being upgraded and expanded through the processes of science and technological invention. Sometimes, it all seems too much.
Episodic memory, then, sits at the top of an evolutionary pyramid, providing the ultimate in fine-tuning our personal lives. By recording specific events, we can react more precisely to similar events in the future, and plan future activities in detail. We can also intentionally sift through different episodes in our lives to extract relevant information—perhaps in a fashion analogous to the insertion of phrases into sentences, or sentences into narratives. By knowing what went wrong on previous occasions, we can avoid a similar catastrophe in the future. On this view, then, episodic memory is not so much a permanent record of past events as a series of YouTube–like sequences, sometimes embellished or distorted, to provide a stock of situations that we can use to construct viable and detailed futures. Clearly it would be wasteful of storage space to record every event that we experience, since our daily lives are full of repetition. Repetitive events may well contribute to semantic memory, the generalities of the world, but episodic memory is concerned with salient particularities. That said, it is still not clear from research precisely which episodes are likely to be remembered and which are not. Episodic memory has a fickleness about it.
But there is more to the story than the examination of the past. As we shall see in the next chapter, episodic memory may also provide us with a sense of time, extending into the future as well as the past.