We Are Our Brains (38 page)

The importance of the temporal lobe for memory became clear
in 1953, when the American surgeon William Scoville removed large portions of the temporal lobe from a patient famously known as H.M., a man who had developed severe epilepsy after a bicycle accident. The operation cured his epilepsy but caused profound amnesia. He was unable to learn or retain information, although his short-term memory was intact. He could, for instance, briefly remember the number seven by constantly repeating it to himself, but if he was interrupted he completely lost track of what he'd just been trying to remember. In other words, the pathway from short- to long-term memory had been cut off. If Brenda Milner, the neuropsychologist who was treating him, reentered the room only a few minutes after speaking to him, he would invariably say, “It's been so long since I've seen you!” His personal history effectively ceased from the time of the operation. In his mind's eye, he remained about thirty years of age; as he grew older, he became unable to recognize recent photos of himself. Right up to his death in 2008, he was firmly convinced that Harry Truman was president. After moving to a new house, he invariably returned to his old home, and he eventually couldn't be trusted to go out alone.

The prefrontal cortex (
fig. 15
) has many functions and also coordinates the various parts of the brain that constitute the working or short-term memory. This is the memory that allows you to keep certain things briefly in mind, like the number you want to call, the plans you're making, and the problems you need to solve. The working memory is also crucial for processing language and is thought to be underdeveloped in children who suffer from dyslexia. The prefrontal cortex works closely with the hippocampus (
fig. 26
) by focusing attention and selecting stimuli. In memory tests, the words that cause heightened activity in both these areas of the brain are the ones remembered best. If we just want to retain a number long enough to make a phone call, our working memory will suffice. But if we repeat that number often enough, we can store it in our long-term memory. The working memory, a short-term storage space for
general use, is crucial for carrying out complex tasks and for functional performance. It enabled H.M. to remember a few words or numbers, but he was then unable to transfer that information in the normal way, through repetition, to the long-term memory. The focus of H.M.'s epilepsy was in the region of the hippocampus, two-thirds of which had been removed. (The name
hippocampus
, meaning “seahorse,” reflects the shape of this structure, with its ridges and curls.) He could still perfectly recall events that had taken place more than three years prior to the operation, which proved that the hippocampus wasn't the site of remote memory storage. It was H.M.'s complete inability to form new memories after the operation that gave clues to the hippocampus's function. Studies of neurological patients have since shown that even partial damage to the hippocampus can cause considerable, long-term impairment in the ability to create memories, or anterograde amnesia.

The hippocampus specializes in combining sensory information. The location of the restaurant you arranged to meet up at, what the person you're meeting looks like, the sounds and smells from the kitchen, and the position of the set table are all fused into a single coherent item of autobiographical memory, the chronicle of your life. And later, at least if the dinner was worth it, this information is transferred to the long-term memory. The hippocampus does all this in close partnership with an area located nearby on the underside of the cortex: the entorhinal cortex (also called the parahippocampal gyrus,
fig. 26
). Scoville also removed this latter structure from H.M.'s brain. But which of the two regions does information reach first? The answer was provided by studies of epileptic patients with electrodes in their brains who performed memory tests while the electrical activity in the two regions was selectively recorded. The entorhinal cortex proved to be activated first, followed by the hippocampus.

It's also in the entorhinal cortex that the first signs of Alzheimer's appear, and the memory problems in the onset of the disease indeed
typically concern recent information. People with Alzheimer's may not know what happened an hour ago, but they can tell you detailed stories about a classmate at elementary school.

The hippocampus isn't only crucial to memory; we also need it for spatial orientation. Brain scans of taxi drivers who spent four years learning London's enormous and complex network of streets by heart showed a gradual increase in the volume of gray matter at the back of the hippocampus. Studies of people with damage to both sides of the hippocampus have shown that the hippocampus is necessary for imagining the future.

Fortunately, not all recent information is stored in the long-term memory. Who would want to retain every single detail of everything experienced during an entire lifetime, including every meal, every conversation, and every word in every book? That would make it incredibly difficult to locate and access really important information. There are people who are capable of remembering and reproducing enormous quantities of trivial information, like numerical series or entire phone books or train timetables. Yet this ability comes at the cost of other functions. These “savants” usually have a form of autism with severe impairments in areas such as social interaction or abstract thought (see
chapter 9
).

So what is normally sieved out for storage in the long-term memory? The deciding factors are the importance of the information and the emotional charge of a particular moment. Everyone knows where they were and what they were doing when they heard about the attack on the Twin Towers in New York on September 11, 2001. The amygdala (
fig. 26
), positioned just in front of the hippocampus in the temporal lobe, imprints memories that carry a strong emotional charge under the influence of the stress hormone cortisol. As a result, a traumatic experience is immediately stored for good in the long-term memory. And that explains why over 80 percent of our earliest memories have negative associations, as the psychologist Douwe Draaisma has shown. Remembering fear, shock, and sorrow
is more important for survival than pleasant memories. However, this mechanism can cause problems. A woman with temporal lobe epilepsy, whose focus was the amygdala, kept having the same hallucinations during her seizures, in which she reexperienced a traumatic period of her youth, causing her terrible distress.

There's a clear evolutionary advantage in imprinting danger in the mind—for instance in wartime—so that when a similar situation occurs you're immediately on the alert. Sometimes this natural tendency becomes pathological, however, like when a soldier returns home from a war zone but can't shake off the feeling of being endangered. If he continues to feel fearful and under threat, if the images of war constantly replay in his brain, and if he immediately dives for cover when he hears a bang, then he's suffering from post-traumatic stress disorder (PTSD). During the First World War this was called “shell shock,” and 306 British soldiers with the condition who refused to go back to the front were executed. PTSD is a sign that the amygdala has done its work too well, preventing the prefrontal cortex from signaling to the veteran that the danger is over. The amygdala is activated to respond to danger by the chemical messenger noradrenaline. So veterans with PTSD are treated with beta blockers (which have the opposite effect) to prevent dramatic experiences from being too strongly labeled by the amygdala and the individual from being overwhelmed by stressful memories. An exaggerated response by the amygdala to negative stimuli also underlies borderline personality disorder, whose symptoms include emotional instability and impulsiveness. In this disorder, negative emotions are linked to such a strong stress reaction that patients run an increased risk of retrograde and anterograde amnesia. H.M.'s amygdala had also been removed, along with other temporal lobe structures crucial to memory.

H.M.'s brain was sliced into wafer-thin sections at UC San Diego, a process that could be followed online. This was the first step in a process of intense microscopic research aimed at discovering exactly which brain structures were removed or damaged in the operation that he underwent fifty-five years earlier.

FIGURE 26.
The route taken by information on its way to long-term memory starts in the entorhinal cortex, located deep in the brain in the parahippocampal gyrus. It's briefly stored in the hippocampus in a process directed by the prefrontal cortex. From there it follows two pathways: one taking it back to the cerebral cortex for long-term memory storage, and the other—much longer—carrying it along the great arch of the fornix, suspended in the septum, to the hypothalamus, where the fibers proceed to the mammillary bodies. The information then travels via the thalamus to various parts of the cortex. The amygdala, an almond-shaped structure positioned just in front of the hippocampus in the temporal cortex, imprints memories that carry a strong emotional charge.

While we sleep, the hippocampus constantly activates memories and transmits them to the cerebral cortex. The jury's still out on whether this mainly occurs during dream sleep (REM sleep) or periods of quiet sleep. The route information takes on its way to the long-term memory starts in the entorhinal cortex. It's then briefly stored in the hippocampus in a process directed by the prefrontal cortex. From there it follows two pathways, one taking it back to the cerebral cortex for long-term memory storage and the other—much longer—carrying it along the great arch of the fornix, suspended in the septum, to the hypothalamus, where some fibers travel to the mammillary bodies (
fig. 26
) and some to the hypothalamus (
fig. 18
). Professional boxers sustain so many blows to the head that these connections are not infrequently destroyed, causing dementia, tremors, an unsteady gait, and extreme behavioral changes—a condition known as dementia pugilistica, or “punch-drunk syndrome.” Examination of the brains of ex-boxers with this syndrome often reveals a ruptured septum, a shrunken fornix, a lack of myelin (an insulating layer) around the fibers of the fornix, undersized mammillary bodies, and an oversized third ventricle due to loss of brain tissue. Other findings include Alzheimer's-type changes, shrinkage to the cerebral cortex, and cell death, mainly in the temporal region and the hippocampus (see
chapter 12
). Plenty of reason, in other words, for serious memory impairment and other malfunctions. Damage of this kind isn't confined to boxing but extends to all contact sports, from kickboxing and rugby to soccer and American football. Infarction or bleeding in the above-mentioned areas and pathways can also cause memory impairment or even dementia. In the case of Korsakoff's syndrome (caused by a combination of alcohol abuse and vitamin B1 deficiency due to poor diet) small hemorrhages and scars are found in the mammillary bodies. People with Korsakoff's have memory impairments similar to those of patients with damaged temporal lobes. They fill in the gaps in their memory with made-up stories. The importance of the mammillary bodies to memory has emerged not just from problems associated with boxing, tumors, or operations
(see
chapter 5
) but also from a bizarre accident that happened to a man during a game of billiards. His opponent's cue was accidentally forced up his nose, penetrating the underside of the brain and damaging the mammillary bodies, leaving the poor man with severe memory problems.

The mammillary bodies pass on information to the thalamus (
fig. 2
). Small infarctions in this area can lead to severe memory problems and even dementia. The information travels on from the thalamus to areas of the cerebral cortex from which memories of facts and events can be consciously recalled. This is known as the declarative or explicit memory.

Different aspects of an event are stored in different sites in the brain. When we try to recall something that happened, the various elements have to be pieced back together again. Any missing bits are filled up by our brains, a process of which we're entirely unconscious. So the common comparison of memory to a computer hard disk that can reproduce everything perfectly isn't quite accurate. A better analogy would be the way in which an archeologist tries to reconstruct an entire skeleton from a few little bones—frequently getting it wrong. Our memory is notoriously unreliable, as is often shown in court cases.

That different types of information—music, images, and faces—are stored in different parts of the cortex emerged from cases of patients with very specific problems of recall. For instance, people who suffer damage to the temporal sulcus sometimes lose the ability to recognize faces, even of the person they're married to, despite there being nothing wrong with their eyesight. But they are able to recognize objects, like their cars, because those memories are stored in
another place. Being able to identify your Ford Fiesta but not your wife must make for some interesting household scenes. This condition is known as prosopagnosia, or face blindness. Oliver Sacks described it in
The Mind's Eye
and
The Man Who Mistook His Wife for a Hat.
Dr. P., the man in question, was so severely afflicted that, instead of his hat, he tried to put on his wife's head. It's hard to conceive that he was meanwhile pursuing a distinguished career as a teacher at a music school. In its extreme form, the condition makes it difficult for people even to identify themselves in a mirror. A case is also known of a soldier who ran into his mother on the street without recognizing her when he came home on leave. Luckily it's not quite that bad in my case, but I've always had trouble recognizing faces, something that often leads to embarrassing situations. I occasionally introduce myself to a person who then gazes back at me in amazement and says, “Yes, I know who you are; we've been on the same committee for three years now.” My father was troubled by this problem too, which does seem to run in families. It's one of the mutations that has been passed down to me. Yet defects in pattern recognition are clearly extremely selective, because I have excellent recall when it comes to microscope samples. More than once I've looked at a sample I haven't seen for several years and thought (rightly as it turned out), “Oh, that's Mr. X or Ms. Y.” Yet if I'd met the individuals in question after a similar interval I would never be able to recognize them.