One of the critical properties that makes the human mind so extraordinarily suited to understanding and dealing with the world is its ability to shift in time—to model the future and reconstruct the past. Reconstruction of the past requires memory, and memory is fundamental to nearly any cognitive skill. It is involved in complex processes such as problem-solving, and it is involved in even what seem to be the simplest skills, such as recognizing a familiar face. The role played by memory in cognition is complex enough that not just a single memory system will do. Humans and other animals have several memory systems with different characteristics and different neural implementations,
Encyclopedia of the Human Brain Volume 2
Copyright 2002, Elsevier Science (USA).
All rights reserved.
and these systems, acting in concert, contribute to the human mind's tremendous adaptability.
At the broadest level, one can distinguish between ''working memory'' and ''long-term memory.'' Working memory refers to the system that stores a small amount of information for a brief span of time. Information stored in working memory is then used in the service of other cognitive tasks. For example, if we were solving an arithmetic problem such as 817 + 723 without the benefit of writing anything on paper, working memory would be used to store the problem, store the intermediate steps in the addition, and store the final solution. In addition to temporary storage, an important component of working memory is what is called ''executive processing:'' the set of operations that permits one to manipulate the contents of working memory. In the previous example, executive processes would be involved in switching attention from one column of addition to another and in organizing the order of steps to arrive at a final sum. Whereas there is as yet no overall agreement about a full list of executive processes, they generally can be thought of as operations that regulate the processes operating on the contents of working memory, processes such as selective attention to relevant information (more about this shortly).
In contrast to the short duration and small capacity of working memory, long-term memory is a system with very long duration memory traces and a very large storage capacity. In our previous mental arithmetic example, long-term memory would be the repository of the facts of addition that would be needed to solve the mental arithmetic problem. Of course, long-term memory stores much more than that. For example, it is the repository of all the words we know in our language, of the sensory information that we all have stored for untold numbers of events (e.g., the taste of a good chocolate), of the spatial information we have stored for navigating around our world, and so on. In addition, many pieces of information are stored that we normally do not retrieve consciously but that nevertheless guide our everyday behavior, such as the rules of language or habitual actions in which we engage every day.
Larry Squire of the University of California and Endel Tulving of the University of Toronto have proposed schemes that summarize the various forms of long-term memory. One way of synthesizing and expanding these schemes is shown in Fig. 1. The figure shows that there are two broad divisions of long-term memory: declarative and procedural. Declarative memory refers to the facts and events that we can retrieve at will, often consciously. By contrast, procedural memory refers to stored information that has an impact on our behavior but that is not willfully retrieved. Consider, for example, the concept of a bicycle. A declarative memory you might have of a bicycle is that it is blue and that it has 21 gears, mountain terrain tires, two handbrakes, and so on. These are all facts that can be willfully retrieved from memory. By contrast, you also have stored information that allows you to ride your bicycle—a task that any 6-year-old child will tell you is not easy. This information is not consciously retrievable; indeed, it is a nontrivial problem in physics and kinesiology to describe just how people are able to ride a two-wheeled bicycle without falling over. The contrast between these two sorts of memory is a contrast between declarative and procedural memory. Perhaps the most compelling evidence that procedural knowledge is different from declarative knowledge is that patients with damage to their hippocampi and surrounding medial temporal lobes can learn new procedural skills, even though they cannot encode where they learned
Items Sources Facts Events
Figure 1 A taxonomy of various forms of long-term memory.
Items Sources Facts Events
Figure 1 A taxonomy of various forms of long-term memory.
the skill or remember any details of having practiced it, even when that practice occurred very recently. Other patients with cerebellar damage can remember the practice sessions, but their skills on most motor tasks do not improve. This pattern of deficits, called a double dissociation, helps to define procedural and declarative processes as distinct types of memory.
Declarative memory itself comes in two forms. One is called episodic memory, or memory for specific events, and it consists of memory traces that are accompanied by memory for the context in which they were formed. Each piece of episodic memory has a source tag associated with it, possibly including the time and place of memory formation and other details about the context. When retrieving an episodic memory, one can retrieve either the item itself, given information about the source, or the source, given information about the item. For example, you may recall where and when you purchased your current bicycle or, given the time and place, you may recall the features of the bicycle that you purchased. The other category of declarative memory is semantic. This type of memory consists of the vast store of facts and events that you have in long-term memory, regardless of whether you can retrieve when and where you learned them. For example, you may remember the fact that bicycles can be mountain bikes, racing bikes, hybrid bikes, and so on, yet you may not be able to recall when or where you learned this semantic fact.
Procedural memories also are of various sorts. There are skills, for example, such as riding a bicycle. There are classically conditioned responses, which entail a previous pairing of an unconditioned with a conditioned stimulus to yield a conditioned response. And there are cases of priming, in which a previously learned piece of information facilitates processing of some new piece of information. Psychological measures of priming, such as decreases in response time to recognize a previously viewed word, indicate that a trace of the previously learned piece of information is affecting current cognitive processing—even if there is no conscious recollection of having seen the word before.
Another important dimension of memory, whether working or long-term, is the type of information being stored. As we shall see later, the brain circuitry involved in a memory task honors the type of information that is stored and retrieved. Perhaps the most frequently studied case of this concerns the distinction between linguistic information (such as letters, words, sentences, and stories) and visual or spatial information (such as a scene, an object, a face, or a spatial environment). By now ample evidence exists that the two hemispheres of the brain are differentially activated by these two types of information, with the left hemisphere specialized for verbal information and the right for visual or spatial information in most humans.
Memory entails three cognitive operations: encoding, storage, and retrieval. These terms refer to the sequence in which memory processes are thought to occur. Entering information is first put into the proper internal code and a new memory trace is formed (encoding). Encoding is followed by storage of the information for some period of time. This stage may include consolidation or alteration of memory traces to make them last longer and ease retrieval. Retrieval is the process of reporting information from storage.
The nature of encoding depends on two factors: the type of material that is involved in the memory task and the task that is performed with that material. The type of material exerts a strong influence on the path of activity in the brain early in the processing sequence. The best example of this is the visual system. Spatial information about a visual stimulus is selectively routed to a dorsal stream of information processing that mainly includes the parietal lobes, whereas information about shape and other nonspatial object features of the same stimulus is processed by a ventral stream in the occipital and inferior temporal lobes. To generalize from this example, we can say that the nature of incoming information will influence the path of processing that the information takes in the brain. Beyond this, though, there is also an influence of the task with which a person is faced, as many different operations may be performed on any given type of material. For example, one can process a word by noting its meaning or by noting whether it is printed in uppercase or lowercase letters. These very different types of processing on the same stimulus yield different patterns of activation in the brain, as we shall see later.
Once encoded, information is retained for some period of time. Consistent with the fundamental distinction between working and long-term memory, the length of the retention interval in large part will determine which of these systems is most heavily involved. Retrieval after short retention intervals—up to, perhaps, intervals as long as 30 sec to 1 min—uses working memory. Retrieval of information stored for longer periods will, under most circumstances, necessitate the involvement of long-term memory storage. Which memory system is engaged will be revealed by the circuitry that is activated. Working memory engages circuitry in frontal and parietal cortices most prominently, whereas long-term memory requires the involvement of frontal and parietal circuitry as well as hippocampal and parahippocampal mechanisms.
Just as encoding different types of material engages different mechanisms, storage of different types of material also requires different mechanisms. This has been demonstrated most handsomely in the contrast between verbal and visual material, which predominantly activate left and right hemisphere structures, respectively. This distinction has been demonstrated for both working memory and long-term memory, as we shall see later.
Once encoded and stored, information in memory can then be retrieved as needed. Suppose, for example, that we ask a person to memorize a list of words. Retrieval can be accomplished in several ways. We might simply ask the person to recall as many of the words as possible (free recall). Or we might guide recall by giving some of the words on the list as hints and asking the person to recall the others (cued recall). Or we might present the person with a longer list of words, some of which were presented on the original list and some not, and ask the person to decide which is which (recognition). Any of these procedures requires the person to access the stored information in memory and produce an explicit response that depends on that stored information. For this reason, these are often called explicit tests of memory. However, there are also implicit tests. Suppose, for example, that we presented someone with a list of words and later flashed the same words and new words, one by one, so briefly that they were difficult to identify. If the person were more accurate in identifying words that had been presented on the original list than ones that had not (which is what happens in this perceptual identification situation), then we could conclude that the original words were stored in memory even though no explicit retrieval of them was ever demanded. The process of storage and use of information without explicit memory is called priming. Evidence from positron emission tomography (PET) and functional magnetic resonance imaging (fMRI) suggests that implicit and explicit tests of memory recruit different brain areas, as reviewed later.
With these preliminaries about memory in place, we are now in a position to review what neuroimaging evidence has contributed to understanding basic mechanisms of human memory.
Was this article helpful?