Figure 7.1 shows several subheadings in the “long-term memory” box. While we often think of memory as a single process, memory storage is actually more than one type of process. As early as 1911, the French philosopher Henri Bergson stated that our past survives in two fundamentally different forms, conscious and unconscious (Schacter, 1996). Scientists usually characterize these two forms (and their subcategories) as procedural or declarative. As we will see, these two forms of memory are localized in different neural systems.
Figure 7.1. Long-Term Memory
Procedural Memory—Skills and Priming
Procedural memory is knowing how versus knowing what. It is sometimes called nondeclarative: You do not need to “declare” anything, and you may not be able to say much about what you are doing, for the information to be stored. The first type of procedural memory is your ability to store automatic processes for routine actions. You can think of these processes as skills, the “how” to do things. They may be simple procedures, such as walking, brushing your teeth, or tying your shoes, or they may be more complex, such as driving a car or decoding words. These procedures have in common their automatic nature. After a good deal of repetition and practice, we perform them without conscious thought. The famous cognitive psychologist Jerome Bruner called procedural memory a memory without record (Squire & Kandel, 2000). The automatic procedures form a sort of unconscious stimulus-response bond. Once we have a skill or habit at this level, however, it becomes difficult to access it in any way except by performing it. Imagine trying to teach someone to tie a shoe, swing a golf club, or write a word without physically demonstrating it. We no longer know how we accomplish the procedure. Its separate parts or its rules of operation are virtually inaccessible to our consciousness.
Most of the skills we have been discussing involve motor activity, but some types of skillful behavior are not based on learned movements. An example of a nonmotor skill is reading. When you first learn to read, your eyes move slowly from word to word; but with a great deal of practice, you move through the words much more quickly. Skilled readers move their eyes about four times a second, taking in the meaning of more than 300 words a minute (Squire & Kandel, 2000).
A second type of procedural memory is known as priming. Priming involves being influenced by a past experience without any awareness of consciously remembering that experience. In a sense, priming is similar to the skill learning mentioned in the previous section. In neither case are you consciously aware of what you are doing—which is why both skills and priming are sometimes called implicit memory, in contrast to conscious recollection, or explicit memory (Schacter, 1996). In priming experiments, researchers show subjects lists of words, then hours or days later show them another list and ask whether they have seen any of the words before. On a second task, subjects are given the beginning of a word from the lists (abs____ for absent and inc____ for income) and asked to complete the word. The subjects perform much better on the fragmentation completion task (nondeclarative or implicit memory) than on the task in which they have to identify whether or not they have seen the word before (declarative or explicit memory). It may be that for a period of time after seeing a word, less neural activity is required to process that word again (Squire & Kandel, 2000). This has been seen in amnesiacs who are able to learn new procedural skills but have no memory of learning them. This is why H. M. is able to improve his performance of new motor skills (such as mirror writing), but does not remember ever having done them before (Amaral, 2000). This type of experiment indicates that our memory can be influenced by experiences that we fail to recollect consciously. Having seen or experienced something previously seems to prime our ability to recall it later. Procedural memory, whether skill learning or priming, provides dramatic evidence that unconscious mental activities do exist.
Declarative Memory—Semantic and Episodic Recall
Declarative memory is our ability to store and recall information that we can declare (speak or write). Unlike procedural memory, declarative memory requires conscious processing; it is reflective rather than reflexive. Instead of the automatic, unconscious recall of how to do something, declarative memory permits us to consciously recall and discuss what something is, or recall and describe an event that occurred in the past. This dual function has led to declarative memory being subdivided into two categories, episodic and semantic memory.
Episodic memory is sometimes called “source memory,” because it involves remembering where and when information was acquired. It allows you to recall a hike you once took, how much you loved your 1st grade teacher, and a surprise party on your 16th birthday. It is your record of faces, music, facts, and your individual experiences, a sort of “autobiographical reference” (Squire & Kandel, 2000). As critical as episodic memory is (it's important to remember where you parked your car), it can at times be problematic. The brain does not store memories in a linear manner, as a tape recorder or video camera does; it stores memories in neural circuits or networks. When we recall an event, we are actually reconstructing it. While many events are important or emotional enough to be remembered, the details often escape us. What the brain does in this case is “fill in” the details. This process is called refabrication; it can be defined as the reconstructing of a memory from bits and pieces of truth. As we tell our stories over and over, we embellish them, add to them, and make them a bit more elaborate. Eventually the refabrication becomes the memory, and it is virtually impossible to distinguish it from what actually happened. Even though the memory of the event is quite vivid, the details may actually be inaccurate.
Semantic memory, on the other hand, is generally fairly accurate. Semantic memory includes words, the symbols for them, the rules for manipulating the words, and their meanings. It also consists of rules of grammar, chemical formulas, rules of computing in mathematics, and your general knowledge about your world. These facts are normally independent of a particular time or place. Knowing that 6 × 7 = 42 is an example of semantic memory; remembering what grade you were in when you learned the multiplication tables is an illustration of episodic memory.
The Cellular Basis of Memory
We have been looking at the big picture of memory and its various types. It is important to remember, however, that underlying our memory (no matter which type) are changes in the neurons and connections between neurons that form the physiological basis of storing and retrieving information. What are the cellular mechanisms that allow information to make the crucial leap from working memory to long-term memory? Endel Tulving, considered by some to be the world's foremost authority on cognitive theories of memory, states:
As a scientist I am compelled to the conclusion—not postulation, not assumption, but conclusion—that there must exist certain physical-chemical changes in the nervous tissue that correspond to the storage of information, or to the engram, changes that constitute the necessary conditions of remembering (Tulving, cited in Gazzaniga,1997, p. 97).
The study of the molecular events underlying memory formation is one of the most exciting fields of neuroscientific study. The introduction to memory in Chapter 5 mentions “Hebb's Law.” In the 1940s, Canadian neuroscientist Donald Hebb proposed that a synapse between two neurons is strengthened if the neurons are active or firing at the same time. His theory is generally accepted in the field of neuroscience today; however, how this occurs is still open to some debate (Squire & Kandel, 2000). One current hypothesis is that the synapses between neurons representing experiences become strengthened or potentiated over a period of time. This is referred to as long-term potentiation (LTP). LTP has been demonstrated in the laboratory with animals and has been the predominant model of the cellular basis of memory for more than two decades. Not all neuroscientists agree that the experiments necessarily reflect what happens during the storage of a memory in humans, but most agree that it is at least one of the important mechanisms involved in changing the synaptic strength between neurons in neural networks.
Long-Term Potentiation (LTP)
How might LTP result in a memory? First, let's review what we understand about how neurons allow us to see or to hear. We know that the experience of seeing a yellow rose or a blue ball is the result of the activation of a particular group of neurons in the visual cortex. Likewise, a group of neurons firing together in the auditory cortex will result in the experience of a certain tone or note of music. A memory appears to entail a similar firing of neurons, but the pattern of firing remains encoded in a neural circuit or network after the stimulation that originally caused the neurons to fire has ceased. You can remember the image of the rose or the ball, and you can hear the melody of the song in your head. It appears that this is possible because when two or more neurons are active at the same time, they become more sensitive, i.e., more likely to fire a second time. The more often the pattern of neurons is activated, the more efficient the synapse becomes. This increased efficacy of the synapses is what many scientists refer to as LTP. Researchers have demonstrated LTP in several parts of the hippocampus and surrounding structures in the medial temporal lobe, which we will see are critical to the formation and storage of memories.
There is some evidence that the chemicals released at the synapse that lead to LTP may result in the modification of proteins, synthesis of new proteins (implicated in memory), and changes in gene transcription (Amaral & Soltesz, 1997).
Growth of Synapses
In the 1960s, Marian Diamond, Mark Rosenzweig, and their colleagues at the University of California, Berkeley, demonstrated that substantial changes in the brain's architecture can be influenced by an animal's environment (Diamond, 1988). Somewhat later, William Greenough at the University of Illinois extended the research on “enriched” environments. Enrichment for the rats in both these studies was provided by placing a colony of rats together in a large cage with toys that were changed every few days. The rats raised in the enriched environment showed growth in the thickness and weight of their cortices, due to larger cortical neurons, heavier branching of dendrites, and larger synapses. Increases of up to 20 percent more synapses per neuron were found in the visual cortex of some of the animals. These structural changes in the rats' brains resulted in their being better able to solve complex maze problems. [Diamond reports that wild rats obtained from their natural environment had even more dendritic growth—and heavier cortices—than those in enriched environments (M. C. Diamond, personal communication, July 2000).]
There is still much to be understood before the mystery of how experiences are stored at a cellular level comes close to being solved. Whatever the process or processes may turn out to be, the fact remains that when we learn, truly amazing changes take place in the neural connections in our brains, and the methods we use to structure learning experiences for our students affect the strength and duration of those changes.
How Are Memories Stored?
Suppose you were asked to recall an event in your life, perhaps a graduation celebration or a surprise birthday party. In all probability you would be able to describe many aspects of that experience: the people who were there, the food that was served, the room you were in, the sound of people singing “Happy Birthday,” and perhaps some of the gifts you received. The memory probably came to you in a fairly complete form, so that it seems that this particular memory must be stored in a special place in your brain, ready to be recalled in its entirety whenever you wish. But in actuality, no complete scenarios or pictures are stored anywhere in the brain; you have to reconstruct these memories every time. While this may seem inefficient and even counter-intuitive, the process by which we encode experience and later recall it really makes a lot of sense.
In his book Inside the Brain, science writer Ronald Kotulak uses the metaphor of eating a meal to represent the encoding and storing of information:
The brain gobbles up its external environment in bites and chunks through its sensory system: vision, hearing, smell, touch, and taste. Then the digested world is reassembled in the form of trillions of connections between brain cells that are constantly growing or dying, or becoming stronger or weaker, depending upon the richness of the banquet (Kotulak, 1996, p. 4).
When you think carefully about it, this is efficient. Our experiences are disassembled into parts and stored in specialized networks of cells. The same brain cells can be used many times to recall similar lines or colors or smells. For example, the cells in the visual cortex that allow us to perceive the color red can be used to see a red rose, a red heart, the red in a sunset, or a red tie. The same is true in the auditory cortex and other sensory areas as well. In a sense, many parts of the brain each contribute something different to the memory of a single event. Our knowledge is built on bits and pieces of many aspects of a given thing—its shape, color, taste, or movement. But these aspects are not laid down in a single place; there is no memory center in the brain that represents an entire event at a single location.
How Are Memories Recalled?
If memories are not stored in specific locations in the brain, how do we retrieve them? Our ability to remember is essentially a process of reconstruction or reactivation. As we have seen, the various elements of past experience reside all over the brain—in the visual cortex, auditory cortex, and other areas. Antonio Damasio (1994), professor and head of the department of neurology at the University of Iowa College of Medicine, describes recall as an activation of all these separate sites in unison, creating an integrated experience. You don't even need all the pieces to reconstruct the total, just the definitive elements. Recall the picture of the dalmatian (Figure 5.3) in Chapter 5. The dog is not clearly defined, but you don't need all the components to reconstruct the total picture, just the definitive elements. When a critical mass of sensory neurons is activated, the brain fills in the missing portions to complete the picture. Keep in mind, though, that the image of the dalmatian must have been previously stored in order to be retrieved; that is, if you had never before seen a dalmatian, you probably wouldn't be able to fill in the blanks.
The same is true when remembering an event. Depending on the cue or reminder, only certain fragments of the total memory may be activated. If the cue is weak or unclear, what is reactivated may differ from the original memory or even belong to another episode. This is why episodic memory details are often fuzzy or even completely inaccurate, and why “eye-witnesses” of events are generally unreliable. Memory researchers Elizabeth and Geoffrey Loftus are well-known for their studies of how memories can be modified or distorted by the type of questions asked in a memory retrieval test. They have also demonstrated that false memories can be planted if the memories contain some aspect that reasonably could have occurred (Loftus & Loftus, 1975).
Brain Structures Involved in Storage and Retrieval
Declarative and procedural memory, while sharing many of the same cellular mechanisms, do not employ the same brain structures for their processing. The two major structures involved in memory processing are the cortex and a part of the brain called the medial temporal lobe. It appears that the brain stores memories in the same structures that are engaged in initially perceiving and processing stimuli; however, these structures differ, depending on whether the memory is procedural or declarative. Understanding the anatomy involved in these two types of memory will further clarify the types of activities and practice best suited to each one.
The Procedural Pathway to Long-Term Storage
You may drive your car on a familiar route, arrive at your destination, and realize that you were not aware of driving there. When you meet someone new, you automatically extend your hand in greeting. You may read a page of text, get to the bottom of the page, and realize that you do not remember what you just read, usually because you were thinking about something other than the text. These motor skills, habits, and perceptual skills are all examples of procedural or non-declarative memory, and all were accomplished without your conscious awareness. As mentioned earlier, trying to consciously express any of these skills while performing them impairs your performance. But if you think back to when you first learned to drive or to read, none of these skills or habits were automatic. They required a great deal of conscious attention and practice.
In the early stage of skill (procedural) learning, three brain areas are involved in laying down the new pathways: the prefrontal cortex, the parietal cortex, and the cerebellum. Their combined activity allows you to pay the necessary conscious attention to the task and ensures that the correct movements are assembled correctly. After practice, however, all these areas show less activity, and other structures, including the motor cortex, become more engaged (Squire & Kandel, 2000).
In nonmotor procedural learning, such as decoding words, the brain area that appears to be most heavily involved is the visual cortex. With extended practice, we improve our ability to discriminate between different line orientations and letter configurations. The ultimate long-term effect is to change the actual neural structure of the visual cortex, which alters the machinery of perception over time. Remember that these changes do not involve understanding word meanings, only the ability to recognize the configurations more quickly. All this occurs outside awareness, as has been demonstrated by amnesiac patients who are able, with practice, to improve their speed in reading a selection of prose, but who do not remember the text in any ordinary sense (Squire & Kandel, 2000).
The Declarative Memory Pathway to Long-Term Storage
The journey from perception to storage of both semantic and episodic memory begins with the sensory receptors receiving stimuli. The stimuli register in the appropriate areas of the cortex (visual, auditory, etc.) and then travel to the hippocampus and an adjacent cluster of structures within the medial temporal lobe. These structures register the stimuli as neural patterns in much the same way as they were registered in the cortex. Note that the hippocampus is not the ultimate storage repository of memory; rather, it acts as an intermediate storage site for cortical representations on their way to long-term memory (Squire & Kandel, 2000). These representations can be reactivated during recall, and each time they are replayed the messages are sent back to the cortex, where the stimuli originally registered. This reactivation of the original neural patterns strengthens them, making them less likely to fade. With repeated activation, the memories form neural links that become more or less permanently embedded in the frontal cortex and temporal cortex. These links remain in long-term memory long after the hippocampal representations have faded. So we can see why the hippocampus is essential for forming new memories, but it becomes less essential over time as these memories are eventually stored in the cortex. This provides an explanation of why amnesiacs with damage to the hippocampus can no longer lay down new permanent memories, but often are able to remember long-ago events that occurred before the brain damage.
Consolidation
Patients receiving electroconvulsive therapy (a controlled series of electric shocks to the brain) often forget experiences and learnings that occurred just before the treatment. This condition is called retrograde amnesia. However, if the treatment is delayed for a period of time after learning new information, the shock is less likely to disrupt recall. The reason for this appears to be that even after an event has been placed into memory, some time must pass for the memory trace to become fully established or organized in the brain.
In the late 19th century, German psychologists Georg Müller and Alfons Pilzecher conducted studies using the nonsense syllables of Ebbinghaus's experiments and found that learning a second list of syllables immediately after learning a first list interfered with later recall of the first list. Without disruption, the newly formed memories gradually became more stable. The researchers labeled this gelling or setting time the consolidation period (Squire & Kandel, 2000). We now know that memory is not formed at the moment information is acquired; memory is not a simple fixation process. Rather, it is dynamic, with unconscious processes (called consolidation) that continue to strengthen and stabilize the connections over days, weeks, months, and years (Gazzaniga, Ivry, & Mangun, 1998). Consolidation is undoubtedly enhanced by rehearsal. When we “replay” our experiences (talk and think about them), we are providing more opportunities for consolidation. Perhaps this is why instruction that allows students to hook new information to previous experiences increases the strength and complexity of their neural connections and, therefore, the retention of the information.
Scientists have studied the consolidation process extensively in rats, mice, and fruit flies. One interesting set of experiments suggests that consolidation requires new protein synthesis. When mice receive an injection of a substance that inhibits protein synthesis just before training, they have a profound loss of long-term memory when tested three or more hours later. Mice given a saline injection show no long-term memory loss.
Consolidation seems to be the result of biological changes underlying the retention of learned information. What are these biological changes? Given what we know about the importance of the hippocampus in the formation of long-term memory, it is not surprising that the function of the hippocampus, and of nearby structures in the medial temporal lobe, is integral to consolidation. Without the mediating effects of the hippocampus, consolidation could not take place. As we practice or repeat our experiences, however, they may become consolidated and the hippocampal structures will no longer be needed.
Sleep and Consolidation
Recent research also points to sleep as another player in the consolidation process. During sleep, particularly during the rapid eye movement (REM) stage, our brains are relieved from processing the continual input of information that occurs during waking. In his book Searching for Memory, Daniel Schacter tells of a hypothesis developed by neuroscientist Jonathan Winson, who suggested that during sleep, the brain continues to work through the experiences of the day. Winson's ideas have recently received support from research on animal brains. Recordings taken of rats' brains during sleep indicate that the hippocampus is particularly active “playing back” recent experiences to the cortex, where they will eventually be stored (Schacter, 1996). You have probably noticed that your own dreams often contain fragments of what you did during the day. It may be that your brain is replaying these experiences, helping to consolidate them just as consciously reviewing information during waking hours does. If this is true, sleep is an important participant in the formation of long-term memories.
Consolidation in Motor Memory
Researchers most frequently discuss the concept of consolidation in terms of declarative memory, which relies on brain structures in the medial temporal lobe. Recent research indicates that learning motor skills (a procedural memory) also involves consolidation. Researchers at the Massachusetts Institute of Technology Department of Brain and Cognitive Sciences have discovered that “learning a motor skill sets in motion neural processes that continue to evolve after practice has ended” (Brashers-Krug, Shadmehr, & Bizzi, 1996). When subjects learned a second motor task immediately after a first skill was learned, the consolidation of the first motor skill was disrupted. This disruption did not occur if four hours elapsed between learning the first and second skills. The researchers propose that motor skill consolidation relies on the same structures in the medial temporal lobe that are necessary for the consolidation of explicit (declarative) memory tasks.
Educational Implications of Consolidation
It is tempting to try to directly apply the research on consolidation to the classroom. It would be helpful for teachers to know just how long students' brains need to consolidate a particular learning before moving on to another. Unfortunately, the research doesn't give us this kind of detailed information. We do know, however, that consolidation occurs and that it takes time. We also know that teaching something new too soon disrupts consolidation of previous learning. What we don't know is how much time is needed for consolidation; and, therefore, we should be wary of specifying time lengths between introduction of concepts or skills. Neuroscience seldom gives us information that can be applied directly to classroom practice, but we need to take what we know about consolidation into account when designing instruction. For example, building elaborative rehearsal strategies into our instruction—allowing students time to process information more in depth—may increase the strength of the learning because these strategies allow consolidation to take place.
Teaching for Long-Term Memory
Most learning in life is incidental. In everyday life, generally, we make no particular effort to record our experiences for later. Our interests, preferences, and survival needs direct our attention and determine how well information is encoded. Although incidental learning has value, we cannot trust that everything we need to remember will be “incidentally” encoded. More often than not, we have to expend some effort to make certain that we'll be able to recall the information when we need it. No one knows more about how difficult this can be than teachers. Students often memorize the information for a test and then promptly forget it. The problem is exacerbated by the demands to cover more curriculum—and covering is often all that happens. Coverage (going over information superficially) does not build strong neural connections and, therefore, is seldom remembered or remembered incorrectly. This problem is difficult to solve, but perhaps the information in this chapter will help educators understand what is necessary to produce long-term retention of information.
The term elaborative rehearsal was introduced in Chapter 6. We are now in a better position to understand why this type of practice is more effective for producing long-term declarative memory than rote rehearsal. The more fully we process information over time, the more connections we make, the more consolidation takes place, and the better the memory will be. In the remaining chapters of this book, we discuss a number of elaborative rehearsal strategies. Most of them require students to reflect on the information being taught, relate it to something they already know, form meaningful mental associations, or employ some other effective elaborative encoding strategy.
Synapse Strengtheners
|
- Based on the information in this chapter about what is necessary for information to be stored in long-term memory, explain why many educators say, “We need to teach a lot less a lot better.”
- Without looking back at the text, draw a diagram of long-term memory with all its subdivisions. Under each subdivision, write at least one example typical of that type.
- If you are reading this book as part of a study group, devote one session to discussing what constitutes an enriched environment for students. You might also want to describe the elements of an enriched environment for teachers.
- Explain to a fellow educator what is meant by consolidation and why it is not a good idea to teach two separate skills close together.
|