MECHANISMS OF MEMORY
Memory is learning that has persisted over time, information that has been stored and can be retrieved. Research on memory’s extremes has helped us understand how memory works. Memory is the persistence of learning over time. The Atkinson-Shiffrin classic three-stage memory model (encoding, storage, and retrieval) suggests that we (1) register fleeting sensory memories, some of which are (2) processed into on-screen short-term memories, a tiny fraction of which are (3) encoded for long-term memory and, possibly, later retrieval. Contemporary memory researchers note that we also register some information automatically, bypassing the first two stages. And they prefer the term working memory (rather than short-term memory) to emphasize the active processing in the second stage.
The Phenomenon of Memory.
Memory function helps fixing of perceived information, keeping it in verbal form or as traces of percept stimuli and recognizing of this information in proper time. Genetic memory keeps information about body structure and forms of its behavior. Biological memory is presented in both philogenetic and ontogenetic forms. The immune memory and psychical memory for instance, belong to ontogenetic memory.
Memory is the persistence of learning over time. Psychologists have proposed several information-processing models of memory. We will use the influential three-stage processing model, which suggests that we (1) register fleeting sensory memories, some of which are (2) processed into on-screen short-term or working memories, a tiny fraction of which are (3) encoded for long-term memory and, possibly, later retrieval.
General characteristics of memory are duration, strength of keeping the information and exactness of its recognizing. In man mechanisms of perception and keeping the information are developed better, comparing to other mammalians.
According to duration is concerned short-time and long-time memory; in relation to kind of information – sensory and logic.
Studying Memory: Information-Processing.
A model of how memory works can help us think about how we form and retrieve memories. One model that has often been used is a computer’s informationprocessing system, which is in some ways similar to human memory. To remember any event, we must get information into our brain (encoding), retain that information (storage), and later get it back out (retrieval). A computer also encodes, stores, and retrieves information. First, it translates input (keystrokes) into an electronic language, much as the brain encodes sensory information into a neural language. The computer permanently stores vast amounts of information on a drive, from which it can later be retrieved.
Like all analogies, the computer model has its limits. Our memories are less literal and more fragile than a computer’s. Moreover, most computers process information speedily but sequentially, even while alternating between tasks. The brain is slower but does many things at once.
Psychologists have proposed several information-processing models of memory. One modern model, connectionism, views memories as emerging from interconnected neural networks. Specific memories arise from particular activation patterns within these networks. In an older but easier-to-picture model, Richard Atkinson and Richard Shiffrin proposed that we form memories in three stages:
1. We first record to-be-remembered information as a fleeting sensory memory.
2. From there, we process information into a short-term memory bin, where we encode it through rehearsal.
3. Finally, information moves into long-term memory for later retrieval.
Although historically important and helpfully simple, this three-step process is limited and fallible. In this chapter, we use a modified version of the three-stage processing model of memory. This updated model accommodates two important new concepts:
• Some information, as you will see later in this chapter, skips Atkinson and Shiffrin’s first two stages and is processed directly and automatically into longterm memory, without our conscious awareness.
• Working memory, a newer understanding of Atkinson and Shiffrin’s second stage, concentrates on the active processing of information in this intermediate stage. Because we cannot possibly focus on all the information bombarding our senses at once, we shine the flashlight beam of our attention on certain incoming stimuli—often those that are novel or important. We process these incoming stimuli, along with information we retrieve from long-term memory, in temporary working memory. Working memory associates new and old information and solves problems.
People’s working memory capacity differs. Imagine being shown a letter of the alphabet, then asked a simple question, then being shown another letter, followed by another question, and so on. Those who can juggle the most mental balls—who can remember the most letters despite the interruptions—tend in everyday life to exhibit high intelligence and to better maintain their focus on tasks. When beeped to report in at various times, they are less likely than others to report that their mind was wandering from their current activity.
Encoding: Getting Information In.
Some information, such as the route you walked to your last class, you process with great ease, freeing your memory system to focus on less familiar events. But to retain novel information, such as a friend’s new cellphone number, you need to pay attention and try hard.
Thanks to your brain’s capacity for simultaneous activity (for parallel processing), an enormous amount of multitasking goes on without your conscious attention. For example, without conscious effort you automatically process information about:
• space. While studying, you often encode the place on a page where certain material appears; later, when struggling to recall that information, you may visualize its location.
• time. While going about your day, you unintentionally note the sequence of the day’s events. Later, when you realize you’ve left your coat somewhere, you can recreate that sequence and retrace your steps.
• frequency. You effortlessly keep track of how many times things happen, thus enabling you to realize “this is the third time I’ve run into her today.”
• well-learned information. For example, when you see words in your native language, perhaps on the side of a delivery truck, you cannot help but register their meanings. At such times, automatic processing is so effortless that it is difficult to shut it off.
Deciphering words was not always so easy. When you first learned to read, you sounded out individual letters to figure out what words they made. With effort, you plodded slowly through a mere 20 to 50 words on a page. Reading, like some other forms of processing, initially requires attention and effort, but with experience and practice becomes automatic.
We encode and retain vast amounts of information automatically, but we remember other types of information, such as this chapter’s concepts, only with effort and attention. Effortful processing often produces durable and accessible memories.
When learning novel information such as names, we can boost our memory through rehearsal, or conscious repetition. The pioneering researcher of verbal memory, German philosopher Hermann Ebbinghaus, showed this after becoming impatient with philosophical speculations about memory. Ebbinghaus decided he would scientifically study his own learning and forgetting of novel verbal materials.
The spacing effect is our tendency to retain information more easily if we practice it repeatedly (spaced study) than if we practice it in one long session (massed practice, or cramming). The serial position effect is our tendency to recall the first item (the primacy effect) and the last item (the recency effect) in a long list more easily than we recall the intervening items.
How We Encode.
Some types of information, notably information concerning space, time, and frequency, we encode mostly automatically. Other types of information, including much of our processing of meaning, imagery, and organization, require effort. Mnemonic devices depend on the memorability of visual images and of information that is organized into chunks. Organizing information into chunks and hierarchies also aids memory.
What We Encode.
When processing verbal information for storage, we usually encode its meaning, associating it with what we already know or imagine. Whether we hear eye-screem as “ice cream” or “I scream” depends on how the context and our experience guide us to interpret and encode the sounds.
Visual encoding (of images) and acoustic encoding (of sounds) engage shallower processing than semantic encoding (of meaning). We process verbal information best when we make it relevant to ourselves (the self-reference effect). Encoding imagery, as when using some mnemonic devices, also supports memory, because vivid images are memorable. Chunking and hierarchies help organize information for easier retrieval.
When processing verbal information for storage, we usually encode its meaning, associating it with what we already know or imagine. Whether we hear eye-screem as “ice cream” or “I scream” depends on how the context and our experience guide us to interpret and encode the sounds. When the students heard the paragraph you have just read, without a meaningful context, they remembered little of it. Research suggests the benefits of rephrasing what we read and hear into meaningful terms. People often ask actors how they learn “all those lines.” They do it by first coming to understand the flow of meaning. We have especially good recall for information we can meaningfully relate to ourselves.
Mnemonic devices can also help organize material for our later retrieval. When Bransford and Johnson’s laundry paragraph became meaningful, we could mentally organize its sentences into a sequence. We process information more easily when we can organize it into meaningful units or structures.
Storage: Retaining Information.
When people develop expertise in an area, they process information not only in chunks but also in hierarchies composed of a few broad concepts divided and subdivided into narrower concepts and facts. This chapter, for example, aims not only to teach you the elementary facts of memory but also to help you organize these facts around broad principles, such as encoding; subprinciples, such as automatic and effortful processing; and still more specific concepts, such as meaning, imagery, and organization. Organizing knowledge in hierarchies helps us retrieve information efficiently.
Information first enters the memory system through the senses. We register and briefly store visual images via iconic memory and sounds via echoic memory.
As information enters the memory system through our senses, we briefly register and store visual images via iconic memory, in which picture images last no more than a few tenths of a second. We register and store sounds via echoic memory, where echoes of auditory stimuli may linger as long as 3 or 4 seconds.
Researcher George Sperling asked people to do something similar when he showed them, for only one-twentieth of a second, three rows of three letters each. After the nine letters disappeared, people could recall only about half of them.
Was it because they had insufficient time to glimpse them? No, Sperling cleverly demonstrated that people actually could see and recall all the letters, but only momentarily. Rather than ask them to recall all nine letters at once, Sperling sounded a high, medium, or low tone immediately after flashing the nine letters. This cue directed participants to report only the letters of the top, middle, or bottom row, respectively. Now they rarely missed a letter, showing that all nine letters were momentarily available for recall.
Sperling’s experiment revealed that we have a fleeting photographic memory called iconic memory. For a few tenths of a second, our eyes register an exact representation of a scene and we can recall any part of it in amazing detail. But if Sperling delayed the tone signal by more than half a second, the image faded and participants again recalled only about half the letters. Our visual screen clears quickly, as new images are superimposed over old ones.
We also have an impeccable, though fleeting, memory for auditory stimuli, called echoic memory. Picture yourself in conversation, as your attention veers to the TV. If your mildly irked companion tests your attention by asking, “What did I just say?” you can recover the last few words from your mind’s echo chamber. Auditory echoes tend to linger for 3 or 4 seconds. Experiments on echoic and iconic memory have helped us understand the initial recording of sensory information in the memory system.
At any given time, we can focus on and process only about seven items of information (either new or retrieved from our memory store). Without rehearsal, information disappears from short-term memory within seconds. Our capacity for storing information permanently in long-term memory is essentially unlimited.
Our short-term memory span for information just presented is limited—a seconds-long retention of up to about seven items, depending on the information and how it is presented. Our capacity for storing information permanently in long-term memory is essentially unlimited.
It was discovered the nervous substrate of long-term memory is mostly cerebral cortex. The most important regions are temporal lobes, prefrontal area and hippocampus. Experimental researches revealed that some thalamic nuclei and reticular formation take part in memory function.
Reticular formation gives ascending stimulatory influences to cerebral cortex, which help in keeping awake condition of cortex and provides voluntary attention.
At the molecular level, the habitation effect in the sensory terminal results from progressive closure of calcium channels through the presynaptic terminal membrane.
In case of facilitation, the molecular mechanism is believed to be following. Facilitated synapse releases serotonin that activates adenylyl cyclase in postsynaptic cell. Then cyclic AMP activates proteinkinase that then causes phosphorylation of proteins. This blocks potassium channels for minutes or even weeks. Lack of potassium causes prolonged action potential in the presynaptic terminal that leads to activation of calcium pores, allowing tremendous quantities of calcium ions to enter the sensory terminal. This causes greatly increased transmitter release, thereby markedly facilitating synaptic transmission.
Thus in a very indirect way, the associative effect of stimulation the facilitator neuron at the same time that the sensory neuron is stimulated causes prolonged increase in excitatory sensitivity of the sensory terminal, and this establishes the memory trace.
Eric Kandel showed initially that weaker stimuli give rise to a form of short term memory, which lasts from minutes to hours. The mechanism for this "short term memory" is that particular ion channels are affected in such a manner that more calcium ions will enter the nerve terminal.
This leads to an increased amount of transmitter release at the synapse, and thereby to an amplification of the reflex. This change is due to a phosphorylation of certain ion channel proteins, that is utilizing the molecular mechanism described by Paul Greengard.
A more powerful and long lasting stimulus will result in a form of long term memory that can remain for weeks. The stronger stimulus will give rise to increased levels of the messenger molecule cAMP and thereby protein kinase A. These signals will reach the cell nucleus and cause a change in a number of proteins in the synapse. The formation of certain proteins will increase, while others will decrease. The final result is that the shape of the synapse can increase and thereby create a long lasting increase of synaptic function.
In contrast to short term memory, long term memory requires that new proteins are formed. If this synthesis of new proteins is prevented, the long term memory will be blocked but not the short term memory.
Storing Memories in the Brain.
The search for the physical basis of memory has recently focused on the synapses and their neurotransmitters; on the long-term potentiation of brain circuits, such as those running through the hippocampus; and on the effects of stress hormones on memory. Studies of people with brain damage reveal that we have two types of memory operating together—explicit (declarative) memories processed by the hippocampus, and implicit (procedural) memories processed by the cerebellum and the amygdala.
Researchers are exploring memory-related changes within and between single neurons. Long-term potentiation (LTP) appears to be the neural basis of learning and memory. Stress triggers hormonal changes that arouse brain areas and can produce indelible memories. We are particularly likely to remember vivid events that form flashbulb memories. We have two memory systems. Explicit (declarative) memories of general knowledge, facts, and experiences are processed by the hippocampus. Implicit (nondeclarative) memories of skills and conditioned responses are processed by other parts of the brain, including the cerebellum.
Researchers interested in the biology of the mind have also looked closely at the influence of emotions and stress hormones on memory. When we are excited or stressed, emotion-triggered stress hormones make more glucose energy available to fuel brain activity, signaling the brain that something important has happened. Moreover, the amygdala, two emotion-processing clusters in the limbic system, boosts activity and available proteins in the brain’s memory-forming areas. The result? Arousal can sear certain events into the brain, while disrupting memory for neutral events around the same time.
“Stronger emotional experiences make for stronger, more reliable memories,” says James McGaugh. After traumatic experiences—a wartime ambush, a house fire, a rape—vivid recollections of the horrific event may intrude again and again. It is as if they were burned in. This makes adaptive sense. Memory serves to predict the future and to alert us to potential dangers.
Conversely, weaker emotion means weaker memories. People given a drug that
blocks the effects of stress hormones will later have more trouble remembering the details of an upsetting story. That connection is appreciated by those working to develop drugs that, when taken after a traumatic experience, might blunt intrusive memories. In one experiment, victims of car accidents, rapes, and other traumas received either one such drug, propranolol, or a placebo for 10 days following their horrific event. When tested three months later, half the placebo group but none of the drug-treated group showed signs of stress disorder.
Retrieval: Getting Information Out.
Recall is the ability to retrieve information not in conscious awareness; a fill-in-the-blank question tests recall. Recognition is the ability to identify items previously learned; a multiplechoice question tests recognition. Relearning is the ability to master previously stored information more quickly than you originally learned it. Retrieval cues catch our attention and tweak our web of associations, helping to move target information into conscious awareness. Priming is the process of activating associations (often unconsciously).
To remember an event requires more than getting it in (encoding) and retaining it (storage). To most people, memory is recall, the ability to retrieve information not in conscious awareness. To a psychologist, memory is any sign that something learned has been retained. So recognizing or more quickly relearning information also indicates memory.
To be remembered, information must be encoded, stored, and then retrieved. Memory is recall, recognition, and relearning. With the aid of associations (cues) that prime the memory, we retrieve the information we want to remember. Cues sometimes come from returning to the original context. We use our senses as cues-a taste, smell, or sight may evoke us to recall a memory. Mood affects memory, too. While in a good or bad mood, we tend to retrieve memories congruent with that mood.
The context in which we originally experienced an event or encoded a thought can flood our memories with retrieval cues, leading us to the target memory. In a different but similar context, such cues may trick us into retrieving a memory, a feeling known as déjà vu. Specific emotions can prime us to retrieve memories consistent with that state. Mood-congruent memory, for example, primes us to interpret others’ behavior in ways consistent with our current emotions.
The process of retrieving a memory follows a similar principle, because memories are held in storage by a web of associations, each piece of information interconnected with others. When you encode into memory a target piece of information, such as the name of the person sitting next to you in class, you associate with it other bits of information about your surroundings, mood, seating position, and so on. These bits can serve as retrieval cues, anchor points you can use to access the target information when you want to retrieve it later. The more retrieval cues you have, the better your chances of finding a route to the suspended memory.
Often our associations are activated, or primed, without our awareness. Seeing or hearing the word rabbit primes associations with hare, even though we may not recall having seen or heard rabbit. Priming is often “memoryless memory”—invisible memory without explicit remembering. If, walking down a hallway, you see a poster of a missing child, you will then unconsciously be primed to interpret an ambiguous adult-child interaction as a possible kidnapping. Although you don’t consciously remember the poster, it predisposes your interpretation. Meeting someone who reminds us of someone we’ve previously met can awaken our associated feelings about that earlier person, which may transfer into the new context.
Putting yourself back in the context where you experienced something can prime your memory retrieval. Associated words, events, and contexts are not the only retrieval cues. Events in the past may have aroused a specific emotion that later primes us to recall its associated events. Sometimes, being in a context similar to one we’ve been in before may trigger the experience of déjà vu (French for “already seen”). Two-thirds of us have experienced this fleeting, eerie sense that “I’ve been in this exact situation before,” but it happens most commonly to well-educated, imaginative young adults, especially when tired or stressed. Our mood states provide an example of memory’s state dependence. Emotions that accompany good or bad events become retrieval cues. Thus, our memories are somewhat mood-congruent. Our mood’s effect on retrieval helps explain why our moods persist. When happy, we recall happy events and therefore see the world as a happy place, which helps prolong our good mood. When depressed, we recall sad events, which darkens our interpretations of current events. For those of us with a predisposition to depression, this process can help maintain a vicious, dark cycle.
We may fail to encode information for entry into our memory system. Memories may fade after storage—rapidly at first, and then leveling off, a trend known as the forgetting curve. We may experience retrieval failure, when old and new material compete, when we don’t have adequate retrieval cues, or possibly, in rare instances, because of motivated forgetting, or repression. In proactive interference, something learned in the past interferes with our ability to recall something recently learned. In retroactive interference, something recently learned interferes with something learned in the past.
This controversy between memory researchers and some wellmeaning therapists is related to whether most memories of early childhood abuse are repressed and can be recovered by means of leading questions and/or hypnosis during therapy. Psychologists now tend to agree that: (1) Abuse happens, and can leave lasting scars. (2) Some innocent people have been falsely convicted of abuse that never happened, and some true abusers have used the controversy over recovered memories to avoid punishment. (3) Forgetting isolated past events, good or bad, is an ordinary part of life. (4) Recovering good and bad memories, triggered by some memory cue, is commonplace. (5) Infantile amnesia—the inability to recall memories from the first three years of life—makes recovery of very early childhood memories unlikely. (6) Memories obtained under the influence of hypnosis or drugs or therapy are unreliable. (7) Both real and false memories cause stress and suffering.
Three sins of forgetting:
• Absent-mindedness—inattention to details leads to encoding failure (our mind is elsewhere as we lay down the car keys).
• Transience—storage decay over time (after we part ways with former classmates, unused information fades).
• Blocking—inaccessibility of stored information (seeing an actor in an old movie, we feel the name on the tip of our tongue but experience retrieval failure—we cannot get it out).
Three sins of distortion:
• Misattribution—confusing the source of information (putting words in someone else’s mouth or remembering a dream as an actual happening).
• Suggestibility—the lingering effects of misinformation (a leading question—“Did Mr. Jones touch your private parts?”—later becomes a young child’s false memory).
• Bias—belief-colored recollections (current feelings toward a friend may color our recalled initial feelings).
One sin of intrusion:
• Persistence—unwanted memories (being haunted by images of a sexual assault).
One explanation of forgetting is that we fail to encode information for entry into our memory system. Without effortful processing, we never notice or process much of what we sense. Age affects encoding efficiency, which explains age-related decline.
Much of what we sense we never notice, and what we fail to encode, we will never remember. Age can affect encoding efficiency. The brain areas that jump into action when young adults encode new information are less responsive in older adults. This slower encoding helps explain age-related memory decline. Without effort, many memories never form.
If children or adults are subtly exposed to misinformation after an event, or if they repeatedly imagine and rehearse an event that never occurred, they may incorporate misleading details into their memory of what actually happened. When we reassemble a memory during retrieval, we may successfully retrieve something we have heard, read or imagined, but attribute it to the wrong source (source amnesia). False memories feel like true memories and are equally durable. Constructed memories are usually limited to the gist of the event.
Memories may also fade after storage—often rapidly at first, and then leveling off. This is the basis for one of psychology’s laws, the forgetting curve. Even after encoding something well, we sometimes later forget it. Cognitive neuroscientists are getting closer to solving the mystery of the physical storage of memory and are increasing our understanding of how memory storage could decay. But memories fade for other reasons, including the accumulation of learning that disrupts our retrieval.
Forgetting also results from retrieval failure. Retrieval-related forgetting may be caused by a lack of retrieval cues, by proactive or retroactive interference, or even, said Freud, by motivated forgetting.
We have seen that forgotten events are like books you can’t find in your campus library—some because they were never acquired (not encoded), others because they were discarded (stored memories decay). But there is a third possibility: The book may be there but inaccessible because we don’t have enough information to look it up and retrieve it. How frustrating when we know information is “in there,” but we cannot get it out.
Learning some items may interfere with retrieving others, especially when the items are similar. If someone gives you a phone number, you may be able to recall it later. But if two more people give you their numbers, each successive number will be more difficult to recall. Likewise, if you buy a new combination lock, your memory of the old one may interfere. Such proactive (forward-acting) interference occurs when something you learned earlier disrupts your recall of something you experience later. As you collect more and more information, your mental attic never fills, but it certainly gets cluttered.
Retroactive (backward-acting) interference occurs when new information makes it harder to recall something you learned earlier.
Misinformation and Imagination Effects.
Memories are not stored as exact copies, and they certainly are not retrieved as such. Rather, we construct our memories, using both stored and new information. Thus, when child or adult eyewitnesses are subtly exposed to misinformation after an event, they often believe they saw the misleading details as part of the event.
In many follow-up experiments around the world, people have witnessed an event, received or not received misleading information about it, and then taken a memory test. The repeated result is a misinformation effect: After exposure to subtle misinformation, many people misremember.
Sigmund Freud might have argued that our memory systems self-censored this information. He proposed that we repress painful memories to protect our selfconcept and to minimize anxiety. But the submerged memory will linger, he believed, to be retrieved by some later cue or during therapy.
People also exhibit source amnesia, by attributing something heard, read, or imagined to a wrong source. Because false memories feel like true memories and are equally durable, sincerity need not signify reality. Thus, we may recognize someone but have no idea where we have seen the person. We may dream an event and later be unsure whether it really happened. We may hear something and later recall seeing it. In all these cases of source amnesia (also called source misattribution), we retain the memory of the event, but not of the context in which we acquired it.
Discerning True and False Memories.
Determining the validity of a memory is difficult. A false memory may feel real and it may be persistent. Interviewers may ask leading questions, contributing to the misinformation effect. True memories tend to be more detailed than imagined ones, which tend to be the gist of the meaning and feelings associated with an event. Because memory is reconstruction as well as reproduction, we can’t be sure whether a memory is real by how real it feels. Much as perceptual illusions may seem like real perceptions, unreal memories feel like real memories. False memories created by suggested misinformation and misattributed sources may feel as real as true memories and may be very persistent.
The psychology of memory suggests concrete strategies for improving memory. These include spaced study; active rehearsal; encoding of well-organized, vivid, meaningful associations; mnemonic techniques; returning to contexts and moods that are rich with associations; recording memories before misinformation can corrupt them; minimizing interference; and self-testing and rehearsal.
Research on memory suggests concrete strategies for improving memory. These include studying repeatedly, making material personally meaningful, activating retrieval cues, using mnemonic devices, minimizing interference, getting adequate sleep, and self-testing.