Chapter 7 Chapter 5 Attention and Memory

"You're not paying attention," said Hayter, "you know, if you don't pay attention, you won't get anywhere." —Adapted from Lewis Carroll Everyone understands the general meaning of the phrase "you didn't pay attention".It could be that you are not concentrating, it could be that you are drowsy, or for some other reason.Psychology distinguishes "arousal" (or alertness) from "attention".Arousal is a general condition that affects a person's whole behavior and is noticed when you wake up in the morning, as William James said, and to a psychologist, noticing means "getting rid of something in order to deal with other things more efficiently".

Our main concern is visual attention, not attention while listening to music or engaging in some kind of activity.We know that attention is thought to contribute to at least some forms of consciousness.One form of visual attention is eye movement (often aided by head movement), and since we see better near the center of our gaze, we gain more attention when our eyes are on an object. Information.Otherwise, if we don't look directly at the object, we can only get rough information (at least about the shape). What mechanism controls eye movement?Such eye movements range from those triggered by a reflex response (such as the eye suddenly jumping somewhere outside the center of gaze) to those triggered by will ("I want to know what he's doing there").All forms of attention may have both reflexive and volitional components.

An example of auditory selective attention is having a subject focus on a sound entering one ear from headphones while trying to ignore a different sound entering the other ear.Many sounds from the non-attentive ear do not reach the level of consciousness, but can leave certain traces in the mind and have an effect on what the attentive ear hears.They are registered at a certain processing level in the brain. Attention, therefore, is about filtering out unnoticed events.The response to the noted event has faster speed, lower threshold and higher precision, and attention can also make the event easy to be remembered.In the past, psychologists didn't care much about what's going on inside our heads; they mostly studied attention by measuring things like reaction speed and error levels.In other words, they study the consequences of paying attention to an event (compared with not paying attention to it) and attempt to infer possible mechanisms of attention from the pattern of experimental outcomes.

Surprisingly, some things just can't be done when your eyes are kept still.For example, a random pattern of dots blinking rapidly on the screen, its presentation time is very short, so it is impossible to generate eye movement.Under these conditions, can you tell the number of random points?If there are only three or four of them, you can correctly report their number; but if there are six or seven or more, you will get an error.This cannot be attributed solely to the brightness of the stimulus.If the flickering dots are very bright, they will leave an afterimage on the retina (the pattern of dots fixed on the retina will move with your eye if you move your eye).For a few seconds, you can see them all the time, but you still can't count them exactly--it's a very strange feeling.When you start counting, you forget which dot you already counted.

Is there some form of attention that doesn't depend on eye movement?Can attention shift between two large eye movements?Michael Posner, a clinical psychologist at the University of Orleans in the United States, has conducted a large number of experiments on this.He and other researchers have shown that such a form of visual attention does exist.In a typical experiment, subjects hold their eyes still by fixating on a particular point.A momentary signal alerts the subject that an object may appear at a certain location (say, to the right of the point of fixation).When seeing the object appear, the subjects are asked to press the switch as soon as possible, and their reaction time will be recorded.If, during a given experiment, the object did not appear where expected (eg, to the left of the fixation point), the response was slower.The delay in reaction time was interpreted as the subject having to switch visual attention from the desired side to the undesired side.According to Posner, this change in attention may involve the following three sequential processes:

Release the original attention -> move the attention point -> implement the attention First, the system needs to disengage attention from what it is paying attention to in the field of view.Then the "attention" point must be turned to the new location, and finally the attention is implemented at the new location.Another important question is whether a person can attend to two separate locations or objects in the visual field at the same time?Evidence suggests that this cannot be done①, although it may be possible to track several ③moving points.But there is solid evidence that attention can be spatially fine-focused or extended over a large scale.For example: When you read a book, you mainly pay attention to the words rather than individual letters.This is not the case when proofreading, where you have to check every letter and punctuation carefully, otherwise small mistakes will be missed.For me personally, proofreading is a difficult thing.Since I usually read very fast, it is difficult to spot small typographical errors unless I pay attention.

Clearly, attention changes the way we see objects.How do theorists explain this phenomenon?I can say bluntly that there is no generally accepted theory of attention.So the best I can do is to describe some of the currently popular, roughly speaking, general consensus that attention involves a bottleneck.The basic idea is that the primary processing process is largely a parallel process, that is, many different activities are carried out simultaneously.Then, there seems to be a bottleneck of information processing at one or more stages.Only one (or few) "objects" can be processed at a time.It does so by temporarily filtering out information from non-attentive objects.Then, the attention system quickly moves on to the next object.As a result, attention is largely serial (i.e., attention to one followed by another) rather than highly parallel (as is the case when a system pays attention to many things simultaneously). ① Later, we will discuss in detail the important differences between parallel and serial processing.

Visual attention is often compared to "searchlights".Inside Searchlight, information is processed in a special way.In this way, we can observe the object of attention quickly and precisely, and make it easier for us to remember it.Information outside the "searchlight" is either processed less, differently, or not at all.The brain's attentional system quickly shifts the imaginary "searchlight" from one part of the field of view to another, just as we move our eyes, only this time at a much slower rate. The searchlight metaphor suggests to us in the simplest way that the visual system pays attention to a place in the field of view.Much circumstantial evidence suggests that this is the case.Another view is that what we pay attention to is not a particular place but a particular object.In some cases, if the object moves (the eyes remain still), attention can track the object instead of staying in one place.It now appears that to some extent both forms of attention (attention to visual objects or attention to visual locations) may co-occur.

Psychologists generally make a strict distinction between preattentive and attentive processing.A striking illustration of some preattentive processing has been given by the Hungarian psychologist Bela Julz, who has worked for many years in the United States.Please see Figure 20.The border between the two "textures" on the left is immediately visible.Now let's look at the right half of the image: at first glance there are no obvious texture boundaries, but a closer look reveals that one area is made up of the letter L in different orientations, and the other is made up of the letter T.But this difference does not immediately pop out (pop-out).To see it requires focal attention.

There's another way to study bounces (or lack thereof).Presents a visual image on the screen for a brief period of time.In this case, the stimulus image often consists of a "target" that the subject is asked to detect and other slightly different objects (called "distractors").For example, there may be a large number of letters scattered on the image, and all but one letter are red, and all others are green.The subjects' task was to press the button as soon as they saw the red letter.We found that the subjects could complete this task very quickly.More importantly, reaction times were not related to having only a few green letters or many green letters.In other words, no matter how many distractors there are, the reaction time is the same.The red letters immediately jumped out in front of my eyes.

Anne Treisman is one of the influential psychologists who studies attention. In 1977, she teamed up with two colleagues to perform a famous experiment.The gist of the experiment is this.She first demonstrated that red letters could jump out on a background of green letters.A single letter T can pop out of the background of the letter S if all the letters are the same color.This means that, for both color and shape, bounce can occur.Then, they gave the subjects a more complex task.One half is a green letter T and the other half is a red letter S. In addition, there is a red letter T.The subject's task is to find the red letter T. At this time, the subject can neither find a single red letter nor a single letter T; because there are too many letters that meet these two conditions.Participants had to look for letters that combined both color (red) and shape (T).And this kind of combination cannot jump out immediately. It takes a while to find the red letter T, and the more the number of interference items, the longer the time required.If there are 25 letters in the pattern, it takes much longer to find a single red letter T than when there are only 5 letters. ① This situation is seen as evidence of a serial search mechanism, that is, in order to determine whether a letter is both red and T-shaped, the attention system can only look at one letter at a time. Notice how much time it takes to move from one place to another?This is a more complicated matter.It seems that the more "prominent" the object (has a greater impact on the attention system), the less time it takes.This situation is possible.For example, if the red letters are very bright, the vision system can detect several letters at a time by extending the "searchlight" to a larger area.This means that all letters can be searched in fewer steps.Therefore, the processing time per letter is reduced.Some people think that it is possible to process an object at a time of about 60 milliseconds.If processing two objects at a time still takes 60 milliseconds per step, each letter (which would have been observing only one letter at a time) now takes only 30 milliseconds to process.And if three objects can be processed at the same time, the processing time for each letter is 20 milliseconds. But there are more complex situations.Perhaps the subjects' brains had been trained to be smarter to only pay attention to red letters (and ignore green ones).This will cause the upper half of the letters to be ignored.This means that he can complete the search faster with the same attentional pace.In this case, a pace of 120 milliseconds yielded the same observations. We also encounter regrettable situations.In some cases, the step time may appear to be less than 20 milliseconds, while the true pace may be as long as 120 milliseconds.This is because before discovering the red T letter, the subject not only only paid attention to the red object, but also processed three letters in batches, thus "deceiving" us.In this case, it is difficult to determine the correct time for the searchlight to move one step. Triisman also explained that jumping out can also be asymmetrical.A gapped circle can pop out in the background of a group of intact circles (Fig. 21a); ​​however, finding a complete circle in the background of gapped circles requires a serial search (Fig. 21b). How do psychologists describe the difference between pre-attentional processing and attentional processing?Initially, Triisman believed that pre-attentional processing is to register simple features such as the orientation, movement, and color of objects in the field of vision in some unique subsystems in a parallel manner.Then, focus on bringing these features together in some way.More careful experiments led her to discover that if the time allowed for feature integration was very short, the brain would err.Sometimes it plays around and falsely integrates features together to give a false composition.In lectures, Triisman uses a quick slide to demonstrate the phenomenon.This slide shows a woman in red with dark hair.However, there were always a few in the audience who said with great confidence that what they saw was a redhead.The color of the girl's sweater was mistakenly "grafted" into her hair, creating an illusion combination. This kind of thing may happen in daily life, but it is only a small number.Triisman gives an example: "A friend walking down a busy street saw a colleague and was about to say hello. Then he suddenly realized that the man's black beard had grown onto a passerby's face, and his Baldness and glasses belong to someone else." What exactly a "simple feature" is, we don't know yet. ① But unfortunately, a lot of research shows that jumping out is not a straightforward matter.It is not my intention here to describe the details of numerous such experiments. In general, many of Triisman's models of attention assume that jumping out is distinct from the sequential search of a longer process.But other psychologists, such as Kyle Cave and Jeremy Wolfe, argue that jumping out is just the first step in the attention process.They hypothesize that attention systems are somehow "noisy" and thus prone to errors.If the object "stands out" enough, move the searchlight of attention to where the object is or to the object as the first step of attention.If the object is not prominent, the system may have difficulty selecting the target.It may take a lot of time before the target is finally found.This mechanism can produce similar results to a simple sequential search mechanism. Duncan (J. Duncan) and Humphreys (G. Humphreys) even denied the existence of searchlights. They believed that different objects in the field of vision all tried to reach short-term memory.If successful, in some cases they become the focal point of the campaign.Their hierarchical model also takes into account the relationship between different distractors.For example, are the distractors the same or are of different types. Further research may lead psychologists to a generally accepted model of attention, though it may not be simple.I suspect that the correct model will be unlikely to be derived from psychological experiments alone, because the system seems too complex. ① Knowledge of the behavior of certain relevant neurons in the brain may be necessary to obtain the correct answer. Thus, we only partially understand visual attention.We don't yet have a universally accepted model of attention in psychology. What about short-term memory?How much do we know about it?Memory may be defined as a change within a system induced by experience that leads to a change in thought or behavior at a later time.But such generalizations don't carry much value.It should apply to conditions such as fatigue, injury, and intoxication, without making a strict distinction between learning and development (early growth).The Israeli neurobiologist Yadin Dudai has come up with a more useful and precise definition.He begins by describing what is an "inner expression" of a "world" (both internal and external).He defines the internal representation of the world as "a structured neural code that effectively guides behavior".It emphasizes that, fundamentally, our main concern is how nerve cells (neurons) affect behavior. "Learning" is the innovation or modification of internal expression caused by experience.This change can be maintained for quite some time (even many years).However, we are more concerned with short-term memory. I'm not interested in such simple forms of memory as habituation or sensitization (if you show a child a picture ten times in a row, at first he'll show interest, but soon get bored. This is "habituation").These processes are classified as "non-associative" processes.They are even shown in some very low animals like sea urchins.We are more concerned with "associative learning", the organism's response to the relationship between stimulus and action. (2) It is instructive to classify memory into several distinct types, although their exact descriptions are debated.A convenient classification divides memory into episodic memory, categorical memory, and procedural memory.A clear memory is the memory of an event, often intertwined with certain extraneous details connected to it. (1) A good example is that you will remember where you were when you heard that President Kennedy was assassinated.An example of categorical memory is the meaning of a word, such as "murder" or "dog".Recalling how to swim or drive a car is procedural memory. Another way of categorizing has to do with time: how long it takes to acquire a memory, how long it generally lasts, and certain memories, especially episodic memories, are called "one-shot" or "flash-shutter" learning.Just one instance can be remembered clearly. (Of course, this kind of memory can also be strengthened by retelling. That is, telling the event again without requiring it to be correct every time.) Another type of memory can be strengthened by repetition of events.From repetition, people extract general properties of something, such as the meaning of a word that is not clearly defined. Procedural knowledge such as driving a car is often difficult to gain from a single experience and often requires repeated practice.It can stay put for quite a long time.Once you learn to swim, you can swim well even if you haven't swam in years.A famous pianist once said to me, when it comes to a familiar piece of music, "Muscle memory is longest." This means that the playing of the piece is automatic, without thinking. Different memories last for different amounts of time.They are often divided into long-term memory and short-term memory.Although the term may mean different things to different people. "Long time" usually refers to hours, days, months or even years; "short time" is from a fraction of a second to a few minutes or longer.Short-term memory is usually unstable and has a limited capacity. Think about something in your dream.When you're dreaming, you can't bring any of the scenes into your long-term memory (or at least recall them vividly).Your brain stores dream scenes in some form of short-term memory.Your long-term memory system isn't switched on until after you wake up (which can be much more often than you realize).Then, what is still held in short-term memory goes into long-term memory.So you don't recall everything you dreamed, but only the last few minutes of the dream.If the phone rings or some other disturbance occurs when you first wake up, the short-term memory of the dream will decay or be completely lost, so that you may not be able to recall the last few minutes of the dream after the phone call. We know that the recall of memory is not a straightforward process.To recall an event often requires a certain clue, although the memory may be confusing at this time.Some memories are weak and require stronger cues to recall.Others fade even before being completely lost.A related memory may interfere and block access to the content of the memory you need. It is clear that consciousness, especially visual consciousness, combines much of what is stored in long-term episodic and categorical memory.We are more concerned with very short-term memory.This is due to the fact that we would probably lose consciousness if we lost all forms of memory of recent events, yet this most important form of memory lasts only a fraction of a second or at most a few seconds.Let's focus on these extremely short-term forms of memory. Please take a look at the scene in front of you, and then suddenly close your eyes.The vivid images you see of the outside world are quickly lost.All that is left of you is a vague memory.It usually disappears within seconds, and attempts have been made as far back as the 18th century to measure how long it takes for it to disappear.A point of light moving in the dark (say a glowing cigarette butt) will leave a trail of light behind it.Modern studies of the length of the light tail suggest that light perception lasts for approximately 100 milliseconds, although some is due to retinal afterimages. How do psychologists study short-term memory of all kinds?American psychologist George Sperlig conducted a classic experiment in 1960.He displays an alphabet set of twelve letters on the screen in a very short time (about 50 milliseconds).The letters are arranged in three rows of four.Due to the short time, the subjects could only recall four or five letters each time.Then in the next experiment, he asked the subjects to report only one of the lines, and he used an audio signal to prompt the subjects which line to report.But this clue is only given right after the rendered graph is closed.In this case, the subject could report about three of the four letters in the cue line. One might conclude from the second experiment alone that since the subject was able to report three of the four letters in any of the three lines, he was able to report nine of the three lines (three times three).But as we have seen, he can actually recall only four or five of the twelve letters.This strongly explains that letters are read out by the brain from rapidly decaying visual traces. This kind of extremely short-term visual memory is called "icon memory", which comes from the word icon, which means icon. There are many other studies on this issue.Whether the field of view is bright or dark before and after stimulus presentation has an effect on the decay time.In the dark field, the decay time is on the order of seconds, while in the brighter field it is much less, perhaps only a fraction of a second.This bright background effect is called "masking".Some patterns can also be used as masking, but these two types of masking are quite different. In short, the masking of bright backgrounds may occur before the information of the eyes is combined, the primary stage of the visual system, and may be at the retinal stage; and Pattern masking relies heavily on the time interval between letter presentation and masking.The data suggest that this presumably occurs at several levels of the visual system after binocular information is combined. Icon memory appears to depend on the retention time of transient visual cues.It is mainly not counted from the trailing edge of the signal but from the leading edge.This suggests that its biological function is to provide sufficient time (approximately 100-200 milliseconds) to process this very transient signal.This means that at least some minimum time is required for adequate visual processing to occur. There is also longer short-term memory.British psychologist Alan Badileley has studied this type of memory in depth, calling it "working memory," and a classic example is recalling a new seven-digit phone number.The number of digits you can recall is called your "number span".For most people, it's usually only six or seven.In other words, the capacity of working memory is limited.This memory appears to take several different forms, and it is related to sensory input.For vision, he calls it the "Visuospatial Scratch Pad".Typically the time involved is several seconds.It also seems to be related to visual imagery when recalling faces or familiar objects.Its characteristics are quite different from the shorter icon memory.Icon memory may involve different processes in the brain. Is working memory necessary for consciousness?There is some evidence that this is not the case.Some brain-injured patients have very small numerical memory spans, unable to recall anything but the last letter they heard, yet remain conscious.In fact, their long-term memory may not have been damaged.To date, no patient has been identified who has lost all forms of working memory (visual and auditory).This is due to the fact that the brain damage that caused this deficiency (and not any other defect) can only be localized in a very precise location (and in a different place), so in practice this may never happen . Long-term memory appears to be distinct from pictorial memory or working memory.A subject who saw about 2,500 different colored slides (10 seconds each) could still distinguish 90% of them after ten days.Because, if the subject is only asked to confirm whether he has seen a certain picture before (not to recall without clues, which will be more difficult), then he only needs to recall a small part of the information of each picture. We do not expend much energy thinking about long-term episodic memory, because a brain-damaged patient who cannot form new long-term episodic memories remains awake and conscious (see Chapter 12).Only short-term memory, especially icon memory, may be closely related to the mechanism of consciousness. ①Here is a piece of evidence for reference. If the corpus callosum is removed, each half of the brain can pay attention to different objects. ③However, it is possible for the brain to perceive these moving points as the corners of a single object that is changing shape.point of view, and mention some of the main points of contention. ①After practice, the brain can track a certain group of special objects (such as a group of letters) as a "chunk". ① Response times vary greatly from one experiment to another.Therefore, to replicate the experimental results, it is necessary to have the subjects respond multiple times and average the response times.In some cases, it is necessary to apply several subjects and calculate their average response time. ①There is experimental evidence that this can happen. ① Someone proposed a research project to explore what visual features can jump out (they should respond to simple features, visual "primitives"); and compound features need to be searched sequentially. ② There are other simple forms of memory not mentioned here, including classical conditioning, operant conditioning, and priming. ①There is evidence that, during the initial period, many people vividly remember the scene when they first heard about Lincoln's assassination.