Crowhurst Donald Donald Crowhurst (1932–1969) was an English yachtsman. The famous lone navigator Sir Francis Chichester called the story of Crowhurst the maritime drama of the century. Donald Crowhurst, engineer, inventor, businessman, decided to take part in the round-the-world race

15. Donald Knuth

15. Donald Knuth Of all the characters in this book, Donald Knuth perhaps needs the least introduction. Over the past 40 years, he has been writing his multi-volume masterpiece, “The Art of Programming,” a bible of fundamental algorithms and data structures. American magazine

Interest in the selective function of attention is reflected both in classical psychology of consciousness (W. James) and in modern cognitive psychology (D. Broadbent, A. Treisman, D. Deutsch and E. Deutsch).

W. James' theory of attention

The American philosopher and psychologist William James characterized consciousness with four main properties: individuality, continuity, variability and selectivity. It was with the last of specified properties he connected the phenomenon of attention: “Being under a constant influx of more and more new impressions penetrating into the area of ​​our senses, we notice only the most insignificant fraction of them. Conscious experience is a stream flowing through a wide meadow.” James believed that it is impossible to continuously focus attention on a constant object of thought. A single act of attention lasts no more than a few seconds in a row. After this, attention is either distracted or directed to other aspects of the same object. This situation can be easily illustrated with the help of two-digit figures (see Fig. 2.2 “Wife or Mother-in-law” in subparagraph 2.2.2). If you just look at such an image, the Wife and Mother-in-law begin to “wink.” Attention on one of the images can be steadily maintained only when you begin to “develop” the object of attention in a certain way, for example, you begin to count the wrinkles on the Mother-in-Law’s neck or the Wife’s eyelashes. It was with the ability to constantly “develop” the object of one’s attention, to find new aspects in it, that James associated the phenomenon of genius. In his opinion, a genius is able to achieve outstanding success in any field because he never gets bored with his activity, he perceives it differently all the time, becoming more and more involved in it. The philosopher Arthur Schopenhauer also pointed out the special intensity of the attention of outstanding people: “Talent hits the targets that simple people they can't get in. And genius hits targets that ordinary people cannot see."

James also proposed one of the most multifaceted classifications of attention (Table 6.1). He identified six types of attention according to the criteria of arbitrariness, focus (external or inner world subject) and the method of connecting the act of attention with the present motivational state (directly or through an associative connection). Note that voluntary attention is always indirect. Involuntary attention, on the contrary, can be either indirect or direct. So, a man rushing to work suddenly freezes at a store window, seeing on the cover of one of the magazines a photograph of a girl who looks like a classmate with whom he was once in love. In this case, his act of attention can be classified as sensory involuntary mediated attention. And when an institute graduate mentally thinks through the order of a conversation with a potential employer, on whose success his future career prospects depend, we are talking about intellectual voluntary mediated attention.

Table 6.1

Classification of types of attention according to V. James

Attention as a filter. Theories of early and late selection

In contrast to the theory of D. Kahneman, where the psyche “struggles” with the limitations of the information processing system, developing the most effective strategy for distributing the attention resource, there is a group of theories that postulate the presence of a special mechanism that saves the psyche from overload. This mechanism is attention. Here attention is seen as filter , which blocks or weakens the processing of redundant information (D. Broadbent, A. Treisman, D. and E. Deutsch). From this perspective, the central topic of consideration becomes the topic of describing the properties of such a filter and its location in the system of information processing operations. This approach to attention can be called structural.

Early selection theories

The fact that attention acts as a filter that cuts off information that is unnecessary at the moment was noticed by W. James: “We know that we can be attentive to the voice of the interlocutor among the hubbub of other conversations that we do not notice, although they are objectively much louder than that speech , to which we listen..." Subsequently, this observation was called the "party effect." Interestingly, a party guest not only easily copes with the task of maintaining a conversation that interests him, without being distracted by other auditory, visual and tactile stimulation, but is also able to switch to a “new channel” of information if he is called by name or neighbors start talking at the table. discuss a topic that is important to him.

Acoustic engineer Colin Cherry (1953) was primarily interested in the basis for the selection and retention of one message in a stream of others. He suggested the following as possible parameters: spatial localization of the sound source, frequency and pitch characteristics of sound, syntactic and content properties of the message. To simulate a party situation and test his hypotheses, Cherry created techniques selective hearing, which are still actively used by researchers to this day. The essence of these techniques is that subjects are presented with two texts through headphones. The text can be presented in dichotic mode (one text is sent to one earphone, and another text is sent to the second) or in binaural mode (both texts are sent to each earphone simultaneously). The subject is asked to attend to one of the messages (relevant channel), saying it out loud, and not pay attention to the other (irrelevant channel). Cherry found that when both texts were read binaurally at the same volume by the same speaker, it took subjects about 25 repetitions to pick out the relevant message. In the case when the messages were presented dichotically, the subjects easily coped with the task on the first attempts. Thus, the important role of the spatial factor in tracking and retaining relevant (i.e., the one to which attention is drawn) information was shown.

In another experiment, Cherry tried to find out what a subject perceives from an irrelevant message (i.e., on the basis of what cues he is able to switch from one message to another). While the subject echoed the relevant message, presented dichotically, the irrelevant message was constantly subject to changes: the male voice changed to a female one, the recording began to scroll in the opposite direction, the announcer switched to a different language, the content of the story changed, and a sharp audio tone sounded. After listening, the subject was asked if he noticed anything unusual in the irrelevant message. The subjects noticed only the change in voice and sound tone. Then the thesis was formulated that the separation of a relevant message from an irrelevant one occurs on the basis of an analysis of rough physical characteristics (direction of the sound source, volume, pitch, timbre of the voice), and properties of a higher level (perceptual and semantic) do not play any role here . Cherry's research formed the basis of early selection models, which postulated that the filter was located at the entrance to the information processing system and had very strict settings.

The first actual psychological model of this type was proposed by D. Broadbent (1958). While generally agreeing with the description of a filter as a device that rigidly cuts out irrelevant information based on the analysis of sensory cues, he provided an explanation of how the filter settings are formed. After all, if the filter were not adjusted in a certain way before receiving information, we would be forced to perceive only one type of stimuli throughout our lives, for example, only female voices or only sounds from sources on the right. Broadbent suggested that the filter does not turn on immediately, but only after there is a threat of overloading the information processing unit of limited capacity. This block is located after the primary sensory analysis system and is intended for perceptual processing of incoming information. The perceptual processing unit can simultaneously “let in” no more than six pieces of information. As soon as it is filled, a filter is turned on, which “lets in” new pieces of information similar to what is already contained in it. This happens as perceptually processed information moves deeper into the system. The perceptual processing unit in Broadbent's model is structurally and functionally similar to the short-term memory subsystem, but has a shorter information storage time (see Chapter 8).

Broadbent illustrated his hypothesis using a series of experiments on “split” memory capacity. In the first series, subjects were given three sets of numbers dichogically with an interval of 0.5 seconds. Immediately after the end of the presentation, they were asked to name the numbers they heard. When reporting freely, the subjects, firstly, coped well with the task and, secondly, always reproduced the numbers channel by channel, i.e. first they named those numbers that were fed to one earphone, and then those that were fed to the other. If they were asked to name the numbers in the order they were received, the percentage of correct answers dropped sharply (to 20%). Broadbendt concluded that switching from channel to drap is possible in principle, but takes time (about 1/3 of a second). The decrease in reproduction efficiency, in his opinion, is due to the fact that information is erased during the report.

In the second series, subjects were dichotically presented with a series of six digits via the relevant channel. A series of two digits was transmitted through an irrelevant channel. Moreover, the numbers could be submitted simultaneously with the beginning or end of the series. Subjects were then asked to reproduce the numbers they heard. It turned out that if the presentation of a pair of digits through an irrelevant channel coincided with the beginning of a relevant series, the subjects reproduced it only in 25% of cases. When it coincided with the end of the series, the subjects were able to reproduce them in 50% of cases. Let's consider what, according to Broadbent, happens in the first and second cases. If the pair of digits that is fed to the irrelevant earphone coincides with the beginning of the series, the subject first reads the pair from the relevant channel, then switches to the irrelevant channel, and then returns to the relevant one again. At this point, the entire limited filter volume is already filled. Consequently, only that information will be further perceived that is consistent (in this case, spatially) with the latter. This will be information from a relevant channel. The irrelevant pair of digits will be erased to make room for the last relevant pair. The result is low reproduction success. Now let's look at the situation where the irrelevant pair appears at the end. Despite the fact that the filter is already full in this case, the system behaves as if it “knows” that the stored information is not in danger. The report starts with an irrelevant bunch of numbers, and of course this pair is reproduced correctly.

Thus, Broadbent considered his original model to be justified: at first, information from both the relevant and irrelevant channels enters the perceptual processing unit, but when the amount of information reaches six units, the irrelevant channel is blocked and the information that arrived through it remains available for a very short time time period (no more than 1/3 s).

Attenuator model A. Treisman

The idea of ​​a filter's rigid properties was soon softened considerably. In an experiment by N. Morey (1959), it turned out that the subjects really did not respond to the commands “Stop!” or “Switch to the other ear!” when they were presented on an irrelevant channel. But if the commands were preceded by addressing the subject by name (“John Smith, stop!” or “John Smith, switch to the other ear!”), the subjects noticed them and carried them out.

American psychologist Anna Treisman (1964) determined that the condition for the intrusion of a message from an irrelevant channel into a relevant message is a semantic connection with the content of the relevant channel, therefore, instead of the hard filter model, she proposed the model attenuator (from the French attenuer - soften, weaken). In one of her experiments, subjects were presented with a short delay via an irrelevant channel a text similar in content to the relevant message, but in a different language known to the subject. The subjects quickly noted the fact that the messages were identical (“But it’s the same thing!”). In another experiment, a text like: “The guests were sitting at three possibilities, waiting for dinner," while an irrelevant channel broadcast text like: "Let us take a look at these dining table explanations of the fact of interest." As a result, the study participants voiced a "mixed" text: "The guests were sitting at the dining table, waiting for dinner." Interestingly, the subjects themselves did not notice the phenomenon of "crossing" messages; it seemed to them that they continued to echo only the relevant channel.

Thus, the data suggested that irrelevant information could be analyzed at the level of meaning. The attenuator model proposed by Treisman assumes that a filter located at the entrance to the system does not completely block the irrelevant flow of information, but only weakens it. In addition to the attenuator, this model includes a system dictionary, largely identical to long-term memory, and vocabulary units – concepts that are elements of the dictionary. The main process on which the fate of information falling on sensory surfaces depends is the activation of the corresponding vocabulary unit, i.e. “meeting” of a stimulus with its psychological referent, necessary for its further processing.

Let us explain how the process of information processing is represented in the Treisman model. After the flow of stimulation overcomes the attenuator, some messages remain unchanged (relevant channel), while others are attenuated (irrelevant channel). In order to undergo further processing, the content arriving through the channel must “meet” its corresponding vocabulary unit (category). A signal traveling along an unattenuated relevant channel will most likely “meet” its semantic unit and undergo further processing. Signals traveling through irrelevant channels and therefore attenuated are less likely to reach their corresponding vocabulary items.

However, things are not always that simple. Vocabulary units have different activation thresholds, i.e. can be activated by a signal of varying intensity. One can clearly imagine the situation in such a way that different units are at different “distances” from the attenuator. For example, categories such as the subject's name, as well as things related to his profession or hobby, are located “close” to the attenuator, and words associated with a traumatic experience are located “far.” Therefore, for vocabulary units that have a minimum activation threshold, a signal attenuated by an attenuator is sufficient. In addition, Treisman introduces the concept contextual lowering of the activation threshold: units that are close in value to those units that are already activated seem to “move closer” to the attenuator and therefore become easier to activate. It is with the effect of contextual lowering of the threshold that Treisman associates the effect of crossover of contents described above. The mechanism of operation of the information processing system in A. Treisman’s model is shown in Fig. 6.3.

Rice. 6.3.

Thus, the attenuator model represents attention as a filter with special properties, which is located close to the input to the information processing system and limits its flow in order to avoid overloading the system.

Theories of late selection: models of D. and E. Deutsch and D. Normann

In 1963, D. and E. Deutsch questioned the existence of an early selection mechanism. But in their opinion, the limitations in the information processing system lie not at the input, but at the output from the system, namely at the stage of awareness, decision-making and response. All units of the dictionary that have been exposed to the influence are activated, but due to the difference in the “reactivity” of the units themselves (with what strength the unit responds to the influence) and the heterogeneity of the influence, only a few stimuli always turn out to be “winners”. They are strengthened by the exit filter and gain access to consciousness.

The main arguments in favor of the late selection model boil down to the following: despite the fact that information coming through an irrelevant channel is not consciously understood by a person, it influences the interpretation of information that is supplied through a relevant channel. Thus, D. McKay (1973) described the effect of “semantic guidance”. The subjects were presented with an ambiguous text via a relevant channel. From time to time, among other words, a key word – a hint – was presented through an irrelevant channel. If this experiment were carried out in Russian, the relevant message would look, for example, like this: “That day he was not in the best shape and therefore lost his glasses.” Alternative keywords then could have been "eyes" and "volleyball". Subjects from both groups were not aware of the content keywords. However, those who had the word “eyes” included in an irrelevant message were more likely to understand the main text as a story about a nearsighted person, and those who had the word “volleyball” in an irrelevant message were more likely to understand the main text as a description of a sports match.

P. Forster and E. Gower (1978) showed that the effect of an irrelevant rope also manifests itself at the level of physiological reactions. At the preliminary stage of the experiment, the subjects developed a conditioned reaction to a certain word (listening to it was accompanied by an unpleasant electric shock). Subjects were then asked to repeat the neutral text. Occasionally, either the target word itself or its synonyms were inserted into an irrelevant message. In both cases, the subjects responded to the appearance of these words with a surge of galvanic skin response, although they did not notice it. It turned out that the person’s consciousness did not seem to “hear” the target word, but his body reacted quite unambiguously.

D. Norman in 1968 supplemented the Deutsch model with the help of the “relevance block”. In Norman's model, the activation of vocabulary units precedes the actual act of perception. The system thus anticipates what will be perceived. The "appropriateness block" determines the level of coincidence between stimulation and the subject's expectations and reinforces adequate information. Norman's model, as is easy to see, inherits the idea of ​​preperception by W. James and the concept of the effect of volitional attention. N. Lange (see subparagraph 6.2.3). By introducing the "appropriateness block," Norman first provides a rational explanation for the problem of illusory perception, where too much activation of the output filter allows the system to make an erroneous decision based on insufficient sensory information.

Thus, the main differences between the early and late selection models are as follows.

• 1. Early selection: selection based on physical characteristics and inhibition of an irrelevant channel at the input to the system.
• 2. Late selection: selection based on significance and relevance for the subject and strengthening of the relevant channel at the exit to consciousness.

The theory of flexible and multiple selection by A. Treisman and the concept of an information processing system as a conglomerate of automatic and controlled processing processes by R. Shifrin

Proposed by A. Treisman a little later (1969) theory of flexible and multiple selection assumes that the information processing system has not one, but many filters. In the absence of a specific information processing task, filters are at rest. The solution to the problem occurs through sequential processing of information at various stages, such as the stage sensory signs, stage perceptual signs and stage semantic signs. Subsequently, another higher stage of information processing was proposed - the stage self-reference , i.e. analyzing information regarding its personal meaning (see Chapter 8). Which filters will be turned on at each stage of processing depends on the conditions of the task and the state of the subject. For example, if a subject needs to find “red A” among a jumble of letters on a screen, at the perceptual stage of processing only a color filter should be turned on, which eliminates all objects of other colors. At the same time, potentially available filters of size, orientation, sound, etc. do not participate in the information selection process. Then you should activate the filter, which will filter out the letters “not A”, leaving only the desired object. An experiment measuring reaction time showed that the time spent solving problems of this type depends linearly on the number of parameters by which an object must be identified.

An important achievement of the model of multiple and flexible selection was the departure of its author from the interpretation of the filter as a rigid mechanism for “eliminating” unnecessary information and the transition to understanding the diversity of selection strategies. Ronald Shiffrin took the next step on this path. In 1988, he proposed another option for “reconciling” the approaches of early and late selection. Shiffrin developed a view of the human psyche as a set of automatic and controlled information processing processes (Fig. 6.4).

Automatic processes are extensive in nature. The human psyche is capable of taking into account the impact huge amount stimuli without bringing the results of their processing to consciousness. Indeed, we take certain postures, move in space, operate with so-called subthreshold stimuli, without being completely aware of it. Most of the information processed through automatic processes never becomes conscious. Automatic processes are evolutionarily more ancient, they are inherent in both humans and animals. When discussing automatic information processing processes, it is more convenient to talk about output filters or apply late selection models to them.

Controlled processes Information processing, on the contrary, is carried out intensively. The scope of controlled processing is extremely limited, but the controlled nature of the process will provide a significant advantage in processing quality. Controlled processes are associated with consciousness and attention in the proper sense of the word, so it is more convenient to describe such processes using the early selection model.

Rice. 6.4.

An interesting phenomenon that demonstrates the constant interchange of controlled and automatic processes is the “popping out” effect (Fig. 6.5).

Rice. 6.5. Illustration of the "pop out" effect (find the horizontal line in the picture)

It is shown that if the target stimulus differs from the distractors (other stimuli) in only one parameter, its detection occurs automatically and does not depend on the number of distractors. Imagine that you need to spot a girl in a white blouse in a photo of many people wearing dark clothes. The feeling is that the desired image will literally “catch your eye.” If the target stimulus differs from the distractors in several respects (for example, you need to find a girl in a white blouse among men in white blouses and women in dark clothes), the psyche will be forced to resort to a controlled process of searching through objects. In this case, the detection time will increase in proportion to the number of distractors.

W. Johnston and S. Haynes proposed a more general model of attention selectivity, integrating models of flexible and multiple selection, R. Shifrin’s position on the separation of controlled and automatic processes and D. Kahneman’s resource model. In their opinion, the selection process requires an investment of resources, with early selection being a resource-saving strategy and late selection being an energy-intensive strategy. Therefore, the specific strategy that the psyche will choose to solve a problem depends both on the properties of the task itself and on the additional power available at the moment. If the task is simple and there is a lot of additional power (for example, a formal conversation with an unimportant person), the late selection strategy will most likely be implemented. If the task is complex and there is little additional power (for example, an exciting conversation that requires intellectual effort), the psyche will choose the strategy of early selection. In other words, in the first case you will actually perceive much more more information, than you realize, and in the second, you actually block extraneous information.

Donald Eric Broadbent(May 6, 1926 - April 10, 1993) was an influential British experimental psychologist, most famous for his work on attention. Broadbent helped nurture what was then the infant field of psychology in England, becoming famous worldwide for his groundbreaking theories and experimental work. His 1958 publication Perception and Communication was radical in its approach, taking the new field of information processing to model unobservable mental processes in a time when Behaviorism was dominant. His career and research work bridged the gap between the pre-Second World War approach of Sir Frederic Bartlett and its wartime development into applied psychology, and what from the late 1960s became known as cognitive psychology.

Broadbent's influence continues not just through his theories, which as he expected have been modified greatly through further research, but through his influence on numerous students and colleagues. His philosophy, scientific rigor, and good character impressed and inspired many to work towards the solution to real human problems even when they appear intractable. Broadbent joins the ranks of those who have contributed to better understanding of human nature.

Life

Donald Broadbent was born on May 6, 1926 in Birmingham, England. His family was quite well off financially. However, this changed when he was 13 and his parents divorced and his home moved to Wales. He won a scholarship to the prestigious Winchester College, an English independent school and completed his schooling there.

As a boy he was fascinated by flying, and at age 17 he volunteered to join the Royal Air Force (RAF). During his time in the RAF, he observed communication difficulties often arose from psychological, not physical, causes. In particular, he noticed that ineffective processes of attention, perception, and memory led to problems, rather than failures of technical equipment. An anecdote he often told to illustrate the importance of psychological processes in practice was counted by his long-time colleague, Dianne Berry:

The AT6 planes had two identical levers under the seat, one to pull up the flaps and one to pull up the wheels. Donald told of the monotonous regularity with which his colleagues would pull the wrong lever while taking off and crash land an expensive airplane in the middle of a field (Berry 2002).

Having made this observation, Broadbent's interests began to zero in on psychology, rather than his previous interest in the physical sciences. Psychology had the "concrete" quality of the physical sciences but it could also shed light on human problems.

Broadbent spent a short time after the war working in the personnel selection branch of the RAF before beginning his studies at Cambridge "s psychology department. Due to its natural sciences orientation and emphasis on practical application, Broadbent found Cambridge ideal. The department was headed by Sir Frederick Bartlett and was eager to apply newfound cybernetic ideals towards understanding human behavior, especially in terms of control systems, practical problems, and psychological theory in general. Broadbent found his place in the Applied Psychology Unit (APU) which had been set up there in 1944, by the UK Medical Research Council (MRC) on Bartlett's persuasion.

In 1958, Broadbent became director of the Unit, a position he held for 16 years. Although much of the work of the APU was directed at practical issues of military or industrial significance, Broadbent rapidly became well known for his theoretical work. His theories of selective attention and short-term memory were developed as digital computers were beginning to become available to the academic community, and were among the first to use computer analogies to make a serious contribution to the analysis of human cognition. His 1958 book, became one of the classic texts of cognitive psychology.

In 1974, Broadbent became a fellow of Wolfson College, Oxford University and returned to applied problems. There, together with his colleague Dianne Berry, he developed new ideas about implicit learning from consideration of human performance in complex industrial processes (Berry 2002). He continued this work until his retirement in 1991.

Donald Broadbent died on April 10, 1993.

Work

Donald Broadbent is best known for his contribution to the development of cognitive psychology. His 1958 book, Perception and Communication, has been rated "the single most influential book in the history of cognitive psychology" (Parasuraman 1996). Broadbent was the first person to bring together the work on information processing with the problem of attention, a radical move at a time when Behaviorism was the dominant paradigm in psychology. Broadbent used data from behavioral experiments and inferred (unobservable) functional stages of processing and their order of occurrence from these data. In so doing, he invented the modern study of attention (Berry 2002).

In all his work, Broadbent never abandoned practical problems. For example, he studied problems caused by communication with gunnery and air control systems, in which many channels of communication were delivered at one time. His work effectively bridged the gap between the laboratory and the field, constantly working on topics that had significance for people and society.

Broadbent contributed both experimental methods and theory to the world of psychology. His best known, and still widely used, method is the dichotic listening experiment, and his filter model of attention is his best known theory. Both were developed during his time at the Applied Psychology Unit at Cambridge University.

Dichotic listening experiments

Although most people spend their lives surrounded by many different types of stimuli, they cannot respond to or describe the majority of them. A practical example of this is found in the "cocktail party effect," described by Colin Cherry (1953) as the ability to focus one"s listening attention on a single talker among a mixture of conversations and background noises, ignoring other conversations. Cherry conducted experiments in which subjects were asked to listen to two different messages from a single loudspeaker at the same time and try to separate them, repeating one but not the other, known as a "shadowing" task. His work revealed that our ability to separate sounds from background noise is based on the characteristics of the sounds, such as the gender of the speaker, the direction from which the sound is coming, the pitch, or the speaking speed. When the messages were similar in these characteristics subjects were unable to complete the task successfully.

Broadbent extended this work by devising what is known as the "dichotic listening" experiment. In these studies, subjects were asked to listen to and separate different speech signals presented to each ear simultaneously (using headphones). For example, in one experimental setup, three pairs of different digits were presented simultaneously, three digits in one ear and three in the other. Most participants in the study recalled the digits ear by ear, rather than pair by pair. Thus, if 496 were presented to one ear and 852 to the other, the recall would be 496-852 rather than 48-95-62.

From the results of such experiments, Broadbent suggested that "our mind can be conceived as a radio receiving many channels at once." The brain separates incoming sound into channels based on physical characteristics (such as location).

Other experiments were concerned with the subject"s ability to answer one of two questions posed at the same time. Subjects with advance knowledge of which question they should attend to scored around 48 percent accuracy. Those informed after the questions had been given had almost no success:

The present case is an instance of selection in perception (attention). Since the visual cue to the correct voice is useless when it arrives towards the ends of the message, it is clear that the process of discarding part of the information contained in the mixed voices has already taken place…It seems possible that one of the two voices is selected for response without reference to its correctness, and that the other is ignored... If one of the two voices is selected (attended to) in the resulting mixture there is no guarantee that it will be the correct one, and both call signs cannot be perceived at once any more than both messages can be received and stored till a visual cue indicates the one to be answered (Broadbent 1952).

Filter model of attention

Broadbent developed his theory of selective attention based on his and other researchers" experimental findings using the information processing model. The major points of his filter theory can be summarized as follows:

• Stimuli presented at the same time are held in a short-term sensory buffer. Information can be retained there for a short period before being processed; after that it disappears from the processing system.
• A filter selects one of the inputs on the basis of its physical characteristics, passing it through a limited capacity channel for further processing.
• The input selected by the filter is analyzed for semantic content (meaning) and comes into conscious awareness.
• Any stimuli not selected by the filter do not receive this semantic analysis and never reach conscious awareness.

This theory provides an explanation of the "cocktail party" phenomenon, since the voice that a person is attending to has different physical characteristics from those of other people in the room. No semantic analysis is necessary to differentiate them. It also explains both Cherry"s and Broadbent"s experimental findings-unattended messages are rejected by the filter and thus receive very little processing.

Later findings, however, raised problems for this "all-or-nothing" filter model. In terms of the cocktail party, hearing one's name spoken by anyone in the room leads to a switching of attention to that speaker. This implies that the content of the message was analyzed prior to the filtering, which was supposed to occur before such analysis. This paradox did not deter Broadbent, and he accepted such data as reason to revise his theory (Craik and Baddeley 1995). His second book on the topic, Decision and Stress(1971) began with his filter model and was modified “to accommodate new findings that the model itself had stimulated” (Massaro 1996). This was typical of Broadbent's approach to scientific research—he regarded all theories as temporary accounts of the current information, likely to need revision and improvement when new data emerged.

Legacy

A lecture in Broadbent's honor is given every year at the annual conference of the British Psychological Society. Broadbent gave the inaugural lecture in 1991. After his death in 1993, tributes and biographical acknowledgments were written in his honor. A special issue of Applied Cognitive Psychology, edited by his long time colleague, Dianne Berry, was written to commemorate his contributions (Berry 1995).

Broadbent is credited with being a major force in the development of cognitive psychology, particularly the study of attention. His 1958 book, Perception and Communication, is a classic that continues to inform the area today.

Broadbent’s contributions to experimental psychology were noteworthy not only for research on attention, but because they also contributed to belief in the need for societal relevance in research—that is, practical application. He believed wholeheartedly that research should not be driven solely by theory but should be guided by important practical problems, and conversely that experimental results should be used to modify theories (Parasuraman 1996). In addition, his informal speaking style and use of commonplace analogies to represent complicated ideas made him memorable to society as a whole, allowing people of all walks of life access to his theories. As noted by Craik and Baddeley (1995), Broadbent's "psychology was intended for society and its problems, not merely for the dwellers in ivory towers."

His influence continues not just through his work but through the influence he had on numerous students and colleagues. He is remembered for the unmistakable image that he projected of himself, as “the man, the scholar, the scientist, the philosopher of science, and of his commitments to empirical psychology, to explicit models or theories, and to the application of psychological knowledge to real-word problems” (Massaro 1996). Unfailingly polite, helpful, and tolerant of the most naive questions posed by students, Broadbent was always approachable and generous with his time (Berry 2002). Yet he made a powerful impression on those who knew him, inspiring in them the conviction that good science would lead to solutions to real human problems.

Major works

• Broadbent, Donald E. 1952. Listening to one of two synchronous messages. Journal of Experimental Psychology 44: 51-55.
• Broadbent, Donald E. 1958. Perception and Communication. Elsevier Science Ltd. ISBN 0080090907.
• Broadbent, Donald E. 1961. Behavior. Basic Books. ISBN 0465005993.
• Broadbent, Donald E. 1962. Attention and the perception of speech Scientific American 206: 143-51.
• Broadbent, Donald E. 1971. Decision and Stress. Academic Press. ISBN 978-0121355500
• Broadbent, Donald E. 1973. In Defense of Empirical Psychology. Methuen young books. ISBN 041676780X.
• Broadbent, Donald E. 1993. The Simulation of Human Intelligence (Wolfson College Lectures). Blackwell. ISBN 0631185879.
• Broadbent, Donald E., and James T. Reason (eds.). 1990. Human Factors in Hazardous Situations. Oxford University Press. ISBN 019852191X.
• Pribram, Karl H., and Donald E. Broadbent (eds.). 1970. Biology of Memory. Academic Press, 1970. ISBN 0125643500.

References

• Baddeley, Alan, and Lawrence Weiskrantz (eds.). 1995. Attention: Selection, Awareness and Control. A Tribute to Donald Broadbent. Oxford University Press. ISBN 0198523742.
• Berry, Dianne C. (ed.). 1995. Special Issue: Donald Broadbent and Applied Cognitive Psychology. Applied Cognitive Psychology 9(7): S1-S215.
• Berry, Dianne. 2002. Donald Broadbent. The Psychologist(15)(8) (August 2002): 402-405. Retrieved October 20, 2008.
• Cherry, Colin E. 1953. Some experiments on the recognition of speech with one and two ears. Journal of the Acoustical Society of America 25: 975-979.
• Craik, Fergus I. M. 2000. Broadbent, Donald E. Encyclopaedia of Psychology 1: 476-477.
• Craik, Fergus I. M., and Alan Baddeley. 1995. Donald E. Broadbent (1926-1993). American Psychologist 50(4): 302-303.
• Hothersall, David. 2003. History of Psychology. McGraw-Hill. ISBN 0072849657.
• Massaro, D. W. 1996. Attention: Yesterday, Today and Tomorrow. 109(1): 139-150.
• Moray, N. 1995. Donald E. Broadbent: 1926-1993. American Journal of Psychology 108: 117-121.
• Parasuraman, Raja. 1996. Profiles in Psychology: Donald Broadbent. C S L Notes 20. Retrieved September 8, 2008.

Credits

New World Encyclopedia writers and editors revote and completed the Wikipedia article in accordance with New World Encyclopedia standards. This article abide by terms of the Creative Commons CC-by-sa 3.0 License (CC-by-sa), which may be used and disseminated with proper attribution. Credit is due under the terms of this license that can reference both the New World Encyclopedia contributors and the selfless volunteer contributors of the Wikimedia Foundation. To cite this article for a list of acceptable citing formats.The history of earlier contributions by wikipedians is accessible to researchers here:

• Donald_Broadbent history

The history of this article since it was imported to New World Encyclopedia:

Note: Some restrictions may apply to use of individual images which are separately licensed.