Although one can point to forerunners, it was not until the 1960s and 1970s that psychology became dominated by the study of errors and biases in cognition, and began to believe that our perception and memory of social reality was for the most part inaccurate. As Jussim (Reference Jussim2012) points out, at the heart of this approach is the idea that humans are cognitively lazy, lacking alertness and irrational in their relating to the social world – in short “cognitive misers” (p. 4). This development coincided with the metaphor of mind as a computer that processes, encodes, stores, and retrieves information; when compared to the standard of decontextualized and literal processing of information on a computer hard disk, the human mind was a poor performer. This focus on identifying error and distortion obscured important, non-mechanical dimensions of mind and the study of accuracy as more than what a computer does with information. Jussim (Reference Jussim2012) analyzes this trend in relation to research on our beliefs and expectations of others, while Gigerenzer et al. (Reference Gigerenzer, Swijtink, Porter, Daston, Beatty and Kruger1990) have shown how it played out in the psychology of thinking. In this commentary, I describe a third area of research, memory, that has also shifted towards a one-sided focus on error. Memory is peripherally discussed in Jussim's book in relation to stereotypes. Here I extend Jussim's treatment to show how memory research began to overemphasis conditions leading to distortion over accuracy.
Like person perception and thinking, memory has become mainly a study of inaccuracy and distortion. Memory errors have been generally described in relation to two processes: expectation-based and post-event misinformation effects. The most well-known concept for analyzing expectation-based errors is that of “schema” and its derivatives “scripts” and “frames” (Wagoner Reference Wagoner2013). Although the concept had been used much earlier (e.g., Bartlett Reference Bartlett1932; discussed further below), schema took on a new meaning in the 1970s as an abstract knowledge structure for representing and storing information – for example, our generic idea of going to a restaurant shapes how we perceive and remember any specific situation of doing so. Studies using this concept showed that we are more likely to remember schema-consistent than schema-inconsistent information, and tend to add information that was schema-consistent but not actually present in the original.
While early studies tended to look at memory for banal situations – such as going to a restaurant (Shank & Abelson Reference Shank and Abelson1977) or a graduate student's room (Brewer & Treyens Reference Brewer and Treyens1981) – it was not long before schema was applied to describe the use and maintenance of stereotypes; for example, in relation to race, gender and social groups. In this context, schemas were guides to action that become self-fulfilling prophecies (Fiske Reference Fiske, Clark and Fiske1982). There has been much evidence put forward to show that schema-based bias is real, but this research has largely sidelined the issues of how schema are developed and changed on the basis of social reality (McVee et al. Reference McVee, Dunsmore and Gavelek2005). Furthermore, Bransford et al. (Reference Bransford, McCarrell, Franks, Nitsch, Shaw and Bransford1977) pointed out that schema set the stage for making visible expectation inconsistent information, an issue that has been largely ignored.
In addition to schema, memory distortion and inaccuracy has also been heavily investigated in terms of “the misinformation effect,” initially demonstrated in a series of pioneering studies by (Loftus Reference Loftus1975; Loftus & Palmer Reference Loftus and Palmer1974; Loftus et al. Reference Loftus, Miller and Burns1978). The experimental set up begins by showing participants complex, fast-moving events (such as an automobile accident or theft) through film clips or slide shows. Immediately afterward, participants are questioned about the event. For the experimental group some of the questions asked include misleading information – for example, “How fast was the white sports car going when it passed the barn while traveling along the country road?” (There was no barn in the film). These participants were more likely to later remember a barn than the control group in which no barn was mentioned in the post-event questioning. Important to note is that the effect was much stronger when a barn was presupposed in the question rather than when participants were asked directly “Did you see a barn in the film?” (Loftus Reference Loftus1975). It is also worth pointing out that participants were fed entirely plausible suggestions by a trusted authority and were only inaccurate a fraction of the time. More recently, it has been found that the misinformation can be neutralized if participants are given reason to mistrust the authority, and may even be reversed if told afterwards that they were given some misinformation (Blank Reference Blank1998). This suggests that humans are not the memory dupes they have often been made out to be, but can actively manage suggestions in remembering (Wagoner & Gillespie Reference Wagoner and Gillespie2014).
Looking back to Frederic Bartlett's book Remembering: A Study in Experimental and Social Psychology, commonly evoked as the locus classicus of memory distortion research, we find it tells a rather different story (Wagoner Reference Wagoner2017). For Bartlett “constructive” remembering meant that it was flexible and adaptable in meeting new challenges: What is of primary interest is what we can do with our memories. In this way, construction was theorized together with conservation and retention of the past. He also believed that it could lead to accuracy in memory, whereas today it has become a synonym for “distortion” (Wagoner Reference Wagoner2015). Bartlett (Reference Bartlett1932) in fact gave several examples of accuracy in recall, including remembering the “gist” of some material, the retention of dominant as well as unimportant details, “almost word perfect reminiscence” (p. 203) functional in stable environments, and “prodigious retentive capacity” of Swazi herdsman in relation to cattle transactions (see also Ost & Costall Reference Ost and Costall2002). The last example is noteworthy in that it illustrates Bartlett's (Reference Bartlett1932) belief that memory is not good or bad in itself but rather that accuracy depends on the context and whether it promotes literal recall. As such, we find in Bartlett (Reference Bartlett1932) a more balanced position towards accuracy and inaccuracy than is typically attributed to him (cf. Allport's warnings about the overemphasis on inaccuracy in perception – Jussim Reference Jussim2012, p. 19ff).
Reflecting on the emerging emphasis on memory inaccuracy in the late 1960s and its relation to his own book, Bartlett commented, “I did not say, I think I did not imply that literal retrieval is impossible, but I did imply that it requires special constricting conditions” (Bartlett Reference Bartlett1968, Note 3), such as in the context of learning by heart in school, witnesses in a court of law, or even in a memory experiment. Moreover, when these “constricting conditions” are not present, memory is still accurate enough to provide an appropriate response to a changing environment. His famous concept of schema (mentioned above as a structure in the head) was actually used by him to describe the well-adapted transaction between person and world (Wagoner Reference Wagoner2013). He also pointed out that human beings were not merely determined by schema but actively manage and reflect on them in order to master and enjoy the world. Thus, we see what was for Bartlett the best means of knowing and acting on the world, became the source of distortion and inaccuracy in later research on memory from the 1970s onwards. In summary, like the research on person-perception described by Jussim, memory research became a study of errors, bias and inaccuracy, downplaying the evidence for possibilities of reflection, accuracy and its different manifestations.
Although one can point to forerunners, it was not until the 1960s and 1970s that psychology became dominated by the study of errors and biases in cognition, and began to believe that our perception and memory of social reality was for the most part inaccurate. As Jussim (Reference Jussim2012) points out, at the heart of this approach is the idea that humans are cognitively lazy, lacking alertness and irrational in their relating to the social world – in short “cognitive misers” (p. 4). This development coincided with the metaphor of mind as a computer that processes, encodes, stores, and retrieves information; when compared to the standard of decontextualized and literal processing of information on a computer hard disk, the human mind was a poor performer. This focus on identifying error and distortion obscured important, non-mechanical dimensions of mind and the study of accuracy as more than what a computer does with information. Jussim (Reference Jussim2012) analyzes this trend in relation to research on our beliefs and expectations of others, while Gigerenzer et al. (Reference Gigerenzer, Swijtink, Porter, Daston, Beatty and Kruger1990) have shown how it played out in the psychology of thinking. In this commentary, I describe a third area of research, memory, that has also shifted towards a one-sided focus on error. Memory is peripherally discussed in Jussim's book in relation to stereotypes. Here I extend Jussim's treatment to show how memory research began to overemphasis conditions leading to distortion over accuracy.
Like person perception and thinking, memory has become mainly a study of inaccuracy and distortion. Memory errors have been generally described in relation to two processes: expectation-based and post-event misinformation effects. The most well-known concept for analyzing expectation-based errors is that of “schema” and its derivatives “scripts” and “frames” (Wagoner Reference Wagoner2013). Although the concept had been used much earlier (e.g., Bartlett Reference Bartlett1932; discussed further below), schema took on a new meaning in the 1970s as an abstract knowledge structure for representing and storing information – for example, our generic idea of going to a restaurant shapes how we perceive and remember any specific situation of doing so. Studies using this concept showed that we are more likely to remember schema-consistent than schema-inconsistent information, and tend to add information that was schema-consistent but not actually present in the original.
While early studies tended to look at memory for banal situations – such as going to a restaurant (Shank & Abelson Reference Shank and Abelson1977) or a graduate student's room (Brewer & Treyens Reference Brewer and Treyens1981) – it was not long before schema was applied to describe the use and maintenance of stereotypes; for example, in relation to race, gender and social groups. In this context, schemas were guides to action that become self-fulfilling prophecies (Fiske Reference Fiske, Clark and Fiske1982). There has been much evidence put forward to show that schema-based bias is real, but this research has largely sidelined the issues of how schema are developed and changed on the basis of social reality (McVee et al. Reference McVee, Dunsmore and Gavelek2005). Furthermore, Bransford et al. (Reference Bransford, McCarrell, Franks, Nitsch, Shaw and Bransford1977) pointed out that schema set the stage for making visible expectation inconsistent information, an issue that has been largely ignored.
In addition to schema, memory distortion and inaccuracy has also been heavily investigated in terms of “the misinformation effect,” initially demonstrated in a series of pioneering studies by (Loftus Reference Loftus1975; Loftus & Palmer Reference Loftus and Palmer1974; Loftus et al. Reference Loftus, Miller and Burns1978). The experimental set up begins by showing participants complex, fast-moving events (such as an automobile accident or theft) through film clips or slide shows. Immediately afterward, participants are questioned about the event. For the experimental group some of the questions asked include misleading information – for example, “How fast was the white sports car going when it passed the barn while traveling along the country road?” (There was no barn in the film). These participants were more likely to later remember a barn than the control group in which no barn was mentioned in the post-event questioning. Important to note is that the effect was much stronger when a barn was presupposed in the question rather than when participants were asked directly “Did you see a barn in the film?” (Loftus Reference Loftus1975). It is also worth pointing out that participants were fed entirely plausible suggestions by a trusted authority and were only inaccurate a fraction of the time. More recently, it has been found that the misinformation can be neutralized if participants are given reason to mistrust the authority, and may even be reversed if told afterwards that they were given some misinformation (Blank Reference Blank1998). This suggests that humans are not the memory dupes they have often been made out to be, but can actively manage suggestions in remembering (Wagoner & Gillespie Reference Wagoner and Gillespie2014).
Looking back to Frederic Bartlett's book Remembering: A Study in Experimental and Social Psychology, commonly evoked as the locus classicus of memory distortion research, we find it tells a rather different story (Wagoner Reference Wagoner2017). For Bartlett “constructive” remembering meant that it was flexible and adaptable in meeting new challenges: What is of primary interest is what we can do with our memories. In this way, construction was theorized together with conservation and retention of the past. He also believed that it could lead to accuracy in memory, whereas today it has become a synonym for “distortion” (Wagoner Reference Wagoner2015). Bartlett (Reference Bartlett1932) in fact gave several examples of accuracy in recall, including remembering the “gist” of some material, the retention of dominant as well as unimportant details, “almost word perfect reminiscence” (p. 203) functional in stable environments, and “prodigious retentive capacity” of Swazi herdsman in relation to cattle transactions (see also Ost & Costall Reference Ost and Costall2002). The last example is noteworthy in that it illustrates Bartlett's (Reference Bartlett1932) belief that memory is not good or bad in itself but rather that accuracy depends on the context and whether it promotes literal recall. As such, we find in Bartlett (Reference Bartlett1932) a more balanced position towards accuracy and inaccuracy than is typically attributed to him (cf. Allport's warnings about the overemphasis on inaccuracy in perception – Jussim Reference Jussim2012, p. 19ff).
Reflecting on the emerging emphasis on memory inaccuracy in the late 1960s and its relation to his own book, Bartlett commented, “I did not say, I think I did not imply that literal retrieval is impossible, but I did imply that it requires special constricting conditions” (Bartlett Reference Bartlett1968, Note 3), such as in the context of learning by heart in school, witnesses in a court of law, or even in a memory experiment. Moreover, when these “constricting conditions” are not present, memory is still accurate enough to provide an appropriate response to a changing environment. His famous concept of schema (mentioned above as a structure in the head) was actually used by him to describe the well-adapted transaction between person and world (Wagoner Reference Wagoner2013). He also pointed out that human beings were not merely determined by schema but actively manage and reflect on them in order to master and enjoy the world. Thus, we see what was for Bartlett the best means of knowing and acting on the world, became the source of distortion and inaccuracy in later research on memory from the 1970s onwards. In summary, like the research on person-perception described by Jussim, memory research became a study of errors, bias and inaccuracy, downplaying the evidence for possibilities of reflection, accuracy and its different manifestations.