Researchers who study person memory examine how perceivers store and recall information about a social target in order to understand how that information is structured in the perceiver’s mind. Understanding the structure and organization of social information is important because it influences the way we perceive and process subsequently encountered targets, making it a crucial part of person perception.
The study of person memory assumes that we have concepts stored in memory. A concept (sometimes called a schema) is a mental representation of a category, where a category is defined as a group of items perceived to fit together (Kunda 1999). Concepts contain the set of features thought to define a category, as well as specific examples of category members. We have concepts for all sorts of objects, including types of people. An “athlete” concept, for example, might contain the notion that athletes have excellent physical skill and a competitive nature, as well as specific examples of athletes like Michael Jordon, Babe Ruth, etc.
The usefulness of concepts is twofold. First, a concept helps us identify new stimuli. When we see a furry, meowing animal, we can identify it as a cat, even if it is a cat we have not seen previously. Second, having applied the concept to the object, we can infer its characteristics to guide how we should interact with the object. Once an animal is identified as a cat, we may infer that it is friendly, is soft, and can be petted, even if we have not directly observed that it has these features. Notice that one can only make these inferences to the extent that the concept contains information about its attributes, and that those attributes are linked together in some fashion. The person-perception process relies on the concepts that are brought to mind when one encounters a person. If features of an interaction partner trigger the concept “university professor,” we may form quite a different impression than if the features trigger the concept “plumber.” This is because these different concepts will likely contain different attributes that inform the impression we create.
The Structure Of Concepts In Memory
Two broad classes of models have been developed to describe how social memories are organized: associative networks and connectionist networks.
Associative networks suggest that each concept in memory is represented by a node, and that nodes that are conceptually related have links between them. Whenever a concept is directly perceived, the corresponding node receives a certain amount of “activation.” When the activation in a node reaches a threshold, that concept becomes part of consciousness. In addition, activation spreads from the perceived concept to any other concepts to which it is linked. If enough activation passes through the links, concepts that are not present in the environment may become part of consciousness. Even if the activation is not enough to bring a linked concept to consciousness, the incoming activation brings the linked node closer to its threshold, making it more likely that the concept will enter consciousness in the near future.
Connectionist networks represent one of the newest ways of thinking about person memory. These networks are composed of nodes and links, just like associative networks, but represent information in a different way. Connectionist networks do not associate particular concepts with particular nodes. Instead, a concept is portrayed by a pattern of activation over the entire set of nodes. This type of pattern is known as a “distributed representation.” This is similar to the way that an image of an actor is portrayed by different levels of electricity in the pixels on a TV set. There is no “Clint Eastwood” pixel stored anywhere in your television. Instead, we recognize that Clint is on our screen by the particular pattern displayed across all the pixels.
One important advantage of distributed representations is that they use the same nodes to represent all possible concepts, so the model never has to build new nodes when it encounters novel objects. As in associative networks, memories in connectionist networks are represented by the links between the nodes. Most connectionist models incorporate rules for adjusting the links after presenting a concept to make the network more likely to instantiate that concept again in the future. This allows the model to recreate the full representation of a concept even when it later encounters a stimulus that generates a partial pattern across the nodes.
Inferring Concept Structure From Memory Measures
Although a single, definitive cognitive architecture of social information has not been identified, researchers have used a variety of measures to uncover what bits of knowledge are related in memory and the strengths of those relations. Below are examples of different types of memory measures and how they illuminate the structure of person memory.
Free Recall And Clustering
In a free-recall measure, participants are presented with information and later asked to write down or recite as much as they can remember. In a seminal work, Hastie and Kumar (1979) used free recall to examine the organization of individual behaviors in memory. They created a personal expectancy by presenting participants with a list of traits that described an individual (e.g., “intelligent”). Participants then were presented with behaviors supposedly performed by that same person. Some behaviors were consistent with the expectation (i.e., were intelligent actions), some were inconsistent, and some were unrelated. Later, participants recalled the behaviors. The proportion of behaviors recalled was greatest for the inconsistent behaviors, followed by consistent and unrelated behaviors.
Based on these and related findings, Srull and Wyer (1989) developed an elaborate model to explain how such social knowledge is organized in memory. This model suggests that when participants have a pre-existing trait expectation about an individual, perceivers will form an “evaluation-based” representation of the person, wherein the social target is represented by a person node that is linked to separate nodes for each behavior the target performed (Srull & Wyer 1989, 64). Links from the person node to the consistent behaviors should be stronger than links to the unrelated and inconsistent behaviors, making them easy to recall. This occurs because perceivers think about the consistent behaviors in relation to the person node to further solidify the original expectancy. However, links will also form between behavior nodes as behaviors are thought about in relation to one another. This will happen most frequently for the inconsistent behaviors, as the perceiver tries to reconcile their incongruence with expected behaviors. Therefore, the inconsistent behaviors will have more linkages emanating from them, improving their memorableness. These mechanisms allow the model to explain why both consistent and inconsistent behaviors should be better recalled than expectation-irrelevant behaviors, and why inconsistent behaviors will sometimes have a recall advantage over consistent ones.
Researchers have also investigated the sequence in which information is retrieved. Hamilton et al. (1980) presented participants with a list of behaviors. Those who read with an intent to form an impression showed more “clustering” in their recall than did those instructed to memorize the behaviors. That is, they showed a greater tendency to recall behaviors that shared the same trait implication at roughly the same time. To explain this finding, Srull and Wyer (1989) suggest that when initially forming an impression, participants encode behaviors in terms of what traits they exemplify. Doing so results in the formation of “trait–behavior clusters,” with a node representing the trait, linked to separate nodes representing each of the associated behaviors (Srull & Wyer 1989, 66).
Free recall tells us a great deal about the structure of social information, but relies on participants’ ability to consciously recollect past experience (e.g., that a person performed an intelligent behavior). However, past experiences may foster connections in memory between social constructs and attributes, even when the perceiver lacks the ability to recall those experiences and is unaware of those connections. Free-recall measures are often not sensitive enough to detect these types of “memories.” Fortunately, so-called “implicit measures” can. Below are examples of implicit measures and how they have influenced thoughts about person memory.
Many implicit measures are used to understand what attributes or evaluations are associated with social groups. As one example, Fazio et al. (1995) exposed participants to photographs of black and white faces on a computer screen. Immediately afterward, an adjective appeared and respondents categorized it as positive or negative. Participants’ response times for these decisions were analyzed. Compared to baseline response times to the words when presented without the photos, white respondents made decisions faster when positive adjectives followed white faces and when negative adjectives followed black faces. Black participants showed the opposite pattern. These findings suggest a linkage in memory between one’s own group and positive evaluations.
Similarly, the implicit association test (Greenwald et al. 1998) has been used to explore unconscious associations with social groups. In this task, participants use a single pair of keys to make distinctions between two types of stimulus. As an example, when using this procedure to examine implicit associations with blacks and whites, the first type of stimulus would be black and white names, whereas the second type would be positive and negative words. In one block, respondents press one key if they see either a black name or a negative word, and press a different key if they see either a white name or a positive word. In a separate block, these key mappings are reversed. White participants typically respond faster to the block pairing white names with positive evaluations than to the block pairing black names with positive evaluations, suggesting an association in memory between negative evaluations and black names.
Implications Of Concept Activation For Impression Formation
Person-memory models imply that the perception of a stimulus is determined jointly by information available in the environment and information available in memory. The extent to which memories of past experiences influence the impression of a newly encountered person or situation depends on several factors.
One factor receiving much empirical attention is accessibility. To examine the effect of concept accessibility on impressions, researchers expose perceivers to a concept (typically a trait) in one context and then ask them to form and report an impression of a social target presented in an unrelated context. Exposure to the trait concept in the initial context is said to “prime” (or make more accessible) that construct. Researchers consistently find that trait-priming influences impressions of social targets. For example, Bargh and Pietromonaco (1982) showed that participants subliminally primed with the trait “hostility” judged the ambiguous actions of another as more hostile than did those not primed.
- Bargh, J. A., & Pietromonaco, P. (1982). Automatic information processing and social perception: The influence of trait information presented outside of conscious awareness on impression formation. Journal of Personality and Social Psychology, 43, 437– 449.
- Fazio, R. H., Jackson, J. R., Dunton, B. C., & Williams, C. J. (1995). Variability in automatic activation as an unobtrusive measure of racial attitudes: A bona fide pipeline? Journal of Personality and Social Psychology, 69, 1013 –1027.
- Greenwald, A. G., McGhee, D. E., & Schwartz, J. L. K. (1998). Measuring individual differences in implicit cognition: The implicit association test. Journal of Personality and Social Psychology, 74, 1464 –1480.
- Hamilton, D. L., Katz, L. B., & Leirer, V. O. (1980). Organizational processes in impression formation. In R. Hastie, T. M. Ostrom, E. B. Ebbesen, R. S. Wyer, Jr., D. L. Hamilton, & D. E. Carlston (eds.), Person memory: The cognitive basis of social perception. Hillsdale, NJ: Lawrence Erlbaum, pp. 121–153.
- Hastie, R., & Kumar, P. A. (1979). Person memory: Personality traits as organizing principles in memory for behaviors. Journal of Personality and Social Psychology, 37, 25 –38.
- Kunda, Z. (1999). Social cognition: Making sense of people. Cambridge, MA: MIT Press.
- Srull, T. K., & Wyer, R. S., Jr. (1989). Person memory and judgment. Psychological Review, 96, 58 – 83.