Affect and judgments – Forgas - 19
Running head: AFFECTIVE INFLUENCE ON SOCIAL JUDGMENTS
Affective influences on social judgments and decisions:
Implicit and explicit processes
Joseph P. Forgas
University of New South Wales,
Sydney, Australia
Draft paper prepared for the Fifth Sydney Symposium of Social Psychology
This work was supported by a Special Investigator award from the Australian Research Council, and the Research Prize by the Alexander von Humboldt Foundation to Joseph P. Forgas. The contribution of Stephanie Moylan, Joseph Ciarrochi, Patrick Vargas and Joan Webb to this project is gratefully acknowledged. Please address all correspondence in connection with this paper to Joseph P. Forgas, at the School of Psychology, University of New South Wales, Sydney 2052, Australia; email . For further information on this research project, see also website at www.psy.unsw.edu.au/~joef/jforgas.htm .
Introduction
Imagine that as you are leaving a cinema having just seen a very funny film, you are approached on the street by a person collecting responses to a public opinion survey. You are asked to make a number of judgments indicating your evaluation of the current Prime Minister, the performance of the main political parties, your views on a various topical issues, your satisfaction with your life, and your expectations about the future. As you ponder your judgments, would your responses be influenced by the fact that you happen to be in a good mood after seeing the entertaining film? Would your judgments have been different if the film was a depressing war drama rather than happy? After some reflection, most people would admit to the unexpected possibility that temporary affect could indeed bias their judgments. And they would be right. We carried out just this experiment some years ago with almost 1000 moviegoers as respondents, and found a highly significant mood effect on the judgments people made (Forgas & Moylan, 1987). But do we know how, when and why affective states can impact on our social judgments and decisions? What are the implicit and explicit psychological mechanisms responsible for such affect infusion phenomena?
This paper will discuss recent evidence and theorising about the nature and causes of affect infusion into social judgments. A number of empirical studies illustrating these phenomena from our laboratory will be reviewed, with particular focus on the differences between implicit and explicit processes that can promote or inhibit affect infusion. We have recently defined affect infusion as the process whereby affectively loaded information exerts an influence on, and becomes incorporated into a person’s cognitive and behavioural processes, entering into their constructive deliberations and eventually coloring the outcome in a mood-congruent direction (Forgas, 1995a, in press). Affect infusion occurs because paying attention to, learning, encoding and interpreting social information necessarily requires high-level, constructive processing, that forces judges to go beyond the information given. Affect can play both an indirect, implicit influence on judgments by impacting on the way past knowledge is accessed and used when constructing a judgment (Forgas, in press), and a more direct, explicit influence when judges simply infer a response based on their current affective state (Clore, Gasper & Garvin, 2001).
Of course, the subtle influence of feelings on thinking and judgments has long been of interest to philosophers including Plato, Aristotle, Epicurus, Descartes, Pascal, Kant and others. Many of these theorists saw affect as a potentially dangerous, invasive force that tends to subvert rational judgment, an idea that was to re-emerge in Freud’s psychodynamic theories. However, during the last few decades important advances in neuroanatomy, psychophysiology and social cognition produced a radically different view. Rather than viewing affect as a dangerous and disruptive influence on our rational judgments, affect is increasingly seen as a useful and even essential component of adaptive responding to social situations (Adolphs & Damasio, 2001; Ito & Cacioppo, 2001). For example, individuals who suffer brain damage in the areas responsible for affective reactions in the prefrontal cortex were often found to make disastrous social decisions, even though their intellectual abilities remain unaffected (Adolphs & Damasio, 2001).
Although most of us are intuitively aware that feelings can have a profound influence on our thoughts and judgments, we do not yet fully understand how and why these influences occur. The challenge for researchers has been to understand and specify why affect sometimes influences judgments, yet at other times the effects are weak or absent, or even reversed (Martin, 2000). This paper will argue that the key to understanding this puzzle lies in the different processing strategies individuals employ when performing different kinds of social judgments. While some kinds of processing strategies allow affective states to exert an implicit or explicit influence judgments, other processing styles reduce or even reverse this effect. Before looking at contemporary research in detail, perhaps a brief overview of early research on affect and judgments should help to provide the necessary background.
Affect and judgments: the background
The importance of affect. Research on social judgments has traditionally emphasized the rational, logical processes involved in combining external stimulus information with internal knowledge structures in producing a response (Wyer & Srull, 1989). Just how important is affect as a component of, and as a determinant of judgments? In an influential article, Zajonc (1980) argued that affective reactions often constitute the primary response to social stimuli, a point he reiterated more recently (Zajonc, 2000). Supporting evidence comes from studies showing that people often display an affective preference towards stimuli long before they have a chance to properly process it, and even when they have no awareness of having encountered it before (Zajonc, 1980; 2000). Affect also plays a crucial role in implicit cognitive representations about common, recurring social experiences (Forgas, 1979). Feelings such anxiety, confidence, intimacy, pleasure or discomfort seem to determine how individuals represent various social episodes (Forgas, 1982). As Niedenthal and Halberstadt (2000) showed more recently, “stimuli can cohere as a category even when they have nothing in common other than the emotional responses they elicit” (p. 381). Similar conclusions were reached by Pervin (1976), who wrote that "what is striking is the extent to which situations are described in terms of affects (e.g. threatening, warm, interesting, dull, tense, calm, rejecting) and organized in terms of similarity of affects aroused by them" (p.471). Thus, affect seems to play a predominant role in cognitive representations and judgments about the social world.
The nature of social judgments. Why are social judgments so sensitive to affective influences? More than four decades ago Bruner (1957) argued that social judgments involve a constructive act of categorisation, where the expectations and states of the perceiver can play a major role. Many social judgments require high-level cognitive processes to infer characteristics that are not directly observable (Heider, 1958). Affect can play an important dual role in social judgments, by (a) influencing both the kind of information processing strategy adopted by judges (Forgas, 1995a, in press), and (b) influencing the sort of information people attend to, and how they interpret, learn, remember and evaluate this information (Bower, 1981; Fiedler, 1990; Forgas, 2001; Forgas & Bower, 1987). In essence, it is the constructive, inferential nature of social judgments that makes affective influences possible.
Constructivist vs. mechanistic approaches. The constructive nature of social judgments has been clearly recognized by theorists such as Heider (1958), Bruner (1957) and Asch (1946), who maintained that even the simplest kind of social stimuli may be subject to constructive biases as the perceiver seeks to categorize the information, and attempt to impose form or 'Gestalt' on a complex and often indeterminate stimuli (Asch, 1946). The tradition of constructivism in judgmental research was counterbalanced by a second, more atomistic and mechanistic approach. 'Cognitive algebra', a field pioneered by Anderson (1974) was based on the psychophysical measurement tradition and conceptualized social judgments as the predictable outcome of simple, arithmetically derived information integration processes. The individual states and constructions of the perceiver were of little interest within this paradigm. This approach assumed that social information has permanent, enduring meanings. However, in most social judgments the information is not 'given' but has to be selected or inferred, and its meaning is not constant but also subject to the constructions of the perceiver (Forgas, 1981). It seems then that the information integration approach and its metaphor of the social judge as a passive processor of stable information may at best be an incomplete account of social judgments.
The social cognition approach. The conflicting assumptions of the constructivist and the mechanistic views of social judgments were ultimately reconciled within the current social cognition paradigm. This approach focuses on the role of information processing strategies and memory processes involved in the translation of information into semantic representations, and the integration of past experiences with new information (Wyer & Srull, 1989). In line with Bruner's (1957) suggestions, it is the process of 'going beyond the information given' that makes judgments open to affect infusion. Principles of semantic and evaluative priming can play an important role in influencing information availability in judgments (Bower & Forgas, 2001; Stapel, this volume). However, the social cognitive approach to judgments has traditionally also assumed 'cold' cognition on the part of the perceiver, where feelings, emotions and preferences were relatively neglected (Forgas, 1981; 1983). The model's focus on the isolated and affect-less perceiver, separated from the social, cultural and emotional context in which judgments are usually made, has been a recurrent point of criticism (Argyle, 1991; Forgas, 1981).
Early evidence for affect infusion into judgments. Several early experiments suggested that affect does play a dynamic role in social judgments. Early accounts of such ‘affect infusion’ emphasized either psychodynamic, or conditioning principles. The psychoanalytic account suggested that affect has a dynamic, invasive quality and can ‘take over’ judgments unless adequate resources are deployed to control these impulses. Feshbach and Singer (1957) in an early study tested the psychoanalytic prediction that attempts to suppress affect should increase the ‘pressure’ for affect to infuse unrelated judgments. They induced fear in their subjects through electric shocks and then instructed some of them to suppress their fear. Fearful subjects were more likely to judge "another person as fearful and anxious" (p.286), and this effect was greater when judges were instructed to suppress their fear. Feshbach and Singer (1957) explained this in terms of the hydraulic principle: attempts to suppress affect increase the pressure for it to be ‘infused’ into unrelated judgments. Thus, "suppression of fear facilitates the tendency to project fear onto another social object" (p. 286).
An alternative account for affect infusion was offered by conditioning theories. Although radical behaviourists denied the value of studying internal constructs such as affect, Watson’s classic ‘little Albert’ studies were among the first to show that judgments of a previously neutral stimulus, such as a furry rabbit, could be influenced by associating fear-arousing stimuli, such as a loud noise, with this target. According to Watson, all our complex affective judgments throughout life are conditioned by such patterns of complex and cumulative associations. An early study by Razran (1940) seemed to support this view. Razran (1940) made people feel bad or good (by exposing them to aversive smells, or giving them a free lunch), and then asked them to make judgments about persuasive messages presented to them. Judges spontaneosly reported significantly more negative or positive attitudes towards persuasive messages depending on whether they made to feel bad or good previously.
The conditioning framework was later extended by Byrne and Clore (1970) and Clore and Byrne (1974) to study affective influences on interpersonal judgments. They argued that affective reactions to a neutral target (such as a previously unknown person) could be conditioned by associating them with positive or negative affect-eliciting situations. In other words, simple temporal and spatial contiguity will be enough to link an affective state to an incidentally encountered person. Several experiments demonstrated just such a conditioning effect, showing that people will be judged more positively when encountered in a pleasant situation, and be judged more negatively in aversive environments (Griffitt, 1970; Gouaux, 1971; Gouaux & Summers, 1973). More recently, Berkowitz and his colleagues (Berkowitz, Jaffee, Jo & Troccoli, 2000) also proposed a neo-associationist account of affective influences on judgments.
Ultimately, neither the conditioning, nor the psychoanalytic theories could provide a convincing explanation for when, how and why affect will infuse social judgments. It was only with the recent emergence of the social cognitive paradigm (Forgas, 1981), and the growing attention paid to information processing mechanisms that more comprehensive explanations of affect infusion began to emerge. Some of these models emphasized implicit, automatic cognitive processes as the main source of affect infusion (such as the affect priming model), whereas other theories proposed a more explicit mechanism to explain affect congruence, such as the affect-as-information model (Clore et al., 2001). We shall now briefly consider the role of different processing strategies in producing affect infusion effects.
Implicit and explicit mechanisms of affect infusion into judgments
Implicit processes of affect infusion: The affect-priming model
It was the affect-priming model developed by Bower (1981) that first highlighted the implicit role of affect in social cognition and judgments. Implicit affect infusion occurs, according to Bower (1981; Bower & Forgas, 2001), because affective states are directly linked to cognitions within a single associative network of memory representations. Affect may thus influence judgments through the automatic priming of associated constructs, as "activation of an emotion node also spreads activation throughout the memory structures to which it is connected" (Bower, 1981, p.135). Affect priming has several important consequences: affect should facilitate the learning of mood-congruent information because of the greater availability of an affect-congruent associative base. Affect should also facilitate the recall of information encountered in a matching rather than a non-matching mood state (mood-dependent memory; Eich & Forgas, in press). Due to the greater availability of affect-consistent associations, judges should evaluate and interpret current social stimuli in an affect-congruent manner in their judgments.