The Role of Imagery in Information Processing: Review and Extensions DEBORAH J. MACINNIS LINDA L. PRICE* Mental imagery is receiving increased attention in consumer behavior theory and research. This article describes imagery, characterizing it as a processing mode in which multisensory information is represented in a gestalt form in working memory and discusses research on the unique effects of imagery at low levels of cognitive elaboration. It specifies researchable propositions for the relationship between high elaboration imagery processing and consumer choice and consumption behaviors Finally, it reviews specific methods for studying imagery. nformation processing research has traditionally fo- 1978; Cohen 1982; Greenwald and Leavitt 1984; Ros- I cused on discursive or descriptive information pro- siter and Percy 1978, 1983; Smith et al. 1984). cessing (Bettman 1979; Greenwald 1968; Olson, Toy, This article addresses the meaning of imagery pro- and Dover 1982; Swasy and Munch 1985; Wright 1974,' cessing, how it differs from discursive processing—par- 1980).' As such, researchers have examined how sym- ticularly under varying degrees of cognitive elaboration, bols (most commonly words and numbers) are com- and what unique effects it has on processing outcomes! bined in working memory to represent and solve prob- It also reviews research on the unique effects of imagery lems. Discursive (symbolic, language-like) information under conditions of low elaboration and explores the processing encompasses a broad range of strategies. potentially unique effects of imagery at higher levels of Compositional choice strategies, counterarguments, elaboration. The article gives specific attention to the attributions, and formulations of choice rules are illus- role of imagery in problem framing, assessing proba- trations of discursive information processing. Increas- bilities, forming intentions, and generating affect. In ingly, however, attention has been given to the role of addition, it considers the use of imagery in consumer mental imagery in information processing (Childers and settings throughout the phases of consumption and de- Houston 1982, 1984; Childers, Houston, and Heckler velops researchable propositions to guide future re- 1985; Lutz and Lutz 1977, 1978; Rethans and Hastak search on elaborated imagery. Finally, it explores issues 1981; Rossiter 1982; Rossiter and Percy 1983; Smith, related to the conduct of research on imagery. Houston, and Childers 1984). Imagery is a conceptually distinct way of representing information, a way that is "very like picturing and very unlike describing" (Fodor THE MEANING OF IMAGERY 1981, p. 76).^ To date, this emergent research stream Imagery is defined here as (1) a process (not a struc- has focused predominantly on imagery processing at ture) by which (2) sensory information is represented low levels of cognitive elaboration—for example, men- in working memory. Imagery processing, and infor- tally picturing a stimulus object. More recently, there mation processing in general, fall on an elaboration have been discussions that refer to imagery processing continuum that ranges from processes limited to the at higher levels of elaboration. These discussions suggest simple retrieval or evocation of a cognitive concept to that elaborated imagery plays a role in influencing (1) affective responses to stimuli, and (2) behavior (Calder As used here, discursive means passing from premises to conclu- sions; proceeding by reasoning or argument (Oxford dictionary). De- scriptive processing is a somewhat broader term suggesting any •Deborah J. Maclnnis is Assistant Professor, Department of Mar- meaningful combination of verbal or other symbolic units. Although keting, University of Arizona, Tucson, AZ 85721. Linda L. Price is the two terms have a slightly different connotation, they will be used Assistant Professor, Department of Marketing. University of Pitts- interchangeably here for ease of presentation. burgh, Pittsburgh, PA 15260. Authors are listed in alphabetical order •This is not to say that the debate of what imagery is and what it to reflect equal contributions. The authors would like to acknowledge is not has been resolved (cf Block 1981). Processing probably often the many helpful comments of C. Whan Park and three anonymous occurs on a continuum between "pictures" and "symbols." For ex- reviewers on earlier versions of this article. ample, maps and stick figures (as mental images) would have features of both (Fodor 1981). 473 © JOURNAL OF CONSUMER RESEARCH • Vol. 13 • March 1987 THE JOURNAL OF CONSUMER RESEARCH 474 processes involving multiple concepts and constructions an individual may draw upon a script of a "romantic (e;g., problem solving, creative thinking, and day- evening" in imagining such an evening. Information dreaming). The following section will describe each of from schemas or scripts can also be processed in a dis- these issues in greater detail, and will serve to (1) dis- cursive mode. Hence, the information structure is viewed as distinct from processing mode. tinguish imagery processing from schemas and scripts, Since imagery processing relies on stored knowledge, (2) distinguish imagery processing from discursive pro- the evocation and vividness of the image is likely to cessing, and (3) describe the elaboration continuum for depend on the level of knowledge development. Con- imagery processing. sistent with this notion. Smith et al. (1984) found that individuals who possessed a script for a given event re- Imagery as a Process, Not a Structure ported evoking imagery experiences when they instan- While the idea of imagery processing is widely ac- tiated that script. Moreover, inclividuals with well-de- cepted, the assumption that knowledge is stored or rep- veloped scripts reported that their imagery experiences resented as images has been hotly debated (Pinker and were significantly more vivid than did individuals with- Kosslyn 1983). Propositional theorists (e.g., Anderson out well-developed scripts. :1978; Anderson and Bower 1973; Lang 1979; Pylyshyn 1973, 1981) argue that knowledge is represented only Imagery as a Sensory Process as a set of verbal propositional networks composed of nodes (representing concepts) and connected by links Imagery processing has several qualities that distin- (representing relationships between concepts; see also guish it from discursive processing. Most fundamen- Kieras 1978). Others have argued that knowledge can tally, imagery processes are evoked as sensory experi- be stored as images. In support of their argument they ences in working memory. Imagery processing includes note that studies on mental rotation of objects, image perceptual or sensory representations in working mem- size, and parallel processing are more consistent with ory that are used in much the same way as perceptions an imagery' perspective than a propositional one (Bug- of external stimuli. Thus, imagery involves concrete elski 1983'; Kieras 1978; Kolers 1983; Kosslyn 1975; sensory representations of ideas, feelings, and memo- Kosslyn and Pomerantz 1977; Kosslyn et al. 1981). ries, and it permits a direct recovery of past experiences These researchers contend that images have emergent (Yuille and Catchpole 1977). The evocation of iniagery properties that cannot be explained by propositjonal may be multi-sensory—involving images that incor- network theories. porate, for example, smell, taste, sight, and tactile sen- sations—or may involve a single sensory dimension, While the debate between propositional theorists and imagery theorists is far from resolved, Yuille and such as sight. In contrast, discursive processing (such Catchpole (1977) have developed a compromise posi- as verbal retrieval, cognitive responding, and verbal en- tion based on the work of Piaget and Inhelder (1973). coding) is more detached from internal sensor>- expe- They propose that the ability to generate images does riences. The absence of sensory dimensions of infor- not necessarily imply that knowledge is stored as images. mation in working memory- makes these discursive They argue that knowledge is stored in an abstract (but processes less concrete (more abstract) than imagery not necessarily verbal) operational code. The well-sub- processes. Moreover, unlike discursive processing, im- stantiated ability to move from words to pictures and agery can be described along several unique sensory- pictures to words suggests that there is a representation related dimensions. For example, vividness refers to the in memory that encompasses both. Once a knowledge clarity of images. And controllability reflects the extent structure has been activated, imagery can be generated to which images can be held in mind and/or altered in from information contained in that structure. For ex- specific ways at will (Marks 1972). ample, after activating the knowledge structure for That imagery involves internal sensory experiences birds, an individual may imagine a prototypical bird has both intuitive and empirical support. Early research such as a sparrow (see Mervis and Rosch 1981 for a suggested that people can confuse imaging with per- review). ceiving (Perky 1910). Even before research on imagery Consistent with the views of Yuille and Catchpole existed, people talked naturally about the "pictures" in (1977), imagery is conceptualized here as a mode of their heads. Research has since indicated that imagery processing information. In other words, imagery pro- and perceiving (or sensing) share the same