Some Further Dimensions of Psycholingistics, Imagery, and Consumer Response

Morris B. Holbrook, Columbia University
ABSTRACT - This paper discusses some aspects of psycholinguistics, imagery, and consumer response that are sometimes neglected in studies of consumption behavior. Specifically, examples from the author's research are used to illustrate the dangers of neglecting (1) the power of language to evoke visual imagery, (2) the role of nonvisual imagery modalities, and (3) the nature of emotional experience. Awareness of these pitfalls and limitations may help guide consumer research toward a better-rounded understanding of the consumption experience.
[ to cite ]:
Morris B. Holbrook (1982) ,"Some Further Dimensions of Psycholingistics, Imagery, and Consumer Response", in NA - Advances in Consumer Research Volume 09, eds. Andrew Mitchell, Ann Abor, MI : Association for Consumer Research, Pages: 112-117.

Advances in Consumer Research Volume 9, 1982      Pages 112-117


Morris B. Holbrook, Columbia University

[The author gratefully acknowledges the support of Columbia University's Faculty Research Fund.]


This paper discusses some aspects of psycholinguistics, imagery, and consumer response that are sometimes neglected in studies of consumption behavior. Specifically, examples from the author's research are used to illustrate the dangers of neglecting (1) the power of language to evoke visual imagery, (2) the role of nonvisual imagery modalities, and (3) the nature of emotional experience. Awareness of these pitfalls and limitations may help guide consumer research toward a better-rounded understanding of the consumption experience.

Imagination is funny;

It makes a cloudy day sunny . . . .

--Johnny Burke and Jimmy Van Heusen,



In the two preceding papers, Rossiter and Percy have examined the visual and psycholinguistic components of advertising. These authors focus primarily on words and pictures as stimuli, verbal and visual modes of processing, and cognitive responses such as those involved in memory. Within these limits, they provide admirable reviews and exegeses. The present discussion is therefore directed at extending their perspective to encompass some additional types of psycholinguistic effects, some further dimensions of imagery, and some neglected aspects of consumer response.



One might extend Paivio's (1971) dual coding hypothesis to arrive at a conceptualization of information processing something like that shown in Figure l. This schema retains the familiar verbal and visual imagery systems (center column) but adds a third category called other imagery, which is intended less as a theoretical statement about how the mind works than as a reminder concerning the many additional types of intuitively familiar but poorly understood processes that occur in mental functioning. Similarly, words and pictures are retained as key stimulus types, but the diagram also recognizes other cues (e.g., auditory, gustatory, olfactory, and tactile) as inputs worthy of consideration. Finally, the figure suggests that cognitions (e.g., memory), problem solving (e.g., creativity), and experiences (e.g., emotions) serve as major response categories of interest.

Conceptually, the aforementioned types of stimuli, processing modes, and response categories could be connected in at least 33 = 27 different ways. As one adds further refinements to this simplest of schemes, the number of implicated phenomena increases geometrically. many such complex linkages are adumbrated in Rossiter and Percy's (1980) study of beer imagery and further elucidated in the excellent review chapter by these same authors (Rossiter and Percy, in press). However, most of the relevant research, including the papers presented at this session (to judge from their titles), has been somewhat more narrow in scope. For example, Percy (1981) focused primarily on the psycholinguistic/cognitive linkages shown by the arrows labeled "PPP" in Figure 1. Rossiter (1976), meanwhile, concentrated on the visual/cognitive linkages designated "RRR." Clearly, other linkages indicated by "HHH" for obvious reasons might also be studied to good effect.

This discussion dwells on three aspects of the schema indicated by HHH arrows. To preview briefly, I shall focus on important aspects of the following three phenomena:

(1) the connection between linguistic inputs (words) and processing by the visual imagery system (mental pictures);

(2) the relationship of sensory cues to mental imagery in other modalities (e.g., auditory, gustatory, olfactory, tactile);

(3) emotional experiences as a neglected aspect of consumption behavior.


Rossiter and Percy (in press) clearly recognize the role of language in evoking visual imagery. This phenomenon relates directly to Paivio's (1971) focus on the abstractness/concreteness of words and the power of more concrete linguistic cues to provide more vivid visual imagery. In other words, colorful language calls up clear mental pictures.

This function of language is central to the art of poetry. As explained by the Bard himself:

The poet's eye, in a fine frenzy rolling,

Doth glance from heaven to earth, from earth to heaven;

And as imagination bodies forth

The forms of things unknown, the poet's pen

Turns them to shapes, and gives to airy nothing

A local habitation and a name.

William Shakespeare,

A Midsummer-Night's Dream,

Act IV. Scene 2

In describing Shakespeare's art, Spurgeon (1935) defines poetic imagery as "the little word-picture used by a poet or prose writer to illustrate, illuminate and embellish his thought. It . . . transmits to us . . . the depth an richness of the way the writer views, conceives or has felt" (p. 9). This function of visual imagery is ubiquitous in poetic discourse. One conspicuous illustration appears in Milton's richly evocative descriptions of Eden Consider, for example, the language used to convey Satan' first glimpse of Adam and Eve:

His fair large Front and Eye sublime declar'd

Absolute rule; and Hyacinthine Locks

Round from his parted forelock manly hung

Clust'ring, but not beneath his shoulders broad:

Shee as a veil down to the slender waist

Her unadorned golden tresses wore

Dishevell'd, but in wanton ringlets wav'd

As the Vine curls her tendrils . . . .

So hand in hand they pass'd, the loveliest pair

That ever since in love's imbraces met.

--John Milton,

Paradise Lost, Book IV

Contrast this vision with the forlorn juxtaposition that expresses Adam's reaction to Eve's announcement that she has eaten the forbidden fruit:

Thus Eve with Count'nance blithe her story told;

But in her Cheek distemper flushing glow'd.

On th'other side, Adam, soon as he heard

The fatal Trespass done by Eve, amaz'd,

Astonied stood and Blank, while horror chill

Ran through his veins, and all his joints relax'd;

From his slack hand the Garland wreath'd for Eve

Down dropp'd, and all the faded Roses shed:

Speechless he stood and pale . . . .

John Milton,

Paradise Lost, Book IX

Here and throughout his work, Milton uses concrete language to evoke visual (and other) images that reflect emotional content. This psycholinguistic process was clearly described in Matthiessen's (1958) account of the techniques employed by T. S. Eliot: "His design is to give the exact perceived detail, without comment, and let that picture carry its own connotations . . . . the images here are 'consciously concrete'; they correspond a closely as possible to something . . . actually seen and remembered" (p. 63). Matthiessen's prime example is a passage from Ash Wednesday intended to convey "the very sensation of his distraction . . . by the enchantment of the senses" (p. 64):

though I do not wish to wish these things,

From the wide window towards the granite shore

The white sails still fly seaward, seaward flying

Unbroken wings


And the lost heart stiffens and rejoices

In the lost lilac and the lost sea voices

And the weak spirit quickens to rebel

For the bent golden-rod and the lost sea smell

Quickens to recover

The cry of quail and the whirling plover

And the blind eye creates

The empty forms between the ivory gates

And smell renews the salt savor of the sandy earth

T. S. Eliot,

Ash Wednesday

As Matthiessen remarks, "It is impossible to divorce the reflection from the imagery" (p. 64). Eliot himself spoke of this poetic property in terms of "transparency":

to write poetry which should be essentially . . . so transparent that we should not see the poetry, but that which we are meant to see through the poetry, poetry so transparent that in reading it we are intent on what the poem points at, and not on the poetry, this seems to me the thing to try for

--T. S. Eliot,

unpublished lecture (quoted by Matthiessen, 1958, p. 90)

My purpose in waxing poetic is to indicate the pitfalls that await any researcher who neglects the important connection between words and visual imagery. Such pitfalls are well illustrated by some of our own work in which we compared the effects of verbal and pictorial stimuli on evaluative judgments (Holbrook and Moore, 1981a, b). Briefly, we argued that words and pictures should tend to be processed by different verbal and imagery systems (e.g., Paivio, 1971), that these tend to reflect contrasts between sequential and parallel processing (e.g., Das, Kirby, and Jarman, 1979), that these phenomena also correspond to differences between the left and right cerebral hemispheres (e.g., Galin and Ornstein, 1972), and therefore that evaluative judgments of pictures should show more cue configurality than those of words so that pictures should produce a greater number of interaction effects (Holbrook and Moore, 1981a). We tested this hypothesis by exposing subjects to verbal descriptions or pictures of sweaters like those shown in Figure 2 (Holbrook and Moore, 1981b).



Specifically, 59 subjects rated 32 factorially designed sweaters on 20 bipolar adjectival scales that were subsequently reduced to four judgmental factors. Contrary to our hypothesis, however, we found no difference in cue configurality (as assessed by number of two-way interaction effects) between the verbal and pictorial treatments.



In light of the present discussion, this disappointing finding raises the possibility that our hypothesis might have failed because subjects did not process the stimulus information in the manner anticipated. For example, if subjects tended to convert verbal descriptions into visual imagery before forming evaluative judgments, this processing strategy might tend to cancel out any differences in cue configurality between the two treatments. This hypothesis (along with some others) was investigated by recontacting the subjects and asking for their degrees of agreement concerning the following statement: "When I rated the sweaters, I tried to form a mental picture of what each sweater would look like." we then found that our data showed a moderator effect consistent with the proposed explanation. As indicated in Figure 3, the hypothesized difference in cue configurality tended to appear in general (p = .025 for the main effect), but to disappear for those subjects who tried to convert the stimulus presentations into mental pictures (p = .005 for this interaction effect). In other words, the original hypothesis could be supported only when controlling for the tendency of words to evoke visual imagery more strongly in some subjects than in others. This interesting moderator effect therefore provided a lesson in the dangers of ignoring the visually evocative powers of language.


Rossiter and Percy (in press) also recognize the important, but neglected role of imagery in other modalities involving the senses of hearing, taste, smell, and touch. This multi-modality of imagery has been clearly articulated by literary critics and was illustrated in the previously cited passages by Milton and Eliot. Spurgeon (1935), for example, argues:

I use the term 'image' here as the only available word to cover every kind of simile . . . . I suggest that we divest our minds of the hint the term carries with it of visual image only, and think of it . . . as connoting any and every imaginative picture or other experience, drawn . . . through any of his senses (p. 5).

Thus, the philosophical definition of "image" as "a sensory quality reinstated by the mind in the absence of sensory stimulation" (Runes, 1980, p. 141) can apply to any of the senses. The same multi-sensory interpretation may be given to Richardson's (1969) delineation of mental imagery as

(1) all those quasi-sensory or quasi-perceptual experiences of which (2) we are self-consciously aware, and which (3) exist for us in the absence of those stimulus conditions that are known to produce their genuine sensory or perceptual counterparts, and which (4) may be expected to have different consequences from their sensory or perceptual counterparts (pp. 2-3).

Indeed, the early psychological work on imagery (e.g., Galton, 1883) did focus on auditory, gustatory, olfactory, and tactile modalities (see the reviews by Kaufmann, 1979; Richardson, 1969). More recently, however, psychologists have tended to dwell increasingly on the visual components of imagery:

The term "imagery" . . . may be used to denote cognitive content . . . both of a sensory and a nonsensory kind. More frequently, perhaps, "imagery" refers to symbolic processes of a specific sensory kind. Usually the reference is to the visual modality . . . . we shall use the term "imagery" in the latter sense, i.e., as substantially equivalent to visual symbolic processes. Such a usage of the term seems to coincide in large measure with current practice in the literature (Kaufmann, 1979, p. 9).

This preoccupation with visual imagery at the expense of other modalities appears reflected in the published work of consumer researchers (Holbrook and Moore, 1981a; Rossiter and Percy, in press). Of course, one can only speculate on the interesting, but proprietary multisensory research that might lurk in the secret files of advertising agencies, food processors, and perfume manufacturers. But those who consult the available literature will find few published studies on such important topics as the role of music in advertising (Wintle, 1978) or the psychophysics of taste (Huber, Holbrook, and Schiffman, in press). Yet, intuitively, one suspects that nonverbal and nonvisual modalities play a key role in consumer responses to products and related marketing efforts e

The danger of neglecting nonvisual and nonverbal modalities is again illustrated by some of our own work (Holbrook and Moore, in press). This study used the previously described data from subjects who rated sweater pictures to develop a spatial representation through the use of canonical correlation in general and discriminant analysis in particular. For example, a two-dimensional product space appears in Figure 4. Here, one can see immediately that the horizontal and vertical axes reflect the presence (absence) of stripes and sleeves, respectively. This point was confirmed by strong fits for stripes and sleeves vectors introduced into the product space (R = .96 and .84). Moreover, the vertical axis (sleeves) was closely aligned with evaluative ratings such as beautiful/ugly (r = .84) and displeasing/pleasing (r = -.80).

Notice that these dimensions are strongly visual in modality and that this property of the product space stems directly from the nature of the experimental design, which used carefully controlled but highly artificial pictures of sweaters. One might therefore wonder what would happen if subjects were exposed to real sweaters varying not only visually but also in other sensors modalities.





This question was explored in a study (Holbrook, 1981) that collected 34 adjectival ratings similar to those discussed earlier from 25 subjects of 20 real sweaters varying on some of the same visual dimensions (pattern, fit, sleeves, neck) but also on some primarily tactile dimensions: heaviness (weight divided by surface area) and stitching (number of stitches per inch). The same canonical correlation procedure was used to construct the product space shown in Figure 5. Here, one of the visuaL characteristics is again represented horizontally by the general drift from solid to patterned, multi-colored, and striped designs. By contrast with the earlier study, however, the vertical dimension now reflects more tactile sensations. As indicated by feature vectors introduced into the product space, sweaters toward the top are heavier and looser-knit (R - .61 and .65) while those toward the bottom are lighter and tighter-stitched with a smoother, softer finish. Moreover, this tactile dimension corresponds closely to a preference vector showing that the lighter-and-softer sweaters toward the bottom are evaluated more favorably than the heavier-and-coarser garments toward the too (R - .50).

These findings suggest the important role played by tactile imagery in the evaluative judgment of sweaters. More generally, they illustrate the danger of being misled by experimental paradigms that emphasize strictly visual and/or verbal processes at the expense of neglecting other modalities. In the present case, preferences appear to be driven by tactile sensations--a point that was obscured by the earlier presentation of stimuli in schematized pictorial form. This finding therefore argues for increased attention to the multi-sensory aspects of the consumption experience.

Further, one might conjecture that, for many product classes, important interactions might occur across imagery modalities. The potential importance of such cross-modality cue configuralities may be illustrated by the case of a liquor company that introduced a new product resembling vodka in color and bourbon in taste. This product flopped miserably on the market. Thus, with 20/20 hindsight, one concludes that this failure may have resulted from the negative synergy between its visual impression (vodka) and gustatory sensations (whiskey).


Much of the psychological work on imagery has focused either on its cognitive effects associated with memory (e.g., Paivio, 1971) or on its relationship to creative and other aspects of problem solving (e.g., Kaufmann, 1979). In consumer research, these preoccupations have been compatible with the dominant view of buyers as information processors who seek, attend to, perceive, remember, and evaluate marketing cues as part of a decision-making task that culminates in brand choice and purchasing activity (e.g., Bettman, 1979). This prevalent information-processing view is clearly reflected in the work of both Percy (1981) and Rossiter (1976).

Recently, we have advocated greater emphasis on some relatively neglected components of consumer behavior associated with what might be called the experiential aspects of consumption (Holbrook and Hirschman, 1981). This perspective takes a more phenomenological look at the consumption experience and draws attention to its symbolic, hedonic, emotional, and esthetic components. In the present context, the experiential approach suggests that we might fruitfully pursue a more careful investigation of the emotions stirred up by visual and other kinds of imagery, especially in an era when much advertising appears to be veering toward a greater emphasis on feelings rather than facts ("Have a Coke and a smile," "Get that great G.M. feeling with genuine G. M. parts," "Marlboro Lights . . . the spirit of Marlboro in a low tar cigarette," "The Sealy Posturepedic morning . . . Feeling so good it shows," "Kodak . . . decorate your home with love."

Rossiter and Percy (in press) mention the emotive function of imagery, but refer primarily to work using the Semantic Differential and its tripartite distinction among evaluation, activity, and potency (Osgood, Suci, and Tannenbaum, 1957). Unfortunately, our own research again provides a lesson in some potential dangers to be avoided in this respect. Specifically, as previously noted, we have tended to use conventional bipolar adjectival or semantic differential scales to tap responses to esthetic and other imagery-evoking stimuli. For example, one study focused on conventional bipolar adjectival ratings of line drawings and architectural photographs (Huber and Holbrook, 1981), prompting Olson (1981) to offer the following criticism:

what is an esthetic response? . . . Are these "esthetic reactions" different from consumers t responses to the more mundane products of everyday life such as toothpaste, chewing gum, breakfast cereal, and automobiles? If so, what is it that is special or unique about them? For instance, is there some particularly distinctive combination of affective and emotional characteristics that comprise what we seem to want to call an esthetic response? (p. 71)

Of course, Olson is right. One cannot necessarily use traditional methods to measure new constructs. Moreover, his critique can be extended beyond the boundaries of consumer esthetics to cover the whole state-of-the-art in studying emotional experience. Here, measurement techniques are still in their infancy. As a field of inquiry, the emotions are just beginning to attract the sustained, systematic attention of psychologists (e.g., Plutchik, 1980). With little to draw upon in the way of established, well-validated methodology, the efforts of consumer researchers in this direction have understandably been tentative and highly exploratory. Nevertheless, we cannot expect to plumb the depths of emotional experience with methodologies developed to test conventional multiattribute attitude models. The need to work out suitable procedures for investigating the consumer's emotional experiences must therefore be considered a challenging but fascinating task along the road toward a fuller understanding of the experiential aspects of consumption.


This special ACR session has provided a welcome opportunity to address the interacting roles of psycholinguistics and imagery in consumer behavior. Rossiter and Percy have performed a valuable service in calling our attention to the importance of these phenomena in advertising and, by extension, in all marketing activities. For these insights, I applaud their contribution. I can only add that, as usual, much still remains to be done. In sum, we need more work on other linkages (e.g., linguistic-visual), other sensory cues and imagery modalities (e.g., tactile), and other kinds of response (e.g., emotional experiences). As consumer research proceeds into these relatively neglected areas, we can look forward to developing a better-rounded understanding of the consumption experience where, in the words of a Nineteenth Century poet,

Thought is deeper than all speech,

Feeling deeper than all thought . . .

-Christopher Pearse Cranch,



Bettman, James R. (1979), An Information Processing Theory of Consumer Choice (Reading, MA: Addison-Wesley Publishing Company).

Das, J. P., Kirby, J., and Jarman, R. F. (1979), Simultaneous and Successive Cognitive Processes (New York: Academic Press).

Galin, David, and Ornstein, Robert (1972), "Lateral Specializations of Cognitive Mode: An EGG Study," Psychophysiology, 9, pp. 412-418.

Galton, F. (1883). Inquiries Into Human Faculty and Its Development (London Macmillan).

Holbrook, Morris B. (1981), "Cue Configuralities and Tactile Sensations in Consumer Judgments of Real Products," (working paper, Columbia University).

Holbrook, Morris B., and Hirschman, Elizabeth C. (1981), "The Experiential Aspects of Consumption: Consumer Fantasies; Feelings, and Fun," (working paper, Columbia University).

Holbrook, Morris B., and Moore, William L. (1981a), "Cue Configurality in Esthetic Responses," in E. C. Hirschman and M. B. Holbrook, (eds.) Symbolic Consumer Behavior (Ann Arbor: Association for Consumer Research).

Holbrook, Morris B., (1981b), "Feature Interactions in Consumer Judgments of Verbal Versus Pictorial Presentations," Journal of Consumer Research, 89, pp. 103-113.

Holbrook, Morris B., (in press), "Using Canonical Correlation to Construct Product Space for Objects with Known Feature Structures," Journal of Marketing Research.

Huber, Joel, and Holbrook, Morris B. (1981), "The Use of Real Versus Artificial Stimuli in Research on Visual Esthetic Judgments," in E. C. Hirschman and M. B. Holbrook, (eds.) Symbolic Consumer Behavior (Ann Arbor: Association for Consumer Research).

Huber, Joel, and Schiffman, Susan (in press), "Situational Psychophysics and the Vending-Machine Problem, " Journal of Retailing.

Kaufmann, Geir (1979), Visual Imagery and its Relation to Problem Solving (Bergen: Universitetsforlaget).

Matthiessen, F. O. (1958), The Achievement of T. S. Eliot An Essay on the Nature of Poetry (New York: Oxford University Press).

Olson, Jerry C. (1981), "What is an Esthetic Response?" (eds.) E. C. Hirschman and M. B. Holbrook, in Symbolic Consumer Behavior (Ann Arbor: Association for Consumer Research).

Osgood, Charles E., Suci, George J., and Tannenbaum, Percy H. (1957), The Measurement of Meaning (Urbana: University of Illinois Press).

Paivio, Allan (1971), Imagery and Verbal Processes (New York: Holt, Rinehart and Winston).

Percy, Larry (1981), "Psycholinguistics: Some Simple Rules for Do-It-Yourself Ad Testing," (working paper, Creamer).

Plutchik, Robert (1980), Emotion: A Psychoevolutionary Synthesis (New York: Harper & Row).

Richardson, Alan (1969), Mental Imagery (New York: Springer Publishing Company).

Rossiter, John R. (1976), "Visual and Verbal Memory in Children's Product Information Utilization," in B. B Anderson (ed.) Advances in Consumer Research, Vol II: (Ann Arbor, Association for Consumer Research).

Rossiter, John R., and Percy, Larry (No. 2, 1980), "Attitude Change Through Visual Imagery in Advertising," Journal of Advertising, 9, pp. 10-16.

Rossiter, John R., (in press), "Visual Communication in Advertising," in R. J. Harris, (ed.) Information Processing Research in Advertising (Hillsdale, NJ: Lawrence E. Erlbaum & Associates).

Runes, Dagobert D. (1980), Dictionary of Philosophy (Totowa: NJ: Littlefield. Adams & Co.).

Spurgeon, Caroline F. E. (1935), Shakespeare's Imagery (Boston: Beacon Press).

Wintle, Regie Rae (1978), Emotional Impact of Music on Television Commercials, Dissertation, University of Nebraska, Lincoln.