The Pan-cultural Semantic Differential. The early work of Osgood and his associates on the semantic differential technique, as summarized in Osgood, Suci, and Tannenbaum (9) clearly established the following fact about the modes of verbal qualification in the usage of English by Americans: if people are asked to rate concepts on bi-polar scales made up of a large number of randomly selected adjectival opposites (e.g. good-bad, sweet-sour, serious-fickle, deep-shallow, young-old, etc.), a substantial portion of the variability in ratings on the various scales (between 60% and 65%) can be accounted for by only three factors. This remarkable finding represents at once the strength and the weakness of the semantic differential technique. It is a weakness, indeed a fatality, for its use as a general measure of the meaning of words since it is obvious that in order to differentiate the many thousands of words that the ordinary speaker of English knows, many more dimensions than three will be needed. It is a strength, for its use as a measure of the affect of words since the small number of factors (only three) insures their highly generalized nature so that the rating instrument gains increased usefulness--and indeed, it has been successfully used with all categories of words and with non-verbal stimuli (geometric forms, color patches, works of art, sounds, objects, people, etc.; see the bibliography in Snider and Osgood (10).)

Having established the American English generality Or the three-factor affective system and having identified the three factors as Evaluation (E), Potency (P), and Activity (A), the next step was to inquire whether this specific language culture generality extended to other language/culture communities. The affirmative answer which was fully expected ran contrary to a broad interpretation Or the Sapir-Whorf Relativity Hypothesis-- and the empirical work which ensued left no doubt as to which hypothesis was correct: each language/culture community, up to 28 in all (so far), exhibited the same pattern (6,4). The E-P-A factor system turned up in every case by independently replicated indigenous (i.e. not by translation) elicitation procedures followed in a carefully standardized method in each location.

One of the useful tools that emerged from this cross-cultural work in the "pan-cultural short form semantic differential" which consists of 12 bi-polar adjectival scales--four scales for each of the three universal affective factors--with each of the 28 communities possessing a version in their own language. The advantage of this tool is that the E-P-A value of stimuli can now be directly compared across the various cultures. For example, in the World Atlas of Semantic Profiles (2) comparisons are made on the affective meaning of 600 concepts as rated by unilingual speakers in each location. These comparisons lead to hypotheses about cultural universals as well as differences and peculiarities proper to a cultural group (for examples, see 3).

The Cross-cultural Generality of Visual-Verbal Synesthetic Tendencies. A demonstration I like to try when discussing with my students the general problem of metaphor goes as follows. I ask them to imagine a collection of stimuli and group them into sets that "fit together"; the collection is made up of various sensory stimuli in different modalities: the loud clang of a bell, the feel of sandpaper, an itchy scratch, a white circle, a straight line, the taste of chocolate, a rising melody, various colors, and so in. Invariably, there is good agreement on how the sub-sets fit together: there is a set that contains a white circle, a straight line, a caressing touch, a sweet taste, a bright color, a rising melody: there is another set containing a loud clang, saturated colors, deep low sounds, hard surfaces, and irritating scratches; and a third set is made up of bright colors, wavy lines, rhythmic sounds, and flashes of light.

What is the basis of these agreements? Osgood's explanation, in summary form, runs as follows: "...the highly generalized nature of the affective reaction system--the fact that it is independent of any particular sensory modality, yet participates with all of them--is...the psychological basis for synthesia and metaphor. It is because such diverse sensory experiences as a white circle (rather than black), a straight line (rather than crooked), a rising melody (rather than a falling one), a sweet taste (rather than a sour one), a caressing touch (rather than an irritating scratch) can all share a common affective meaning that one can easily and lawfully translate from one modality into another in synesthesia and metaphor...I would suggest that this meaning system is intimately related to the nonspecific projection systems from the hypothalamic, reticular, or limbic systems and their cortical connections in the frontal lobes--both are gross, nondiscriminative, but highly generalizable systems and both are associated with the affective, purposive and motivational dynamics of the organism" (8, p.21).

Is it just chance, or universal affect, that is at work in our linguistic metaphors when we say that we move up in society when we better our social standing, when we form an inner circle of friends to keep strangers out, when we refer to happy faces as bright, when our lot is sweet when we are lucky, when a pleasant voice is caressing, and so on? There is evidence (1) that the up-down and in-out analogies of social standing appear in other languages as well, and the "laws of good Gestalt" which Wertheimer and Kohler talked about were certainly meant to be a universal property of the human visual system. Osgood's (5) study on the cross-cultural generality of certain visual verbal synesthetic tendencies demonstrated good agreement on the part of subjects drawn from three different linguistic/cultural backgrounds: American college students, Japanese foreign students in the US, Mexican-Spanish speaking persons in the Southwest, and Navajo Indians. These subjects were shown cards on which were drawn two simple figures: e.g. a white and a black circle, two arrows (one pointing upward, the other downward), two lines (one thick, the other thin), two circles (one large, the other small), etc. subjects were given words in their language, one at a time (e.g. heavy, Man, yellow, sad, EXCITEMENT) and they were to point to one or the other of the visual alternatives on the card according to what "seemed most appropriate to that concept." Significant and substantial agreements were found for most of the answers both intra-culturally and cross-culturally.

The Cross-cultural Graphic Differential. The preceding discussion tried to establish the two presuppositions mentioned in the Introduction as they constitute the rationale for attempting to construct a "graphic differential" (henceforth GD). If one is convinced that the E-P-A affective system is universal and that it mediates the synesthetic reactions to visual stimuli, then it makes sense to attempt to construct a visual equivalent of the pan-cultural short form semantic differential (the latter will henceforth be referred to as SD). Thus, the GD was to become a visual equivalent of the SD in all its respects: it should be made up of bi-polar terms (henceforth to be referred to as "pictographic opposites"); these should have the property of a unidimensional scale with seven polarity positions including a mid-point; there should be a small number of them and of a size suitable to fit a single page.

There are a number of reasons why such an instrument would be desirable. The following appear to me to be the most salient: (1) It would represent an alternative form to the present SD. Several advantages are to be had by this. Some of our colleagues are bothered by the fact that the various indigenous forms of the SD are not translation equivalents of each other, yet we infer that they all measure the same affect. They might find the GD more comfortable to work with since the forms for all languages will now be (visually) identical. Other colleagues have worried about the effects on SD ratings in situations where the experimental design called for repeated testing; the availability of an alternative non-verbal form should relieve some of this anxiety on their part. In some situations, verbal written forms of the SD are inappropriate; e.g. with pre-school children, with illiterate adults or those from a pre-literate society, with aphasics who have lost the ability to read the GD, should it prove practicable with these subjects on other grounds (e.g. the concept of a scale, the rating task), would then allow for applications now impossible with the SD. (2) It would provide insights into the nature of synesthesia and metaphor. Arguments to this effect have already been presented above. Osgood's neurophysiological speculations about synesthesia, while just that, nevertheless provide a rationale for interpreting a large variety of disparate phenomena.

The specific visual characteristics of the pictographic scales that have cross-culturally similar affect-- should such an outcome turn out to be feasible--would provide specific hypotheses about the nature of the human visual system and its dynamic relation to the human motivational system. Such an outcome would also be relevant to certain semantic univer sals that arc grounded in the perceptual system such as visual metaphors in language. (3) It would contribute to the psychology of art and aesthetics. A demonstration of the existence of visual forms that have an unlearned, universal meaning (of the affective sort) would contribute to the long standing controversy between critics who maintain an "absolutist" position vis a vis "good" art and proponents of the "relativist" school. There is keen interest today among graphic designers in constructing international symbols to denote public facilities and services; a cross-cultural GD should prove useful in identifying characteristics of visual symbols that carry "emotional tones" acceptable in all cultures.