Mental Models of Contagion

What are the properties that the mind attributes to the contagious entity? We enumerated the principles of contagion previously; these constrain but do not fully dictate the manner in which the contagious entity is conceived. Furthermore, given the wide range of contagious entities, from germs to individuals with compromised morals, it is likely that there might be different mental models, varying within individuals across domains of contagious entities, among individuals within a culture, and between cultures. We explored this possibility in an in-depth interview study with 36 more or less representative American adults (Nemeroff, 1988; Nemeroff & Rozin, 1994). Participants were asked to imagine various source people coming into contact with a sweater (and various other objects), and then imagine their reaction to wearing this sweater. We then asked them to further imagine a set of “purifications” of that sweater (such as laundering or changing its color or appearance) and asked participants to estimate their willingness to wear the sweater after each of the designated “purifications.” Our logic was that one could deduce the nature of the contagious entity by discovering how it could be neutralized.

The source people were one’s lover, a good person, and a sex symbol (positive interpersonal sources); one’s enemy and an evil person (negative interpersonal sources); a person with contagious hepatitis and a person with AIDS (physical illness sources); and an imperceptible amount of dog feces on the sweater (physical disgust). Purification methods included physical cleansing (e.g., deodorizing, washing by hand, sterilizing); physical/symbolic transformations (e.g., unraveling the sweater and reknitting it into a scarf; gashing it); and spiritual “opposite-contacts” (i.e., having an opposite-valenced source wear the sweater, such as Mother Theresa wearing Hitler’s sweater). Templates were devised for five possible models of contagion:

1. A physical germ model, in which contagion is seen as carried by a living invisible entity (particularly vulnerable to boiling, and to a lesser degree to washing).

2. A physical residue model, in which contagion is dependent on residues such as sweat, dandruff, and so on (particularly vulnerable to washing).

3. A symbolic interaction model, in which contagion is the result of the meaning implied by interaction with an object (symbolic actions like gashing or otherwise destroying or reconstituting should be effective).

4. An associative model, in which the reminding value of an object is the key to effects (change in appearance cues rather than substance, which would be expected to be moderately efficacious).

5. A spiritual essence model, in which the source’s “nonmaterial essence” or “soul-stuff” are believed to be in the object (here, the only effective purifier might be an opposite force of the same type; that is, Mother Theresa’s positive essence neutralizing Hitler’s negative essence).

We had a separate set of subjects read a description of each model and rank order the set of purifications in terms of their effectiveness for/within each of the models. We then tested these template orders against the actual rank orders provided by participants in the study to determine the best-fitting model for each subject and each source.

The predicted purification potency for the two physical models (germ and physical residue) turned out to be highly similar, as did the potency ordering for the three nonphysical models (associative, spiritual, and symbolic interactional). Because of this relative similarity and the small sample size, we combined the models into two broad classes, henceforth termed physical (material) and nonphysical (spiritual).

On average, the nonphysical models were the best match for all interpersonal sources, both positive and negative, and the physical models were the best match for physical sources (illness and disgust). Thus, the idea of at least two broad models of contagious essence was supported. One is material/physical and is effectively moderated by washing; the other is spiritual/nonphysical, is reduced much less by washing, is very difficult to erase, and is most effectively reduced by opposite valence contact. Because positive contagion sources are almost entirely interpersonal, positive contagious essence is typically spiritual/nonphysical.

The associative model of contagion, which we have included in the nonphysical category, is problematic. It does not capture many of the properties of contagion, particularly the physical contact principle. Although physical contact may promote association, it is hardly necessary, and unlike the symbolic and spiritual essence models with which it is grouped, there is no “spiritual” aspect to it. In a sense, the association model is an alternative to contagion because it does not depend on physical contact and the transfer of essence. Unfortunately, it is a difficult model to distinguish from other nonphysical models. Two observations support the claim that there is something besides association going on in cases of non-physical contagion. First, ownership without contact, although heavy on association, seems less potent in transferring negativity than brief contact without ownership (Nemeroff, 1988; Rozin et al., 1992). Along the same lines, similarity, based on images or representations, and hence on association, is generally less potent than contagion, in which reminding cues are less salient. In the extreme, a book about Hitler’s life with his photograph on the cover and quotes of his words inside is less offensive to many people than his worn sweater would be, even though the sweater is not uniquely linked to Hitler in its sensory properties.