Americans are Barbarians or Allies? Images of Americans in the Italian Social Context

Abstract

Image theory was tested in the context of Italian/American relationships. Participants were Italian students. In addition to the images of ally, barbarian, imperialist, enemy, a new image was suggested: that of father. We predicted that political orientation would affect the structural perceptions (goal compatibility, relative power, relative status), and, as a consequence, the image of Americans: right-wing participants should prefer the images of father and ally, while left-wing participants should associate Americans with barbarian. Findings generally supported our predictions; also the figure of father obtained some support. The theoretical and practical implications of results are discussed.

ACKNOWLEDGMENTS

We thank Annalisa Bigarello for her help in collecting and coding data for this survey.

Notes

¹For left-wing participants, one of the four items measuring the image of ally was dropped owing to its low correlations with the other items: “Americans care only about their own interests, and not about Italian interests.” In the left group, reliability of items, measuring the two cooperative images, was somewhat low. Probably, left-wing participants endorsed ambivalent attitudes toward Americans, believing that, while in the past Americans supported Italians, now American and Italian goals are incompatible. Ambivalence can affect the response consistency.

²As to the ts, because of missing data, the degrees of freedom were sometimes lower than 101, for left-wing participants, and lower than 91, for right-wing participants.

Note. On the 7-step scale, the higher the score the higher the endorsement of the image. A different subscript, in the same column, indicates the two means are significantly different, p < .001.

∗p ≤ .05. ∗∗p < .01. ∗∗∗p < .001: difference from the neutral point of the scale (4).

³Concerning the barbarian image, right-wing participants were neutral, and they endorsed this figure much less strongly than their left-wing counterparts.

∗p < .01. ∗∗p < .001.

⁴We analyzed semantic differential data also to explore the correspondence between the direct and indirect assessment of images. For each participant, we calculated the correlation between the profile of American and that of each of the five images (scales were used as the units of analysis); these correlations were z′transformed. Then, for each image, the correlation was computed between its direct (agreement/disagreement scale) and indirect (z′; semantic differential) assessment; participants were the units of analysis. Findings showed that, for each political group and each figure, the direct and indirect measures were significantly related, rs ≥ .22, ps < .04, the only exception was imperialist, for the right-wing group, r = .02. These findings represent a kind of concurrent validation for both types of measure; moreover, they indicate that, when answering the direct measure, participants expressed a personal belief and not only the position of their political in-group.

^a Item taken from Alexander, Levin, and Henry (2005).

^b Item taken from Alexander, Brewer, and Livingston (2005).