Frequency and specificity of referents to violence in news reports of anti-gay attacks

N. Henley, M. Miller, J. Beazley, D. Nguyen, D. Kaminsky, R. Sanders

Two studies were conducted to test the hypotheses that: (1) news reports of anti-gay attacks would use fewer and vaguer referents to the violence than similar stories about attacks against straight persons; and (2) this milder use of referents would cause readers to perceive less harm done and to blame the perpetrator less. A content analysis of two newspapers found that one used far fewer, less specific nominals to refer to anti-gay than to anti-straight violence, whereas the other, based in a more gay-friendly community, did not differentiate significantly by sexual orientation. An experimental study in which frequency and specificity of referents were systematically varied in mock newspaper stories found that greater referent frequency, but not specificity, caused readers to perceive greater harm to victims. The results are interpreted in terms of cognitive processing and within the context of the use of linguistic variation to encode and enforce power differences.