Alcohol Shifts Emotional Language in Social Interaction Toward Positive and Away from Negative Content: A...

Although alcohol is largely consumed socially, experiments rarely test its effects on social language. Guided by the premise that language is a fundamental social behavior that shapes alcohol's socio-emotional effects, this preregistered study addresses critical gaps by testing alcohol's verbal effects in unstructured small-group interactions using the largest sample to date, proper controls, and reliable/replicable text-processing measures.

Sayette et al. (2012) randomly assigned unacquainted triads of social drinkers (360 males, 360 females) to alcohol, placebo, or control conditions, and recorded them during an unstructured interaction. Here, extensive transcription efforts permitted examination of alcohol's impact on approximately half-a-million words, using the Linguistic Inquiry and Word Count (LIWC) system.

Hypotheses aligned with general verbal disinhibition were unsupported. Instead, alcohol, vs. placebo and control, decreased the proportion of negative emotion words (B = -0.19, OR = 0.83, p < .001) and increased occurrence of positive-emotion words (B=0.26, IRR = 1.30, p = .002). Further, there was a trend such that alcohol, vs. placebo and control, reduced verbal authenticity (B = -4.32, β = -0.19, SE = 2.37, t(233.23) = -1.82, p = .07). Post-hoc analyses suggested this effect was driven by males, with alcohol decreasing authenticity (B = -9.85, β = -0.43, p = .040) and increasing confident/dominant language (B = 9.63, β = 0.42, p = .045), selectively in all-male groups.

Results indicate that alcohol alters the emotional focus of conversation and illuminate potential sex-differences in its effects on authenticity/bravado. This study underscores the value of incorporating formal language measures into alcohol (and other drug) research.