Most models of trade in speculative markets assume that agents interpret public information identically. We provide empirical evidence on the relation between the volume of trade and stock returns around public announcements, and we argue that the evidence is inconsistent with this assumption. We then develop a model of trade around public announcements that incorporates differential interpretations and is consistent with the observed volume-return relation. Then we test the standard model of belief revision underlying most models of trade using stock brokerage research analysts' earnings forecasts. The hypthesis of identical interpretations seems inconsistent with the forecast revisions in these data.