Past work on social risk amplification has concentrated on studying large-scale, aggregated societal responses to risk issues as the outcome of social processes. An alternative approach, explored in this article, is to regard amplification as an attribution. Social risk amplification is something that social actors attribute to one another as they try to explain their systematic differences in response, not an objective characterisation of a response that is somehow disproportionate to its stimulus. This avoids the problem that the idea of a social amplification of risk can be taken to imply that a risk external to the social system can somehow be distorted by it. The attributional view also helps illustrate differences in risk response as much as commonalities. In order to explore risk amplification as an attribution we analyse the explanations and descriptions used by actors discussing recent outbreaks of zoonotic disease. We present a grounded analysis that produces a classification of these explanations and descriptions. It is evident from the clichés that informants used – such as ‘crying wolf’, ‘scare-mongering’ and ‘jumping on the bandwagon’ – that social actors have had a concept resembling social risk amplification that long predates the social amplification of risk framework. Moreover, they use it in a strongly normative way: sometimes saying this amplification is explicable and excusable, sometimes not. It is therefore a basis of social judgment. The idea of amplification as an attribution offers the particular advantage that it helps deal with situations where social actors develop their risk responses in reaction to the risk responses of other social actors.