There are a number of accounts of how best to understand illocution. However, for the sake of argument, I will not
only assume that illocution makes sense, I will borrow Hornsby’s (1994) useful characterization. Hornsby suggests
that other accounts explaining illocution have been on the whole obscure, reliant on doubtful tests and
over-dependent on the notion of convention. For instance, consider an account that tries to explain illocution in terms
of consequences alone. This is problematic. Consequences seem to be as much a part of the picture when we
consider illocutions as when we consider perlocutions. Consider, for example, ‘warning’, which is an illocution. For
someone to successfully warn someone there will necessarily be certain consequence(s), such as that they are warned.
So, according to Hornsby, to demarcate illocution, focusing on consequences is not sufficient.
There is then a danger of losing the distinction between illocution and perlocution altogether, and our project will fall
at the first hurdle. However, Hornsby suggests a plausible way to proceed. We should not rely on linguistic tests,
consequence or convention. The key to understanding illocution is reciprocity. If I warn someone – an illocutionaryspeech-act – then what does there need to be for this to take place? Well, it seems that what is required is simply
certain receptiveness on the audience’s part, i.e. that the audience takes me to have said what I meant. Reciprocity is
the condition under which people can recognize one another’s speech as it is meant to be taken. Such a background
of reciprocity allows for Hornsby (1994, p. 194) to stipulate a working definition of illocution. Let us consider an
example to illustrate this definition. Imagine we are trying to work out whether the mathematics teacher asserting
that ‘2 + 2 = 4’ is an illocutionary act. According to Hornsby, it is if (and only if) the conditions are right. The
conditions are right if they allow the class to take the teacher to be asserting ‘2 + 2 = 4’. If these conditions are not in
place, then asserting ‘2 + 2 = 4’ in this context is not an illocutionary act.