The argument is manifestly circular. It sets out from the facts: threats are not
obviously directed towards agreement, but pertain to the category of strategic
acts. However, a threat is a speech act, governed by the interlocutory structure
of language. A threat has a speaker, an addressee, a propositional content,
and exercises illocutionary force over the addressee, producing a perlocutionary
effect on her. It is therefore impossible to distinguish it at the level of the
structure from the most common and consensual speech acts, except as regards
the content of the illocutionary force, which is agonistic. Habermas thus finds
himself confronting a manifest counter-example, which calls into question
the universality of his consensual structure. And he gets out of it by a sleight
of hand. He denies that threats possess an illocutionary force (they have ‘lost’
it) and postulates that they are a kind of indirect speech act, in which the
agonistic character is the second product of an irenic illocutionary force
exploited for the purposes of strategic action. But, in addition to the fact that
it seems difficult, in any imaginable context, to attribute an irenic ‘illocutionary
meaning’ to ‘if you don’t shut up, I’m going to smash your face in’, there is
clearly a slippage in Habermas as regards the concept of ‘illocutionary force’.
In Austin and Searle, the term refers to the force exercised by the speech act:
it can just as easily be agonistic as irenic. And they distinguish illocutionary
force from perlocutionary effect: the force exercised by my threat can produce
various effects – fear, indignation, amusement – depending on the power
relationship with my addressee. But Habermas reduces illocutionary force to
perlocutionary effect: his ‘force’ is in fact nothing other than the goal of a
consensual effect, it is determined by it, and is therefore consensual, since the
intended effect is always to secure the listener’s rationally motivated agreement.
We can see why threats cannot be a speech act, but only the exploitation or
travesty of a consensual speech act. The argument is manifestly circular: it
claims to discover in speech acts a consensual interlocutory structure, but
only counts as speech acts those of them that conform to this structure. This
leads to counter-intuitive analyses: threats exercise no illocutionary force
because, for Habermas, this ‘force’ is not a force exercised, but an attempt at
agreement.
Habermas is therefore condemned to deny certain phenomena, since he
excludes all agonistic speech acts – threats, insults, various forms of aggression.
These are not the only ones. Habermas is an exception among his colleagues
(from Derrida to Lyotard, from Heidegger to Gadamer) in that he very rarely
Critique of the Philosophy of Language • 53