Originally, the term "illocutionary act" was introduced by JL Austin. Austin uses the term for actions which are characterised by two features:
(1) "Illocutionary acts" are 'conventional' in that their performance involves the creation of (what Searle later calls) 'institutional facts'.
(2) "Illocutionary acts" are acts which involve what Austin calls the "securing of uptake", which amounts to the actor's communicating the fact that the act is (being) performed to someone.
At first, this may strike us as a rather strange combination of features. But closer analysis reveals that a rather great number of actions have these two features--including offering, buying, marrying, baptizing, bequeathing, betting, along with many other common-or-garden action types.
So these two features often occur together; they are characteristic of a certain peculiar kind of action of its own. Oxford philosopher and early "speech act theorist" PF Strawson apparently overlooked this fact, however. In his article "Intention and Convention in Speech Acts" he argues that ...
It seems perfectly clear that, if at least we take the expressions "convention" and "conventional" in the most natural way, the doctrine of the conventional nature of the illocutionary act does not hold generally. Some illocutionary acts are conventional; others are not. (445)
(1) On the one hand, Strawson argues, there are cases of attempted communication in the sense of Grice's (1957, xxx) analyses of 'speaker meaning'. In such cases, Strawson argues (quite plausibly), the act performed is "not essentially a conventional act" (456).
(2) On the other hand, there are also acts which actually "are essentially conventional" (457), cases where it actually is "as conforming to an accepted convention [...] that [the] act is performed" (443).
Strawson suggest calling both types by the name "illocutionary act". Thus, roughly, he suggests to use the two criteria which in combination characterise Austin's "illocutionary act" singly, define by either of them one separate variety of "illocutionary acts".
Strawson thereby polysemizes the term "illocutionary act" in a double way. He abandon's the definition which Austin had given.
Austin: That an act is "illocutionary" means that its performance involves the creation of (what Searle later calls) 'institutional facts', and that its performance requires the "securing of uptake", i.e., the communication of the fact that the act is (being) performed.
In the place of this sense of being "illocutionary", Strawson introduces two new senses:
Strawson I (roughly): That an act is "illocutionary" in the first sense means that its performance involves the creation of (what Searle later calls) 'institutional facts'.
Strawson II (roughly): That an act is "illocutionary" in the second sense means that its performance involves the speaker's meaning something (in the sense famously analysed by Grice).
...............
ERROR: That "illocutionary act" refers to acts of speaker meaning.