Andreas Kemmerling's approach to "illocutionary acts" follows the general idea that these acts and speaker meaning in the sense of Grice (1957) are the same thing. Yet Kemmerling does not accept the definition given by Schiffer (who follows the same idea). Nor does he comply with Austin’s definitions (for obvious reasons: Austin does not share the general idea). Indeed, Kemmerling warns us not to conflate “linguistoid communication”, which on his account amounts to Gricean 'speaker meaning' plus being understood, with “the class of actions which Austin had in mind when he spoke of ‘illocutionary acts’” (Kemmerling 2001: 76 f.).
Now, for whatever reason, Kemmerling seems to presume that the term "illocutionary act" should be defined as directed at "linguistoid communication", rather than to that class of actions which Austin had in mind. He thus continues to give his own definition of "illocutionary acts"--as cases of "linguistoid communication".
For an action type to be ‘illocutionary’, he suggests, is for it to be “gricy”. For an action type x to be "gricy", again, is to satisfy the following condition: To make it clear that the acting person wants (or intends) to x in doing what they do, is, by conceptual necessity, sufficient for their therein x-ing (Kemmerling 2001, 84).
------------------------------------------------------------------------------------------------------------------------------
Kemmerling, Andreas, "Gricy Actions", in G. Cosenza (ed.), Paul Grice's Heritage, Brepols 2001, 69–95.