In J. L. Austin’s theory a “speech act” is an utterance that achieves some action by virtue of its being spoken, “I thee wed,” “I hereby open the 23rd International …”, etc. but not “The king of France is bald,” “Did you go to the party yesterday?”, etc.
I have recently witnessed an inclination at workshops in religious studies to refer to magic as speech acts. The motivation is to not dismiss magic as hocus pocus, but instead take an emic perspective; the effort fails. To claim that hoc est corpus … constitutes a speech act is to admit that nothing essential happens to the bread and thereby rejects the emic. The more respectful participant observer would admit the magic is magic and admit it works.
Linguists have stretched “speech acts” yet further; to them it is simply a synonym for ”utterance”. Speaker and hearer are dubbed “speech act participants”, which, in the perennial quest to make prose opaque, become SAPs. This choice of terminology is particularly unfortunate because in Austin’s formulation a third person, such as the witness at a wedding, is often a participant in a speech act. If somehow Linguists cannot bear the tired old words “speaker” and “hearer” they could try ”interlocutors”, but please let Austin have speech acts, without attempting to bathe in his reflected glory.
Here is Haspelmath in support of your preference https://dlc.hypotheses.org/2430?fbclid=IwAR059Ato5AqZZtSspaOjsVOI4h7U-6PLqs40_WH2idvBpgsgilNkkQa7R7c
Many speech acts, in a narrow sense, qualify as such because they constitute a legal or quasi-legal commitment (“I swear”, or the rather unceremonious “deal”).
Now aren’t _some_ magic spells meant as treaties with spirits or deities?
The example of an oath is particularly striking:
We may today consider it a legal act. However, I understand that it originates from a conditioned self-curse (“if this is not true, then …”).
Insofar this applies, magic words are contracts or offers, directed to spiritual addressees, and therefore speech acts in a narrow sense as comparable deals were with human partners.
Of course, this says nothing about the case where the spell constitutes a plea, or a command, or is considered as a measure that “mechanically” has an effect on the world.
> Linguists have stretched “speech acts” yet further; to them it is simply a synonym for ”utterance”.
A major point Austin makes in How to Do Things With Words is precisely that all utterances are performative speech acts. He only sets performatives apart in the early part of the book. As he elaborates, he argues that saying, for example, “There’s a cat by the window” is also performative (you’re “doing something with words”: warning your friend, or trying to get them to close the window, or drawing their attention, or…). So is “The king of France is bald”.
It is all true. Nonetheless, the use ‘speech act’ to mean ‘act of speaking’ rather than ‘speaking which constitutes an act’ still shows a discontinuity with Austin. In particular, even if all utterances are speech acts to use ‘speech act participant’ as an equivalent for ‘speaker and hearer’ is inappropriate since third persons can and do participate in speech acts, as I say in the post.