O'Rourke, Searle on Speech Acts

Searle creates the concept of rules created from utterances, and divides rules into constitutive and regulative.

Constitutive rules are rules which, if you changed them, would "change the game". The existence of an activity is logically dependent on these rules.

Regulative rules are rules which changing would not "change the game", i.e. an activity doesn't logically depend on them specifically; in fact the rules' existence might depend on the activity's shape, i/.e. on the constitutive rules.

Searle's example is traffic on a two lane road. Examples of each type of rule:

• constitutive: everybody going in a particular direction has to drive on the same side (which each driver will always experience as either left or right from the vehicle's frame of reference).
• regulative: everybody has to drive on their respective left side.

One might notice that this is a way to factor rules-as-stated: in the above case, usually the constitutive part of the rule is implicit in the regulative part, because:

• the regulative part is necessary for it to be a coherent ruleset - i.e. once you've decided that everybody drives on one side, you have to pick a side in order to make the ruleset follow-able.
• the regulative version of the rule implies the constitutive part.

So it's easy to package these things together, and the separation is kind of artificial.

The constitutive part of a ruleset is what makes up the frame in which a further discussion occurs, however. This is because:

• Logical compatibility is what makes rules compatible or incompatible.
• Extensions to a given ruleset are easy or hard depending on their compatibility with constitutive rules specifically.
• A ruleset's applicability - what I've above called its follow-ability – can be thought of as the soundness and completeness of its relationship to the agents and environment it's trying to describe (NB: norm? Instrument? Perform things about? Attach to? What's the correct verb here?). This is wholly derived from its constitutive rules.

How you factor a rule into constitutive and regulative parts is not straightforward or objective. For one thing it's dependent on the presumed ontology of the world you're trying to write rules about – presumptions/axioms which constitutive rules have baked into them, which can't be factored out except in relation to some other worldview. It's language games in every direction, baby!

For the same reasons, these same problems arise for action under a description.

The lexical content of an illocutionary speech act. The thing that can participate in syllogism. Attaches to reality by describing it. Actually it attaches to a virtual reality, a conjuration in the speaker's and listener's imaginations. What I elsewhere call unreality.

How do you intend your utterance to attach to reality? Basically a generalization of the Gricean account of meaning (which is, "does your utterance attach cleanly and cooperatively to reality?").

(Hmm. Grice does seem to say something like "being honest is being nice." Is this a downstream or upstream claim to the theory?)

When you say a proposition in order to convey the intention of rendering it true, and therefore constraining at least yourself to behave in ways that aim towards that outcome, that's a promise. A signaled intention – description as part of the action.

The lecture notes go into the logical structure of promises under this framework.

You kind of have to learn, and verify, the grammar of your mother tongue, and of your idiolect – subjective scientific investigation. You've got to take a similar approach to looking at rules which arise from performative speech – by observing the context of the speech, the downstream consequences, and the relation of the lexical meaning to the action which seems to have been performed, you arrive at the rules as they have been "performed into being". You can then factor these using contextual information to get the constitutive and regulative factor rules.