I gave a talk on this topic at Fictionalism. The Bucharest-Budapest Workshop in Philosophy, a few days ago. My slides are here. A first draft of my paper is here. Some of the notes I have made while working can be read here (although they will not make much sense by themselves).

I have received some objections with respect to my PP1 (“In order to pretend to be doing A one must not be doing A”).

1) Daniel D. Hutto: An actor can play the role of someone being an actor. So the actor pretends to be an actor, although she/he actually is an actor. So one can pretend that X is the case even if X is the case.

Reply: Perhaps this is possible due to the fact that the description which we provide for what the actor is doing is incomplete. Let C be our actor, pretending to be D in play E. Now, if D is the character of an actor, then D would either play no role, or she would play role F in play G. According to PP1, C cannot play the character of an actor playing D in E. So PP1 holds.

Of course, one could ask: “What if C would utter ‘I am an actor.’ in the play? Wouldn’t C be saying something true?”.

To this the answer would be: No, since “I” in that utterance would not refer to C.

“What if D was real and an actor?”

We could still compare the case when C says “I am an actor” while pretending to be D (who does exist) in a play with the case when C says “I am an actor” while translating what D says in another language. In the second case, “I” stands for (or refers to) D. What is said is true iff D is an actor (if it is false, then it was D the one who lied, since C has no commitment to the truth of what she says). In the first case, according to SFF, at least, “I” does not refer to anybody.

“Right, but PP1 still seems to be false for the first case.” According to PP1, if it was true that D is an actor, C shouldn’t be able to pretend to assert that D is an actor, but only to actually assert it.”

The reply could be, again, that it matters how we describe A. We could describe A as a generic action or as a particular action. The second type of description should be of the form: “this and that, done by X”. That is, if we consider A under a description which also specifies the agent performing the action (it could be argued that we must do this since actions are not individuated in the same way in which events are), we would say: “C is pretending to perform the assertion that D is an actor done by D and she is able to do this (to pretend to perform the assertion etc.) precisely because she is not actually doing it.

“What about the case in which C says some simple truth during a play?”

The case is similar. The action performed by C is to pretend to perform the assertion of that simple truth by someone else.

“What if C was acting in a play or a movie about her own life, playing her own character?”

I am not sure that we should call what C was doing “pretending”. Anyway, the next objection I want to discuss might throw some light on such a case as well.

2) Matti Eklund: NT (from my analysis of Revised Force Fictionalism: “Z pretends to assert that p in uttering S only if p is not true”) does not hold (so PP1 and the first part of PP2 – “Given the purpose X of an action A, in order to pretend to be doing A one must not achieve X” – also do not hold). Suppose my neighbor is an alien and in order to deceive my friend I blatantly pretend to assert “My neighbor is an alien”. It seems that I succeed to pretend to assert that my neighbor is an alien even if it is true that my neighbor is an alien.

My first reply to this was to distinguish between fiction producing pretension and pretension which does not necessarily produce any fiction, but Matti replied that it is the second type of pretension that could be more important for a fictionalist (and even a force fictionalist). Perhaps it could also be said that my pretended assertion that my neighbor is an alien also produces a fiction (although via a different pragmatic mechanism and in the manner of deceptive pretenses), namely the fiction that my neighbor is not an alien.

In any case, now I think this case could be better described by saying that I do not actually pretend to assert that my neighbor is an alien, but pretend to pretend to assert it. Since pretending to perform A is itself an action, there is no reason why one could not pretend to perform such an action.

In the same way, one could say that if C is an actor who plays her own role in a movie, then C pretends to pretend to be C, but does not pretend to be C.

This is a way in which such cases could be conceived without giving up PP1. However, if this move is not strong enough to escape counterexamples of this sort (someone could perhaps devise a case in which one was pretending to pretend to do A while pretending to do A), then a distinction between the kind of action performed and the manner in which an action is performed might be useful.

‘Pretense’ could, then, either mark the kind of action one is performing or the manner in which one was performing an action. For instance, I could either pretend to undress myself (while keeping my clothes on), or undress myself in a pretended manner (as if I was doing a striptease or with exaggerated gestures, or with the gestures of a particular person I would imitate, or even with the gestures of a caricature version of myself).

It would be clear, then, that I am interested only in cases of the first and not of the second kind.

3) Matti Eklund also had an objection directed at the first part of my PP2 (Given the purpose X of an action A, in order to pretend to be doing A one must not achieve X). The main idea, if I understood him correctly, was that goal X could be achieved by performing a different action than A and one could still be pretending to perform A. For instance, suppose two different switches close to each other could turn on the light in a room. I pretend to turn on the light, so I move my hand near the first switch but touch the second one and actually turn on the light. So I did not perform the action which I was pretending to perform (turning on the light by using the first switch), but the purpose of the action I was pretending to perform was nevertheless achieved. I find this case quite interesting, since it brings out the difference between PP1 and PP2.

The case can be brought down to two different sub-cases:
(a) I accidentally touch the second switch.
(b) I touch the second switch on purpose.

Sub-case (b) seems to be excluded if my aim was to pretend to turn on the light. However, case (a) still points to a situation in which one seems to be able to pretend to do A although the goal of A is achieved. I think, however, that the impersonal formulation used here (contrast “the goal of A is achieved” with “one achieves the goal of A”) holds the key to my answer.

I still have to think about this, but I suspect that the difference between the way I see such situations and the way Matti sees them boils down to the difference between using “pretense” to mark a kind of action and to use it to mark the manner in which an action is performed.

If we were talking about illocutionary acts, for instance, the difference between an actor uttering p on a scene (pretending to assert that p) and in her daily life (asserting that p) would be, according to the view I attribute to Matti, not a difference between two illocutionary points (or residing in different propositional content, sincerity or preparatory conditions), but rather a difference between two modes of achievement (in the same way in which a request and a command differ not in their illocutionary point or other conditions, but in their mode of achievement).

4) My argument against SFF does not essentially depend on PP1 or PP2(a), but on PP2(b). However, in developing it I focus on a force fictionalism based on Searle’s theory of fictional talk (pretending to assert that p is not to perform any other illocutionary act, but only to pretend to perform one). For this I have argued that the proposals made by Currie and Martinich cannot support a version of SFF, since they allow the content p to be “presented” (speaking like Frege from Das Gedanke), so then one could ask “What if one actually asserted that p?”.

Matti’s reply was that one could still be a standalone force fictionalist and accept that p is presented in a way when one pretends to assert that p, while answering my question by saying: “If I was actually asserting that p, then p would have the ontological commitment it usually has (so we would not need content fictionalism to analyze p). Luckily, I do not assert that p.”

My initial reply was that if the fictionalist would give such a reply, then she could say that if p was properly asserted, either (i) p was false, or (ii) p was meaningless. If the fictionalist says (ii), then my argument against SFF should apply to such a case as well (since my argument shows that one cannot pretend to assert that p if p is meaningless). If the fictionalist says (i), then she still needs some form of content fictionalism to show how can p be false even if some of p’s assumptions do not hold.

If I remember well, what Matti said was that the fictionalist could reasonably claim that p is not meaningless in the same way in which a word salad or “Green ideas sleep furiously” are meaningless.

This is something I still have to think about. However, the worst case scenario would be that my argument is effective only against Searlian SFF (although I still suspect that the theories of fictional discourse provided by Currie and Martinich cannot offer good support for a SFF view).

LE: 5) Iulian Toader suggested that a version of the Frege-Geach problem could be used against SFF, thus rendering my argument useless. In short, the Frege-Geach problem seems to force the standalone force fictionalist to start speaking of the content she pretends to assert, since D (which she considers fictional) can also contain modus ponens inferences. Perhaps the standalone force fictionalist could still reply that they are not proper inferences, but only resemble inferences (in the same way in which a move in the pretended game of chess might seem to have been made because it was a necessary condition for shouting ‘Chessmate!’, for instance). I still have to think about this.