When Failure is Success: Counter-Performative Speech Acts

Andrew P. Porter

1993; AAR Western Regional Meeting, Santa Clara, March 1994
Formerly available by ftp at phil-preprints.L.chiba-u.ac.jp, in /pub/..../preprints;
this is revision 1 (format changes necessary for html, and punctuation errors corrected), 2001 July 30.


1. John L. Austin(1) saw that where assertive speech acts (to use Searle's term) are true or false, other performatives can be happy and successful, or they can be infelicitous, defective, or even outright failures. In Searle's systematization of speech act theory(2) the emphasis was always on articulating the conditions for non-defective success. Self-defeating speech acts were explored initially by Daniel Vanderveken,(3) but nothing was remarked beyond the logical structure of their failure as illocutions. It was not suspected that they could be effective and successful as perlocutions. To turn to such speech acts as successes is to turn from illocutionary theory to the study of perlocutions, but the perlocutions in question turn on their illocutionary structure, and so require attention to that structure. Such speech acts need not be vicious; irony is in some sense a self-defeating speech act.(4) When they are objectionable, remedying them usually requires dissecting their illocutionary incoherence; this is a problem, because the inconsistency of the parts of a compound and self-defeating speech act is usually concealed. (It must be concealed, if the illocutionarily self-defeating speech act is to succeed as a non-ironic perlocution.) Let us call self-defeating performative speech acts that work at some level as perlocutions counter-performative speech acts. When the counter-performative character of a speech act is obvious, it is ironic; when it is not, the speech act is usually pathological in some way. I look for the most part at speech acts with concealed counter-performative character.

2. The investigation of speech acts other than assertions has moved from the simple paradigmatic cases of promises and orders etc. to indirect speech acts, metaphor, fiction, and eventually to internally inconsistent speech acts. In the course of that exploration some quite peculiar peformatives have surfaced. It has been noticed that some performatives are analogous to self-contradictory statements,(5) such as "Do (not) obey this order," "I promise (not) to keep this promise," and others of similar construction. It does not matter whether the negative is present or not; these performatives are intrinsically infelicious, misfires. There is no way to keep this self-referring promise, or to obey this self-referring order. In a more applicable vein, Daniel Vanderveken has noticed what he called "self-defeating" speech acts.

3. To see how some counter-performatives work, consider the following. In a legendary example of a counter-performative, it is said that one of the Three Great Lies is

I'm from the government, and I'm here to help you.
This purports to be more than a statement, an offer of help, though it is incidentally also a statement. It is taken as a classic example of a lie, but the problem does not arise with its being contra-factual. S, the speaker, is in fact from the government. And he intends business in the life and affairs of H, the hearer. But not what the hearer would call "help."

4. It is performative, in as much as the social worker does something in saying it (by implication, he offers help), and it goes awry in ways that are characteristic of performatives that are not just assertions. This much has been noticed before, though this sort of utterance has not attracted much attention, but has been taken as a theoretically marginal and degenerate instance of performative language. Performatives that work were treated as more interesting than those that don't.

5. In fact, it does work: It does exactly what it is intended to do, which is to apply persuasion to the welfare client in a way that is more economical and more effective than candid reasoning, orders, or threats. (And if there is no compliance, the one making this "offer" appears to be in a much better position to apply coercion.) While appearing to be an offer, an offer of help, this utterance is in fact not an offer at all, but a form of pressure, manipulation. It is a performative that purports to do one thing, but in fact does something quite opposite: a counter-performative. Its effectiveness, its performative force, requires its counter-performative sense; its perlocutionary effect of being coercive pivots on its illocutionary appearance of being an offer of help, and on the silent failure of at least some of the conditions for the non-defective performance of such an offer.

Preliminary Expansion of the Thesis

6. The notion of utterances that are systematically and intentionally counter-performative (if perhaps not consciously so) has not attracted focused attention. But the problem has already been encountered outside of speech act theory. Popular psychological literature focuses, if without philosophical precision, on the counter-performative discourse that creates dysfunctional family structures. In a related way, the philosophical literature on self-deception does achieve precision, though without explicit attention to the speech-act theoretical features of such discourse. Examples can be found also from law, politics, and religious apologetics.

7. While counter-performatives are not theoretically central to the logic of illocutions, they are crucial to the pragmatic understanding of the same utterances when considered as perlocutionary acts. Formal performatives, whose illocutionary sense cannot be twisted after the fact, are a defense against counter-performatives. Formal performatives commit the speaker in one way or another, whether sincere or not, and sometimes even without happy preparatory conditions. It is because of the generally understood possibility of counter-performatives that formal performatives are necessary at critical commissive junctures in life.

8. The speaker who engages in counter-performative discourse knows how this sort of speech act works, even though he may not be willing or able to spell it out or explain it.(6) He has the skill of counter-performative speech acts, included in which is the opposition between the ostensible illocutionary force and the probable (and intended) perlocutionary effect. All this may be "unconscious" -- he does not spell it out to himself -- but it is still done with great skill, and so has to be accounted as intentional, responsible. In no way does the skill of counter-performative speaking require being able to explain (even to oneself) that one has misfired in one performative act, and has instead effectively performed some other speech act. It is not that the illocutionary force has been literally transformed. But when the speech-act turns on its implications, by way of filling the preparatory conditions for yet other speech acts, its perlocutionary working may indeed not only extend beyond but in fact be in conflict with its illocutionary force. Indirect speech acts, as Searle has observed, are accomplished when the conditions for one speech act are supplied in the performance of another.(7) If a statement or question provides the preparatory conditions for a request or other directive, it may be taken as such. The essential condition for a stronger directive is satisfied, and by convention, the question "Can you pass the salt?" counts as a request to do so. Thus an apparently simple speech act may, in its implications, count for much more.

9. Searle and Vanderveken list a variety of ways in which a compound speech act can become internally inconsistent, self-defeating. The success of one member of the compound may be inconsistent with the illocutionary point of the other.(8) One member of a compound may be inconsistent with the mode of achievement of the other: one cannot simultaneously command and plead with another to do something. And one member of the compound may be inconsistent with the propositional content of the other, or the presupposed (preparatory) conditions of the other, or with the psychological state required by the other. We shall see general circumstances in which each of these modes of counter-performative speech can be highly effective.

10. It is true that self-defeating speech acts are such by virtue of inconsistency, but the inconsistency can arise in various ways, which may be noted at this point, and which will appear in the discussion that follows. The model for a counterperformative is a compound speech act in which the several members are inconsistent. In one way or another, I think all counter-performatives can be rephrased as compounds. A speech act may not be an explicit compound; crucial parts may be only implied, or enacted only by indirection. In addition to patently compound illocutionary acts, simple acts may be counter-performative in the failure of a condition; that condition presumably can be spelled out, thus supplying the missing element of what would then be a compound speech act. The implied additional speech acts may arise from features internal to the uttered speech acts, or they may arise only given the context known to both speaker and hearer. Complex speech acts are possible in which multiple speakers participate, and the cumulative implications are counter-performative, because one speaker appears to presuppose agreement with what the other has said. When all the implied but unstated members of a speech act are spelled out, most simple failures and multi-speaker acts can be understood as compound speech acts whose several parts are inconsistent.(9)

11. In what follows, I shall consider three major examples of counter-performative speech acts; the one with which we began (inconsistent illocutionary points), one of the form "do A & don't do A" (inconsistent propositional content), and lastly one that arises in a complex interchange between a telephone salesman and his prospective customer. The critical importance of context will emerge: it is necessary to supply context simply in order to make it plausible that people could say things such as these, and especially in order to show how such counter-performatives could be effective; in each case, in ways concealed from at least one conversant. In reliance on context in demonstrating the perlocutionary workings of such speech acts, we are on the threshold of pragmatics. The first two speech-act examples will instantiate some of Searle and Vanderveken's catalog of self-defeating speech acts; the conversation with the telephone salesman is less clear, though I will hazard some taxonomic guesses for it. At this point, rather than search for more examples to complete a bestiary of counter-performatives, it is more useful to return to Searle and Vanderveken's own catalog, and draw some elementary inferences from it by way of general recommendation for what to look for in ferreting out counter-performatives in the wild.

Inconsistent Illocutionary Points

12. Now it is possible to re-examine the example with which we began, "I'm from the government and I'm here to help you." Both the social worker and speaker, S, and the welfare client and hearer, H, know that H is in trouble, that H has no bargaining power, the appearance of no reasonable options, except, that is, to accept the "help" that S offers. It is background information that is played upon in the counter-performative, and it characterizes one act as another: what it calls help is meddling, interference, directing the life of H, forcing H to comply with the requirements of the Welfare Department, imposing S's hierarchy of ends and order of means on H. It is persuasive because the social worker is here to take charge; that is the way welfare works. This persuasion is effectively coercive, because of the limited options of the prospective welfare client. As such, it is directive. In effect, the inconsistency can be exhibited quite simply: "I'm from the government [the preparatory condition for a directive, reminding the hearer of the government's intrinsic power to coerce], and I'm here to help you [a commissive whose illocutionary force is inconsistent with coercion]." A social worker once admitted the truth, when dealing with elderly clients, for whom "help" means total loss of independence in a nursing home: "I can't tell you the countless numbers of people we've had to, well, brainwash to get them to accept services. They always think it's a step toward nursing homes."(10) In choosing the word "brainwash", the social worker has come as close as it is possible to do without using the technical language of speech-act theory to admitting that his speech is counter-performative.

13. Human life in any context is to a great extent a matter of commissives and directives, especially the latter. Institutional structures and power relations are largely a matter of directives. It seems to be a feature of present-day rhetorical life that the most effective way to accomplish a directive, and to elicit compliance, is to disguise it as an offer of help: a distinctly different sort of speech act. In this sense, the familiar "Can I help you?" does not mean what it says, but rather, "You are on my turf, I will lead you through it, you will do as I say." This locution has migrated from the script in which the speaker is a salesperson, nominally at the service of the hearer, a prospective customer, to any situation that can be characterized as an encounter between an insider and an outsider. The illocution "Can I help you?" by presupposition asserts a claim of power and dominance, and asserts the relations of insider and outsider; quite the opposite of its ostensible meaning of service and subservience.

Inconsistent propositional content: "Do A & Don't do A"

14. It is difficult to believe that one could utter a performative of the form "I command you to do A and I forbid you to do A" to any useful perlocutionary end. But the only thing standing in the way of the usefulness of such a counter-performative is its transparently obvious character. Searle and Vanderveken argue that ". . . a speaker cannot perform an act of illocutionary denegation of the form not-A at a time when he performs an act that commits him to A."(11) But it is only true that the speaker cannot consistently illocute an utterance of this form; the same utterance, considered in its perlocutionary aspect may depend precisely on its illocutionary failure. A speaker may find it advantageous to appear to commit himself to A, when his real commitment is to not-A; it is possible to craft utterances which do just this. I once heard in public debate the following:

(a) I respect the religious views of those who disagree with ballot initiative X,
(b) and they should vote against the initiative;
(c) at the same time, I request that they not impose their morality upon others.
I have tagged the three parts of this performative (a), (b), and (c). In context, it was implied that to vote against the initiative was exactly to impose one's morality on others. The initiative was put on the ballot to promote a practice that some had moral objections to. Why does this counter-performative work so well? (b) is consistent easily enough with (a); but (b) is radically incompatible with (c). If (c) is accepted as sincere, (a) cannot be. Respect for another's commitments involves admitting those commitments to deliberation and debate; (c) is precisely an attempt to rule those others' commitments out of order. (b) implies a directive, "vote against the initiative". In this implication, the conjunction (b)&(c) is precisely of the form "do A & don't do A." My conjecture as to how this performative works is that (a) allows those potentially in agreement with the initiative X but not yet firmly convinced to see themselves as not in the sort of disagreement that actually requires a difficult decision. Instead, they are offered a compromise that ought to satisfy both parties, by offering to their opponents all that their opponents could reasonably ask for (which by implication does not include allowing their opponents to vote their consciences). Such potential supporters of the initiative could then in easy conscience vote for what the speaker hoped they really wanted all along anyway. When this sort of counter-performative occurs, (b) is usually ommitted, in as much as it tends to give the game away.

15. This utterance can have different strategies for its two classes of hearers. It is probable that while the apparent intended hearers (i. e., those to whom it was addressed) were those opposed to the initiative (they were requested not to impose their views on others), the real intended hearers were those potentially in favor of it (the context was a public debate), and they were implicitly encouraged to conclude that they could give all they owed to their opponents and still vote for the initiative.

16. A more bald example is provided in legends of the city politics of Cambridge, Massachusetts, where the dividing issue is rents, and the parties landlords and tenants. Occasionally a politician tries to promise to lower rents for tenants and raise rents for landlords. It is reported that this platform meets more success than one might expect in a city of such sophistication. A counter-performative of the form "do A and don't do A" can be quite effective if the speaker faces conflicting demands from different constituencies; if he can appear to satisfy one, or at least neutralize its opposition, he can later gratify the other.

17. Applications of this form are hardly limited to politics, but politics and even court decisions seem to be particularly rich in them. Antonin Scalia, in a dissent in Johnson v. Transportation Agency, alleges that the Civil Rights Act of 1964 has been converted by court decisions "from a guarantee that race or sex will not be the basis for employment determinations, to a guarantee that it often will."(12) Merits of the case and of the opinion aside, this is a clear charge of a counterperformative interpretation of the law. That the law was even susceptible of such an interpretation raises the question whether not only the Court's opinion but also the statute itself is counterperformative. It has been informally observed that canny legislators, who know that the final determination of the meaning of a law will come only in the courts, from friendly and activist judicial review, insert contradictory material into the legislative history of a bill, planting grounds for judges to rewrite or even reverse the sense of the act as passed. In another example, Congressional appropriations for the Tellico Dam (which would have destroyed the habitat of the snail darter on the Little Tennessee River) continued while Congress deliberated on and passed the Endangered Species Act; the resulting counter-performative implications of the several acts taken together had to be resolved in court.(13)

18. Are there other reasons why a speaker could issue a performative that is implicitly and covertly of the form "do A and don't do A"? It relieves him of responsibility, no matter what the hearer does. Which member of the counter-performative conjunction is concealed will be chosen so as to maximally conceal the speaker's responsibility. If A is dangerous, and S does not want responsibility for harm to H, it is easy to imagine a plausible context for such a counter-performative.

19. Related to this is "I dare you to do A." It appears to be a directive, coupled with a disavowal of the responsibility that goes with a directive. This is, at a minimum, a somewhat peculiar illocutionary force. But how can a directive be issued without committing the speaker to some responsibility for his injunction to the hearer? It seems unlikely to me that a single word could in its own intrinsic explicit illocutionary sense be a counter-performative, but this verb raises such a suspicion. Certainly the prudent hearer is well-advised to treat it as a counter-performative.

Invitations to commit a counter-performative

20. In a common multi-speaker counter-performative, it is a grave strategic mistake to give a telephone salesman reasons for declining his offer. Who has not had a telephone salesman call, offering "Eight weeks of the Tri-Valley Gossip free, you pay only for the Sunday edition . . . "? After declining the offer, the recipient will next hear, "Why don't you want our bargain trial subscription offer?" It is a fatal mistake to give reasons at this point -- because the respondent (who received the call) and the speaker (the salesman) will treat the reasons quite differently. The salesman can always treat reasons as an offer to bargain, and reply with counter-reasons, and counter-demands for more reasons. But the respondent presumably has no intention of bargaining; he (or she) just doesn't want the paper, but he also wants to be polite. (Note the performative intentions!) But to give reasons at all is to presuppose that under some conditions, the recipient of the call would subscribe. When the recipient has initially declined the offer, the salesman's move in asking for reasons is to get the recipient of the call to concede exactly this presupposition. In effect, he has asked the recipient to treat his own refusal not as the starting point of a chain of practical reasoning, but as its end point. Out of politeness, the recipient usually obliges. But from the new starting point implicit in whatever reasons the recipient gives, the salesman can twist reasons to his own desired end point, a sale. In the end, if the recipient really doesn't want to subscribe, he has to say so, without giving reasons. The direct approach is so rarely taken with telephone salesmen that it tends to produce stunning results: "I'm sorry, I don't give reasons."

21. The salesman is inviting the recipient of the call to commit a counter-performative: that is, in giving reasons, to offer to bargain, when the recipient has no intention of bargaining at all. The logical form of such a request is, "please justify your starting point." This is to construe an argument from X as an argument to X. By definition, it is not possible to justify a starting point. Consider, for example, some of the commitments inherent in doing science: openness to criticism, empirical encounter with the world as it is (rather than with some platonic ideal world), science open to all and for all. These are not things that could be justified; least of all justified from the fruits they bring. Yet one could answer the question "Why these commitments?" by citing their fruits. Another could then interpret such an answer as an argument of expediency from the benefits of science, rather than as the confession of one whose commitment to science is axiomatic, and to whom these other benefits have been given in addition. But an argument of expediency can be modified or suspended at convenience. (It would have been better not to answer the question at all.)

22. To guess the structure of the self-defeating speech acts in this sort of a conversation, the problem appears to lie in the illocutionary force of the reasons given by the respondent to the salesman. The salesman is asking the respondent to make a commissive whose force is ambiguous, and so can be twisted: a confessional commissive has a force crucially different from that of an offer to bargain. My suspicion is that confessional speech acts (as in confessions of faith), have features which, despite all the attention they have attracted, still leave room for significant work.

23. At this point, we have seen clear examples of two sorts of illocutionarily self-defeating speech acts which are nevertheless quite effective as perlocutions, so long as their working can be concealed. The welfare worker engages in the appearance of making an offer, a commissive, while effectively issuing a coercive directive; in this context, the two illocutionary points are incompatible. (It is possible to combine a commisive and a directive in compatible ways.) The politician can find it rhetorically very effective to appear to come down on both sides of an issue (inconsistent propositional contents). The problems in the encounter with the telephone salesman arise in the subtleties of the illocutionary strength of the speech-act of giving reasons. Rather than exhibit examples of the remaining ways in which the illocutionary forces of members of compound speech acts can be inconsistent on analytic grounds, grounds of the form of the speech acts, it seems more economical simply to note the remaining possibilities for trouble, in as much as the problems which do arise usually do so only with respect to context; that is, they are not evident simply on the form of the speech acts, but the pathology shows itself only after knowledge of the pragmatic context.

24. The mode of achievement may be at stake in conflicted ways when an assertive takes on the color of a directive, and the role-authority of the one speaking is in question. Such problems arise in the psychological literature, especially as it treats dysfunctional family systems.

25. Speech acts presuppose a psychological state on the part of the speaker, and this too can be the locus of inconsistency. Searle's own definition is a most economical example:

"Finally, an illocutionary act whose preparatory conditions cannot be presupposed simultaneously with the expression of the sincerity conditions of another illocutionary act is also relatively incompatible with that other illocutionary act. For example, a speaker cannot both recommend that the hearer carry out a certain course of action and simultaneously complain under the same aspects that he will carry it out because one cannot consistently both presuppose that a course of action is good and express dissatisfaction about it under the same aspects and for the same reasons."(14)
But people do this all the time! As an illocution, such a speech acts are inconsistent, but as a perlocution, they can be surprisingly effective. The challenge is in dissecting from the conversation the ends to which it is effective. That virtually always depends on context, and it usually also turns on interpreting "under the same aspects" in apparently analogous but in fact conflicting senses.

26. By far the richest occasion for counter-performatives lies in the required presuppositions for speech acts. One can assert a description by presupposition, and so conceal the assertion. And one can insinuate, by presupposition, that the world is such that it can be treated as it is in the ensuing speech acts. Analogies may be distorted, one party may co-opt the position of another, maintaining the other's speech acts, but to quite opposed ends, and one party may put out a "conceptual Trojan Horse," which the other party accepts at its own risk. Obviously, these tend to be multi-speaker speech acts of considerable complexity, in addition to being intricately context-dependent.

27. Non-philosophers have had to deal informally with counter-performatives for some time, as the psychological literature and oral tradition in politics and law can amply testify. And philosophers other than speech act theorists have recognized the problem:

Instead of aiming directly at propositions believed or practices performed in order to challenge their truth or value, suspicion is aimed at the individual or community who believes and performs in order to challenge their integrity. It looks for that discrepancy between professed meaning and actual use which renders life ironical; for it is the essence of the ironical (speech) act that it performs a function quite at odds with its surface meaning. Thus an ironical compliment functions to express a criticism. For this reason suspicion is less interested in the official meaning of beliefs and practices than in their operative meaning, the clue to which is the life-world from which they arise and which, in turn, they legitimize.(15)
Searle's and Westphal's philosophical outlooks are worlds apart -- showing that Searle and Vanderveken's typology of illocutionarily self-defeating speech acts is a useful and precise analytic tool in dissecting perlocutionarily counter-performative speech acts far beyond its original philosophical home. It is surprising that Vanderveken's original insight has not found more use.


1. How To Do Things With Words, 2nd ed., (Cambridge: Harvard Univ. Pr., 1975).

2. Speech Acts: An Essay in the Philosophy of Language (London: Cambridge Univ. Pr., 1969); Expression and Meaning (Cambridge Univ. Press, 1979). Terrence Tilley's The Evils of Theodicy (Washington, DC: Georgetown University Press, 1991) was my guide to some of this literature, and Tilley has been helpful in private correspondence.

3. "Illocutionary Logic and Self-Defeating Speech Acts," in John R. Searle, Ferenc Kiefer and Manfred Bierwisch, eds., Speech Act Theory and Pragmatics (Dordrecht: D. Reidel, 1980), and John R. Searle and Daniel Vanderveken, Foundations of Illocutionary Logic (Cambridge University Press, 1985), esp. pp. 148 ff.

4. When Searle analysed irony, in "Metaphor" (in Expression and Meaning, esp. pp. 112-116), he defined it from the hearer's knowledge in context that the speech act is to be interpreted in a sense opposite to its ostensible meaning.

5. George Lakoff, "Performative Antinomies," Foundations of Language 8 (1972) 569.

6. Herbert Fingarette explores the paradoxes of having an "unconscious" policy of not spelling out an engagement with life in Self Deception (London: RKP, 1969).

7. "Indirect Speech Acts," in Expression and Meaning (Cambridge University Press, 1979).

8. This and following types of pathology are from Searle and Vanderveken, pp. 148-152.

9. Contextual implications may be further complicated by the fact that one or both of the speaker and hearer are able to deal practically with the context in life (though maybe not happily so), but are unable to spell out all of its revelant features.

10. Wall Street Journal cxxvii, no. 110 (1992/12/03), p. 1.

11. Searle and Vanderveken, p. 153.

12. 480 US 616, at p. 658.

13. See TVA v. Hill, 437 US 153; the chronology of events can be found at pp. 153 and 197. I am indebted to Phillip E. Johnson for notice of these cases.

14. Searle and Vanderveken, p. 150.

15. Merold Westphal, "Phenomenologies and Religious Truth," in Phenomenology of the Truth Proper to Religion, ed. Daniel Guerrière (Albany: SUNY Press, 1990); p. 120.