DRAFT VERSION [to appear in: Dialogue and Universalism, 1, 2013] Maciej Witek THREE APPROACHES TO THE STUDY OF SPEECH ACTS* ABSTRACT The paper reconstructs and discusses three different approaches to the study of speech acts: (i) the intentionalist approach, according to which most illocutionary acts are to be analysed as utterances made with the Gricean communicative intentions, (ii) the institutionalist approach, which is based on the idea of illocutions as institutional acts constituted by systems of collectively accepted rules, and (iii) the interactionalist approach, the main tenet of which is that performing illocutionary acts consists in making conventional moves in accordance with patterns of social interaction. It is claimed that, first, each of the discussed approaches presupposes a different account of the nature and structure of illocutionary acts, and, second, all those approaches result from one-sided interpretations of Austin’s conception of verbal action. The first part of the paper reconstructs Austin's views on the functions and effects of felicitous illocutionary acts. The second part reconstructs and considers three different research developments in the post-Austinian speech act theory—the intentionalist approach, the institutionalist approach, and the interactionalist approach. Keywords: Austin; illocutionary acts; communicative intentions; constitutive rules; verbal interaction. 1. AUSTIN ON THE FUNCTIONS, EFFECTS AND CONVENTIONALITY OF ILLOCUTIONARY ACTS According to John L. Austin, describing the world in not the only, or even not the central aim of language; we use words to do things, i.e., to perform certain actions that bring about changes in our social environment. For example, we (a) inform our interlocutors that something is the case, warn them against * Work supported by Tadeusz Szubka’s research grant „Filozofia analityczna: historia i najnowsze wyzwania” from the Foundation for Polish Science (Programme „Mistrz”). DRAFT VERSION certain perils, ask them to do something, promise them to do something, greet them, apologize them for our incorrect behaviour, and so on; we also (b) get our interlocutors to believe that something is the case, cause them to be on the alert for certain things, get them to do something, cause them to expect us to do something, insult or amuse them, get them to feel sympathy for us, and so on. According to Austin, the actions listed in point (a) are illocutionary acts; their function is to produce or—as Jacek Jadacki (Jadacki 2010, 117–122) puts it—to generate conventional or normative states of affairs. The actions listed in point (b) are, in turn, perlocutionary acts; to perform a perlocutionary act is to “produce certain consequential effects upon the feelings, thoughts, or actions of the audience, or of the speaker, or of other persons” (Austin 1962, 101). In short, both the illocutionary and perlocutionary acts change the context of their performance. However, the former, unlike the latter, have conventional rather than natural effects. In other words, one could not perform an illocutionary act—and, by the same token, one would not be able to produce its normative effect—if there were no “accepted conventional procedure having a certain conventional effect, that procedure to include the uttering of certain words by certain persons in certain circumstances” (Austin 1962, 14); but one could produce what counts as the perlocutionary consequence of one's utterance even if there were no language and linguistic conventions. The effect of an illocutionary act is necessarily conventional: it would be impossible to produce it if there were no accepted conventional procedures for performing the act in question. By contrast, what counts as the perlocutionary consequence of a speech act is a natural state of affairs that happens to be produced by performing a conventional act. To make an illocutionary act, then, is to change the context of its performance by producing certain effects. The effects so produced, let us stress, should be carefully distinguished from the act’s perlocutionary consequences. According to Austin, the illocutionary act—as distinct from the perlocutionary one— affects the context of its production in the following three ways: first, (e1) “the performance of an illocutionary act involves the securing of uptake” (Austin 1962, 116) on the part of the hearer; second, (e2) the “illocutionary act ‘takes effect’ in certain ways, as distinguished from producing consequences in the sense of bringing about states of affairs in the ‘normal’ way, i.e. changes in the natural course of events” (Austin 1962, 116); third, (e3) “many illocutionary acts invite by convention a response or sequel” (Austin 1962, 116). In short, there are three types of effects that can be ascribed to successful illocutionary acts: (e1) the securing of uptake, (e2) the taking of effect and (e3) the inviting of a response or sequel. Generally, securing uptake consists in getting the hearer to recognize the force and meaning of the speaker’s utterance. Normally, the securing of uptake involves the hearer’s forming a conscious mental representation of the speaker’s act. For an act to take an effect, in turn, is to produce certain normative states of affairs conceived as the commitments, obligations, entitle- DRAFT VERSION ments and rights of the communicating agents. For instance, a successful promise results in the speaker’s being committed to do the action she refers to (Austin 1962, 102) as well as in the hearer’s being entitled to expect the speaker to do this. A binding act of ordering, in turn, creates the hearer’s obligation to do what he is told and the speaker’s right to expect the hearer to do what he is told. What is more, the successful illocutions under discussion produce effects of the (e3) type. For instance, the binding promise invites the speaker’s response of fulfilment and the felicitous order invites the hearer’s obedience or at least his explanation why he cannot comply with the speaker’s order. The mechanisms whereby the effects (e1), (e2) and (e3) are produced involve, in one way or another, operating of illocutionary conventions. Austin claims that „the illocutionary act is a conventional act: an act done as conforming to a convention” (Austin 1962, 105). Unfortunately, he offers no systematic analyses of the concept of convention that could be used to ground his account of illocutionary acts and their effects (Harnish 2005, 13–14). Let us assume, however, that what he has in mind by claiming that illocutionary acts are conventional is the rough idea that the performance of an illocutionary act involves the invocation of a conventional procedure that has a conventional effect and includes the employment of certain conventional means. Assume, next, that a procedure is conventional if it is commonly accepted; by the same token, an effect is conventional if it is tacitly and collectively accepted by the interacting agents; finally, some illocutionary acts are conventional because they are performed with the use of conventional means—performative formulas, ritual phrases, grammatical moods—whose function is to indicate the illocutionary force of an utterance. There are, then, at least three different concepts of conventionality: procedure-conventionality, effect-conventionality, and means-conventionality. It is not clear which one of them Austin has in mind when he claims that „the illocutionary act is a conventional act.” Is it conventional be cause of the conventionality of the procedure invoked by the speaker, because of the conventionality of its effect, or because of the conventionality of the means by which it is performed? I return to these questions in section 2.2. 2. THE POST-AUSTINIAN SPEECH ACT THEORY AND THE DISPUTE ABOUT THE NATURE OF ILLOCUTIONARY ACTS There are three research traditions in the post-Austinian speech act theory: intentionalist, institutionalist, and interactionalist. Each of them seems to result from a one-sided interpretation of Austin’s idea claiming that illocutionary acts are context-changing actions. Austin defines illocutions as conventional acts that affect the context of their performance by producing three types of effects: (e1) the securing of uptake, (e2) the taking of effect, and (e3) the inviting of a re- DRAFT VERSION sponse or sequel. According to the intentionalist approach, most illocutionary act types—such as statements, warnings, promises, requests, and so on—are communicative rather than conventional. It is claimed, namely, that to perform a successful illocutionary act of the communicative sort is to issue an utterance with a communicative intention, i.e., with the intention to produce the effect of the (e1) type by getting the hearer to recognize this intention. The proponents of the institutionalist approach claim that illocutionary acts are to be explained and typed by reference to their institutional effects conceived as commitments, obligations, rights, duties, and so on. In other words, they assume that all illocutionary acts are institutional and as such they produce effects of the (e2) type. Finally, according to the interactionalist approach, the performing of an illocutionary act consists in initiating the reproduction of a conventional pattern of social interaction, and thereby in inviting the complementary action on the part of the hearer; the invitation in question is tantamount to the effect of the (e3) type. Let us assume, following Robert M. Harnish, that the „utterance of a sen tence in a context is not sufficient for the performance of a speech act. The theories of speech acts can be organized in terms of what must be added” (Harnish 2005, 11). In other words, the structure of a successful illocutionary act includes at least three elements: words uttered by the speaker, the context of their utterances, and a third element whose nature is a matter of dispute. This element plays a decisive role in determining the force of an act and links the illocutionary practice with other domains of human activity, e.g., mental, institutional or interactional. According to the intentionalist approach, the force-determining element is the speaker’s communicative intention. The proponents of the institutionalist approach identify it with the normative state of affairs produced by the speaker’s act. Those who adopt the interactionalist approach claim that it should be described by reference to a response invited by the speaker’s act. The purpose of the present section is to reconstruct the three approaches in question. Before we go into detail, however, let us note that the approaches presuppose different accounts of the nature of speech acts. According to the intentionalist approach, communicative illocutionary acts form a subclass of intentional actions, and as such they can be explained within the framework of belief-desire psychology (Harnish 2005, 16). In the institutionalist approach it is assumed that illocutions form a special class of institutional acts conceived as moves made by socially accepted rules and procedures. Finally, according to the interactionalist approach, illocutionary acts form a subclass of natural acts which help to achieve coordination between interacting agents. In short, we are faced with the following trilemma: illocutionary acts are in their nature either (i) communicative, (ii) institutional, or (iii) interactional. DRAFT VERSION 2.1. The intentionalist approach According to the intentionalist approach, the structure of most illocutionary acts—such as statements, warnings, requests, promises, and so on—involves three elements: the words uttered by the speaker, the context of their utterance, and the speaker’s communicative intention. It is claimed that the third element determines the force of the act. The proponents of the intentionalist approach are Peter F. Strawson, Stephen Schiffer, Kent Bach and Robert M. Harnish. They claim that what underlies the practice of producing and interpreting illocutionary acts is the agent's mental ability to form and identify certain intentions rather than the existence of certain conventional procedures and rules. In his Intention and Convention is Speech Acts Strawson distinguishes between two categories of illocutionary acts: (c1) communicative and (c2) conventional. The performing of an act of the (c1) type consists in issuing an utterance with a certain communicative intention: with the overt intention to get the hearer to believe that something is the case (statements and other assertive acts), to get the hearer to perform certain action (requests and other directive acts), to get the hearer to be on the alert for certain things (warnings), and so on. To perform an act of the (c2) type, by contrast, is to make a move that forms a part of a convention-governed and convention-constituted practice: to pronounce a couple man and wife, to pronounce the verdict of divorce, to redouble at bridge, to checkmate in the game of chess, and so on. According to Strawson, only acts of the (c2) type are conventional in Austin's sense; acts of the (c1) type are Gricean acts of non-natural meaning that can but do not must be performed by conventional means. Following Grice, Strawson offers the following definition of non-natural meaning: “S nonnaturally means something by an utterance x if S intends (i1) to produce by uttering x a certain response (r) in an audience A and intends (i2) that A shall recognize S's intention (i1) and intends (i3) that this recognition on the part of A of S's intention (i1) shall function as A’s reason, or a part of his reason, for his response r. (The word “response,” though more convenient in some ways than Grice's “effect,” is not ideal.)” (Strawson 1964, 446). He claims that the definition provides a partial analysis of the concept of understating: the understanding of the speaker’s act is the recognition of intention (i1) behind his utterance or, in other words, the satisfying of intention (i2). According to Strawson, the concept of understanding is tantamount to Austin’s notion of uptake : in order to secure uptake the hearer ought to recognize one’s intention (i1). Strawson claims: DRAFT VERSION “If the identification were correct, then it would follow that to say something with a certain illocutionary force is at least (in the standard case) to have a certain complex intention of the (i4) form (…)”. (Strawson 1964, 449) More specifically, if two acts are illocutionarily equivalent, i.e., have the same illocutionary force, then they are equivalent with respect to the response r the speakers intend to produce on the part of their hearers by getting them to recognize the speakers’ intentions (i1). In their Linguistic Communication and Speech Acts Bach and Harnish offer a refined version of the intentionalist account of illocutionary acts. They accept Strawson’s distinction between the acts of the (c1) type and the acts of the (c2) type. They claim, however, that Strawson’s definition of communicative intentions is not adequate and requires a substantial revision. According to Bach and Harnish, Strawsonian force-determining intentions are, in fact, perlocutionary: they are intentions to bring about certain consequential effects upon the feelings, thoughts, or actions of the audience. It turns out, however, that in at least some cases one’s utterance can legitimately function as a statement, warning or request even though one fails to intend the hearer to, respectively, accept the proposition one communicates, get the hearer to be on the alert for the danger one describes or get him to do what he is told. To do justice to such illocutions and to avoid explaining illocutionary acts in terms of perlocutionary consequences, Bach and Harnish develop their own definition of illocutionary communicative intentions. To perform a communicative illocutionary act, they claim, is to express an attitude. The attitude expression consists, in turn, in uttering a sentence with the reflexive intention to get the hearer to take one’s utterance as a reason of thinking that one has that attitude (K. Bach and Harnish 1979, 15). The distinctive feature of communicative reflexive intentions is that “their fulfilment consists in their recognition” (K. Bach and Harnish 1979, 13). For example, to state that p in uttering sentence s is to reflexively intend the hearer to take one’s utterance of s as reason to think that one has (a) the belief that p and (b) the intention to get the hearer to form the belief that p; the hearer’s recognition of one’s reflexive intention consists in his taking one’s utterance as reason to think that one has attitudes (a) and (b). Note that the intention (b)—which plays a decisive role in determining the force of one's act—is perlocutionary and as such is equivalent to the Strawsonian intention (i1) for statements. According to Bach and Harnish, however, making a successful statement involves expressing rather than having the intention (b); the point is, namely, that one can express an attitude no matter one has it or not. 1 1 Bach and Harnish’s account of illocutionary acts involves the so-called non-achievement use of “express.” For a discussion of this topic, see (Harnish 2005, 16–17). DRAFT VERSION In short, Bach and Harnish’s version of the intentionalist conception allows for successful though insincere illocutionary acts. In this respect, the account of illocutionary acts in terms of attitude-expression has an advantage over Strawson’s conception that requires the speaker to have the intention that determines the force of his act. Bach and Harnish’s account has a few further theoretical merits. First, it defines successful communication in terms of the hearer’s recognition of the speaker's reflexive intention, and thereby distinguishes the speaker’s communicative success from the achievement of his perlocutionary goals. Second, it gives rise to a theoretically based taxonomy of speech acts, since illocutionary acts can be typed by reference to kinds of atti tudes expressed by speakers (K. Bach and Harnish 1979). Nevertheless, the intentionalist approach faces two serious problems. Recall, first, that according to Strawson, Bach and Harnish, what underlies the illocutionary practice is the communicating agents’ ability to form and recognize communicative intentions. One can doubt, however, whether communicative intentions are thinkable representations at all, i.e., whether it is possible for our limited minds to think their complex contents. It seems, for example, that Strawson’s analysis of communicative acts is not complete or, in other words, that the list of intentions (i1), (i2), (i3) and (i4) can be continued ad infinitum. The intention (i4) has to ensure the overtness of intention (i2). The point is, namely, that every communicative intention—i.e., the intention whose fulfilment constitutes the speaker’s communicative success—is by definition overt. For the same reason, however, intention (i4) is to be overt too, since its fulfilment is part of the speaker’s communicative success. To make Strawson’s analysis complete, therefore, we should ascribe to the communicating agent intention (i5) thus enabling the hearer to recognize intention (i4). The speaker’s communicative success, however, involves the fulfilment of intention (i5). For this reason we should ascribe to him intention (i6) which has to ensure the overtness of intention (i5), and so on ad infinitum. It turns out, therefore, that the proponent of Strawson’s account has to concede that to perform a communicative illocutionary act, infinitely many intentions (ik) should be formed in order to enable the hearer to recognize intention (ik-1). Bach and Harnish avoid this problem by replacing the iterative account of communicative intention with the reflexive account. They assume, namely, that communicative intentions are reflexive in the sense that their fulfilment consists in their recognition. One can doubt, however, whether it is possible for the human mind to produce a representation whose content contains such a token-reflexive element (Recanati 1986; Bach 1987; Harnish 1991; Siebel 2003; Witek 2009). Second, one can object to the idea that performing successful illocutionary acts such as statements, requests, warnings, and so on, comes down to uttering a sentence with the intention to produce the effect of the (e1) type, i.e., with the intention to secure uptake on the part of the hearer. According to Austin, how- DRAFT VERSION ever, the central function of illocutionary acts—including statements, requests, warnings, and so on—consists in producing effects of the (e2) type, i.e., in generating normative states of affairs conceived as commitments, obligations, rights, entitlements, etc. In other words, an illocution is successful if it is binding. The proponents of the intentionalist approach ignore this norm-producing function of successful illocutions, and, in this connection, they seem to redefine the Austinian concept of illocutionary acts (Doerge 2009). 2.2. The institutionalist approach According to the institutionalist approach, the structure of a successful illocution comprises the words uttered by the speaker, the context of their utterance and a normative state of affairs the speaker produces in making his act. The state consists in the speaker’s becoming responsible for the fulfilment of a certain condition. For example, making a statement involves expressing a proposition and taking responsibility for its truth, making a promise involves committing oneself to performing the action one describes, and so on. In short, the issuing of a successful illocutionary act involves the production of an effect of the (e2) type—let us call it the “institutional effect” of the act—which determines the force of the speaker’s utterance. The producing of this effect, in turn, involves following certain rules or procedures that exist in virtue of their being collectively accepted by the interacting agents. In short, what constitutes the possibility of the illocutionary practice, i.e. what makes it possible to produce institutional or normative states with words, is the collective acceptance of certain institutional rules or procedures. The proponents of the institutionalist approach are Robyn Cameron, William P. Alston and John R. Searle. In his paper Sentence-Meaning and Speech Acts Cameron claims that illocutionary acts form a subclass of effective institutional acts . The distinctive feature of the latter is that: “they effect, in a non-consequential fashion, some change (…) in the “institutional” or “conventional” world—the world of rights, duties, commitments, roles, status, and other social facts of the institutional kind—as opposed to the natural world. If I make a promise to someone, I create a promissory commitment between myself and him; similarly a marriage ceremony, an ordination, inauguration, or investiture, or a command issued within a command structure, effects something, makes a change within the realm of institutional fact. (In each case the change amounts to the fact that certain actions or sorts of actions, on the part of certain people, are now in order, or required, or proscribed.)” (Cameron 1970, 101) DRAFT VERSION In short, effecting institutional acts, in general, and illocutionary acts, in particular, produce effects of the (e2) type: they “take effect” by changing the normative situation of the interacting agents. According to Cameron, they can perform the norm-changing function only against the background of socially accepted “constitutive conventions” (Cameron 1970, 98). In his Illocutionary Acts and Sentence Meaning Alston claims that performing a successful illocution requires to utter words in accordance with a certain rule and take responsibility for the fulfilment of the condition specified by the rule. Contrary to the proponents of the intentionalist approach, he claims: “The utterance is made the illocutionary act it is, apart from any conventional effect production that is essentially involved, not by any “natural” facts about the speaker—his beliefs, perlocutionary intentions, or whatever—but by a “normative” fact about the speaker—the fact that he has changed his normative situation in a certain way by laying himself open to the possibility of censure, correction, or the like in the case the conditions in question are not satisfied” (Alston 2000, 70–71). The most popular version of the institutionalist approach comes from Searle, who in Speech Acts defends the idea claiming that “speaking a language is engaging in a rule governed form of behavior” (Searle 1969, 22). He also claims that illocutionary acts are basic units of human communication and that one’s performing of a successful illocution involves one’s producing of a certain institutional state of affairs: one’s commitment to do something (promises), one’s responsibility for the truth of the proposition one expresses (statements), one’s attempt to get one’s interlocutor to do something (requests), and so on (Searle 1969, 66–67). According to Searle, the rules whose existence creates the possibility of the illocutionary practice have the form “X counts as Y in context C.” They are illocutionary constitutive rules; their totality forms the institution of language. Generally speaking, every institution is a system of constitutive rules of the form “X counts as Y in context C” that provides a structure within which one creates institutional facts Y by performing action X in context C; in other words, the job of human institution is to create the possibility of producing new sorts of deontic powers : rights, entitlements, commitments, obligations, and so on (Searle 2005). For example, the utterance of an explicitly performative sentence of the form “I promise to do A” (the action X) in a certain context (the context C) counts as the undertaking of the obligation to do A (the institutional state Y). According to Searle, language is the most basic institution; it is, namely, a system of illocutionary constitutive rules that makes it possible to create other, more specific and extra-linguistic institutions. DRAFT VERSION The institutionalist account allows for the intuitions underlying the intentionalist approach, and avoids its weaknesses. First, viewed from the Serlean perspective, the Strawsonian distinction between the acts of the (c1) type and the acts of the (c2) type comes down to the distinction between the acts whose performance involves nothing but the linguistic constitutive rules and the acts whose performance requires the existence of some extralinguistic institutions. Second, like Bach and Harnish, Searle assumes that performing a successful illocution involves one’s reflexively intending to produce the effect of the type (e1) on the part of one’s audience. Unlike Bach and Harnish, however, Searle claims that one can form such an intention only against the background of certain illocutionary rules that jointly define the institutional effect of one’s act, i.e., define its effect of the (e2) type: the intention of producing the hearer’s uptake is equivalent with recognizing that the institutional states specified by the rules in question obtain (Searle 1969, 50). In other words, the proponents of the institutionalist approach allow for the idea stating that illocutions are intentional actions and they justify the fact that the central function of illocutionary acts is generating normative states of affairs. There are, however, at least two serious challenges to the institutionalist approach. First, as Harnish notices, commit ments and entitle ments are higher-level properties that can hardly be accommodated within the naturalistic picture of the world. In other words, the theory of institutional facts can hardly be regarded as an adequate basis for a naturalistic reduction of speech act theory (Harnish 2005, 23). Second, one can object that the theory of constitutive rules can be used to account for a narrow class of conventional illocutionary acts, i.e., those that form a part of conventionalized and ritualized practices. The problem is, one may add, that the most ordinary illocutionary acts are indirect and as such, it seems, they cannot be accounted in terms of the rules of the form “ X counts as Y in context C.” To meet the first challenge, Searle attempts to account for the ontology of institutional fact in terms of collective acceptance. He, namely, assumes that the rules of the form “X counts as Y in context C” exist in a given community if they are collectively accepted by its members. To meet the second challenge, in turn, one can refer to Marina Sbisà’s distinction between two concepts of conventionality: means-conventionality and effect-conventionality. One can admit, namely, that few acts are conventional because of the conventionality of the means by which they are performed; but all illocutions are conventional because of the conventionality of their effect, i.e., because of the fact that their effects are institutional states that exist as they are collectively accepted by interacting agents. DRAFT VERSION 2.3. The interactionalist approach According to the interactionalist approach, performing a successful illocutionary act consists in uttering certain words in a certain context and in initiating, in uttering these words, the reproduction of a certain conventional pattern of verbal interaction. Normally, the pattern consists of two parts: the speaker’s part and the hearer’s part. Therefore the initiation of its reproduction consists in the production of the speaker’s part and thereby in inviting the production of the complementary hearer’s part. The main idea of the interactionalist approach is that illocutionary acts are typed by reference to their invited responses or cooperative outcomes. The category of conventional patterns of interaction has been proposed by Ruth G. Millikan. In her Language Conventions Made Simple she claims that natural conventions, in general, and language conventions, in particular, consist of reproduced patterns of activity whose forms are arbitrary relative to their functions (Millikan 1998). For example, using forks and using chopsticks are two equally effective strategies of placing food in one’s mouth. In other words, the form of any of these two types of activity, i.e., using forks in Europe and using chopsticks in Asia, is arbitrary relative to its function. For this reason we can call them conventional patterns of behaviour because, first, they proliferate by reproduction, and, second, they have been reproduced because of the weight of their cultural precedent. Some conventional patterns are reproduced by counterpart-reproduction rather than by direct copying. Assume, following Millikan, that within every counterpart-reproduced pattern one can distinguish two aspects or parts whose joint reproduction is guided by the need to fit one another. According to Millikan, the function of a counterpart-reproduced pattern is to help to achieve coordination between interacting agents: the leader and the follower or, if the interaction involves the performance of speech acts, the speaker and the hearer. In Proper Function and Convention in Speech Acts Millikan claims: “In the case of conventional directive uses of language such as paradigm uses of the imperative mood, the pattern that is conventionally reproduced begins with an intention or desire of S’s that H should act in a certain way. It is completed when H has acted that way as a result of guidance, in accord with conventional rules for guidance, from conventional signs made by S. That the pattern is not completed until H has acted as directed is clear, for new instances of the pattern would not be initiated by speakers were it not that hearers sometimes complete such patterns. The first part of the pattern is conventional, is reproduced, only because both parts are sometimes reproduced. […] DRAFT VERSION Similarly, when S tells H that something is the case in a conventional way, it is conventional for H to believe it. That the pattern is not completed until H has been guided into belief in accordance with the conventional rules is clear because new instances of the pattern would not continue to be initiated by speakers were it not that hearers sometimes believe what they are told” (Mil likan 2005, 152–153). In short, performing an assertive illocutionary act consists in initiating the reproduction of a conventional pattern whose speaker part involves the uttering of an indicative sentence; the hearer’s part of the pattern involves the hearer’s cooperative response, i.e., his believing what the speaker asserts. To perform a directive act is to initiate the reproduction of a complex conventional pattern whose speaker part involves the uttering of an imperative sentence; for the hearer to complete the reproduction is to do what he is told. Illocutionary acts, therefore, can be typed by reference to the responses they conventionally invite, i.e., by reference to what can be called their interactive effects (Witek 2010). Note that the invitation of the interactive effect of an act—that Millikan calls the acts conventional outcome —is the production of the effect of the (e3) type. Undoubtedly, the idea of interactive effects of illocutionary acts requires more detailed elaboration. Note, for example, that in many cases our interlocutors respond to our assertive acts by correcting or challenging the propositions we express rather than by believing them. Nevertheless, their responses can still be regarded as cooperative and conventional in the light of certain pat terns of interaction. By the same token, the hearer, who, instead of complying with what he is told, negotiates the conditions under which the speaker’s request can be fulfilled, can still be regarded as behaving cooperatively and conventionally. To allow for these forms of conventional cooperation, I propose to distinguish between the pri mar y and secondar y interactive effect of an illocutionary act and to stipulate that it is the former, not the latter, that determines the act’s force and type-identity. I assume that the main function of illocutionary acts is the evoking of their primary effects in accordance with pri mar y illocutionar y conventional patterns (e.g., accepting information, carr ying out a request, etc.) whereas the occurrence of their secondary effects (e.g., challenging what the speaker states, negotiating the carrying out of a request, etc.) results from reproducing secondar y illocutionar y conven tional patterns and allows for the cooperation in order to proceed even if the hearer refuses or delays the production of the primary effect of the speaker’s act. The interactionalist approach offers an original account of what Austin calls the hearer’s uptake . According to the standard reading—suggested by Austin and assumed by both the intentionalist approach and the institutionalist one— the securing of uptake consists in getting the hearer to understand the force DRAFT VERSION and meaning of the speaker’s act. Let us consider, however, what it is for the hearer to understand the speaker’s act, i.e., what it is for him to respond to the act by producing the effect of the (e1) type? According to one answer, this is the forming of a mental representation of the force and meaning of the speaker’s utterance. It seems, however, that in at least some cases the hearer’s uptake consists in the way he reacts to the speaker’s act rather than in his having certain mental states. In some cases, for example, taking one’s utterance to be a binding order consists in simply responding by complying with what one says, no matter whether the response is produced consciously or not. What matters here is that it is produced non-accidentally by completing the reproduction of the relevant pattern of social interaction. The most serious challenge to the interactionalist approach is an explanation of the norm-generating function of an illocutionary act, i.e., the investigation of the relationship between the inviting of the act’s interactive effects and the producing of its normative effects. In other words, the idea of illocutionary patterns of interaction should be used to account for the discursive mechanisms responsible for bringing about changes in the social domain of rights and commitments. DRAFT VERSION REFERENCES Alston, William P. 2000. Illocutionary Acts and Sentence Meaning. Ithaca and London: Cornell University Press. Austin, John L. 1962. How to Do Things with Words. Oxford: Oxford University Press. Bach, K., and R. M. Harnish. 1979. Linguistic Communication and Speech Acts. Cambridge, Mass.: MIT Press. Bach, Kent. 1987. “On Communicative Intentions: A Reply to Recanti.” Mind & Language 2 (2): 141–154. Cameron, Robyn. 1970. “Sentence-Meaning and Speech Acts.” The Philosophical Quarterly 79 (20): 97–117. Doerge, Friedrich. C. 2009. “A Scholarly Confusion of Tongues, or, Is Promising an Illocutionary Act?” Lodz Papers in Pragmatics 5 (1): 53–68. Harnish, Robert M. 1991. “Review of Meaning and Force: The Pragmatics of Performative Utterances by F. Recanati.” The Philosophical Review 100 (2): 297– 300. ———. 2005. “Commitments and Speech Acts.” Philosophica 75: 11–41. Jadacki, Jacek J. 2010. Spór o granice języka: elementy semiotyki logicznej i metodologii. Warszawa: Semper. Millikan, Ruth G. 1998. “Language Conventions Made Simple.” The Journal of Philosophy 95 (4): 161–180. ———. 2005. Language: a Biological Model. Oxford: Oxford University Press. Recanati, Francois. 1986. “On Defining Communicative Intentions.” Mind & Language 1 (3): 213–241. Searle, John R. 1969. Speech Acts: An Essay in the Philosophy of Language. Cambridge, Mass: Cambridge University Press. ———. 2005. “What Is an Institution?” Journal of Institutional Economics 1 (1): 1–22. Siebel, Mark. 2003. “Illocutionary Acts and Attitude Expression.” Linguistics and Philosophy 26 (3): 351–366. Strawson, Peter F. 1964. “Intention and Convention in Speech Acts.” The Philosophical Review 73 (4): 439–460. Witek, Maciej. 2009. “Scepticism About Reflexive Intentions Refuted.” Lodz Papers in Pragmatics 5 (1): 69–83. ———. 2010. “Naturalising Illocutionary Rules.” In Beyond Description: Naturalism and Normativity, ed. Marcin Miłkowski and Konrad Talmont-Kaminski, 243–264. London: College Publications.