Against Illocutionary Archaeology, For Illocutionary Negotiation
How do we decide what illocutionary classification a specific utterance should be classified as? This essay is structured in three sections. In the first section, I trace a lineage of answers to this question, including ratification (McDonald, 2021, 3506), constitution (McDonald, 2021, 3507) (Kukla, 2014, 454), and social constitution views (cf. Harrison and Tanter, forthcoming, 17). I argue that each theory goes some way towards solving the problems of its predecessor, but that there are still some loose ends present by the time we get to the social constitution theory. In the second section, I argue that all three theories are waylaid by the implicit premise of “illocutionary monism” (Johnson, 2019, 1151) the idea that all utterances must be classified by a single illocutionary classification, which leads to unintuitive outcomes, such as speech acts retroactively changing classification. I argue that we should instead accept the premise of illocutionary pluralism, in which it is possible for an utterance to be correctly classified as multiple different illocutionary acts. Having argued for this, I argue for the stronger position, that illocutionary classes should be understood as mid-level abstract descriptors, which necessarily highlight some relevant features of an utterance whilst leaving others unreferenced, and therefore all utterances may be validly labelled as multiple illocutionary classes, depending on what an agent wishes to highlight. As such, I argue for a position of illocutionary relativism (Johnson, 2023, 132), in which illocutionary classification of an utterance is relative to an agent, a locution, a set of illocutionary classifications, and their corresponding felicity conditions. In the third and final section, I address the problem of relativist and pluralist notions of illocutionary classification, namely that there are some situations in which we vehemently disagree and find it very important to combat someone else’s ‘incorrect’ classification of an utterance. I argue that it is still possible to persuade within the bounds of an illocutionary relativist framework; that we may negotiate on behalf of our preferred illocutionary classification by appealing to the pragmatic benefits it poses for an agent’s pre-committed-to ends. By ‘plugging-in’ to the pre-committed ends of an agent, we can make an argument that rationally and pragmatically prescribes how they ought to class the utterance, thereby settling which speech act is agreed to have occurred.
Section 1 - A Lineage Of Answers
What is meant by a speech act? J.L. Austin, in How To Do Things With Words (1975), argues that utterances are not merely used to convey information or propositions, but rather, also to perform specific actions, which he refers to as ‘performatives’ or speech acts. Furthermore, he proposes a tripartite model of analysis; each speech act may be separated into three parts. A locution or locutionary act r_efers to a specific utterance; the words and the sentence itself that was uttered. An _illocution or illocutionary act is the class, or type of conversational move that is made when a specific locution occurs. Asking a question, insulting someone, commanding, ordering, requesting, and flirting, are all examples of illocutionary acts. Finally a perlocution or perlocutionary act refers to an effect that is accomplished in virtue of making an illocutionary act. As such, in someone saying “That’s a nice shirt” to a friend, the locutionary act would be the words, tone and pitch used in making the utterance, the illocutionary act might be a compliment, and the perlocutionary effect would be the friend’s reaction to the statement. Performatives, or speech acts, Austin argues, are not like propositions in that they are true or false, but can rather succeed or fail, what he calls being felicitous or infelicitous. The conditions by which a particular utterance succeeds are called felicity conditions and are unique to each illocutionary act or type. As such, the felicity conditions for a successful order are different from the felicity conditions for a successful request. It is furthermore held that some illocutionary acts require appropriate hearer ‘uptake’, meaning the hearer ‘takes up’ the act, providing it with recognition. As such, Austin argues, one cannot be said to have ‘warned someone’ unless they actually are warned; “an effect must be achieved on the [hearer] if the illocutionary act is to be carried out” (1975, 152). In this sense, an actor who, while on stage, attempts to warn the audience by yelling “There’s a fire!”, may fail to have their speech act count as a warning, because the audience may not give it the appropriate uptake. The concepts, then, locutionary act, perlocutionary act, illocutionary act, felicity conditions, and uptake, set the stage for the rest of the analysis of this essay, which shall investigate the question of how we should illocutionarily classify a particular utterance; how we should settle which speech act occurred.
Ratification theories (McDonald, 2021, 3506), answer this question, by combining the phenomenology of how it feels to do a speech act, in that we intend some communicative action, with Austin’s requirement of uptake for us to properly say that a particular speech act occurred. In this sense, a speech act is said to occur if a speaker makes an utterance with some communicative intention (e.g. to warn, to insult, to praise) and this communicative intention is properly recognised and thereby ‘ratified’ by a hearer (McDonald, 2021, 3509, cf. Grice, 1989, Searle, 1969, 47). In this sense, we might say that if the actor’s audience did properly recognise his communicative intention to warn them of a fire, and reacted appropriately, then when asked the question, “What speech act did the actor perform?” we should illocutionarily classify his utterance as a warning. Ratification theories, therefore, split the authority of illocutionary classification between the speaker and the hearer, in that the speaker puts forth a communicative intention, and the hearer’s reception determines whether it is ratified and successful, or not. Importantly, under ratification theories, the correct illocutionary classification of an utterance always stems from what the speaker intends, whereas whether the speech act succeeds or fails is determined by the hearer’s uptake. In this sense, if the actor intends to make a warning with their utterance, but the audience treats the utterance as though it is a joke, or delivering the next line of the play, under the ratification view, we should not say that ‘A joke occurred’, or ‘A line was delivered’ but rather that ‘The actor’s warning failed’. The ratification view, therefore, answers the question of how we should settle which speech act occurred, by appealing to the speaker’s intention of which speech act they intended to perform, which may then be ratified and succeed if the communicative intention is correctly recognised by a hearer.
Whilst ratification theories capture our intuitions about speaker intentions and combine them with Austin’s insights about proper uptake being required, they yield strange results in situations in which it appears that an illocutionary act occurs which is different from what a speaker intended. It seems that on ratification theories, one cannot accidentally perform a speech act other than what they intend. Subsequently, it is also not appropriate for us, the classifiers, to class an illocutionary act as something other than what is intended. And yet, consider the following case:
Coffee Order: An Australian businessman has been practising his Chinese in preparation for a business meeting in Shanghai. In a rush before the meeting, he pops into a small coffee shop and makes the utterance: “Coffee with tāng, please. Make sure it has tāng. I can’t drink it otherwise.” The barista looks surprised. On the way to the meeting, he takes a sip, and, spluttering, realises he has ordered coffee with “tāng” (soup) in it, instead of coffee with“táng” (sugar) in it.
What speech act should we say occurred here? Under the ratification view, it is inappropriate to declare that ‘despite intending to speech act x, speech act y occurred’, because for a successful speech act occurrence, there must be the recognition of the speaker’s actual communicative intention. It seems more appropriate, instead, to say that despite intending to order coffee with sugar, he accidentally used his words in such a way that that he ordered coffee with soup. Perhaps one might defend the ratification view by arguing that in both interpretations of this utterance, the same speech act occurred, namely ‘a request’, and it occurred in virtue of the barista recognising the businessman’s intention to request. Furthermore, one might find the language-barrier nature of this case contrived, and hold that the ratification view still works under ideal conditions in which the same language is shared. However, both of these objections are unsatisfying. This case simply demonstrates that it is possible to accidentally use our words to press ‘social buttons’ and leverage norms in ways that we do not intend. Whilst this case demonstrates an accidental variation in arguably the same ‘illocutionary type’, we can easily imagine similar scenarios, in which someone intends to praise, but ends up insulting (“Your hair looks so much better than usual!”), or intends to insult but ends up complimenting (“You barely qualify as a good engineer”). On the language point, this case simply makes even more apparent that the norms that we activate, the ‘social machines’, by which we input speech and receive output in terms of reactions, are not within us, but rather socially dispersed. In all of these cases, it is the hearer’s classification and response, that seems to settle which speech act occurred, contrary to the ratification view.
In contrast to ratification theories, constitution theories (McDonald, 2021, 3507) of uptake afford greater authority to hearer response, settling which speech act occurred by the classification and pragmatic response of hearers (Kukla, 2014, 454). In this sense, what illocutionary act a given locution is, is constituted by the pragmatic uptake it receives; it is what it is treated as by the hearers. Quill Kukla (writing as Rebecca Kukla, 2014, 445) uses the example of a female factory boss, Celia, who finds her utterances being given uptake as requests by the male factory workers, instead of as orders, as she intends. According to the constitution view, we should settle what speech act occurred in virtue of the pragmatic uptake it received, and therefore, on this view, Celia is accidentally making requests. Equally, the businessman in Coffee Order has accidentally leveraged norms such that the hearer, the barista, makes him coffee with soup - in this sense, he did order coffee with soup. The constitution view, therefore, would have us settle what speech act an utterance is classified as, by how the pragmatic norms and behaviour of hearers were influenced, and what uptake they gave the utterance.
In taking the hearer’s response to define what illocutionary act is performed, however, the constitution theory eliminates the possibility that the hearer might be wrong and that they responded inappropriately to an illocutionary act. The constitution view therefore begs the question; To someone with an extremely jovial disposition, is it possible to tell them anything but jokes? That is to say, is it truly the case that if a person were predisposed to give all utterances the uptake of non-serious jokes, should we illocutionarily classify all utterances said to them as jokes? We might worry that the constitution view reduces our ability to disagree with someone’s classification. This concern becomes especially pertinent in cases in which tremendous importance is placed on a particular illocutionary classification of an utterance, such as sexual consent (Harrison and Tanter, forthcoming, 2). Under the constitution view, if an utterance is given pragmatic uptake as consent, then it is correctly illocutionarily classified as an act of consent. We might worry then, that the constitution view, in placing the locus of illocutionary classification exclusively on hearer uptake, robs us of the ability to disagree with what we perceive to be an incorrect illocutionary classification of an utterance, especially when the stakes are high.
A social constitution theory (Harrison and Tanter, forthcoming, 17) offers a solution to these problems, by settling which speech act occurred not by the uptake of a single hearer, but by the pragmatic and potentially overriding uptake of the broader discursive community. Uptake is not limited solely to the hearer of an utterance, Harrison and Tanter argue, but is given by a broader discursive community including “bystanders, accidental overhearers and a range of people who come to hear about the speaker’s utterance second-hand” (forthcoming, 14). The concept of uptake, then, is extended under the social constitution view, far beyond the initial speaker and hearers of the utterance. If a particular illocutionary classification is spread by those with the power to do so, and supplies a particular pragmatic structure, then that uptake should be considered overriding. It is therefore possible for a discursive community to apply their own felicity conditions to an utterance, and in the case of sexual consent, classify it as an illocutionary act of consent or refusal, regardless of how it appeared to the hearer at the time. In this sense, then, according to a social constitution view, we should settle what illocutionary act a particular utterance is, by the uptake that the broader discursive community gives it.
The social constitution view presents an attractive solution to the problem of disagreeing with an ‘incorrect’ illocutionary classification, by vesting the constitutive power in the broader discursive community. However, one might argue that this kicks the problem down the road, from disagreeing with the illocutionary classification of a single hearer, to disagreeing with the illocutionary classification of a discursive community. In this sense, a speaker and a hearer could even verbally agree that an utterance is a particular illocutionary classification, such as a compliment or act of consent, and yet they could be wrong, the utterance is actually a different illocutionary act, because the broader discursive community classifies it, or would, or will classify it, differently, providing overriding uptake. The shift of constitution from an individual hearer’s uptake to the discursive community’s uptake should only be comforting to the extent that we think the discursive community is more likely to ‘correctly’ illocutionary classify the utterance*.* This concern is noted by Harrison and Tanter, that there is “no guarantee that a discursive community will provide just overriding uptake” (forthcoming, 16). In this sense, a social constitution theory goes some way towards resolving the problem of disagreement, by providing the ability for a broader discursive community’s uptake to override initial hearer uptake. We are, however, still beholden by the discursive community’s illocutionary classification, even in potentially unintuitive situations, such as when hearer and speaker both agree.
Section 2 - Illocutionary Monism, Pluralism and Relativism
One premise, that all three views implicitly hold, I argue, is illocutionary monism (Johnson, 2019, 1153). This is the premise, Casey Rebecca Johnson puts forth, that “each utterance has, at most, one [illocutionary] force”. Illocutionary monism is therefore the belief that each utterance, is and only is a single illocutionary act; that there exists a single correct illocutionary label for it. This is implicit in all three theories, each of which tries to find an answer for which illocutionary act an utterance actually is, at its core. It is this premise, I argue, that leads to unintuitive upshots for all three views. Firstly, for the ratification view, it means that the only correct illocutionary classification is one that is in line with the speaker’s actual communicative intention. Secondly, for the constitution view, it means that the only correct illocutionary classification is one that is in line with the uptake the hearer grants it. Finally, for the social constitution view, it means that the only correct illocutionary classification is that which the discursive community’s uptake constitutes.
Furthermore, the premise that there is only one correct answer results in the very unintuitive case of “retroactive undoing” of speech acts (Langton, 2018, 157). ‘Retroactive undoing’ occurs as an upshot of holding the premise of illocutionary monism, along with any theory of illocutionary classification that allows for one classification of an utterance at time t1 and another ‘more correct’ classification at time t2. This may be accomplished either by extending the felicity conditions of certain speech acts into the future (Langton, 2018, 157), or by temporal separation of utterance and classification, such as a social constitution view (Harrison and Tanter, forthcoming, 15). As such, when a new classification is applied, due to the premise of illocutionary monism, the consequential claim is that the most recent claim is the correct one; that the true force of the utterance has been archaeologically discovered. The previous classification is therefore retroactively undone, it is revealed that even at time t1 the utterance would always eventually result in the later classification, and therefore, any alternative classification at time t1 is false, it has always ‘actually’ been the most recent classification. This results in the intuitively uncomfortable scenario in which any utterance’s proper illocutionary classification is deferred to the future, the ‘true core’ of the utterance may always be archaeologically revealed from out underneath us, as though the next layer down of a gobstopper. It is prima facie unintuitive, I argue, to hold that a speaker and a hearer, who both agree on the illocutionary force of an utterance, might actually be wrong about it. Is there a viable alternative, then, to the premise of illocutionary monism, that comes at a lower intuitive price?
I argue that there is. Illocutionary pluralism is the position diametrically opposed to illocutionary monism, and instead of holding that each utterance may have only a single unique illocutionary force, holds that it is possible for an utterance to exhibit multiple illocutionary forces simultaneously (Johnson, 2023, 121). One intuitive advantage to illocutionary pluralism over illocutionary monism is that it easily accounts for conversations in which there are multiple participants (Johnson, 2019, 1158)(Tanesini, 2019, 760). This is demonstrated by Johnson’s example, a derivative of Kukla’s Celia case, in which one hearer, Mark, illocutionarily classifies a given utterance as a request, and another hearer, Mary, illocutionarily classifies the same utterance as an order (Johnson, 2019, 1158). In such a scenario, it certainly appears as though a single locutionary utterance has had multiple illocutionary forces. Furthermore, as an utterance may be taken multiple ways by multiple hearers, this may be capitalised on and intended by a speaker, exemplified with dogwhistles (Saul, 2018, 362). Based on these examples then, we should reject illocutionary monism, which holds that a given utterance may only be classified as a single illocutionary type, and instead accept an illocutionary pluralist position, in which it is possible for a single utterance to exhibit multiple different illocutionary forces.
I propose to take the argument even further, however, such that we should accept the stronger premise: that all utterances may be validly described by multiple different illocutionary classifications. Motivating this is a conception of illocutionary classifications as a mid-level abstraction. Imagine a joke is told to a friend group. To fully and exhaustively communicate the event to someone who was not present, with all of its implications and consequences, one would have to exhaustively explain the exact pitch pronunciation of the joke, the exact material conditions, including who was there, why they were there, where they were, an entire history of the group’s dynamics, the effects that the utterance was likely to produce etc. But, is the case that the most exhaustive description is always preferred? No! There is utility in being able to ask for a “Teriyaki chicken sub” instead of “3.5*10^6 mols of carbon, hydrogen, oxygen and nitrogen in a convex lattice structure”. Perhaps more succinctly, there is utility in being able to vary the level of analysis, evidenced in the variation from physics, to chemistry, to biology. There is an ecological niche, I argue, in speech descriptors for a mid-level abstract description which conveys a rough idea of a typical utterance, uttered in a conventionally typical way, with conventionally typical aims, and conventionally typical results. Fortunately, illocutionary classifications fit this bill. Under this conception then, an illocutionary classification must always fail to entirely capture a locutionary act, there must always remain something uncaptured by it. Consequently, this necessary failure to capture an entire locutionary act opens the door for one to later highlight different aspects of the same utterance by referring to it via a different illocutionary classifier. Therefore, on the basis of illocutionary classification being a mid-level abstraction necessarily leading to many valid classifications of a given utterance, I argue that we should accept a strong illocutionary pluralism.
We might further develop the position, to one described as an ’illocutionary relativism’ (Johnson, 2023, 132). The upshot of such a view is that which speech act occurred is settled relative to an utterance, a classifying agent, their available illocutionary classifications, and those classification’s felicity conditions. As such, regarding the Celia case, we would say “For Mark, given his set of available illocutionary classifications {A,B,C…}, and their felicity conditions {A-1,2,3, B-1,2,3, C-1,2,3…}, locutionary act x was an A, a request” whereas we might say “For Mary, given her set of available illocutionary classifications {A,B, D…}, and their felicity conditions {{A-1,2,4, B-1,2, D-1,2,3…, locutionary act x was a B, an order.” This conception denies that felicity conditions are metaphysically free-floating, and denies utterances having an underlying ‘true’ illocutionary core, instead, placing felicity conditions for illocutionary classifications in the minds of classifying agents. We might therefore accept an illocutionarily relativist conception of illocutionary classification, distinguished from the constitution view, by the premise of strong illocutionary pluralism, in that multiple classifications are considered valid, each relative to a particular locution, agent, set of illocutionary classifications, and their felicity conditions.
Section 3 - Pragmatic Illocutionary Prescriptivism
An illocutionary relativist position, however, brings the problem of disagreement even more to the forefront. Again, it certainly seems like in cases such as sexual consent, it matters distinctly whether a locutionary act is classed as this illocutionary classification rather than that. It may seem that whilst the illocutionary relativist position validates our intuitions in that our ‘correct’ classification is not beholden to anyone, not speaker, nor hearers, nor discursive community, it in fact goes too far in the other direction, validating any illocutionary classification, and thereby destroying its utility as a useful conventional mid-level descriptor. Is it possible to disagree, or convince another person to change their illocutionary classification, within the bounds of an illocutionary relativist position?
I argue that it is. Admittedly, I am buoyed with hope by how close the model of illocutionary relativism sounds to disagreeing in productive arguments with family members and friends. “Ahh,” one sometimes says, “See, when you made locutionary utterance x at the gathering, I interpreted it as illocutionary classification C due to felicity conditions C-1,2,3.” “_Now, however,” one sometimes adds, “due to _considerations {XYZ}, I think we are aligned in considering it illocutionary classification D”. But what are the elusive ‘considerations {XYZ}’ that might cause someone to change their illocutionary classification? Both Johnson, and Harrison and Tanter provide answers to this question. Johnson argues that there are social costs associated with holding an illocutionary classification that is not the dominant classification (2023, 133), and furthermore that certain illocutionary classifications may be better pragmatically suited towards appropriately coordinating behaviour (2019, 1164). Consequently, these considerations might serve to persuade one to change their illocutionary classification. Equally Harrison and Tanter argue that those in positions of power might employ it to adjust the pragmatic structure assigned to an utterance, thereby changing the most viable illocutionary classification. Furthermore, they present that we might appeal to a prescriptivist illocutionary classification, arguing that we might persuade someone by arguing that we ought to consider am utterance a particular illocutionary classification (forthcoming, 15). Both of these answers are intuitive, and could predictably lead to change in illocutionary classification, but they also prompt more questions. For Johnson, why are there social costs with holding an alternative classification? And to what ends are certain illocutionary acts beneficial for coordinating behaviour? On the Harrison and Tanter front, in what sense ‘ought’ an utterance be given a particular illocutionary classification, given the premise of illocutionary relativism means we cannot simply answer, ‘because it fulfils the correct felicity conditions’. If these questions can be satisfactorily answered, then we will have hopefully vindicated illocutionary relativism, by providing a means of accounting for disagreement and persuasion.
I put forward an answer to this question, by stitching the two proposed answers together, and appealing to a higher-order principle. I argue that under illocutionary relativism, one may persuade another to change their illocutionary classification by prescriptively appealing to their pre-committed-to ends. One way of understanding in what sense someone ought to give a particular utterance illocutionary classification as one classification and not another, is to prescribe it in a manner similar to Harrison and Tanter, but to understand ought as an instrumental ought, which functions relationally in regards to some end (Finlay, 2008, 315). By pragmatically ‘plugging in’ to someone’s pre-defined end, or goal, and making a link between a particular illocutionary classification and the satisfaction of that goal, one provides them with a rational reason why they ought to classify it in a particular way. The prescription would look something like “Classifying acts like locution x, as illocutionary classification A, by accepting felicity conditions for A as A-1,2,3 instead of A-4,5,6, will forward your end E.” Equally, this explains why agents may rationally apply social sanctions to those that illocutionarily classify an utterance differently to them; their different classification may impede the satisfaction of the agent’s own end. Furthermore, this explains why the exercise of power in terms of threats or rewards, by someone such as a manager, may provide pragmatic reasons for one to change their illocutionary classification, as their own ends are jeopardised or moved forward. Therefore, I put forth that even under illocutionary relativism, we may still disagree and persuade, by making pragmatically prescriptivist cases for why a particular illocutionary classification over another will better serve one’s ends.
There are two potential downsides to this conception. One is that inclusion of the premise of illocutionary relativism may simply contradict one’s intuition that a particular locutionary utterance arrives as a single unique illocutionary act (Johnson, 2023, 133), and that this classification is the only correct classification. While I do not share such intuitions, we might potentially explain them by arguing that human perception is innately ‘goal-oriented’, and therefore locutionary utterances ‘arrive’ already pre-sorted into our goal-directed matrices, and are therefore already assigned an illocutionary classification that seems ‘proper’, ‘legitimate’, ‘correct’, or ‘innate’ to us. The second downside is that in situations in which there is no common end or higher principle to instrumentally appeal to, it seems there are no grounds for persuasion. I am optimistic about the possibility that there is almost always some shared end, principle or goal that might be found to unite an illocutionary classification, however, in practice these may be buried so deep and require so much other preference-revisionism that it may be nigh-impossible to utilise them. While this is an unfortunate state of affairs, I argue that it accurately reflects the world. That agents can reach stalemates in which each refuses to budge from the illocutionary classification that best serves their ends, seems to be an advantage of the theory, and explains situations such as why Grandma Wilkinson’s post-divorce Italian boy-toy Andrea insists that her utterance, “If I could, I’d give you the world” should be understood as a bequest of her estate, whilst the rest of the Wilkinson family vehemently disagrees.
To apply this conception to the examples mentioned so far: Celia may present to the factory workers that they ought to modify their felicity conditions to count her utterances as orders, as this will serve their ends of not getting fired. The workers might equally present some justified conventional pragmatic reason that a particular felicity condition is in place, such that she should modify her utterances to match it; perhaps orders are typically yelled on the factory floor, so all can hear them over the din of the factory machines, serving the end of coordinating work. In Coffee Order, perhaps the barista’s boss might present that the barista ought to adjust her felicity conditions for a successful order of ‘coffee with táng (sugar)’ to account for foreigners with poor pronunciation, serving the end of acquiring happier customers. Equally, perhaps ‘coffee with tāng (soup)’, is a local delicacy, and therefore the distinction between the two is warranted to serve the end of good business, and the businessman ought to adjust his utterances to the end of acquiring non-soup-infused coffee. For the case of sexual consent, it may be that we judge that our interests and ends are best served in maintaining a just society and in evaluating future cases by classifying an utterance either as an act of consent, or an act of refusal. Regardless of how an utterance was illocutionary classified by a hearer, we may hold them responsible for holding inappropriate felicity conditions. In less serious cases, we can imagine partners discussing afterwards “When I said locution x, I was intending an illocutionary act C, according to felicity conditions C-1,2,3” and a partner responding on the appropriateness of such felicity conditions, whether they serve their shared ends of being easily communicable. In more serious cases, a judge or legislator must perform this role, prescribing the illocutionary classification that serves the ends of citizens. In all of these situations, any particular illocutionary classification is permissible, but not all are equal. What settles what speech act we should say occurred is the illocutionary classification that we are satisfied best serves our chosen ends.
This essay therefore concludes, having argued that we should settle which speech act occurs, by pragmatically prescribing an illocutionary classification that best serves our chosen ends. This position was reached, by tracing a lineage through three existing views. The ratification view was presented first, in which what speech act occurs is settled by whether the speaker’s communicative intention is ratified or given uptake by a hearer (McDonald, 2021, 3506). This view, it was argued, accounts for the ability for speech acts to fail if they do not receive appropriate uptake, but fails to account for situations in which it intuitively appears as though one performs an illocutionary act other than what they intend. This concern, it was argued, is mitigated by constitution theories (McDonald, 2021, 3507), in which what illocutionary act occurred is settled by the pragmatic uptake that a locutionary utterance is given by hearers (Kukla, 2014, 454). Constitution theories, however, in assigning illocutionary-defining authority to hearers, were argued to necessitate disregarding speaker intention even in cases in which we find it important to regard, such as sexual consent. Social constitution views, such as Harrison and Tanter’s, in which what settles what speech act occurs is the uptake of the broader discursive community, were argued to mitigate these concerns, by potentially increasing the likelihood that democratic, overriding uptake would be provided (Harrison and Tanter, forthcoming, 17), however, ultimately we would still remain beholden to the broader discursive community’s uptake. I argued that an implicit premise in all three views is illocutionary monism; that each utterance has only a single correct illocutionary classification (Johnson, 2019, 1151). This premise, I argued, resulted in unintuitive concepts, such as ‘retroactive undoing’ and is responsible for ratification, constitution and social constitution views making us beholden to the illocutionary classification of the speaker, hearer, and discursive community respectively. In place of this, and in virtue of speech acts being considered a mid-level abstract description, I argued that we should adopt an illocutionary pluralist conception, in which utterances can necessarily be defined by multiple illocutionary classifications. Furthermore, I put forward that we should consider felicity conditions to be within the minds of classifiers, and not metaphysically detached. Therefore, we should adopt an illocutionary relativist position (Johnson, 2023, 132), in which what speech act occurs is settled in reference to an utterance, an agent, a set of illocutionary classifications, and a set of corresponding felicity conditions. Illocutionary relativism, I argued, came up against the same problem of disagreement, in that all illocutionary classifications being valid removes our ability to disagree with or persuade others to change their illocutionary classifications, at odds with our intuitions that it is sometimes very important. Finally, I argued this tension can be resolved within an illocutionary relativist framework, by negotiating and settling which speech acts occur by appealing to which illocutionary classification best serves one’s ends.
References
Austin, John Langshaw. 1975. How to Do Things with Words. Oxford: Clarendon Press.
Finlay, Stephen. 2009. “Oughts and Ends.” Philosophical Studies 143 (3): 315–40.
Grice, H. Paul. 1989. Studies in the Way of Words. Cambridge, MA: Harvard University Press.
Harrison, Rebecca E., and Kai Tanter. forthcoming. “Whose Uptake Matters? Sexual Refusal and the Ethics of Uptake.” Philosophical Quarterly.
Johnson, Casey Rebecca. 2019. “Investigating Illocutionary Monism.” Synthese 196 (3): 1151–65.
Johnson, Casey Rebecca. 2023. “Some Varieties of Illocutionary Pluralism.”
Kukla, Rebecca. 2014. “Performative Force, Convention, and Discursive Injustice.” Hypatia 29 (2):
440–57.
Langton, Rae. 2018. “Blocking as Counter-Speech.” In New Work on Speech Acts, edited by Daniel Fogal, Daniel W. Harris, and Matt Moss, 144–64. Oxford: Oxford University Press.
McDonald, Lucy. 2021. “Your Word against Mine: The Power of Uptake.” Synthese 199 (1): 3505–
Saul, Jennifer. 2018. “Dogwhistles, Political Manipulation, and Philosophy of Language.” In New Work on Speech Acts, edited by Daniel Fogal, Daniel W. Harris, and Matt Moss, 360–83. Oxford: Oxford University Press.
Searle, John R. 1969. Speech Acts: An Essay in the Philosophy of Language. Cambridge: Cambridge University Press.