Critical Pragmatics: Nine Misconceptions

Abstract

In this paper, we focus on some misconceptions about Critical Pragmatics, what it is, what it assumes and what it proposes. Doubtless, some of these misconceptions are due to clumsy writing on our part; perhaps others are due to inattentive reading. And some may be due to an effort to shield us from the apparent implausibility of what we said—and in fact meant. It does not matter much. We focus on those misunderstandings that most matter to us, either because, by repetition, they have ended up being annoying, even if they are not, perhaps, that important; or because they are substantial enough to represent a distortion of the basic picture of Critical Pragmatics and its theoretical foundations, namely, Critical Referentialism—also known as the Reflexive-referential Theory—and, more generally, Perry’s fundamental views on meaning and content(s).

1 Introduction

It is now almost twelve years that Critical Pragmatics was published.¹ It is some years more since it was conceived. Korta and Perry published their first joint article, “Three demonstrations and a funeral,” in 2006, and, given that it was the product of some years work, it is now almost twenty years or so that they started thinking together about issues concerning language, mind, action and the world. In the last few years, a new member has joined the authors’ team, María de Ponte, and new papers have been added to the oeuvres of Critical Pragmatics.² During the past years, our theory has received quite a bit of attention. Some philosophers and linguists have adopted the framework to deal with this or that philosophical or linguistic issue. They have been exciting and satisfying years for us, no doubt.

In this paper, we are not concerned with celebrating our achievements; we focus on some misconceptions detected during these years about—and around—Critical Pragmatics, what it is, what it assumes, and what it proposes. Doubtless, some of these misconceptions are due to clumsy writing on our part; perhaps others are due to inattentive reading. And some may be due to an effort to shield us from the apparent implausibility of what we said—and in fact meant. It does not matter much. We focus on those misunderstandings that most matter to us, either because, by repetition, they have ended up being annoying, even if they are not, perhaps, that important; or because they are substantial enough to represent a distortion of the basic picture of Critical Pragmatics and its theoretical foundations, namely, Critical Referentialism—also known as the Reflexive-referential Theory—and, more generally, Perry’s fundamental views on meaning and content(s).

Critical Pragmatics is a natural development of the approach developed by Perry in Reference and Reflexivity (2001 [2012]), which, in turn, has its roots in his seminal papers “Frege and demonstratives” (1977) and “The problem of the essential indexical” (1979). Arguably, the origin of some of the misunderstandings can be traced back to those early papers, perhaps not so much by action but by omission. Most of the examples given by Perry on those papers are well-remembered—Hume/Heimson, Rudolf Lingens, the messy shopper, the tardy professor, the lost hiker...—but his positive claims have been often distorted, and often overlooked.

A central thesis defended by Perry that has been largely ignored is his rejection of the doctrine of propositions, and, in particular, his rejection of the view according to which beliefs—and “propositional” attitudes in general—³ consist in a relation between an agent and a proposition. Perry forcefully argued against this in his earlier papers—one might even say, with Falk, that the rejection of the doctrine of propositions is actually the main revolutionary idea of those papers (Falk 2015). Many people, however, seem to have either misunderstood, or simply forgotten this.⁴ Otherwise, it would not be easy to understand how they can attribute to Perry’s Critical Referentialism, or to Critical Pragmatics, the thesis that beliefs consist in a relation between an agent and multiple propositions. Without the rejection of the doctrine of propositions, Critical Pragmatics cannot be understood.

Some of the misconceptions, then, have a relatively long history, but some emerged later, either associated to the publication of Reference and Reflexivity (2001 [2012]), or to the series of papers that started with “Three demonstrations and a funeral” (2006) and led to Critical Pragmatics (2011). Then again, some of the misconceptions might have to do, in part, with the title of the latter book, which is also (one of) the label(s) we use for our approach. We start discussing those.

2 Misconceptions due to the title

The term “pragmatics” in the title refers to the third level of semiotics according to Charles W. Morris in Foundations of the Theory of Signs (1964). Roughly, syntax is about the way words going together makes phrases and sentences; semantics is about meaning, and pragmatics deals with the use of language. Morris was a pragmatist, in the tradition of Peirce and James. But our use is not meant to suggest any particular connection with that movement.

We didn’t realize at the time that there were terms like “critical pragmatism” out there, that pointed to various critical takes on Dewey’s pragmatism, or a sort of synthesis between discourse analysis and critical theory, or to illustrate “a new critical relevance of pragmatist reflection for science and ethics” (Ulrich 2007), among other things.⁵ Our term has nothing to do with any approach like that. Thinking that it does is the first misconception, perhaps a natural one for readers with broader knowledge of all that is going on in philosophy than the authors had (or have).

Misconception one: Critical Pragmatics (CP) is related to Critical Pragmatism.

Our approach started as a natural development of the application to issues of pragmatics of Perry’s Reflexive-referential Theory or Critical Referentialism. These two terms are used in his Reference and Reflexivity to refer to the theory there. Critical Referentialism accepts the main thesis of referentialism, as developed by Donnellan, Kaplan, Kripke and others: That utterances of sentences with names, indexicals, and demonstratives express singular propositions with the references of those terms as constituents.

Straightforward referentialism is well-argued for, and plausible as far as it goes. But utterances involving indexicals, proper names, and the like seem to convey more than what is captured by singular propositions. To suppose that the singular proposition exhausts the content is to commit to what Perry called “the subject matter fallacy.”⁶ This is clearest in the case of identity statements. One might utter “Donostia is San Sebastian” not to convey the information that the city is self-identical, but that the two names refer to the same city. This might be dismissed as “mere pragmatics,”—i.e., a matter of Gricean conversational implicatures—, but Perry thought that if we are more critical about the role that propositions play in semantics, there is a better explanation.⁷ If we shift our attention to truth-conditions, we see that the truth of “Donostia is San Sebastian” requires a bit more of the world that the truth of “Donostia is Donostia.” He argued that by introducing “reflexive” truth-conditions or, more generally, contents, we can explain the cases where the singular proposition doesn’t capture all of the information that an utterance can be used to convey, without giving up the central insights behind referentialism. Hence the word “critical.”

But that doesn’t mean that pragmatics drops out of the picture. Critical Referentialism takes some of the pressure off pragmatics, while at the same time offering it some new tools. Perry was working in the second and expanded edition of Reference and Reflexivity at more or less the same time that he and Korta were completing Critical Pragmatics. It just seemed natural to use “Critical Pragmatics” to refer to the approach to pragmatics that incorporated Critical Referentialism, especially for two authors ignorant of the uses to which the phrase had already been put. Also, they were being critical about some basic theses assumed by the contemporary pragmatic theories they were familiar with. In particular, they thought that pragmatic theories didn’t fully assume the theoretical consequences of adopting Austin’s and Grice’s lessons on language as action; that they were mostly focused on the hearer’s understanding rather than the speaker production of linguistic utterances; and that they committed the fallacy of the unique truth-conditional content of utterances.

Critical Pragmatics was critical with other theories and offered an alternative. That was another reason for the adjective in the title. Also, Korta and Perry think that pragmatics is critical in the study of language, in the sense of being crucial. The title was intended to exploit the ambiguity of “critical,” which provided several things they advocate.

Korta and Perry said something along those lines in the book, but apparently not very explicitly, since, during this dozen years, they have received multiple messages asking about the reasons for the title. We have also found some citations to our work in papers that have apparently little to do with our framework.⁸ So apologies are due, in this case, for using a possibly misleading label. It is probably too late to rectify, and a bit of an overstatement to call this a “misconception,” but we thought we should clarify things.

An important theme in CP is our distinction between the truth-conditions of an utterance or a belief and the proposition that is expressed or believed. The next six misconceptions are all connected with various ways of grasping what we are getting at with this distinction.

Misconception Two: CP might better be called “Multi-propositionalism” or “Pluri-propositionalism.”

Perry’s views on Reference and Reflexivity were a development of his rejection, in “The problem of the essential indexical” (1979), of what he called “the doctrine of propositions.” This is basically the view that propositions are the objects of the “propositional” attitudes, and that belief, for example, consists in a relation to a proposition.

Critical Pragmatics incorporates the rejection of the doctrine of propositions, and explores further its implications. Beliefs and other so-called intentional states do not consist of minds standing in relations with propositions. Beliefs, desires, intentions and the like are states of the brain/mind that are typically caused by objects that play various roles in our lives: objects seen, heard, touched, tasted, remembered, and so forth. Combinations of such states typically cause actions that affect these objects and other that are related to them. For example, the belief that there is a cookie in front of one and the desire to eat a cookie typically causes one to pick up a cookie and eat it. It is these states, and combinations of them, that have typical causes and effects. (See Armstrong 1968; Lewis 1972).

The working theory of attitudes, sometimes called “folk-psychology,” is available to most humans, who can’t see brains and know relatively little about them in other ways. It developed long before people were aware of the importance of brains. What Armstrong (1968) calls “The causal theory of mind” classifies mental states in terms of their typical causal roles.⁹ An important part of the theory is assigning truth-conditions to beliefs, satisfaction-conditions to desires, etc. This is what propositions are for. They are abstract objects that we use to classify attitudes in terms of truth- and satisfaction-conditions. Utterances are intentional activities, motivated by beliefs, desires and intentional states, and inherit the feature of being usefully classified with propositions.

These truth- and satisfaction-conditions arise from the typical causes and effects of a given intentional state. Centrally and roughly, beliefs have truth-conditions and desires have satisfaction conditions, and the combination of a belief and a desire will motivate actions the results of which will satisfy the desires if the beliefs are true. Such truth- and satisfaction-conditions can be represented by propositions, but intentional states do not consist in having relations to such propositions. The truth-conditions of utterances derive from the truth-conditions of the intentional states that motivate them, or the truth-conditions of intentional states in others the speaker desires to cause. This is all oversimplified, of course, but we hope it clarifies the role we think propositions have in talking about mind and action, as opposed to the picture that underlies “the doctrine of propositions.”

Looked at in this way, different propositions may be useful in classifying the same attitudes. For example, if Kepa’s utterance u is of the sentence “I am a Basque,” we can characterize u’s truth-conditions with the proposition that the speaker of u is a Basque, or that Kepa is a Basque, and in various other ways. It is natural to call this account “multi-propositionalism,” since there are multiple propositions that can do a classification job, depending on interest, what is given, and so forth. But the phrase “multi-propositionalism” is a bit misleading. In keeping the focus on propositions, it suggests to many that our view is simply a modified doctrine of propositions where the relation between intentional state and proposition is “one-many” instead of “one-one,” a simple step from mono-propositionalism to multi-propositionalism.

Actually, we have used this misleading terminology at times. Korta (2007), for instance, argued against mono-propositionalism, emphasizing the virtues of a multi-propositionalist approach, closely connected to the Reflexive-Referential Theory or Critical Referentialism. Jesus M. Larrazabal told him that he should use “pluri-” instead of “multi-,” because, while the former prefix means a possibly austere several, more than one, the latter means many and can suggest a multitude of propositions populating the space around an utterance or intentional attitude. “Pluri-propositionalism” was the term chosen then, and it made its way to Critical Pragmatics (on pages 92, 94, 138; “mono-propositionalism” appears on pages xii, 158 and 160).

Since then, and after realizing the potential confusions that it might provoke, we have tried to avoid these terms altogether. Apparently, however, “pluri- propositionalism” and “multi-propositionalism” are catchier than “Critical Pragmatics” or its predecessors “Reflexive-Referential Theory” and “Critical Referentialism,” because their use is quite common. These terms are regularly used, for instance, by brilliant philosophers like Eros Corazza (2012), Carlo Penco (2010) and Richard Vallée (2018), who like our view.¹⁰

But whatever their virtues, we think now we should discard the terms “multi-” or “pluri-propositionalism” for designating our view, and use “Critical Referentialism” (CR) or “Critical Pragmatics (CP)” to avoid confusion. So, to emphasize, CP does not endorse the doctrine of propositions. Intentional attitudes do not consist in a relation, neither to one proposition nor to many of them.

3 The expression of propositions

Actually, one of the most common misconceptions about CP, possibly invited by the use of the mentioned misleading tags, is the following:

Misconception Three: An utterance expresses multiple propositions.

The verb “express” is part of ordinary language and we use it regularly in different ways. We talk about expressing ideas, opinions, feelings, and even expressing oneself. But it is very problematic when it is used to talk about the relation between a belief or an utterance, and a proposition. As we see it, in philosophical conversations, at least, the default use of “express” is for a relation between an utterance and its referential content, clearly suggesting that this relation is constitutive of saying or believing, rather than an important fact about an assertion or belief. When used in this way, we are likely to fall into the “doctrine of propositions”: Saying that so-and-so consists in expressing the proposition that so-and-so.

If, however, by “expressing a proposition” one simply means saying something that has certain truth-conditions, which are captured by a given proposition, that’s fine. But this is not the usual meaning of that phrase. It seems pretty clear to us that, in its uses in philosophy of language, the phrase “express a proposition” has been co-opted by the traditional doctrine of propositions.

This being so, saying that, according to CP, an utterance expresses several propositions, is at best very confusing and, at worst, wrong. For instance, despite his highly interesting applications of CP to a number of important linguistic phenomena, Vallée has sometimes attributed to Perry a “Multiple Proposition View,” which “introduces a suggestion according to which an utterance of a sentence expresses many propositions for semantic, rather than syntactic, reasons” (Vallée 2005 [2018]: 60); or a “Manifold View of Propositions,” according to which “an utterance of a sentence expresses many propositions.”(Vallée 2008 [2018]: 105.)¹¹

CP assigns to the referential content of an utterance or belief an important role—CP incorporates Critical Refentialism, after all. It is the default for the content of beliefs, utterances, and the like, and it is usually what the speaker intends the hearer to come to believe. We could use the term “express” for the relation between a belief or an utterance and its referential content. And the term can be used in other cases for other contents. But, to emphasize, if we take “expresses” to be the relation between an attitude and its propositional object, in the spirit of the doctrine of propositions, this is wrong.

What CP has repeatedly asserted is that the referential content of a statement is what corresponds by default to what referentialist philosophers call “what is said” or “the proposition expressed.” But CP has also repeatedly asserted that this is not always the case. When making identity statements, for instance, we’ve claimed that what the speaker says is not the referential content, but some utterance-bound or “hybrid” content. This is what else has to be the case, given not only the utterance, but also some additional facts, though not enough to yield referential contents. And likewise for what the speaker says via the that-clause in an attitude report. Often times it is not the referential content but some hybrid content which better explains what the speaker is talking about. In the case above the person who says “Donostia is San Sebastian” intends to convey the reflexive content of the utterance with the reference of “is (the same city as)” fixed, but not the reference of the names. Furthermore, as we have discussed more than once, there are various roles the notion of “what is said” is supposed to play, which sometimes leads to replacing it with other notions like the “locutionary content.”¹²

Given all this, it might be advisable to talk about the belief expressed rather than the proposition expressed by a statement. On our view, a speaker does not express abstract theoretical entities. Rather, she expresses particular mental states: she expresses beliefs—with statements; intentions—with promises; desires—with requests; and so on. These mental states play crucial causal roles in the production of a statement or other sort of utterance. A causal role that propositions, or any other abstract theoretical entity, cannot play. In a paradigmatic communicative situation, for example, with a statement, a speaker expresses a belief with the same referential content as her statement; a belief she holds, if she is being sincere; a belief she does not hold, if she is being insincere, but a belief about which she is publicly committing to, as for its truth.¹³

A speaker, however, does not express multiple beliefs with a statement, or other type of utterance. Paradigmatically, with a statement, the speaker expresses one belief, and only one belief. So, once again, CP does not claim that a single statement expresses many propositions, and it doesn’t claim that one single statement expresses multiple beliefs either. What CP claims is that the belief expressed has more than one kind of truth-conditions. In particular, that it has reflexive, referential, and hybrid truth-conditions. But we should be careful, because this last part of our claim has sometimes been misunderstood, leading to the following two misconceptions about CP.

Misconception Four: A statement and the belief expressed in making it have exactly the same truth-conditions.

It is a common assumption that the statement and the belief expressed by the speaker when/by/in making it share the same content, that is, that they have the same truth-conditions. According to CP this is true, but with important qualifications. Suppose Perry uttters u: “Korta is a Basque.” His belief and the statement it motivates have the same referential truth-conditions, that Korta is a Basque. But the reflexive truth-conditions of the motivating belief and of the utterance are not the same. The reflexive truth-conditions of the motivating belief b will have b as a constituent, and not u. The reflexive content of the utterance u will have u as a constituent, and not b.

In other words, different statements and beliefs can, and often do share their referential truth-conditions. But they do not share their reflexive truth-conditions. Statements have statement-bound truth-conditions,—truth-conditions with the statement itself as a constituent. Beliefs have belief-bound truth-conditions—truth-conditions with the belief itself as a constituent. The belief-bound truth-conditions of the belief expressed by a statement, which when dealing with singular terms are what we sometimes call notion-bound truth-conditions, are relevant to explain issues like the cognitive significance of statements, but it is important to keep in mind that they are contents of the belief motivating the statement, not of the statement itself.¹⁴

Another thesis on beliefs and other propositional attitudes that has been (wrongly) attributed to us, often to then argue (correctly) against it, is the following one:

Misconception Five: A belief state involves believing multiple propositions.

Beliefs—or cognitions, as Perry sometimes calls the mental states under discussion in Reference and Reflexivity—have several layers of truth-conditions or contents—from reflexive to notion-bound to referential. But this does not mean that, being in a a belief state, an agent believes multiple propositions. Any reader of the first few pages of CP will likely come to believe that Kepa Korta is Basque. They will believe this via their notion of Kepa Korta, which will be part of their belief-state. According to CP, for that belief to be true, their Kepa Korta notion will have to be of Kepa Korta, the person their belief is about; i.e. the person who is part of the referential content of the belief. But the agent, the reader in our case, may not believe this is the case, at least not in the ordinary sense of “believe.” For one, they may not know what a notion is. After doing the reading, they will be attuned to the fact that this is so, that their notion is of Kepa Korta. But people are attuned to many facts, about which they don’t have the concepts necessary to believe, in the ordinary sense.

If the reader masters the whole book, acquires the necessary concepts, and is convinced of the approach, they will believe, not only that Kepa Korta is Basque, but also that the person their Kepa Korta-notion is of is Basque. However, if this reader finds the theses of the book unconvincing, they may still not believe this. According to us, they would still be attuned to it.

The distinction between belief and attunement can be illustrated by John Searle. He uses proper names competently, indeed his language is usually quite elegant. That means that he is attuned to the way proper names work—that an utterance of a proper names refers to the object to which the utterance has a rather complex historical connection (in our humble opinion). But he doesn’t believe that; in fact, he has denied it vigorously. Philosophers are typically attuned to many truths they deny. No doubt including us.

Misconception Six: Believing consists in a relation between a subject and multiple propositions.

To quote ourselves from a recent paper:

Saying, believing, hoping, and other “propositional attitudes” do not consist in relations of a speaker or thinker to propositions, although they induce such relations. One of the authors is 79 years old. This fact puts him into a relation to the number 79 and the numeral “79” because they are the number and numeral used to classify and refer to his age. But being 79 years old doesn’t consist in having these relations to numbers and numerals. It involves living for a long time, however we measure or refer to lengths of time. Similarly, having the belief that the Canary Islands are beautiful induces a relation to the proposition that the Canary Islands are beautiful. The belief induces the relation to a proposition, just as a person’s height induces a relation to the number assigned to that height by some system of measurement. But the belief does not consist in the relation to this abstract object. It consists in having memories and images of the Canary Islands and a concept of beauty that these memories and images fit. (de Ponte, Korta and Perry (forthcoming a): 7–8)

Taking into consideration this and the previous misconceptions, it might be a good healthy measure to stop using the term “proposition” altogether when doing CP or Critical Referentialism. Not only because it leads to confusions about the name of the theory. Innocent talk about propositions expressed easily and usually leads to talking of propositions asserted, believed or grasped, and of abstracts objects being the bearers of truth-values and holding causal relations.

Using this jargon, we tend to forget that it is particular cognitive episodes, such as utterances and beliefs, which are true or false, satisfied or unsatisfied, and so on. Also, we might forget that these cognitive episodes are the ones that are causally efficient, the ones that have causes and effects. This is important to incorporate our theory about language and communication with both folk psychology and speech act theory; i.e. our best theories about human action and communication. Cognitive episodes, such as beliefs or desires, play an important causal role in our actions and, among those, in our linguistic actions—me wanting a cookie, and believing that the small round object in front of you is a cookie, will normally cause me to either move my body in appropriate ways to get the cookie or to say “please, give me a cookie.” Cognitive episodes such as utterances are usually part of the cause of others doing things—my utterance of the sentence “please, give me a cookie” will normally, assuming there are cookies around and the hearer is a nice person, cause the hearer to give me a cookie.

Sticking with the use of “propositions,” and with phrases like “expressing propositions,” “grasping propositions,” “believing propositions,” or the like, we either end up embracing the old doctrine of (one) proposition, rejected by Perry in the 70’s, or embracing the newer doctrine of multiple propositions, rejected by CP. Therefore, we think it is best to leave proposition-talk aside for the moment.¹⁵

4 Two levels

Despite the labels sometimes used to refer to CP that involve a plurality or a multiplicity of truth-conditions, several people take CP as a variant of a two-dimensionalism of sorts. According to this view, CP essentially postulates two—and only two—kinds of content, truth-conditions: reflexive and referential.

Misconception Seven: There are only two kinds of truth-conditions: reflexive and referential.

Accompanying this claim, it is assumed that each of these two levels would account for problems of its own. For instance, the reflexive level would account for epistemological phenomena like learning from hearing utterances, while the referential level would be the level relevant for issues involving metaphysics (Clapp and Lavalle Terrón 2019). Consider Frege’s puzzle about identity statements. On the one hand, “Hesperus is Phosphorus” is a necessarily true statement, as much as “Hesperus is Hesperus.” At the referential level, we have the self-identity of a single planet, Venus. The apparent contingency of “Hesperus is Phosphorus” is just epistemic: it is possible for a competent speaker to believe that it is false. It is a truth known a posteriori, and this fact is reflected in its reflexive truth-conditions.

Once again, this is not entirely correct. According to CP, there are more than just two kinds of truth-conditions, and it is not right to assign issues of metaphysical and epistemic modality (or semantic versus cognitive issues) to referential and reflexive truth-conditions, respectively. Similarly, it is not always correct to identify the referential content with what the speaker states. Often times, other kinds of hybrid truth-conditions turn out to be the relevant ones to account for these and related issues. So, it is not right to reduce the array of levels of truth-conditions to just two, or to identify CP with a two-dimensionalism of sorts.

There might be, however, some historical reason for this wrong association between CP and two-dimensionalism. At an early stage of his reflexive-referential theory, Perry (1988 [2000]) formulated a twofold distinction between

• the proposition expressed by an utterance; and

• the proposition that the truth-conditions of the utterance are satisfied;

adding that the former—the proposition expressed—could be true if the utterance had never occurred, and that, since the other has the utterance itself as a constituent, its existence is contingent: it is, in a sense, a proposition created by the utterance.

Both can be regarded as singular propositions ...But ...we should equate the cognitive significance [i.e. cognitive content] with the proposition created by an utterance, not the proposition the utterance expresses (Perry 1988 [2000]: 197).

Interpreted literally, one could say, not only that issues regarding cognitive significance should be accounted for appealing to reflexive truth-conditions, but even that, according to Perry, the cognitive significance should be equated with them.

We won’t say much about this, we believe it is pretty clear, both in other works by Perry and in CP, that this is not the intended interpretation. But it is worth making some remarks, to clarify the quote above:

• 1. Perry’s distinction between these two kinds of propositions is framed, among other things, by the discussion of Wettstein’s important essay “Has semantics rested on a mistake?” (1986) and his distinctions there. In this context, there are various notions of “cognitive significance” at stake. Wettstein’s notion is not Kaplan’s notion, for example. In that paper, Perry uses “cognitive significance” in Wettstein’s sense as something like the cognitive content that could be characterized, or even equated with a proposition:

  It seemed to me that the way Wettstein used the term [“cognitive significance”] and its close cousins like “cognitive content” in discussing his examples required that a cognitive significance be a proposition, that having one be a property of utterances, and that the cognitive significance of an utterance be something a competent speaker recognizes. Given those requirements, I think the concept I develop in this paper does pretty well. But the concept that meets these requirements will not be the right object to individuate thoughts by their psychological role. (Perry 1988 [2000]: 205 (Afterword)).¹⁶

• 2. With the development of CP, it should be clear that there is not just one reflexive “utterance-created” propositional content, but rather several. Also, that they are all used to resolve problems of cognitive significance, and that we do not, and cannot, equate one single content with the cognitive significance of the utterance.

CP assumes an array of hybrid contents between reflexive and referential—and, also, beyond referential content, as it is the case with designational (Perry 2001[2012]) or referential* (Korta and Perry 2011) contents—, which are the result of incrementally taking on board—from the theorist’s perspective—facts about the utterance—related to the sentence used, the circumstances of the utterance, and the speaker’s plan. It can be concluded that the two dimensions of two-dimensionalism are included among the layers of truth-conditions of CP, but reducing the latter to the former is a reductive misconception.

5 Metalinguistic truth-conditions

In logic and linguistics, a metalanguage is a language used to describe another language, often called the object language. The idea is that one can speak and understand a language without having the capacity to use a meta-language, so it is implausible to suppose that the contents of our utterances at any level are metalinguistic.

In discussions, we have often heard that the use of reflexive contents in understanding the pragmatic reasoning is dubious, since these contents are “meta-linguistic.” But pragmatic reasoning does not require meta-linguistic beliefs or knowledge. This wrong attribution can also be found in some interpretations of CP or of Perry’s views.

Marga Reimer (2002) says, about Perry’s account of proper names in Reference and Reflexivity:

Perry’s reflexive content (...) captures the insights of meta-linguistic accounts of proper names promoted (at one time or another) by Frege, Russell, and even Mill himself. But is Perry’s “cognitive significance” what Frege was so worked up about in the opening paragraph of “On sense and reference”? (...) Frege’s cognitive significance is not explicable by appealing to meta-linguistic considerations of the sort Perry appeals to via his reflexive content.

Ray Elugardo (2013), discussing Korta and Perry’s (2011) account of sub-sentential utterances, claims that “reflexive content is problematic insofar as it meant to be a meta-linguistic, truth-conditional, content of sub-sentential utterances” (Elugardo 2013: 106). More recently Botterell and Stainton (2017) follow the same thread.

In a nutshell, they misunderstood us: Our reflexive truth-conditions are not metalinguistic. This is the eighth misconception in our list:

Misconception Eight: Reflexive contents are metalinguistic.

Consider the statement, “Donostia is a city.” The referential content of the statement is the singular proposition that Donostia is a city; it is about Donostia, not about the name “Donostia” or the other words in the sentence. It is true, so it states a fact, a geographical fact. The utterance uses language, but it does not state a linguistic fact. But consider,

“Donostia” is a name.

Quotes are used to indicate that a name is being mentioned rather than used; they are meta-linguistic devices. The utterance uses metalinguistic devices to refer to the bit of language it is about, the name “Donostia.” But the fact stated is a linguistic fact, not a meta-linguistic one (whatever that might mean exactly). We use quotes as metalinguistic devices to refer to names and other expressions which are constituents of reflexive and hybrid contents. But that does not make the contents metalinguistic rather than linguistic, any more that the fact that we use words to refer to cities makes it a linguistic fact, rather than a geographical one—i.e. that San Sebastian is a city.

Suppose María tells Kepa, who is looking for his copy of Critical Pragmatics:

(u) It is on the top shelf.

Following CP, and simplifying a bit, the reflexive truth-conditions of (u) are

(u$_{rx}$) That there is an object x that the speaker of (u) is referring to and x is on the top shelf.

According to Elugardo, and other interpreters, the fact that the reflexive truth-conditions are about (u) itself make them meta-linguistic. But (u) is not a piece of language; it is an act that involves a piece of language. In other words, (u$_{rx}$) are not truth-conditions of a sentence, not even a sentence in a context, but an utterance, an act that might or might not involve a complete sentence. They can involve sub-sentences or no sentence at all—in CP, we use “utterance” à la Grice, in a broad sense, including non-linguistic communicative acts.¹⁷

6 Metaphysics versus epistemology

Adding to his remark about the metalinguistic nature of reflexive contents, Elugardo notes that

Very young English-speaking children do not have difficulty understanding utterances of “on the top shelf,” but it is unlikely that they possess the requisite meta-linguistic concepts needed for understanding utterances of (5.13) [(5.4)].¹⁸ So, just on empirical grounds alone, it is doubtful that the language faculty generates anything like complex reflexive contents as outputs. (Elugardo 2013: 101, fn 17.)

We think that there is another misconception of CP here. One that has to do with taking reflexive truth-conditions to be a representation of the linguistic meaning of an utterance; a representation that would constitute the first step of what the hearer has in mind in his utterance comprehension process. Once again, this is not right.

To begin with, the theoretical notions and concepts we use to characterize the various levels of truth-conditions of an utterance are not restricted to those that are typically supposed to be known by competent users of the language.

They are the concepts and notions that the theory uses to make explicit the truth-conditions that competent speakers and communicators, including very young ones, are attuned to. The apparent complexity of reflexive truth-conditions is due to the complexity of our formulation of the truth-conditions, but these are not representations in the hearer’s mind.¹⁹

The possible source of this sort of misunderstanding is the conflation of the metaphysics of meaning and the epistemology of interpretation, (See Devitt 2013, 2019; Korta and Perry 2019a, b). In fact, Elugardo, among others, has taken CP to be a theory about the epistemology of utterance comprehension, such that the various levels of truth-conditions represent the hearer’s inferential steps in that process. That is another misconception.

Misconception Nine: The variety of contents of an utterance reflect the series of steps in the utterance comprehension process.

The hierarchy of contents we supply, from reflexive contents through hybrid contents to referential contents, is not a theory of the steps involved in understanding an utterance. No doubt, the various levels of contents play a role in the hearer’s understanding process, but we do not give a precise account of that process as a serial inferential process. Actually, CP may be compatible with various theories. In Chapter 11 of Critical Pragmatics, we show its compatibility with relevance theory, but it might be compatible with others. It is important to keep in mind, in any case, that hierarchy of contents should not be considered as a hierarchy of language-of-thought sentences in the speakers-hearers minds.

In conversation, we often start by identifying the referential content without much thinking. We consider lower levels when we try to figure out why the speaker said in a particular way. And sometimes, when we can narrow down the referential content possibilities, it is helpful to consider what they imply about the lower levels. In speaking with people who mumble, for example, we often grasp the likely referential content intended, and then use it to figure out which words were used. And sometimes there is a back and forth to come up with the likeliest hypothesis.

For example, Paul Grice once came up to Perry after Perry gave a talk at an APA meeting and said, “That was really an ex-mumble-mumble talk.” Perry thought he intended to say either that it was an excellent talk or that it was an excrement of a talk. Perry decided he meant that it was excellent, because he had pleasant expression and why would he walk all the way to the podium to say that it was an excrement of a talk? Also, the thought expressed by the latter would not require such a crude term; “lousy” would fine. So Perry’s reasoning went from likely referential contents to likely reflexive contents, and back.

Perry, of course, may have been wrong. And he did not think these contents; nor did he consciously make inferences from one content to another. His reasoning was pretty automatic, and it involved many different observations and previous beliefs, from Grice’s facial expression, to his believe that Grice was a nice person, not prone to rudeness or unnecessary cruelty.

The variety of truth-conditions of an utterance is a theoretical tool, aiming at encapsulating the different levels of truth-conditions that are relevant on each case. A means to reconstruct, theoretically, what is said by an utterance, how is said and, perhaps, what is understood, which is compatible with various neo- and post-Gricean theories of communication and theories about what happens in the speaker-hearers’s minds in speech production and comprehension.

7 Conclusion

We are certain that there are many valid criticisms to be made of Critical Pragmatics, understood just as we meant it. Philosophers need to have somewhat conflicting attitudes towards the theories they put forward. They have to believe what they are saying is correct, otherwise they lose the motivation to write and explain what they think. But, at least as we conceive it, the history of philosophy consists of brilliant people putting forward ideas that turn out to be not quite right, and in many cases dramatically wrong. What are the chances of a team of not-so-brilliant philosophers coming up with ideas that have everything—or even most things—right?

We try to explain what we believe to be true, at one level, as best we can. At the same time, or at least later in the day when we go out to get a drink, we have no doubt that we don’t have everything right, and hope that we have at least put forward some ideas that will add to the understanding of the phenomena we discuss.

So we are happy when others confront our ideas and explore how they measure up to their own concerns and insights. Most times, these confrontations turn into fruitful debates and exciting new philosophical challenges, either because they show us new aspects of our theory we had not thought of—new applications or new implications—or because, oftentimes, they bring about new limitations that need to be overcome.

The point of this paper is to help clarifying some aspects of Critical Pragmatics, focusing on those that have been most often misunderstood. We also incorporate some of the insights we’ve gained in discussion with other philosophers and linguists during the last decades. The focus, however, is put on clear misinterpretations of the theory, in the hope this helps avoiding criticisms to things we didn’t mean.