Attitudinal objects and referential interpretation for performative utterances

Abstract

The present paper explores resources provided by ontology of satisfiable objects (an ontologico-semantic framework proposed by F. Moltmann) for development of a referential analysis of performative utterances. In order to position ontology of satisfiable (including attitudinal and modal) objects among other ontological options for theories of meaning, I compare it to ontology of propositions and ontology of acts and events. Then I address the notion of performative utterances and its ontological background in Austin, as well as assertion-oriented accounts of performative utterances, contending with the initial Austinian account. I put forward the hypothesis that adoption of satisfiable objects ontology is a way to an integrative vision of truth and felicity of performative utterances. I discuss the logical form ascribed to performative utterances by Moltmann and then outline my proposal for their analysis within the framework of ontology of satisfiable objects.

Keywords:
Attitudinal Objects; Modal Objects; Natural Language Ontology; Performative Utterances; Sincerity

1. Introduction

The aim of this contribution is to investigate capabilities of ontology of satisfiable objects (an ontologico-semantic framework developed in the recent works of F. Moltmann) for obtaining a better account of the classical topic on the border of philosophy and linguistics - ‘performative utterances’¹. The strategy of my work is to address this topic in a wider context of fundamental problems of ontology and philosophy of language.

I state that ontology of satisfiable objects (including attitudinal, modal and intensional objects) is a natural solution to some problems concerning reference, which makes it a good alternative to the standard Austinian views. However, the categories of entities postulated in this ontology may be regarded as ‘suspicious’ nominalisations from the points of view of more habitual contemporary ontologico-semantic doctrines. I juxtapose ontology of satisfiable objects to, on one hand, ontology of classical propositions and, on the other hand, ontology of acts and events, in order to show that addressing satisfiable objects is plausible and promising.

The discussion of linguistic performativity starts with the Austinian promotion of felicity/infelicity as the parameter for qualification of performative utterances counterposed to truth/falsehood (the idea being supported by some recent works, e. g. Jary 2007, Corredor 2011, Stojanović-Prelević 2019). Based on the works of J. Austin, I address the ontological background of this idea, entitling it ‘ontologicalised behaviourism’, which is to much extent opposite to ontology of satisfiable objects. Considering the efforts of supporting the assertive character of performative utterances (e. g., Lewis 1970; Bach & Harnish 1979, 1992; Ginet 1979; Recanati 1987; Searle 1989), I share their motivation but point out that they, nevertheless, preserve the Austin-style double vision of parameters for qualification of utterances. I put forward the hypothesis that adoption of satisfiable objects ontology is a clue to an integrative account of truth and felicity, by virtue of the more general notion of satisfaction.

Addressing the analysis of performative utterances in Moltmann, I propose the way of further development of this account. In my present proposal, I distinguish performative meaning, possible modification by performative inference and evaluation parameters (such as truth, sincerity, felicity) as the aspects of which linguistic performativity is comprised.

2. Reference, ontology and performativity

A sign is an entity that, in the eyes of some interpreter, stands for some other entity called referent…

…or that’s what we are told by standard textbooks on semantics. However, one may face difficulties just searching for a simple illustration of this in the real world. Sometimes it seems to be the phenomenon known as inscrutability of reference, addressed in the works of V. W. Quine², D. Davidson³ and H. Putnam⁴. But sometimes the issue may be just ontological prejudice, caused by some biased metaphysical views. Anyway, as the mentioned philosophers insisted, a reflexion on ontology is very important for settling problems with reference.

Consider a simple example of a sign: a green traffic light turned on at a crossroad. Or another simple example: a shot of a signal pistol at a running competition, after which participants start to run.

What is the referent of the sign in these examples? Most naturally one would answer here, maybe after trying some clumsy other options, that the referent is a permission of certain kind.

However, such answer may be regarded as unacceptable from the ontological point of view: things like permissions hardly fit the usual philosophical conceptual schemes of physical objects, or sense data, or even Platonic ideas.

The aim of the present paper is to investigate the resources for referential analysis of performative utterances, provided by ontology admitting such things as permissions - and others referred to with the umbrella term satisfiable objects, - following a proposal of F. Moltmann.

It’s particularly interesting to address performative utterances this way, since the very notion of linguistic performativity, as it was proposed by Austin and preserves in some recent literature, contends with the referential interpretations of this kind. The non-verbal signals from the examples are, in a sense, performative utterance. With the referential interpretation mentioned above, they are easily convertible to a verbal referential utterance ‘There’s the permission’ (with the implied details about the type of the permission, as well as localisation of its validity in space and time), which can be qualified as true or false. However, the standard Austinian view deny this possibility, claiming that such utterances should be evaluated only as felicitous or infelicitous.

Before proceeding to analysis of performative utterances by means of satisfiable objects ontology, I’ll discuss the position of this conception among other important ontological options for referential theories of meaning.

3. Ontological options for referential theories of meaning

3.1. Classical propositions in theories of meaning

The standard Fregean view on the relation of propositions to reference is a bit complicated. It states that in direct linguistic contexts propositions aren’t referents of sentences, but, being the senses of sentences (‘thoughts’ expressed by sentences), propositions determine their referents (which are truth values). As structured intelligible entities, consisting of individual and predicative components, propositions determine also referents at subsentential level. In indirect linguistic contexts, propositions become referents themselves.

In the contemporary literature the notion of proposition is connected rather to a number of heterogenous views than to a single idea. In some of them, propositions are treated as structures (‘structured propositions’) constituted by objects, their properties and relations. There’s a discussion about the

ontological nature of that constituents: whether they are extensional⁵ or intensional⁶. In some views, propositions are reduced to sets of possible worlds, with various suggestions about the ontological nature and relation between these two categories⁷. In some views, propositions are regarded as primitives⁸.

However, looking for the common feature behind all these options, one can point out that the classical notion of proposition typically presupposes abstract, inter-agentive (or rather non-agentive) and cross-temporal (or even atemporal) nature of semantic matter of sentences. It’s that invariance that (according to the classical theories of propositions) can be grasped in different synonymic sentences, by different agents, in different spatiotemporal locations and circumstances, thus being in a sense independent of such factors.

The Frege’s conception of thought (Gedanke in German), to a great extent grounding the further theories of propositions, claims that ‘thoughts’, which are senses of sentences, aren’t created but revealed by humans⁹. Frege argues pointing, as an example, to the theorem of Pithagoras: as he believes, (1) the theorem of Pithagoras is a though and

(2) it is an objective feature of the reality, not a voluntary product of human fantasy. Besides generalisations like the theorem of Pithagoras, he considers to be ‘thoughts’ the contents of univocally formulated factual utterances, like ‘The population of Germany in 2025 is about 84 million people.’

For Frege, the task of the project which one may now call semantics is to relate linguistic expressions to objective features of reality grasped by ‘thoughts’. All ‘subjective’ components of linguistic and communicative activities (e. g., the subtle difference between ‘and’ and ‘but’) are outer of scope of semantics so understood.

There’s a widespread tradition to consider Frege to be a Platonist. However, this stance may be not so well-founded as it could seem at first glance, see Ricketts (1986), Reck (2005). In my view, the strategy of ascribing Platonism to Frege is hardly winning mainly due to we don’t know exactly what Platonism is (if we seriously mean by that the philosophy of Plato). I would rather suggest a Cartesian interpretation with the following theses: (1) the reality can be adequately grasped by the intellect; (2) there’s only one proper way to approach the reality; (3) and there’s the common (standard of) intelligence, the conceptual and methodological invariant of correct thought, grasping anything that is objectively true and throwing away subjective aberrations. Frege actually overtly argues for conceptual invariance of thought in ‘Negation’, telling that a judgement doesn’t actively divide the reality into pieces (as some other thinkers think) but grasp what is already given as constituents of the reality¹⁰.

Thus, there’s only one science, meaning by that the theory accumulating all true thoughts. The latter thesis is, as far as I see, implied by the following Frege’s argumentation by contradiction: ‘If every thought requires a bearer […] then it would be a thought of this bearer only and there would be no science common to many […] Each of us occupies himself with the contents of his own consciousness […] and it would really be idle to dispute about truth’¹¹. So, for Frege, the task of what we call semantics is to correlate linguistic expressions with the contents of the Science. However, the idea of the united science today turns rather into an argument contra, since the development of logic and philosophy of science shows rather persuasively that there’s no theory accumulating all objective truths, that sciences develop non-cumulatively, and that conceptualisations of reality may vary.

As for the argument concerning the theorem of Pithagoras, it’s useful to notice that there are entities which are intersubjective, in a sense - objective, and still they are agent-dependent and constructed. Actually, there is a large number of well-known things to be described this way: pieces of music, dances, juridical laws, etc. E. g., there may be a musical composition to be played on guitar, composed once by a concrete person, but objectively existing (say, as a musical notation or a record) for other persons.

Someone may say that it’s just an illusion of reproducing one and the same musical piece, as, ultimately, each of us hears and plays music ‘in a personal way’ each time. Nevertheless, at least from the common sense point of view, there is the idea of a musical piece as it is, though it’s not eternal but created once by a musical composer. On the other hand, someone may say that all music is made in Heaven and there is an eternal idea of each musical piece, but it’s hardly compatible with the common sense too.

Should we consider the theorem of Pithagoras to be a constructed informational entity too or to be completely agent-independent, as it expresses an objective feature of the reality, as Frege believes? Actually, we don’t need to be involved in this hard metaphysical discussion now. The problems of meaning and reference addressed here, which are connected with the notion of performative utterances, quite likely belong to the ephemeral and agent-dependent side of being, to changes and events produced by human activity.

So, should we take the notion of act as basic for a theory of meaning?

3.2. Acts and events in theories of meaning

One may consider the category of acts as basic for semantics in two different (but not mutually excluding) ways:

• (1) Addressing acts as an ontological category allowing to describe the eventual side of being, along with other ‘eventualities’: events, states, processes, so on.

• (2) Regarding (cognitive) acts to be ontological basis for propositions.

The first option seems to be trivial today, due to development of Davidsonian semantics and its relatives. Considering acts as semantic matter of some sentences is rather natural. E. g., one may analyse the sentence ‘Davidson wrote a paper’ as a description of a certain action¹². Such analysis is (apparently) easily compatible with the believe that acts are a class of events.

However, there is a question of how acts (as well as other eventualities) relate to non-eventual kind of semantic matters - say, to the contents of a sentence expressing the theorem of Pithagoras. The examples provided above illustrate this question yet another way: they certainly address the eventual side of being (such things as road traffic or run competitions). However, the signs in question doesn’t refer to acts or events as such: a green traffic signal doesn’t refer to a movement, the shot of a signal pistol or a shout ‘You may run!’ doesn’t refer to a start of running. They may be connected to such acts but refer to something else: as said above, it’s naturally to think that the referent is a permission.

Probably, one may think that there is some complex way to reduce such things as permissions to acts and events; but if it’s there, it should be explicated. Also, if one addresses acts and events as ontological primitives, logically just as objects sui generis, a question arises: what are ontological correlates for linguistic constructions of sentential level? One may still adopt here the classical notion of proposition; but, as shown above, it rests on certain rather problematic metaphysical presuppositions. Moreover, it would lead, in a sense, to internal semantic tension between eventual entities at the level of semantic primitives and atemporal propositions at the sentential level.

The second option mentioned above is to regard (cognitive) acts as an ontological basis for propositions (which is anyway a departure from the classical notion of proposition). There is a considerable amount of recent research directed so. E. g., King (2007) claims that ‘Propositions exist and have the truth conditions they do because of the intentions and actions of conscious agents’ (p. 130). Soames (2010) states that ‘propositions are representational because of the relations they bear to inherently representational mental states and cognitive acts of agents’ (p. 7). Hanks (2015) suggests a view that propositions are a type of acts.

I’m not addressing details of these works, as they deserve more place than is available here. However, the very aim of grounding the notion of propositions by addressing cognitive activity, in my view, is interesting but not the most plausible way of developing cognitively realistic picture of how language and communication work. The very notion of proposition now bears an influence of metaphysical theories not friendly to such development.

The conception of satisfiable objects, to which I turn now, is an interesting way of developing semantics on completely new basis, not just nominally substituting the notion of proposition but putting forward an essentially new account of ontology for natural languages and communication. One of its major merits, in my view, is elucidation of relations between acts (including cognitive ones) and their products in linguistic and communicative activities.

3.3. Attitudinal objects and other ‘satisfiables’ in a theory of meaning

The concept of attitudinal objects is developed by Friederike Moltmann in a number of works, such as Moltmann (2013, 2020, 2023, 2024)¹³. The following account is based mainly on her book ‘Objects and Attitudes’ (2024), containing up-to-date exposition of the theory.

The term attitudinal objects is intended to refer to such things as claims, judgments, beliefs, assumptions, hopes, requests, decisions, desires, intentions, ideas, hypotheses. Attitudinal objects belong to the more general category of satisfiable objects, or satisfiables, which includes also modal objects and intensional objects. Modal objects include such social phenomena as obligations, permissions, laws, rules, orders, invitations, strategies, options and such natural phenomena as abilities, dispositions, essences. Intensional objects are such things as searches, debts, purchases.

In my interpretation, the general aim of the Moltmann’s project is to provide a system of ontological categories consisting of entities that, on one hand, has informational nature, one the other hand - are agent-dependent and causally active. Thus, in a sense, it’s an attempt to find the middle way between theories of meaning oriented to classical propositions and act-oriented ones. This account, though rejecting the standard propositional semantics with its notion of propositions as abstract and, so to say, completely objective entities, nevertheless leaves a room for objecthood of ‘propositional’ content.

Attitudinal objects are understood ontologically as (1) particulars (concrete entities) which (2) are agent-dependent (3) and, moreover, mind-dependent, (4) have informational nature (‘come with a content’, by the expression of Moltmann), (5) may exist for a limited time, (6) are involved in causal relations and (7) may be perceivable.

They are divided into three subcategories:

• (1) illocutionary objects - entities corresponding to illocutionary acts, such as claims, requests, promises, suggestions;

• (2) mental objects - entities corresponding to (or being themselves) various cognitive, volitional and certain emotive states of mind, such as beliefs, thoughts, intentions, decisions, hopes, doubts;

• (3) locutionary objects - entities corresponding to ‘Austinian locutionary acts’ - ‘roughly, acts of merely presenting or entertaining a content’ (Moltmann 2024, p. 26).

Semantically, they can be:

• (1) entities for which (some) that-clauses stand for;

• (2) entities denoted by (some) nominal parts of internal structure of a sentence;

• (3) entities corresponding to whole utterances of independent sentences.

Some attitudinal objects are generatively connected to modal objects. As Moltmann puts it: ‘Someone makes a promise and hence is under an obligation to fulfill it; someone makes an assertion and puts herself under a commitment to maintaining its content’ (Moltmann 2024, p. 3).

Attitudinal objects have features in common and in contrast with what is traditionally understood by ‘propositions’, as well as with acts and events:

• - Attitudinal objects are similar to ‘propositions’ as they ‘come with a content’ and have satisfaction conditions (which for such objects as claims, judgments and believes are truth conditions). But, unlike classical ‘propositions’, they are concrete and agent-dependent. Also, as Moltmann stresses, semantically attitudinal objects should not be regarded as senses of sentences; they are ‘entities that can just bear the satisfaction conditions given by sentences’ (ibid., p. 14).

• - The features of concreteness and agent-dependence are what attitudinal objects have in common with acts. But attitudinal objects are content-bound entities having satisfaction conditions, such as conditions of truth, fulfilment, realisation or answerhood (depending on their type); while acts, as well as events in general, have no such; they also have no informational structure, only temporal structure.

An important aspect of satisfiable objects ontology is the distinguishing of non-contain-based and contain-based causation, as well as perception and memory. To illustrate this idea of Moltmann, let’s consider a situation when X tells you that your answer surprised her. What could be the cause of the surprising effect? Actually, there are two (compatible) options:

• - non-contain-based causation: The surprising effect was caused by the very act of answering (it was not expected that you answer at all).

• - contain-based causation: The surprising effect was caused not by the act of answering (which could be quite predictable) but by the content of the answer.

Another very important reason for distinguishing attitudinal and other artificial satisfiable objects from acts is that such objects may relate to acts, by virtue of which they emerge, in a complex way. E. g., a claim can be made by virtue of many subtle movements of articulation apparatus, as well as cognitive acts connected with it and contextual behaviour. Of course, one may say that all that sums in a complex act of claiming. But considering that complex act to be an act of claiming, we begin with the claim produced by it and characterise so the entire complex act due to this result.

What is the relation of such result to a sum of underlying acts, due to which it emerges? Moltmann states that it’s a relation of generic ontological dependence, another example of which is the dependence of a musical composition from a sum of various heterogenous acts summing in a complex process of composing music. The relation of such ontological dependence is considered to be a non-causal one.

Thus, in a sense, attitudinal objects have priority to the underlying acts. E. g., one takes some activity to be an act of claiming only because it results (actually or potentially) in a claim.

That is a good argument for the framework of semantic analysis of natural language sentences, proposed by Moltmann, called ’attitudinal objects semantics’. In this framework, the sentence (3.1)

• (3.1) John made the claim that Mary is a genius

• is analysed as (3.2)

• (3.2) ∃d(make(John, d) & claim(d) & prop([that Mary is a genius])(d)),

• (‘there is an attitudinal object d such that d is produced by John, d is a claim and the satisfaction conditions for d is ‘that Mary is a genius’’).

• Importantly, sentences like (3.1) are considered to be

• semantically basic compared to sentences like (3.3)

• (3.3) John made the claim that Mary is a genius,

• as verbs like ‘claim’ are regarded to be shortenings for more semantically explicit constructions like ‘make a claim’.

One may ask: what is the clause ‘prop([that Mary is a genius])’ if not a formulation of a proposition? Once again: in attitudinal object semantics it stands for satisfaction conditions of a satisfiable object, which for a claim are truth conditions. Supposedly, given one is particularly interested in this part of the analysed sentence, it’s possible to provide a deepen interpretation for it. However, focusing on the structural role of an attitudinal object, one may leave that part without further analysis.

4. Performative utterances

4.1. The idea of linguistic performativity and its ontological background

One of the central claims of speech act theory(ies) is that some sentences in some speech acts are used not for description of current realities but for establishing of new realities, i. e. they are used with performative force. E. g., by saying ‘I pronounce that you’re married’ or ‘I promise that she will agree’ one not (only) indicates something happening but makes it happen (however, only provided that one has such a power).

The origination of the idea of linguistic performativity is commonly ascribed to J. Austin, in who’s works it was presented initially as the distinction of ‘constatives’ and ‘performatives’, which then was claimed by him unsatisfactory and replaced by a more general theory of different speech acts, see Austin (1962, 1979). Yet the latter theory was criticised and modified in the later literature, e. g. see Searle (1969). Nevertheless, the very idea of performativity remains nowadays quite important for speech act theory and its applications, including investigations in spheres very distant from pure philosophy of language, such as cultural and social studies.

Despite of the wide recognition, the conception of performativity is far from being clear and consistent. That can be seen already from the above mentioned discrepancies in the works of its proponents, as well as from a number of critical contributions. The above mentioned sociocultural expansions of speech act theory are apparently excited by some profound insights in Austin and Searle about social norms and types of rules-governed behaviour, but it’s still needed to examine thoroughly the semantic nature of performativity and, so to say, the interface between ontology, semantics and actions, which can in turn affect sociocultural applications of the theory.

To take a quick look on the ontological motivation of the idea of linguistic performativity, as proposed by Austin, let’s consider a quite characteristic example of his argumentation: ‘it’s very easy to think that the utterance is simply the outward […] sign of […] inward spiritual act of promising, and this view is certainly been expressed in many classic places. […] There is the case of Euripides’ Hippolytus, who said ‘My tongue swore to, but my heart did not’ […] Now it is clear from […] example that, if we slip into thinking that such utterances are reports, true or false […] we open a loophole to perjurers and welshers and bigamists’ (Austin, 1979, p. 223).

I would like to call the stance expressed in this citation ontologicalised behaviourism. It appears to be a metaphysical (or at least methodological) denial to admit psychic entities beyond overt behavioural manifestations. As one can see, Austin prefers it also for ethical reasons. However, even if we regard ethical motivation as significant for the matter discussed, Austin’s argument is not decisive, because his theory of performative utterances opens loopholes for perjurers of its own kind, such as those who say ‘with best regards’ as a performative, without the feeling.

Contrary to ontologicalised behaviourism, the existence of the notion of sincereness is an evidence that there are mental (or appearing to be such) objects as correlates of commissive utterances (see more in Section 5.4). Or at least that our descriptive metaphysics, in the sense of Strawson (1959), and natural language ontology is bound to such picture of reality.

4.2. In search of truth for performative utterances

The discussion of performativity, since its beginning in Austin, resulted in a range of positions towards the topic. One margin is the stance of the earlier Austin, according to which performative utterances are (1) actions themselves and (2) are not assertive, thus cannot be evaluated as true or false, but only as felicitous or infelicitous. There are similar views in the recent literature, e. g. Jary (2007), Corredor (2009), Stojanović-Prelević (2019).

This view is problematic due to the issue of inconsistency, which is pointed out in a number of works, e. g.:

‘…this view introduces an intolerable discrepancy between the semantics of certain first-person present-tense verbs and their other-person other-tense variants’ (Davidson 1979, p. 16);

‘The sentence ‘She thereby promises to be there’ can be used to state of another person’s current act […] Why should a mere shift in person or tense, and from ‘thereby’ to ‘hereby’, deprive such a sentence of its power to state that a certain (indexically referred to) act is of a certain sort?’ (Ginet 1979, p. 246).

Such considerations ground the request for common referential interpretation of performative utterances, which is supported by a number of works, intended to justify the assertive status of performative sentences (i. e. that 'Performatives are statements too’, as claimed by the title of one relevant paper¹⁴).

However, these attempts face the opposite problem: when regarding performative utterances to be on a par with (other) descriptive utterances, how can one explain (at least, apparent) specificity of them, in virtue of which the very notion of linguistic performativity has emerged? One of the possible options would be just to deny it. However, usually the specificity in question is more or less acknowledged. Given this, how could it be explained?

One possible way is to claim that performative utterances have double nature, being together the statement of an act and the very act (belonging to certain range of kinds of acts). Thus, this stance differs from the Austin’s one because the latter doesn’t regard such utterance to be a statement, only an act. According to Ginet (1979), such possibility emerges due to the relation between the meaning of a sentence and the context in which it is used, the conditions are the following:

• (1) the context provides a stage-setting needed for фing;

• (2) the utterance ‘I (hereby) ф’ conforms to the natural requirements or rules for фing;

• (3) the utterance ‘I (hereby) ф’ have the external relations required for фing in that context;

• (4) ‘the full, specific content of the objectives required by фing is, in that context, implied by the verb phrase ’ф’’ (p. 259)

• (5) it’s at least theoretically possible to reach the objectives by uttering ‘I (hereby) ф’ in that context.

Worth noting is that, despite of the stressed divergence with Austin, what Ginet provides with this list is rather conditions of felicity, not conditions of truth for performative utterances.

Generally, there are various accounts ascribing the assertive status to performative utterances. They may differ (sometimes in a rather subtle way) in such questions as how a recipient understands and supports the message of a speaker, what inferential mechanisms are involved, etc. An important line is to consider pragmatic force to be a phenomenon emerging in virtue of pragmatic factors, such as contextual inferences.

In Searle (1989), a performative utterance is characterised as a sentence used in the ‘present present’ tense which ‘describes an event that is simultaneous with its utterance’ (p. 557). It’s understood as a declaration, e. g. a locution having performative force, which turns to be a true assertion at the very time of its (felicitous) uttering.

Recanati (1987) also adopts the notion of declaration. He regards performative utterances to comprise of two different illocutionary acts. A declaration is concerned to be a force-neutral act performed directly, while the simultaneous another act happens due to certain pragmatic inferences in a communicative situation.

In Bach & Harnish (1992), the performative force is also considered to be the second component of an utterance’s meaning in a concrete pragmatic context. According to them, facing an utterance as an assertion, the recipient infers additionally its performative reading, assuming attitudes and intentions of the speaker.

In my view, the problematic feature of the theories mentioned above is the very tendency to analyse performative utterances (or their pragmatics-driven interpretations) as double-component. This feature inherits the Austin’s separation of truth/falsehood and felicity/infelicity, despite of the declared intention to abandon that position.

My hypothesis in this context is that ontology of attitudinal and modal objects is suitable for development of an integrative view on truth and felicity conditions, by virtue of the more general notion of satisfiability. Below, I firstly address the analysis of performative utterances suggested by Moltmann, then I proceed to my proposals in that regard.

4.3. The treatment of performative utterances in Moltmann (2024)

Moltmann makes a claim that attitudinal objects ontology ‘allows for a straightforward account of performative sentences’ (Moltmann 2024, p. 128). Her account is based on the presumption that some utterances produce modal objects (such as obligations or permissions), see: ibid., p. 95-96. The relation between utterances and objects produced by them is denoted as ∠. This presupposition allows to ascribe to the sentence (4.3.1), in its usage as a performative utterance, the logical form (4.3.2):

• (4.3.1) You must leave the room!

• (4.3.2) λu[∃d(u∠d) & must(d) & prop([you leave the room])(d)],

which is a lambda-function λu, taking an utterance as the argument, assigning to it the following meaning:

• - there is a satisfiable (in this case, modal) object d produced with the utterance u;

• - d is an obligation (‘must’);

• - the satisfaction conditions of d is ‘you leave the room’.

It should be noticed that this analysis is applied to performative reading of (4.3.1), according to which the obligation emerges in the moment when the utterance is pronounced.

In my view, (4.3.1) may also be read non-performatively, as denoting an obligation that takes place independently of the utterance. To this reading, one should assign the logical form like (4.3.2) but without the first clause, which apparently could be (4.3.3):

• (4.3.3) ∃d[must(d) & prop([you leave the room])(d)],

• but it leaves no place for a lambda-function.

The ability to account for this rather subtle difference between two readings, in my opinion, is certainly an argument for this framework of analysis. However, in the work of Moltmann this distinction is not drawn or at least is not clear, since the analysis proposed for an assertive utterance of the sentence (4.3.4), quite similarly to (4.3.2), is (4.3.5):

• (4.3.4) Mary is a genius.

• (4.3.5) λu[∃d(u∠d & M(d) & prop([Mary is a genius])(d)],

which again is a lambda-function λu, taking an utterance as the argument, assigning to it the following meaning:

• - there is a satisfiable object d produced with the utterance u (but in this case it is understood as an attitudinal object);

• - M(d) denotes that d has ‘word/mind-to-world direction of fit’ - a category, used by Moltmann, that may be roughly taken as standing for assertions;

• - the satisfaction conditions of d is ‘Mary is a genius’.

An issue is that, following this pattern, one apparently should to analyse the non-performative reading of

• (4.3.1) You must leave the room!

• once again as

• (4.3.2) λu[∃d(u∠d) & must(d) & prop([you leave the room])(d)],

with the only non-formalised difference that d in this case is regarded to stand for an attitudinal object of different type. Such treatment of the logical forms of performative and non-performative reading of the same sentence hide the difference between them, pointed out above.

Worth noting, in another place Moltmann states that, in some cases, with a single utterance both attitudinal and modal object can be produced: ‘Modal objects may be produced by the very same acts that produce illocutionary objects, such as acts of requesting, promising, and permitting. A modal object produced by an illocutionary act shares its satisfaction conditions with the illocutionary product […] However, it generally has a different lifespan’ (ibid., p. 30). Thus, it may be a good idea to explicitly denote both attitudinal (illocutionary) and modal objects produced with an utterance.

Now, let’s consider the logical form (4.3.7) suggested by Moltmann for a typical performative sentence (4.3.6):

• (4.3.6) I am hereby making the claim that John is guilty.

• (4.3.7) λu[∃d(u∠d & make(S(u), d) & claim(d) & prop([that John is guilty])(d))],

where S(u) denotes the one who produces this utterance, other as described above.

Comparing (4.3.7) to the examples cited above, such as (4.3.5), one should raise the question about, so to say, counting of attitudinal objects. If for utterances like (4.3.4) ‘Mary is a genius’ the logical form shows that there is an attitudinal object corresponding to the whole utterance, why it’s not so for (4.3.6)? Possibly, the idea behind the analysis proposed by Moltmann is that the claim mentioned in (4.3.6) is the very utterance (4.3.6), which is a claim. However, in this case it seems to be correct to state explicitly this identity. Now, one may see the logical form for (4.3.6) as the more complicated (4.3.8):

• (4.3.8) λu[∃d(u∠d & make(S(u), d) & claim(d) & prop([∃d’[make(S(u), d’) & claim(d’) & prop([that John is guilty](d’) & d∠d’] ) (d))].

As, obviously, here is actually no identity of d and d’, I propose to regard the relation between them as the relation of production, denoted as ∠. Is it a plausible approach, if above this relation was understood as the relation not between two attitudinal objects but between an utterance and an attitudinal object? This question needs to be discussed. To support the possible affirmative answer, one may say that attitudinal objects, ultimately, are utterances - or their individual informational correlates, i. e. utterances again in certain understanding of the nature of utterances.

5. The present proposal for analysis of performative utterances

5.1. Performative meaning, performative inference and evaluation parameters

In order to clarify the relations between semantics and pragmatics in performative utterances, as well as between assertions and acts in them, I propose to distinguish the following aspects of communicative situations involving performative sentences:

• - semantically encoded performative meaning;

• - modification by performative inference (pragmatically determined reading of an utterance in some cases, ascribing performative force to it);

• - evaluation parameters, such as satisfaction conditions, including conditions of truth, sincerity and felicity.

• The main advantages of adoption of attitudinal and modal objects ontology for analysis of performative utterances are the following:

• - it allows to systematically analyse and describe parameters of performative meaning and performative inference;

• - it allows to find an integrative view on the conditions of truth and felicity, regarding them to be aspects of satisfaction conditions.

Before proceeding to explanation, I would like to notice again that the word ‘performative’ will be used here in certain conventional way, to denote primarily such phenomena as commissives and declaratives, since in a sense Austin and Searle were right saying that performativity is a characteristic of any utterance.

5.2. Parameters of performative meaning

With attitudinal objects ontology, we can naturally characterise performative meaning, being an attribute of explicit performative sentences, as specification of types of attitudinal/modal objects. E. g., consider the sentence (5.2.1), which is a canonical example of explicit performative sentences with core grammatical features and ‘hereby’ (or, in the Searle’s terms of speech acts, an example of commissives):

• (5.2.1) Hereby, I promise to solve that problem.

What is the semantic structure of this sentence? It consists of the following parameters:

• (p1) - Producer of the attitudinal object. In (5.2.1), it’s a person denoted by ‘I’.

• (p2) - Type of the attitudinal object. In (5.2.1), it’s a promise as an illocutionary attitudinal object (a statement about an obligation of certain type).

A promise as an illocutionary attitudinal object is an attitudinal object of such kind that it necessarily generates a promise as a modal object. The parameters p3 and p4, described below, are related to the satisfaction conditions of the promise as a modal object, however, they are introduced by the promise as an illocutionary attitudinal object.

• (p3) - Denoting of the sentence encoding the satisfaction conditions. In (5.2.1), it’s (5.2.1), which is expressed by the word ‘hereby’ used in this sentence.

• (p4) - Specification of the satisfaction conditions . In (5.2.1), it’s the specification of a modal object generated by that attitudinal object. This modal object may be characterised as an obligation of certain kind, for which the following satisfaction conditions are introduced: ‘to solve that problem’.

The values of all these parameters can be varied (in some cases, with an influence to some other parameters), resulting in various sentences.

Variations of (p1):

• (5.2.2) You promise to solve that problem.

• (5.2.3) Julia promises to solve that problem.

Variations of (p2):

• (5.2.4) I swear to solve that problem.

• (5.2.5) I ask to solve that problem.

Variations of (p3):

• (5.2.6) I promise that.

• (5.2.7) I promise what is mentioned above.

Variations of (p4):

• (5.2.8) I promise to jump to the Moon.

• (5.2.9) I promise to make it clear.

The word ‘promise’, by its lexical meaning, denotes an attitudinal object of a commissive kind, no matter (1) by whom it’s produced, (2) how it was produced and (3) what is the content of the promise.

Moreover, it’s also possible to state a promise in a conditional way:

• (5.2.10) if you manage to catch that bottle in the stormy sea, I’ll promise to do what’s required in the letter in it.

One may put forward here an objection that (5.2.10) is equal to the direct promise in (5.2.11).

• (5.2.11) I promise to do what’s required in the letter in that bottle, if you manage to catch it in the stormy sea.

However, there is no such equality, see Section 5.3.

5.3. Modification by performative inference

To figure out the role of possible performative inference, one should consider modifications of meaning from its initial invariant (locution) to various illocutionary options. E. g., such sentence as ‘I promise to solve that problem’ may be used seriously and literally, or as a joke, or in a theatrical play, or as a linguistic example, etc.

To make things simple, I propose to view such intentions of speakers (and probable understandings of recipients) just as translations derived from an initial locution, with some pragmatic mechanisms, resulted in a meaning assumed in a given context. E. g., if the sentence as ‘I promise to solve that problem’ is used as a joke, it may be seen just as translated to ‘I’m joking saying that I promise to solve that problem’.

Various kinds of pragmatic effects (implicatures, subauditions, modes of ‘unserious’ presentation like fairytales and theatre, etc.) stay at a different level than the assertive status of an initial utterance and its evaluation as true or false. The latter is given anyway, while the matter of pragmatics is answering the question why a speaker says a certain truth or a certain lie in such and such circumstances.

A topic important here is that sometimes a locution is assumed to be a performative utterances only by pragmatic inference, being not that by its locutionary meaning. Namely, it’s what happens when someone regards an utterance like ‘I’ll solve that problem’ (i. e., of the kind often qualified in the literature as so-called ‘implicit performative utterances’) to mean ‘I promise to solve that problem’.

My thesis is that there’s no implicit performative utterances, since proper performative utterances contain an expression denoting an attitudinal object of performative kind. One may promise, but one may swear, which is a different attitude (as well as a different ‘step in a social game’), and nor of these is just to state that doing something will be the case.

However, though ‘implicit performative utterances’ don’t bear a performative meaning at the level of locution, they may be assumed as such by a performative inference of a speaker or a recipient.

5.4. Evaluation parameters: truth, sincerity, felicity

As I stated above, adoption of attitudinal and modal objects ontology is, hypothetically, a way to integrative account of truth and felicity conditions for performative utterances. In my view, a more comprehensive account should address yet another parameter, which is sincerity. To outline my proposal about this, I’ll continue to discuss commissive utterances of a promise.

The analysis of evaluation parameters and their values for promises will be supported much by distinguishing of the meanings of the word ‘promise’. We need to distinguish ‘promise’:

• (1) as an illocutionary attitudinal object;

• (2) as a mental attitudinal object;

• (3) as a modal object.

The subtle difference between sentences denoting promises in the first person and in other persons is connected to the polysemy of (1) and (2).

When one talks about someone else’s promise, it’s typically a reference to an illocutionary object. Nevertheless, it’s possible to address for someone else other aspects of this polysemy, like I do here.

When one claims in the first case to promise something, this person at the same time:

• (1) produces a promise as an illocutionary attitudinal object, which is inevitable given the conditions for felicity is fulfilled (i. e., roughly, that utterance is taken to be serious and pertinent by the participants of a communicative situation);

• (2) refers to a promise as a mental attitudinal object;

• (3) becomes subjected to certain obligation - a promise as a modal object - depending on the communicative (and broader social) context of that promise as an illocutionary attitudinal object.

A performative utterance of a promise is true if and only if there is a corresponding mental attitudinal object in mind of the speaker (i. e., if and only if it’s sincere).

However, insincerity of a promise (and, thus, falsehood of the corresponding utterance) doesn’t dismiss the promise as an illocutionary attitudinal object, neither it neglects the promise as a modal object generated by that illocutionary object. This may seem to be a surprising conclusion, but actually it corresponds to rather usual situations in life. E. g., a boy named Hippolytus may promise to his mom to return home by supper, knowing in advance that he will not do so. Thus, his promise is insincere, and the utterance about it is untrue. Nevertheless, by producing this utterance he inevitably takes the mentioned obligation. For Hippolytus, it may be just a question of cost, if there’s no other way to stay outside the home as long as he wants and he internally agrees to be subjected to the penalty for breaking the obligation.

This example shows that ontology of attitudinal and modal objects helps to clarify the structure of communicative situations involving performative utterances and conditions of truth, sincerity and felicity of them.

6. Conclusion

The present contribution shows that ontology of attitudinal and modal objects is a powerful framework allowing to understand better some features of natural languages and communication. On one hand, it provides additional distinctions we need for analysis. On the other hand, it makes some things simpler. The application of ontology of attitudinal and modal objects to performative utterances substantiates these theses. In the present paper, I put forward basic ideas for such application and consider in more details commissive utterances. The future work may consist in further elaboration of this account and applying it to other types of performative utterances.

Research Data Availability:

All research data are available within the main text of the article.

References

• Austin, J. (1962). How to do things with words. Oxford: Clarendon Press. vii, 167 p.
• Austin, J. (1979). Philosophical papers. Eds.: J. Urmson, G. Warnock. Oxford: Clarendon Press . 242 p.
• Bach, K. (1975). Performatives are statements too. In: Philosophical Studies, 28, p. 229-236.
• Bach, K. & R. Harnish. (1979). Linguistic communication and speech acts. Cambridge (Mass.): The MIT Press. 352 p.
• Bach, K. & R. Harnish. (1992). How performatives really work: A reply to Searle. In: Linguistics and Philosophy, 15, p. 93-110.
• Corredor, C. (2009) The reflexivity of explicit performatives. In: Theoria: An International Journal for Theory, History and Foundations of Science, Segunda Epoca, Vol. 24, No. 3 (66), p. 283-299.
• Cresswell, M. (1985). Structured meanings: The semantics of propositional attitudes. Cambridge (Mass.): The MIT Press . xii, 202 p.
• Davidson, D. (1977). Reality without reference. In: Dialectica, 31 (3-4), p. 247-258.
• Davidson, D. (1979). Moods and preformances. In: Meaning and Use. Ed.: A. Margalit. Dordrecht: Reidel. P. 9-20.
• Eckardt, R. (2012). Hereby explained: an event-based account of performative utterances. In: Linguistics and Philosophy , Vol. 35, No. 1, p. 21-55.
• Frege, G. (1919a). Der Gedanke. Eine logische Untersuchung. In: Beitrage zur Philosophie des deutschen Idealismus, Bd. 1, 1918-1919, s. 58-77.
• Frege, G. (1919b). Die Verneinung. Eine logische Untersuchung. In: Beitrage zur Philosophie des deutschen Idealismus , Bd. 1, 1918-1919, s. 143-157.
• Ginet, C. (1979). Performativity. In: Linguistics and Philosophy , Vol. 3, No. 2, p. 245-265.
• Hanks, P. (2015). Propositional content. Oxford: Oxford Univ. Press. 227 p.
• Jary, M. (2007). Are explicit performatives assertions? In: Linguistics and Philosophy , 30, p. 207-234.
• Keller, L. (2013). The metaphysics of propositional constituency. In: Canadian Journal of Philosophy, Vol. 43, No. 5/6, p. 655-678.
• King, J. (2007). The nature and structure of content. Oxford: Oxford Univ. Press . x, 230 p.
• Lewis, D. (1970). General semantics. In: Synthese, Vol. 22, No. 1/2, p. 18-67.
• Marsili, N. (2023). Lies, common ground and performative utterances. In: Erkenntnis, Vol. 88, p. 567-578.
• Moltmann, F. (2013). Propositions, attitudinal objects, and the distinction between actions and products. In: Canadian Journal of Philosophy , Vol. 43, No. 5/6, p. 679-701.
• Moltmann, F. (2020). Truthmaker semantics for natural language: Attitude verbs, modals, and intensional transitive verbs. In: Theoretical Linguistics, 46(3-4), p. 159-200.
• Moltmann, F. (2023). Attitudinal objects and propositions. In: Routledge Handbook of Propositions. Eds.: A. Murray, C. Tillman.
• Moltmann, F. (2024). Objects and Attitudes. Oxford: Oxford Univ. Press . xii, 218 p.
• Putnam, H. (1981). Reason, truth and history. Cambridge Univ. Press. xii, 222 p.
• Recanati, F. (1987). Meaning and force: The pragmatics of performative utterances. Cambridge (Mass.): Cambridge Univ. Press. x, 278 p.
• Reck, E. H. (2005). Frege on numbers: beyond the Platonist picture. In: The Harvard Review of Philosophy, Vol. 13, Iss. 2, p. 25-40.
• Ricketts, Th. G. (1986). Objectivity and objecthood: Frege’s metaphysics of judgement. In: Frege Synthesized. Eds.: L. Haaparanta, I. Hintikka. Dordrecht: Reidel . P. 65-95.
• Salmon, N. (1986). Frege's puzzle. Cambridge (Mass.): The MIT Press . xi, 194 p.
• Searle, J. (1969). Speech acts: An essay in the philosophy of language. Cambridge: Cambridge Univ. Press. x, 203 p.
• Searle, J. (1989). How performatives work. In: Linguistics and Philosophy , Vol. 12, No. 5, p. 535-558.
• Soames, S. (2010). What is meaning? Princeton & Oxford: Princeton Univ. Press. x, 132 p.
• Stalnaker, R. (2019). Knowledge and conditionals: Essays on the structure of inquiry. Oxford: Oxford Univ. Press . ix, 252 p.
• Stojanović-Prelević, I. (2019). Are all performatives statements? In: Theoria, Vol. 62, Iss. 4, p. 17-26.
• Strawson, P. (1959). Individuals: An essay in descriptive metaphysics. London: Routledge. 260 p.
• Twardowski, K. (1911). O czynnościach i wytworach: kilka uwag z pogranicza psychologii, gramatyki i logiki. Kraków. 33 s.
• Quine, V. W. (1960). Word and object. Cambridge (Mass.): The MIT Press . xv, 294 p.

• 1
  A reviewer of the present paper stated that the notion of performative utterances is problematic, since Austin and Searle rejected the distinction of performatives and constatives, claiming that every utterance is in a sense performative. However, the term ‘performative utterance’ remains widely used in the literature; just to provide a couple of recent examples: Eckardt (2012), Marsili (2023). In the present paper, it’s used to cover a range of linguistic phenomena, primarily (in other terms) commissives and declaratives.
• 2
  See Quine (1960).
• 3
  See Davidson (1977).
• 4
  Who claimed that the Quine’s account may seem to be too weak, leaving the impression that all possible reinterpretations of a term’s reference are at least interconnected, while actually one shouldn’t even exclude the possibility that the word ‘cat ‘refers to cherries (Putnam 1981, p. 35).
• 5
  E. g., see Salmon (1986) and Soames (2010).
• 6
  E. g., see Cresswell (1985).
• 7
  E. g., Stalnaker (2019) says that ‘propositions (and possible worlds, which are maximal propositions) are something like properties that reality as a whole might have’ (p. 119).
• 8
  E. g., see Keller (2013).
• 9
  See Frege (1919a).
• 10
  See Frege (1919b).
• 11
  Frege (1919a), as translated in: Frege, G. The Thought: A Logical Inquiry. In: Mind, New Series, 1956, Vol. 65, No. 259, p. 301.
• 12
  I use here the words ‘act’ and ‘action’ as interchangeable.
• 13
  There is a number of historical predecessors of some aspects of this conception, especially Twardowsky (1911).
• 14
  See Bach (1975).
• Funding:
  The article was prepared within the framework of the Basic Research Program at HSE University.
• Article info
  CDD: 401

Edited by

• Editor-in-Chief
  MARCO RUFFINO
  ORCID: https://orcid.org/0000-0001-5619-0016

Publication Dates

• Publication in this collection
  09 Jan 2026
• Date of issue
  2025

History

• Received
  15 Aug 2024
• Reviewed
  26 Aug 2025
• Accepted
  24 Nov 2025