Skip to main content
Log in

Lies, Common Ground and Performative Utterances

  • Original Research
  • Published:
Erkenntnis Aims and scope Submit manuscript


In a recent book (Lying and insincerity, Oxford University Press, 2018), Andreas Stokke argues that one lies iff one says something one believes to be false, thereby proposing that it becomes common ground. This paper shows that Stokke’s proposal is unable to draw the right distinctions about insincere performative utterances. The objection also has repercussions on theories of assertion, because it poses a novel challenge to any attempt to define assertion as a proposal to update the common ground.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Similar content being viewed by others


  1. In a footnote, Stokke specifies that he only wants to capture “the aspects of assertion that are relevant for defining lying”, and explicitly denies that his account of assertion is a definition of assertion (Stokke 2018: 47). However, since DLS is meant to spell out what an ‘insincere assertion’ is, Stokke is de facto committed to define assertion as the joint satisfaction of condition (a) and (b) of DLS (cf. also Stokke 2018: 90).

  2. Explicit performative utterances are utterances of the form “I (hereby) [performative verb] that p”, in which the speaker performs a given speech act (promising, asserting, betting, etc.) by declaring that she is performing it. In the examples, the starred propositions represent the content of the speech act (what the speaker is promising, asserting, commanding), identified in a ‘non-descriptivist’ way. I will clarify what a non-descriptivist interpretation is in §2.A.

  3. See Marsili (2016) for empirical evidence that laypeople overwhelmingly judge insincere promises to be straightforward cases of lying. Only Austin (1962: 11) and Meibauer (2014) have defended the opposite view that promises cannot be lies. However, Austin presents this view only to proceed to challenge it (1962: 50, 70–1, 135–6), as part of a reductio ad absurdum of the distinction between constative and performative utterances (and the descriptivist semantics associated with it). Meibauer (2014: 76) merely states, in passing, that “we do not want to speak of an insincere promise or warning as a lie”, but provides no justification for this claim. Marsili (2016) has since challenged Meibauer’s view on theoretical (2016: §2) and empirical (2016: §6) grounds. In addition to this, Meibauer recognises that “it is possible to lie by using explicit performatives” (2014: 94), so that he would still agree that (4) and (5) can be lies.

  4. To be sure, Stokke is in good company: as I argue in Marsili (2020), this problem is shared by some other influential definitions, namely deceptionist ones (e.g. Mahon 2008; Lackey 2013), and some other assertion-based views, such as Fallis's (2012, 2013).

  5. This is not to deny that there may be some truth to the descriptivist view, considered as a semantic theory of performative utterances: clearly, there is some technical (and perhaps philosophically interesting) sense in which I would be telling the truth in uttering (10) (since I would in fact be guaranteeing that I have already booked a restaurant) and not in uttering (8). My point here is simply that, unless we are aiming for a revisionist account of lying, we should classify (8–12) as any competent speaker would—that is, as lies, and not as truthful statements. This point can be acknowledged independently of whether one is convinced by a descriptivist analysis of these utterances (e.g. Bach and Harnish 1979, 1992) or by the arguments against it (Harris 1978; Searle 1989; Reimer 1995; Jary 2007).

  6. Authors who reject the force-content distinction propose comparable analyses of this sort of cases (e.g. Hanks 2007: 150, Barker 2004: §1.5.1). For a more technical discussion of non-finite to-clauses, see the analysis of PRO elements in Binding Theory (Chomsky 1981).

  7. For the record, this analysis is at odds with Stokke’s own account of how we assign truth-conditional content to sentences (Schoubye and Stokke 2016, rehearsed in 2018:§4). When it comes to determining which sentences have semantic content at all, Stokke takes mood, rather than force, to be decisive. He holds that “all declarative sentences […] are associated with a minimal content” (Schoubye and Stokke 2016: 773), where ‘minimal content’ is understood to be truth-evaluable. Since (7) is in the declarative mood, Stokke’s own criterion would assign it truth-evaluable content. That noted, this solution still is worth exploring, to see if DLS (so understood) can accommodate our desiderata.

  8. While English only seems to accept infinitive to-clauses for claims about the future, this restriction is not universal. It does not hold in Romance languages, such as French or Italian, where claims about the present and the past can be made by means of performatives that accept both finite and non-finite clauses:

    • Affermo [di avere diciotto anni/che ho diciotto anni]

    I affirm that I am 18 years old

    • Je jure [d’avoir dit la vérité/que j’ait dit la vérité]

    I swear that I have told the truth.

  9. We have already seen that a syntactic criterion based on the presence of that-clauses would not succeed. A criterion based on direction of fit would fail for similar reasons: since promises, pledges, advices and orders have the same direction of fit, such a criterion would either incorrectly exclude them all, or incorrectly include them all. A somehow related alternative, delineated by Portner (2004), would be to differentiate between illocutionary acts that update the common ground and those that update to-do-lists. But also this criterion will not do: it will incorrectly rule out promises like (6), which update to-do-lists, as well as assertions and guarantees about one’s future actions (“You can chop the onions, and I will peel the tomatoes”; “I guarantee that I will lower the taxes during my term”).

  10. Of course then can be contexts in which Adolf’s command (7) is not successful (e.g. because Adolf is a patient in a mental hospital, with no authority of Beatrice). In these cases, it will be infelicitous to presuppose (7*) by (14). But this does not prove that DLS passes the test. For DLS to pass the test there must be no circumstances in which (7*) can be felicitously presupposed as a result of Adolf’s command. For this reason, discussion in the main text will be restricted to felicitous (successful) commands, which pose a more direct challenge to DLS.

  11. Could we adopt instead a 'strong' notion of proposals, that captures only common ground updates that do not need approval from the audience? This revision would  exclude suggestions, which is a positive result, since suggestions cannot be lies. And it would preserve a distinction between proposals that need approval from the audience and proposals that need no such approval. However, this solution is not preferable, for several reasons. First, it would not address the main problem faced by DLS, since DLS would still incorrectly classify insincere orders as lies. Second, we would be considering a revision of the notion of 'proposal' that is inconsistent with most of what Stalnaker and Stokke say about assertion, so that we could not possibly regard it as an amendment of their view. Finally, this solution feels ad hoc: one thing is to adopt a technical understanding of ‘proposals’, broader than its ordinary meaning (as Stokke does); another is to carve this notion around the counterexamples, in a way that explicitly contradicts both its ordinary and its technical meaning.

  12. The negative case against Stokke’s proposal has grown in size since I first redacted this manuscript. For equally critical stances on Stokke’s characterisation of lying and assertion, see Van Riel (2019) and Van Elswyk (2019). Further objections are in Fallis (2013, 350–2) and Keiser (2016: 476), but see Stokke (2017a, b) for replies, and Marsili (2017, 69–71) for a vindication of Fallis's objection.


Download references


I would like to thank Jennifer Saul, Grzegorz Gaszczyk, and two anonymous reviewers for very helpful comments on earlier versions of this article.


Research funding was provided by Ministerio de Ciencia, Innovación y Universidades (Grant No. FFI2016-80636-P, AEI/FEDER, UE).

Author information

Authors and Affiliations


Corresponding author

Correspondence to Neri Marsili.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Marsili, N. Lies, Common Ground and Performative Utterances. Erkenn 88, 567–578 (2023).

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: