Truthful Insincerity


1. Introduction

In this essay I want to defend the claim that at least some insincere speech acts do not involve lying. This claim is based on an intuition concerning the use of the pronoun I and the nature of insincerity. The goal of the essay is therefore to test this intuition by constructing and analysing a thought experiment. For the definitions of the terms lying and insincerity, I rely on a basic introduction to the philosophy of lying by Italian philosopher Franca D’Agostini (2012). To be clear, my aim is not to argue against D’Agostini. Rather, I hope that her relatively uncontroversial take on these notions can serve as a jumping-off point for the discussion below.

Thus, for the purpose of this essay, lying is understood as a speech act whose goal is to make someone accept as true what the speaker or writer believes to be false (D’Agostini, 2012, p. 54). In other words, I am lying if I believe that non-p but say that p because I want you to believe that p. “Say that p” is understood to include both direct and indirect speech acts, where p is communicated by conversational implicature.

As for insincerity, I will provisionally identify it with what D’Agostini calls lying de se, i.e. lying about one’s mental states (2012, p. 38). In other words, I am being insincere if I lie about what I believe, feel, or intend to do. This is in contrast to lying de re, i.e. lying about things, events or properties of the world that are not the speaker’s mental states.

As D’Agostini’s presents things, lying de se has the same relation to truth as lying de re: in both cases, the liar says that p while believing that p is not the case. The only difference between the two types of lying, according to D’Agostini, is that the veridical content in lying de se is less accessible to the liar’s audience; it “leaves no traces” (p. 86). I will argue, however, that lying de se, henceforth insincerity, has a more complex relationship to truth and, as a result, sometimes does not involve any lying at all. Consequently, by the end of the essay I hope to offer a revised definition of insincerity that does not treat insincerity as a kind of lying.

2. Presentation of the thought experiment

Suppose that Kajsa, a Swedish microbiologist, has recently started dating Dasha, a cello player from St. Petersburg who moved to Sweden because of the growing homophobia in her home country. As a professional musician, Dasha has sophisticated music tastes ranging all the way to Schönberg and Penderecki, but her favourite composer is Shostakovich. Dasha admires the rhythmic and melodic brilliance of the Soviet composer’s work. Moreover, she feels that his music – in particular his First Violin Concerto, full of anxiety and premonition, – reflects the story of her own Leningrad intelligentsia family, who barely survived the Nazi blockade only to be targeted in the post-war wave of purges.

Kajsa, by contrast, is no fan of Shostakovich, or indeed of anything that can be labelled “classical music”. As a teenager, she listened to her fair share of soulful Swedish folk singers and Marxist punk; now, as an adult and a busy researcher, she mostly listens to chill-out playlists on Spotify. When Dasha, her eyes ablaze with excitement, invites her to a performance of the First Violin Concerto at the Stockholm Philharmonic, starring a celebrated German violinist, Kajsa accepts the invitation but feels no enthusiasm whatsoever. Just as she fears, she hates the concert; the music seems at times deafening, at times mind-numbingly boring, and always disjointed and utterly depressing. However, she joins in the thunderous applause at the end. When Dasha asks her if she liked the music, Kajsa musters her substantial middle-class emotion-feigning skills and makes the following assertion:

(1) I loved it!

Given the preliminary definition of insincerity above, Kajsa’s assertion is clearly insincere. Her description of her mental state during the concert does not match the mental state she was actually in while the music was still playing. As a case of insincerity, (1) can be contrasted with the following blunt assertion, which Kajsa could have made if Dasha and she were at a later stage in their relationship:

(2) I didn’t like it.

(1) can also be contrasted with a partly insincere assertion to the same effect as (2) but involving strategies aimed at saving Dasha’s positive face (as defined by e.g. Pfister, 2010) and thus maintaining the relationship between the speakers:

(3) I’m afraid classical music is still not my thing. I’m happy I gave it another try, though!

Finally, (1) can be contrasted with a face-saving assertion that gives Kajsa even more wiggle room thanks to its ambiguity:

(4) It was really interesting.

It is not immediately clear what kind of mental state (4) is describing. Depending on the prosodic, facial and gestural cues, it could be taken to mean that Kajsa found the experience genuinely interesting, i.e. intellectually engaging, while it lasted. However, (4) could also be understood as an act of acknowledging that Shostakovich’s Violin Concerto No. 1 is interesting in some objective, or at any rate generally accepted, sense which has little to do with Kajsa’s experience of the music. Given the sociocultural background of the speakers, this latter interpretation is then likely to give rise to the conversational implicature that Kajsa did not enjoy the music. If this implicature was intentional, (4) would also qualify as a largely sincere speech act on a par with (3).

On the definition of lying as a speech act aiming to make someone accept as true what the speaker believes to be false, the four assertions would seem to fare as follows. (4) is not a lie, provided that Kajsa intended to create the implicature that she did not enjoy the concert. (3) is not a lie or only partly a lie, depending on whether Kajsa was in fact happy to give classical music another go. (2) is not a lie. (1) is a lie.

3. Discussion of the thought experiment

As indicated earlier, I want to argue that, to a significant extent, this analysis of the thought experiment presented above is inadequate. In particular, “I loved it!” is not necessarily a lie. By the same token, “I’m happy I gave it another try” does not have to be a lie even if Kajsa has not felt a single positive emotion about giving classical music another try. My argument for this pivots on the reference of the pronoun I and the corresponding variation in the content of assertions that contain this pronoun.

It is well beyond the scope of this essay to defend or even present a particular theory of how exactly linguistic signs refer to things in the real world. For the purpose of my argument, I will simply assume that speakers can, as a matter of fact, use linguistic signs to refer to things in the real world. I will further assume that, with the exception of reported speech and certain idiomatic expressions, the pronoun I is normally used to self-refer to the entity using I. In other words, when I say that an insincere assertion, such as Kajsa’s “I loved it!”, may not be a lie depending on the reference of I, my claim is that the pronoun I can be used on different occasions by the same human being to self-refer to different entities.[1]

Consider why Kajsa may choose to reply “I loved it!” to Dasha’s question even though she intensely disliked the music while it lasted. One likely reason is her desire to maintain her relationship with Dasha. Because Kajsa knows how important Shostakovich’s music is to Dasha, she may be afraid of losing Dasha’s respect, or she may want to please her new partner. Alternatively, she may simply want Dasha not to feel bad about having dragged her to the Philharmonic and subjected her to something she does not like.

Considered in isolation, none of these reasons suggest that Kajsa’s insincerity does not involve lying. Suppose, however, that Kajsa’s insincerity is part of a long-term strategy. Suppose that she wants to love Shostakovich.[2] She is willing to do whatever it takes to learn to appreciate his works, both because she cares about Dasha and because she has come to believe that developing one’s taste in music is a moral imperative. After the concert, Kajsa starts listening to Shostakovich and other mid-twentieth century composers while jogging or doing household chores. She orders and reads two biographies of Shostakovich, a hefty volume recounting the dramatic story of his Seventh Symphony, and The Noise of Time by Julian Barnes. She completes an online course in classical music appreciation as well as a course in Soviet history. All these efforts eventually bear fruit. Kajsa does begin to enjoy much of Shostakovich’s work, including the magnificent scherzo of the First Violin Concerto.

Let us now revisit Kajsa’s assertion at the Philharmonic. When she says (1) “I loved it!”, it still appears that she is being insincere. As far as Kajsa knows, no mental state that could be accurately described as “loving it” obtained in the human being named Kajsa at any point during the concert. Therefore, her speech act fits our provisional definition of insincerity as lying about one’s mental states.

However, what I want to call into question is the assumption that Kajsa is talking about mental states that actually obtained during the concert. It can be argued that the reference of I in (1) is not Kajsa the human being, or whatever state Kajsa the human being was in during the concert. Instead, the I in (1) may well be referring to an entity that Carol Rovane describes as “a set of rationally related intentional episodes” striving to achieve “overall rational unity over time” (Rovane, 1998, p. 239). Rovane calls such entities “persons”. To avoid using this loaded term, I will simply refer to the relevant set of rationally related intentional episodes within Kajsa as Kajsa (1). It is worth emphasizing here that Kajsa (1) is not identical with Kajsa the human being (henceforth Kajsa HB). Rather, Kajsa HB is the site of sets of rationally related intentional episodes such as Kajsa (1) (Rovane, 1998, p. 185).

A key element of Kajsa (1) is the implicit desire to have Kajsa HB act in a way that matches both the intentional episodes constituting Kajsa (1) and their rational unity. For example, Kajsa (1) includes a strong conviction that racism is morally reprehensible. Whenever Kajsa HB has a pre-reflective racist thought, Kajsa (1) strives to make sure that this thought is not translated into observable action with potential consequences for other human beings. Another intentional episode that Kajsa (1) includes is, as we have seen, the determination to appreciate Shostakovich. This determination is rationally related to Kajsa (1)’s desire to be not just a partner but also a true soulmate to Dasha. Thus, Kajsa HB’s dislike of the First Violin Concerto does not match any of the intentional episodes making up Kajsa (1), and it would gravely threaten their overall rational unity if adopted as part of the set.

If we accept this account of the reference of I in (1) “I loved it!”, does (1) still fit the definition of lying offered in the introduction? It is far from certain that it does. Recall that, for (1) to fit the definition, Kajsa (1) has to be asserting something she believes to be false in order to lead Dasha to accept it as true. However, Kajsa (1) is not asserting anything she believes to be false. For Kajsa (1), “I loved it!” is not a description of what happened to Kajsa HB during the concert. Instead, it is a description of what Kajsa (1) contains. In other words, what it asserts is that Kajsa (1) contains the intention to have loved the concert, rationally related to Kajsa (1)’s overall commitment to appreciating Shostakovich.

One immediate objection to this conclusion is that it seems to threaten the very concept of lying. If (1) is not a lie because Kajsa (1) intends to have loved the concert, what is to stop any given lie from collapsing into an “intention” for things to have turned out the way they did not? If we were to accept this point of view, would we not end up with a thoroughly Trumpian world, where nothing is a lie as long as it fits someone’s agenda and everything is a lie when it does not?

However, this objection misses an important epistemic distinction between publicly observable events and mental states. As D’Agostini notes, mental states often “leave no traces”. As long as the speaker is adept at inhibiting behavioural manifestations of her mental states, they are sealed off epistemically from the rest of the world. If Kajsa HB yawns and fidgets her way through the concert, we will be rightfully sceptical of Kajsa (1)’s claim that she loved the experience. Suppose, however, that Kajsa (1) has Kajsa HB masterfully feign attention and interest throughout the concert. When she then claims “I loved it!”, we have to look elsewhere for evidence that this assertion is a lie. Admittedly, we might be able to find such evidence later, provided that Kajsa (1) fails to achieve her goal of turning Kajsa HB into a Shostakovich fan, in which case behavioural slips are quite likely. If she succeeds though, no such evidence may ever be available.

Moreover, even if Kajsa confesses to Dasha many years later that she hated that first Shostakovich concert, (1) might still not retroactively qualify as a lie. Kajsa (1) – who, as we remember, eventually succeeded in training Kajsa HB to enjoy Shostakovich – could still be justified in claiming that the entity that hated the concert was never her, while the speaker of “I loved it!” was.

There is another immediate objection to the view of non-lying insincerity defended above. One may point out that, even if Kajsa (1) rejects Kajsa HB’s dislike of the concert as a part of her, and even if no evidence of this dislike is ever available to anyone else, (1) “I loved it!” is still a lie because Kajsa (1) is lying to herself about who she really is. This objection, however, presupposes that there is such a thing as Kajsa’s real self, waiting to be discovered. To put this in terms of Rovane’s sets, the assumption is that there is one correct set of rationally related intentional episodes corresponding to “real Kajsa”. This view sounds intuitively appealing but, as I have argued elsewhere (Andreev, 2017), it is hard to defend. Until it is shown that “real selves” exist, we have to assume that “real” Kajsa (1), understood as a set of intentional episodes, is whatever Kajsa (1) happens to contain. If Kajsa (1) is fully aware that Kajsa HB hated the concert but she rejects that reaction, it is not clear in what sense Kajsa (1) is lying to herself.

4. Concluding remarks

This essay has argued that at least some insincere speech acts do not involve lying. If this is correct, then the provisional definition of insincerity offered in the introduction needs to be revised because it identifies insincerity with a kind of lying, namely lying about the speaker’s mental states. It is now time to suggest a new definition of an insincere speech act, based on the discussion above: an insincere speech act communicates false information about mental states of the human being performing the speech act.[3] This definition assumes that, in an insincere first-person assertion, the reference of I is always the human being making the assertion. In other words, it assumes that the speaker is fully identical with the human being that is speaking.

It is this fixed reference of I that makes insincerity fundamentally distinct from lying. As I have tried to show, an insincere speech act is not necessarily a lie because the definition of lying cannot assume that the reference of I is the human being performing the speech act. At least in some cases,[4] the reference of I in first-person assertions is best understood as a set of rationally related intentional episodes occupying the human being rather than the human being as such. I have argued that this feature of first-person reference is crucial in assessing whether an insincere speech act is a lie.

I would like to conclude this essay with a brief note on ethical implications of this account of insincerity. In many cultures the world over, insincerity is strongly associated with lying, and for this reason it tends to be judged just as harshly or nearly as harshly. However, the possibility of truthful insincerity suggests that the moral status of insincerity is strictly neutral rather than strictly negative or even largely negative. In fact, it can be argued that certain acts of insincerity, such as Kajsa’s suppression of her knee-jerk racist reactions (or indeed her fake-it-till-you-make-it appreciation of Shostakovich out of love for her partner) are crucial for moral behaviour. This is a potentially fascinating line of inquiry that I would like to return to in the future.


[1] By the same token, the pronoun you may be understood to refer to different entities by the same human listener.

[2] This part of the thought experiment is inspired by Emanuele Arielli’s book on the philosophy and practice of changing one’s tastes (Arielli, 2016). Arielli describes his efforts to appreciate atonal music.

[3] This definition can be modified to accommodate nonhuman beings capable of having mental states and performing speech acts.

[4] My intuition here is to say “in most cases”.


Andreev, K. (2017). Constraining agency in the name of agency: How can we know what not to want? (Unpublished C-level essay). Uppsala university, Uppsala, Sweden.

Arielli, E. (2016). Farsi piacere. La costruzione del gusto. Milan: Cortina Raffaello.

D’Agostini, F. (2012). Menzogna. Turin: Bollati Boringhieri.

Pfister, J. (2010). Is there a need for a maxim of politeness? Journal of Pragmatics, 42(5), 1266-1282.

Rovane, C. (1998). The bounds of agency: An essay in revisionary metaphysics. Princeton: Princeton University Press.