Consider this case (derived from Krstić 2020, 758–759):

Pinocchio intends to cause the police to realize (i) that Tony is in Pinocchio’s house, (ii) that Pinocchio is lying by saying that Tony is not in Pinocchio’s house, and (iii) that he is lying because he intends to cause them to deductively infer the relevant true proposition from his assertion and the behaviour of his nose. Since he intends to cause them to learn the whole truth—i.e., where Tony is, that he (Pinocchio) is lying, and why—Pinocchio does not count as intending to deceive the addressee and, because the nose is an indicator of lying rather than uttering something he believes to be false, Pinocchio seems to be genuinely lying.

Pinartio appears to be an excellent counterexample to the view according to which intending to deceive is necessary for lying. Mahon (2015) names this view deceptionist. Some proponents of deceptionism are Davidson (1998), Williams (2002), Derrida (2002), Faulkner (2007, 2013), Lackey (2013, 2019), Keiser (2016), Meibauer (2014a, 2014b, 2016), Maitra (2018), and Harris (forthcoming). Pinartio is a counterexample to this view because Pinocchio does not intend to cause the police to believe what he asserts (or to make them more confident in this proposition) and he does not intend to cause them to believe that he believes what he asserts (or to make them more confident in this proposition), which are the standard ways of deceiving discussed in the literature on the nature of lies. In fact, Pinocchio does not intend to mislead the police with respect to anything or to conceal any information.¹ He is trying to help them.

Pinartio suggests that the non-deceptionist analysis, according to which asserting what you believe (or judge) is false is (necessary and) sufficient for lying, is correct.² Some proponents of the non-deceptionist view are Aquinas (Summa Theologica, II-II, Q.110, article 1), Johnson (1755a), Carson (2006, 2010), Sørensen (2007, 2010, 2022), Fallis (2009, 2012, 2013, 2015), Saul (2012), Stokke (2013, 2016, 2017, 2018), Rutschmann and Wiegmann (2017), Krstić (2018, 2019, 2020), Marsili (2021), Sneddon (2021), and Michaelson and Stokke (2021). And while Pinocchio’s lie is a so-called bald-faced lie, i.e., knowingly undisguised lie, it is different from all other cases of bald-faced lies in a very important way. Standardly, bald-faced lies involve situations in which the liar believes that it is common knowledge that what the liar says is false. The common knowledge bald-faced liar does not intend to deceive their addressee because they think that the addressee already knows the truth. Say that a gambler asserts to his wife that he was not gambling when she caught him with the betting tickets from that afternoon’s races (Arico and Fallis 2013); it seems sensible to think that he did not try to make his wife believe him, since he should think that she already knows everything.

However, the deceptionists promptly responded to the argument from the existence of these bald-faced lies by saying that the proposed examples (1) are either lies intended to deceive in some of the senses I mentioned above or (2) are not genuine lies in the sense in which they do not involve genuine assertions—and you need to assert in order to lie.³ These replies do make the common knowledge bald-faced lies much less effective. A desperate gambler may hope that his lie could cause his wife to become slightly less confident in her true belief and he could add that the ticket belongs to a friend. Alternatively, it might be that, since he does not intend his wife to believe him, the gambler does not assert what he says but rather merely makes it look like he does; he could be playing a kind of a (language) game, he could be doing something similar to acting or even to being verbally aggressive (e.g., Keiser 2016; Maitra 2018; Harris forthcoming; Meibauer 2014a; against, e.g., Marques 2020; Viebahn 2019; Marsili 2021).

These two general objections are well-known and, in one form or another, they have been put forward in detail by many contemporary philosophers (e.g., Faulkner 2007, 2013; Kenyon 2010; Lackey 2013, 2019; Meibauer 2014a, 2014b, 2016; Leland 2015; Dynel 2011, 2015; Hawley 2018; Keiser 2016; Maitra 2018; Harris forthcoming). I do not intend to discuss their application to common knowledge bald-faced lies for two reasons. The first is that some replies already exist (e.g., Fallis 2015; Stokke 2017, 2018; Marques 2020; Viebahn 2019; Marsili 2021). The second, and more important, is that I designed cases like Pinartio to avoid both deceptionist objections. Pinartio involves what I named a “tell-tale kind” bald-faced lie (Krstić 2020), in which the addressee not only does not know the truth (no common knowledge) but rather learns the truth by observing the behaviour of the lie-disclosing sign and, vitally, Pinocchio intends this to happen. He wants the police to deductively infer the truth from his false assertion and the behaviour of his nose. Therefore, the lie is not intended to deceive the addressee in any sense. Since it is clear that the speaker does not intend to deceive, Pinartio seems to be an importantly different counterexample to the deceptionist analysis of lying; it gives us a new perspective on the issue and thus it opens the door for a new and promising debate.

This paper is designed to fill this void and bring the debate regarding lies and intending to deceive closer to a fruitful end. In section 1, I argue that cases such as Pinartio are not uncontroversial but that more convincing cases can be developed from them. In section 2, I discuss two objections to my argument. In section 2.1, I reject the popular objection according to which bald-faced lies are not genuine lies on the count of them not involving genuine assertions. In section 2.2, I argue against the interpretation that tell-tale liars (indirectly) assert not the literal meaning of the descriptive sentence they utter but rather the proposition they want the hearer to infer from their behaviour. In section 3, I conclude my argument.

1 Finding the Right Case

My original cases do avoid some standard objections but I see now that they also generate new problems. In this section, I discuss these problems and offer a case that avoids them. In the next section, I discuss two further objections that arise.

In my analysis of Pinartio, I write:

The non-[deceptionist] definition counts this [Pinocchio saying “Tony definitely isn’t in my house”] as lying because Pinocchio asserts what he believes is false, I count this as lying because the nose grows, and the [deceptionist] definition does not count this as lying because Pinocchio does not intend to deceive his addressee notwithstanding the fact that the nose indicates that Pinocchio is lying—this is why (Pinartio) is a counterexample to the [deceptionist] definition. [Krstić (2019), 653; italics added]

In short, I argue that Pinartio is a counterexample to the deceptionist analysis because (i) Pinocchio does not intend to deceive his addresses and (ii) the nose indicates that he is lying. Pinartio cannot be dismissed on the count of it being a fairy-tale case. True, Pinocchio is a fictional character and the situation I put him in is uncommon. However, this is not relevant to the question under discussion. My main idea is to present a situation in which a speaker believes (correctly or incorrectly, it does not matter) that their lie will undoubtedly be disclosed to the addressee and they use this circumstance to communicate the truth. In real life, police may convince their suspect that their lie detector makes no mistakes and there could be a real-life person who mistakenly believes that they always blush when they lie, or that their interlocutor can read their thoughts, or is skilled enough in detecting lie-betraying cues. In Meet the Parents (Universal Pictures, 2000), Jack Byrnes convinces his future son-in-law, Greg Focker, that he will unmistakably detect when Greg lies just by feeling Greg’s pulse. I merely flesh this situation out using the character of Pinocchio as a communicative device.

Moreover, the issue of whether the nose will grow or not is irrelevant: the deceptionist definition assumes that the intention to deceive is necessary for lying. If the nose does not grow, Pinocchio will still lie; he will just fail to communicate the truth. The examples are effective as long as (i) the liar believes that something will show that he is lying and (ii) the liar and their audience are in a standard context (i.e., unless some further conditions apply) in which it is common ground that one asserts what one says. Thus, even if Pinocchio misleads the police by asserting what he says, this would be against his intention, which still sits uneasily with the deceptionist analysis. Consider the following case.

In Artocchio, Artie believes that there is a perfectly reliable giveaway of lying and he intends to use it to cause the police to infer the truth from his lie. Artie’s blushing is not a real tell-tale sign of lying but it worked as one nonetheless. Furthermore, Artie and the police are in a standard context, he thinks that he will be understood as asserting what he says and they expect him to assert what he says, and thus there seem to be no reasons to think that he did not assert what he said. Importantly, Artie’s blushing is not analogous to winking or finger-crossing: the function of winking is set by a convention that is a part of common ground (the context is not standard) and Artie’s blushing is not. Therefore, we should think that Artie is understood as asserting what he says and his plan is not unwise: if the police are sufficiently attentive to detail, they will realise that he is lying and why he is lying. Nevertheless, there is an important concern that Tony was intentionally caused to believe falsely that Artie deceived the police. Tony was misled, that is; just as Artie planned.

In Krstić (2019, 656), I argue that the issue of whether Artie intended to deceive Tony is irrelevant for the question of whether lies must be aimed at deceiving—since Artie was addressing the police, not Tony. However, it may be that Artie was addressing Tony after all: Artie needed Tony to hear what he is saying. Hence, Tony does appear to be Artie’s intended hearer and, it seems to follow, Artie did intend to deceive someone by lying—Tony.⁵ The conclusion that Artie intends to deceive Tony by lying, however, does not follow. Tony expects and coerces Artie to say to the police that Tony is not in Artie’s house and Artie says this because he knows that Tony expects him to do it. Therefore, Artie cannot reasonably intend to cause Tony to believe as true a proposition for which Artie knows that Tony knows is false. Thus, even if Artie addresses Tony, he cannot be lying to Tony in the sense in which this requires intending to deceive him. In fact, he does not seem to be lying to Tony at all: he just says what Tony wants him to say.⁶ Say that a company manager orders his assistant to present false data at the board meeting in exchange for a promotion. This assistant will lie to other board members but not to his boss. The same applies to Artie.

So, it seems as if Artie lies while not intending to deceive anyone by asserting what he says. However, whether Artocchio is enough to reject the deceptionist view depends on how broad our analysis of lying is. On the view that one may lie only by asserting something one believes is false, Artocchio vindicates the non-deceptionist definition. However, broader analyses exist (see footnote 3) and so one may argue that Artie’s lie is intended to deceive Tony—only in a sense that does not involve asserting. Artie may, for instance, intend to deceive Tony by falsely implicating that he led the police off Tony’s track. On this analysis, then, Artie both counts as asserting what he says and as intending to deceive by lying (i.e., by implicating false information to Tony). Therefore, while Artocchio does appear to put reasonable pressure on the deceptionist analysis, there still are some controversies about it. Most of these controversies can be eliminated with simple modifications. Consider the following case.

Witnessio and Artocchio involve a rather plausible idea that some people may think that something will unmistakably show that they are lying and they are similar to the context of two recent movies. The first is Knives Out (Lionsgate, 2019) in which the character of a nurse, Marta Cabrera, cannot lie without vomiting (“Just the thought of lying […] It makes me puke,” she says), the detectives that are questioning her know this, she knows that they know, and indeed she does vomit every time she lies. The second is Meet the Parents (Universal Pictures, 2000), in which Jack Byrnes convinces his future son-in-law, Greg Focker, that he will unmistakably detect when Greg lies just by feeling Greg’s pulse (the pulse is a lie-betraying sign). The main difference is that, on the one side, Marta avoids lying by giving true but incomplete answers to questions when asked and Greg simply goes for telling the truth, whereas, on the other side, Artie comes up with a plan to make lemonade when life gives him lemons: he decides to use what he thinks is his otherwise unfortunate reaction to his advantage and cause the police to learn the truth by lying to them. Another difference is that, if Artie blushes at the right moment, this will be not because blushing is somehow connected to whether he lies but rather because of the highly stressful situation he is in.

In Witnessio, Tony did not end up with a false belief; rather he was left without any belief regarding Artie’s conversation with the police. More importantly, not only is Artie not addressing Tony, he hopes that Tony never learns about the conversation. Therefore, neither did Artie intend to deceive Tony, nor was he lying to Tony. Even if one can lie by making false promises, implicatures, or presuppositions, Artie does not lie to Tony in any of those senses; he is not addressing Tony. While Artie does not lie intending to deceive the police, misleading them about Tony’s whereabouts or Artie’s intentions is possible in Witnessio (they do not believe that Artie blushes always and only when he lies) but this is irrelevant for our discussion. We are not analysing whether epistemic harm will be caused but rather whether the liar intends to cause it or whether he expects that he might cause it (see Krstić 2020, sec. 2.1) and Artie clearly lacks the intention to deceive them and he does not expect them to end up misled (he believes that they will understand why he blushes). Finally, because the belief that Artie blushes always and only when he lies is not common ground between Artie and the police, we cannot say that blushing implies that Artie does not assert what he says.

Witnessio avoids the most obvious objections, but it fails to avoid all objections. I anticipate two. The first is that, because he does not intend to give the police a reason to believe what he says, Artie does not assert what he says and thus does not lie. The second is that Artie (indirectly) asserts the proposition he intends the police to infer from his behaviour and the literal meaning of the uttered proposition (“Tony committed the crime”) and, because he believes this proposition to be true, he does not count as lying. I will resolve these objections in turn.

2 Objections

2.1 Not an Assertion

According to one influential analysis of assertion (Gricean in nature), I asserted that \(p\) by uttering \(x\) if and only if I uttered \(x\) intending to induce in you the belief that \(p\) or give you grounds for believing it by means of your recognition of the full content of my intention (e.g., Bach and Harnish 1979; Récanati 1987; similarly, Peirce 1935; Grice 1989).⁷ Accounts of this sort are typically called Gricean or Neo-Gricean, Keiser (2016) calls them epistemic, and Harris (forthcoming) and Siebel (2020) call this approach to communication intentionalism. I will refer to the view as Neo-Gricean analysis of assertion.

According to one influential argument based on the Neo-Gricean analysis of assertion, because he does not intend to give his audience grounds for believing what he says because he says so, Artie does not count as asserting what he says in Witnessio and therefore—on the popular assumption that lies are a subset of assertion—he does not lie (e.g., Chisholm and Feehan 1977; Meibauer 2014a, 2014b, 2016; Keiser 2016; Harris forthcoming, 13, 15). That is to say, I assume that one would object that, while there is nothing in the context that could prevent Artie’s utterance to count as an assertion, Artie did not act on an intention constitutive of asserting and this failure (rather than pragmatic considerations or linguistic conventions) disqualifies the utterance from counting as an assertion and a genuine lie (which requires asserting what you say).

Many influential philosophers think that this is a very serious problem for any non-deceptionist analysis of bald-faced lies. The idea is that, because Neo-Griceans think that the ordinary language concept of lying is too ambiguous, they are looking for a definition of lying that fits neatly within their definition of assertion, which in turn fits neatly within a relevant broader theory of speech acts and communication. Therefore, they will maintain that bald-faced lies are not genuine lies because they do not involve genuine assertions. While Artie believes that his blushing signals that he is lying when he is lying, his conception of lying is too broad and he does not intend to lie in the relevant sense (which involves asserting what you say).

I will reject this objection in three steps, where each step gives my argument a premise. In step 1, I argue (by analogy) that the argument from the objection generates an unfalsifiable position. In step 2, I argue that it entails that competent language users unreliably track assertions, which is a very bold and empirically unsupported claim. The first two steps only show that the objection is much less serious than what initially seems; they do not show that it fails. The third step, however, shows that the argument fails. In step 3, I argue that, if understood as not allowing exceptions (such as bald-faced lies), the Neo-Gricean analysis of assertion misclassifies some sincere assertions as not assertions: some sincere asserters do not act on the intention to give their hearers grounds for believing what they say. I conclude that, considering steps 1–3, this specific argument claiming that bald-faced lies are not genuine assertions fails. I now proceed to step 1.

In his Introductory Lectures on Psycho-analysis (1916–1917), Freud defends his dream theory from the concern that dreams do not reveal our unconscious mental life. His defence is strikingly similar to the comeback presented above: he dismisses counterexamples by reinterpreting them. The following is one such case.

A woman dreamer says: “Am I supposed to wish that my husband were dead? Really that is outrageous nonsense! Not only is our married life very happy, though perhaps you won’t believe that, but if he died I should lose everything I possess in the world.” (Freud 1929, 121)

This woman directly challenges Freud’s diagnosis in the same way Witnessio challenges the deceptionist account of lying (Artie thinks that he is lying; the woman thinks that she wants her husband alive). Freud’s answer is very interesting.

Assuming that unconscious tendencies do exist in mental life, the fact that the opposite tendencies predominate in conscious life goes to prove nothing. […] What does it matter if you [the woman] do find the results of dream-interpretation unpleasant, or even mortifying and repulsive? “Ça n’empêche pas d’exister” [“It doesn’t prevent things from existing”]. [Freud (1929), 122; italics added]

Freud practically says that the woman incorrectly believes that she does not want her husband dead because her desire is unconscious. This reply raises two important problems (see Derksen 2001). According to the reply, Freud’s theory is correct no matter what the woman says: the only difference is in whether the desire is conscious or unconscious. A theory defended in this way can never be disproved since any testimony becomes evidence in support of it. Therefore, the reply makes Freud’s position unfalsifiable; the first flaw. The second flaw is that the comeback is viciously circular (i.e., it begs the question). Freud correctly says “assuming that unconscious tendencies do exist in mental life [i.e., that his theory is correct],” the fact that this woman’s testimony contradicts his theory “goes to prove nothing.” But, whether his theory is correct is exactly what is at stake. This testimony is a counterexample to his theory and the theory cannot be used as a reason to disregard it.

The argument to the conclusion that bald-faced lies do not involve genuine assertions involves the same fallacious line of reasoning. This is the analogy between the two replies: Freud says that, if the woman concurs with his diagnosis, her desire is conscious and, if she denies it, the desire is unconscious, she just thinks that she does not have the desire. Analogously, according to the given argument, if speakers intend to deceive by lying, then their lies involve genuine assertions and, if they do not intend to deceive, then their “lies” do not involve genuine assertions; these speakers just think that they are genuinely lying.

And here is the immediate problem with this comeback: if we cannot trust the speakers’ judgements as to whether they are lying or not, then no testimony can be used as a counterexample to the given analysis of assertion. Therefore, the reply causes the view to become unfalsifiable. This reply uses the idea that intending to give grounds for beliefs is necessary for asserting to discredit cases of bald-faced lies but whether this intention really is necessary for asserting is exactly what is at stake in this debate. Therefore, the reply begs the question.

The circularity is actually very visible in this argument. What the reply is actually saying is that the deceptionist definition of lying fits neatly within the Neo-Gricean definition of assertion, which in turn fits neatly within a relevant broader Gricean theory of speech acts and communication, but this is the same theory. In effect, then, the argument says that assuming that Neo-Griceanism is correct, the fact that other people (e.g., Artie, Pinocchio, Marta) see bald-faced lies as genuine lies goes to prove nothing; their testimonies do not matter. This is not to say that Neo-Griceanism is incorrect but this specific defence is problematic and, as philosophers, we should be basing our views on good arguments. Therefore, because it is viciously circular and unfalsifiable, this argument should not be accepted unconditionally. I now proceed to argue that it also makes a very daring claim; this is step 2 in my argument.

This dispute is not only about what people (Artie, Marta Cabrera, etc.) recognize as lying but also about what they recognize as asserting. Therefore, in discrediting bald-faced lies in this particular way, one is not just saying that common folks have a broader conception of lying, but also that they cannot recognize when a proposition is being asserted.⁸ This strikes me as a rather bold position, which is the second premise in my argument, step 2. In a standard context, assertion simply seems to be a default interpretation of a declarative sentence, and this view seems to be common ground in the debate. Williamson (2000, 258), for instance, writes: “In natural language, the default use of declarative sentences is to make assertions.” That being said, even though the idea is bold, it is not completely unwarranted. Consider the following argument made by Keiser (2016).

In The Godfather 2 (Paramount, 1974), “Frankie Five Angels” Pantangelli is called in as a surprise witness in a Senate hearing against the mob boss Michael Corleone but, to everyone’s utter shock, he goes against the agreement and claims under oath that he has no knowledge of any wrongdoings committed by Michael Corleone. Keiser (2016, 471) argues that, since he does not intend to give his audience grounds for believing what he says, Frankie does not assert what he says; rather, he is playing a “courtroom [language] game,” a game in which a speaker can avoid asserting what they say but still achieve a specific intended effect (e.g., go for the record).

People standardly think that witnesses assert statements they make while testifying under oath—witnesses assume many assertoric commitments (they, e.g., guarantee that what they say is true)—and the terms “lying on the stand” or “lying under oath” are standardly understood as lying by asserting what the witness believes is false. According to Keiser, however, Franky was just making a move in a courtroom game. If he lied, this was not in the sense we are discussing here (it does not involve the default use of a declarative sentence) and our intuitions about the case are incorrect. Following this analysis of Frankie, one may say that, while we may allow that common folk can recognize asserting in standard contexts, the cases I discuss are all fictional and very unusual (nose growing, blushing). Therefore, even if it is true that people can reliably detect assertions in standard contexts, this is of limited value for my argument.

Keiser does not offer an unreasonable interpretation of this particular case, people’s intuition may go in the wrong direction in the “courtroom” context, but the idea that bald-faced lies are not genuine assertions is not only bold, it is also controversial. For one, choosing Frankie to support a very general claim—namely, that no bald-faced lie involves a genuine assertion—is rather unfair: Frankie involves a specific, non-standard context that allows Keiser to apply her “game” analogy but people predominantly lie in standard contexts. Therefore, we cannot apply insights from Frankie to all cases of bald-faced lying: the analogy breaks. People would have been equally shocked if Frankie had said that he knew no Godfather in a standard context far away from the courtroom (e.g., in a private conversation), they would have been shocked because they would think that he is lying, and Keiser’s argument would not apply here as easily as in the courtroom context.

Gambler, for example, involves a standard context and there are good reasons to say the same about Witnessio. After all, Artie is not in a courtroom, and the police did not arrest him or bring him in for questioning; they were just talking. Consider the following combination of Gambler and Witnessio.

Gamblessio preserves the virtues of Witnessio while making the stakes lower. Therefore, even though the position according to which our intuitions about lying may be unreliable may make sense when applied to cases such as Frankie, it fails to easily generalise to all cases of bald-faced lying.

As I argued in step 2, the claim that no bald-faced lie is an assertion is very bold: it entails that many people are not competent speakers. Of course, there is nothing wrong with making bold claims per se but, other things being equal, we should go for less demanding claims. And other things are not equal: this bold argument not only begs the question and generates an unfalsifiable position, it also suffers from three additional problems. Two can be immediately noticed. One problem is that this position cannot be easily generalised to all cases of bald-faced lies. I discussed this problem here. It is not really obvious that we can say that Artie and Carmela do not know what it is to assert a proposition in the context of Gamblessio. Another problem is that Artie and Pinocchio assert according to many successful accounts of assertion and so we do not need to commit ourselves to very demanding positions.

Other views will say that Artie and Pinocchio assert what they say because they take themselves as being in a warranting context (Saul 2012), because they propose that what they say be added to official common ground (Stalnaker 1984, 1999, 2002; Stokke 2013, 2016, 2017, 2018), because they represent themselves as believing what they say (Black 1952; Davidson 1998; Fallis 2013) and even as knowing what they say (Unger 1975, 250–270; DeRose 2002, 185). They also count as asserting what they say because they make many assertoric commitments: they warrant the truth of what they say (Carson 2006, 2010), they undertake the responsibility of justifying their assertion and what follows from it (Brandom 1994, 173–175), they commit themselves to act in accordance with what they say (Dummett 1981) or that they will withdraw it if the proposition is shown to be untrue (MacFarlane 2005; similarly, Dummett 1991, 165), they commit themselves to the truth of what they say (Marsili 2021; similarly, Dummett 1981, 300), and so on.⁹

With all of this in mind, while one need not think that Neo-Griceanism delivers a failed analysis of assertion, one must wonder whether making bold claims just to keep a particular interpretation of Neo-Griceanism is justified. It may just be that Neo-Griceanism allows for exceptions.¹⁰ If this is correct, we get to keep both the Neo-Gricean analysis of assertion and the idea that some lies not intended to deceive are genuine assertions. This is why, in my final step 3, I argue that the intention to give grounds for believing what you say cannot be necessary for asserting what you say; this is the additional third problem this bold position faces. Consider a real-life case involving a sincere speaker.

Because she is well aware that, whatever she says, Miller’s credence in the proposition will not be shifted one bit, Arangio cannot reasonably intend to give Miller grounds for believing what she says based on her say-so. Therefore, according to the given argument, because she cannot rationally intend to give Miller grounds to believe what she says, Arangio cannot rationally assert to Miller a proposition she believes to be true. In other words, according to a consistent application of the argument claiming that no bald-faced lie is an assertion, in this situation, it is impossible for Arangio to assert a proposition she believes is true. This result is surely counterintuitive: the issue of whether it is possible for me to sincerely assert something to you should depend on me (i.e., on whether I can utter the proposition, etc.), not on you (i.e., on whether you will believe me or not). Therefore, we should think that a consistent application of the deceptionist argument misclassifies some sincere assertions.

Given the arguments from steps 1–3, this particular argument cannot be used as a reason to say that bald-faced lies are not genuine assertions. In particular, the line of reasoning is such that it misclassifies some sincere assertions, because the argument begs the question, generates an unfalsifiable position, and is difficult to generalise to all bald-faced lies, we can safely assume that it is not a reason to think that bald-faced lies are not genuine assertions.

However, my argument needs to resolve one more issue: we need to see whether Artie asserted the proposition he uttered (\(\neg p\)) or the proposition he intended his hearers to infer (\(p\)). For, if he asserted the latter, then he asserted what he believed was true and thus did not lie. I discuss this interpretation below.

2.2 Indirect Assertion

The sentence “He’s (She’s) a friend of Dorothy” in the early 20^th century US and British homosexual subculture made a claim about a person who was a homosexual.¹² Because expressing their sexual orientation was a criminal offence, homosexuals had to hide it. This sentence made it possible for people to say that a certain person is gay without uttering that proposition. With that in mind, consider this situation.

Arguably, Grace asserts that Bill is gay because she means “Bill is gay” when she utters “Bill’s a friend of Dorothy” and because they are in a context in which it is common ground that she asserts the proposition she means (this is what she says) rather than the proposition she utters (the literal meaning of the uttered sentence). Our intuitions seem to correspond with this interpretation: it is natural to think that Will will think that Grace lied to him if Bill turns out not to be gay but not if it turns out that Bill does not know a girl named Dorothy.

Dorothy highlights a difficulty that may arise concerning my cases: in these cases, someone is intending to communicate proposition \(p\) by means of uttering \(\neg p\) under certain circumstances, and hence it may be that they are asserting \(p\), rather than \(\neg p\). That is, it may be that Artie actually asserts that Tony committed the crime (\(p\)) by uttering “Tony definitely did not commit the crime” (\(\neg p\)). The means by which Artie asserts \(p\) (i.e., by uttering \(\neg p\)) is unusual, but it is hardly impossible to assert one proposition by uttering another.¹³

Indirect assertion is not an uncontroversial concept (see, e.g., LePore and Stone 2014; Garcı́a-Carpintero 2018) but this interpretation of Witnessio is plausible and interesting enough to be seriously considered. I assume that the relevant analysis of my cases would go something like this. Just as Grace intends that her utterance “He’s a friend of Dorothy,” in the light of the relevant subculture’s linguistic conventions, means “Bill is gay,” Artie intends that his utterance “Tony did not commit the crime,” in the light of his blushing, be understood as meaning “Tony committed the crime.” And, because he asserts what he says (standard context) and he says what he means rather than what he utters, Artie is not lying—since he believes that the meant proposition is true.

It is pertinent to note that this interpretation is not consistent with the main idea behind the Pinartio-style examples. Artie believes that his blushing signals that Artie believes the opposite of what he says (i.e., it signals that Artie lies). He does not believe that blushing signals that he means the opposite of the literal meaning of his utterance.¹⁴ However, this is not a reason not to consider this interpretation, my descriptions of the cases could be misguided. Let us therefore consider how the received analyses of assertion explain Artie’s and Grace’s behaviour.

Artie warrants the truth of the uttered proposition (the literal meaning of the uttered declarative sentence) rather than the truth of the proposition he intends the audience to infer from his behaviour. He proposes that the uttered proposition be added to official common ground, he represents himself as believing or knowing the uttered proposition (“Tony is not the murderer”), he commits himself both to the truth of this proposition and to act in accordance with this proposition. This is in clear contrast with Grace’s behaviour. Grace warrants the truth of the “Bill is gay” proposition, she proposes that “Bill is gay” be added to official common ground, she represents herself as believing that Bill is gay, she commits herself both to the truth of this proposition and to acting in accordance with this proposition. Therefore, the fact that it does not seem odd to think that Grace indirectly asserts “Bill is gay” is not a reason to think that Artie indirectly asserts “Tony is the murderer.”

We see that neither Pinocchio nor Artie is willing to accept any assertoric responsibility for the communicated propositions but only for the literal meaning of the uttered declarative sentences. Grace, however, does seem to be taking assertoric responsibility for the “Bill is gay” proposition. Therefore, while the idea that Artie indirectly asserts the proposition he intends the police to infer from his blushing and the uttered proposition is rather interesting, it does not seem to capture the relevant cases in the right way. Artie and Pinocchio intend to cause their hearers to realise that they are non-deceptively lying to them. This is vital for the success of their plan: the hearers should infer the truth from the fact that the blushing and the nose growing show that Artie and Pinocchio are lying.

3 Concluding Remarks

The idea that some lies can be intended to communicate the truth by having the hearer recognize that the speaker is lying is both plausible and important. By showing that one can lie without intending to deceive anyone, tell-tale sign bald-faced lies also vindicate the non-deceptionist interpretation of the common knowledge bald-faced lies. Since we now know that lying without intending to deceive is possible, we can think that common-knowledge bald-faced liars—the gambler, for instance—could be genuine liars who did not intend to deceive. The tell-tale kind of bald-faced lies, thus, takes the debate out of the impasse and suggests that the deceptionist analysis of lying should be abandoned.

We can now move on and focus on other aspects of lying. For example, we can start analysing scenarios in which people typically lie to themselves and try to identify their motivation for such behaviour. Lying to myself is an intrapersonal analogue of interpersonal tell-tale sign bald-faced lies: I will immediately know when I form the intention to lie to myself, I will know that I will know this, etc. Therefore, tell-tale sign bald-faced lies can help us to understand a much bigger class of human behaviour. Because a bald-faced self-liar will probably have similar motives as a bald-faced interpersonal liar, understanding other people’s behaviour—namely, why others bald-faced lie to us—will help us to understand our own behaviour. And vice versa, we will be able to understand why other people lie better if we investigate our own motives for lying to ourselves. So, I suggest that this is the direction in which our analysis of lying should take.