In ideal conditions, communication is honest and cooperative. A message is exchanged between a sender to a receiver; as a result of the process, truthful and relevant information is shared. The study of human communication has often focused on investigating such ideal cases. Unfortunately, however, communication does not always happen under ideal conditions. Speakers are frequently insincere, and they sometimes aim to deceive and manipulate their audience for selfish goals. To develop a full picture of how human communication works, non-ideal cases need to be studied too, and our theory of how honest communication works needs to be complemented with a theory of how dishonest communication works, and of its ethical, epistemological and political implications. The collection of essays Lying, Knowledge, Ethics and Politics, edited by Eliot Michaelson and Andreas Stokke, is a welcome contribution in this emerging field of studies, which both expands on the existing research and opens up new pathways for future investigation.

Several of the essays in the volume deal with definitional questions about lying, insincerity, and related concepts. Much contemporary discussion has revolved around whether Deception is a necessary condition for lying, in addition to Assertion and Insincerity:

In recent years, many examples of lying without intent to deceive (e.g. bald-faced lies) were discussed in the literature, leading many philosophers to conclude that Deception is not a necessary condition for lying.

Don Fallis's opening essay ‘What is Deceptive Lying’ looks at the debate on Deception from a different angle. Even if Deception does not hold (and not all lies are intended to deceive), we can still be interested in defining what makes a lie deceptive, when it is. This task proves difficult: Fallis shows that existing formulations of Deception fail to capture all and only deceptive lies. To address this problem, Fallis introduces two alternative formulations of Deception. But while these alternative formulations fare better than existing ones, they also fail to accommodate all the examples summoned in the essay. Proponents of Deception now face one additional challenge: in addition to failing to capture all and only lies, it seems that they are also unable to capture all and only deceptive lies.

Ishaini Maitra's essay ‘Lying, Acting and Asserting’ is more optimistic about Deception, and attempts to vindicate it as a necessary condition for lying. Maitra argues that, if one accepts a specific definition of assertion (such that you assert p only if you are sufficiently responsive to your total evidence that bears on p), some putative counterexamples to Deception (e.g. lies under coercion) should not be regarded as genuine assertions (but rather pretend-assertions), so that they cannot be genuine lies.

Some commentators before me (Viebahn 2019, Marques 2020) have noted that Maitra's argument is problematic. I would like to highlight two weaknesses myself. The first is that the argument relies on a rather unorthodox conception of what assertion is: only if we follow Maitra in defining assertion as she does (tying it to the evidence-responsiveness rule) is her argument fully persuasive. The second difficulty is that, even if successful, the argument only seems to show that some counterexamples to Deception (e.g. lies under coercion) are not genuine lies. But there are many other counterexamples to Deception that would not easily admit the same treatment (cf. Arico & Fallis 2013, Rutschmann & Wiegmann 2017). Lacking a strategy to generalise to other counterexamples, the argument falls short of rescuing Deception from existing objections (to be sure, this limitation is shared by other attempts to rescue Deception from counterexamples, such as Leland 2015 or Harris 2019).

Katherine Hawley's essay takes a different route to argue that coerced lies are not genuine lies. When it turns out that a monetary gift was made under coercion, Hawley notes, it is natural to think that its recipient hasn't really earned the right to keep the money. A similar point would apply to coerced promises and assertions: we are not bound by their normative consequences. If this is right, and coerced statements lack the full normative import of genuine assertions, ‘lies’ told under coercion are not genuine lies.

Hawley's argument is convincing. However, I am not sure that coercion always defeats assertoric force. Consider a context where no member of the audience is aware of the threat: for instance, a witness forced by a mafia gang to lie on the stand. In this context, everybody (including the speaker) regards the statement as an assertion. My intuition is that here the speaker has asserted and lied (but cf. Hawley's essay, p. 242). If this is right, some coerced utterances can be genuine assertions and lies after all (whenever the audience is unaware of the coercion).

Jessica Pepp's ‘Truth Serum, Liar Serum’ focuses on the Insincerity condition for lying. It discusses some classic reference-fixing puzzles (by Frege, Kripke and Salmon), showing that they generate problems for Insincerity. To complicate matters, Pepp introduces an even more complex scenario. A thief mistakenly thinks that she has been given a truth-serum. She forms the belief that she can only say the truth, and that even when it seems to her that she is saying something false, she will actually be speaking truthfully. When she lies to her captors, she believes (incorrectly) that she is saying something true. Despite lying, she will not satisfy Insincerity. Potential revisions of Insincerity are then considered, concluding that none is fully satisfying – although the following fares better than others:

Pepp does not acknowledge any previous formulation of this condition. This is despite the fact that her historical review begins from Augustine and Aquinas, and a constraint equivalent to INS* is already found in Aquinas: ‘ratio mendacii sumitur a formali falsitate, ex hoc scilicet quod aliquis habet voluntatem falsum enuntiandi’ [the definition of lying is taken from formal falsity, i.e. from the fact that someone has the intention to state what is false] (ST, II, II, 110 my italics; cf. also Fallis 2012, 577).

I have expressed sympathy for this constraint myself (Marsili 2018: 174). According to Pepp, however, INS* still fails to classify the liar-serum example as a lie, because the thief does not believe that what she is saying is false. Still, this problem is not insurmountable. The thief thinks that what she is saying is true, but she also believes that the proposition that she is trying to assert is false. If (with a slight amendment) we tie ‘that’ in INS* to the latter proposition, INS* will correctly predict that the thief is insincere.

Some essays in this collection are foundational works that open up new avenues for research. Jennifer Lackey's ‘Group Lies’ asks under which conditions a group can lie. Lackey elaborates a convincing strategy to extend her account of individual lying to groups. While the proposal is surely promising, I suspect that it inherits some weaknesses of Lackey's account of individual lying. Lackey's (2013) reformulation of Deception notoriously fares better than other accounts, but is subject to known counterexamples (Fallis 2015). Lackey's formulation of Insincerity is known to be too restrictive (Carson 2006: 286; Saul 2012: 20; Marsili 2016, 2018: 175). And Lackey's formulation of Assertion is too weak, because it requires that the speaker state, rather than assert, the relevant utterance, which leads to incorrect classifications (Marsili 2020). That noted, Lackey's strategy for translating a definition of individual lying into a definition of group lying is surely groundbreaking, and a noteworthy contribution to the literature.

Roy Sorensen's ‘Lying to Mindless Machines’ offers a pioneering discussion of lying to mindless machines (AIs). Sorensen argues that we can lie to machines, but machines cannot lie to us. If he is right about the first claim, and we can lie to mindless machines, Deception must be false: we need not intend to change someone's mind (since machines have none) in order to lie.

Most of the volume's essays focus on how individual statements can be used to deliberately convey false propositions. Jennifer Saul's article examines other ways in which falsehoods spread through communication. Sometimes falsehoods are produced by the aggregation of different discourses (Saul calls these aggregate falsehoods). For instance, a journal that disproportionately covers crimes by black people may create the false impression that black people are more likely to commit crimes. Other times, media propagate falsehoods unintentionally, by mere negligence, because their journalists and editors did not check their sources as carefully as they should have (Saul calls these negligent falsehoods). Although aggregate and negligent falsehoods are not strictly speaking ‘lies’, they are common and can be immensely damaging to the communities that they target. To understand how nefarious stereotypes are created and spread, Saul concludes, researchers need to pay more attention to these neglected forms of disinformation.

The volume is divided into four main sections, each dealing with one of the four topics mentioned in the title (lying, knowledge, ethics, politics). Due to space constraints, I have been unable to comment on the sections of knowledge and ethics, which indeed contain some excellent essays. If I have not discussed essays on politics, it is because (perhaps fittingly for a volume that deals with deception), none of the essays grouped under ‘politics’ discusses (or touches upon) the political implications of dishonest communication (with the exception of Saul's essay). This may be the main weakness of the volume: despite being published in the ‘Engaged Philosophy’ series of Oxford University Press, it offers very little material that fits this label. That said, this is an important collection with several excellent contributions, some of which will contribute to redefining research in this field in the forthcoming years.