A physician shall … strive to report physicians … engaging in fraud or deception to appropriate entities.
American Medical Association, Principles of Medical Ethics
From the days of Plato and Socrates, via John Gregory and Thomas Percival in the 18th century and William Osler in the 20th, the most esteemed doctors have deceived their patients with good intentions. Today, however, the consensus on the legitimacy of deception in Western medical practice is clear: Doctors should never deceive their patients. Whether this absolute rule is morally sound or translated into clinical practice is doubtful, but I shall not, in this paper, address the ethical and empirical issues on truth telling and deception. Instead, I shall dissect the concept of deception, evaluating various possible definitions and their limitations, and propose a new definition that embraces both acts and omissions.
What Is Deception?
Behind the apparent simplicity of the question lies a tangled web of definitions. Sociobiologists, ethologists, anthropologists, child psychologists, linguists, and philosophers have differed in their interpretation of deception, each employing the term to describe behavior specific to their respective fields of study.1
Mitchell R, Thompson N, eds. Deception: Perspectives on human and nonhuman deceit. New York: State University of New York Press; 1986.
See note 1, Mitchell, Thompson 1986:3.
Galasiński D. The language of deception; a discourse analytical study. London: Sage; 2000:18.
Truth Value
The truth value (i.e., the truth or falsity) of an assertion is, in itself, inadequate to determine whether or not an act is deception. A person can utter a sequence of true statements whose irrelevance or obscurity is such that they deceive as convincingly as a lie. Truthfulness clearly requires more than the mere avoidance of lying, and an utterance can deceive although it is neither true nor false. Galasiński gives the famous “Have you stopped beating your wife?” example as a sentence that has no truth value but which may still be deceptive.4
See note 3, Galasiński 2000:19.
In what Bilmes calls the “world of total meaning,” where every utterance or pause can be meaningful, silence itself can be deceptive.5
Bilmes J. Constituting silence: Life in the world of total meaning. Semiotica 1994;98:73–87.
Saville-Troike M. The place of silence in an integrated theory of communication. In: Tannen D, Saville-Troike M, eds. Perspectives on Silence. Norwood, N.J.: Ablex; 1995:6.
See note 6, Saville-Troike 1995:7.
Deception, then, can be nonverbal. A smile (to feign liking someone), a silence (to confess to a false accusation), or a nod (although you strongly disagree with the speaker) can constitute deception.
False Belief
Much of the literature on deception incorporates the notion of “false belief” into the definition of deception.8
Russow L-M. Deception: A philosophical perspective. In: Thompson N, ed. Deception: Perspectives on Human and Nonhuman Deceit. New York: State University of New York Press; 1986:48.
Deception (a): Deception is a communicative act intended to induce or maintain what the agent believes to be a false belief in the target.
Galasiński, however, identifies a problem with this definition: A renowned liar may deliberately induce a true belief in the hearer by saying something false.9
See note 3, Galasiński 2000.
[T]he target of the deceptive act is manipulated into thinking what suits the deceiver; that this happens to be the truth is, in my opinion, of minor importance.10
See note 3, Galasiński 2000:20.
A concrete example will clarify this: Imagine I witness a gun-toting criminal pursuing someone I deeply dislike. The criminal, who distrusts me greatly, asks me for the whereabouts of my nemesis. Intent on seeing my fleeing foe injured by the thug and expecting that the latter will do the opposite of what I tell him, I point to the opposite direction and support this verbally. In short, my intention is to make him believe something that is true, although the statement itself is false. This would constitute a case of deception, although I am not inducing a false belief. The “false belief” definition therefore fails to cover all types of deception. But is this so?
The “traditional” definition (Deception (a) above) can still work if we say that I have deceived the criminal not about the whereabouts of the fugitive but about my own truthfulness. In other words, the criminal (C) believes the proposition Φ (= Sokol will lie). I know that C believes Φ (= Sokol will lie), therefore I decide to ¬ Φ (Sokol will not lie), thus inducing a false belief in C who erroneously thinks Φ.
Galasiński adopts another approach to overcome the problem: He alters the definition of deception by introducing the notion of manipulation.11
See note 3, Galasiński 2000.
Deception (b): Deception [is] a communicative act that is intended to induce in the addressee a particular belief, by manipulating the truth or falsity of information.12
See note 3, Galasiński 2000:20.
Under this definition, a liar who deliberately says something false to make someone believe the truth is still deceiving the “target” by manipulating factual information to attain a hidden goal.
The main problem with this definition is the inclusion of the equally unclear concept of manipulation, which itself needs to be elucidated. What does “manipulating” mean? In what way is it distinct from straightforward assertion?13
Galasiński does provide a definition of manipulation: “an attempt to affect the target in such a way that her or his behavior/action is an instrument of attaining the goals of the manipulator, who acts without using force but in such a way that the target does not know the goal of the manipulator's action.” I am not convinced that this definition captures what people typically understand by the notion of “manipulation.” We often do not know the goals of people's actions or utterances, but this does not necessarily mean we are manipulated. Like deception, manipulation is a complex notion, heavily dependent on context.
Misleading
Like Galasiński, Ekman avoids using the phrase “false belief,” and focuses instead on the intention to mislead.14
Ekman P. Telling Lies. New York: Norton & Company Ltd; 1985.
Deception/lying (c): Deception/lying is a deliberate attempt to mislead a “target”, without prior notification of this purpose and without having been explicitly asked to do so by the “target”.
His broad definition of a lie encompasses concealment. A doctor who deliberately conceals information about his terminally ill patient is, according to Ekman, lying although no false information has been uttered. The concealment constitutes lying because it is (a) intentional, (b) misleading, and (c) has not been requested by the unsuspecting patient. While acknowledging that liars generally prefer concealment to falsification, Ekman suggests that both types are equally reprehensible. The targets may be harmed to the same extent in both cases. Concealers who believe that their lies are morally less bad than those of falsifiers are indulging in self-deception. They can persuade themselves that the addressee either already knows the truth or does not want to know it. They can also believe that they are not, by their definition, lying to the addressee.
Ekman's attempt to view concealment as a form of lying is, in my view, flawed, at least because his definition may suggest that most of us are constantly lying. If I notice that the person standing next to me has his zipper undone, and I deliberately withhold this fact, am I lying? Certainly I am not in the commonly used sense of the word. I am intentionally withholding the information and have not been asked to do so by my unzipped neighbor. The question then is: Am I misleading him by withholding the information? This raises the definitional problem of what it means to mislead a person, which again is similar to my initial question, “What is deception?” The inclusion of “mislead” in a definition of deception amounts to a tautology. For this reason, I am not convinced that Ekman's definition clarifies our understanding of deception. Replacing a problematic term with an equally problematic one does not resolve the definitional conundrum.
I am also skeptical of Ekman's “request condition,” which states that a lie can only be a lie if it has not been requested by the addressee. Consider the following (I hope) far-fetched scenario: When I first meet my new GP, I grant him permission to lie to me if he believes this is in my best interests. He agrees. Two years later, and two days before my Ph.D. viva, I travel to the Congo and contract a virulent strain of the Ebola virus. Aware of my fervent, life-long desire to obtain this Ph.D. and the 1-week incubation period of the virus, the GP lies to me about the severity of the disease when, slightly unwell, I ask him about the prognosis. Now, although I have asked him to lie, it seems clear that the GP has lied to me. In my view, determining an utterance's status as a “lie” is not dependent on external factors (such as an explicit request or the “success” of the lie) although of course this is not necessarily so of its moral status. The GP may well be morally justified in lying to me in response to my request.
Concealment and Deception
Although lying does not, in my view, rely on an expectation condition to qualify as a lie, withholding information does rely on such an expectation to count as deception. At any one time, people hold hundreds of false beliefs. They do not, however, expect others to correct them unless there is good reason to do so. Just as we hold many false beliefs about the world, so do we withhold true beliefs from others. Because there is generally no expectation to reveal those true beliefs, it would be odd to say we are concealing them from others, let alone that we are deceiving them. Benn writes that an agent only deceptively conceals “if the context is such that, were it the case that p, people could reasonably expect me [the agent] to reveal that p.”15
Benn P. Medicine, lies and deception. Journal of Medical Ethics 2001;27:130–4 at p. 132.
Under my chosen interpretation, determining whether an act of concealing is intentionally deceptive will rest on (1) the agent's intention, (2) people's reasonable expectations in the circumstances, and (3) the success of the attempted deception. If I invite the nosy Jones to my house for dinner, expecting that he will want to peruse my collection of rare books on Mongolian dung beetles, and consequently hide my books in the attic, have I (intentionally) deceived him? As long as I succeed in concealing the books, the answer will depend on my intention and whether Jones' expectation to inspect my books is reasonable.16
In the sense used here, an expectation is “reasonable” if it is supported by good reasons as judged from an external perspective. Subjective beliefs about the reasonableness of an expectation do not entail that the expectation really is reasonable. One can be mistaken about the reasonableness of one's beliefs or expectations. The notion of “reasonable expectation” is discussed in the section below.
Smith: “I hid the books from Jones, although I knew he wanted to see them. I deceived him!” A: No, you didn't. He didn't have any right to see your books. B: Yes you did, but he didn't have any right to see your books.
The first answer (A) represents my preferred interpretation of deception. Answer B represents the interpretation relying on the perceived, rather than the actual, reasonableness of expectation.
Reasonable Expectation
To gain a clearer understanding of the phrase “reasonable expectation,” I briefly examine the notions of “expectation” and “reasonableness.”
Although an expectation can be merely descriptive (e.g., I expect to come last in the race because I am unfit), the expectation in the proposed definition of deception (see below) is normative, indicating that the “expector” believes the speaker ought to disclose the information. Because it is normative, a failure to meet this expectation would generally be met with disapproval by the expector.
An important point relates to the meaning of the phrase “the target's reasonable expectation” of truthfulness. This does not mean, as Paprzycka notes, that the target actually does expect me, the agent, to be truthful. He may not expect anything. Rather, it should be understood in the conditional sense: “If the target were to expect of me that I tell him the truth, my target's expectation would be reasonable.”17
Paprzycka K. Social anatomy of action: Toward a responsibility-based conception of agency. Ph.D. dissertation. Pittsburgh: University of Pittsburgh; 1997.
The notion of reasonableness can be divided into two distinct concepts: agent reasonableness and normative reasonableness.18
See note 17, Paprzycka 1997:79.
The appeal to “(normative) reasonableness,” by referring to an external standard of appropriateness, serves to invalidate the normative force of inappropriate expectations (such as John's expectation to look at my books, in the earlier example). If John's expectation of truthfulness is based on irrational beliefs, for example, then I am under no obligation to meet his expectation.
Deciding whether or not an expectation is (normatively) reasonable, then, will rely on the strength of the reasons for and against the expectation. Paprzycka writes that “because an expectation involves placing a demand on another person, such a demand must be justified and weighed against various kinds of considerations.”19
See note 17, Paprzycka 1997:85.
Invoking external reasons to justify the reasonableness of an expectation is unlikely to yield clear-cut answers in all cases. There may be situations, as in moral dilemmas, where more than one expectation is reasonable. There may also be disagreements over what constitutes a reasonable expectation, due to agents holding different (or differently prioritized) moral values. For this reason, calling an act “deception” is not as straightforward as calling an act of coughing “coughing.” Little interpretation is needed to recognize an act of coughing, whereas labeling an act “deception” needs to be justified by appealing, in part, to the reasonableness of the agent's expectation and thus to the reasons supporting the expectation.
Although the example I offered earlier (of the nosy John and the dung beetles) deals with deception by omission, the notion of “reasonable expectation” can also be applied to deception by commission. A doctor who enthusiastically tells his dying patient “I have seen people with your condition recover miraculously!” without mentioning that he has only seen two cases in his long medical career may be deceiving the patient if one could reasonably expect this important information to be shared.20
I owe this example to Raanan Gillon (personal communication, October 17, 2005).
The decision to include the notion of reasonableness acknowledges the moral dimension of deception. It also stresses the highly contextual nature of deception.
Settling on a Definition
As the discussion above shows, defining deception is an arduous task. Ekman and Galasiński's efforts to redefine the term do not, in my view, settle the definitional question. Galasiński's reliance on “manipulation” creates as much as resolves problems, and Ekman's suggestion that all traditionally deceptive acts are in fact “lies” fails to draw distinctions between clearly different types of deceptive strategies.
Although I initially attempted to arrive at a definition of deception that included no morally laden terms, I eventually changed my mind. The notion of “reasonable expectation” of truthfulness seemed essential to capture the meaning of deception.21
The notion of “reasonable expectation” is also relevant to aspects of U.S. law, notably eavesdropping and privacy. In Katz v. United States (389 U.S. 347 [1967]), the Supreme Court ruled that FBI agents who attached a listening device to the exterior of a public telephone booth to record the conversations of someone unlawfully transmitting gambling information to other states was unconstitutional. The Court agreed that the “petitioner's” reasonable expectation of privacy was violated and, in the words of Mr. Justice White, that the petitioner “spoke under circumstances in which a reasonable person would assume that uninvited ears were not listening.” Available at: http://caselaw.lp.findlaw.com/scripts/printer_friendly.pl?page=us/389/347.html (accessed Nov 22, 2005). I thank Ronald P. Sokol for pointing this out to me.
I thus propose the following definition of deception:
Deception (d): Deception is a communicative act intended to induce or maintain what the agent believes to be a false belief in the target when (1) the target's expectation of truthfulness is reasonable and (2) the agent is successful in producing the intended deceptive outcome.
Perhaps less confusingly: “I deceive Jones if I intend to induce or maintain what I believe is a false belief in Jones when Jones expects me, with good reason, to be truthful and I succeed in inducing or maintaining that perceived false belief.”
I do not claim the definition to be watertight. Although I opted to include the second clause—the “success” condition—into my definition, thereby distinguishing between attempted but unsuccessful deception and successful deception (in much the same way as attempted murder is distinguished from murder), I acknowledge that others may not view a successful outcome as a necessary condition for deception. Under this last interpretation, an act or omission that is intended to deceive is sufficient to label that act as “an act of deception” even if it fails to deceive. My suggested account of intentional deception, on the other hand, requires both an intention to deceive and a successful outcome. The advantage of this definition is that it can account for semantically odd sentences of the type “I was deceived, but wasn't fooled” or “I deceived him about the location of the treasure but he found it,” which surely prompt the response “so you weren't deceived/didn't deceive him!” These examples lend support to the “success” condition included in my definition.
How does the proposed definition cope with the problematic “gun-toting criminal” example, where I direct the distrustful pursuant in the right direction by pointing the opposite way? Assuming I successfully direct him in the right direction, the definition rules that I did not deceive the pursuant about the fugitive's whereabouts (because I did not intend to induce a false belief about this), but I did deceive him about my intention, which he falsely believed was to lead him away from the fugitive. In my view, this is not a counterintuitive result.
Conclusion
Whatever one's preferred definition, clarifying the concept of “deception” should logically precede the normative question about the moral legitimacy of deception in clinical practice. I have shown in this paper that dissecting “deception” is a complex and daunting procedure that yields varied results. In light of the definitional disagreements and the strong negative connotations associated with “deception,” it may be advisable to avoid the term altogether, replacing it with a clearer, less loaded description of the act in question. For example, instead of asking, “Is it morally acceptable to deceive a dying cancer patient?” it would be better to ask, “Is it morally acceptable to withhold bad news from a dying cancer patient?” This phrasing is more conducive to rational debate than one that includes the contentious term “deception.” As it stands, labeling an act “deception” is often used as a smoke screen to condemn the act without elaborating on the reasons for its moral wrongness.
