*This is the second installment of Sam Ponders Epistemology. As noted before, I am not a philosopher, so take this with a grain of salt.

Sam Seitz

Perhaps the most incisive examination of bullshit is Harry Frankfurt’s On Bullshit, which attempts to delineate the distinction between lies and bullshit more clearly. Frankfurt concludes there are two criteria that distinguish bullshit from lies. The first is a lack of knowledge of the truth: While a liar knows the truth and tries to conceal it, a bullshitter does not know the truth yet attempts to convince an audience that she does.[1] The second criterion is that the bullshitter “does not care whether the things he says describe reality correctly.”[2] These two distinctions go a long way in providing a more rigorous definition of bullshit and helping to distinguish bullshit from lies and other forms of deception. However, Frankfurt’s two criteria – a lack of knowledge of the truth and a lack of concern about describing reality faithfully – fail to account for instances of inadvertent bullshit where the bullshitter does not realize that he lacks access to the truth.

Before further exploring inadvertent bullshit and its implications, it is worth detailing Frankfurt’s conception of bullshit more closely. To do this, it is helpful to consider an example offered by Frankfurt in On Bullshit. The example provided involves an incident between the philosophers Ludwig Wittgenstein and Fania Pascal. When Pascal describes her physical condition after having her tonsils removed, she compares herself to “a dog that has been run over.”[3] This prompts Wittgenstein to retort that she does not know “what a dog that has been run over feels like.”[4] As Frankfurt points out, Pascal is not lying: She knows that a dog hit by a car does not feel well and she is simply trying to convey that she also does not feel well. However, in her attempt to convey her feelings, she is imprecise in her use of language. As Frankfurt puts it, “she does not presume that she knows the truth, and therefore she cannot be deliberately promulgating a proposition that she presumes to be false: Her statement is grounded neither in a belief that it is true nor, as a lie must be, in a belief that it is not true.”[5] This example also meets Frankfurt’s second requirement that the bullshitter lacks the desire to describe reality correctly. After all, it is not literally correct that Pascal feels the same as a dog hit by a car, but she does not seem to be concerned by her use of imprecise similes.

This example is useful in highlighting the differences Frankfurt sees between lying and bullshitting. For him, lying requires knowing the truth and hiding it; bullshit merely requires fakery or phoniness, but not necessarily falsehood. It is, in fact, perfectly possible that a bullshit assertion turns out to be correct for entirely fortuitous reasons. Importantly, liars and bullshitters both adopt deceptive practices, but the aim of the deception differs between the two. A liar seeks to deceive people about an objective truth; a bullshitter seeks to deceive people about his knowledge, making them think he has certain knowledge or beliefs that he does not.

This distinction is compelling and works to significantly clarify the difference between lies and bullshit. Yet, it seems clear that Frankfurt’s conditions for bullshit are incomplete. Consider, for example, a hapless Twitter user who stumbles across a bot designed to spread false information and polarize people politically. Let us assume that this Twitter user is an older individual not fully comfortable with social media and easily misled by clever troll programs. This individual will very likely be duped by the disinformation and truly believe the lies propagated by the bot. However, this unwitting person is a politics junkie who cares deeply about the political opinions she holds; she is not indifferent to the truth as Pascal is. This victim of fake news begins to go around making claims based on the false information that are utter nonsense and, by most definitions, indistinguishable from bullshit. Indeed, it would be impossible for someone listening to her to differentiate her bullshit, characterized by a lack of knowledge of the truth but a desire to describe reality correctly, with Frankfurtian bullshit, characterized by a lack of knowledge of the truth and no desire to describe reality correctly.

To clarify the difference, it is worth looking at one more example. Consider a conspiracy theorist who became convinced through exhaustive, though ineffective, research that human civilization would end in 2012. With hindsight, we can say that this belief was false, and thus the conspiracy theorist lacked knowledge of the truth. Therefore, Frankfurt’s first criterion is satisfied. But given that the individual dedicated a significant amount of time to researching things like the Mayan calendar and other pieces of “evidence,” it cannot be said that he did not care about accurately describing reality correctly. Thus, this individual would not have been propagating bullshit in the Frankfurtian sense. It is nonetheless hard to argue that he would not have been spreading bullshit, as he would have been making grandiose and false claims based in no small part on his lack of knowledge. If someone heard his views, it is in fact probable that he would exclaim “I’m calling BS!”

Of course, one might object that my counterexamples are not real instances of bullshit. After all, the people in them do not seek to misrepresent their knowledge in the same way as, for example, a student who seeks a good grade despite not having done the reading, a bona fide Frankfurtian bullshitter. But this criticism falls flat for two reasons. First, one cannot determine another’s interest in accurately representing the truth. It is simply impossible to know if the 2012 conspiracy theorist genuinely believes his views or is merely trying to sound smart by discussing an idea he understands only tangentially. From the perspective of a hearer of bullshit, therefore, identifying whether the bullshitter cares about accurately portraying the truth is academic. It is a distinction without a difference. Second, both individuals in the counterexamples are, in fact, being deceptive: By promoting their beliefs in an authoritative way, they are attempting to suggest they are credible and have good reason for their beliefs when in fact they do not. In this sense, they aren’t that different from a student bullshitting an essay prompt.

So how might Frankfurt improve his definition to account for the two counterexamples? One approach would simply be to tighten the definition by describing bullshit solely as assertions made without knowledge. This seems promising at first glance, but it falls apart under further scrutiny. For example, a policy advisor might lack sufficient knowledge to advise the president but still be forced to make a call on which option to take. Much like the poor Twitter user described above, she would care about describing reality correctly in order to make the right call, and she would also lack access to the truth, just like the Twitter user. But it seems implausible to consider her a bullshitter. She is aware of her limited understanding and is trying to make the best of a bad situation, not spread bullshit, deceive the president, and falsely portray herself as credible. Indeed, it is likely that she would phrase her recommendation to the president in probabilistic terms, directly flagging her uncertainty. This is something a bullshitter would never do.

An alternative, and better, approach would be to modify the second condition. Instead of saying that bullshit requires a lack of care for describing reality accurately, one could instead say that it requires speaking authoritatively without cause. This would capture every instance of Frankfurtian bullshit while, at the same time, accounting for instances of inadvertent bullshit. Pascal speaks as though she is an authority on the experience of run-over dogs; the conspiracy theorist speaks with authority about the apocalypse. One does not care about accuracy, the other does. Both are bullshitters. Importantly, this new condition would exclude people like the policy advisor, as they recognize and note their insufficient access to the truth. The new condition would also fulfill the purpose of Frankfurt’s requirement that, in order to not be a bullshitter, one must care about portraying reality accurately. This is because it would exclude people who, due to their concern for accuracy, caveat their assertions with admissions of uncertainty.

Frankfurt makes a valuable distinction between lies and bullshit. Yet, by focusing only on the process of bullshitting at the expense of the product, he awkwardly excludes instances of inadvertent bullshit. It is simply not the case that all bullshitters know that they lack access to the truth; many people spread utter nonsense while being thoroughly convinced that their claims are truthful and fair. Happily, this deficiency can be overcome by modifying the second criterion identified by Frankfurt for determining instances of bullshit. It is not a question of whether someone cares about accurately describing reality that is important. Instead, it is a question of whether they are willing to speak authoritatively, and portray themselves as credible experts, on topics about which they lack knowledge. By embracing this new criterion, one is able to capture all instances of Frankfurtian bullshit while also accounting for inadvertent bullshit.



[1] H.G. Frankfurt, “On Bullshit,” Raritan 6(1986), 89.

[2] Ibid., 97.

[3] Ibid., 88.

[4] Ibid.

[5] Frankfurt, 90.


Frankfurt, H. G. “On Bullshit.” Raritan 6(1986): 81–100.