For example, describing the cause for life as aliens seeding it fails to account for the the cause of the aliens that did the seeding.

The explanation that something is "art" because it's like other things that are "art," fails to account for how/why those other things are "art."

Another example is "it's turtles all the way down," wherein the claim that "it's resting on a turtle" requires the very same kind of explanation as it purports to offer.

The specific kind of mistake I'm thinking of is claiming that the given answer IS the root cause, despite it demanding exactly the same type of explanation that it claims to offer.

What is the name for this mistake or fallacy? Is this a "Homunculus Fallacy?"

  • 3
    This is not a mistake or fallacy, it is inevitable, any explanation fails to account for something in it. After any because there is always another why, as parents find out from their kids. It is called infinite regress. The same happens with justifications, see Agrippa's trilemma. The standard is not whether an explanation fails to account for something, but rather whether it makes a substantive contribution to understanding what was asked about. If aliens did, indeed, seed life, establishing that would make such a contribution.
    – Conifold
    Feb 10, 2023 at 20:55
  • True. But I don't mean any given failure of explanation, but specifically entirely shifting the original question to a new target, without answering anything. The value of knowing that aliens seeded life on Earth notwithstanding, it would merely shift the question of "how did life begin" from us to aliens. The art example is a better one. Offering the explanation that "art is art because it's like something else that's art" fails to answer much of anything, and merely shifts the original question to a new target. It's like a tautology, but not circular.
    – Matt
    Feb 10, 2023 at 21:15
  • 1
    Your art example might fall under the homunculus fallacy, "whereby a concept is explained in terms of the concept itself, recursively, without first defining or explaining the original concept". However, the "deeper explanation" may be lacking because none exists. Many concepts are formed by family resemblance, overlapping similarities with no one feature in common. Giving prototypical examples that it is like is just the explanation called for. It is common for emerging/vague concepts.
    – Conifold
    Feb 10, 2023 at 22:42
  • With aliens no real knowledge exists so it’s an enigma without right or wrong answer. With art it’s a big area which is too general and therefore not wrong. A person, is a teacher, a person is a student.
    – estinamir
    Feb 11, 2023 at 4:55
  • @Conifold's answer is probably the most relevant. I'm not interested in claims specifically about aliens or art. I'm interested in the nature of making a claim (as though it were final) that requires the very same explanation that it purports to offer. Not exactly circular, but unendingly linear.
    – Matt
    Feb 11, 2023 at 15:53

3 Answers 3


Homunculus argument/fallacy it is.

So, as per the linked article above, a homunculus fallacy occurs when the explanans (that which does the explaining) is identical to the explanandum (that for which an explanation is sought). So if e.g. I explain rainbows by claiming each droplet of water in the air produces its own tiny rainbow, I've committed the homuculus fallacy as I still need to explain how rainbows form (in the droplets).

  • A precondition for the "homunculus fallacy" to be a fallacy, is a claim that it is NECESSARY for an internal observer to exist for seeing to happen. Few dualists make this claim, instead arguing that the dualist model of an internal observer fits our internal experience and the rest of the data of consciousness. YES, all justifications then require their own justifications, but the structure and content of these other justifications can in principle always be different from the first. Regress is therefore not always a homunculus fallacy, and is not in at least some of the examples asked.
    – Dcleve
    Feb 11, 2023 at 19:28
  • 1
    @Dcleve, the homunculus fallacy, to my reckoning, is to "explain" x with y such that y is just another x.
    – Hudjefa
    Feb 12, 2023 at 7:47
  • 1
    Smith, link-only answers are sub-opimal. Would you mind fleshing out your answer with a description of the Homunculus Fallacy, as it relates to this particular question?
    – Matt
    Feb 12, 2023 at 20:11

I take it from your example that what you really are asking about is explanations that leave themselves open to questions of the same sort as the original explanation.

It is not, of course, a mistake or fallacy for an explanation to fail to account for its own cause in general. Quite the contrary, the vast majority of explanations are like this.

Q1: Why did Jack and Jill go up the hill?
A1: To fetch a pail of water

In the above, A1 is a perfectly valid explanation that reasonably accounts for the data in Q1. It does not however, explain why Jack and Jill wanted a pail of water, nor does it explain why there was water at the top of the hill (an unusual location for a well), nor would it be expected to.

Explanations always lead to further questions. For example,

Q2: Why do the planets orbit sun in roughly elliptical orbits?
A2: Because that's the shape you get from an orbit controlled by an inverse square force like gravity.

This answer obviously suggests additional questions such as why an inverse square force leads to an elliptical orbit and where gravity comes from. But the existence of follow-on questions does not mean that the original explanation is a mistake; it only means that no answer can answer everything.

However, your two examples were a bit different; each is an example of an answer to a question that leads naturally to another question of the same type as the first one.

Q3a: Where does life on earth come from?
A3: Aliens brought it.
Q3b: Where did the alien life come from?

Q4a: Why is that painting art?
A4: Because it it like other paintings that are considered art.
Q4b: Why are those other paintings art?

However, this is also not, in general, a mistake or fallacy, because there is no reason to expect that an answer to a question about one specific instance would apply to all instances.

Q5a: Why did domino A fall over?
A5a: Because domino B fell over and pushed it.
Q5b: Why did domino B fall over?
A5b: Because domino C fell over an pushed it.
Q5c: Why did domino C fall over?
A5c: Because I pushed it.

In this example, A5a is a correct answer even though it leads to more question. Whether it is a proper answer depends on extra-logical features of language and social interaction. That is, the intention of asking Q5a may not be satisfied by answer A5a, because the person asking the question already knows the answer A5a and is really asking why all of his hard work setting up dominos has come to naught. In such a case, the questioner is really asking:

A5x: Why is my domino setup all fallen over?
A5x: Because Joe pushed over the first domino after I told him not to.

I suspect your dissatisfaction with A3 and A4 is that you think the conversational implicature demands a more comprehensive answer than the one you received. In other words, you thought it was apparent from the conversation that you weren't merely asking Q3a and Q4a; you were asking more generally:

Q3x: What is the ultimate origin of life on earth?
Q4x: What is the ultimate feature that distinguishes art from non-art?

However, in cases like this, your demand for a more comprehensive answer may be unreasonable. The suggestion that life came from aliens, for example is a response to the difficulties of explaining how organic life could arise from non-living chemicals on earth. It is reasonable to speculate that organic life appeared elsewhere where the chemical conditions were more suitable or that there are other kinds of life, not based on organic chemistry, which might be more likely to arise spontaneously (it is reasonable so long as it is only speculation).

In cases like this, a partial answer might be useful and suitable even if it does not answer the ultimate question because it can lead to research opportunities.

In the case of the art example, it is reasonable to suggest that what is considered art today is based on nothing more than the history of art in the past. That is a valuable insight even if it does not answer the detailed historic question of how all art originally attained that status. The person who offers A4 as an answer is providing a literal answer to Q4a and a partial answer to Q4x, which may be perfectly reasonable if the ultimate answer to Q4x is unknowable or is beyond the scope of the conversation.

  • This is a good discussion of how an unsatisfying answer can nonetheless be correct, or how coincident causes can be similar... until we find the ROOT cause.
    – Matt
    Feb 11, 2023 at 15:49
  • The mistake I'm trying to get at is the mistake of claiming that an answer is final, when in reality it merely shifts the very same question to its own claim. Like the mythological claim about the orbit of the earth. The answer "it's resting on a turtle" is only technically an answer, but it insists that I ask the VERY SAME question about its claim. More to the point, the ostensible final answer about it being "turtles all the way down" merely prompts me to ask the very same question again. There is no claim of a "Joe" as a ROOT cause.
    – Matt
    Feb 11, 2023 at 15:50

This is not a "mistake" it is instead a feature of our world and what we have to live with. It is impossible, in principle, to satisfy either the Justified True Belief, or The Principle of Sufficient Reason. The general problem is called the Munchausen Trilemma. All justifications must terminate either un unjustified claims, infinite regress, or circularity. The name "Munchausen" is a reference to a German fairy tale in which Baron von Munchausen while riding his horse got stuck in a bog. The horse was too heavy to lift directly, but by pulling on his own pigtail, and then lifting the horse thru his stirrups, he was able to lift both out of the bog! The name ridicules the "web of belief" that some philosophers have favored as a solution, by pointing out that circular logic is still wrong even when the web is made large/complex.

Here is a discussion of the Trilemma: Is the Münchhausen trilemma really a trilemma?

Yes, claims that one has done either JTB or PSR are false, and may involve any of a variety of fallacies in their rationalization.

  • This Munchausen Trilemma also seems like a good candidate for the type of thing I'm trying to name...
    – Matt
    Feb 12, 2023 at 20:13

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .