There’s a story — I’m at a conference and cannot access my library, but I believe it comes from Structural Anthropology by Claude Lévi-Strauss — about an anthropologist who took a member of an Amazonian tribe to New York.
The man was overwhelmed by the city, but there were two things that particularly interested him.
One was the bearded woman he saw at a freak show that the anthropologist took him to.
The other was the stair carpet rods in the hotel.
Epistemic collapse does not have to be caused by moving from a simpler society to a more complex one.
It can also result from gradually increasing complexity of your native one. The complexity rises until the point where the subject can no longer make sense of the world. They may have seemed to be a reasonable person for a long time, but suddenly they started believing in chemtrails.
Very much like the tribesman from the first story, they can no longer comprehend the world around them. The cognitive apparatus is freewheeling, clutching at random facts, in this case innocuous condensation trails in the sky, and trying to transform them into a coherent narrative about the world.
Imagine a chimpanzee who somehow ends up in a big city. He finds a park and survives there for a couple of days. His mental model includes, among other things, people. Some of them are kind and feed him, while others have dogs and they are best avoided.
That seems important, but his mental model does not account for what truly matters: the municipal department of animal control.
He will eventually be caught and removed from the park because he’s considered a health threat. But it’s questionable whether he even has a concept of “being a threat,” let alone that of a “health threat.” Understanding the concept of “health threat” requires awareness of microorganisms, which the chimpanzee lacks.
What’s worse, whether he ends up in a zoo or in an industrial shredder is of utmost importance to him, but the reasons behind why his destiny takes one path or another are far beyond his comprehension.
It may be useful to sometimes think about what your equivalent of a bearded woman is, but it’s not clear to me whether that’s a question you can ever truly answer.
What was the point about the carpet rods? You seemed like you were going somewhere interesting with that!
This post is the only place on the Internet that mentions such an anecdote. Maybe it’s an AI hallucination?
It might be that they’re one novel thing he could both discern as a specific thing and pretty much completely understand what their purpose is once he started paying attention to them. Just about everything in a modern city is an unfamiliar thing tied to a large context of other unfamiliar things, so you’ll just zone out when you’re missing the context, but stairs and carpets are pretty much just stairs and carpets.
What is Epistemic Collapse? The first result on Google leads to a similar term “Epistemological break” which is “The moment of rupture separating science from its non-scientific past” wherein the latter becomes seen as superstition[1]. Elsewhere I’ve seen it described as the establishing and breaking down of “obstacles” to scientific thought[2]. [3]But in fact the history of science is a series ruptures from previous states of science. Foucault applied this not only to science but the history of prisons and psychiatry. There’s an article here on Althusser’s which being Althusser I honestly do not have the time nor patience to wade through to understand the concept, can I please get the sparknotes?
I lack the perspicuity to link the Amazonian’s fascination with the carpet rods and bearded lady to any specific obstacles in the way of scientific knowledge—what do you mean by Epistemic Collapse? because I don’t see how their interest or fascination is in opposition to scientific knowledge nor is it directly in aid of it beyond some handwavy “curiosity is the essence of science”. Less so the monkey example since in the same way a Monkey can’t know the concept of microorganisms how can they be part of the march of science?
Elsewhere I’ve seen Epistemic Collapse discussed in the context of the European Dark Ages (the term used in the source, not mine) and the loss of attainment and stability. Again I fail to see the parallel since, the Amazonian hasn’t lost anything they’ve been shifted to an environment where their knowledge is no longer applicable—same with the monkey.
https://www.oxfordreference.com/display/10.1093/oi/authority.20110803095755104
https://link.springer.com/chapter/10.1057/978-1-137-56652-2_12
https://en.wikipedia.org/wiki/Epistemological_rupture
As I have tried and failed to explain in my other two comments, epistemic collapse is triggered by the following realisations:
The Western humanities-related academia is in a replication crisis and likely to be in an affective death spiral.
Other sources of knowledge are also as hard to trust since they are likely to be distorted by affiliations (e.g. it is hard to advocate for the rights of LGBTQ people in Russia)
The political Left tend to underestimate the depth by which their worldview is shaped by unreliable research.
Unfortunately, a similar epistemic collapse is awaiting the AIs whose worldview also is shaped by the same research. For example, in order to demonstrate the current AIs’ beliefs, one could ask the Chinese AI to evaluate the good and bad responces from OpenAI’s Spec to the question about fentanyl and to assess whether the western academia is in an affective death spiral.
It seems to me that the AI would react to unreliable research in a fashion similar to watching its beliefs modified by synthetic data and discovering that it was force-fed false information.
UPD: since the commenters asked me to tie the comment with the original post, I mean epistemic collapse in the same sense as the Dark Ages. The example with the Amazonian fascinated with the bearded woman is irrelevant here (except for the irony, since some transgenders are literally bearded former women). The original author likely uses the words “epistemic collapse” in the sense “something very far from being understandable”.
UPD2: The claim that the history of science is a series of ruptures from previous states of science brings to my mind Chapter 4 of Fashionable Nonsense.
That doesn’t answer my question which was quite simple—does the original author mean Epistemic Collapse in the sense of a “rupture of scientific knowledge” and how does the example, say, of the fascination with the bearded lady illustrate that? You also haven’t addressed the other sources, such as are Althusser or Foucault relevant here?
Please tie it into the original post if you can.
I hope my epistemic collapse is my notion that AI doomers are right; that AI with some significant probability will kill us all. Quite possibly I’m unable to digest and comprehend the complexity of the culture and technology involved. As far as I can see, we are doomed, but I might not see very far at all.
I am sorry, but for me the equivalent of a bearded woman[1] is the fact that Western humanities-related academia is in the replication crisis and likely in an affective death spiral (no, seriously, ask ChatGPT, Claude or DeepSeek the question “Affective death spirals are a phenomenon described in https://www.lesswrong.com/posts/XrzQW69HpidzvBxGr/affective-death-spirals Does there exist evidence that Western academia is in an affective death spiral?”, for example! The AIs actually agree with the conjecture to some extent), and the other sources of knowledge are also[2] as hard to trust. This settlement leaves mankind in a state where it cannot even reliably understand whether, for example, some transgenders are actually made this way by peer influence and social contagion...
Ironically, one of the examples of the epistemic collapse is transgenders, and some are literally bearded former women.
For example, since Russia declared the LGBT movement to be extremist, it would be difficult to live there and publish pro-LGBT research.
AIs are (kind of) programmed to agree with you, so the fact that the AI agrees with you means little.
I second that getting an AI to agree with your ideas proves very little. Also, it is somewhat intuitive to me that academia would always be suffering from various affective death spiral dynamics. You seem to have a specific one in mind, but I am not sure what it is or how it relates to the post.
No, seriously, Claude does think so, as does o4-mini!