Fabric of Reality Book Club
We discuss David Deutsch’s first book, The Fabric of Reality.
Log in or sign up to participate in this discussion.
With an account, you can revise, criticize, and comment on ideas, and submit new ideas.Can't think of how it could be otherwise. Do you have any examples of inexplicit explanations?
Let's fuck with your intuitions a little bit:
Say "stop" when it's no longer an explanation:
Didactic chapter in plain English with examples and edge cases, distilled into a concise technical note with formal definitions, invariants, and pseudocode.
Literate program interleaving prose and code, or a heavily commented Python implementation with docstrings and tests.
The same code stripped of comments/tests and then minified or obfuscated (e.g., Python one‑liner, obfuscated C), up through esolangs and formalisms (Brainfuck, untyped lambda calculus with Church numerals, SKI combinators).
Operational specifications with minimal labels (Turing machine tables), then hand‑written assembly without labels and self‑modifying tricks, down to raw machine code bytes/hex and binary blobs with unknown ISA or entry point.
The same bits recast as DNA base mapping with unknown block codec, unknown compression, encrypted archives indistinguishable from noise, arbitrary bitstrings for unspecified UTMs, or physical media (flux/RF) without modulation specs.
Haha not a programmer so understood maybe half of it, but I think I see what you mean. There'll always be inexplicit parts to every explanation. My concept of explanations is that there must be at least some explicit part for it to be called an explanation. That's why genes aren't explanations.
My point is rather that it's not so clean a line between explicit and inexplicit. You're a doctor, so imagine the steps being something like:
- Extensive description of patient's symptoms, test results, conclusion, etc, in English.
- Same as above but mostly made out of quick notes by attending doctors and nurses.
- Only a collection of test names and test results. Test results accompanied by Chinese.
- Just a collection of numbers coming out of tests, without saying which test.
Arguably all the information is always there, and can be read off, but with increasing difficulty, requiring you to learn another language, or do a series of deductions.
Yeah nice, seems true. There's no objective explicit/inexplicit ratio for knowledge, it depends on the person's background knowledge.
The important thing is to be able to make predictions about images on the astronomers’ photographic plates, frequencies of spectral lines, and so on, and it simply doesn’t matter whether we ascribe these predictions to the physical effects of gravitational fields on the motion of planets and photons [as in pre-Einsteinian physics] or to a curvature of space and time.
I’m getting conflicting results online for this quote. Some sources that quote the same passage say singular ‘effect’, others use the plural like Deutsch does.
I don’t have access to the original text, so I can’t say for sure if this is possibly a slight misquote or if different people are just quoting different editions.
A gene doesn’t have problems in any conscious sense, but it always faces the problem of how to spread through the population at the expense of its rivals.
Maybe that answers your question, Erik.
How could we integrate that vision with Popper's definition (paraphrased): a tension, inconsistency, or unmet explanatory demand that arises when a theory clashes with observations, background assumptions, or rival theories, thereby calling for conjectural solutions and critical tests.
The rival theories and clashes sound like competition between genes – or more precisely, between the theories those genes embody.
Basically, genes contain guesses (in a non-subjective sense) for how to spread through the population at the expense of their rivals. Those guesses are met with selection pressure and competition.
Yeah, thanks! Are ideas also guesses of how to survive in the mind and across substrates, or is there more to ideas?
In the neo-Darwinian view, any replicator’s primary ‘concern’ is how to spread through the population at the expense of its rivals. This view is what Dawkins (IIRC) calls the gene’s eye view, and it applies to ideas as much as it does to genes. Any adaptation of any replicator is primarily in service of this concern.
So I think the answer to your question, “Are ideas also guesses of how to survive in the mind and across substrates …?”, is ‘yes’.
… any replicator’s primary ‘concern’ is how to spread through the population at the expense of its rivals.
Why “through the population”? Doesn’t this presuppose a replicator needs to exist within a population to do what it does? The first replicator spread with no population to spread into.
I suppose it’s theoretically possible for the very first replicator to exist in isolation until it replicates for the first time. But that’s what it does right away anyway.
Accounts of the origin of replicators (such as RNA World) involve proto-replicators. By the time the first ‘full-fledged’ replicator came on the scene, it was already part of a larger population of proto-replicators.
… any replicator’s primary ‘concern’ is how to spread through the population at the expense of its rivals.
Why “at the expense of its rivals”? Isn’t the concern to spread at all, regardless of the outcome of rivals?
Rivalry means competition, win/lose outcomes. If one replicator spreads, it will be at the expense of its rivals (if any), eg taking up niches that rivals would otherwise have taken up.
I’m using standard neo-Darwinian phrasing. Compare, for example, BoI chapter 4:
The most general way of stating the central assertion of the neo-Darwinian theory of evolution is that a population of replicators subject to variation (for instance by imperfect copying) will be taken over by those variants that are better than their rivals at causing themselves to be replicated.
And, same chapter:
[T]he knowledge embodied in genes is knowledge of how to get themselves replicated at the expense of their rivals.
See also several instances in chapter 15 in the context of meme evolution.
Richard Dawkins’ The Selfish Gene has a ton on rivals (alleles), too, for example (chapter 2):
Ways of increasing stability and of decreasing rivals’ stability became more elaborate and more efficient. Some of them may even have ‘discovered’ how to break up molecules of rival varieties chemically, and to use the building blocks so released for making their own copies.
Most people (except in Alzheimer's, etc.) don't run out of memory in the brain. The reason most people don’t (permanently) run out memory (of either kind) isn’t that memory isn’t scarce, but that there’s a pruning mechanism in the mind. And there’s competition.
In your revision, you asked me to let you know if you are doing things incorrectly.
You can revise ideas the way you did, it’s not wrong per se, but revisions are better for incremental changes. They’re not really meant for taking back criticisms or indicating agreement. If a criticism of yours is successfully counter-criticized and you would like to abandon it, I would just leave it counter-criticized and not revise it further.
If you are looking for a way to indicate agreement (with a counter-criticism, say), it’s something Dirk and I have been discussing offline, see #2169. I hope to implement something to that effect soon.
#2325 serves as an example. I had submitted a criticism which is now outdated and remains counter-criticized. It’s actually better that way because it shows that an error has been corrected, and makes it less likely for others to submit a duplicate criticism.
Since you’re a doctor, Erik, let me ask: is there a possibility Alzheimer’s could be explained in terms of bad software? Correct me if I’m wrong, but it seems like the prevailing view is limited to bad hardware.
Hmm never thought of that, interesting! I think since the disease involves continuous loss of brain volume, hardware decay seems like the best explanation.
In general I think it makes sense to speak of diseases in neurology (e.g. Alzheimer's, Parkinsons, stroke) as bad hardware and psychiatric disease as bad software. But it could very well be that some of those diagnoses are miscategorised.
Not a doctor. But it's not hard for me to imagine untainted memory but a script with an error such that it can't manage to look up the information.
Yeah that's definitely a possible medical condition, e.g. in psychosis or after having ECT. Don't think it's the best explanation for Alzheimer's though, where the loss of brain volume is so apparent.
Deutsch says our body of knowledge keeps growing both deeper—better explanations—and wider—new fields, more facts, rules of thumb. He thinks depth is winning. It might be interesting to assess what that balance looks like in 2025.
Perhaps it’s premature, but I’d love to discuss:
why DD thinks the four strands already amount to a theory of everything
why DD presents quantum mechanisms as having already subsumed general relativity
what other (proto)strands we could envision and why they are indeed a meaningful addition to the 4 strands
Yeah (3) is interesting. Constructor theory is the contender I can think of for a future fifth strand. Any other suggestions?
Yes, but that inhirent in biology (evolution) right? I see it as part of the evolutionary strand for this reason.
In that same vein, why couldn't we class biology (evolution) under epistemology?
I still see epistemology as distinct, and I'll try to make my case for it. Epistemology explains how humans create explanatory knowledge — unlike biological evolution, which also produces knowledge, but not explanations. Explanatory knowledge is special because it allows us to understand the world. Deutsch even suggests that this kind of knowledge tends toward convergence — a unified theory of everything — implying a deep connection between reality and its capacity to be explained.
Economics, on the other hand, isn’t distinct in the same way. It deals with trade-offs and scarcity — principles already fundamental to biology. Life itself is about managing limited resources and the trade-offs that come with them. Evolution, in turn, discovered increasingly effective strategies for doing so — including cooperation, exchange, and other relationships between and across lifeforms that facilitate these trades.
May have misunderstood, but do you mean that explanatory knowledge corresponds to truth, whereas biological/evolutionary knowledge doesn't?
I think that was refuted by Lucas Smalldon and others: https://barelymorethanatweet.com/
Undestanding does not flow from explanatory knowledge the way you imply. I understand Dutch and English, but a lot of my understanding of it is inexplicit.
While a lot of what’s involved in understanding a language is inexplicit, it is not possible to come to understand a language without ever dealing with it explicitly.
This is part of what separates explanatory knowledge from other types of knowledge.
You say that trade-offs and scarcity are fundamental to biology. I agree, and this implies economics as a more fundamental science than biology or evolution. It still applies in our computer models, where biological details may not.
Guess: We can generalise economics further and let it be subsumed by epistemology.
When we choose to try to solve certain problems, we always make trade-offs from a place of scarcity. Likewise, our conjectures wouldn't evolve without the competition enabled by scarcity in our minds.
Economics is simply at the intersection of evolution and epistemology.
I currently see Constructor Theory as a meta-theory. A different mode of explanation. But it raises an interesting question: does CT actually qualify as a deeper theory than the four strands? Even if we were to express all four strands in constructor-theoretic terms, that alone wouldn’t make it explain more or have greater reach. So when would it truly deserve to be considered a strand/theory of everything?
By the same logic, wouldn't neo-Darwinism also disqualify as a strand, since it's subsumed by Popperian epistemology?
Why does neo-Darwinism qualify as a strand, if it can be understood as a component of Popperian epistemology?
Consequently (they say), whether or not it was ever possible for one person to understand everything that was understood at the time, it is certainly not possible now, and it is becoming less and less possible as our knowledge grows.
If something already isn’t possible, how could it become less possible?
Isn’t possibility a binary thing? As opposed to difficulty, which exists in degrees.
“([T]hey say)” presumably means he is paraphrasing people who get it wrong.
My charitable interpretation:
“Less and less possible” is a loose way of saying something like “more and more difficult to achieve”, or “occurs less and less often in the multiverse”.
That’s fine if you want to interpret it charitably, but that isn’t a criticism. Maybe you’re implying that I’m not being as charitable as I should be. That would be a criticism, but it should be made explicit.