The Weakness of Anthropic Arguments
ineffectivetheory.com
ineffectivetheory.com
These are the questions where “Because those are the only conditions suitable for your existence to happen” is so obviously correct the questions make you roll your eyes. Applying the anthropic principle to fundamental constants is a speculative extension that might be right or wrong, but there are plenty of mundane coincidences that have no other answer.
Add in the "need" for observers in Wheeler's Participatory Anthropic Principle[3], I would say that some extraordinary minds have recognized and proposed some strong anthropic arguments.
In any case, the older I get, the more I realize I don't know a damn thing. As a poet once sang:
Now, I wish I could write you a melody so plain
That could hold you, dear lady, from going insane
That could ease you and cool you and cease the pain
Of your useless and pointless knowledge
[1] https://www.youtube.com/watch?v=2cT4zZIHR3s&fbclid
That is not because just about everyone is too dumb to realize that it stands as a true statement about these circumstances; it is because we have better answers - answers that explain not only why we could not exist in these alternative scenarios, but which also give a coherent explanation of how the actual circumstances came to be... at least until we extend those explanations to the point where we run into facts such as the fine structure constant needing to be just about exactly what it is for any of this to work. The anthropic principle is only invoked where we do not have anything better in way of explanation.
The anthropic principle is arguably an observation that things are the way they are, rather than an explanation of why they are that way.
That's a different question than you initially posted. The anserr is trivial, so nobody would ask this question, except in the futuristic sense, "what's stopping me", or the naive sense, "why is the sky blue, who am I, and why does it hurt when my teeth grow".
The initial question entsils a lot more, depending on how far you want to take it, at least, "why am I", but I would argue that independent of your or my own endeavour, it is ultimately deep. You can try to answer the "how", but not the "why" without circular reasoning.
Circular reasoning (which is tautologic, but not necessarily a formal fallacy) is rather human indeed, insofar we ascribe reason ability exclusively to humans--because we are egoists, since we are limited by our event horizont. So in contrast to assuming the anthropic principle as an answer, I'd rather hold it with Plato.
tl;dr: Here's the fallacy. "Why am I"--"because I am not not" is an inadmissable argument, if nonexistence (not being) cannot be the cause for anything. It's nothing, pretty much by definition. Coralary: If you take the short cut and reduce the double negation, "Why am I"--"Because I am" is a step back from descartes, "I think therefore I am", who managed at least to include a necessary argument for the definition of being human. Although, this only answers the how.
Beyond that, I am only able to argue on linguistical semantic grounds. If "why" is as good as "what for", while "to be for something" means preference, the real content of the question becomes apparent. What am I to do, write more philosophical absurdity, get a cheese cake, or try to have my cake and eat it too?
It isn’t, though. I gave three questions as examples, and this was the first one. I added “intergalactic” as shorthand for “not near any star”, but it’s the same question. I don’t understand what contrast you’re making between the two.
Putting aside a) the fact that there are a few people who suspect that life did (or could) start in the emptiness of space, b) the possibility of Boltzmann Brains, and c) that in your original post, you claimed that "there are plenty of mundane coincidences that have no other answer [than the anthropic principle]" [my emphasis], there are better arguments for why we do not find ourselves in the emptiness of intergalactic space, based on what we know of the complexity of both our bodies and of the biological environment they require to sustain them, what it took for all this to evolve, and the thermodynamic improbability of such assemblages forming in intergalactic space.
This is a better answer because, instead of taking “life can’t form in the emptiness of space” as an axiom, it explains why in terms of more general principles.
In fact, without some sort of explanation, your claim that “life can’t form in the emptiness of space” would be a speculative non-answer (ironically, the rest of your comment here is a demonstration of what I am saying: you are using these better explanations to justify this claim!)
Similarly, the precise value of the fine structure constant (and even that there is such a thing) would not even be recognized as an issue if it were not for our understanding of how much is dependent on it. Clearly, we did not, and could not have, reached that level of understanding by invoking the anthropic principle in response to all the "why is...?" questions that came up along the way (for example, "Why do the chemical elements have the properties that allow life to form from them?" "Because if they did not, we would not be here.")
It is only when we reach the end of our current ability to explain that the anthropic principle is invoked.
My point is that however specific your arguments are for exactly why life forming in intergalactic void is impossible (or “vastly unlikely” if you want to be pedantic), if you reject the notion that you should only expect to find yourself in places where life is possible (or “not vastly unlikely”), what does it matter? By appealing to the particular reasons why intergalactic space is inhospitable to life, you’re implicitly assuming that the plausibility of life is somehow pertinent to what conditions we should expect to observe in our local environment, given that we exist. That assumption is the anthropic principle.
I am not claiming you said that, I am pointing out that there are better answers (more fundamental, more specific) than “life can’t form in the emptiness of space” to the question “why am I not in intergalactic space right now?”, which is what you asked for in the post I was replying to.
Note that when people begin a sentence "Putting aside..." they do so precisely because they do not think the things being put aside are relevant to the discussion.
> By appealing to the particular reasons why intergalactic space is inhospitable to life, you’re implicitly assuming that the plausibility of life is somehow pertinent to what conditions we should expect to observe in our local environment, given that we exist. That assumption is the anthropic principle.
This is, to put it mildly, a highly idiosyncratic definition of the anthropic principle, which does not seem to accord with the usual definitions of either the strong or weak anthropic principle, as given, for example, in Wikipedia [1].
Both versions of the anthropic principle are explicitly cosmological ones, and you appear to be trying to extend it to cover more mundane issues. The problem with your attempt to do so is that, as discussed previously at some length, there are better explanations for the cases you have given, that do not depend on the anthropic principle. As an additional premise, your extended AP is unnecessary; these better explanations have no need of it, while still yeilding more fundamental and specific explanations than your extended AP can. As an alternative explanation, your extended AP is inferior, as you demonstrated when you resorted to the better explanations in an attempt to justify your use of it!
Furthermore, several of your examples are trivial, in that no-one seriously thinks that the scenarios you present are feasible; invoking a somewhat epistemologically-suspect cosmological principle to tackle them is like using a sledgehammer (with a loose head!) to crack a nut.
Your cases may be consistent with your extended AP, but that does not make it the best explanation of them, let alone that they have no other answer, as you claimed in your original post.
But we can't observe parts of the multiverse with different fine structure constants, which puts it in a different category of entities that Occam's razor recommends not to multiply.
More speculatively, as we’ve gathered more observation, we’ve repeatedly found a world consistent with the theory, “Things could be different; your local conditions are unrepresentative; there was a de facto parameter sweep and you’re in an odd corner of it.” Humans are in a very specific place with respect to their surroundings on many, many scales. Admittedly, though, not on the very largest scales we can observe. The cosmos does appear homogenous eventually. Anthropic cosmology is tantamount to the claim that this homogeneity isn’t the end of the story, and that on some “larger” scale (which might not correspond literally to spatial position), the non-homogeneity resumes and explains why our region is the way it is. If there were some a priori theoretical reason to suppose that constants actually do vary, just so slowly that they appear constant within any given Hubble volume, it would add credence to the explanation that we’re just once again in an odd part of a parameter sweep.
For example within a certain range of values Newton's law of motion hold but outside of that they don't. We can simulate outside of that range as if they do but luckily can observe that those simulations don't hold experimentally.
The ultimate experiment would be to start simulating at the big bang with slightly different parameters and see if intelligent life evolves on one of the simulated planets after several billion years of simulated time. But we're very far from such an experiment being feasible.
We shouldn’t confuse the map for the terrain in other words.
Axioms aren't helpful in that regard.
Yes - it's called mathematical statistics.
The difficult case is the one where direct observation isn't possible. The neutron star example is good: they exist in our universe, but who has strong evidence or argument that there's no exotic life in/on the crust?
I think the solution to the Fermi paradox is that intelligent life is just spectacularly unlikely, and that we're the only intelligent life in the observable universe. I think most people reason that we're probably not that special, that our conditions aren't too rare, and that the likelihood of intelligent life given those conditions must be somewhat good, but I don't think any of this reasoning is valid given the anthropic principle. I think the fact that we find ourselves in a world where we're alone is evidence for the sheer unlikelihood of life.
If there were many values of the fine structure constant that could create a universe with any life at all, then I would guess that across that range of values, there would be more variation where some values would create a universe with rare lonely life and some values would create a universe where life was more plentiful. In that case, we would be more likely to find ourselves in a universe where life was more plentiful. But if there were a tighter range of possible constant values that create a universe with any life at all, then I would expect less variation with only universes with rare lonely life. Given that we find ourselves in a universe with rare lonely life, that seems like evidence that there isn't much room for variation in the constants while staying life-supporting.
And that concludes the circle in your circular argument in support of the antrhopic principle :-)
(And then in the last paragraph, I wasn't trying to argue for the anthropic principle, but to argue that given the anthropic principle and the observation we're alone, then I think it's correct to believe that the specific values of physical constants that we observe are likely to be necessary for intelligent life.)
> Loads of systems exhibit Turing universality, so a small change in alpha shouldn't break that property of our universe.
(I consider this relatively indefensible, and the weakest part of my argument.) The other is
> Turing universality => intelligent life is inevitable, or at least likely.
I think that this second piece, although not discussed explicitly in the post, can be defended. It's not a highbrow argument, just that often-made observation that, if a "computer" requires finite size and finite energy to create, then somewhere in the universe, a computer will come into being by random chance.
It's possible that life is just likely enough that none of them have yet chosen to or have been alive long enough to create self-replicating probes, but I don't think it's correct to assume that the likelihood value for life would be on the extreme upper end of the range that's compatible with our observations.
Both values of α look identical to me, but the text implies that they're subtly different - is this a typo in the article, or am I missing something?
The author intents to make a point, this values, which approximate to the same degree of x decimals, are supposed to be explained by "if the first wasn't true we wouldn't be here to ask this question".
From the linked article: "Suppose the fine structure constant were in fact 10% different — enough to prevent stellar fusion from producing carbon."
This is an old idea and is probably false (I am not an astro/nuclear physicist) but the author of this in-depth paper is[1]. Fred Adams is at the Physics and Astronomy departments at the University of Michigan and in this 2019 paper[1], with over 500 references, states the fine structure constant can vary by 4 decades/orders of magnitude (decade=log10(max value/min value)). See table two on page 140 for that range and ranges of some other fundamental values. The most constrained value ranges still vary by an order of magnitude (down quark mass and the up-down quark mass difference).
* The terms life and intelligence lack formal definitions.
* The predicate why lacks a formal definition.
* The argument does not address the maximum number of possible configurations of the universe, which dilutes any quantitative arguments. Most importantly, the cardinality of the maximum number of possible configurations is not addressed. For example, the possibility of an aleph number [0] was not addressed.
* The argument does not address qualitatively different structures of the universe, such as universes with discontinuous lengths or types of affine spaces [1] where measurement as we understand it does not exist.
The list of concerns above is not exhaustive, and I don't think addressing them individually will necessarily salvage the argument. While stronger arguments for or against the anthropic principle are possible, I think they deserve much more rigor.
The base probability is, as always, based on some form of Solomonoff induction, which is invariant to ‘scale’, hence why we can explain away seemingly-arbitrarily low probability events (α≈0.007297352569) by talking about universes that contain a continuum of different α. In the lottery example, the hypothesis ‘lots of people tried’ is simple, even if the probability of each individual winning the lottery is low.
You might ask instead, but what's the probability that that specific person won the lottery? Surely that is exponentially less likely? Yes, indeed, but unlike the Derren Brown case, where P(Derren Brown cheats the lottery) is fairly high prior to him winning the lottery given the public attempt, P(that random guy cheats the lottery) is very low, as most people don't cheat the lottery.
Thinking about alternative observers and question-asking is a distraction. Other people asking other questions are not you, so they are not compatible with your observations.
It is a fair point if nobody has proven (or indeed, can prove) that something in physics is a random number. But alas that does not mean the theory is weak. It is perfectly consistent with the evidence, and mathematically satisfying.
It would be quite exciting if someone can find a reason why we observe specific values and rule out the random number idea. But at some point we know we'll run up against the impossible questions of "why does anything exist at all?", "is it possible for something to not exist?" and "what separates existence from non-existence?". We can't expect to find causal answers for everything.
That's a bloody big if, but let us let it ride for now
> we should be able to find the best(whatever definitions is used)
Agreed it hinges on that, assuming the preceding if is accepted. What 'best' do you nominate? I have no idea what it might be.
I don't nominate any best. I don't have an idea either. I merely used a word used by a parent. Best could be minima, maxima or any other criteria of the function they choose to refer here.
My point is if you have a belief that your model is good enough, you don't need a large number of trials to calculate a probability on whatever characteristic you choose.
The accuracy of this result is a question but the question would remain similar to the scenario where you calculate the probability from a larger number of trials.
1. This place is perfect for us
2. Why did we end up in a place that's perfect for us?
3. Perfect places for us are possible
4. We could only evolve in one of those perfect places
Number three is the (often unstated) assumption. The anthropic principle doesn't (even attempt to) answer why perfect places are possible; it merely states that because we're in one, they are possible.
The article seems to just be complaining that anthropic arguments cannot explain the precise _value_ of the fine structure constant. Well, nor can Occam's Razor, but we don't seem to have a problem quoting that!
We wouldn't be debating it if the universe were substantially different, but that doesn't explain why the universe we observe is the way it is.
In other words, for some reason they expect the universe to be more "understandable" or "sensible" than this . . .
Of course the strong AP is ludicrous in the extreme ;)
There's no number system you can invent where the fundamental constants turn into simple integers or have obvious ratios.
So why do they have those values? What gave them those values? The anthropic principle is disliked because it rejects further inquiry and progress.
Unfortunately, that doesn't explain how these values are generated. That is the important question.
No, their opposition is to the fact that a physical constant is a mere accident, a mere facet coming with the universe as a set, and not explained by some more fundamental mechanism.
And anthropic arguments are like: "if they weren't apples, then they wouldn't be apples..."
Well, Occam's Razor is just a heuristic, not a logical necessity.
Sometimes the more complex answer is actually the correct.
Occam's Razor is not an argument or explanation. It's a heuristic for discriminating among different arguments/explanations.
And anthropic arguments amount to "there isn't any explanation for the precise value of the fine structure constant!" Or "you should just give up trying to explain why it has the value it has!"
I don't actually think the anthropic principle is capable of tackling that one.
( Original intention of the backstory of https://andromeda.fandom.com/wiki/Trance_Gemini )
(Yeah, not quite anthropic principle, but definitely popped into my mind as something to give a 2-year-old)
If you use the cellular automata analogy that is being alluded to, ‘intelligence’ is being essentially equated to some limit of ‘complexity’
This is all very hand-wavey so it doesn’t appear to add much to the discussion?
That doesn't mean AP is wrong, it is a tautology, correct by definition. Neither is AP useless. It is the correct counter to teleological arguments such as fine-tuning. However, treating it as a scientific explanation is not correct.
That being said, that confusion is totally understandable. The current state of anthropic reasoning is quite messy. There are contending theories on many related paradoxes. The sleeping beauty problem, the doomsday argument, the fine-tuned universe, the simulation argument, etc. Like before, I find these paradoxes are caused by ignoring the role of perspectives as well. I explained my solution to the related debate on my website https://www.sleepingbeautyproblem.com if anyone is interested.
[0] https://en.wikipedia.org/wiki/Linguistic_relativity
[1] https://plato.stanford.edu/entries/linguistics/#SoCalSapWhoH...
[2] https://www.sciencedirect.com/science/article/pii/B008043076... , https://doi.org/10.1016/B0-08-043076-7/03042-4
Here’s an interesting explanation from Socrates:” Socrates I will tell you. The name “man” (ἄνθρωπος) indicates that the other animals do not examine, or consider, or look up at (ἀναθρεῖ) any of the things that they see, but man has no sooner seen—that is, ὄπωπε—than he looks up at and considers that which he has seen. Therefore of all the animals man alone is rightly called man (ἄνθρωπος), because he looks up at (ἀναθρεῖ) what he has seen (ὄπωπε).”
http://perseus.uchicago.edu/perseus-cgi/citequery3.pl?dbname...
And yet the English “man”, Spanish “hombre”, Russian Мужик mean the same yet different thing, namely a (working) hand. As in “how many hands do you have in your factory?”
We are yet to find another primitive carbon based life form. Anthropic principle works just fine with Occam's razor.
I mean, it doesn't have to bring in "we wouldn't be here" questions of survival, it seems to be just of the form of...
(1) Question whether A is true. (2) If A were not true, then it would necessitate that B is true. (3) B is clearly impossible. (4) Therefore... A must be true.
It's useless for statements about the physical universe.
Edit: IIRC the form is
!(!A || B) && !B
(A || !B) && !B
(A && !B) && (!B && !B)
A && !B && !B
A && true && true
A> The universe is the way it is because if it weren’t, we wouldn’t be here to discuss the issue.
That can't be right. That statement says we are here as a consequence of the way the universe is. It says nothing about causes (of the universe being the way it is).
What's a better statement of the principle?
Crank the value up to .1 and you get no fusion at all. This one is super-interesting, because no fusion means no photonic pressure keeping silly large accumulations of gases up and so you will crank your way to neutron stars and black holes much faster, as well as resulting in a universe whose only light is the remaining cosmic background radiation. No more supernovae creating shockwaves and the like.
In the other direction and you have a different problem -- molecules won't molecule and you will just have a bunch of disconnected atoms incapable of covalent bonds.
Aside from common sense ("this being negative would be self-contradicting") for a given number and sets of inter-relations, that's out of my wheelhouse. I very dimly recollect some work on allowable sets of constants somewhere in the nineties, something to do with superstrings, but that was never a huge interest of mine. I will admit that such a generating theory may exist and I am just unaware of it.
In some sense, each constant represents a bit of embarrassment to physics. Having this rather large set of adjustable parameters calls to mind the von Neumann quote "With four parameters I can fit an elephant, and with five I can make him wiggle his trunk."
Why this seemingly arbitrary number? Can we take multiple constants and find some smaller set from which they are derived? Can I make a constant emerge from a physical law, devoid of measurement?
This puts all realms of the discussion in the metaphysical, which is fine, but it should be noted :)
Anthropic arguments seem to provide a reason to be uninterested in a large class of apparent surprises. Should we invest in investigating those surprises, or not? The discussion may be metaphysical, but the decision has real-world consequences.
I am by no means trying to downplay metaphysical discussions, most lived human experience is in that realm (“what should I wear/eat/do/etc.”), so one could argue it is more important than many scientific/engineering endeavors looking to find new ways to serve advertisements :)
I love the discussion of complexity and automata, and I struggle to think how our limited minds will ever work around the problem of unbounded complexity...