Fine-tuning and intrinsic value

I’ve been thinking a bit about the fine-tuning argument for design, and in particular about this Scientific American piece by Philip Goff. I’ve also been talking to Philip about it. Here are few thoughts arising from the process.

(Note added 8/11/21: Below I talk about fine-tuning as evidence for design — a word which might be taken to imply the existence of a personal creator god of the kind believed in by the Intelligent Design community. Philip has asked me to clarify that he does not align himself with that community and that he prefers to talk of fine-tuning as evidence for ‘teleology’ or ‘goal-directed activity’ in the early universe rather than design. The points made below equally to this weaker claim.)

The fine-tuning argument starts with the uncontroversial claim that if the fundamental physical constants of our universe had been different, then life would have been impossible. There is only a tiny range of possible values within which stable complex physical structures, and hence life, could arise. That is, it looks as if our universe has been designed to support life. More precisely: given the evidence (the actual nature of the constants — call this E), the hypothesis that the universe was designed (D) is more probable than the hypothesis that it arose by chance (C). Schematically, Pr(E/D) > Pr(E/C).

One reply to this is that is there is another, equally probable, hypothesis — namely, that there is a multitude of universes (a ‘multiverse’), each member of which has different constants. Our universe was not specially designed for life; it just happens to be one in which life is possible. It seems special to us because it enabled us to exist. Compare a lottery winner who thinks there must have been something special about their ticket. There wasn’t really anything special about it. It seems special to them because it made them a winner. It doesn’t seem special to anyone else, and our universe doesn’t seem special from other perspectives either.

I think Goff would agree that if there were a multiverse, then the existence of our universe would be no more improbable than if there were a designer. But he argues that the existence of our universe does not make it probable that there is a multiverse. To think it does, he argues, is to commit the reverse gambler’s fallacy (the fallacy that an observed lucky throw of the dice must have been preceded by long series of unobserved unlucky ones). To underline the point, he describes a scenario in which a deranged kidnapper will kill you unless a certain monkey types out an English sentence within a set time (an analogy for the universe having just the right constants for us to exist). Goff points out that if the monkey comes through with the goods and you survive, you won’t be tempted to infer that there are lots of other typing monkeys around, on whose productions the lives of other, less fortunate, victims depend (analogous to a multiverse).

The issues here are tricky, and I don’t claim to be clear about them. But even if Goff is right about the multiverse hypothesis, I don’t think the result is a win for the design hypothesis. For the obvious alternative to the design hypothesis is a simpler one: that our universe arose by chance, with random constants (analogous to the monkey just happening to type English). Our universe had to have some constants, and they just happened to be ones that made life possible. It seems amazingly lucky to us, of course, but that is because we are living beings and wouldn’t be around if the constants had been different. The specialness is in our eyes only — analogous to the specialness of the lottery’s winner’s ticket. Why is this hypothesis less probable than the hypothesis that the universe was designed to support life? Why should we think that a designer tuned the constants to make our existence possible, any more than a lottery winner should think that someone fixed the lottery to make them win?

Now we can easily imagine a case where design would be more probable than chance. Suppose our universe had some feature that made it objectively special, not just special in our eyes. Suppose that, in virtue of having the constants it has, it supported something intrinsically valuable, which a universe designer would plausibly want to promote. Then, other things being equal, design would be more probable than chance. Analogously, if there were something about the lottery winner that made their winning independently significant — say, that they were married to the chief executive of the lottery company — then we would be warranted in suspecting that the lottery had been fixed. Proponents of the fine-tuning argument can claim that this is in fact the case, and that life is the specialness-conferring feature.

In short, the fine-tuning argument requires the assumption that life itself, or the experiences of living beings, is intrinsically valuable. (This is in fact Goff’s view.) This isn’t to say that that the assumption is sufficient for the argument to work — there might be other problems with the argument, particularly with the design hypothesis itself — but it is necessary to get it going.

Should we make the assumption? It is a big issue, but I think not. Of course, life and the experiences of living creatures matter enormously to us, but I don’t see any reason to think they matter in the way required by the fine-tuning argument — that they have an intrinsic value, independent of the interests of living creatures. Indeed, I find it hard to make sense of the claim. As I see it, value and meaning are rooted in the interests of living creatures, even if we sometimes speak of them as intrinsic properties of things. But that’s another story. The moral I want to draw here is that the fine-tuning argument needs the assumption that life is intrinsically valuable.

Posted in Tricks of the mind and tagged , .


  1. I think there’s a lot of nuance hidden behind “intrinsically valuable”. Valuable implies some sort of evaluation function, the default of which is the human race’s definition of valuable. If we are talking about this default, it is a moot point. Of course we think life is valuable. Perhaps we don’t mean valuable but rare, in that it arises in a small percentage of possible, or existing, universes. That seems like a more precise idea but we have no way of saying much more as we have no idea of the parameter space from which universes are drawn or the requirements on those parameters for life to arise.

  2. The argument that life is intrinsically valuable feels to me, for fuzzy reasons, like the Cartesian argument that it’s the nature of the Designer that they would not mislead us. Descartes was making an epistemological argument, not an ontological one, but still…

  3. I think both Philip’s argument against the multi-verse, your own argument against a designer, and the multi-verse argument as a response to fine-tuning are all are beside the point.

    There is … [gestures broadly at all around me]… everything we know of, the Universe.

    And there is physics, which is our =model= of how the Universe works.

    Our model happens to includes a bunch of numbers, some of which we call ‘constants’. Humans came up with these constants while building the model because they helped make the model a good fit to the reality, where we measure ‘good fit’ in the practical sense by how useful it is for making accurate predictions.

    But a cool thing about models and numbers is that we can ‘turn the dials’ and tweak the model. We can change the numbers and observe what happens when we re-run the model.

    Changing numbers in a model has nothing to do with reality. Tweaking the model is a vehicle for imagining in a principled way how things could have been different.

    But we are talking about modelling the Universe here, not just modelling some little part of the Universe, like a climate model, or a statistical model. The numbers we are considering tweaking are part of our base understanding of everything. Without explicitly realizing it, when we ponder our base physical models and how the numbers can be tweaked, we have adopted the ‘god stance’.

    To attempt to answer questions like the ones following is to adopt the god stance:
    – What if there were a Universe where the physical model is most accurate with a different gravitational constant?
    – What if the Universe was different to the way it is?
    – What if there were more than one universe?
    – What if there was some sort of process (whether a conscious design process, or a series or random happenings, or whatever) that played out and resulted in the creation of the Universe?

    The god stance is a perspective that imagines itself to be outside of the Universe, looking at it from an external perspective. That’s the seductive illusion created by modelling the Universe – when we consider the model, we get the wrong impression that we are outside the universe observing from a distance. Of course this is not possible, we are always in, and never out.

    It’s an impossible perspective, and therefore we must be mindful of the assumptions we are making to render the perspective conceivable.

    The Universe is all we really know. We dont =and can’t= really know how the Universe seems from a perspective outside looking in.

    In an objective sense, we can’t answer questions about the probability of the universe, or the probability of it being different to how it is. We can certainly make assumptions and then concoct a view based on those assumptions, but that view will always be contingent on those assumptions.

    To ponder fine-tuning is to first assume that the universe could be any different to what it is. This already is a huge step towards needing to consider whether or not the universe was designed to be the way it is. It already assumes a perspective outside the universe, where our universe comes to seem to be only a token of the general type: “Universe”. There is no evidence/reason to believe that the universe is merely an instance of the category of universe. From our true perspective within the universe, it is everything, not just a token of a type.

    Instead of assuming the universe could have been different, we could instead assume that the Universe is all there is, there is no such thing as ‘outside’, there is no possibility of things being different, there is no fine tuning, there is no designer, there is no random process that brought the universe into existence, there is no multiverse.

    There is just the Universe: always is, always was, encompassing all things, all stuff, all possibilities in its one self.

    Of course, this is also an assumed perspective, but it is no less valid/possible/impossible than the idea the universe is designed, or the outcome of a random process, or just one of a multitude of multiverses.

    I guess I’m suggesting that if we identify the assumptions made in preparing our questions, we can stop trying to search for an impossible ‘correct’ answer, and instead just recognize that we can come up with different answers just based on the assumptions we choose to begin with. It’s a question of faith/subjectivity alone if you want to argue that one set of starting assumptions is better than another.

    I blame modal realism for not identifying that possibility/impossibility is always contingent on starting assumptions. A possibility isn’t real unless you commit to the assumptions that allow the possibility to be considered.

  4. Ok but what if this universe was designed by living beings like us? Surely that would be sufficient to close the loop here. If it’s natural for life to view life as very valuable — and if it’s possible for life to create new universes — then you’d expect that living beings would design habitable universes.

  5. This conundrum seems almost formally invalid (and so unfortunately otiose).

    First (and most trivial, but let us start basic), if the universe is defined as all that there is, there can be no explanation of it. Anything will inevitably bottom out in circularity, because there is nothing else to refer to. Hence uninformative, unexplainative.

    Secondly, if the response is that the explanations seek to conjecture something outside the universe, we run into the uniqueness problem: you cannot have evidence of something there is only one of. Evidence requires a model that it fits, and a model requires multiple instances of what it models. So a unique thing cannot have a model. And without a model, any evidence means nothing.

    Thirdly (from another angle), the fine-tuning argument is a fallacious use of probability ‒ we only ever have one instance. Also it is a fallacious inference informationally ‒ the possibility of different physical constants makes the universe a message we are trying to decode, but we cannot know about the info source from (only) its message content.

    And fourthly, for the question to be meaningful it must be part of a consequential proposal: if X is true, we should do this thing, and if not, we should do something else. But in this case, of what the universe is, there is never any change in outcome, the ‘if’ conditional is never activated.

    The general arguments around this and their more everyday examples confuse by implying/assuming contextual info that is not present in the whole universe question. The lack of that info is what makes it seem an especially intriguing question, but simultaneously that lack of info is what makes it impossible, and pointless, to answer.

  6. Yes. Goff is comfortable with anthropocentric arguments. After all, if Panpsychism is true than we are the finest known example of the fundamental intrinsic universal stuff. Now he adds this further arrogance that not only are we the stars of the show, and the stuff out of which the show is made, we are also the reason there is a show at all.

    It’s very unlikely we are thinking correctly about the range of possibilities when it comes to the set of wannabe universes. On this epistemically shaky ground it seems we must pick a preference of Dirty Harry lines: do you feel lucky or special, punk?

  7. “As I see it, value and meaning are rooted in the interests of living creatures, even if we sometimes speak of them as intrinsic properties of things.”

    But living creatures are things and it seems like it is intrinsic to your definition of living creatures that they have interest and so intrinsic to their existence as living things. Therefore if as you seem to mean value and meaning are rooted in the interests of living things then they are rooted in an intrinsic property of (actual) things.

    I think what you mean is that the argument relies on the posited extrinsic value of living creatures. The things living things value (including living things themselves) would be valuable even if no living thing existed, even if the intrinsic properties of all the things were wildly different. This is actually weirded then it sounds “Best of all never to be born, but who is so lucky not 1 in 10.” This is silly no one who was never born has an interest in not being born or any other interest because in order for their interest to exist they would need to exist, it is intrinsic to their existence etc.

    Whatever universe existed would have some very unlikely properties (any actual value of a real number is infinitely improbable after all) and who knows what might follow from. No doubt there would some highly (but not perfectly) regular pattern or other that would follow, one might see that pattern as a tendency towards its realization (given it is not perfectly regular it does not always happen). To call it even beautiful would be an importation of the interests of living creatures that do not exist, but still such tendencies seem like the strivings of living creatures to me. If one chose to single out that pattern it would seem awfully improbable to have ended up in the universe with that pattern. So if we valued all those unique patterns equally, no special status to our universe.

    Personally I think fine tuning arguments do signal something worth thinking about. Under Ptolemaic geocentrism it was a weird coincidence that Venus and Mercury always stick close to the Sun and that the major epicycle of every planet has the same period. Under heliocentrism (even Tychonic variations) these have an easy explanation. The major epicycles are all accounting for the relative motion due to Earth moving relative to the Sun, Venus and Mercury have orbits closer to the Sun than Earth and so are always near it relative to the Earth under heliocentrism etc. The fine tuned constants may represent a set of apparent coincidences that in another description become far less coincidental and this would make that description more explanatory less brute etc. But jumping straight to design or teleology seems like a non-sequitur the possible explanations are endless.

Leave a reply (comments will be held for moderation)

This site uses Akismet to reduce spam. Learn how your comment data is processed.