In his book Time Reborn: From the Crisis in Physics to the Future of the Universe theoretical physicist Lee Smolin takes a shot at the anthropic principle. He is not in favor, because the principle supports cosmological theories (such as eternal inflation) that cannot be falsified by observations.

I think one can agree with the sentiment that theories should not just rest on a principle alone but still view the anthropic principle as self-evident. In any case, the following reasoning by Smolin for refuting the principle's role (in reasoning about the cosmological constant, but it could be X) seems flawed:

I’ll start with a fallacious argument that goes like this:

  1. Galaxies are necessary for life. Otherwise stars would not form, and without stars there is no carbon and no energy to promote the emergence of complex structures, including life, on the surfaces of planets.

  2. The universe is full of galaxies.

  3. But the cosmological constant has to be smaller than the critical value if galaxies are to form.

  4. Hence, the anthropic principle predicts that the cosmological constant must be smaller than the critical value.

Can you see the fallacy? Point no. 1 is true, but it plays no role in the logic of the argument. The real argument starts with point no. 2. The fact that the universe is filled with galaxies is evident from observations; it is irrelevant whether or not life would be possible without them. So the first point can be dropped from the argument without weakening the conclusion. But point no. 1 is the only place life is mentioned, so once it’s dropped, the anthropic principle plays no role.

I would argue that point no. 1 cannot be dropped because the anthropic principle (in point no. 4) is about "life": anthropic stands for "of or pertaining to mankind or humans, or the period of humanity's existence". Without human life (if the argument starts with point no. 2) there is no observation and nothing is "evident from observations".

So my question is this: is Smolin's argument indeed flawed or am I missing something?

  • why don't you ask him your question?
    – nir
    Commented Sep 18, 2014 at 21:36
  • @nir If I could reach out to him with my SO-identity (and invite him to the forum, so to speak), I would ...
    – Drux
    Commented Sep 20, 2014 at 13:16
  • Send him an email and let us know his response.
    – nir
    Commented Sep 20, 2014 at 13:18

4 Answers 4


The weak anthropic principle essentially states that when we observe the universe, we have 100% probability of observing a universe in which it is possible for humans to exist - because we already know that we do. If the universe is such that it is impossible for humans to exist, we obviously wouldn't be here to do the observing!

This is sometimes important to remember when we're observing the universe that we're in - just because it is impossible for humans to observe a universe filled with 'Smergs'1 instead of galaxies doesn't mean that the alternate universe is impossible, just unobservable. Sometimes the answer, unsatisfying as it may be, to the question of 'Why do we observe the universe to be like this?' is that it's impossible for humans, specifically, to be in that different universe.

So that said, Lee Smolin's mistake comes in at point #2. Smolin states:

The universe is full of galaxies.

when more accurately, when discussing the anthropic principle, he should have used:

Human beings observe that the universe is full of galaxies.

So for human beings to be able to observe galaxies, humans need to 1) exist (to do the observing) and 2) galaxies need to exist (to be observed). If humans can't exist without galaxies, then anything that disrupts galaxy formation will thwart us humans from doing the observing right at that first step. Thus, once we have proved the necessity of galaxies for human existence we can conclude that the cosmological constant must be in the range of values that allow galaxies to form, thanks in part to the weak anthropic principle.

1 Where a 'Smerg' is any arbitrary 'Something that might emerge rather than galaxies' in our thought-experiment universe.

  • Isn't "Human beings observe that the universe is full of galaxies" a consequence of Dr. Smolin's premises 1&2?
    – Dave
    Commented Sep 18, 2014 at 18:39
  • 1
    @Dave: It is, but by omitting the part where humans do the observing in point #2 Smolin can proceed to drop point #1 and reach the conclusion that the anthropic principle has no bearing on the cosmological constant. Smolin says But point no. 1 is the only place life is mentioned, so once it’s dropped, the anthropic principle plays no role, ignoring that human observers are an (unstated-by-Smolin) requirement for #2.
    – Dave B
    Commented Sep 18, 2014 at 19:23

I think that's a pretty bad presentation of the anthropic principle on his part. But I wouldn't say the flaw rests in just saying he can lop off the first claim. It seem more flawed in that he's misunderstood the entire idea of the anthropic principle in presenting it that way.

I take it the anthropic principle is not one thing is coincidentally right for life of the sort we are but rather that an extremely large number of cosmic values and equations and placements and configurations are so well-suited to the sort of life we are. And that the sum of all of these makes coincidence less likely than a coordinated basis.

Or to put it another way, we can apply the anthropic principle to a lawn:

1. Grass is necessary in a human-managed lawn
2. This yard is full of grass.
3. But grass would be necessary to a human-managed lawn
4. Therefore the anthropic principle predicts this yard will be full of grass.

I think the place he's putting the anthropic principle is what's broken there. The anthropic principle is not intended to prove things based on wish but rather to suggest that the data supports a hypothesis.

So then the grass one would be:

1. Grass is necessary  for a human-managed lawn
2. Humans and nature can both water grass
3. Cut grass is unlikely to happen by nature
4. Only humans use brick edges to place along the lawn
5. Since the lawn contains cut grass, brick edges, was recently watered [despite no rain], odds on there's someone engineering that.

And that's more like the anthropic principle.

  • +1 Do you have an example where a cosmological theory would involve sg. like "cutting grass" -- i.e more than "observation" by humans?
    – Drux
    Commented Sep 18, 2014 at 6:41
  • I never used the word "observation" in my answer, but I'm not sure how any argument humans generate can involve more than the sum of reasoning and observation by humans... I take the anthropic principle to be that it should count for something that all the constants and occurrences make life possible considering even minor changes in one kill the whole thing.
    – virmaior
    Commented Sep 18, 2014 at 6:49
  • 1
    Thus I take the presentation in his book to be flawed because it works from one thing and misunderstands how the principle is supposed to work at all. Shoddy philosophy being offered by a scientist?!
    – virmaior
    Commented Sep 18, 2014 at 6:50
  • That's a nice way to put it: "I take the anthropic principle to be that it should count for something that all the constants and occurrences make life possible considering even minor changes in one kill the whole thing"; quoted "observation" was from my text BTW
    – Drux
    Commented Sep 18, 2014 at 6:58

Like Smolin, I'm a critic of the anthropic principle, but I don't find his attack on it (at least as so quoted) compelling. As I understand the principle, it's largely concerned with statistical illusions. If I flip a fair coin 50 times and get heads each time, it might seem like the chances of getting another head on the 51st flip are tiny, but in fact they are 50%, the same as any other flip.

It seems unlikely that we would find ourselves on a planet compatible with life, but given that life exists, it is certain we would find our own planet compatible with it.

The principle is compelling in (and I would argue, only in) the case where one can either presuppose a range of conditions that includes some compatible with life, or where we can assume that something like life can be compatible with any of a range of available conditions.

We can restate Smolin's version of the principle as "given that X exists, conditions must be compatible with the creation of X." In his second version of the argument, we substitute "galaxies" for X instead of "life." He now considers this to no longer be anthropic, since life isn't a part of it at all. However, if the generic version of the principle is good, then it remains good whether we call it "anthropic" or not.

It's not clear, at least from this brief excerpt, where he's taking the larger thrust of his argument --perhaps this is enough to serve his larger aims, perhaps it isn't.

  • According to the wikipedia the anthropic principle is not about optimisation but compatibility.
    – nir
    Commented Sep 18, 2014 at 16:13
  • @nir you are correct, I will edit Commented Sep 18, 2014 at 16:24
  • In your 1st paragraph you may be mistaking the anthropic principle for confirmation bias (or some black swan phenomenon), which it is not.
    – Drux
    Commented Sep 19, 2014 at 7:04

I dunno, this all seems to be very similar to "I think, therefore I am."

Here's my problem with the Anthropic Principle. Somewhere buried in there it mentions the vast size of the universe. It talks about the Multiverse. And to some aspects of the principle, it calls on the assumption, that the galaxies are filled with endless possibilities.

I'm not sure I agree. Even if there were a quintillion galaxies, I don't think that automatically means that the odds are in our favor that there is more life out there and therefore, it's more proof that the system was build for life.

I in fact think, that if anything is possible, it's possible that in the real scheme of things, this is but a blip, a brief moment in time. And we may just be the first life (of millions of life-forming galaxies to come later) here.

After all, if there is a beginning, there has to be a first. And no- I don't believe there is anything inherently special about being first. So a guess as to life that's out there based on odds to further hold up another guess.. I believe is standing on a false premise.

  • The anthropic principle talks about the multiverse? -- not that I would know.
    – Drux
    Commented Sep 19, 2014 at 7:00

You must log in to answer this question.

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