Is the cosmos "fine-tuned"?

[LegionOnomaMoi]

The answer is...yes.

Before I am besieged by an angry mob all responding with posts concerning the problems with ID, who designed the designer, probability arguments, etc., I must clarify. Yes, the universe is “fine-tuned” in that there are certain constants (sometimes called fine-tuned constants or FTCs) that, were they ever-so-slightly altered, we wouldn’t exist. For some, no life would exist and possibly no universe. This interpretation of fine-tuning is similar to the “weak anthropic principle” (WAP), which is essentially a tautology. It asserts that because we exist, the nature of the cosmos must have properties such that we can exist, for if it did not, we wouldn’t be here.

But I wouldn’t start a thread just to note that some uses of two terms have an entirely non-religious context. To me, the fine-tuned argument (FTA), which uses fine-tuning as evidence for design, provides the best evidence for the existence of a “god” or “designer” available, but it is generally misunderstood. I will seek to remedy these (hopefully) over the course of this discussion, but I can’t just reference the FTA without addressing what it is.


The FTA has, essentially, to components. One is not disputed: there are a number of properties of physics, such as the strength of gravity, which, had its force been stronger or weaker by about one in 10,000,000,000,000,000,000,000,000,000,000,000,000,000, life wouldn’t exist (no stars). Then there is the big bang, which requires a much, much larger number to represent how little a change would be needed in order for the big bang to either immediately turn into a “big crunch” or expand far to rapidly for life. The list goes on and on, but here I am just introducing issues, and the list is for debate/discussion.


The other component of the FTA is that so many properties of the universe require such extreme, unimaginable precision indicates design (and thus a designer). Normally, this part of the argument is (usually badly) justified on the basis of probability. Probability is a deceptively and deeply philosophical matter with important implications for most of the sciences. I bring this up because one counter-argument to the FTA is simply that we have no idea what the “probability space” is such that we can determine the probability that e.g., gravity would have the force it does. This interpretation of probability is frequentist. It assumes that events/outcomes are some subset from a set of all possible outcomes and the probabilities of these are determined in advance just the way we determine the probability of a coin flip to be 1/2.


The Bayesian approach is different. It is different in application, but more importantly it is also different philosophically, in that it does not posit probabilities to be viewed in absolute, idealized terms that are never actually realized, but in terms of likelihood given our state of knowledge. Thus we need not necessarily ask about the probability of a particular FTC, but rather its likelihood (in the technical sense of the term).


With as minimal information as I could manage to start this thread, I invite comments, questions, positions, arguments (for or against), criticisms, credit card information, and donations.

 

[Kirran]

I think that it's not, because if it wasn't the way it is, we wouldn't be able to ask why it is like this. WAP all the way.

 

[Rick O'Shez]

Then there is the big bang, which requires a much, much larger number to represent how little a change would be needed in order for the big bang to either immediately turn into a “big crunch” or expand far to rapidly for life.

Is it possible that there have been many "unsuccessful" big bangs before the one we refer to now? Either big crunches or extreme expansions?

 

[LegionOnomaMoi]

I think that it's not, because if it wasn't the way it is, we wouldn't be able to ask why it is like this. WAP all the way.

The WAP is completely compatible with fine-tuning, and again nobody is really debating (well, physicists aren't) whether the universe is fine-tuned, including those who opt for alternatives because, like Susskind, they believe this removes the potential need to invoke a creator:

 

[lovemuffin]

If the probabilistic arguments are usually badly made, how should the "fine-tuning" aspect be characterized? The term itself seems to indicate that sort of appreciation for the apparent metaphysical space of possible values of those constants.

Do you think the logical possibility of "life" arising with physical properties much different from our own (but presumably necessitating at least a very complex universe) is a problem for the fine tuning argument? It's presumably not a problem if changing a parameter slightly leads to a universe which has far far less complexity, but it's a problem if the argument relies only on the idea of the parameters being finely tuned for specifically carbon?

 

[LegionOnomaMoi]

Is it possible that there have been many "unsuccessful" big bangs before the one we refer to now?

Sure. But generally there is an easier and preferred method to explain why it appears that our universe is "designed". Namely, multiverse cosmologies. The argument in simple form is that the reason it appears that the seemingly impossible odds exist because there are many universes (possibly infinitely many) and almost none of them are capable of sustaining life, but in the tiny minority (perhaps even one) that is, that is indeed where we find it.

 

[Rick O'Shez]

Namely, multiverse cosmologies. The argument in simple form is that the reason it appears that the seemingly impossible odds exist because there are many universes (possibly infinitely many) and almost none of them are capable of sustaining life, but in the tiny minority (perhaps even one) that is, that is indeed where we find it.

Is there any evidence for the multiverse theories do you know, or is it a purely theoretical idea?

 

[LegionOnomaMoi]

If the probabilistic arguments are usually badly made, how should the "fine-tuning" aspect be characterized?

Likelihood (or propensities, subjective probability, or a number of approaches to probability that don't assume a frequentist interpretation). For example, maybe 2 years ago or so a paper was published in PNAS on the probability of life existing elsewhere in the universe. However, the approach was Bayesian. It starts with what we know about life originating (life occurred here, and did so under certain circumstances, some of which are little known and others well understood), about the nature of the universe and in particular "habitable zones", etc., and then plugs all of what we know into a model that tells us the likelihood (or a range of likelihood values) which takes as given what we know about the universe and life and uses Bayesian statistics/inference to compute the probability of life occurring elsewhere based upon what we know. It turns out that the result is arbitrarily small.

That is, when we start with what we know and use this information to calculate the likelihood of a range of scenarios based upon what we know, the fact that life exists here is controlled for rather than biasing in the extreme any calculation of the probability of life existing elsewhere. All we can say is that we know life arose here and it seems to have been highly improbable for various reasons, and thus the likelihood of it existing anywhere else approaches 0.

This kind of argument can be and repeatedly has been used as part of the FTA. In broad strokes, we can begin as if we knew nothing about the values of various parameters, constants, forces, etc. This is our non-updated position (the "new" information, i.e., the information about these values as they exist, is left out as it is essentially a bias). We can then imagine all sorts of possible values and at best we have no reason to think any set of values is more likely than another (it's actually more complicated than this, and not all outcomes are equal). If that is the case, then given a universe we'd expect the values to be randomly distributed. Then we receive our "updated" information, or the values that actually exist. These values are not randomly distributed. They are about as far from randomly distributed as it is possible to get. Given our initial ignorance about the values, we were justified in assuming a random distribution. Now we are at the very least far, far less justified. Additionally, the values aren't just non-random, but are extremely precise in just the right ways to enable life and us. If we calculate how likely it is that the values we find were random given our initial assumptions and updated beliefs, the likelihood here is infinitesimal.

Do you think the logical possibility of "life" arising with physical properties much different from our own

Excellent point, but the problem is that so many of these constants would result in no universe, or a universe of only hydrogen. So while something like the surface tension of water can be seen as fortuitous for us but not necessarily for life (as we can speculate a universe in which life as we know it couldn't exist but in which life itself could), these parameters are rather insignificant.

 

[outhouse]

The answer is...yes.

Unsubstantiated personal opinion.

As is mine on this topic


FTA is based on design. Nature shows no design what so ever in any aspect of any part of the universe. The arguments for are all rhetorical in nature that include any aspect of design.

The universe is in fact, not friendly to life. Most is devoid of life because of this reason.


Life is hard to hold back. Abiogenesis is a product within aggressive environments.

Concrete is not a suitable place for plant life, it would be considered a place where plant life does not exist. Yet it only takes a crack and plant life takes hold. The crack was not designed for life to take hold. That is where I see your argument failing.

 

[lovemuffin]

Is there a criticism of the Bayesian approach in that the outcome is almost all in the prior probability, which is constructed in a somewhat arbitrary way? That is, since we don't have other universes to observe there is only one data point in that sense. So while the frequentist approach is nonsensical, the bayesian approach just slaps a mathematical veneer on what are mostly assumptions that can't easily be justified?

I'm asking because I'm not sure, rather than in a rhetorical way

 

[LegionOnomaMoi]

Is there any evidence for the multiverse theories do you know, or is it a purely theoretical idea?

Actually, fine-tuning is considered by some to be evidence of the multiverse. However, there does exist other evidence. One strong piece of evidence is the fact that it may potentially explain quantum physics (the "many-worlds interpretation" or MWI). However, the kind of multiverse that we'd need is one in which our universe is incredibly vast, and indeed so vast that from our perspective it exists as some huge number of so-called pocket universes, each closed off from any other and in which the laws of physics may differ radically. The evidence for this is largely mathematical, but not entirely (nor is mathematical evidence necessarily a poor basis, at least to begin with). If we are able to learn more about the nature of the big bang and (if it exists) the pre-big bang period, are able to find empirical evidence for supersymmetry, and in general are able to show a particular GUT is superior to all known others and is both theoretically sound and empirically supported, we will be able to say more about multiverse cosmologies.

 

[outhouse]

However, the kind of multiverse that we'd need is one in which our universe is incredibly vast, and indeed so vast that from our perspective it exists as some huge number of so-called pocket universes, each closed off from any other and in which the laws of physics may differ radically.

Doesn't have to be that way, the laws could be similar.

If our universe is a product of a supermassive black hole expanding, other possible universe could have the exact same laws.

Until we can fully define a singularity, including fully understanding gravity, we just wont know.

 

[Parsimony]

One is not disputed: there are a number of properties of physics, such as the strength of gravity, which, had its force been stronger or weaker by about one in 10,000,000,000,000,000,000,000,000,000,000,000,000,000, life wouldn’t exist (no stars).

I want to see the calculations and assumptions which justify this assertion. The way I see it, gravity's strength could be vastly different and we'd still have stars. The difference being that, in a world with a higher gravitational constant, the minimum mass required to initiate fusion in a star would be lower and the diameter of the resulting star would be smaller than in our universe. The reverse would be true for a lower gravitational constant.

 

[LegionOnomaMoi]

Is there a criticism of the Bayesian approach

Sure. It's practically the longest running debate in the philosophy of probability. In application, things are much easier. Machine learning, computational intelligence, AI, pattern recognition, etc., all involve algorithms or classes of algorithms that use Bayesian inference because it works, while to use the frequentist approach would require that we had the problem solved. On the other hand, given a fair coin, as each toss is independent of the one before it we can continually update our information and never arrive at the intuitive answer we get from the frequentist approach: H's and T's both have a 50/50 chance.

That is, since we don't have other universes to observe there is only one data point in that sense.

True. This presents a serious problem, particularly (as you note) for the frequentist position, but also for us. What strengthens the argument considerably is that, while have no other data points, that only limits how much we can "learn" given a single data point. Normally, it would limit things to the point of being useless (in machine learning, whether Bayesian statistics is used or not, some pattern/classification/whatever is learned through repeated input, each one called an "experiment" and the result of each an hypothesis; normally, the program will run through thousands of hypotheses before it learns to correctly classify).

Luckily, the single data point is offset both by the number of constants and parameters and (FAR, FAR more importantly) the possible ranges of each. Also, starting from a position of no knowledge as to the probabilities of particular values is independently justified in that there is no reason (unless one invokes a designer) for particular values. That is, we have no justification for limiting the strength of gravity just because of its value in the universe we find ourselves. From a frequentist perspective, we have no justification for positing anything about probabilities as we have only the one "experiment", but the frequentist approach is contradicted almost all the time (it has to be, as it treats single outcomes in terms of idealized frequencies such that a large or infinite number of identical "experiments" are posited to determine the probability of anything using this approach).

From a Bayesian perspective, we aren't actually justifying anything so much as we are declaring ignorance. Even if we determine that, given our updated information, we can't determine how likely we were to find values that we find did (even approximately), we can determine how likely it is that we (or life) would exist given different outcomes.

 

[LegionOnomaMoi]

Doesn't have to be that way, the laws could be similar.

That's true. However, arbitrarily small departures from certain constants or parameters would mean no life or perhaps no universe.

 

[LegionOnomaMoi]

I want to see the calculations and assumptions which justify this assertion.

Found a copy of the paper for free: The significance of numerical coincidences in nature

 

[outhouse]

That's true. However, arbitrarily small departures from certain constants or parameters would mean no life or perhaps no universe.

Luck does not deduce designed though.

 

[Valjean]

I think it's we who are fine tuned by the Reality we find ourselves in, not vice versa.

 

[LegionOnomaMoi]

Luck does not deduce designed though.

Sure. After all, if I thought the evidence that convincing, I wouldn't be agnostic. It is, though, a considerably stronger argument (or superior evidence) than it is often credited as being, and in particular because it is often misunderstood, or combined with the other "design" argument that has to do with biology and evolution and is not a good argument in the slightest.

Unlike with e.g., ID, fine-tuning isn't some at best fringe theory and more likely just pseudo-science, but pretty universally accepted. Also, it is accepted in such a way that a number of physicists find it to be a rather serious problem that requires an explanation. The standard explanation given by non-believers is multiverse theory, which borders on the religious all by itself.

 

[Guy Threepwood]

The answer is...yes.

Before I am besieged by an angry mob all responding with posts concerning the problems with ID, who designed the designer, probability arguments, etc., I must clarify. Yes, the universe is “fine-tuned” in that there are certain constants (sometimes called fine-tuned constants or FTCs) that, were they ever-so-slightly altered, we wouldn’t exist. For some, no life would exist and possibly no universe. This interpretation of fine-tuning is similar to the “weak anthropic principle” (WAP), which is essentially a tautology. It asserts that because we exist, the nature of the cosmos must have properties such that we can exist, for if it did not, we wouldn’t be here.

But I wouldn’t start a thread just to note that some uses of two terms have an entirely non-religious context. To me, the fine-tuned argument (FTA), which uses fine-tuning as evidence for design, provides the best evidence for the existence of a “god” or “designer” available, but it is generally misunderstood. I will seek to remedy these (hopefully) over the course of this discussion, but I can’t just reference the FTA without addressing what it is.


The FTA has, essentially, to components. One is not disputed: there are a number of properties of physics, such as the strength of gravity, which, had its force been stronger or weaker by about one in 10,000,000,000,000,000,000,000,000,000,000,000,000,000, life wouldn’t exist (no stars). Then there is the big bang, which requires a much, much larger number to represent how little a change would be needed in order for the big bang to either immediately turn into a “big crunch” or expand far to rapidly for life. The list goes on and on, but here I am just introducing issues, and the list is for debate/discussion.


The other component of the FTA is that so many properties of the universe require such extreme, unimaginable precision indicates design (and thus a designer). Normally, this part of the argument is (usually badly) justified on the basis of probability. Probability is a deceptively and deeply philosophical matter with important implications for most of the sciences. I bring this up because one counter-argument to the FTA is simply that we have no idea what the “probability space” is such that we can determine the probability that e.g., gravity would have the force it does. This interpretation of probability is frequentist. It assumes that events/outcomes are some subset from a set of all possible outcomes and the probabilities of these are determined in advance just the way we determine the probability of a coin flip to be 1/2.


The Bayesian approach is different. It is different in application, but more importantly it is also different philosophically, in that it does not posit probabilities to be viewed in absolute, idealized terms that are never actually realized, but in terms of likelihood given our state of knowledge. Thus we need not necessarily ask about the probability of a particular FTC, but rather its likelihood (in the technical sense of the term).


With as minimal information as I could manage to start this thread, I invite comments, questions, positions, arguments (for or against), criticisms, credit card information, and donations.

That general observation used to be considered a 'theistic argument' as the big bang used to be, but I think it is ever more accepted even in atheist cosmology nowadays, it's the basic foundation of multiverse theories that, enough random universes 'would create this one eventually' in Hawking's words. Presumably the multiverse comes equipped with a built in safety mechanism to guard against ever 'creating God eventually' or that would defeat the entire point!

I agree that the odds do suggest creative intelligence, not simply because the odds of chance are low, but that they are so easily beaten if we allow only the merest possibility of creative intelligence. Like finding 'help' written in rocks on the beach. It's always possible the waves did it, but you would have to practically rule out the possibility of purpose before chance becomes the best answer, and we have no basis to rule out an intelligent creator of our universe.

I think there are also the beginnings of signs of preparations to accept this is atheist cosmology also, Andre Linde for example, where the creator would be an alien and or future intelligence, something as far removed from God as possible