I wrote this essay in January, 1995 for the Freethought Debate FAQ. I have since made some minor modifications to that original version. The essay represents my "take" on the Anthropic Principle. I first came across the Principle in 1987 (probably in some of Isaac Asimov's writings), and then again in my physics classes in 1988. My first opinions on the Principle have not changed at all over time, even as I continued to read more about it. If anything, my thoughts have become more concrete and confident.
Krishna.
One of the most popular, widely-discussed and important arguments for the existence of God (or Gods) is the Teleological Argument (from the Greek telos meaning end or purpose), more commonly known as the Argument from Design---important not because teleology has any special inherent merits (it is in fact based on poor analogies), but for the following two reasons:
For these reasons, it is important to debate and refute all theistic arguments based on intelligent design, and show that they are not the wonderful explanations the theist claims them to be.
As human knowledge has increased, past examples of alleged design have been shown to have naturalistic explanations. However, knowledge always has a boundary, and some version of teleology is always being applied at or beyond the current boundary of knowledge.
One of the most widely known, read and quoted teleological argument comes from William Paley---he used biological systems as evidence of intelligent design. The theory of Darwinian evolution and the many theories of abiogenesis have provided more than adequate explanations for how complex biological systems could evolve from simple ones, and how the simple systems could have arisen through entirely natural processes in the first place. Paley's version of teleology (however clever it once might have been) is no longer a productive point for debate, although some (I am referring here to the creationist cretins) would prefer to think otherwise.
However, clever ideas rarely disappear completely. The argument from design as applied to biological systems has been resurrected in the recent past as the argument from the Anthropic Principle as applied to the Universe as a whole.
The essay will examine the Anthropic Principle in its many incarnations, the attempted uses of the Principle to prove the existence of God, and refutations of such arguments. Readers who are familiar with the Anthropic Principle can directly skip to the section dealing with the refutations.
There is a lot of deep physics and mathematics behind some of the following discussions. I am not even remotely qualified to pass judgement on these subjects or try to explain the precise implications of these ideas. However, it is not critical to fully understand the physics and the math to be able to critique the theology associated with the Anthropic Principle.
It is first quite important to recognize that the term Anthropic Principle (henceforth abbreviated as AP) has been much abused. Different authors offer different:
Perhaps disgusted by this proliferation, Martin Gardner once proposed his own Completely Ridiculous Anthropic Principle (i.e. CRAP). This essay will focus on one set of popular definitions; other formulations can be found in the references.
Any reasonable definition of the AP depends on the truth of some basic propositions. Let us first examine these proposition.
A corollary of the last two postulates is sometimes stated as a question. "Why is the Universe so conducive to life?". The rhetorical answer is that "if the Universe were not conducive to life, we would not be around to ask that question". And interestingly enough, this seemingly cheeky answer lies at the heart of the proper uses of the Anthropic Principle (and forms part of the refutation of the incorrect uses of the principle).
Now we come to the crucial propositions that are the foundation of the Anthropic Principle.
The reason we do consider the AP worthy of discussion is that an increasing body of scientists thinks that the above proposition is true, namely that life (and intelligent life) is possible only in a small range of values for the various fundamental physical constants and initial conditions.
For example, if the coupling constant for electromagnetic interactions was greater than its current value, chemical reactions would be so violent that large molecules would not be stable. If the coupling constant was smaller (and thus, the strength of electromagnetic interactions was weaker) than its current value, chemical reactions would occur far too slowly to allow complex molecules to form.
Similarly, if the electromagnetic force was comparable in strength to the nuclear force, the Universe would consist of neutrons. If it were weaker than what it is now, the Universe would consist of hydrogen.
These two examples are somewhat simplistic, but literature on the subject provides more detail and more examples.
The various anthropic principles are explanations for the choice of initial conditions and the values of the fundamental physical constants. We can define four different kinds of Anthropic Principles.
Few scientists would quarrel with the observations leading to the WAP, but few believe that the SAP is true. Fewer still think that there is any truth or predictive power in the PAP and FAP. PAP and FAP find more acceptance among abstract philosophers.
This document will discuss the uses of only the WAP and SAP in arguments for the existence of God.
The APs (especially WAP) are not idle speculations; they have been put to non-trivial uses. The WAP has explanative and predictive power.
For example, the WAP provides an explanation for the current age of the Universe---it takes time for first-generation stars to form, die, for second-generation stars with planetary systems to form, and for life to evolve and observe the Universe.
The WAP also has predictive power. Fred Hoyle used the WAP to postulate the existence of energy resonances in the nucleo-synthesis of carbon from simpler elements---a fact confirmed by subsequent observations.
Every scientific theory must:
The WAP is a scientific theory in the sense that it has predictive power (as the above examples indicate) and in the sense that it can be falsified (for example, one could simply show that life of some sort will inevitably emerge in all possible Universes). On the other hand, many philosophers of science have criticized the SAP for failing the Popper test of falsifiability. Some scientists feel that if the SAP were true, it only makes a case for the existence of deeper natural laws that lead to constrained values for the fundamental physical constants. Many scientists feel that the WAP is true, but that every other AP is false.
To recap, the APs all affirm that the Universe we live in is the result of the combined inter action of a set of factors, which if randomly and independently chosen, would make the emergence of our Universe exceedingly improbable.
From this affirmation, theists seek to infer proofs for the existence of God (or Gods), and further, the existence of God (or Gods) as described by one or more of the religions known to humans.
All uses of APs to affirm the existence of God are based on probability arguments. We therefore move beyond the realm of propositional logic and enter the realm of statistics and probability. We move from proof to plausibility. The probabilistic arguments against incorrect uses of the Anthropic Principle are the most comprehensive, but they are also the most abstract.
Theists often quote large numbers (I mean really large), often from physicists, as odds against the chance formation of the Universe that can support life. As mentioned before, a posteriori arguments are not proof, we may only talk about likely and unlikely scenarios.
It is also worth noting that none of the arguments from APs is valid as a defense of a personal Judeo-Christian God. This certainly does not prevent theists from claiming otherwise, but that is fiat not logic.
Assume that we can create (at least in our minds) Universes with random values for the different fundamental physical constants. Since there is only a small range for each constant that is conducive to life, it is extremely improbable that we would choose the proper values for the constants. An overwhelming number of these created Universes would not support life. Our Universe does. Therefore, it is extremely improbable that our Universe was created by chance, i.e. its creation was directed by a Creator of cosmic intelligence whom we call God.
Such a view of God is Deistic. God sets the control knobs of creation to the proper values and then let the machinery of the Universe work unhindered. A small variation of the Deistic view might be a God who also fine-tunes the laws of nature during the inflation epochs of a Universe. The AP does not in any way support the notion of a personal God who constantly interferes with the laws of nature via miracles that are supposedly detect able by a small group of humans on a tiny planet orbiting an average star in an average galaxy, in the Universe.
Arguments for the existence of God that are based on the Anthropic Principles can be refuted in many different ways. Here are some.
Refutation from Rhetoric is the simplest refutation of the argument from the AP. The refutation goes as follows:
It may be true that life in the Universe is exceedingly improbable, that the conditions necessary for the emergence of life as we know it are exceedingly improbable, and the selection of such favorable initial conditions and physical laws of the Universe is exceedingly improbable. So what! It may well be true that our Universe and our existence is a vastly improbable accident. It may be true that if the machinery of the Universe were to be run again, (and again and again) a Universe like ours (and with us in it) would never emerge. As long as our Universe is not impossible, we have no reason to ascribe transcendent purpose or deliberate design to its origins.
Wishful thinking that our existence cannot be accidental does not make our desires valid; they are simply evidence of human vanity and anthropocentric biases in our thought. While an explanation that denies purpose to the Universe and importance to humankind may be unsatisfying to many, that does not legitimize the construction of incorrect but satisfying falsehoods.
As an example, consider the toss of a billion coins. While a toss of a billion heads is very unlikely, it is not forbidden. If the very first toss results in a billion heads, there is no reason to suspect that the toss was carefully orchestrated by a master Creator. In fact, any given configuration of coins is as likely as any other and requires its own explanation and can thus be construed as proof for manipulation by a master Creator. Therefore, the selection of one particular final configuration (and thus the rejection of all other possible final configurations) is not a proof of any coherent statement about the process of selection itself.
If one assumes the existence of a God, then this God may serve as an explanation for the current Universe. However, God is not the only explanation. To use the postulate of the existence of God to prove the existence of God is circular reasoning.
Refutation from Regress is based on the question that most 5-year old children ask when told about the concept of God as a creator i.e. "who then created God?".
Let us concede that the Universe is improbable, so improbable that it requires fine-tuning by a creator God. This means that this God must be capable of analyzing and solving the complex problem of constructing anthropophilic (conducive to humans) Universes. This means that the creator God must itself be complex (and this is what theists claim as well). Therefore, if God is complex (as is necessitated by the existence of an anthropophilic Universe) and that all complexity derives from pre-existing complexity and requires explanation based on design arguments (which is the theme underlying theistic uses of the anthropic principle), it leaves us with the new problem of who or what designed the complex, creator God.
It is not possible to wish away this problem by claiming that God is unexplainable or does not require an explanation. One might as well claim that the Universe simply is and requires no explanation. It is also not possible to claim a hierarchy of increasingly powerful Gods since this leads to infinite regress.
Thus the invocation of God as a solution to the problems posed by the Anthropic Principle is no solution at all. At best, it is an attempt gives the label God to our ignorance of the conditions leading to the formation of our particular Universe. In any case, it obfuscates the issues at hand.
So far we have assumed that the Anthropic Principle is true i.e. life in our Universe is improbable (or at least the classes of Universes where life can emerge are improbable). Implicit in the preceding statement is the notion of a random selection of one particular Universe from among the set of possible Universes. But what if not all Universes (and not all choices of natural laws, initial conditions, and values for fundamental physical constants) are equally probable.
Such powerful arguments come from physicists who use general relativity and quantum theory to investigate the Big Bang. Stephen Hawking is among those who think that when a consistent formulation of quantum gravity is found, we will be able to provide explanations for the current axioms of physics (natural laws, initial conditions, and values of fundamental physical constants) in terms of a simpler theory.
Of course, one may say that such a formulation only leaves us with the problem of explaining this simpler theory. But one also needs to remember that all teleological arguments are applied at the boundary of current knowledge. Explanations have to stop at some point, but theists are constantly shifting the position of this stopping-point to suit their needs for debate. At some level of detail, it becomes valid to ask if there is really a legitimate need to explain or justify the postulates of a system in terms of simpler ones.
Alan Guth proposed the Inflationary model to explain features of the post-Big Bang evolution of the Universe. A very simplistic view of inflation is that it is a phase transition in the Universe that obliterates all (or most) traces of history (in addition to producing long range structure in the Universe, among many other things). Andrei Linde has proposed more advanced Chaotic inflationary models that are consistent with general relativity and quantum theory. The existence of inflationary phases in the Universe has two consequences:
Thus we have no valid basis to claim that our Universe is in some sense special. The Universe, by means of the bubbles, actually does explore all possibilities of existence; it is a self-selection bias that makes us wonder in amazement at the tailoring of our bubble to support our form of life. Other bubbles can evolve in very different ways, with or without any possibility of life, and with exotic life forms that may in turn wonder at their privileged role in creation.
Thus chaotic inflation essentially destroys the basis for all the APs other than the WAP. The WAP itself is reduced to a statement about self-selection, i.e. our Universe seems so conducive to us because if it were not, we would not be around to ponder this conduciveness.
Stephen Hawking has proposed that the quantum idea of sum-over-histories may be applied to the Universe, and that not all possible histories are equally probable. Thus a theory of quantum gravity may provide a fully self-contained explanation of why the Universe evolved along the direction that it did without recourse to any probabilistic arguments or appeals to God---such a no-boundary condition theory is all that is needed to explain the Universe.
Such a theory would destroy the basis of all APs (including the WAP) since it no longer makes sense to talk about the extreme improbability of the Universe when the Universe is actually constrained by its own existence to evolve along certain directions.
As with all theories, quantum gravity can only shift the granularity of explanation down to a finer level. If and when a theory that explains quantum gravity is found, the level of explanation would shift down another level. The existence of a set of unexplainable axioms at some level does not validate the use of design arguments to explain axioms. To claim so is as meaningless as claiming that there is always the need for a non-scientific, un-analyzable metaphysics sitting below science.
Most of the previous refutations can be summarized in a simple mathematical form. The refutations based on conditional probability uses the Bayes Theorem. Let us define the following symbols:
and equations:
Then we have:
prob(U) = prob(G).prob(U|G) + prob(N).prob(U|N)
i.e. the probability of the current Universe is the sum of two probabilities (the two scenarios are further mutually exclusive i.e. the event of an intelligent designing God renders natural laws irrelevant in creating our Universe (although not in sustaining the Universe) and the event of natural laws capable of creating our Universe renders God irrelevant. Therefore, it is mathematically valid to add the probabilities using the equation given above without accounting for the simultaneous occurrence of the events G and N).
The theistic argument assumes that prob(U) is close to 1. The scientific (non-theistic) arguments assume that either prob(U) is close to 0 (the Anthropic Principle) or that prob(U) is close to 1 (the no boundary condition theory). From this assumption of what prob(U) is, each camp attempts to derive prob(G) or prob(N) based on further assumptions about prob(U|G) or prob(U|N).
The theists claims that prob(U|G) is close to 1 i.e. that an intelligent designer God would want to create life like ours (and human life in particular). This is an unsupported assumption-using religious books to show God's special preference for humanity is worth less as evidence. Even more troublesome for theists would be the discovery of non-human extra-terrestrial advanced civilizations that do not have religions or have religions very different from Earth-based ones. Theists also use the Anthropic Principle to claim that prob(U|N) is close to 0.
Given prob(U) close to 1, prob(U|G) close to 1, and prob(U|N) close to 0, it is little wonder that theists derive a prob(G) close to 1 and prob(N) close to 0 i.e. if our Universe is almost a certainty and if a God is almost certain to create our Universe, then it is almost certain that such a God exists (the God/Nature dichotomy is implicit).
The AP-accepting scientist can make no claims about prob(U|G) since this question is not scientific (and involves probing the mind of God)! Such a scientist simply accepts low values for prob(U|N) and thus for prob(N) i.e. our existence is an accident.
The quantum-gravity theorist (using no-boundary-condition theories) would claim that both prob(U) and prob(U|N) are close to 1. As with other scientific statements, no claims can be made about non-scientific questions like prob(U|G). Invoking the Occam's razor, we may claim that prob(N) is close to 1 i.e. the Universe arose through entirely naturalistic causes.
The Anthropic Principle is invalid as a proof of the existence of a God (or Gods), much less a proof for the notion of a personal God as described by any of the religions known to humans.
Arguments from the Anthropic Principle can be refuted based on Rhetoric, the problem of Infinite Regress, a denial of the Anthropic Principle itself, and from Conditional Probability and the Occam's razor.
[ Miscellaneous | Krishna Kunchithapadam ]
Last updated: Sun Jun 27 17:00:19 PDT 2004
URL: http://geocities.datacellar.net/krishna_kunchith/misc/anthropic.html