Cosmic Fine-Tuning

Several physical constants and initial conditions fall within ranges that permit complex chemistry and structure. Changing some of them appreciably produces a universe without stars, atoms, or chemistry.

This page covers the physics: which quantities are described as fine-tuned, by how much, and how the ranges are calculated. The inference from fine-tuning to a designer is a philosophical argument and is treated under the teleological arguments in Philosophy.

The commonly cited cases#

The cosmological constant is the case usually presented as strongest. The discrepancy is between the observed value and a naive quantum field theory estimate of the vacuum energy, which exceeds it by roughly 120 orders of magnitude.

How the numbers are produced#

The sensitivities are calculated by varying a constant in the relevant physical model and computing the result — stellar structure equations, nucleosynthesis codes, or cosmological simulations.

This is legitimate physics, and the calculations are published. Several important qualifications apply to how the resulting numbers should be read.

One constant at a time. Most published sensitivities vary a single parameter while holding the rest fixed. Where multiple parameters are varied together, the life-permitting region is often larger, because changes can compensate.

Fred Adams's 2019 review in Physics Reports surveys the stellar case and concludes that a substantial fraction of parameter space — by his estimate roughly a quarter, when gravitational constant, fine structure constant, and nuclear reaction rates are varied jointly — permits stars that burn stably for long periods.

The probability of a range is undefined without a distribution. Saying a constant is fine-tuned to 1 part in 10⁴⁰ describes the width of a window relative to some assumed span. Without knowing the possible range of values and how likely each is, the ratio is not a probability. Whether the constants could have taken other values at all is unknown.

Some constants may not be free. A deeper theory could fix values that currently appear arbitrary. This has happened before: the ratios of chemical element abundances once looked arbitrary and are now derived from nuclear physics.

The Hoyle resonance#

The most cited historical case.

Fred Hoyle reasoned in 1953 that carbon production in stars required an excited state of the carbon-12 nucleus near 7.68 MeV, because the triple-alpha process would otherwise be too slow to produce observed carbon. He predicted the resonance, and it was found at 7.65 MeV.

This is a genuine successful prediction and is often presented as evidence of fine-tuning. Hoyle himself remarked that the universe looked like a "put-up job," while remaining an atheist.

Two clarifications. Hoyle's reasoning was anthropic in form — carbon exists, therefore the mechanism must work — and it was a prediction about nuclear structure rather than about the constants. Later work has found the resonance tolerates variation of a few percent in the strong force rather than being knife-edge, though the tolerance is genuinely narrow.

Proposed physical explanations#

These are physical rather than theological, and each has difficulties.

The multiverse is often characterised as an ad hoc escape. It is not proposed to solve fine-tuning: eternal inflation was developed in the 1980s to address the horizon and flatness problems, and the landscape emerged from string theory's vacuum structure. That both predict varying constants is a consequence rather than a motivation. Neither is confirmed.

Common objections#

"The fine-tuning numbers are exaggerated"

Some published figures do overstate the case, and the criticism has force for the more dramatic quotations.

The main issues are varying one parameter at a time, assuming an arbitrary range over which to compute the ratio, and assuming carbon-based life is the only possibility.

Not all cases dissolve under scrutiny. The cosmological constant remains anomalous under any treatment, and Adams's analysis — which is the most careful attempt to widen the parameter space — still finds constraints, just weaker ones than the strongest claims suggest.

The honest summary is that the physics identifies real constraints whose severity is disputed and depends on how the question is set up.

"The cosmological constant problem shows physics is incomplete, not that it's tuned"

Largely correct, and this is the standard view among physicists.

The 10¹²⁰ figure is a comparison between the observed value and a naive quantum field theory estimate. Most physicists regard the estimate as wrong rather than the value as miraculous, since the calculation ignores possible cancellations from unknown physics.

Supersymmetry was expected to reduce the discrepancy substantially, though it has not been observed at accessible energies.

The observed value is still small and non-zero in a way no theory predicts, so something requires explanation. Whether "fine-tuned" is the right description of an unsolved calculation is the point at issue.

"If the constants were different, different life would evolve"

A common reply, and it has limited reach in the extreme cases.

Some variations produce universes with no atoms at all, or no structure larger than elementary particles. It is not that carbon-based chemistry fails while something else succeeds; there is nothing complex of any kind for an alternative biochemistry to be made from.

The reply is stronger for moderate variations, where different chemistry or different stellar physics might still permit complexity. Assessing this requires knowing what forms complexity could take, which is not known.

"The multiverse is untestable, so it's no better than design"

Fair as stated for direct observation, and physicists disagree among themselves about whether multiverse proposals are science.

The distinction is that eternal inflation makes other testable predictions — about the CMB, the geometry of space, and the spectrum of primordial fluctuations — and the multiverse follows from it rather than being posited separately. Searches for collision signatures in the CMB have been conducted, and found nothing, which is at least an attempted test.

Whether that is sufficient is genuinely disputed. The comparison to design is a philosophical question rather than a physical one.

What the physics shows#

Several constants and initial conditions fall in ranges that permit complex structure, and varying them substantially produces universes without stars, atoms, or chemistry. The constraints are real and are computed from standard physics.

The severity of the constraints is disputed. Published sensitivities usually vary one parameter at a time, and joint variation widens the permitted region considerably in the stellar case.

Converting a range into a probability requires knowing what values were possible and how likely each was, and neither is known.

The inference from these constraints to a designer is a separate argument, examined under the teleological arguments in Philosophy. This page covers what the physics establishes, which is that the constraints exist and that their tightness is a live scientific question.