Universe family tree

An interactive toy for one question on the consciousness map: can selection work on whole universes?

Universes that make black holes have more children

Toy model, invented numbers

Lee Smolin's cosmological natural selection, reduced to its logic. Each universe has two made-up constants. The landscape sets how many black holes a universe makes; each black hole starts a child universe with its parent's constants, nudged at random. Watch which constants take over.

Seed

The family tree

Founding universes on the left, one column per generation. Each line joins a parent to a child.

Black holes per universe, population average

Each generation sits under its column in the tree. The dashed line is the most this landscape allows.

Where the constants are

Every universe in the chosen generation. Rings mark equal black-hole yield; the cross is the most. The white line traces the population average.

What Smolin claimed

  • In 1992 he proposed that black-hole singularities bounce into new universes, where the dimensionless parameters of particle physics and cosmology undergo small random changes. Universes whose parameters make more black holes leave more descendants. Smolin, "Did the Universe Evolve?", Class. Quantum Grav. 9, 173 (1992). doi.org/10.1088/0264-9381/9/1/016
  • The prediction that follows: if ours is a typical universe, its parameters sit near a local maximum of black-hole production, so changing them should tend to decrease the number of black holes. Same paper, abstract.
  • In 2006 he restated it as his master prediction: almost no change in the parameters of the standard model from the present values will increase the numbers of black holes produced. Smolin, "The status of cosmological natural selection" (2006). arxiv.org/abs/hep-th/0612185

The test, and what pulsars show

  • In 1997 he named a clean test: the observation of a pulsar with mass greater than 2.5 solar masses would cleanly refute the theory. He leaned on Bethe and Brown's upper limit of about 1.5 solar masses, and wrote that if one trusted it, any heavier value would be troubling. All well-measured neutron stars were then below 1.5. Smolin (1997). arxiv.org/abs/astro-ph/9712189
  • In 2006 he put the predicted upper limit at about 1.6 solar masses, for neutron stars made of kaon condensate. Smolin (2006), section 3.2. arxiv.org/abs/hep-th/0612185
  • Since then: PSR J0740+6620 at 2.08 ± 0.07 solar masses (Fonseca and others, 2021) and PSR J0952-0607 at 2.35 ± 0.17 (Romani and others, 2022). Both are above the 1.6 limit and below the 2.5 line Smolin called a clean refutation. arxiv.org/abs/2104.00880 · arxiv.org/abs/2207.05124

The main objections

  • Vilenkin (2006): The rate of black hole formation can be increased by increasing the value of the cosmological constant, which he says falsifies the claim that the constants maximise black-hole production. Smolin replied the same year that the argument is much weaker than may appear at first, because it applies laws tested on accessible time scales to scales vastly larger. arxiv.org/abs/hep-th/0610051 · reply in hep-th/0612185, section 4
  • Rothman and Ellis, "Smolin's natural selection hypothesis", Q. J. R. Astron. Soc. 34, 201 (1993), an early critique that Smolin cites. We could not retrieve its text for this figure, so its argument is not restated here; the map's entry for the theory summarises it. ui.adsabs.harvard.edu/abs/1993QJRAS..34..201R · the map's entry

How the toy maps onto the theory

  • Smolin's two hypotheses: singularities bounce into new universes, one per black hole, and each new universe's parameters differ from its parent's by a small random change. The toy assumes both and models neither. Smolin (1997), section 1; restated as hypotheses I and II in Smolin (2006).
  • Two peaks shows his own caveat: selection climbs to a local maximum, not necessarily the highest one. Flat is the control: with no difference in yield the constants only wander.

Smolin's claims, his neutron-star test, the pulsars measured since and the main objections, each with its source: open the full figure page.

What this toy does not show

  • No physics of the bounce. Whether a black hole can start a universe at all is assumed, not modelled.
  • Inheritance with small changes is the contested core of the theory, and here it is simply built in. Given heredity, variation and more children for more black holes, the climb is guaranteed by construction. The toy illustrates the argument; it cannot test it. Measured neutron-star masses, the test Smolin named, can.
  • The landscapes, the two constants and every number (one to three black holes per universe, the step sizes) are invented. Real yield depends on many parameters at once, and nobody knows its shape.
  • Only up to 120 universes per generation are followed, picked at random, so the tree stays drawable. Picking at random does not change the average mix of constants.