The selected answer on this question provides an extremely good, and detailed, description of what is generally accepted to be the main (only?) way a terrestrial planet can form, i.e. through accretion of the gaseous disk around a forming star.

My question is: Assuming that some discovery is made that concludes that this accretion process is false (this 'hand-waved' discovery is for the sake of argument, and it need not have any other bearing on our understanding of physics beyond this unique aspect of planet formation, though it can have other repercussions on known physics if it helps generate an answer), no planets, ever, anywhere in our known universe, were created using this accretion method, what is (are?) the next most likely method(s?) of planetary creation?

Best answers will use a minimum of change to known physics whenever possible, and those changes need not be described or explained. I'm interested mainly in identifying the process itself, rather than why one process doesn't work and a new one does, I just want to know what the new process is. In other words, hand-waving known physics is allowed, but should be avoided as much as possible.

Answers do not need to consider habitability of planets, except to the extent that we know of at least 1 (Earth) in the universe that ended up habitable due to this planetary formation process. Likewise, answers should not be limited to a process that obviously excludes the possible formation of any planet that we know to have formed in reality (don't describe a process that can only account for gas giants, but cannot produce terrestrials, or vice versa).

  • 2
    $\begingroup$ "that concludes that this accretion process is false" - problem is, we actually see protoplanetary disks at various stages. $\endgroup$
    – Mołot
    Sep 27, 2018 at 17:35
  • $\begingroup$ To "hand wave" away the accretion process, you either need to take away some real-world physics or add some fictitious physics or magic. What you take away or add will influence/define what answers are available, so I think you need to give a reason why the current planet-forming method wouldn't work. So...what happens to all the gas and dust in a star's accretion disk, such that it doesn't form planets...? $\endgroup$
    – Qami
    Sep 27, 2018 at 17:36
  • $\begingroup$ @Mołot : Yes, the "Hand-Wave" reference was meant to allow answers to ignore those observations completely, or explain them away as they see fit. $\endgroup$
    – Harthag
    Sep 27, 2018 at 17:47
  • 1
    $\begingroup$ @Ummdustry If we get philosophical, one could argue that the current accretion process might actually be the result of divine intervention itself, barring evidence to the contrary, such evidence not being possible for a being that is not divine themselves (this is simply the "Any sufficiently advanced technology is indistinguishable from magic" concept taken to the extreme). $\endgroup$
    – Harthag
    Sep 27, 2018 at 18:02
  • 2
    $\begingroup$ @Dalila - You might want to explain why "the planets were ordered on Amazon" is an invalid answer. $\endgroup$ Sep 27, 2018 at 18:06

3 Answers 3


Gravity is not "generated" by mass, instead mass gathers in gravity wells.

This is from wikipedia:

A gravity well or gravitational well is a conceptual model of the gravitational field surrounding a body in space – the more massive the body, the deeper and more extensive the gravity well associated with it.

But what if scientists "found out" that the gravitational field is just there and matter just accumulates in the "dips" the same way rainwater forms puddles and lakes. Stars form where the gravity well is deep and planets form where the well is shallower, but it can only accumulate matter until it "reaches the well's rim" and then the overflow will go to different dips. (And where there isn't enough surrounding matter, we get those unexplained gravitational effects which scientists today call "dark matter".)

I'm not sure this answers the whole question. For example, I have no idea how scientists would go about discovering this or disproving the accretion model. Or even if it would actually disprove it and not just exist alongside it. And I definitely don't know enough physics to tell you how big a handwave you'd need to create this theory.

  • $\begingroup$ It answers the whole question, yes. It's a described process that is decidedly different from current explanations, and the 'hand-wave' is just a new understanding of existing physics, not even an addition or subtraction from them. I like it, thanks. $\endgroup$
    – Harthag
    Sep 28, 2018 at 12:49

We can go with an old disproven explanation that I found in an outdated science textbook I saw in high school:

Star Farts

During star formation, as the star's spin increases, it throws off matter for its equator in bursts. The heavier elements don't travel as far and form the inner rocky planets. The lighter elements travel farther, forming the gas giants.

(quote block to indicate that this thought does not come from me)

This same textbook said that Jupiter had only four moons.

The sad thing is that they were still teaching the non-AP science class with that book.

No wonder why there are flat Earthers with that kind of teaching.

  • $\begingroup$ Amalthea was discovered in 1892. I guess it was a very old textbook :) $\endgroup$
    – Alexander
    Sep 27, 2018 at 18:09
  • $\begingroup$ @Alexander, No, it wasn't that old! (young'uns have no respect for their elders...) That was the sad part. That textbook was in use in 1979. There, now everyone knows how old I am. $\endgroup$
    – ShadoCat
    Sep 27, 2018 at 18:18
  • $\begingroup$ +1 just for having the chutzpah to write the words "star farts" in an answer. $\endgroup$
    – JBH
    Sep 27, 2018 at 19:03

Gravitational instabilities

There's actually a second idea for planet formation that's been around for some time (see Kuiper 1951). It requires the nebula hypothesis, like accretion, but it's a top-down process, not a bottom-up one, involving the fragmentation of accretion disks around young stars to form giant planets.

Protoplanetary disks can be unstable under specific conditions. In certain regions of overdensities inside a disk, clouds of gas may fragment into gaseous protoplanets, soon to evolve to become gas giants. The timescales of formation via this method are much shorter timescales than those required by the accretion hypothesis by an order of magnitude or so - tens or hundreds of thousands of years,

Fragmentation would likely happen farther out in a protoplanetary disk (Boss 1997, and volatiles like ice would form part of the core. The hotter inner portions have been pretty firmly ruled out as potential sites. Current views hold that fragmentation is certainly a possibility, and may even account for the formation of planets with large semi-major axes (including, I believe, the system around HR 8799, but it is unlikely that it contributes much to the general exoplanet population. Nonetheless, it requires no changes to the laws of physics.

Carbon planets and binary accretion

Another set of methods involve interactions between a star and a companion object, where the star accretes much of the companion's outer layers, leaving a remnant that may appear to be a gas giant or a carbon planet, depending on the precise mechanism and the companion's original form:

  • EF Eridani B may be a star that had much of its outer layers siphoned off by its companion star, leaving an object that might be about as massive as a massive giant planet, but isn't a brown dwarf.
  • PSR J1719-1438 b, the "diamond planet", may be a very low-mass white dwarf that was mostly accreted by its companion (see Bailes et al. 2011), leaving a dense object that would appear to be a carbon planet - terrestrial, but far from habitable.

These processes can form either terrestrial or giant planets - an advantage the instability model doesn't have. Another upside is that we have solid candidates for "planets" that formed this way. However, these pathways aren't conducive to forming habitable worlds (although I noticed you didn't require habitability).


You must log in to answer this question.

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