I am mostly active on Physics Stack Exchange but a question occurred to me that I think is best submitted here.

This question is about a thought experiment: whether the circumstances of the intended thought experiment are physically possible. For this thought experiment it's not necessary for the circumstances to be physically possible, but it would be nice bonus.

In the history of astronomy the fact that the fixed stars were available as reference of celestial motion was hugely helpful.

For example: my understanding is that Kepler's approach to finding the shape of the orbit of Mars was as follows. From astronomical records extending over a long time the period of the orbit of Mars can be reconstructed. Kepler would take pairs of Mars position observations, spaced one Mars year apart. These pairs of Mars positions were from two different Earth positions, allowing Kepler to triangulate.

The intended thought experiment:
A solar system in a region of its galaxy with such a density of interstellar gas that all the stars are obscured. Without the fixed stars as reference of celestial motion the Kepler problem would be much harder.

(The purpose of this thought experiment is to argue that while it would have been harder to obtain the inverse square law of gravity, it would not be impossible. This can then serve as demonstration as to what theory of motion is.)

Some of my guesses:
My understanding is that all parts of the galaxy can be observed, it's just that for some parts the observed luminosity of the stars is diminished because the light coming from those stars traveled through regions with interstellar gas. That would make it very unlikely that interstellar gas can be dense enough to obscure absolutely all stars.

That brings me to guessing that this thought experiment would require rather contrived circumstances. For instance, maybe it would require a star to be out in intergalactic space. That would mean that star has had the misfortune, in its distant past, of being ejected from its originating galaxy.

[Later edit]

Assuming the planet is rotating and that the axis has an inclination with respect to the orbital plane:
Once the astronomers make the transition to a heliocentric model:
The astromers have the measurement of duration of the solar day, and from the fact that the duration of the year is constant over time they can, it seems to me, infer the ratio of the solar day to the intrinsic day. The astronomers can opt to use the line of intersection between the plane of the equator and the planet's orbital plane as a reference of zero rotation rate. (Of course, a rotating planet that has an inclination will undergo gyroscopic precession. However, the precession of the equinoxes is slow enough that on the assumption that the reference is stationary highly fertile science will be possible.)

It seems to me that once such a reference system is in place a Kepler has the opportunity to find Kepler's laws of planetary motion.

The purpose of the thought experiment is to demonstrate this: the very motion of the planets of a solar system is sufficient to establish a reference of motion. I submit that the conservation of the angular momentum of a rotating planet is sufficient to provide a solar system wide reference of orientation. Over time astronomers can iterate to an ever more accurate reference of motion.

Motion itself is sufficient to provide reference of motion.

  • 3
    $\begingroup$ Didn't Isaac Asimov write a famous story on this premise? $\endgroup$
    – davidbak
    Commented Feb 28, 2021 at 20:37
  • 4
    $\begingroup$ So did Douglas Adams (Krikkit) $\endgroup$
    – James K
    Commented Feb 28, 2021 at 21:31
  • 2
    $\begingroup$ The Asimov story @davidbak mentions is Nightfall, about a world constantly illuminated by at least one of the suns of its multiple star system. Without night, they don't see any other stars. Could this happen? Possibly: space.com/40234-alien-planet-with-no-nightfall-kalgash.html $\endgroup$ Commented Mar 1, 2021 at 9:35
  • 3
    $\begingroup$ Niven & Pournelle wrote "Mote in God's Eye" where a planetary system, which has evolved intelligent life, is in a gas cloud that obscures the rest of the universe. $\endgroup$ Commented Mar 1, 2021 at 13:12
  • 2
    $\begingroup$ Does it have to be the whole solar system? If it's just one planet, you could easily obscure the view of all stars just by having a very thick atmosphere. Venus would be an example of this - if you could survive on the surface of Venus long enough to look at the sky, you would not see any stars (except perhaps the Sun) through its atmosphere. $\endgroup$ Commented Mar 1, 2021 at 20:35

3 Answers 3


A planet with no stars?

Our solar system could certainly pass through a Bok Globule, which is a dense cloud of opaque interstellar dust and gas. They are typically about a light year across, so we could be stuck in one for quite some time. While this would obscure the other stars while we passed through, I don't think Earth would be a very good place to live during that time. Gas and dust falling into the Sun would probably strip or change our atmosphere and end life on Earth.

When stars form in a Bok Globule, the gas in the star's sphere of influence gets pulled into the interstellar disk and then either absorbed by the star and planets or ejected out of the stellar system. I suspect that a star couldn't form without leaving big holes in the globule.

But suppose life emerged on a tidally locked planet. If only the sun-facing side of the planet was warm enough to support life, but the dark facing side was fatally cold (even near the day/night transition zone), then the inhabitants might never see the other stars due to light pollution from their sun.

Another possibility could be that a planet was in a binary star system. If the primary star was much more massive than the other star, and the planet was in an L1 orbit with the less massive star around the more massive star, then even if the planet was turning, there would be no night. Here is a picture from wikipedia to illustrate the orbit.

enter image description here

Unfortunately the L1 orbit is unstable, so a planet in such an orbit wouldn't stay there for long.

On a planet in constant daylight, the other planets would also be difficult or impossible to see. Highly reflective gas giants in nearby orbits would certainly be visible, though.

Orbital calculations without a fixed celestial sphere?

It would be much harder to deduce a model for the motion of the planets without fixed stars in the back ground.

But say one has a single observatory, and the only things one measures are the times and azimuth angles for the high magnitude solar system bodies (the Sun, Moon, Mercury, Venus, Mars, Jupiter, and Saturn) as they ascend into the sky and descend out of it (for a fixed elevation angle).

If one gathers these for a few years, they will have enough data to have an overdetermined system to calculate the latitude of my observatory, the Earth's axial tilt, the Earth's rotation rate, and the orbital elements of all the bodies they can observe.

One could do it as Kepler did, by matching geometric shapes to the observations. Better yet, one could do it as Newton did, with a theory of gravity.

  • $\begingroup$ So a dust cloud is not a good option, as it will create a violent environment. On the issue of 'without a fixed celestial sphere': I added a discussion to the question, on how I expect a reference of motion can be established. $\endgroup$
    – Cleonis
    Commented Feb 28, 2021 at 20:51
  • 1
    $\begingroup$ Dust cloud IS a good option. You just need one of the larger, less dense clouds. The stated Bok Globule is very dense, it is actually a star trying to form. $\endgroup$
    – PcMan
    Commented Mar 1, 2021 at 14:40

The Coalsack nebula is likely a fairly typical dark nebula, reducing starlight by about 10 magnitudes over about 35 lightyears. A solar system in the middle of the nebula would hence get a reduction of outside stars by 5 magnitudes, which would hide much of the galaxy but would just reduce a bright star like Polaris to a faint one at the limit of unaided vision. While denser clouds likely exist, they are likely not opaque enough for the scenario unless they are very extensive (a galaxy like M64 may have big enough dust clouds to be more suitable, but there are going to be stars mixed in with them).

Runaway stars may be closer to the scenario. They do not have to undergo very radical accelerations (e.g. they could be part of a tidal tail ripped out by a passing galaxy), although the most noticeable ones have undergone some drama. If separated far enough into a void the distance modulus can bring the magnitude of all galaxies down to the level where you need a good telescope to see them.

  • 1
    $\begingroup$ For runaway stars: the universe is constantly expanding, aka the distance between galaxies is increasing (and therefore their stars dimming, when seen from a different galaxy). This could be used as an explanation as to why none of them are visible $\endgroup$
    – Hobbamok
    Commented Mar 1, 2021 at 11:28
  • $\begingroup$ Would the light of the parent star inside such nebula also create reflections (like when you wear a head lamp in a fog), drowning out even more of the outside star light? $\endgroup$ Commented Mar 2, 2021 at 0:26
  • $\begingroup$ @PaŭloEbermann - Possibly. Also, it may ionize gas to make it shine. Still, depending on what wavelength you look through the nebula may still be fairly transparent. $\endgroup$ Commented Mar 2, 2021 at 9:26
  • 1
    $\begingroup$ Comment on my acceptence of your answer: I will be using the scenario of a tidal tail star that is ripped out, and has had a couple of billion of years travelling out into intergalactic space. $\endgroup$
    – Cleonis
    Commented Mar 3, 2021 at 20:58

Runaway stars are stars with enough speed to escape the galaxy. It is possible for a runaway star to hold onto its system of planets, especially if they orbit close to the star.

Getting the dynamics just right that the process that ejected the star doesn't also sterilize any planets, and that the planets are not striped from the star in the ejection, is difficult. Runaway stars are not the first place that a SETI mission would consider.

However, such stars will travel outside the galaxy. Once the star is far enough away from the disk, no individual stars will be visible, only the hazy form of the milky way. Getting precise astronometric coordinates would be more difficult without any point sources of light.

  • 1
    $\begingroup$ As I feared: too contrived. To have a runaway star not lose it's planets would require the runaway state to be the end product of multiple mild gravity assists, rather like the series of gravity assists that conferred escape velocity to the Voyager space probes. But of course, the distances between stars are so large that even a single gravity assist is extremely unlikely to occur. $\endgroup$
    – Cleonis
    Commented Feb 28, 2021 at 20:57
  • 3
    $\begingroup$ The galaxies aren't obscured; and individual stars with a telescope as well. Last time I remember seeing the math, I think a moderate strength telescope can see from the middle of the largest voids we can see to the edge, and those voids probably have galaxies in them (just fewer of them) $\endgroup$ Commented Mar 1, 2021 at 1:58
  • 1
    $\begingroup$ @Yakk-AdamNevraumont true, but with the universe ever expanding, those other galaxies are dimming constantly. While working in hypotheticals, they therefoure could be far enough away already $\endgroup$
    – Hobbamok
    Commented Mar 1, 2021 at 11:27

You must log in to answer this question.

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