# Are multiple stars actually more common than singles?

In the book by Neil Tyson & Michael A. Strauss & J. Richard Gott titled "Welcome to the Universe: An Astrophysical Tour" there is a sentence that starts: "Since more than half the stars in the galaxy are in binaries, ... ". But is it really true?

The reason I cast some doubts is as follows. Light-weight stars happen to be in pairs (or otherwise multiple) with a lower chance than heavy stars: for a light-weight star the chance to be in a pair is estimated roughly about 20%, whereas for a heavy star the chance is about 80%. However it is also a well-known fact that light-weight stars represent the majority of all known stars in our universe. For instance, stars with masses up to 0.9 of solar mass (incl. M-G classes) represent 95% of all known stars. Even if we consider only the 30 nearest stars to our Sun, the binaries and trinals amongst them are only 13. In fact, the chance for a neutron star (they all are relatively heavy) to be a binary is estimated as 5% only.

So the question I have to ask is how valid was the statement from the book by Tyson & Strauss & Gott?

Thanks in advance.

• Are you counting per star, or per star system? If half the stars are in a binary system, then only 1/3rd of systems are binary, with the other 2/3rds being unary. Commented Dec 12, 2023 at 23:09

## 3 Answers

Too long for a comment:

There is one ambiguity you are going to have to deal with, because nobody agrees what “star” means.

Suppose that you see three points of light in the sky, and then with a telescope or good binoculars you see that one of those three points is a pair of stars very close together. If you like, name the three things you see as A B and CD.

Then both the following statements are true:

1. One in three stars you see in the sky is a binary: A, B, CD.
2. One in two stars is in a binary. A and B aren’t, C and D are.

Different writers will use either (1) or (2) depending on the impression they want to give you. Writers also copy from each other and some of them will not know or care which of the two approaches they are copying.

Thus some of the discrepancies you are in your sources may be linguistic, not real.

• The linguistics could maybe explain why someone would say more than half of stars are single. That is because it is true that more than half of star systems consist of single stars. I don't think that confusion can attach to someone claiming that more than half of stars are in binary systems, as here. Commented Dec 11, 2023 at 8:53

That's a good question with a non-intuitive answer. The main point is: one has to distinguish stellar systems from stars themselves. Consider an essemble of 30 stellar systems. 33% or 10 of them are double stars. Then we have overall 40 stars where 20 of them (50%) are stars in multiple systems.

The reality here is complicated. There is a review article by Gaspard Duchêne and Adam Kraus on stellar multiplicity. They provide a table of the frequency of multiplicity:

• $$< 0.1$$ solar masses: 22%
• $$0.1 - 0.5$$ solar masses: 26%
• $$0.7 - 1.3$$ solar masses: 44%
• $$1.5 - 5$$ solar masses: >50%
• $$8 - 16$$ solar masses: >60%
• $$>16$$ solar masses: > 80%

These probabilities have to be folded with the initial mass function to see how much these probablities impact the overall number. While I didn't find a paper which does this exactly, looking at the IMF diagramme, and the fact, that multiplicity is already above the threshold of 33% for masses > ~0.5 solar masses, the majority of stars will be in multiple systems while most systems are single star systems. This argument also so far did not consider multiplicity higher than two - which further helps this argument.

However, there is the RECONS census from 2018 of the solar neighbourhood within the nearest 10 parsec. This sample consists of 317 systems, which of 232 are found to be single and overall 378 stars and 50 brown dwarfs are found. The numbers from this survey give us that there are 232/(378+50) = 54% single stars (or brown dwarfs). There is an update regarding this sample using the Gaia data from 2021 by Reyle et al which conclude

Almost half of the stars and brown dwarfs are in multiple systems. As summarised in the bottom part of Table 3, our 10 pc sample contains 246 single, 69 double, 19 triple, three quadruple, and two quintuple systems

Thus the thorough sample, and assuming that our solar neighbourhood is representative, gives us the idea that it might possibly be so that both, most stars systems are single systems and that about half of the stars are single stars. Surveys for a slightly larger range (but less rigorous) point into the same direction. Thus it might actually be that the initial assumption of half the stars in the Galaxy being binaries does not quite hold. Instead likely both, most stars and most stellar systems, might be single.

• But the peak of the IMF is at around 0.25 solar masses. Commented Dec 11, 2023 at 8:49
• @terdon Say the multiplicies of the systems are {1, 1, 98}. Most stars (98 out of 100) are in a multiple star system but most systems (2 out of 3) are single. The real world isn't as extreme as that of course but the same thing applies. Commented Dec 11, 2023 at 16:23
• @terdon read my first paragraph, I try to illustrate that problem there with an example. Or similar: Take 30 stellar systems. Make 15 systems double star systems. Then you have 15 single stars systems for 15 stars and 15 double star systems for 2*15 stars = 30 stars, for a total of 45 stars. Thus 50% of the star systems are single star systems, but 2/3 (30 of 45) of the stars are in multiple systems Commented Dec 11, 2023 at 16:41
• According to the RECONS census of 2015, reported at meeting #227 of the American Astronomical Society, of 270 'primary' stars within 10 parsecs, 28% have companions that are stars and 2% have those that are brown dwarfs. This may have changed a little with more recent data. 366 stars were found, so that tells me that just about half of the stars within 10 parsecs are single stars. This, of course, is weighted toward lower mass stars, so to determine the numbers for F,G and K stars you have to go to farther distances to get a better sample. Commented Dec 11, 2023 at 20:54
• @JackR.Woods that is the answer (within the assumption that there isn't anything special about the solar neighbourhood). Commented Dec 13, 2023 at 8:38

You might also like to consider the probability that a number of the singleton stars that are visible might have once been part of a binary system and have devoured their companion. There may also be a number of singleton stars that are going to be part of a binary system in the future.

• This doesn't really answer the question but speculates on how multiple star systems form. That said, capture of stars to form multiple systems is nearly impossible considering conservation of momentum. Commented Dec 16, 2023 at 13:16