What is a galaxy? (Vote now!)

Article written: 22 Jan , 2011
Updated: 24 Dec , 2015


Classification is key to all sciences, but can often cause debate. Within astronomy, fierce debates have raged over the definition of a planet, both on the low-mass end, as well as the high-mass end. A recent paper explores definitions on a larger scale, pondering the definition of a galaxy, particularly, what separates the smallest of galaxies, the dwarf galaxies, from star clusters.

A working definition for dwarf galaxies was proposed in 1994 based on the brightness of the object in question as well as it’s size. For brightness, the cutoff was taken to be an absolute magnitude (MB) of -16. The size would need to be “more extended than a globular cluster.”

As with many definitions, they seem to work initially, but as new technology became available, objects were discovered around the cutoff line, blurring the distinction. These objects, which were first discovered in the late 90’s, are generally referred to with names like “ultra-faint dwarf spheroidals” (dSphs) and “ultra compact dwarfs” (UCDs). Regarding these small fragments, a 2007 study noted that they may “contain so few stars that they can be fainter than a single bright star and contain less stellar mass than some globular clusters”.

To help reconsider the definition of a galaxy, the authors looked at several commonly used criteria that have been applied (often inconsistently) to these questionable cases previously. This included requirements that the system be gravitationally bound, which would keep stellar streams and other ejected objects from being considered galaxies in their own right. Obviously, most galaxies will slowly bleed away stars due to random interactions, giving rise to hypervelocity stars which will leave the galaxy, so the team proposes a threshold that the galaxy have a “relaxation time” greater than the age of the universe. This would allow dSphs and UCDs to be considered galaxies, but would keep out objects that have generally been considered globular clusters.

Another proposed constraint is based on the size of the object. The team proposes a cutoff where the effective radius be greater than or equal to 100 parsecs. This cutoff would exclude dSphs and UCDs.

The types of stars is another consideration proposed since this can be used to achieve somewhat of an understanding of the history of the object. While clusters usually form in a single instance, galaxies are generally considered to have their own, internal machinations leading to complex stellar populations. Thus, the presence of multiple populations of stars. This would include dSphs and UCDs, but may allow some globular clusters to slip in as well since studies have shown that some of our more massive globular clusters in the Milky Way have interacted with gas clouds, triggering star formation which was absorbed by the clusters.

Dark matter is another criteria that is examined. Since galaxies are proposed to form within dark matter halos and be intrinsically tied into them, the requirement that dark matter be present would fit well with the theory. However, this criteria also poses many difficulties. Firstly, measuring the presence of dark matter in small objects is a challenging task. It is also questionable as to whether or not dSphs and UCDs would contain dark matter as a general rule since their formation is not well understood and the possibility remains that they may have been ejected from our own galaxy during formation and recoalesced, possibly without a dark matter halo.

The last possible criteria is much along the same lines as the nebulous definition for planets that they dominate the local gravitational field. The team considers the possibility that objects would be required to have stellar satellite systems as globular clusters of their own. This would include some dwarf galaxies, but may exclude others.

Even with many of these criteria, classification will still be a treacherous issue. Objects like Omega Centauri may fit some definitions but not others. According to the paper’s lead author, Duncan Forbes, “many amateur astronomers know Omega Cen as massive star cluster, some professional astronomers regard it as a galaxy. This is a stellar system that could be upgraded or downgraded by this exercise, depending on your point of view.”

To help gather opinions on the topic, the authors have set up an online survey to gather opinions on this definition and hope to reach a satisfactory conclusion by collective wisdom. This poll is open to the general public and results will be presented at a future astronomical conferences allowing participants to help take part in the astronomical process. Forbes hopes that this public interaction will help garner public interest in much the same way as the Galaxy Zoo project has.


28 Responses

  1. vagueofgodalming says

    If they think they’re going to gin up a Pluto level of controversy on this, they’ve got another think coming.

  2. antoniseb says

    This is probably not as potentially controversial as whether Pluto is a planet.

    I think that *IF* an object contains a super-massive black hole, it is a galaxy. If not it is either a proto-galaxy (if it will eventually have an SMBH), or it is a star-cluster.

    I choose this definition to identify galaxies based on how they are formed. I don’t want to include tidal debris from galaxy collisions.

    • Excalibur says

      M33 is suspected to not have a SMBH, yet i believe it should be considered a galaxy. Im not arguing with your attempt at a definition, I understand what you are trying to achieve, i’m just pointing out that it can get tricky.

      • Lawrence B. Crowell says

        A central supermassive black hole is not important, but a dark matter halo is. A galaxy without a SMBH might have a low rate of starburst, which might actually prolong its stellar life, but it is still a galaxy.


  3. Greg says

    It is no surprise to me that the IAU made these criteria rather hastily. A simple way to assess how good the classification crtieria are is to look at extreme known objects and see if are properly designated. In the case of the definition of a minor dwarf planet, a Jupiter mass object can be classified as a dwarf if it is in such a distant orbit (such as in the oort cloud) that it cannot possibly clear its orbit. There is no other way to resolve that conundrum than to add an additional criterion to the classifiication. (I suggested a hard diameter cutoff, above which an object is classified a planet)
    Similarly with this question I have my reasons for my choice. Naturally I would include the current criteria that a galaxy be gravitationally bound and contain stars. This eliminates the controversial and rarely sighted “dark matter galaxies.” I agree with using a two body relaxation time greater than the age of the universe as it eliminates globular clusters. I like the >100 parsec criterion in much the same way as I like applying it to the definition of a planet. This knocks out ultra compact dwarfs which I don’t believe deserve galaxy status.
    I don’t like using the presence of complex stellar populations, since that will not stand the test of time. Many billions of years down the road all current galaxies will have stopped churning out complex stellar populations. Using the presence of non-baryonic dark matter is almost ludicrous considering that we have not had a confirmed detection/measurement of it. Also it is inevitable that there will be some galaxies where dark matter or the effect we attribute to it (meaning impossibly stellar rotational velocities) is somehow absent. Using similar logic to the last two, I would also exclude having satellite galaxies as inclusion criteria.

  4. Hon. Salacious B. Crumb says

    Greg here somewhat misses the whole point.
    There is only one galaxy called “The Galaxy”, being the Milky Way. Historically it is called this after the Galactic Circle, and was named from the Greek word first given by Hipparchus of Rhodes. In reality it means just a “system of stars” which was fine until it was realised that the extragalactic nebulae were of similar to the structure of our Milky Way. The problem is we are using antiquated notions applied to astronomical objects, that were named, then found to be different from what was originally thought. (Another example is the term now applied to planetary nebulae.)

    So in the end, all the rest of the ‘galaxies’ are merely imitators of ours.
    These small objects, perhaps should be called sub-galaxies, which could include these “ultra-faint dwarf spheroidals” (dSphs) and “ultra compact dwarfs” (UCDs), and even the globular star clusters. They are defined by their size (<100pc.), and being objects directly associated with larger collection of galaxies.

    IMO there is only one solution, is that we have discovered most of the kinds of objects in the Universe. Perhaps it is time to toss all the old naming systems, and redefine the system from the top down. In the end at least this would be less of a hotchpotch of system that it is now. The new system should be based on dynamics

  5. SteveZodiac says

    The word galaxy derives from the Greek term for our own galaxy, galaxias (????????), or kyklos galaktikos, meaning “milky circle” (Wikipaedia) Just like “planet” came from a word meaning “wanderer”, it was originally a convenient, descriptive word from the undertanding at the time. Trying to plot the precise scientific boundaries of such old words suggests to me a certain reversal of astronomy where the navel is more important than the cosmos. You only have to look at the lessons of history to see what happens if you let people with this attitude take over your organisation.

    • Hon. Salacious B. Crumb says

      “The word galaxy derives from the Greek term for our own galaxy, galaxias (????????), or kyklos galaktikos, meaning “milky circle” ”

      This actually comes paraphrased from Richard Allen “Star Names: Their Lore and Meaning” of 1899. pg. 474.

      As for ; “Trying to plot the precise scientific boundaries of such old words suggests to me a certain reversal of astronomy where the navel is more important than the cosmos” is completely wrong.

      Words are in context of something, and understanding the derivation gives some idea of the thoughts going on at the time. The use of the units and types of celestial objects has evolved over time. These older notions haven’t changed, and the word ‘galaxy’ will be still used in the vernacular. However in science descriptive words have less importance in meaning but are needed to be categorised by a coherent system of systematised nomenclature. (It is one of the roles of the IAU and the reason for its existence.) [The best example are planetary nebulae, which have nothing to do with planets. A better name would be, say, neons, because of the process of fluorescence make the characteristics of the nebulosity. It is both descriptive and accurate.]

      In the end, perhaps there should be a similar system like the diversity of biological animal kingdom. A common name and a Latinised name, etc, described in some taxonomic order. I.e. Sizes of the universe down to atoms. (at least in the menagerie of astronomy objects novices might have a better concept of where celestial bodies fit into the scheme of things. It may also clear up the confusion.)

      Admittedly, this is unlikely to happen, but there is no reason to think more logically with what currently amounts to a huge jigsaw puzzle whose pieces on their own make relatively no sense.

      • SteveZodiac says

        “Words are in context of something, and understanding the derivation gives some idea of the thoughts going on at the time.” Words are convenient symbols for entire concepts and it is human nature to refer to a new or altered concept using words for an existing one, some stay that way, others mutate or are replaced. “Horseless carriage” springs to mind along with computer “files” .The biological taxonomy was what I had in mind for concise linguistic labels for the concept of various aggregations of stars but, like you, I doubt it will happen. I still think that this is in the realms of how many angels can dance on the head of a pin (however dubious the origin of that expression) and that energies are better spent elsewhere.

  6. Hon. Salacious B. Crumb says

    Did you paraphrase your words from this in SciLogs running the survey? I.e.;

    “The word “galaxy” comes from the Greek word for “milky circle”. The discovery of galaxies as extra-galactic objects is a fascinating part of science history and is outlined for example in the Wikipedia article on Galaxy.”

    This also suggests in the Swinburne Press Release; “While the term galaxy is bandied around regularly by scientists and the broader community, no clear definition exists for what constitutes a galaxy.” This is also not quite true. I.e. A galaxy is an object structurally similar to the Milky Way.

    • SteveZodiac says

      No, I Googled “etymology galaxy” and decided wikipaedia was reliable enough, I had always thought it just meant “milky” a memory from my organic chemistry days of playing with Galactose

  7. Hon. Salacious B. Crumb says

    Galaxies should firstly include this as the primary criteria, then divided by secondary considerations, like population, size, dynamics, evolution, etc. Thinking about it, they should be called sub-galaxies in order to avoid confusion by non-scientists. The whole division should be based on dynamics of these systems, similar to the stellar motions with globular star clusters (random movement of stellar orbits) with galaxies (having organised structures and stellar motions.)
    So. Are all elliptical galaxies like all other galaxies or are they like globulars?

  8. Torbjorn Larsson OM says

    I have learned to look at biology for this, because they have a serious basic and age old problem in classifications (naming), relationships (here, phylogeny) and populations (species problem).

    What biologists do is to separate classifications, which may be artificial and convenient, from natural relationships which follows from the processes the populations obey.

    Having that analog, and being clear that it is an analogy, much of the same happens in astronomy. I must then agree with HSBC when he proposes that dynamics (processes) could and, looking at how it works rather well in biology should, decide.

    First, we can look at the more clear (I take it) status of planets. Obviously IAU made a correct choice out of the possible ones there, since the “clearing the neighborhood” criteria separates out a natural population (planets) from others (belt objects). This is in the same way that a “biology species” definition would pick out eukaryote species of the 26+ possible species conceptualization candidates.

    And as for species conceptualizations there is a fuzziness since the processes which separates populations act gradually, so there _should_ be funny cases or it isn’t nature. (Say, planets-to-be while undergoing clearing).

    Second, we can look at that list of galaxy criteria for processes naturally resulting in (for the purpose) distinct populations. Gravitation results in several populations, so isn’t enough. A dark matter halo is a possible criteria, it fails in some cases as expected. The so called “nebulous” definition of “clearing the neighborhood” seems like a keeper, since galaxies interact and coalesce while growing; and it fails in some cases as it should.

    Since the latter criteria seems like a candidate for a more general astronomical relationship classification, looking at behavior of, and processes resulting in, sufficiently isolated systems, I will go vote for that one.

    [And, I note, IAU “for the win”! Are astronomers good at what they are doing, or what!? :-D]

  9. HeadAroundU says

    I think it should be harder for an object to become a galaxy. In the end, we live in a serious galaxy, so why not to be a bit haughty.

    So, I voted for >100 pc and dark matter.

    • Hon. Salacious B. Crumb says

      My apologies if my comments appear to be superior or disdainful. It was not my intent.
      Did you mean naughty, instead?

      • HeadAroundU says


        No, but now I see it’s close to N on keyboard. Originally, I wanted to use snobbish but since my name is HAU…

        Anyway, Milky Way is a hottie galaxy.

      • SteveZodiac says

        So we call them all “MilkyWays” eh? or go latin and use “Vialactea”. I like your chauvinistic (galacto-centric?) streak but the trouble is “way” is used as in the concept of “path” which encompasses the ignorance of those looking at only part of a very large disk and seeing it as a path in the sky, I can’t see the IAU accepting that.

      • HeadAroundU says

        No idea what you’re talking about. Looks like a strawman. Why we are talking about the name of our galaxy? 😀

        But if you have a problem with the name, we can call it Milky Spiral or Chocolate Cookie. Dark matter is from chocolate. 😀

  10. SteveZodiac says

    Hmm looking at the Forbes and Kroupa paper I now see the mess created by ad-hoc pigeon-holing of the universe – sorry, cosmos, no, don’t go there……
    @SBC I may have to eat some words when I vote 🙂

  11. Member
    Jon Voisey says

    I’m glad so many people are discussing this seriously. Internet polls are serios bizness.

  12. shifty says

    When i think of a galaxy i think about stars rotating a center this be a (super masive) black hole or bigger singularities.
    But the real definition may just be a collection of stars that make up solar systems and star clusters wether it rotates around a circle or not.
    I also think it being held together by gravity and dark matter being present.

    • Greg says

      Not every galaxy has a supermassive black hole which I agree with leaving that criteria out. For Example:

      As for HSBC: The question from the article: http://arxiv.org/PS_cache/arxiv/pdf/1101/1101.3309v1.pdf
      is how to classify a galaxy, not whether to call it a galaxy or something else. The frame of reference, of course, is the system of stars that surrounds us. That is as far as the historical references should go. Trying to preserve historical logic when trying to to define a galaxy to me borders on ridiculous considering the Greeks also thought those dust clouds that surround us came from a lactating Hera. So on this point I presume we agree.
      The classification system that the authors propose is in fact much like a Linnaean bilological system which you referred to. It is based on using visual characteristics, naturally since that is what astronomy is all about. Intriguingly the Linnaean system is under considerable and justifiable fire for inherently disregarding DNA evidence which can yield completely unexpected and conflicting results. Nevertheless I do not see such a problem coming to light for astronomy. If you wish to propose a new functional classification system, then you have quite a bit of work ahead to make sense of it and then get it accepted by everyone.

  13. shifty says

    This brings up defining when interstellar space starts.

  14. Member
    Aqua says

    I vote yes.. there are no doubt galaxies out there in any configuration, size or composition imaginable and probably some that are unimaginable.

    Short story subject: An extremely faint radio beacon is detected at the Arecibo telescope. It is determined that the signal contains a message from an alien civilization! Eventually the location of the transmitter is found. Apparently it is being emitted from a cluster of molecules located somewhere within the focal point/cone of the telescope! A galaxy far, far away, yet underfoot?

Comments are closed.