Definition of a Planet

(I wrote this article back in 2007. It was written in response to the astronomical debate the previous year, which resulted in the IAU "demoting" Pluto to a dwarf planet. While that in itself did not bother me too much, the definition of a planet that the IAU produced did. I decided to shed some logic on the subject.)

The next time you want to organise a classification system for something, for goodness sake, ask a programmer! I mean, this is what we do.

In object-oriented programming and database design, we organise objects according to their attributes, or the specific information that's connected to or describes that object.  When you approach the problem this way, it becomes very simple.

Discounting specks of dust, clouds of interstellar gas, artificial satellites, spacecraft and black holes, amongst the common objects in the Universe there are 3 easily distinguishable groups:

Category A:  Lumpy objects that are not large enough for gravity to pull them into a roughly spheroidal shape (hydrostatic equilibrium).

Category B:  Objects that are large enough for hydrostatic equilibrium, but not yet large enough to have achieved stellar nucleosynthesis (fusion).

Category C:  Objects that are large enough for stellar nucleosynthesis to occur.

These categories are clearly defined by mass. No other property is needed. It is very easy to look around the Universe and say what category any particular object would belong to, just by looking at it. This is how a child would organise objects in the Universe. If it gives off light and heat, then it's a Category C; if it doesn't, but it's round, then it's a Category B, and if it's lumpy then it's a Category A. Very simple.

If these categories were represented in a computer program, they would each require different information to describe them. Category A would require 3 diameters, whereas B and C would require only 2. Category C would require a "luminosity" attribute, which A & B wouldn't.

Further categories, also defined by mass and gravity, could possibly be defined for neutron stars and black holes. But for this exercise I am interested in objects that fit into category B, which I use as the definition for a planet:

Planet: Any astronomical body massive enough to have achieved a roughly spheroidal shape due to hydrostatic equilibrium, but not massive enough to have achieved stellar nucleosynthesis.

Put simply, a planet is a round thing that doesn't shine.

The main point is that the definition has nothing whatsoever to do with characteristics of the object's orbit. This is simply because an object's orbit can change without the object itself changing. The definition is purely based on mass.

According to this definition, "planets" would therefore include terrestrial planets, gas giants, dwarf planets, round moons, and orphan planets (planet-like objects without a parent star). It would not include objects from the category known as "minor planets", which are more commonly called asteroids, planetoids, trojans, Kuiper Belt Objects (or Trans-Neptunian Objects) or comets.


There are various types of "planets" to examine.

Dwarf planets

We don't need a category called "dwarf planet".  However, if we decide that we do need one then it should be based on the planet's size, and nothing to do with its orbital characteristics.

"Dwarf planets" are currently defined by the following four rules:

  1. In orbit around the Sun.
  2. Has sufficient mass to have assumed a near-spherical shape (hydrostatic equilibrium).
  3. Has not cleared the neighbourhood around its orbit.
  4. Is not a satellite.

According to this definition, dwarf planets can only be found in orbit around Sol, which therefore limits the life expectancy of this definition up until we discover similar objects around neighbouring stars. Rules 1 and 4 should both be discarded. It isn't useful to classify an object according to what it orbits. If an object is captured such that it orbits something different, there would be no reason to change its category, since the object itself hasn't changed. No new attributes are required to describe it. Was Triton a planet that then became a moon (or satellite) when captured into Neptune's orbit?

Rule 3 is highly ambiguous since even Jupiter hasn't cleared the neighbourhood around its orbit, hence the existence of trojan asteroids.  According to the current definition, Jupiter is technically a dwarf planet!

Nonetheless, Rule 3 could perhaps not be discarded but clarified. Dwarf planets such as Ceres and Pluto do appear to be clearly different from other larger planets in the Solar System, since they are much smaller, and share their orbits with many thousands of other objects.

The question is this – how clear does the orbit have to be for the object to be considered a major planet rather than a dwarf planet? We could define a torus based on the object's orbit, then count up all the other objects whose orbits are either partially or wholly within this torus. The total mass of the orbit could be calculated by summing the masses of the objects that use this orbit, and then the mass of the planet-like object in question could be calculated as a percentage of this total. If this percentage is greater than some cut-off, say, 90% or 99%, then we might call the object a major planet – otherwise it would be a dwarf planet.

The only problem with this idea is that it's complicated, not very scientific, and requires the collection of a lot of data. It's based on an arbitrary, human selected cut-off, but not on any specific attributes of the object itself.

Another method of defining a dwarf planet could be by size, for example, we could say that any object with a diameter less than 2000km is a dwarf planet. But again, this is an arbitrary cut-off with no scientific basis.

Two of the things that distinguish Pluto from the major planets are its orbital eccentricity and orbital inclination, which are much greater than those of major planets. But even if you said "only objects with an orbital inclination less than 5°, or with an orbital eccentricity less than 0.1, can be considered a major planet" – again, these are arbitrarily-selected cut-offs.

Enter the database designer or object-oriented programmer.  A database table designed to store information on major planets would store information on diameter, parent object, aphelion, perihelion, orbital eccentricity, orbital inclination and so on. Now, let's say the same database now has to store information on dwarf planets. Does the database designer need to create a new table? No, because the attributes for dwarf planets are exactly the same.

Still not convinced?  Consider these scenarios:

  • As the result of a colossal mining effort, all the asteroids are removed from the main belt, leaving only Ceres. Is Ceres now promoted to "planet" or is it still a dwarf planet? According to the new definition, Ceres would now be considered a planet, even though nothing about Ceres has changed. Unless you specify that because Ceres didn't clear its own orbit (it was done by humans) then its still a dwarf planet. Gah!
  • What if our Solar System is still forming, and over the next 5 billion years, Ceres grows as it gradually collects most of the mass of the asteroid belt into itself. At what point does it become upgraded from dwarf to major planet? When it reaches a certain size, or when the orbit is cleared to a certain point?
  • What if we discover another star system which has a belt around it containing hundreds of round objects the size of Mercury, Venus, Earth and Mars. Would these be considered dwarf planets or major planets?  They're the same size as major planets in our star system, they just haven't cleared their orbit! Jupiter and Saturn's orbits could once have been like this.

Obviously, it doesn't make any sense at all to define a dwarf planet based on whether or not its orbit is shared with other objects. Even the adjective "dwarf" makes it clear – a dwarf planet is small, therefore its definition, if necessary (which it isn't), should be based on size.

One way to define a dwarf planet, then, could be like this: "an object massive enough to have achieved hydrostatic equilibrium, which has a mean radius less than 2000km".  It does not refer to any other objects or its orbital characteristics. It is a definition that will make sense in other star systems. This definition would include Ceres, Pluto, Charon and Eris, yet would exclude Mercury. Seems simple enough? Ah, but this definition also includes Luna, Triton, Io, Europa and other small round moons, while excluding Ganymede, Callisto and Titan. Somehow this doesn't seem right.

The only solution?  Discard the definition of "dwarf planet". We don't need it. It serves no-one except schoolchildren who can't count above 10. A planet is a planet.

Planet-sized Moons (Natural Satellites)

In the Solar System (and presumably other star systems), there are certain objects which orbit planets, yet which are massive enough to have achieved hydrostatic equilibrium and formed into a spheroidal shape. These should be called planets as well. At present there is no official definition for this category of objects – they are merely called satellites, as are any other type of objects orbiting a planet.

An object's classification should not be based on its orbit. Consider this scenario: a round moon is orbiting a planet about 13 times the mass of Jupiter. A comet impacts the planet, causing it to become a brown dwarf. Has the moon now become a planet because it's parent object no longer is? The object itself hasn't changed – only the object it orbits has.

What if the object is orbiting the Sun, then is captured by a gas giant – something which may have happened to Triton. Should it be demoted from planet to moon?

Surely an object's classification should not change unless it changes.

Binary Planets

Charon is commonly classified as a moon (or natural satellite) of Pluto. However, this is not really fair on little Charon, because technically speaking Charon does not orbit Pluto. Pluto and Charon both orbit their barycentre, or common centre of mass, which is a point in space that lies between them (Earth and the Moon orbit a barycentre as well, however, this point is located inside Earth). Pluto and Charon spin around each other like two figure skaters. It isn't reasonable to say that Pluto is the "planet" because it's bigger, and Charon is the "moon" because it's smaller. Pluto and Charon are therefore often referred to as a binary planet or double planet.

With this new definition, Pluto and Charon both qualify as planets and the terminology problem goes away.

Orphan Planets

These are planets that do not orbit a parent star. They just kind of wander through the Universe on least-energy paths defined by the curvature of spacetime. But they should still be called planets if they fit the definition. After all, one day they might get adopted (captured by a star) – and it would not make sense for the object's classification to change if the object itself had not changed.

Planets in the Solar System

According to this new definition, there are at least 30 planets in the Solar System. Other potential candidates include Vesta, Pallas, Hygiea, Orcus, Sedna, Quaoar, Varuna, Ixion, and a few others – it remains to be seen whether these objects have achieved hydrostatic equilibrium.

# Name Parent Equatorial Diameter (km)
1 Mercury Sol 4,879
2 Venus Sol 12,104
3 Earth Sol 12,756
4 Luna Earth 3,474
5 Mars Sol 6,805
6 Ceres Sol 975
7 Jupiter Sol 142,984
8 Io Jupiter 3,643
9 Europa Jupiter 3,122
10 Ganymede Jupiter 5,262
11 Callisto Jupiter 4,821
12 Saturn Sol 120,536
13 Mimas Saturn 397
14 Enceladus Saturn 504
15 Tethys Saturn 1,066
16 Dione Saturn 1,123
17 Rhea Saturn 1,529
18 Titan Saturn 5,151
19 Iapetus Saturn 1,472
20 Uranus Sol 51,118
21 Miranda Uranus 472
22 Ariel Uranus 1,158
23 Umbriel Uranus 1,169
24 Titania Uranus 1,578
25 Oberon Uranus 1,523
26 Neptune Sol 49,528
27 Triton Neptune 2,707
28 Pluto Sol 2,390
29 Charon Sol 1,206
30 Eris Sol 2,400


I like to read, write, teach, travel, code, lift weights, play music, listen to music, make things out of wood, watch scifi movies, and play board games and computer games. My interests are broad, spanning science, engineering, architecture, technology, nutrition, environment, psychology, health, fitness, finance, business, and economics, but my main passions are spirituality, space settlement, and veganism. My ambition is to be a successful writer and speaker, and to create a company to produce awesome science fiction books, movies, and games that inspire people about the future. Eventually, I would also like to create vegan cafes and urban farms.

Tagged with: , ,
Posted in Space

Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Follow Shaun's Blog on
Let’s connect!

You can reach me via any of the options below:

Email Facebook Twitter LinkedIn Instagram YouTube Reddit Quora GoodReads Amazon
Read my book!

It's free to download or read online, and very good. If you're into Mars exploration, you will love it.

You can also buy the Kindle version or a print copy.

Watch my TEDx talk
%d bloggers like this: