Late editing: Skip to the summary at the end of this post if you get bored while reading it.
Hi again,
Thanks for all your comments Yor_on. The scope of this thread is growing beyond anything I imagined. The links you provided are interesting but I can't respond to all of them today. I also think it will go far beyond what was originally described as a "school-level question" in my title.
So what are you thinking of ES?..... I thought you might refer to laws and properties but if it is something else it's time for you to state it, as nothing seems to fit?
It probably is time to say something about what I think and perhaps explain why I don't think gravitational potential energy is required.
1. Potential energy (of any kind) is something that tends to be invented (in the sense of given a name) whenever existing forms of energy are insufficient to account for the total energy in the system.
Alancalverd presented one example of this in his/her earlier posts. For simple Newtonian systems with gravity it is noted that
= constant.
and then the quantity mgh was called gravitational potential energy. [see reply #22 and #25 on this thread for more details]
There are many examples where the term "potential energy" is introduced just because it's convenient. It represents the most abstract part of a conserved quantity that exists in some systems. Specifically, since Total Energy is conserved, we tend to sum all the forms of energy that can be seen and then subtract this from the total energy. This "bit left over" is given the name "potential energy" but it is abstract and (in my opinion) quite misleading to suggest that this potential energy is any sort of fundamental or tangible thing that exists in the universe.
2. Historically, the existence of a conserved quantity that could be considered as energy was a conjecture (a generalisation or extrapolation) based on what we knew for only a small number of special cases.
3. The first mathematically rigorous proof that systems have conserved quantities which we could call Total Energy (and also momentum and angular momentum) was provided by Noether's theorem circa. 1918. This assumes that a "system" can be described by a Lagrangian and a least action principle governs the evolution of the system. I am not aware of any proof that all things we might want to consider as a physical system can be formulated in this way (with a Lagrangian). However, many of them can and it is now so widely accepted that this is the correct way to model physical systems that we will just go along with the idea that all physical systems can be described with a suitable Lagrangian for the remainder of this post.
Noether's theorem showed that symmetries were required in the physics of the system to produce these conserved quantities. There is a correspondence between the symmetries and the conserved quantities. Given a symmetry we can find a conserved quantity but also we can reverse this - given a conserved quantity we can find a symmetry that the Lagrangian must obey.
This is all inherently abstract mathematics. It describes energy as a mathematical expression involving several canonical variables that were sufficient to describe the Lagrangian of the system. These expressions for Energy don't always break apart into easily identified components. It is just a relationship between some canonical variables that are unique to the individual system.
In particular, Noether's theorem does nothing to identify fundamental forms of energy that may exist in the universe. In many systems there isn't a quantity that looks like or behaves like gravitational potential energy. Furthermore, Noether's theorem does not define energy as the "ability to do work", it is just an abstract relation between some variables.
4. Our best theory of gravity is still General Relativity (but I'm biased because I like it). In this, it is important to identify all sources of energy-momentum since this will be the source of gravitation. In particular, the energy density due to the content of the manifold must be specified.
(i) You are not required to include gravitational potential energy as a form of energy-momentum that contributes to the stress-energy tensor. (This was discussed in an earlier reply).
(ii) You are not able to include gravitational potential energy as a source of mass-energy even if you wanted to. Some of this was also discussed in earlier replies. There is a problem knowing where in space this energy would be located (for example, is it in the earth or in the apple, or spread out in the space between them). You don't actually have a working model of gravity as described by GR yet so any method of calculating gravitational potential energy would need to be based on some other theory available to you - such as Newtonian gravity. There is also a problem trying to determine how much gravitational potential energy there would be due to each particle in your system. For example, if you were using Newtonian gravity to calculate the potential energy then you have this formula to work with:
Gravitational potential energy = so that grav. p.e. → ∞ as r→0.
We can then place two particles of matter close together and provide arbitrarily large (large negative) energies. If we assume that this energy is located in the vicinity of those particles then strange things would happen. It would make two small particles of mass m that were close together the single most important source of gravitation in the manifold, far more important than say one particle located further away with a mass of 1000m. Now, it seems that this is not what happens in reality. Newtonian gravity describes the situation well enough for most purposes. If we put a test mass half-way between the 1000m mass particle and the pair of smaller particles of mass m that are close together, then the pair of close particles just behave like one particle of mass 2m. The test particle is not attracted to the close pair of small particles, instead it is pulled toward the more massive 1000m particle. Just in case you were concerned that the energy is a large but negative in the vicinity of the two close particles, the test mass isn't repelled by those two close particles either - it just sees them like one particle of mass 2m.
We have already discussed the idea that gravitational potential energy should NOT be included as a contribution to the stress-energy tensor but this makes gravitational p.e. stand out from all other forms of energy. In the theory of GR all forms of energy are sources of energy-momentum for the stress-energy tensor and we've got to ask why gravitational p.e. wouldn't be included - perhaps it is not a fundamental form of energy after all.
Another reasonable possibility is that if gravitational potential energy exists and should be included as a source of energy-momentum for the stress-energy tensor then it is not localised in or around two objects that are a certain distance apart. Instead that energy is spread uniformly throughtout the manifold. We already acknowledge that we cannot measure the absolute value of most forms of energy that contribute to the stress-energy tensor. The electric and magnetic fields are a common example since day-to-day physics only involves the differences between these potentials not the absolute value of those potentials. Anyway, if gravitational potential energy is a form of energy that is spread uniformly throughout the manifold then it would seem to be a component of the vaccum energy.
OK, I've over-run my time here again. Let's just say that I don't think gravitational potential energy is required, we already have vaccum energy in GR.
(iii) Finally, I should mention that, unlike the Newtonian theory of gravity, gravitational potential energy does not emerge as well defined quantity after GR. There are some situations where time-like Killing vectors can be identified and symmetry conditions like those required in Noether's theorem are met - but there are many situations where this can not be done.
In general relativity gravitational energy is extremely complex, and there is no single agreed upon definition of the concept. It is sometimes modelled via the Landau–Lifshitz pseudotensor.... [Taken from Wikipedia https://en.wikipedia.org/wiki/Gravitational_energy#General_relativity]
Summary
1. Energy is an extremely complicated thing.
2. Gravitational potential energy is more likely to be an emergent property. Some systems have an abstract relation between some variables that we can identify as and call gravitational potential energy but this is a property of that system. Gravitational potential energy is unlikely to be a fundamental form of energy that exists in the universe.
3. If we step back and re-examine something that @Halc mentioned in reply #5 we can start to make sense of it another way. If we consider a system that is the earth with all it's apples on the trees compared to a system which is earth with all it's apples on the ground, then we can attempt to consider how much force would need to be applied to these entire systems in order to start to move them. The system with the apples on the trees, has a slightly higher inertial mass then the system with all the apples on the ground. So the increased potential energy of the system seems to show up as a change in inertial mass of the system. This is one interpretation for the quantity Halc described as "system mass". It's perfectly sensible but only required because we were considering the earth-and-apples system as if they were one combined particle. If we take a suitable reductionist approach to analysing these systems then gravitational potential energy is not there to be found and indeed neither the earth nor the apples have changed their mass at all just because their positions have changed.
4. I only started the thread to discuss some basic science that we start teaching children at school and mainly just to pass a few evenings in discussion. If the thread has made anyone pause to reconsider what they thought they knew about Energy and how they should present to children then that'll be a bonus.
Thanks to everyone who has spent some time here. You are all free to continue commenting of course and indeed you can rip apart anything I've said apart if you want to.
Best Wishes.