When it comes to that I prefer the example of statistical mechanics, which is of course closely related.
Consider a set of molecules constituting a body of gas. Each of those molecules has a mass, a position vector, a velocity or momentum vector, an angular momentum vector, a chemical species etc. etc. None of them has a pressure, a volume, or a temperature. The molecule-by-molecule description of the gas does not contain values for the pressure, volume, or temperature of the gas. But it does imply them. You can in principle calculate those values as specialised statistics of the distribution of values of position, mass, and velocity.
In principle you could calculate all the physics of that gas from the dynamics of the individual molecules that make it up¹. The Newtonian dynamics of the molecules in the gas are fully sufficient to cause all its phenomena, such as exerting a force on the walls of its contain, its buoyancy in water and the buoyancy of a helium balloon within it, the lift and drag of an airfoil passing through it, the freezing of icecream or scorching of pizza placed within it, the sensation of a specific sound transmitted through it to a specific microphone. There is nothing to the physics of that body of gas that is not explicable in principle and caused in fact by the dynamics of its molecules. There are no supervening or wholistic causes that reach down though layers of abstraction to cause any of those molecules to do anything other than what individual dynamics cause them to do.
However, the body of gas does have properties that none of its molecules possess: pressure, volume, temperature, a speed of sound². These are what I call “emergent properties” in what I consider to be the straightforward meaning of the term, but I note that philosophers have defined such things to be “not truly emergent”, for clarity and convenience. We could calculate the pressure, volume, density, and temperature of a particular parcel of gas from information about its molecules, but we could never confirm by particular calculations, only repeatedly fail to falsify, that it had a speed of sound. What’s more, no detailed consideration of the molecular dynamics of a particular parcel of gas could derive universal generalisations about all gases.
But we can step away from the specific details of a specific parcel of gas. We can consider the Newtonian physics of molecules in general, and derive generalisations about pressure, volume, density, and temperature that apply to any parcel of gas. From general principles of Newtonian dynamics we can deduce the Universal Gas law, Bernoulli’s Equation, the speed of sound in a gas, the specific heat capacity of a gas. The emergent properties, or statistical properties if you prefer to say that they are Not Truly Emergent, have ironclad regularities to their physics that apply regardless of the particulars of the distribution of molecular properties, and that can be worked with confidently and precisely without having to know the position and velocity of even one molecule, without having to predict a single elastic collision. We can study the Carnot Cycle, design airfoils and rocket nozzles, turbines and piston engines, derive laws of acoustics and so forth without having to consider molecules any further. Once you’ve got the Universal Gas Law you can charge off into aerodynamics, thermodynamics, and acoustics by treating pressure, volume, and temperature as things subject to their own laws, and pay the Newtonian dynamics of molecules no further heed. What’s more, you can discover those laws empirically and independently of their derivation from dynamics: Boyle‘s Law, Charles‘ Law, and Gay-Lussac‘s Law were all discovered empirically before the molecular theory of gases was developed, and together they imply the Universal Gas Law.
Anyway: pressure and temperature are in no way independent of the positions and motions of molecules. If we already knew the positions and motions of all its molecules, measuring its pressure and temperature would not tell us anything that we could not have worked out from information that we already had. But when we step away from the particular to the general, forming the concept of pressure and temperature allows us to deduce or empirically investigate general propositions about all nearly-ideal gases that no amount of information about particular gases could have told us. Armed with the Universal Gas Law and Bernoulli’s Equation you can go on to do further science and engineering that are completely intractable if you do not use the abstractions “temperature” and “pressure”. Gas dynamics is completely consistent with and explained by Newtonian dynamics: it deals with pressure and temperature that Newtonian dynamics knows nothing of, not in supervention of Newtonian dynamics but because Newtonian dynamics implies that it must.
Is this what philosophers would mean if they said that pressure and temperature are real? Is it what they would mean if they said it wasn’t real? Is this what they mean when they say that the Universal Gas Law supervenes the dynamics of molecules or what they mean when they say it doesn’t. I find their jargon so confused that I can’t tell.
So anyway. In Flat Black There is no spooky stuff. No Cartesian dualism, no vital principle. Everything is physical³. Just as in the extended example above, “wholistic” principles of high levels of abstraction do not reach down into lower levels of abstraction to make particles do anything, particles just follow particle physics, and the higher-level physics, chemistry, biology, neuroscience, psychology, and economics is all summaries of what particles do for particle reasons. When a sophomore philosopher says “I think, therefore I am” that event is completely but not usefully explicable by fundamental-particle physics. A psychological explanation is more helpful than, completely consistent with, and just as correct as the particle physics.
Chemistry does what quantum mechanics implies that it must. Biochemistry does what chemistry implies that it must. Neurophysiology does what biochemistry implies that it must. Neuropsych does what neurophysiology implies it must. Psychology does what neuropsych implies it must. Sociology and economics do what psychology implies they must. It’s a single solid block of causal consistency, but each level is approachable on its own terms and can be applied without calculation from prior principles, as gas laws were discovered before molecular dynamics and can be applied directly.
¹ Including its deviations from the ideal gas laws.
² The distribution of chemical species, masses, positions, velocities, angular momentums, vibrational excitations etc. of its molecules also permits the calculation of other statistics. Some (such as for instance the proportion of molecules of each chemical species that are travelling faster than escape velocity) may be of interest in some situations but not others. Others (such as the proportion of molecules in the gas that as of chemical species whose formulas begin with “C”) are calculable but of no interest.
³ I share @whswhs’ objection to the usual formulation of materialism, which would require me to abjure belief in electrical fields, photons, and the curvature of space-time.