Whewell, Newton, Kant, and Hershel's Double Stars

We may observe, however, that the law of gravitation according to the inverse square of the distance, which thus regulates the motions of the solar system, is not confined to that province of the universe, as has been shown by recent researches. It appears by the observations and calculations of Sir John Herschel, that several of the stars, called double stars, consist of a pair of luminous bodies which revolve about each other in ellipses, in such a manner as to show that the force, by which they are attracted to each other, varies according to the law of the inverse square. We thus learn a remarkable fact concerning bodies which seemed so far removed from us that no effort of our science could reach them; and we find that the same law of mutual attraction which we have before traced to the farthest bounds of the solar system, prevails also in spaces at a distance compared with which the orbit of Saturn shrinks into a point. The establishment of such a truth certainly suggests, as highly probably, the prevalence of this law among all the bodies of the universe. And we may therefore suppose, that the same ordinance which gave to the parts of our system that rule by which they fulfil the purposes of their creation, impressed the same rule on the other portions of matter which are scattered in the most remote parts of the universe; and thus gave to their movements the same grounds of simplicity and harmony which we find reason to admire, as far as we can acquire any knowledge of our own more immediate neighbourhood.--William Whewell Astronomy and General Physics, Considered with Reference to Natural Theology, 1864, (Bridgewater Treatise, "New Edition, with New Preface"),pp 229-230 [HT David Haig]

I quote from the edition of Whewell's Bridgewater Treatise in which he adds a new preface in order to respond to Darwin's Origin (which he quotes, but doesn't name). In the preface, Whewell makes a point of associating Darwin with the system of Epicurean chance, which at the cosmogenic level is to be associated with the nebular hypothesis as developed by Kant and Laplace. In the Bridgewater Treatise Kant is not mentioned only Laplace. (I suspect that Whewell, who was one of the greatest historians of science of his age, was not unfamiliar with the Kantian provenance because he alludes to it in his treatment of copernican turn of the critical philosophy elsewhere.*) So, arguments against one (the nebular hypothesis) reinforce the arguments against the other (Darwinism), and vice versa. 

One may wonder why, after after the empirical success of the Principia, and its Rule 3, anybody would have doubted the universality of the law of gravity. But, as a matter of fact, Newton himself had opened to the door to this possibility, in the 1706 Latin edition of the Opticks which ultimately ended up in Query 31 of the later English edition.+ It reads:

Since Space is Divisible in infinitum, and Matter is not necessarily in all places, it may also be allowed that God is able to create Particles of Matter of several Sizes and Figures, and in several Proportions to Space, and perhaps of different Densities and Forces, and thereby to vary the Laws of Nature, and make Worlds of several sorts in several Parts of the Universe. 

Newton sometimes uses ‘world’ in a metaphysically neutral sense, to mean a solar system or, more generally, a closed (or virtually closed) system of interacting celestial bodies. In the above passage, however, ‘world’ is a metaphysically richer notion. A world in this sense is constituted by the kinds of particles and forces it contains, particles and forces that are ontologically prior to the world they constitute. Moreover, the above passage suggests that within ‘worlds’ of this kind particles and forces are prior, in some sense, to the laws of nature.

For Newton laws are clearly contingent. They are not grounded in God’s immutability (as Descartes’s were), but depend on God’s will. Their contingency stands in stark contrast to the necessity of Newtonian space and time. For Newton, space and time could not have been otherwise and are a consequence of God’s necessary existence. As he writes in the General Scholium to the Principia, “by existing always and every where, [God] constitutes Duration and Space . . . ’Tis allowed by all that the supreme God exists necessarily; and by the same necessity he exists always and every where.” Physical laws do not possess such necessity. (Although Newton’s laws are contingent, they are necessary in relation to God’s will. Given God’s edicts, they could not be otherwise. ) As Roger Cotes wrote in the editor’s preface to the second edition: “From this source [God], then, have all the laws that are called laws of nature come, in which many traces of the highest wisdom and counsel certainly appear, but no traces of necessity” (Newton 1999: 397; the denial of necessity distances Newton from the charge of Spinozism and its attack on design arguments.)

In the above passage, the contingency of Newton’s laws is also coupled to the claim that they may not be universal. They may hold of some worlds (in the metaphysical sense), but not all. Their variability has at least two consequences. First, a universe with metaphysically distinct worlds entails causal disconnectedness among (some of) its parts. This disconnectedness appealed to Newton, because it could prevent different solar systems from collapsing in on each other. It thus supports the argument from design of the General Scholium. 

So, while in the eighteenth century the Huygensian claim that the law of gravity only held for celestial bodies in our solar system was defeated by the stronger claim that it held throughout the solar system, including terrestrially, it was not obvious it would hold in other solar systems throughout the galaxy. Huygens (recall here; here) and Newton seemed to posit a universe with tenuously connected solar systems (light could pass among them), even though Newton encouraged us to take or hold true the laws universally in context of inquiry. 

In 1802, Herschel wrote (in the Transaction of the Royal Society):

If, on the contrary, two stars should really be situated very near each other, and at the same time so far insulated as not to be materially affected by the attractions of neighbouring stars, they will then compose a separate system, and remain united by the bond of their own mutual gravitation towards each other. This should be called a real double star; and any two stars that are thus mutually connected, form the binary sidereal system which we are now to consider. (Vol. 92, 481)

Herschel's language here is conditional. But on p. 486 he goes on to claim there is overwhelming empirical evidence for the reality of binary systems. Interestingly, on. p. 485 Herschel claims that Binary systems are themselves causes by law of gravity because their production and the phenomena they give rise to cannot be the effect of "casual situations" that is, chance.

So, from the point of view of physical astronomy, Herschel's discovery of binary systems was analogous to the seventeenth century discovery that the satellites of Jupiter and Saturn obeyed Keplerian motion. But from a cosmological perspective, Herschell provided empirical evidence for the hitherto unavailable "simplicity and harmony;" whatever else might be discovered, Hershel showed "a remarkable fact" that the motions of bodies obeyed at least one law or set of laws. This claim which was taken for granted by Descartes on metaphysical grounds, and left open empirically by Huygens and Newton, could now be taken as established, or at least "as highly probable." 

Let me close with three observations. First, Whewell clearly thinks that (strong evidence for) the actual university of the law of gravity points to a single, unified cause of the universe. In itself that does not settle the debate with the systems of necessity and chance, in favor of theism. But it does burden shift in one non-trivial way against the system of chance (but not the system of necessity). 

For, Whewell thinks, not entirely unfairly, that the system of chance must assume that the present laws of nature are themselves the effect of historical evolution. That's the "supposition that the universe has gradually approximated to that state of harmony among the operations of its different parts," (p. 30).** For, the system of chance has, if it accepts the principle of sufficient reason (which it may not, of course) no right to positing, without further explanation, the existence of natural law that governs motions of bodies.++ (The system of necessity gets this posit for free.) 

Second, and this is the burden shifting against the necessitarian, if one accepts the principle of sufficient reason, the now universal law of gravity appears without foundation at all: "no [sufficient] reason, at all satisfactory, can be given why such a law must, of necessity, be what it is." (216) In particular, while gravity is a universal quality of matter there is no reason to think (here Whewell explicitly disagrees with Newton's editor, Cotes) it is a necessary quality (p. 222).

Third, while Whewell's language is still voluntarist ("ordinance"), Whewell changes the character of God's voluntarism. For Newton, theological voluntarism is compatible with God choosing different laws of motion for different parts of the universe. Newton allows the genuine possible that God settled for a dappled universe. Whewell's nature, by contrast, obeys a single ordinance/decree by God, who could choose differently in different parts of the universe, but did not.

 

 

 

 

*Of course, Kant himself rejects the association with the Epicurean system of chance, although he recognizes that some readers might be tempted to attribute it to him. My own view is that Kant embraces the (Spinozist) system of necessity in his early cosmogony. 

+What follows draws on joint work with Zvi Biener.

**In context Whewell is discussing plant adaptation. But if we follow the guidance of new preface we should be willing to recycle arguments from cosmological to terrestrial contexts and vice versa.

++If I recall correctly, in Darwin's Dangerous Idea, Dennett grants that if we apply the universal acid to cosmology, we might be tempted toward a view of eternal creation and destruction of universes selecting toward universes that appear more fine tuned, etc.

Comments

[Marius Stan]

Thanks, Eric. Very interesting, as always. I have a few questions.

I don't understand the idea that the System of Chance (SoC) doesn't get to posit a law for bodily motions. Without laws of motion, you can't infer to anything past the "chance" state (=distribution of masses and velocities) you start with. Laws are rules of inference, or else the chance state entails nothing but analytic truths. So, if SoC is expected to explain the *current* world-state from some "chance" *past* state, surely it must be allowed rules of inference. Namely, laws. Kant knew that. I don't see why you have to explain the laws themselves. Their "explanation" is that they suffice -- together with the chance state -- to account for the present state.

Second, it seems the dialectic is between *three* sides, not two: the Chancer, the Dappler, and the Uniformer. If God could be a dappling creator, the anti-chancers run into some problems. 1) to sort out whether the Dappler or the Chancer is right, you have to do a complete induction, with evidence from all over the universe, not just finite regions. That exceeds whatever evidence-gathering technologies anyone could hope to have in the 19th c. 2) short of an induction by complete enumeration, the debate between Dappler and Uniformer would have to get settled partly from non-empirical evidence. Before the end of inquiry, the empirical facts are compatibel with either side above. 3) more generally, because these positions are explanatory projects, I have a hunch they expected to carry the day partly on extra-empirical grounds. Namely, not from astronomical evidence, but somehow from reasons falling out of the "best"-part in 'inference to the best explanation.'

What say you? :-)

[Alan Nelson]

An interesting development of this comes from the distinction between 1) different laws being legislated in different places for one matter, and 2) different laws applying because different kinds of matter are created in different places.
In the indented quote above from Newton, "...different densities and forces" is ambiguous. I might mean that the quantities of force are different because the densities of composite bodies are different, or it could mean that the forces are different because the nature/essence of the matter is different.

[Eric Schliesser]

Ah this is interesting, Alan.
In the first edition of the Principia, Newton is clearly committed to a homogeneous account of matter compatible with your 1). But he drops that hypothesis in later editions, and I think he shifts to something like 2) in the queries of the optics or, at least, that's how I interpret the passage from the queries.
I don't think Newton thinks the essence of matter is different in different worlds (because he seems to think the essence of matter is a conceptual truth); although he might think that other qualities are.....

[Eric Schliesser]

First a clarification (in light of your second comment, Marius), prior to Whewell, the debate is primarily among the system of chance, the system of necessity, and the system of mind/orderer. [I think mature Kant and Hume complicate the story but let's leave that aside.]
Second, of these three only the system of necessity requires uniformity. But both the system of chance and mind are compatible with dappled world or uniformity.
Third, indeed Whewell thinks that such debates cannot be settled merely by empirical evidence. (He clearly and explicitly appeals to underdetermination --it's funny how Duhem/Quine are given credit--and extra-empirical/scientific principles.)
On your first point. I have to think about it. But I think one can have more primitive inference rules that one does not identify with the laws of nature. Now, (a) that does leave us with an interesting question if the laws of nature are different in kind or merely degree from the more primitive inference rules one uses in reasoning; (b) it does leave an interesting question whether any of the three approaches can explain that difference between primitive inference rules and more substantive laws of nature; (b) Whewell thinks God is a good explanation for the character of laws of nature (this style of argument is also in Clarke, who does allow that one can ask a further question about explaining God). Descartes and Spinoza also offer, quite different, accounts of the way laws are ground in God. So, I think the explanatory demand can be recognized.