1 Einstein, the reality of space and the actionâreaction principle
Harvey R. Brown and Dennis Lehmkuhl
For me it is an absurdity to ascribe physical properties to âspaceâ.
Albert Einstein to Ernst Mach1
That a real thing has to be presupposed as the cause for the preference of inertial systems over non-inertial systems is a fact that physicists have only come to understand in recent years.
Einstein2
Introduction
It was striking in the 2012 seminar âThe Nature of Realityâ at the Indian Institute of Advanced Study, Shimla, that so many speakers independently referred to the quip by Albert Einstein in his 1930 dialogue with Rabindranath Tagore that he, Einstein, was more religious than the great Bengali polymath. Of course, what Einstein was getting at is the fact that at the heart of the physical sciences, at least as he saw it, is a commitment to the reality of an external world whose goings-on are governed by laws that contain no fundamental reference to conscious agents, and in particular to human observers. For Tagore, all truth is human truth, if one is to take his claims literally.3 For Einstein, the ultimate truth about the physical world transcends the human realm. Einstein was a ârealistâ, but his realism was of a modest, or non-metaphysical, kind. It is the job of physicists, he argued, to come up with models of a mind-independent reality that are explanatory of our phenomenological experiences of natural regularities, within the laboratory and without. However, whether such models, when judged successful, correspond to the way the world âreallyâ is, is a question Einstein thought best to leave aside. There are some philosophical questions for which Einstein thought the best response is a smile, and this was one of them. But Einstein stressed that even this weak, pragmatic take on truth involved a leap of faith. He made it clear, particularly in his 1949 Autobiographical Notes, that Natureâs connivance in allowing for the success in the scientific venture as he conceived it could not be a foregone conclusion.4 It was conceivable for Einstein that the mind, say, could be the bedrock of reality, but he felt that there was no good reason to start with that premise, and good reasons not to. Realism for Einstein was more a programme than a doctrine; it was a dogma about which he was careful not to be dogmatic.
Several years after the development of his 1915 general theory of relativity (GR), Einstein began to stress that physical space, or rather, the metric field, not only constitutes a fundamental, autonomous element of objective reality, it plays a causal role in accounting for the inertial motion of bodies.5
He compared this with the active role of space in the cases of Newtonian mechanics (NM) and special relativity (SR). In these cases, such putative action is clearly not reciprocated by bodies or fields: they do not act back on space-time structure, so the so-called actionâreaction principle is violated. In contrast, in his relativistic theory of gravity (GR), Einstein was to see the vindication of the principle. The metric can have a dynamical life of its own in the absence of matter fields (though, as we shall see, this goes against Einsteinâs original expectations), but, more to the point, when the latter exist, the metric affects and is affected by them.6
Whether Einsteinâs reasoning is correct is open to doubt. What is debatable is not the claim that GR is consistent with the actionâreaction principle; it is the claim that the older theories involving absolute space-time structures are not. More specifically, it is the assertion that such structures act in the relevant sense. Clearly, if they do not (as Newton himself argued regarding space; see below), then the fact that material bodies do not act back on them constitutes no violation of the actionâreaction principle. The case for the view that absolute space-time structures of the kind that appear in NM and SR are not fundamental causal entities in their own right, but rather codifications of certain bare facts concerning the movement of bodies or behaviour of fields has been made a number of times, in different ways.7 One curious instance is arguably in a letter Einstein himself wrote to Ernst Mach in late 1913;8 another is due to Moritz Schlick in correspondence with Einstein in 1920.9 We shall return to the second case below; it turns out to be directly related to the main concern of this chapter, which is, why would Einstein start systematically emphasising the actionâreaction principle in his defence of GR only in the 1920s?
The actionâreaction principle
It is a venerable tradition in natural philosophy to assert that a substance is the seat of actions on other substances, and in turn subject to the actions of these other substances: the actionâreaction principle (AR). In his pre-Principia manuscript De Gravitatione et Aequipondio Fluidorum10 (or De Grav for short), Newton insisted that natural philosophers tacitly understand substance as an entity that acts on things, even if they donât state this explicitly. He distinguished two kinds of action, one that stimulates the perceptions of thinking beings, and one between material bodies, as in collisions. (Later, he would extend this second kind to action at a distance.)11 This distinction today seems of little import, and the fundamental premise, that it is interactions between systems that count in physics and not their intrinsic properties, is close to the heart of the âstructural realistâ position that has been much discussed in recent years in the philosophy of science. Be that as it may, Newton is clearly appealing to a principle in the De Grav that is more fundamental and general than what he would later designate as his third law of motion in the Principiaâthough the latter is often referred to as the law of actionâreaction. (We shall see shortly how space, for Newton, is a kind of exception to this fundamental principle.)
Leibniz, whose views on the nature of space and time were so different from Newtonâs, nonetheless shared the same intuition. In fact, when defining substance as that which acts and can be acted upon, he understood that he was adopting the view of the scholastics.12 Caution must be exercised, however, in attributing AR, as it is standardly understood today, to Leibniz. In the light of his doctrine of pre-established harmony, the meaning of causation, or rather interaction, within his deep metaphysics is almost certainly at odds with those views adopted by the majority of current metaphysicians,13 not to mention physicistsâand its scope is still controversial.14
If there is a questionable aspect of AR, it is less the claim that substances act (how otherwise could their existence be known to us?) than the notion that they are necessarily acted back upon, that action must be reciprocal. If all substances act, they do so in relation to other substances; these other substances therefore cannot be immune from external influences. Now it might seem arbitrary on a priori grounds to imagine that the âsensitivityâ of such substances is not universal. That is to say, it might seem arbitrary to suppose that not all substances react to others. But no such abstract qualms can be entirely compelling; Nature must have the last say.
Nowadays, it is well known in the foundations of physics that the de Broglie-Bohm (âpilot waveâ) interpretation of quantum mechanics, in its standard form, involves a dynamical agent (the wave function) that acts on corpuscles (point particles) but is not acted back upon, at least by the corpuscles. It is noteworthy that David Bohm himself clearly found such violation of AR uncomfortable in his (first) 1952 formulation of the theory and attempted to remedy it, as have others after him.15 While defenders of the original pilot-wave theory can legitimately argue that AR is not a logical necessity, others see its violation as a blemish in the theory.16 It seems fair to say that currently within the physics community, there is no consensus supporting the failure of AR in quantum theory.
For his part, Einstein himself had already stated in 1922 that it is âcontrary to the mode of scientific thinking to conceive of a thing ⊠which acts itself, but which cannot be acted uponâ.17 The object of Einsteinâs ire in 1922 was NM and his own creation, SR. Yet there is no hint in his writings around the time of the development of SR in 1905 that Einstein considered either of these theories to incorporate a violation of the actionâreaction principle; at any rate, the explicit condemnation came later. Why? In all probability, because it was part of an honest sales pitch for GR, his greatest and most radical contribution to science, after Einstein was reluctantly forced to concede, because of results by de Sitter, that the theory as a whole was not consistent with âMachâs Principleâ, even though special solutions are. It seems that this change of tack on Einsteinâs part was consolidated in the mentioned 1920 correspondence with the physicist-philosopher Moritz Schlick.
But before examining the evolution of Einsteinâs views on the causal role of space-time, it is worth briefly visiting Newtonâs own views on the nature of absolute space, especially as expounded in the De Grav. This manuscript contains a hard-hitting critique of Descartesâs relational theory of motion and the positive reasons why Newton felt compelled to posit the existence of absolute space; these important lines of reasoning have been discussed extensively in the literature.18 What concerns us is Newtonâs insistence that space, despite its reality, does not act on either our sense organs (which is patent) or on our bodies. If it is a substance, it is by Newtonâs own reasoning sui generis in its causal inefficacy, and ultimately, Newton had no interest in classifying it as either substance or accident.19
When, in the context of discussions relevant to this chapter, Newtonian space is assigned a causal role, it is usually to account for inertia, i.e., the privileged existence of inertial frames, or equivalently, the special motions of force-free bodies. In the De Grav, Newton explicitly renounced such a role. He stated that the reason why projectiles that are not being acted upon by other bodies moving in straight lines and at uniform speed is precisely that space has no ability to help or hinder any change in the motion of bodies!20
Einstein on absolute space
Einstein and Mach
Einsteinâs tortuous road to the 1915 field equations of his unhappily named general theory of relativity followed a number of fundamental, partially overlapping, conceptual signposts, apart from th...