THREE SPECULATIONS FROM PHYSICS

Let me begin with three examples from the history of physics, together with claims of their detractors who reject or at least criticize them not because they are false or refuted but be­cause they are speculations.

a. Thomas Young’s Wave Theory of Light

In 1802, Thomas Young published “On the Theory of Light and Colours,”¹ in which he resuscitated the wave theory of light by introducing four basic assumptions: first, that a rare and highly elastic luminiferous ether pervades the universe; second, that a luminous body excites undulations in this ether; third, that the different colors depend on the frequency of the vibrations; fourth, that bodies attract this medium so that the medium accumulates within them and around them for short distances. With these and other assumptions, Young shows how to explain various observed properties of light.

In 1803, Henry Brougham, a defender of the particle theory of light, wrote a scathing review of Young's paper, in which he says:

As this paper contains nothing which deserves the names ei­ther of experiment or discovery,... it is in fact destitute of every species of merit A discovery in mathematics, or a

successful induction of facts, when once completed, cannot be too soon given to the world. But... an hypothesis is a work of fancy, useless in science, and fit only for the amusement of a vacant hour.^{[4] [5]}

Brougham defends the Newtonian particle theory of light on the grounds that it is inductively supported by experiments, and he rejects Young's wave theory on the grounds that it is mere speculation.

b. William Thomson (Lord Kelvin): Baltimore Lectures on Molecules and the Wave Theory of Light

In 1884, Sir William Thomson delivered a series of lectures at Johns Hopkins University on “molecular dynamics and the wave theory of light.” His aim was to provide a molecular interpretation for the luminiferous ether postulated by the wave theory.

It seems probable that the molecular theory of matter may be so far advanced sometime or other that we can understand an excessively fine-grained structure and understand the lumi­niferous ether as differing from glass and water and metals in being very much more finely grained in its structure.^[6]

He proceeds by offering various mechanical models of the ether to explain known optical phenomena, including recti­linear propagation, reflection, refraction, and dispersion.

In his 1906 classic The Aim and Structure of Physical Theory,^[7] Pierre Duhem excoriates Thomson for presenting a disorderly series of contradictory models (as, he claims British minds, incapable of continental (meaning French) orderliness, are wont to do), for invoking occult causes, and for not producing a “system of principles, which aim to represent as simply, as completely, and as exactly as possible a set of experimental laws” (p. 19). Duhem writes:

The multiplicity and variety of the models proposed by Thomson to represent the constitution of matter does not astonish the French reader very long, for he very quickly recognizes that the great physicist has not claimed to be fur­nishing an explanation acceptable to reason, and that he has only wished to produce a work of imagination.^[8]

Again, the complaint is that we have a theory, or set of them, that are pure speculations, and ones of the worst kind, since they lack order and simplicity.

c. String Theory

Characterized by some of its proponents as a “Theory of Everything,” it attempts to unify general relativity and quantum mechanics into a single framework by postulating that all the particles and forces of nature arise from strings that vibrate in 10-dimensional spacetime (according to one prominent version) and are subject to a set of simple laws specified in the theory. The strings, which can be open with endpoints or closed loops, vibrate in different patterns giving rise to particles such as electrons and quarks.

The major problem, or at least one of them, is that there are no experiments that show that strings and 10- dimensional spacetime exist. The theory is generally regarded, especially by its critics, as being entirely speculative. Steven Weinberg, once an enthusiastic supporter of string theory (as the “final theory”) in 2015 writes:

String theory is... very beautiful. It appears to be just barely consistent mathematically, so that its structure is not arbitrary, but largely fixed by the requirement of mathematical consist­ency. Thus it has the beauty of a rigid art form—a sonnet or a sonata. Unfortunately, string theory has not yet led to any predictions that can be tested experimentally, and as a result theorists (at least most of us) are keeping an open mind as to whether the theory actually applies to the real world. It is this insistence on verification that we mostly miss in all the poetic students of nature, from Thales to Plato.^[9]

As these examples illustrate, speculations are assumptions normally introduced in the course of activities such as explaining, unifying, predicting, or calculating. Young sought to explain, or at least to see whether it is possible to explain, known phenomena of light by a theory other than the particle theory. Kelvin was attempting to provide a molecular account of the ether, and in terms of this, to explain the known op­tical phenomena. String theorists want to explain and unify the four known fundamental forces and calculate the funda­mental constants of nature. In the course of doing so, they introduce speculative assumptions.

There are two sorts of speculations I want to distinguish. The first, and most common, are made by speculators who, without knowing that there is evidence (if there is), intro­duce assumptions under these conditions: (a) They believe that the assumptions are either true, or close to the truth, or possible candidates for truth that are worth considering. (b) They introduce such assumptions when explaining, predicting, unifying, calculating, and the like, even if the assumptions in question turn out to be incorrect.^[10] I will call (a) and (b) “theorizing” conditions.

In other cases, assumptions, without evidence, are introduced in the course of explaining, predicting, unifying, etc., but their introducers do not believe that they are true, or close to the truth, or even possible candidates for truth. Indeed, it is often believed that they are false and cannot be true. A good example, which I will discuss in section 12, is Maxwell's imaginary fluid hypothesis introduced in his 1855 paper “On Faraday's Lines of Force.” Here, to represent the electromagnetic field, Maxwell describes an incompressible fluid flowing through tubes of varying section. The fluid is not being proposed as something that exists or might exist. It is, as Maxwell says, purely imaginary. Its purpose is to pro­vide a fluid analogue of the electromagnetic field that will help others to understand known electrical and magnetic laws by employing an analogy between these laws and ones governing an imaginary fluid.

Another prominent example of this second type of speculation is atomic theory as viewed by some nineteenth­century positivists. They employed the assumptions of atomic theory not as ones they believed to be true, or close to it, or as possible candidates for truth, but as fictions useful for explaining, predicting, and unifying certain observable phenomena.⁸ For them, as for Maxwell in the imaginary fluid case, no evidence is given for the truth of the assumptions introduced. Indeed, evidence is irrelevant, since truth is. I will call these truth-irrelevant speculations.

Both truth-relevant and truth-irrelevant speculations contain assumptions about objects and their behavior for which there is no known evidence. Both are introduced for purposes of explaining, predicting, and organizing phe­nomena.⁹ That is why I call them both speculations.

8. As with Maxwell's incompressible fluid, the explanations are not meant to be causal. In Maxwell's case, we explain, not what causes the phe­nomena but what they are, as well as unify them, by invoking an analogy between these phenomena and others, real or imagined (see section 12, this chapter). In the atomic case, according to some positivists, we ex­plain not what causes—e.g., Brownian motion—but how the observed Brownian particles are moving. We explain that they are moving as if they are being randomly bombarded by molecules (without committing our­selves to the claim that they are being so bombarded). For my own ac­count of explanation in general, and non-causal explanation in particular, see Achinstein, Nature of Explanation. For a much more recent account of non-causal explanations, see Marc Lange, Because without Cause: Non- Causal Explanations in Science and Mathematics (New York: Oxford University Press, 2016).

9. There are non-speculative cases when assumptions are introduced without knowing that there is evidence for them—e.g., introducing an as­sumption known to be false in the course of giving a reductio argument, or in the course of giving an historical account of a discarded theory, or just to see whether it is consistent with what we know. But here the as­sumption does not satisfy “theorizing” condition (b), required for both truth-relevant and truth-irrelevant speculations. It is not introduced with the purpose of explaining, predicting, etc., but in the first case, just with the purpose of showing that it is false; in the second, just doing some his­tory of science; and in the third just determining consistency. Nor, in such cases, does the assumption satisfy the “theorizing” condition (a) required difference between them stems from “theorizing” condition (a). Truth-relevant speculations satisfy it, truth-irrelevant speculations do not. However, both kinds are anathema to writers such as Brougham, who demands inductive proof based on experiment, and Duhem, who rejected atomic theory construed either realistically or as a useful fiction.

My discussion of truth-irrelevant speculations and how they are to be evaluated will appear in section 12. The main focus of this chapter will be on truth-relevant speculations. Until section 12, when I speak of speculations I will mean just these. (Some readers, indeed, might prefer to restrict the term “speculation” to these, using a different term—e.g., “im­aginary construction”—for truth-irrelevant ones. Because of similarities just noted, I will continue to classify them both as speculations, while recognizing an important difference be­tween them.) Truth-relevant speculations have sparked the for truth-relevant speculations. It is not introduced with the idea that it is true, or close to the truth, or a candidate for truth worth considering. To be sure, in such cases the epistemic situation of the introducer may change, and the assumption may come to be treated by the introducer in a way sat­isfying (a) and/or (b). But that is a different situation.

most controversy among scientists and philosophers. I turn to three contrasting views about them next.

3.