November 24, 192 1] 



NATURE 



3<7^ 



Letters to the Editor. 



[T/je Editor does not hold himself responsible for opinions 

 expressed by his correspondents. Neither can he undertake 

 to return, or to correspond with the writers of. rejected 

 manuscripts intended for this or any other part of Nature. 

 No notice is taken of anonymous communications.^ 



Metaphysics and Materialism. 



Ix his article in Nature of October 20 Prof. Carr 

 maintains that the principle of relativity has ■"re- 

 formed " the concept of physical reality, and has 

 finally made untenable the doctrine that matter is 

 real. I dissent most strongly from his conclusion. 

 If "real " is used in the only sense in which a belief 

 in the reality of matter ever has been part of physics, 

 that belief is entirely unaffected by the principle of 

 relativity, which involves the belief as much as any 

 other physical prof>osition. There may be senses of 

 the word "real" in which the doctrine that matter 

 is real is affected by the acceptance of the principle 

 of relativity, but they are either repugnant to physics 

 or irrelevant to it. I have explained my reasons for 

 this opinion at length elsewhere, but at such length 

 that nobody seems inclined to read them. May I be 

 allowed to set them out rather more concisely? 



First, then, Prof. Carr will admit that '"real" is 

 used in at least two senses. He presumably smiled 

 when first he heard the old storA* of the mongoose 

 which was as unreal as the snakes it was kept to kill. 

 If he could understand "real" only in the meta- 

 physician's sense, he would not, being an idealist, see 

 any point in the story, for he would assert that every 

 mongoose and every snake was unreal. The reality 

 which the traveller denied to his mongoose was the 

 reality of common sense, not the reality of meta- 

 physics. The criterion of this common-sense reality 

 is universality of experience. The snakes were not 

 real because the experience of them was limited to 

 the intemjjerate friend ; other people could not experi- 

 ence them. The mongoose was unreal because if the 

 box were opened nobody would experience it. If Prof. 

 Carr were to say that he does not know what I mean 

 by '"universality of experience " in this connection, I 

 fear I should simply disbelieve him ; for the whole 

 of his normal life shows that he recognises and 

 applies the criterion like anybody else, materialist, 

 idealist, or physicist. 



The "real " of common sense is the fundamental 

 real " of science. A belief that there are real things 

 in this sense (which might be, but is not usually, 

 expressed by the assertion that matter is real) is 

 necessary to science. It is necessary not so much in 

 order that science should be true as in order that there 

 should be any science. For real things in this sense 

 are the subject-matter of science. Science selects the 

 material it will study by applying the criterion of 

 universal experience. It rejects from its subject- 

 matter the snakes which only the traveller's friend 

 could see ; it accepts only such snakes as everA-body 

 can see. It is as absurd to ask whether students of 

 science are right to employ this criterion as to ask 

 whether Greek scholars are right to accept the uni- 

 versal practice of Greek writers ; if we applied anv 

 other criterion our study would not be experimental 

 or natural science. 



In so far as the principle of relativity leads to anv 

 conclusions relevant to experimental science, it depends 

 on this criterion as much as other scientific theorv ; 

 the only observations with which its conclusions must 

 be in accordance, if the principle is to be accepted as 

 true, are obser\'ations concerning which evervone 

 agrees. Sir Oliver Lodge does not differ from Prof. 



NO. 2717, VOL. 108] 



Eddington concerning the distribution of the spots on 

 the Principe plates, and both of them would agree 

 that, if they did differ as to that distribution (say, 

 because the spots were very diffuse), the plates would 

 not afford any evidence on which to base any scientific 

 conclusion whatever. They agree that the plates must 

 satisfy the criterion of scientific reality before they 

 can proceed to differ in their interpretation of the 

 plates. Nor would any difference between them in 

 the matter of scientific reality arise even if the hypo- 

 thetical experiments of which Prof. Eddington is so 

 fond could actually be tried (I hold that it is most 

 imix)rtant to insist that they cannot be tried), and if 

 it were found that two observers travelling with great 

 relative velocity differed concerning the simultaneity- 

 of events. They would both agree that the observers 

 did differ ; and they would agree that, because of this 

 difference, the judgment of neither could be accepted 

 as a valid basis for scientific argument unless one or 

 both could be "corrected " in some manner so as to 

 remove the discrepancy. The position which would 

 arise would be precisely similar to that which arose 

 when first it was noticed that observers at different 

 distances from a gun differed concerning the simul- 

 taneity of flash and report. Some observations which 

 had been expected to be "real" would have turned 

 out to be "unreal," but enough real observations, 

 concerning which agreement could be obtained, would 

 remain (if anything can be predicted about an event 

 which cannot occur) to make science possible, even 

 as applied to the discrepant observations (cf. Phys. 

 Zeit., vol. 13, p. 126, 1912). 



So much for "real" in its fundamental sense. 

 But physicists also use the word in another sense 

 when they speak of electrons or atoms or the aether 

 being real. This sense is not the fundamental sense, 

 because nobody has perceived or could perceive an 

 electron, and because people do actually differ about 

 the reality of some of these things (e.g. the aether). 

 Fortunately, there is no need to discuss in detail here 

 the significance of this kind of "reality," for no 

 phvsicist would assert in this sense that "matter is 

 real." If matter is something different from, but 

 common to, electrons, atoms, and aether, then science 

 does not assert, but most strenuously denies, that such 

 matter has any "reality." And that conclusion, again, 

 is perfectly independent of the principle of relativity'. 



On the other hand, the principle of relativity might 

 have a bearing upon reality in this sense, though not 

 on the reality of matter. For to assert that some- 

 thing is real in this sense is to say that some theon*- 

 is believed to be true ; if we believe the theory* of 

 relativity to be true, we may assert that some things 

 are real which we shall assert are not real if we do 

 not believe the theory. But it is not certain that we 

 shall do so, for not all ideas of true theories are 

 asserted to be real. For example, I do not think 

 that anyone, speaking carefully, would assert that the 

 frequency of light is real ; he would not say it was 

 unreal ; he would say that the conception of reality 

 was inapplicable to a frequency, just as is the con- 

 ception of triangularitA'. Personally, I should adopt 

 that attitude towards the ideas of the theory- of rela- 

 tivity ; others might possibly differ. 



But it must be noted that anv influence the theorv* 

 of relativity mav have on our views of realit}' in this 

 sense is shared by other theories. Prof. Carr is quite 

 right when he suggests that the theorv establishes 

 some rather special relation between mathematics and 

 phvsics ; it gives realit}-, in the second sense, to ideas 

 which derive their attractiveness, not from analogv 

 with material laws, but from connection with mathe- 

 matical form. But so does the Bohr-Sommerfeld 



