AJP Volume 40 "A. Schild, Amer, Math. Monthly 68, 1 (1959); J. Bronowski, Sei, Amer. 208, 134 (1963) ; and papers given in Amer. J. Phys. 80, 462 (1062). For example, R. Resnick (Indroduetion to Special Relativity (Wiley, New York, 1968) ] suggests that the paradox in easily resolved since “Back twin ean earry an accelerometer.” Not quite universal, however. See M. Sachs, Phys, ‘Today 24, 23 (1971) 4K, Schwarzschild, Sitzber. Prous. Akad. Wiss, PhysikeMath, Ki, 189 (1910); R. Skinner, Relativity (Bisisdel, Waltham, Mass., 1968), p. 316, “HE Goldstein, Classical’ Mechanica (Addison-Wesley, Reading, Mass., 1950), p. 77 A, Kinstoin, Naturwiss, 6, 697 (1918); HL. Bergson, Duration and Simultancity (Bobbs-Merrill, New York, 1005), Appendices 1-3 The Reality of the Twin Paradox Effect DANIEL M. GRBENBERGER Department of Physics City Collegeof the City University of New York New York, New York 10081 (Reosived 6 Deoomber 1971) We discuss the assumption put forth recently by Sachs, that the proper time of a particle is singlesued func tion of the space-time coordinates. We demonstrate that swhile it eliminates the age difference between the twins in the twin paradaz, it also eliminates every other relativistic effect and implies Newtonian physics. We then discuss exactly how the nature of this assumption runs counter to the existence of « geometrical space-time manifold, and point out that the twin paradox is an integral part of relativistic physics, and that fo deny ite reality is to deny relativity. 760 | May 1978 Of all the consequences of special relativity, probably the most tnintuitive is the “twin para- dox’” effect, necording to which the passage of time for an observer is a path dependent phe- nomenon. ‘Thus two observers whose watehes agree when they are relatively at rest, may find at a later time when they are again relatively at rest, that their watches no longer agree. This result of relativity makes a violent assault upon “common sense” and has led to a rather large literature of dissent, much of it from people who claim that. they are willing to accept most of relativity, but not the twin paradox. The main point of this paper will be to show that this is not 2 consistent point: of view. If one rejects the twin paradox, then one must reject all of relativity. First we shall give a simple statement of the twin paradox as interpreted by special relativity. Consider two twins, A and B, at rest initially in an inertial system. Twin A remains at rest throughout the experiment, but twin B accelerates up to speed v in negligible time, and continues traveling. at constant velocity. After a while he quickly decelerates, taking negligible time, until his speed is v in the opposite direction, He then re- turns to A at constant velocity, at which point he rapidly decelerates to rest. According to special relativity, a clock at rest with respect to B and reading his proper time, will run slowly with respect to a clock at A, according to the formula Ar= A¢(1—0/ct)!®, and therefore when the twins again are at rest at the same point, B's clock will read less elapsed time than A’s.‘The moving twin will have aged more slowly than the one who remained behind, ‘The “paradoxical” nature of the result: oceurs when one analyzes the problem from the point of viow of B, who ean consider himself at rest. He then sees A as having taken the trip and therefore should find that A has aged more slowly. ‘The standard answer of special relativity is that, the descriptions offered by A and B aro not sym- metrical since B accelerates on turning around, and is therefore not always an inertial observer. One can assume that the relativistic interpreta- tion of the clock paradox is wrong and give a precise formulation to this point of view.! ‘Then fone would say that the clock on B will agree with one on A alter B retums. Furthermore, this should be true for an arbitrary path taken by B, not merely for motion at constant. velocity. ‘This hypothesis of equal elapsed time implies that the proper time r(z, t) is a unique analytic function of x and ¢. It follows that whenever an observer finds himself at the point x at time t, his proper time as read by a clock always at rest with respect to him, will read r(x, ?). Any other observer who also finds himself at z at time ¢ will read the same time r(x, t), independent of his previous history. This flatly contradicts the result of relativity, that the proper time 7 depends on the trajectory of the observer. We shall call the assumption that 7 is an analytic funetion of x and the Sachs assumption. He states it essentially without proof in his article but claims that it can be proven from his quarternion formalism? We shall assume the truth of the Sachs assumption and from it prove that it actually implies classical Newtonian physics and completely denies any relativistic effects whatsoever. Sachs ends his article with the statement of his assumption, as though it eliminated the twin paradox, without otherwise affecting relativity theory. But the statement that a particular differential is exact and leads to a state function is a very powerful and restrictive one (witness the existence of entropy) and in this ease not only seals the doom of the twin effect but of all rela istic effects. In fact, the assumption can be used as a criterion for classical Newtonian physics to hold For simplicity, we shall merely give three ex- Reality of the Twin Paradox Eifeet Fig. 1. Demonstration of the lack of time dilatation, Observers A and C are at rest. C emits radial light pulses which are picked up by A and B. B quickly accelerates, travels around the circle at constant speed v, and then quickly decelerates, According to the Sachs assumption, bis elapsed time is the same as 4’s, and therefore since he rooeives the same number of pulses, there is no transverse Doppler effect on frequencies, or time dilatation, amples, which will serve to” wipe out most of relativity. Howover, it is easy to construct many similar examples. The technique is to set up a situation in which two observers are initially at rest; in an inertial system and one takes off in a, rocket, follows some preseribed path, returns, and comes to rest. Then the two observers compare their clocks and find, according to the Sachs assumption, that they agree. EXAMPLE 1 Consider an observer A at rest in empty space. Another observer B is initially coincident. with observer A. B enters a rocket, takes off, and accelerates very rapidly to speed v, taking negligible time to do s0. B thon travels in a circle at constant speed v, as seen by any observer at, rest with respect to A. (See Fig. 1.) If the radius. of the circle is &, the trip takes time ¢=25R /v. When B again reaches A, he decelerates in negligible time, so that he and A are again relatively at rest. Then we must assume with Sachs that according to B the trip also takes time ¢ Now imagine that at the center of the circle another observer C is stationed, who is at rest AJP Volume $0 / 751Daniel M. Greenberger —— © o Fic. 2, Demonstration of the variability of the speed of light. (a) Tn the rest frame of observers A and A’, A sends 8 light signal to A’, who reflects it back. As A sends out tho signal, B quickly accelerates to speod », which he maintains until A’ receives the signal, at which point B quickly decelerates to speed ein the opposite direction. The reflected signal arvives at A coincidentally with B, who comes to rest. (b) ‘The same experiment as seen by B. Here A moves from As to A and back. Because the elapsed time is the same for both A and B, according to Sachs, but the distance traveled is different, they must see the light travel at different velocities. with respect to A and who is emitting light pulses at_a constant rate whieh propagate radially, with spherical symmetry, at velocity c. ‘Then, if pulses have been received during Bs circuit sinee ‘they will all be received by both A and B, and in ‘the same time ¢, it follows that both A and B observe the same pulse frequency Thus it follows that there is no transverse Doppler effeet. And sinee the transverse Doppler effect is merely the time dilatation effect in its purest form, it further follows that Sachs? assump- tion implies that there is no time dilatation. ‘There is a further conclusion to be drawn from this argument. if we accept the equivalence principle, which statos that an observer at rest in a gravitational field is locally equivalent to one at rest in an aceclerating frame, According to the equivalence principle, combined with special relativity, the observer B can consider himself to be at rest, and as such, instead of the inertial centrifugal force, he feels a gravitational accelera- tion g=w'r outward, due to a gravitational potential ¢= Sut, He then interprets the differ- ence in rates between his clock and the one at @ as due to the gravitational red shift. However, according to Sachs’ assumption, when A and B V2 { May 1978 coincide again, the clocks on A, B, and C will all be equal. So we may conclude from this assumption that there is no gravitational red shift. EXAMPLE I Tmagine two observers A and A’ who are at rest relative to each other in an inertial system, a distance Z apart. A emits a light signal to A’ who reflects it back to A. The total elapsed time is, .Lc. Now imagine another observer B, who is, initially coincident with A. At the moment the experiment is begun B accelerates up to speed 0 in negligible time, in a direction perpendieutar to the line AA’, (See Fig. 2.) B proceeds at constant speed v until the instant t=L/e, when A’ receives the signal from A. At that moment B reverses his direction in negligible time as seen by A and returns to A at speed v. Ashe again coincides with 4 he deeclerates to rest in negligible time. According to Sachs, his clock will also read time r=t=2L/c. Also, since the trips out and in look quite symmetrical to B (being eonneeted to each other by a reflection and time reversal), they will each take time 4=Z/e. But the total distance traveled by the light on one trip is, according to B, [17+ (oh) W#=d (as suming that distance is unaffected in the trans- verse direction). Thus the light signal will be traveling at a speed d/ly=(ct-+v4)'" with respect toB. ‘Therefore, the Sachs assumption contradiets the principle of the constancy of the velocity of light, and in fact implies the classical veetorial addition of velocities, and consequently the complete ‘Newtonian view of space and time as separate entities. EXAMPLE UT ‘We will now show that the assumption r=r(2, t) actually implies r=7(0), that is, ris a funetion of alone. And in fact this universal funetion is just, what plays the role of “” in classical physics. It is the absolute, universal time of Newton. Imagine two observers, initially at rest at the origin of an inertial frame. The first remains at rest at the origin, A, throughout the experiment. ‘The second leaves the origin at ¢=0, moves in an arbitrary trajectory, arriving at the point B, located at 21, at time t, at which moment he