Abstract
In this history I have examined the Fifth Force hypothesis from its origins, through its proposal and further investigation by other scientists, to its ultimate rejection by the physics community. These are what philosophers of science have called the contexts of discovery, of pursuit, and of justification. In previous work (Franklin 1990) I have argued that science follows an “evidence” model in which questions of theory choice, confirmation, and refutation are decided on the basis of valid experimental evidence. I have applied this model to various episodes in the history of science, including the discoveries of parity violation (the violation of left-right symmetry in nature) and of CP violation (combined parity and particle–antiparticle symmetry violation) and argued that this evidence model applies to the context of justification. I believe that this history has not only provided us with another illustration that the evidence model works in the context of justification, but it has also allowed us to examine the contexts of discovery and pursuit and to investigate the role that evidence may play in these contexts.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
I have also argued that there is an epistemology of experiment, a set of strategies for arguing for the validity of experimental results.
- 2.
It may not always be possible to clearly separate these contexts, but I believe that we should do so where it is possible. I believe it adds clarity to the discussion. For example, in the discussion of Fischbach’s work on the development of the Fifth Force hypothesis, I discussed it as the context of discovery because there was no hypothesis being investigated. It might also have been discussed as the pursuit of a solution to the problem of CP violation and its possible connection to gravity.
- 3.
I refer here to the original proposal of a force with a strength approximately 1 % that of gravity and a range of about 100 m.
- 4.
For other illustrations of this see the discussions of the interaction of experiment and theory in the case of weak interactions and atomic parity violation in Franklin (1990).
- 5.
This refinement of the theoretical model allowed more sensitive experimental tests of the Fifth Force to be both designed and implemented.
- 6.
Another method will be discussed later.
- 7.
“Different” experiments are classified by the theory of the apparatus. See Franklin and Howson (1984) for details.
- 8.
It could also be the case that both experimental results were correct, or they could both have been wrong.
- 9.
In the 1930s it was found that experiments on beta decay using thick and thin sources, respectively, gave consistently different results, and a systematic error was later found in the thick target experiments. (See Franklin (1990, Chap. 1) for details.) There is a similar contemporary problem that has not yet been resolved. Scientists using one type of detector have found suggestive evidence for a neutrino with a mass of 17 keV/c2. No such evidence appears when another type of detector is used. The question remains whether one of these results is an artifact of the detector used and the other a valid result, and if so, which one is valid.
- 10.
It is a fact of experimental life that experiments rarely work when they are initially turned on and that experimental results can be wrong, even if there is no apparent error. It is not necessary to know the exact source of an error in order to discount or to distrust a particular experimental result. It’s disagreement with numerous other results can, I believe, be sufficient.
- 11.
In the intermediate range (of the order of hundreds of meters) the local gravity measurements near Eckhardt’s North Carolina tower were taken at too high an average elevation because of the local swampy ground. Near Stacey’s Australian mine site, the local measurements in this range were too low because they thought that the mineshafts would be located in valleys. This also explained the difference in sign of the two observations.
- 12.
- 13.
There is an advantage for the historian or philosopher of science to be present while the science is being done and discussed. There is also an obvious danger that one will identify too closely with the participants. In this case I believe the danger has been avoided because I had no preference as to whether or not the Fifth Force actually existed. The history would be just as interesting and instructive in either case.
- 14.
- 15.
The constructivists depend here on philosophical arguments such as the underdetermination of theory by evidence, the incommensurability and the theory ladeness of observation, and the Duhem-Quine problem (the problem of assigning blame when a theory is apparently refuted). For a discussion of these issues see Nelson (1994) and Franklin (1990, Chap. 7).
- 16.
Ultimately experiment provided evidence of the superiority of the charm model.
- 17.
One might, of course, claim that no such evaluation took place. After all, there was very little published criticism, but this overlooks the large amount of unpublished criticism such as that which took place at the Moriond Workshops.
- 18.
Millikan’s value for e actually differs from the modern value. The difference can be explained by differences in the values used for the viscosity of air.
- 19.
The experimental refutation of such strongly held beliefs can, in fact, have beneficial career effects. Thus, Cronin and Fitch, the experimenters who found CP violation, won the Nobel Prize. Similarly, Lee and Yang, theorists who suggested parity violation and its experimental tests, also won the Nobel Prize.
- 20.
This hybrid model allowed both the neutrino results at high energies and the low energy atomic parity violation results to be correct.
- 21.
Because the Fifth Force involves the exchange of a different particle than that exchanged in existing gravitational theory one might very well consider it a new force.
- 22.
This is, in fact, the second point plotted in the later group of results. Although it was reported earlier at a conference it was the second published result.
- 23.
I am not denying that experimenters consider previous results when they are analyzing their own experiments, but rather that this consideration alone does not, in the long run, lead to convergence. See the histories of the measurements of other quantities given in Franklin (1986, Chap. 8).
- 24.
These probabilities are sometimes explicated by betting behavior, what the scientist thinks would be fair odds on the truth of the hypothesis. This should not be interpreted as real betting behavior because, as illustrated below, someone may not be willing to bet at what they actually consider to be fair odds. See Howson and Urbach (1989, Chaps. 1–2) for details.
- 25.
Such differing estimates of probability also lead to more hypotheses being investigated.
- 26.
Earman (1992) shows that for distinguishable hypotheses, those for which there are different empirical consequences, there are convergence of opinion results for Bayesianism.
- 27.
I am really assuming here that they were the only two hypotheses with any significant prior probability. This seems reasonable. By this time experiment had eliminated the Brans–Dicke theory as a serious competitor.
- 28.
For P(h 2) = 0. 00001 we find \(\mathrm{P}(h_{1}/e) = 0.9998\) and 0.996 for two and three standard deviation effects, respectively. P(h 2∕e) is reduced accordingly.
- 29.
Some critics might remark that I could always have chosen the prior probabilities to give the results I wanted. This is true, but I believe that the assignments I have made are reasonable given the fact that some scientists were indeed working on such theories when the hypothesis was offered, and the other evidence available at the time. I have, of course, already suggested that pursuit does not necessarily involve belief, but, even so, I think the calculations are reasonable.
- 30.
Scientists do make such judgments although they tend to be qualitative, i.e., large, small; or comparative, i.e., larger, smaller, etc.
- 31.
There were others who probably assigned the hypotheses a larger prior probability, such as Fischbach. I am not asserting that this analysis applies to any single member of the physics community, but only that it represents reasonable estimates of community beliefs.
- 32.
I am here associating plausibility with probability, although I do not believe that there is a uniform threshold of probability for pursuit, but rather that different scientists will make different judgements as to whether or not a hypothesis is plausible enough to be worth pursuing. Recall the earlier discussion on interest, plausibility, and ease of test in the context of pursuit.
- 33.
Ramsey never performed the experiment because his collaborator became involved in another experiment that seemed more important at the time.
- 34.
I note that the real bet differed from what they thought were the fair odds.
- 35.
For those who might believe that authority and power play a major role in such decisions, I note that Feynman, Ramsey, and Pauli all won the Nobel Prize in physics.
- 36.
At the moment, this is due to the difficulty of the theoretical calculations.
- 37.
- 38.
Both Fischbach and Newman have also raised this question with me. I suspect others have also asked it.
References
Ackermann, R.: Allan Franklin, right or wrong. In: Fine, A., Forbes, M., Wessels, L. (eds.), pp. 451–457 (1991)
Anderson, P.: The reverend Thomas Bayes, needles in haystacks, and the Fifth Force. Phys. Today 45, 9–11 (1992)
Bell, J.S., Perring, J.: 2π decay of the K2 0 meson. Phys. Rev. Lett. 13, 348–349 (1964)
Bernstein, J.: A Comprehensible World. Random House, New York (1967)
Earman, J.: Bayes or Bust. MIT, Cambridge (1992)
Fine A., Forbes, M., Wessels, L. (eds.): PSA 1990, vol. 2. Philosophy of Science Association, East Lansing, MI (1991)
Fischbach, E., et al.: Reanalysis of the Eötvös experiment. Phys. Rev. Lett. 56, 3–6 (1986)
Franklin, A.: The Neglect of Experiment. Cambridge University Press, Cambridge (1986)
Franklin, A.: Experiment, Right or Wrong. Cambridge University Press, Cambridge (1990)
Franklin, A., Howson, C.: Why do scientists prefer to vary their experiments? Stud. Hist. Philos. Sci. 15, 51–62 (1984)
Galison, P.: How Experiments End. University of Chicago, Chicago (1987)
Howson, C., Urbach, P.: Scientific Reasoning: The Bayesian Approach. Open Court, La Salle (1989)
Lynch, M.: Allan Franklin’s transcendental physics. In: Fine, A., Forbes, M., Wessels, L. (eds.), pp. 471–485 (1991)
Nelson, A.: How could scientific facts be socially constructed? Stud. Hist. Philos. Sci. 25, 535–547 (1994)
Pickering, A.: The hunting of the quark. Isis 72, 216–236 (1981a)
Pickering, A.: The role of interests in high-energy physics: the choice between charm and color. In: Knorr, K.D., et al. (eds.) The Social Process of Scientific Investigation. Sociology of the Sciences, vol. IV, 1980, pp. 107–138. Reidel, Dordrecht (1981b)
Pickering, A.: Constructing Quarks. University of Chicago, Chicago (1984a)
Pickering, A.: Against putting the phenomena first: the discovery of the weak neutral current. Stud. Hist. Philos. Sci. 15, 85–117 (1984b)
Pickering, A.: Reason enough? More on parity-violation experiments and electroweak gauge theory. In: Fine, A., Forbes, M., Wessels, L. (eds.), pp. 459–469 (1991)
Pinch, T.: Confronting Nature. Reidel, Dordrecht (1986)
Thieberger, P.: Thieberger replies. Phys. Rev. Lett. 62, 810 (1989)
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 2016 Springer International Publishing Switzerland
About this chapter
Cite this chapter
Franklin, A., Fischbach, E. (2016). Discussion. In: The Rise and Fall of the Fifth Force. Springer, Cham. https://doi.org/10.1007/978-3-319-28412-5_3
Download citation
DOI: https://doi.org/10.1007/978-3-319-28412-5_3
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-28411-8
Online ISBN: 978-3-319-28412-5
eBook Packages: Physics and AstronomyPhysics and Astronomy (R0)