Abstract
In this chapter we continue the task of defending Scientific Realism against threats and challenges. In particular, in this chapter we consider:
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.
The theory of phlogiston is given as an example creating a difficulty for realism by Larry Laudan in his “A Confutation of Convergent Realism” in Jarrett Leplin Scientific Realism (University of California Press, 1984), pp. 218–249, especially p. 231.
- 2.
See Alan Musgrave “Why did oxygen supplant phlogiston?” in Method and Appraisal in the Physical Sciences, edited by Colin Howson, Cambridge University Press, 1976, pp.181–210.
- 3.
See Musgrave, op cit.
- 4.
For the advocates of phlogiston theory the combustion of some substance C was not oxygen combining with C, but phlogiston being given off by C. This way of viewing things led Joseph Priestley to offer his own interpretation of what happened when mercury oxide (what he called “the precipitate per se) was heated to produce pure mercury. Since mercury oxide was, in his view, dephlogisticated mercury, the process of heating mercury oxide to produce pure mercury must have been the process of dephlogisticated mercury re-absorbing phlogiston. And so the resultant “air” produced by this process, he reasoned, must be air from which the phlogiston had been removed. He referred to this as dephlogisticated air. So, for Priestley, what we now call “oxygen” was “dephlogisticated air”. But now, for Priestley, in combustion phlogiston is given off by burning substances: it stops once the air has become saturated with phlogiston and cannot absorb any more. On Priestley’s view, therefore, “dephlogisticated air” (oxygen) ought to have a capacity for supporting combustion greater than that of ordinary air: since it is dephlogisticated air it ought to have more “room” to absorb phlogiston than ordinary air. And this was observed to be the case – things were found to burn more easily and energetically in dephlogisticated air. Thus phlogiston theory led to a novel prediction that was subsequently found to be correct.
- 5.
One way in which phlogiston is wrong is that it misrepresents what is going on when something burns. We know now that when something burns it combines with oxygen: burning is a form of (chemical) combination. But phlogiston theory says that in combustion something (phlogiston) is given off from the burning substance. For phlogiston theory, burning is not combination but expulsion or separation.
There is a sense in which a substance that burns does lose at least something in combustion: the electrons in its outer shell. But I take it that no one would seriously suggest this vindicates phlogiston theory. Presumably advocates of phlogiston theory conceived of phlogiston as a kind of substance, on a par with substances such air or water or carbon. A quantity of electrons is presumably too dissimilar from what advocates of phlogiston theory had in mind to count as a candidate for what they were getting at.
- 6.
This is discussed in James Ladyman “Structural Realism versus Standard Scientific Realism: The Case of Phlogiston and Dephlogisticated Air”, Synthese 2011, volume 180, pp.87–101. The idea that phlogiston could be viewed as “anti-oxygen” is also raised in Mikhail Volkenstein Entropy and Information, Birkhauser Physics (2009), p.6.
- 7.
See Ladyman, op cit.
- 8.
This is explored in Ladyman, op cit.
- 9.
“Reflections on the motive power of fire” by S. Carnot (Paris, Bachelier, 1824).
- 10.
See S. Psillos “A philosophical Study of the transition from the caloric theory of heat to thermodynamics: Resisting the pessimistic meta-induction, Studies in the History and Philosophy of Science, 25 (1994), pp.159–190. For an opposing point of view, see Hasok Chang.
- 11.
See Hasok Chang “Preservative Realism and Its Discontents: Revisiting Caloric” in Philosophy of Science vol. 70 (2003), pp.902–912.
- 12.
Perhaps it is more correct to say that “particle of caloric” is to be replaced by “quantity of mass possessing kinetic energy equivalent to work capable of being done by a particle of caloric”. But as the latter expression is somewhat cumbersome, I will simply speak of “unit of kinetic energy”.
- 13.
A summary can be found in Eugene Hecht, Optics (Pearson Education Limited, 2014), pp.496–497.
- 14.
Laudan op cit, p.225.
- 15.
More specifically, Fresnel’s theory is based on the idea – originally due to Christiaan Huygens – that each point in a wave of light acts as the source or origin of another, spherical wave of light emanating out from that point. This is known as “Huygens’ Principle”. The propagation of a wave of light is the result of the “adding together” of all these spherical waves. (More precisely, it is the adding together of these waves in the direction in which the beam of light is travelling.) Huygen’s Principle is accepted in modern physics. See, for example, M. Born and E. Wolf, Principles of Optics: electromagnetic theory of propagation, interference and diffraction of light (Cambridge University Press, 1999), p.986.
- 16.
Born and Wolf, loc cit.
- 17.
Rankine, W. J. M. “On the mechanical action of heat, especially in gases and vapours” in Transactions of the Royal Society of Edinburgh, 20, pp.147–190. (1853).
- 18.
See Keith Hutchison “Miracle or Mystery?: Hypotheses and Predictions in Rankine’s Thermodynamics” in S. Clarke and T. Lyons Recent Themes in the Philosophy of Science (Kluwer Academic Publishers, 2002), pp.91–120.
- 19.
The account of Rankine’s theory given here is derived from Hutchison, op cit.
- 20.
This is a part of Newton’s First “Rule of Reasoning in Natural Philosophy”, from his Mathematical Principles of Natural Philosophy.
- 21.
One student of mine remarked that Rankine’s mechanism had a “steampunk” quality about it. “Steampunk” is a style in fashion and art that imagines what our modern world might be like if we had continued to rely on the same general types of mechanical device as those prevalent in the ninetheenth century, such as the steam engine. So, an imagined “steampunk” world might have flying machines, computers, communications systems and so on powered by steam engines, clockwork etc. rather than jet-turbines and electronics. The mechanical devices imagined (or constructed) in “steampunk” can be of fantastical and grotesque complexity.
- 22.
Quine argues for this thesis in a number of places, including “On the reasons for the indeterminacy of translation” Journal of Philosophy, 67 (1970), pp.178–183, Word and Object, Cambridge, Mass., MIT press, 1960.
- 23.
W. V. Quine “On Empirically Equivalent Systems of the World” in Erkenntnis 9, (1975) pp.313–328.
- 24.
One author who has consistently defended the thesis that Quine’s thesis that theory is underdetermined by all possible data is Lars Bergstrom. See, for example, his “Underdetermination and Realism” Erkenntnis, 21, pp.349–365, 1984, and “Quine, underdetermination and skepticism” Journal of Philosophy, 90, pp.331–358, 1993.
- 25.
It is worth noting that in what was perhaps his fullest discussion of the issue, Quine refrained from actually asserting the thesis the there will be more than one theory capable of explaining all possible data. See Quine op cit. Quine argues that for the undedetermination thesis to be true, there must be two or more genuinely different theories that can explain all the data. But this, of course, raises the question: “Under what conditions are two apparently different theories genuinely different rather than merely “notational variants” of each other?” Quine develops an account of the conditions under which two theories are to be regarded as “genuinely different”. He then professes agnosticism on the question of whether or not there would be two or more such genuinely different theories capable of explaining all possible observations.
- 26.
See L. Laudan and J. Leplin “Empirical Equivalence and Underdetermination” in Journal of Philosophy, 88, (1991), pp.449–472.
- 27.
See Jarret Leplin A Novel Argument For Scientific Realism (Oxford University Press, 1997), esp. 153–157.
- 28.
See Andrei Kukla “Empirical Equivalence and Undertermination”, Analysis, 53, (1993), pp.1–7.
- 29.
See P. Kyle Stanford Exceeding Our Grasp: Science, History and the Problem of Unconceived Alternatives (Oxford University Press, 2006), esp. pp. 9–17.
- 30.
G. E. Moore “Proof of an External World”, in Proceedings of the British Academy, 25, (1939).
- 31.
See Roger Jones “Realism About What?” Philosophy of Science, 58, (2), pp.185–202 (1991).
- 32.
A number of authors have recognised that, in one way or another, the existence of equivalent descriptions constitute a challenge to realism. The idea is a theme in much of the work of Hilary Putnam in the 1970s and 1980s. See, for example, Reason, Truth and History (Cambridge: Cambridge University Press, 1981) and “Realism and Equivalence” in Philosophical Papers vol 3 Realism and Reason. It should be noted, however, that the object of Putnam’s attack is perhaps “metaphysical realism” rather than “scientific realism”. In his What is this thing called science?” Alan Chalmers argues that the existence of equivalent descriptions constitutes a problem for scientific realism.
- 33.
The locus classicus of Structural Realism is John Worrall “Structural Realism: the best of both worlds” in Dialectica, 43, pp.99–124 (1989). An application of Structural Realism to the specific case of different equivalent formulations of Newtonian mechanics can be found in John Wright “Realism and Equivalence” in Erkenntnis, July 1989, vol 31, pp.109–128.
- 34.
This option is defended in Alan Musgrave “Realism About What?” in Philosophy of Science, 59, (4), pp.691–697. (1992).
- 35.
One formulation of this argument is given in Karl Popper The Logic of Scientific Discovery, p.364.
- 36.
It should be noted that there are other ways of replying to Popper’s argument. One influential approach to the problem was developed by Colin Howson. See Howson’s “Must the logical probability of a theory be zero?” in British Journal for the Philosophy of Science 24, 2, (1973), pp.153–163. However, the position adopted here is perhaps slightly stronger than that defended by Howson. Howson argued it is possible to assign non-zero probabilities to universal generalisations. On the view defended here, any assignment of probabilities to universal generalisations via Bayes’ Theorem must be indeterminate. This leaves it open, however, that we might be justified in making assertions about the probability by means of some other route.
- 37.
It is worth here reminding ourselves of the distinction between a priori probability and “prior probability”. The a priori probability of E is the probability we are entitled to ascribe to E on the basis of purely a priori considerations. The “prior probability” of E, on the other hand, is the probability a speaker might assign to E on the basis of general background beliefs concerning the probability of E, rather than more specific evidence bearing directly on the probability of E.
- 38.
We can perhaps get a clearer intuitive grasp of the situation by considering some “disembodied spirit” considering epistemic probabilities from behind a Rawlsian “veil of ignorance”. The disembodied spirit is told that shortly it will experience something, it is not told what it might be. Amongst the possibilities are: the needle on a meter pointing to “7”, a tropical typhoon, a cow being milked, Tony Abbott delivering a speech in Swahili and so on. There would appear to be indefinitely many possible experiences, and the disembodied spirit has, a priori, no reason to regard any one of them as any more likely than any other.
It might possibly be argued that there could only be finitely many possible experiences the spirit might have. An analogy might be given with a computer screen. Suppose a computer screen has N pixels and each pixel can exhibit M colours, where both N and M are finite. Then the maximum number of possible images will be NM. This may be a very large number, but necessarily it will be finite. In a similar way, it may be argued, the spirit could only have finitely many experiences. However, it is plain that this argument is based on assumptions about the possible nature of the experiences the spirit will have. It is based on the assumption that the experience will be produced by some analogue of a finite number of pixels each of which can have only finitely many states. But what entitles us to make such an assumption a priori? From a strictly a priori point of view, we are not entitled to make any such assumption.
- 39.
There is perhaps a more basic reason why it might be feared that in maintaining that the probability of all observations is zero we have “thrown out the baby with the bath water.” If the probability of all observations is zero, it might be thought, it must surely follow that the probability of all theories is zero too, since surely no theory can be more likely than any observation. However, this objection is not correct. On the view advocated here, the a priori probability of all observations is zero, and the a priori probability of all theories is also zero. But the a posteriori probability of a theory in the light of evidence need not be zero.
We can bring this out with a simple example. Suppose that a coin is tossed ten times. What is the a priori probability that ten heads come up? It is tempting to think that the a priori probability of this will be 1/210. But, on the view advocated here, this is not so. It is, perhaps, the epistemic probability of ten heads coming up if we know in advance that the only two possible outcomes are a “heads” coming up or a “tails” coming up, and we have no reason to believe one outcome is more is more likely than another. But we do not know a priori that these are the only two possible outcomes. Perhaps the coin might turn in to a bowl of petunias: we do not know a priori that this will not happen. The a priori probability of a heads coming up on just one toss is, on the view advocated here, zero. So, the a priori probability of ten heads coming up in a row is also zero.
But now, let us assume that the coin has been tossed ten times and ten heads have come up. We now surely have good, if defeasible, reason to believe:
-
“The physical probability or propensity of a heads coming up on a single toss is greater than ½.”___(1)
We would have more reason to believe (1) than to believe:
-
“The physical probability or propensity of a heads coming up on a single toss is less than ½.”____(2)
But now, if we have more reason to believe (1) than (2), it surely follows that the epistemic probability of (1) must be greater than zero. Of course, (1) is not an example of what we ordinarily think of as a “theory”. But it is a non-observational statement that (together perhaps with the law of large numbers) would seem to give us probabilistic reason to believe indefinitely many other statements. The point is that, even if the a priori probability of all observation statements is zero, the a posteriori probability of a non-observational statement can be greater than zero.
Still, it may be felt, there is something “fishy” going on. On the view advocated here, the a priori of:
-
“The coin was tossed ten times and came up heads every time”_____(3)
is zero. But the epistemic probability of (1) in the light of (3), is greater than zero. But, it may be protested, this surely means that the a priori probability of the conditional statement:
-
“If the coin was tossed ten times and came up heads every time the propensity for heads to come up is greater than ½.”___________________________________________(4)
must be greater than zero. But, it might be thought, there is something implausible or ad hoc about claiming that the a priori probability of (3) is zero while that of (4) is greater than zero. Both (3) and (4) are synthetic, contingent statements. Why should one of them have an a priori probability of zero while the other does not? However, the solution to this is given by Bayes’ Theorem. On the view advocated here, both (3) and “The coin has a propensity greater than ½ of coming up heads” have an a priori probability of zero. Hence, by Bayes’ Theorem, the probability of (4) will be the indeterminate 0/0. There is, therefore, no inconsistency in maintaining that the value of (4) may be greater than zero.
-
- 40.
The “Experimentalist’s Regress” and its significance for science, is discussed in Harry Collins Changing Order: Replication and Induction in Scientific Practice, Chicago, IL: University of Chicago Press, 1985.
- 41.
Perhaps one of the most famous alleged examples of a scientist using only the data that seemed to favour their preferred theory was Eddington’s observations in 1919 of an eclipse as observed from the island of Principe. However, recent investigations have found that Eddington did not selectively use data. See Daniel Kennefick “Not Only Because of Theory: Dyson, Eddington and the Competing Myths of the 1919 Eclipse Expedition” Cornell University Library (2007), http://arxiv.org/abs/0709.0685
- 42.
See Gerald Doppelt “Empirical Success or Explanatory Success: What Does Current Scientific Realism Need to Explain?” Philosophy of Science, v.72, (2005), pp.1076–1087.
- 43.
A similar point is made by Alan Musgrave is response to Arthur Fine’s defence of the “natural ontological attitude”. See Musgrave “The Ultimate Argument for Scientific Realism” in Robert Nola (ed), Relativism and Realism in Science (Kluwer Academic Publishers, 1988), pp.229–252.
- 44.
The notion of the theory laden-ness of observation was perhaps raised in prominence as a topic of discussion in the philosophy of science by Thomas Kuhn’s The Structure of Scientific Revolutions, esp. p.111, pp.113–114, pp.115–121. It is not claimed here that the considerations given by Kuhn fail to show that observation statements do have a risky and theory-laden character. Rather, the claim is that even though they do have that character, Moorean arguments show they are firmer than theoretical claims, and that under ordinary conditions of utterance they are things we can be said to know.
- 45.
See P. Kyle Stanford Exceeding Our Grasp: Science, History and the Problem of Unconceived Alternatives (Oxford University Press, 2006)
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 2018 Springer Nature Switzerland AG
About this chapter
Cite this chapter
Wright, J. (2018). The Skeptical Arguments – 2. In: An Epistemic Foundation for Scientific Realism. Synthese Library, vol 402. Springer, Cham. https://doi.org/10.1007/978-3-030-02218-1_3
Download citation
DOI: https://doi.org/10.1007/978-3-030-02218-1_3
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-02217-4
Online ISBN: 978-3-030-02218-1
eBook Packages: Religion and PhilosophyPhilosophy and Religion (R0)