Abstract
The aim of this chapter is to explore and defend a route to realism about unobservable entities that does not use inference to the best explanation.
A version of this chapter has been given as a seminar in a number of places, including the University of Newcastle NSW, the University of Melbourne, Bristol University and the University of Athens. I am indebted to Joe Mintoff, Russell Blackford, Howard Sankey, Michel Ghins, James Ladyman and Stathis Psillos for helpful comments.
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.
See Arthur Eddington The Philosophy of Physical Science, Tarner Lectures, (Cambridge University Press, 1939) pp. 16–18.
- 2.
See Eddington, loc cit. Eddington writes “Anything uncatchable by my net is ipso facto outside the scope of ichthyological knowledge. In short, what my net can’t catch isn’t fish.”
It is perhaps worth stressing that Eddington himself did believe in things too small to see with the unaided senses. For example, he believed in the existence of protons and neutrons, and much of his cosmological work concerned how the numbers of these particles might determine the nature of the universe. His story of the fish net is not intended to cast in to doubt the existence of things not detectable by our senses. In his story, the fish net does not correspond to our eyes or ears, but to our language or concepts. Eddington is (perhaps) merely pointing out that any claim we make must always be made from our scheme of concepts. The position he is opposing is neither instrumentalism nor what we might nowadays call constructive empiricism. What he is opposing is possibly rather closer to what is sometimes called “Metaphysical Realism”, or the idea – rejected by Nelson Goodman, for example – of a real world underlying all our versions of it.
- 3.
What is widely regarded as the classic statement of this position is given in G. E. Moore’s papers “A Defence of Common-Sense” in Contemporary British Philosophy edited by J. Muirhead (1925) and “Proof of an External World” in Moore’s Philosophical Papers (Routledge, 1959) Chapter 7, pp. 126–148.
- 4.
The approach advocated here has some points of similarity, as well as some points of difference, to the approach sketched by M. Ghins towards the end of his “Putnam’s No-Miracle: A Critique” in Lyons and Clarke (2002), pp. 121–138. Ghins accepts that existence-statements such as “This table exists” are rational. But, for Ghins, it is a philosophical task to understand or explicate just what it is that makes such existence statements rational. He suggests that the way to establish scientific realism is to first develop an understanding of that in virtue of which statements like “Tables exist” is rational, and then proceed to argue that a justification of the same sort can be given for “Electrons exist”.
- 5.
In this book the position is adopted that induction and Eddington inferences are on firmer ground than inference to the best explanation. And the reason for this is because it was argued that both induction and Eddington inferences can be justified, whereas in Chap. 5 it was argued that we are not as yet in possession of a justification of IBE. However, in an influential paper, Gilbert Harman argued that enumerative induction ought to be seen as a special case of inference to the best explanation. (See G. Harman “The Inference to the Best Explanation” The Philosophical Review, vol 74, (1965), pp. 88–95.) On the face of it, if Harman’s claim is correct, it would appear to be potentially very damaging to the position adopted here. However, it seems to me that the thesis Harman is arguing for is perhaps consistent with, and not a threat to, the position advocated here. When Harman suggests that enumerative induction is best seen as a special case of IBE, what he is suggesting is that when a typical speaker (of English, say) offers what looks like an enumerative inductive argument, what the speaker is actually, perhaps implicitly, doing is using IBE. What Harman is doing, that is, is offering something like a thesis about what is implicitly “going on in the head” of the typical speaker when that speaker gives an argument. Harman is not, as far as I can see, offering any kind of thesis about rational justification. But in this book no claim is made about what is “going on in the head” of a typical speaker when they argue. Our concerns are with rational justification. So, as far as I can see, the claims made by Harman and the claims made in this book are distinct, and compatible.
- 6.
It might be suggested that the view of probability given here is inconsistent with the view given earlier. It has here been suggested that it is a priori unlikely that the size of the smallest fish in the sea should coincide with the holes in our net. From this it follows that it is a priori more likely that should be fish smaller than the holes than that there should be no such fish. But it might be felt this is inconsistent with the position adopted in Chap. 2, where it is stated that from an a priori point of view it is not more likely that our apparatus should display a particular numerical result (say, the needle pointing to “7”) than it is that the apparatus should turn in to a bowl of petunias. However, it will be argued that there is in fact no inconsistency here.
First let us look more closely at the claim it is a priori unlikely that the size of the smallest fish in the sea should coincide with the size of the holes on our net. Suppose the holes in our net are 2 inches across. Then, given that our net contains fish of many sizes greater than 2 inches, there are many lengths the smallest fish might have. One of these possibilities is:
-
(1)
The smallest fish are 2 inches
-
(2)
But the following are also possibilities:
-
(3)
The smallest fish are 1.9 inches
-
(4)
The smallest fish are 1.8 inches
-
(5)
The smallest fish are 1.7 inches
And so on.State of affairs (1) – that the smallest fish should be exactly 2 inches – is of course only one possibility amongst many. On these grounds it is therefore a priori less likely that this particular state of affairs given in (1) should obtain rather than that the disjunction “(2) or (3) or (4) and so on” should turn out to be true.
Let us now consider the a priori probability of the outcome of some experiment. Amongst the possible outcomes are:
-
(a)
The needle points to “7”.
-
(b)
The device turns in to a bowl of petunias.
Of course, given our (empirically obtained) knowledge of how the world works, (a) would surely be more likely than (b). But such an assessment of likelihood is, of course, a posteriori. From a purely a priori point of view, there would seem to be no evident reason to regard (a) as more likely than (b). Crucially, the comparison between (a) and (b) is a comparison between two specific outcomes or states of affairs. In this respect, it is unlike the other case just considered which involves comparison between a specific state of affairs and a disjunction of states. So, the reason for holding that it is a priori unlikely that the smallest fish in the sea should be exactly two inches does not hold in the “needle pointing to “7”” versus “bowl of petunias case”.
-
(1)
- 7.
Of course, there will no doubt be other considerations (to do with physiology etc.) that will tell us there cannot be fish below a certain size. But here we are only concerned with the relation that exists between one specific item of data (the fact that we have found fish of four inches in our trap) and hypotheses about the existence of smaller fish.
- 8.
See Nelson Goodman Fact, Fiction and Forecast (Harvard University Press, 1954) p. 74.
- 9.
There are some Eddington inferences using “grue” that produce paradoxical results. The following is an example:
-
All fish observed before D-day have been observed to be grue.
-
Therefore, there exist fish after D-day that are grue.
This inference in effect takes us from an observation that all fish observed before D-day have been green, to the conclusion that there exist fish after D-day that are blue, and so the premise here certainly does not seem to support the conclusion. As far as the present author can see, this paradoxical result does arise if we allow grue/bleen type predicates. But it is possible, within the general framework advocated here, to disallow such predicates. The issue is discussed in Explaining Science’s Success, pp.70–81.
-
- 10.
Paradoxical inductions of this sort are discussed in W. V. Quine and J. S. Ullian The Web of Belief (McGraw Hill, 1978).
- 11.
See Carl Hempel “Studies in the Logic of Confirmation” Mind, 54, (1945), pp. 1–26 and pp. 97–121.
- 12.
Although, as Musgrave has pointed out, this need not be the case if the only explanation we possess is very bad.
- 13.
Perhaps the first use of the agreement between different methods of determining the value of Avogadro’s in defence of scientific realism is Wesley Salmon Scientific Explanation and the Causal Structure of the World. (Princeton University Press, 1984).
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 2018 Springer Nature Switzerland AG
About this chapter
Cite this chapter
Wright, J. (2018). On the Inference to Unobservables. In: An Epistemic Foundation for Scientific Realism. Synthese Library, vol 402. Springer, Cham. https://doi.org/10.1007/978-3-030-02218-1_5
Download citation
DOI: https://doi.org/10.1007/978-3-030-02218-1_5
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)