Abstract
Several writers have attributed to Carnap an important attempt to solve the traditional sceptical problem of our knowledge of the external world.1 That Carnap’s oeuvre should contain a substantial response to traditional scepticism is on first sight a surprising claim. There is little in Carnap’s work, compared with Russell or Wittgenstein’s, say, by way of a direct and sustained explication of the concepts of knowledge and certainty or of other issues traditionally associated with scepticism. Moreover, there is a sense in which Carnap is a sceptic. Carnap, at least in his later writings, rejected “knowledge” as a vague term — like “bald” or “big” — and as of limited use in the sciences in favor of an account of rational belief in terms of subjective probability and rules of acceptance and rejection. If knowledge requires subjective certainty and infallibility, as the traditional account has it, then no one has knowledge, except where the sentence in question is analytic. Richard Jeffrey has placed Carnap’s position broadly within the tradition of the academic scepticism of Carneades of Cyrene.2
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
Preview
Unable to display preview. Download preview PDF.
Notes
Barry Stroud, The Significance of Philosophical Scepticism, Oxford: Oxford University Press 1987;
Christopher Hookway, Scepticism 1990;
Michael Williams, Unnatural Doubts. Princeton: Princeton University Press 1996.
See Ilkka Niiniluoto, “Scepticism, Fallibilism, and Verisimilitude,” Acta Philosophica Fennica 66, 2000, pp. 145–169. Niiniluoto has suggested in conversation that Carnap may have opted for suspension of judgment with respect to the realism-phenomenalism issue, or any of the conflicting existential statements that are the object of “Empiricism, Semantics, and Ontology.” Suspension of judgment, however, appears to presuppose that there is a meaningful question waiting for an answer, an alternative that has a well-defined truth-value, albeit an unknown one. Carnap’s critique is directed against this very assumption.
One may add to the two works mentioned Carnap’s “Überwindung der Metaphysik durch logische Analyse der Sprache” Erkenntnis II, 1932, and “On the Character of Philosophical Problems” Philosophy of Science I, 1934, pp. 5–19, translated and reprinted in: Richard Rorty (ed.), The Linguistic Turn, University of Chicago Press: Chicago, 1967, chapter 2.
Rudolf Carnap, Scheinprobleme in der Philosophie, (Günter Patzig, ed.), Suhrkamp: Frankfurt a.M., 1966 [1928]. Carnap’s retrospective statement seems to go against Bird’s claim, in part 2 of his essay in this volume, that Stroud attacks a much earlier and by 1950 obsolete and largely irrelevant verificationism Carnap’s. Below I argue that Stroud’s argument is invalid even if Carnap were the verificationist Stroud (wrongly) takes him to be.
Barry Stroud, The Significance of Philosophical Scepticism, Oxford University Press: Oxford, 1987, p. 174.
Ibid., p. 205.
Ibid., p. 170.
Ibid., pp. 205, 201, 202, 177–178.
Ibid., p. 201.
Rudolf Carnap, Scheinprobleme in der Philosophie, (Günter Patzig, ed.), Suhrkamp: Frankfurt a.M., 1966 [1928], p. 47.
Ibid., p. 50.
Ibid., p. 53. This is not much of an argument, but it is a significant step away from “scientism.” Meaning is separate and prior to the issue of certainty (p. 52). Cp. Moritz Schlick, “Meaning and Verification,” in: Richard Rorty (ed.), The Linguistic Turn, University of Chicago Press: Chicago, 1967 [1932], pp. 48–49.
Ludwig Wittgenstein, Vorlesungen 1930–35, Suhrkamp: Frankfurt a.M., 1989, p. 97.
Stroud seems to distinguish the two criteria: ibid., p. 200. However, in a footnote, p. 185, he identifies Carnap’s criterion as verificationist.
Carnap’s liberal criterion appears to be consonant with one advanced by Kant, whose “principle of significance” (Peter Strawson) for concepts is encapsulated in the following passage: “Possible experience is that which can alone give reality to our concepts; in its absence a concept is a mere idea, without truth, that is, without relation to any object.” (Immanuel Kant,Critique of Pure Reason, Norman Kemp Smith (transi.), MacMillan: London, 1787 [1990], A489/B517.)
I pretend for the sake of the argument that verificationism, as a linguistic account of meaning, faces no other problems. (There are many forms of verificationism and many difficulties with a precise statement of the doctrine, but the differences are irrelevant here and like Stroud (ibid., 204) I put them aside.) Stroud wants to create an internal difficulty for the anti-sceptic solution.
The given is an “Erlebnis.” “Erlebnisse” are analyzed into two kinds of elements: a “sensory core” and a “Nebenerlebnis.” The latter is characterized as a (logically, not psychologically) redundant element, open to the possibility of error and deception. The former, the “founding” core private to the subject is not open to error or deception, i.e. is indubitable or certain. (Rudolf Carnap, Scheinprobleme in der Philosophic (Günter Patzig, ed.). Suhrkamp: Frankfurt a.M., 1966 [1928], p. 26.)
Carnap offered this explanation of “associations that resemble the genuine understanding of a propositional statement” without elaboration again in: Paul A. Schilpp (ed.), The Philosophy of Rudolf Carnap, Open Court: La Salle, IL, p. 874.
Ibid., p. 205. “The conclusion of Descartes’ reflections as I presented them is said to be meaningless [...] but it certainly does not seem meaningless. [...] Of course that initial appearance of intelligibility might turn out on reflection to be illusory. But it also may turn out not to be illusory.”
Barry Stroud, The Significance of Philosophical Scepticism, Oxford University Press: Oxford, 1987, p. 202.
Rudolf Carnap, Scheinprobleme in der Philosophie, (Günter Patzig, ed.), Suhrkamp: Frankfurt a.M., 1966 [1928], p. 52. Compare Moritz Schlick, “Meaning and Verification,” in: Richard Rorty (ed.), The Linguistic Turn, University of Chicago Press: Chicago, 1967 [1932], p. 49.
The independence becomes evident when we take seriously the suggestion that meaning varies with the method of confirmation. Knowledge surely is not so variable.
Under the premises of phenomenalism it is plausible to take actual verification (of p by S) as a necessary and sufficient condition for knowledge (of p for S). For one, complete verification of sentences like “this square is red,” or “this square now appears red,” is possible in this framework. Utterances (to oneself) of this form are immediately, directly verified by the subject. They are taken to be self-justifying sentences (utterances), independent of other information. Carnap’s phenomenalism is a thoroughgoing foundationalist view of knowledge. However, the intrinsic difficulties of a phenomenological language are well known and need not be rehearsed here.
The example is Christopher Hookway’s (1990). Note, in the quote above Stroud slides from having the verificationist require confirmation of p to the requirement that she knows that p. Usually, for instance in the way Carnap used the notion of “confirmation” there is no implication that a state of knowledge is attained.
Knowledge is, as many examples have taught, more than true justified belief. An additional clause, for instance a counterfactual one, is problematic from a strict empiricist point of view. In fact, Carnap nowhere claimed that knowledge is a by-product of verification. Hempel wrote of setting standards of rational belief (Carl G. Hempel, “Empiricist Criteria of Cognitive Significance,” in: Aspects of Scientific Explanation, Free Press: New York, 1967, p. 8).
Incidentally, Carnap could have redefined (restricted) the concept of knowledge so that it would fit the relevant alternative model.
Rudolf Carnap, “Replies and Systematic Expositions,” in: Paul A. Schilpp (ed), The Philosophy of Rudolf Carnap. Open Court: La Salle, IL, p. 874.
Carnap’s formulation of the “Principle of Confirmability” is not better off in this respect: “If it is in principle impossible for any conceivable observational result [...] then expression A is devoid of cognitive meaning.” Ibid., p. 874.
Christopher Peacocke defends a solution, based on his “discrimination principle,” that, despite his criticism of verificationism, is similar to the criterion (1992, A Study of Concepts, MIT Press: Cambridge, Mass.). His argument is intricately interwoven with an intriguing, controversial theory of possession conditions for concepts, and cannot be examined here.
In Paul A. Schilpp (ed.), The Philosophy of Rudolf Carnap, Open Court: La Salle, IL, Carnap wrote (p. 877) that the arguments for the weak criterion can be found in Wittgenstein’s Tractatus (in 4.26, 6.5 and 6.51 perhaps) and in his Scheinprobleme in der Philosophic These arguments, however, address the strong conception of cognitive meaning, not the weak criterion or the Principle of Confirmability.
Descartes argued for the likelihood of his brand of corpuscle theory in the Principles by noting that it is highly improbable that the mechanistic theory he had developed would fit all known independent phenomena if it weren’t true and would not describe the true causes. By way of analogy Descartes asked: what is the probability of discovering a complete and meaningful decipherment of a coded Latin text that is wrong, i.e. does not agree with the original, uncoded text? It can be shown that if the coded text is sufficiently long and varied, and if the set of letters and vocabulary is sufficiently large, the probability is very low.
I can be brief in my exposition of “Empiricism, Semantics, and Ontology” since the same ground is covered in Bird’s essay in this volume.
The inclusion of inductive rules into the framework creates a difficulty: it renders trivial the claim that all justification is framework “internal” and that the choice of a linguistic framework is not justifiable.
See part 1 of the essay by Bird.
Rudolf Carnap, “Empiricism, Semantics, and Ontology,” in: Meaning and Necessity, University of Chicago Press: Chicago, 1956 [1950], p. 74.
Many have noticed affinities here. For instance: Michael Williams, Unnatural Doubts, Princeton University Press: Princeton, 1996, p. 30, 160.
Here I follow Mary McGinn, Sense and Certainty: A Dissolution of Scepticism. Blackwell: New York, 1989, p. 138.
Ludwig Wittgenstein, On Certainty, (G.E.M. Anscombe, G.H. von Wright, eds.). Harper & Row: New York, 1969 [1949–1951], § 153.
Ibid., §370.
Ibid., §655.
Ibid., pp. 81–82.
But does the invariance of factual truths in different frameworks not accord those facts a special metaphysical status, in need of explanation? I return to this question below.
Michael Williams, Unnatural Doubts, Princeton University Press: Princeton, 1996, pp. 20, 150–151. The Achilles’ heel of Carnap’s approach, it is said, is the assumed separability of the “factual” from the “formal” in any empirical area of discourse. Carnap distinguished between sentences framework-internal and sentences that are framework-external in terms of analytic versus synthetic truth and thereby opened up a familiar assortment of problems. In his paper (this volume) G.H. Bird argues that the argument of “Empiricism, Semantics, and Ontology” depends only on the weaker assumption that we are able to identify languages and separate languages from theories proper. Since this matter is only tangentially related to the discussion of Stroud’s objections I will not consider this issue further.
I like to raise here concerns about the applicability of the ideas of “Empiricism, Semantics, and Ontology” to a staple of contemporary scepticism: the “brain-in-the-vat” scenario. The thought-experiment has scientists stimulate a human brain, kept separated from its body in a “tank,” in a way so that it perceives a wholly fictional “external” world, in which it has all sense organs, a body, a fake history and no knowledge of its predicament. A widely discussed solution is based on what is known as semantic externalism. Can one show on Carnap’s terms that we are not brains-in-a-vat? Part of the beauty of the thought-experiment is that it requires nothing but a physicalist language. The sceptical alternative is not obscure or ill-defined. It seems difficult to maintain that all that is involved here is a wistful alternative framework decision.
Barry Stroud, The Significance of Philosophical Scepticism, Oxford University Press: Oxford, 1987, p. 192.
Ibid., p. 193.
Ibid., p. 194.
Some “facts” will not be recovered. Take, for instance, the number of objects in a given spatial region. Various ways of counting are possible (mereological sums, etc.) giving rise to different counts. Does this contradict the claim of invariance? Methods of counting can differ. What matters are observation statements, and insofar as they are about the objects in that location, they must be invariant under a change of counting rules as much as distances are invariant when replacing yards by a metric equivalent.
Ibid., p. 195.
Ibid., p. 195.
Compare Morman’s essay in this volume.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2003 Springer Science+Business Media Dordrecht
About this chapter
Cite this chapter
Bonk, T. (2003). Scepticism Under New Colors? Stroud’s Criticism of Carnap. In: Bonk, T. (eds) Language, Truth and Knowledge. Vienna Circle Institute Library [2003], vol 2. Springer, Dordrecht. https://doi.org/10.1007/978-94-017-0151-8_8
Download citation
DOI: https://doi.org/10.1007/978-94-017-0151-8_8
Publisher Name: Springer, Dordrecht
Print ISBN: 978-90-481-6258-1
Online ISBN: 978-94-017-0151-8
eBook Packages: Springer Book Archive