Abstract
Logical consequence is typically construed as a metalinguistic relation between (sets of) sentences. Deflationism is an account of logic that challenges this orthodoxy. In Williamson’s recent presentation of deflationism, logic’s primary concern is with universal generalizations over absolutely everything. As well as an interesting account of logic in its own right, deflationism has also been recruited to decide between competing logics in resolving semantic paradoxes. This paper defends deflationism from its most important challenge to date, due to Ole Hjortland. It then presents two new problems for the view. Hjortland’s objection is that deflationism cannot discriminate between distinct logics. I show that his example of classical logic and supervaluationism depends on equivocating about whether the language includes a “definitely” operator. Moreover, I prove a result that blocks this line of objection no matter the choice of logics. I end by criticizing deflationism on two fronts. First, it cannot do the work it has been recruited to perform. That is, it cannot help adjudicate between competing logics. This is because a theory of logic cannot be as easily separated from a theory of truth as its proponents claim. Second, deflationism currently has no adequate answer to the following challenge: what does a sentence’s universal generalization have to do with its logical truth? I argue that the most promising, stipulative response on behalf of the deflationist amounts to an unwarranted change of subject.
Article PDF
Similar content being viewed by others
Avoid common mistakes on your manuscript.
Change history
22 November 2019
The original version of this article unfortunately contains mistakes introduced by the publisher during the production process. The mistakes and corrections are described in the following list
Notes
This requires meta-arguments to have a non-empty metapremise set (though this is compatible with a metapremise—an argument—having an empty premise set). But this seems unobjectionable. A “meta-argument” with nothing above the line is just the assertion of some entailment. Even if we allow these degenerate kinds of meta-argument, the result still holds. Proof. For reductio suppose some zero-metapremise meta-argument is K-metavalid but J-meta-invalid. Let its conclusion be Δ ⊩ β. Then Δ ⊩Kβ but \({\varDelta } \nvdash _{J} \beta \). So \(\beta \in Cn_{\vdash _{K}} ({\varDelta })\) and \(\beta \notin Cn_{\vdash _{J}} ({\varDelta })\). Hence the K- and J-closures of Δ are not the same set. Contradiction with the closure equivalence of K and J.
This requires that J is not the empty logic. The WLOG assumption a few lines above requires that K is not empty as well. Note that Γ could be the empty set, however.
References
Cobreros, P., Égré, P., Ripley, D., van Rooij, R. (2013). Reaching transparent truth. Mind, 122(488), 841–866. https://doi.org/10.1093/mind/fzt110.
Dicher, B. (2018). Variations on intra-theoretical logical pluralism: internal versus external consequence. Philosophical Studies. https://doi.org/10.1007/s11098-018-1199-z.
Etchemendy, J. (1999). The concept of logical consequence. CSLI Publications.
Fine, K. (1972). Vagueness, truth and logic. Synthese, 30(3/4), 265–300. http://www.jstor.org/stable/20115033.
Haack, S. (1978). Philosophy of logics. Cambridge University Press.
Hjortland, O.T. (2016). Anti-exceptionalism about logic. Philosophical Studies, 174(3), 631–658. https://doi.org/10.1007/s11098-016-0701-8.
Maddy, P. (2014). The logical must: Wittgenstein on logic. Oxford University Press. https://doi.org/10.1093/acprof:oso/9780199391752.001.0001.
Priest, G. (1979). The logic of paradox. Journal of Philosophical Logic, 8(1), 219–241. https://doi.org/10.1007/bf00258428.
Priest, G. (2006). Doubt truth to be a liar. Oxford University Press. https://doi.org/10.1093/0199263280.001.0001.
Quine. W.V. (1986). Philosophy of logic, 2 edn. Harvard University Press.
Read, S. (1988). Relevant logic: a philosophical examination of inference. Blackwell.
Russell, G. (2015). The justification of the basic laws of logic. Journal of Philosophical Logic, 44(6), 793–803. https://doi.org/10.1007/s10992-015-9360-z.
Shapiro, L. (2011). Deflating logical consequence. The Philosophical Quarterly, 61(243), 320–342. https://doi.org/10.1111/j.1467-9213.2010.678.x.
Tarski, A. (1936/1956). On the concept of logical consequence, chapter 16. In Woodger, J.H. (Ed.) (trans) Logic, semantics, metamathematics: papers from 1923 to 1938 (pp. 409–420): Clarendon Press.
Williamson, T. (1994). Vagueness. Routledge.
Williamson, T. (2013). Modal logic as metaphysics. Oxford University Press. https://doi.org/10.1093/acprof:oso/9780199552078.001.0001.
Williamson, T. (2017). Semantic paradoxes and abductive methdology, chapter 13. In Armour-Garb, B. (Ed.) Reflections on the liar (pp. 325-346): Oxford University Press.
Williamson, T. (2018). Supervaluationism and good reasoning. THEORIA, 33 (3), 521–537. https://doi.org/10.1387/theoria.18490.
Acknowledgments
This paper has benefitted from discussion with Jared Henderson, Nathan Kellen, Gurpreet Rattan, and participants of the Society for Exact Philosophy annual conference at the University of Connecticut in 2018. I am also grateful to Joanna Lawson, Erik Stei, and Simon Varey for reading drafts. Special thanks are due to Gillian Russell. The idea for the paper came from participating in her Philosophy of Logic graduate class at UNC in the spring semester of 2017. She also provided extremely helpful feedback, and much-needed encouragement, at multiple stages of the writing and rewriting process. Finally, thank you to two anonymous reviewers, whose generous comments improved the paper greatly.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix
Appendix
Meta-Argumentative Equivalence (MAE)
If two logics are closure equivalent, then they are (globally) meta-argumentatively equivalent.
Proof
Let K and J be closure-equivalent logics. Let Γ1…Γn and Δ be arbitrary (possibly empty) premise sets, and α1…αn and β be arbitrary conclusions. Consider the following arbitrary meta-argument:
We show that Arg is K-metavalid iff Arg is J-metavalid. Consider the left-to-right direction.
Suppose for reductio that K metavalidates Arg, but that J does not. There are two cases to consider.
Case 1. K metavalidates Arg because both its metapremises and metaconclusion are K-valid, and J fails to metavalidate Arg. More precisely:
- (i)
Γ1 ⊩Kα1⋯Γn ⊩Kαn
- (ii)
Δ ⊩Kβ
- (iii)
Γ1 ⊩Jα1⋯Γn ⊩Jαn
- (iv)
\({\varDelta } \nvdash _{J} \beta \)
By (ii), \(\beta \in Cn_{\vdash _{K}} ({\varDelta })\). But by (iv), \(\beta \notin Cn_{\vdash _{J}} ({\varDelta })\). Hence \(Cn_{\vdash _{K}} ({\varDelta }) \neq Cn_{\vdash _{J}} ({\varDelta })\). But, since K and J are closure equivalent, \(Cn_{\vdash _{K}} ({\varDelta }) = Cn_{\vdash _{J}} ({\varDelta })\). Contradiction.
Case 2. K metavalidates Arg because at least one of its metapremises is invalid (thus the conditional “if all metapremises, then metaconclusion” obtains), and J fails to metavalidate Arg. More precisely, and (iii) and (iv) hold as in Case 1, and:
- (v)
For some Γi and αi, \({\varGamma }_{i} \nvdash _{K} \alpha _{i}\).Footnote 1
By (v), \(\alpha _{i} \notin Cn_{\vdash _{K}} ({\varGamma }_{i})\). But by (iii) \(\alpha _{i} \in Cn_{\vdash _{J}} ({\varGamma }_{i})\). Hence \(Cn_{\vdash _{K}} ({\varGamma }_{i}) \neq Cn_{\vdash _{J}} ({\varGamma }_{i})\). But since K and J are closure equivalent, \(Cn_{\vdash _{K}} ({\varGamma }_{i}) = Cn_{\vdash _{J}} ({\varGamma }_{i})\). Contradiction.
Either way, we get a contradiction. It follows by reductio that if Arg is K-metavalid, then it is J-metavalid.
The proof for the right-to-left direction of the biconditional is almost exactly the same.
Hence, Arg is K-metavalid iff Arg is J-metavalid.
Since this was proved on the assumption that K and J are closure equivalent, and since the logics and meta-arguments were arbitrary, it follows that if two logics are closure equivalent, then they are meta-argumentatively equivalent. □
Converse of MAE
If two logics are (globally) meta-argumentatively equivalent, then they are closure equivalent.
Proof
Suppose for reductio that K and J are meta-argumentatively equivalent, but they are not closure equivalent: for some Δ, \(Cn_{\vdash _{K}} ({\varDelta }) \neq Cn_{\vdash _{J}} ({\varDelta })\). Without loss of generality assume that:
- (i)
\(\delta \in Cn_{\vdash _{K}} ({\varDelta })\)
- (ii)
\(\delta \notin Cn_{\vdash _{J}} ({\varDelta })\)
By (i), Δ ⊩Kδ. Hence:
- (iii)
Any meta-argument to the metaconclusion that Δ ⊩ δ is K-metavalid.
Consider in particular the following meta-argument, where:Footnote 2
-
(iv)
Γ ⊩Jα
By (iii), Arg1 is K-metavalid. So, by the assumption of meta-argumentative equivalence, Arg1 is J-metavalid too. Either (a) the metapremise is J-invalid, or (b) the metaconclusion is J-valid. But (a) contradicts (iv), because then both \({\varGamma } \nvdash _{J} \alpha \) and Γ ⊩Jα. And (b) contradicts (ii), because then Δ ⊩Jδ. Hence \(\delta \in Cn_{\vdash _{J}} ({\varDelta })\) and \(\delta \notin Cn_{\vdash _{J}} ({\varDelta })\). Either way, contradiction.
It follows by reductio that if two logics are meta-argumentatively equivalent, then they are closure equivalent. □
Rights and permissions
About this article
Cite this article
Blake-Turner, C. Deflationism About Logic. J Philos Logic 49, 551–571 (2020). https://doi.org/10.1007/s10992-019-09529-5
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10992-019-09529-5