Reductive Theories of Modality
Sider (Ted)
Source: M. J. Loux and D.W. Zimmerman, eds., The Oxford Handbook of Metaphysics (OUP, 2003): 180–208
Paper - Abstract

Paper StatisticsBooks / Papers Citing this PaperColour-ConventionsDisclaimer

Author’s Introduction1 (Modality)

  1. Logic begins but does not end with the study of truth and falsity. Within truth there are the modes of truth, ways of being true: necessary truth and contingent truth. When a proposition is true, we may ask whether it could have been false. If so, then it is contingently true. If not, then it is necessarily true; it must be true; it could not have been false. Falsity has modes as well: a false proposition that could not have been true is impossible or necessarily false; one that could have been true is merely contingently false. The proposition that some humans are over seven feet tall is contingently true; the proposition that all humans over seven feet tall are over six feet tall is necessarily true; the proposition that some humans are over seven feet tall and under six feet tall is impossible, and the proposition that some humans are over nine feet tall is contingently false.
  2. Of these four modes of truth, let us focus on necessity, plus a fifth: possibility. A proposition is possible if it is or could have been true; hence propositions that are either necessarily true, contingently true, or contingently false are possible.
  3. Notions that are similar to the modes of truth in being concerned with what might have been are called modal. Dispositions are modal notions, for example the disposition of fragility. Relatedly, there are counterfactual conditionals, for example "if this glass were dropped, it would break." And the notion of supervenience is modal2. But let us focus here on necessity and possibility.
  4. Modal words are notoriously ambiguous (or at least context-sensitive). I may reply to an invitation to give a talk in England by saying "I can't come; I have to give a talk in California the day before". This use of "can't" is perfectly appropriate. But it would be equally appropriate for me to say that I could cancel my talk in California (although that would be rude) and give the talk in England instead. What I cannot do is give both talks. But wait: it also seems appropriate to say, in another context, that given contemporary transportation, one can give a talk in California one day and England the next. It may be very exhausting, but one can do it. What one cannot do is give a talk in California and then give a talk in England the next hour. But in yet another context one could say the following: "Given the limits on travel faster than the speed of light, one cannot give a talk on Earth, and then another on Alpha Centauri an hour later. But one could give a talk in California and then an hour later give a talk in England." Finally, even this performance seems appropriate: "The laws of nature could have been different. Supra-luminal travel might have permitted by the laws of nature. One could (if the laws had indeed allowed supra-luminal travel) have given a talk on Earth, and then another an hour later on Alpha Centauri, 4.12 x 1013 km away. What is impossible is giving talks on Earth and Alpha Centauri at the very same time."
  5. There are, therefore, different "strengths" of necessity and possibility, which can be signified by modal words (like 'can') in different contexts. Philosophers have tended to concentrate on a very broad sort, so-called "metaphysical" possibility and necessity. According to many, it is metaphysically possible that the laws of nature be different, that the past be different from what it actually was, and so on3. All of the scenarios in the last paragraph — giving a talk in England, giving a talk in California one day and England the next, giving a talk in California at one moment and a talk in England an hour later, giving a talk on Earth one moment and on Alpha Centauri an hour later — are metaphysically possible. What is not metaphysically possible? Almost everyone agrees that contradictions are metaphysically impossible — it is metaphysically impossible to both give a talk in California and also not to give a talk in California. And everyone who accepts the legitimacy of the notion of analyticity — of truth that is in some sense guaranteed by meaning — agrees that the negations of analytic sentences like 'all bachelors are unmarried' are impossible. But it is usually thought that there exist further impossibilities. Examples might include the existence of a round square, someone's being taller than himself, someone's being in two places at once, George W. Bush's being a donkey, there existing no numbers, and there existing some water that is not made up of H2O. Exactly what is metaphysically impossible beyond logical and analytic contradictions is unclear; this unclarity is what makes the analysis of metaphysical possibility and necessity so difficult. But it is metaphysical possibility and necessity that most concerns philosophers, and so from now on it is on the metaphysical sense of the modal notions that I will focus. It is common to distinguish between de re and de dicto modality.
  6. The contrast may be brought out with this example:
    • (De dicto) Necessarily, the number of the planets is odd
      □ [(the x: Nx) Ox]
    • (De re) The number of the planets is such that it is necessarily odd
      (The x: Nx) □ [Ox]
    The de dicto sentence is false. It claims that it is necessary that the number of the planets is odd, whereas there clearly might have been 6 or 8 planets. The de re sentence, however, is presumably true. It claims of the number that actually numbers the planets — namely, 9 — that it is necessarily odd. Assuming with orthodoxy that mathematical facts are necessary, this is true: the number 9 itself is necessarily odd. The de dicto sentence claims that a certain descriptive claim is necessary: it is necessary that the number picked out by the description 'the number of the planets', whatever that might turn out to be, is odd. In each possible world, whatever number is the number of the planets in that world must be odd. In contrast, the de re sentence uses the description 'the number of the planets' to single out a certain individual, the number 9, but then goes on to make a modal claim about that number itself; the description used to single out 9 plays no role in evaluating the modal claim about 9. In each possible world, 9 itself must be odd, never mind whether 9 is the number of the planets in that world.
  7. There is a grammatical contrast between the de re and the de dicto sentences that is made clearer by the symbolically regimented versions of those sentences. In the de re sentence there is a variable in the scope of the modal operator □ (symbolizing 'it is necessary that') that is bound to a quantifier outside the scope of the □ whereas in the de dicto sentence no quantification into the scope of modal operators occurs. A further example: the false sentence4 'Possibly, some bachelor is unmarried', or '◇∃x(Bx&~Mx)' is de dicto, whereas the true sentence 'Some bachelor is possibly unmarried', or '∃x(Bx & ◇~Mx)' is de re, since the variable x occurs inside the scope of the ◇ but is bound to the quantifier ∃x which occurs outside the scope of the ◇. This grammatical or syntactic way of drawing the de re / de dicto distinction is common, and can be extended to natural language given the existence of natural language analogs of modal operators and variable binding. However, for present purposes it will be useful to (somewhat stipulatively) draw the distinction slightly differently. Specifically, in addition to sentences with quantification into modal contexts, let us count as de re modal sentences in which "directly referential terms" occur within the scope of modal operators. Directly referential terms are terms whose prepositional contributions are simply their referents, for example proper names and indexicals. The reason for counting these sentences as de re is that they attribute modal properties to objects simpliciter, rather than under descriptions5.
  8. Modality is important to philosophy for many reasons. A first reason derives from philosophy's traditional association with logic. Advances in modal logic in the middle of the 20th century provided a reason to be interested in the modalities. Moreover, propositions that are logically true seem necessarily true. Another source of modality's importance is that necessary truth, according to one tradition, demarcates philosophical from empirical inquiry. Science identifies contingent aspects of the world, whereas philosophical inquiry reveals the essential nature of its objects; philosophical propositions are therefore necessarily true when true at all.
  9. But the most important source of importance derives from modality's connections with epistemology and philosophy of language. These connections are at the core of analytic philosophy. The propositions identified by traditional epistemology as those that can be known a priori, independent of sensory experience, seem necessary. These are generally agreed to include the propositions of logic as well as analytic truths. Whether there are other a priori propositions was one of the great questions of 17th and 18th century epistemology, and the debate continues to this day. But it was generally agreed until recently that all a priori propositions are necessarily true. Indeed, before the publication of "Kripke (Saul) - Naming and Necessity", it was not uncommon to identify a prioricity with necessity.
  10. Given the compelling examples of necessary a posteriori propositions given by Kripke, and by Hilary Putnam (1975), as well as Kripke's examples of contingent a priori propositions, this identification is no longer made. And given "Quine (W.V.) - Two Dogmas of Empiricism" critique of analyticity, some have doubted the connection between analyticity and necessity, others the sense of the notion of necessity itself. But despite this, many of the important traditional connections remain. It is still relatively common to claim that some necessary propositions are a priori, thus, the nature of necessity is relevant to epistemology, for what is necessary truth, that it can be ascertained without sensory input? And despite Quine, there remains an overwhelming temptation to think that the notion of linguistic convention has some legitimate application, and some connection with the traditional notion of necessity.


See Ted Sider - Reductive Theories of Modality.

In-Page Footnotes

Footnote 1: Footnote 2: Footnote 3: Footnote 4: Footnote 5:

Text Colour Conventions (see disclaimer)

  1. Blue: Text by me; © Theo Todman, 2019
  2. Mauve: Text by correspondent(s) or other author(s); © the author(s)

© Theo Todman, June 2007 - Jan 2019. Please address any comments on this page to File output:
Website Maintenance Dashboard
Return to Top of this Page Return to Theo Todman's Philosophy Page Return to Theo Todman's Home Page