Skip to main content

That solution to Prior’s puzzle


Prior’s puzzle is a puzzle about the substitution of certain putatively synonymous or coreferential expressions in sentences. Prior’s puzzle is important, because a satisfactory solution to it should constitute a crucial part of an adequate semantic theory for both proposition-embedding expressions and attitudinal verbs. I argue that two recent solutions to this puzzle are unsatisfactory. They either focus on the meaning of attitudinal verbs or content nouns. I propose a solution relying on a recent analysis of that-clauses in linguistics. Our solution is superior, as it not only avoids the problems faced by previous solutions, but it also brings developments in linguistics in line to solve an old puzzle in philosophy.

This is a preview of subscription content, access via your institution.


  1. I take propositions to be meanings of well-formed sentences, whether these are sets of situations, possible worlds, or other entities that can be indices of evaluation for a given sentence. I take sentences to be strings of symbols arranged according to the rules of a given language. Throughout the paper I use italics to mention linguistic expressions. I use expressions such as means P and refers to P interchangeably, as refers to sounds more natural when paired with entities, whereas means P sounds better when paired with sentences. I take the meaning of an expression to be what it refers to. The meaning/reference of a sentence is the proposition expressed by the sentence, whereas the meaning/reference of an entity description, e.g., the proposition that P, is the entity described, e.g., the proposition that P, unless otherwise stated.

  2. Such considerations became even more forceful with analyses of attitudes as formal relations (e.g. Fodor 1987).

  3. Substituting the that-clause in (3-a) with the proposition description in (3-b) does not affect grammaticality; but this is not always the case:



    Sally hopes that Fido barks.



    *Sally hopes the proposition that Fido barks.

    This is Rundle’s puzzle (Rundle, 1967). Nebel (2019: §2) shows it can be solved by certain assumptions about how to translate attitudinal verbs when they are complemented with content nouns. With Nebel’s solution, Rundle’s puzzle reduces to Prior’s puzzle.

  4. Some conclude from the puzzle that the analysis of attitude verbs that proposes that attitude verbs express relations between individuals and propositions is wrong (Bach, 1997; McKinsey, 1999). I will bracket these views in my discussion.

  5. Prior’s puzzle is old (Prior, 1963) and naturally these two views are not the only solutions offered. For instance, Parsons1993) , Mofett (2003) and Harman (2003) offer solutions that take that-clauses to sometimes refer to entities other than propositions. The solution offered here is distinct from theirs, because we deny that that-clauses refer to entities. We will see that our proposal is intended to cover where their proposals go right while also being more general.

  6. By contentful entities I mean entities that felicitously take that-clauses as complements such as the rumor that P, the evidence that P, the thesis that P, the statement that P. The expressions that refer to these entities are called content nouns in the linguistics literature (Moulton, 2009: §2.2).

  7. Of course, fearing the proposition that Fido barks sounds nonsensical. How can one fear a proposition? This is exactly the point. According to King, the relation expressed by fear in (3-b) generates a nonsensical meaning, whereas the relation expressed by fear in (3-a) does not.

  8. Does this establish (rather than merely provide evidence) that fear is not polysemous? The zeugma test for establishing or rejecting polysemy might not suffice to settle the existence of polysemy. The regular examples of zeugma to establish polysemy are straightforward as seen in (4). However, the lack of oddity or zeugma might not be sufficient to establish univocality, since there seem to be examples where there is no zeugma, but polysemy might still be needed:



    I believe my mother.



    I believe that my mother helped the neighbors.



    I believe my mother and that she helped the neighbors.

    Here (i-a-b) unify into a single felicitous instance; yet Nebel’s account of the synonymy of (2-a-b) in Prior’s puzzle depends on a claim of polysemy for the verb believe. Nebel himself (2019: 97) finds a variant of (i-c) strange: “I believe my mother and that Fido barks”. I am tempted to think this is due to the irrelevant-sounding complements in the examples he uses. If we have some anaphora in the that-clause complementing believe, his examples sound fine. I am therefore inclined to think that zeugma test might not suffice to establish univocality, even though it might suffice to establish polysemy. If this is also reader’s suspicion, they can read my argument as specifically targeting Nebel, who takes the zeugma test to be conclusive against King’s account. Thanks to Ian Phillips for questioning the zeugma test and for detailed discussion.

  9. Nebel attributes this solution jointly to Montague (1973) and Higgins (1973).

  10. van Elswyk (2022) notes but does not develop this worry for Nebel, with reference to Moltmann (2003).

  11. For the objection against Nebel to succeed, all that is required is a distinction between the role propositional concepts play in linguistic theories, which are abstract and the types of things that cannot prima facie stand in causal relations, and other contentful entities which can. Of course, there is still the metaphysical question about what exactly contentful entities are. What kind of a thing is a rumor or a statement such that it can hurt or surprise someone, while propositional concepts cannot? Are they utterance tokens of the type rumor or a more general type? (For the type-token distinction, see Wetzel, 2018). Do they belong to a social ontology where the society postulates the existence of these categories that can stand in these relations, because they find them useful enough to organize the structure of their community? (For an introduction to social ontology, see Epstein, 2018).

  12. In Montagovian type theory (1973; for a textbook exposition, see Gamut, 1990: 81–122) where \(e\) is the type of entities, \(s\) is the type of possible worlds and \(t\) is the type of truth values, causal verbs can either be of type \(\left\langle e \right.,\left. {\left\langle {e,t} \right\rangle } \right\rangle\), which takes a pair of entities to a truth value or they can be \(\left\langle {\left\langle s \right\rangle } \right.,\left. {\left\langle {s,t} \right\rangle } \right\rangle ,\left. {\left\langle {e,t} \right\rangle } \right\rangle\) which takes a pair of propositional concepts and entities to a truth value. However, in order for causal predicates to be univocal in being of either type, the type \(e\), the set of entities, and type \(\left\langle s \right.,\left. {\left\langle {s,t} \right\rangle } \right\rangle\), the set of propositional concepts, must be identical. This is absurd.

  13. The proposal has its roots in Kratzer (2006). It has been further developed in Moulton (2009).

  14. There are technical fixes that make this possible (see especially Potts, 2002). But such fixes lead to problems that are analogous to the problems we have raised for Nebel (see Moulton, 2009: 36).

  15. The distinction goes back to Frege (1951) and was sharpened in Strawson (1959).

  16. \(\left[ {\left[ \cdot \right]} \right]\) is the interpretation function which maps expressions to their meanings. I use as an exception Roman fonts in the interpretation function to mention expressions to avoid confusion with italics in the metalanguage and the lambda expressions to denote predicates that can be saturated by entities of the specified type. For instance, the meaning of the predicate run is expressed as \(\lambda\left( s \right)\) where s stands for the type of entities. Also the assumption that xc is a contentful entity is for convenience. It is clear that the relation expressed by fear can take non-contentful entities as arguments, too, as in: Carlos fears snakes.

  17. An anonymous reviewer helpfully inquires what happens to Partee’s puzzle under the current account, since sentences like the number of insect species on Earth surprised Mary seems to give rise to similar puzzles. I believe we can give a similar treatment for the example mentioned. For the expression the number of in particular, we might help ourselves to a metalanguage function # similar to cont in (11) that takes the actual size of a collection of countable entities (instead of contentful entities) and returns a contextually specified cardinal number. The meaning of the number of would be a predicate equating the size of the collection to the cardinal number specified. The predicate surprise is then saturated by the size of the collection in question, which seems to be the right kind of entity for a causal predicate. This analysis gives rise to an ambiguity, which can be confirmed to exist. The interpretation that the size of the collection of insect species on Earth is such that its number is n and it surprised Mary is true. However, the interpretation that the number n is such that it is the number of the size of the collection of insect species on Earth and it surprised Mary is false. I take the prediction of such ambiguity a good result for our analysis.

  18. The finest-grained way of specifying such subsets would be to let each attitudinal verb determine its own subset of contentful entities. However, as Anand and Hacquard (2014) show, there may be generalizations that hold among verbs, too. For instance, cognitive factives such as know, discover, and realize presuppose the truth of the content of their that-clauses, so their domains will consist of factive contentful entities.

  19. Kratzer (2006: §9) divides that-clauses into categories according to their role, e.g., logophoric that-clauses, factive that-clauses. Our treatment here captures this by postulating a restricted existential closure of the predicational form.

  20. Assumption of distinct contentful entities for distinct attitudinal verbs may seem like our solution to Prior’s puzzle is closer to Parsons (1993), Mofett (2003) and Harman (2003) who argue that that-clauses sometimes refer to entities other than propositions than I acknowledge. However, this is not the case. Under our solution that-clauses are not referring expressions at all. They are predicates that associate the propositional content of many different entities selected by their attitudinal verbs with their propositional content. Our analysis assigns a uniform meaning to that-clauses rather than stipulating that they refer to different entities in different contexts. I believe that our analysis is more explanatory, because it assigns the same meaning to an expression which occurs verbatim in different contexts, while varying the implicit contentful entities with the varying attitudinal verbs. Thanks to an anonymous referee for pressing me to clarify the difference among these analyses and ours.

  21. Thanks to an anonymous reviewer for rightfully questioning if the present solution extends to proper name variants of Prior’s puzzle. Examples are from Nebel (2019, p. 93).

  22. Thanks to an anonymous reviewer for drawing my attention to this development and the following discussion.

  23. Thanks to an anonymous reviewer for pressing on this question.

  24. Britannica, T. Editors of Encyclopaedia (2011, August 5). Propositional attitudeEncyclopedia Britannica.


  • Anand, P., Hacquard, V. (2014). Factivity, belief and discourse. In The art and craft of semantics: A Festschrift for Irene Heim, vol 1. (Eds.) C. Luka & U. Sauerland. MITWPL 70, (pp. 69–90).

  • Bach, K. (1997). Do belief reports report beliefs? Pacific Philosophical Quarterly, 78(3), 215–241.

    Article  Google Scholar 

  • Epstein, B. (2018). Social ontology. In The Stanford encyclopedia of philosophy (Summer 2018 En), (Ed.) E. N. Zalta.

  • D’Ambrosio, J. (2020). Ramsification and the ramifications of Prior’s puzzle. Noûs.

    Article  Google Scholar 

  • D’Ambrosio, J. (2021). Prior’s puzzle generalized. Philosophy and Phenomenological Research.

  • Fodor, J. (1987). Psychosemantics: The problem of meaning in the philosophy of mind. MIT Press.

    Book  Google Scholar 

  • Frege, G. (1951). On concept and object. Mind, 60(238), 168–180.

    Article  Google Scholar 

  • Gamut, L. T. F. (1990). Logic, language, and meaning volume 2: Intensional logic and logical grammar. University of Chicago Press.

    Book  Google Scholar 

  • Harman, G. (2003). Category mistakes in M&E. Philosophical Perspectives, 17(1), 165–180.

    Article  Google Scholar 

  • Higgins, F. R. (1973). The pseudo-cleft construction in English. PhD dissertation, Massachusetts Institute of Technology.

  • King, J. C. (2002). Designating propositions. Philosophical Review, 111(3), 341–371.

    Article  Google Scholar 

  • Kratzer, A. (2006). Decomposing attitude verbs. Talk in honor of A. Mittwoch. The Hebrew University, Jerusalem.

  • Ladusaw, W. A., & Chung, S. (2003). Restriction and saturation. MIT Press.

    Google Scholar 

  • Mofett, M. A. (2003). Knowing facts and believing propositions: A solution to the problem of doxastic shift. Philosophical Studies, 115(1), 81–97.

    Article  Google Scholar 

  • Moltmann, F. (2003). Propositional attitudes without propositions. Synthese, 135(1), 77–118.

    Article  Google Scholar 

  • Montague, R. (1973). The proper treatment of quantification in ordinary English. In Philosophy, language, and artificial intelligence, (Eds.) J. Kulas, J. H. Fetzer, T. L. Rankin, (pp. 141–162).

  • McKinsey, M. (1999). The semantics of belief ascriptions. Noûs, 33(4), 519–557.

    Article  Google Scholar 

  • Moulton, K. (2009). Natural selection and the syntax of clausal complementation. PhD dissertation, UMass, Amherst.

  • Nebel, J. M. (2019). Fears, fears, and other grammatical scarecrows. Philosophical Review, 128(1), 63–105.

    Article  Google Scholar 

  • Parsons, T. (1993). On denoting propositions and facts. Philosophical Perspectives, 7, 441–460.

    Article  Google Scholar 

  • Partee, B. H. (1974). Opacity and scope. In Semantics and philosophy (Eds.) M. K. Munitz, P. K. Unger. New York University Press, (pp. 81–101).

  • Potts, C. (2002). The lexical semantics of parenthetical-As and Appositive- which. Syntax, 5(1), 55–88.

    Article  Google Scholar 

  • Prior, A. N. (1963). Symposium: Oratio obliqua. Proceedings of the Aristotelian Society, Supplementary Volumes, 37, 115–146.

    Article  Google Scholar 

  • Rundle, B. (1967). Transitivity and indirect speech”. Proceedings of the Aristotelian Society, 68, 187–206.

    Article  Google Scholar 

  • Strawson, P. (1959). Individuals: An essay in descriptive metaphysics. Routledge.

  • van Elswyk, P. (2022). The linguistic basis for propositions. In The Routledge handbook of propositions (Ed) T. Chris.

  • Wetzel, L. (2018). Types and tokens. The Stanford encyclopedia of philosophy (Fall 2018 Edn), E. N. Zalta (Ed.).

Download references


This paper has been developed for Proseminar taught during Fall 2020-Spring 2021 at JHU. First and foremost, I thank Hanna Pickard and Ian Phillips for their excellent organization. Special thanks to Ian Phillips for reading multiple drafts and very helpful comments and two anonymous reviewers for Philosophical Studies for detailed comments and inquiries that improved the quality of paper immensely. Thanks to Hanna Pickard and Jacob Lettie for detailed discussion of many points in the paper. For their generous feedback, thanks also to Elsie Campbell, Rooke Christy, Oscar Ro- driguez Cruz-Haker, Nick Eggert, Fergus Horan, Matthew Morgado, Sonya Ringer and Hannah Winckler-Olick.

Author information

Authors and Affiliations


Corresponding author

Correspondence to Hüseyin Güngör.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Güngör, H. That solution to Prior’s puzzle. Philos Stud 179, 2765–2785 (2022).

Download citation

  • Accepted:

  • Published:

  • Issue Date:

  • DOI:


  • Propositions
  • Propositional content
  • That-clauses
  • Substitution failures
  • Propositional attitude reports