Preprint
Abstract
In “Avoiding Metaphysical Pratfalls and Positivist Pitfalls” (Synthese 2026), Paul A. Roth discusses Virmajoki’s critique of the role narrative sentences play in historical explanation. Roth charges Virmajoki with contradiction, category mistakes, and non sequiturs. But these charges rest on a misreading. Virmajoki’s argument is not about whether narrative sentences are true – he grants they are – but about whether they attribute properties to past events or express relations between events across time.
1. The Property-Relation Distinction
Consider “V is the father of U.” This is true. But fatherhood is not a monadic property of V. It is a relation between V and U. V does not carry fatherhood as an intrinsic attribute. It comes into existence when U is born, yet no property of V changes at that moment. Likewise, when historians say “The Thirty Years War began in 1618,” they are not attributing a new property to the events of 1618. They are recognizing a relation connecting those events to what unfolded over subsequent decades. The sentence is true without the events of 1618 acquiring an intrinsic characteristic called “being the beginning of the war.”
2. Roth’s Misreading
Roth claims Virmajoki is committed to a contradiction: “Although it is true that The Thirty Years War began in 1618, it is not true that The Thirty Years War began in 1618.” But Virmajoki would never deny that the sentence is true. Virmajoki denies that the sentence predicates a property of 1618-events. Roth generates his contradiction only by treating “true of a time” and “attributing a property to events at that time” as equivalent – but that equivalence is exactly what Virmajoki rejects.
Roth’s principal positive argument is a substitution. He sets up three steps: (1) Philadelphia was founded in 1682; (2) Philadelphia is identical with the city that served as the first capital of the United States; (3) therefore, the city that served as the first capital of the United States was founded in 1682. Step (3) follows from (1) and (2) by Leibniz’s Law – the rule that terms in an identity statement can be interchanged in any sentence without changing its truth value. Since (1) is true and (2) states an identity, (3) must also be true. And (3) is a narrative sentence: it uses a description (“the city that served as the first capital”) that only became available long after 1682. Roth concludes that narrative sentences can state truths about past times, and that no backward causation is needed – just logic.
This, however, does not answer Virmajoki’s point. Roth is right that narrative sentences can be true, but that does not show that they attribute intrinsic properties to past events. The problem is that Roth’s Accumulation Thesis trades on an ambiguity between truth about an earlier time and intrinsic alteration of that earlier event. Virmajoki would accept (3) as true without hesitation. His point is that the logical form of such sentences is relational: “was founded in 1682” connects a city to a date; “served as the first capital” connects a city to a political role that came later. The substitution changes which referring term occupies the argument position but does not alter the predicate’s arity – a two-place relation remains a two-place relation after the swap. Leibniz’s Law preserves truth values; it does not convert relations into properties. Roth himself concedes in a footnote that the argument goes through on either reading – property or relation – but fails to see that this concession renders the argument powerless against Virmajoki’s position, which is precisely about distinguishing the two.
3. Confusing Properties, Relations, and Truths
Let’s go back to basics. Frege’s “The Thought” (1956; originally published in 1918) provides a framework that directly undercuts Roth’s central assumption – that if a narrative sentence is true, it must state a truth at the earlier time.
Frege argues that thoughts (his term for what we now call propositions) belong to a “third realm”. They are neither physical objects nor mental ideas. “Thoughts are neither things of the outer world nor ideas. A third realm must be recognized” (1956, p. 302). A thought is the sense of a declarative sentence, and it is “timelessly true, true independently of whether anyone takes it to be true. It needs no bearer. It is not true for the first time when it is discovered, but is like a planet which, already before anyone has seen it, has been in interaction with other planets” (1956, p. 302; emphasis added).
Crucially, this timelessness applies not just to mathematical truths but to thoughts about temporal events. Frege addresses the apparent objection that some thoughts seem to change truth value over time – for example Frege argues how, “this tree is covered with green leaves” might seem true now and false in six months. His reply is that these are not the same thought: a sentence without a time-indication does not express a complete thought at all. “Only a sentence supplemented by a time-indication and complete in every respect expresses a thought. But this, if it is true, is true not only today or tomorrow but timelessly” (1956, p. 309; emphasis added). The time-indication belongs to the content of the thought, not to its truth. “The truth, whose recognition lies in the form of the indicative sentence, is timeless” (1956, p. 310). Frege also identifies facts with true thoughts: “A fact is a thought that is true” (1956, p. 307).
This basic foundation about timelessness of truths is directly relevant on the confusions in the philosophy of history.
First, it dissolves the framework that generates Roth’s argument. Roth insists that narrative sentences must state truths at past times, and he takes the denial of this to require backward causation or some other metaphysical absurdity – which, to be fair, critics have pointed out as a reductio ad absurdum (Virmajoki 2025). But once one understands Frege’s account, truths are not the sort of thing that have temporal locations. The thought that the Thirty Years War began in 1618 is timeless; it is true simpliciter, not true “in 1618.” The time-indication “”1618” is part of the thought’s content – it tells what the thought is about – but the truth of the thought is not itself located at any time. The entire framework of truths being temporally located, which generates both Roth’s defense of the Accumulation Thesis (“that statements true about a moment in the past continue to accumulate even after that moment has passed” [Roth 2026]) and his charge of contradiction against Virmajoki, results from confusion between the content of a thought with the status of its truth.
Roth’s argument depends on treating truth as if it were temporally located. He slides from the claim that a sentence is true about1618 to the stronger claim that this truth somehow belongs to 1618. But on a view like Frege’s, a complete thought with a temporal indication is true timelessly. The date belongs to the content of the thought, not to the truth-status of the thought. Roth therefore confuses the temporal location of the events described with the alleged temporal location of truth itself (if that makes any sense in the first place).
Second, Frege’s framework preserves the property-relation distinction at the level of logical form. For Frege, the logical structure of a thought – whether it involves a one-place concept applied to one object (“x is red”) or a two-place relation applied to two objects (“x began y”) – is a genuine feature of the thought’s structure, not a surface-grammatical artefact that substitution can dissolve. Roth’s Philadelphia argument changes the referring terms but does not change the arity of the predicate. Substituting “the city that served as the first capital” for “Philadelphia” does not convert a relational predicate into a monadic one. The difference between attributing a property to an event and expressing a relation between events is a difference in the logical structure of the thought, and no amount of term-substitution can alter it.
4. Conclusion
Roth succeeds in showing narrative sentences are true. Hardly anyone denies this. Roth’s actual non sequitur is: from the premise that narrative sentences are true, he concludes they attribute properties to past times. The premise is correct; the conclusion does not follow. What is needed – and what Roth never provides – is a direct argument that the property-relation distinction is inapplicable to narrative sentences. This is, it seems, due to confusions about properties, relations, and truths that have plagued the philosophy of history for a long time.
Return to the father example. When U is born, it becomes true that V is the father of U. No one would infer from this that fatherhood is a property V possessed before U’s birth, or that U’s birth retroactively altered V’s intrinsic nature, or that some metaphysically suspect mechanism must have acted on V’s past to make it so. What happened is simpler: a relation came into existence. The statement “V is the father of U” is true – timelessly true, on Frege’s account – but its truth consists in a relation between two individuals, not in a property of one of them. Likewise, “The Thirty Years War began in 1618” is timelessly true, but its truth consists in a cross-temporal relation between the events of 1618 and what followed, not in a property those events acquired. No backward causation is needed, no metaphysical absurdity arises, and no Accumulation Thesis – in the strong sense of new properties accruing to past moments – is required. We simply gain new relational knowledge about how events are connected as time passes and history unfolds.
What is needed is a direct argument that the property-relation distinction is somehow inapplicable to narrative sentences. Until that argument is given, the charge of metaphysical absurdity falls not on Virmajoki but on a framework that conflates (i) the timeless truth of a relational proposition with (ii) the temporal location of the events it describes. V’s fatherhood is true but does not exist where V exist. Claiming that the fatherhood follows wherever V goes is puzzling thought, to say the least.
There might be reasons to reject Fregean ideas of timelessness of truths, ideas about distinction between properties and relations, and so on, but these arguments must be clearly presented and argued for, not based on implicit assumptions that most fields – other than the philosophy of history – do not share.
References
Frege, G. (1956). The thought: A logical inquiry. (A. M. & M. Quinton, Trans.). Mind, 65(259), 289–311. (Original work published 1918)
Roth, P. A. (2026). Avoiding metaphysical pratfalls and positivist pitfalls: A user’s guide to analytical philosophy of history. Synthese, 207, 210. https://doi.org/10.1007/s11229-026-05603-z
Virmajoki, V. (2025). Defeating the ideal chronicler: The problem with the thought experiment. Journal of the Philosophy of History, Advance Online Publication. https://doi.org/10.1163/18722636-12341562