Extensional Taylor Expansion
Pith reviewed 2026-05-24 08:55 UTC · model grok-4.3
The pith
The equivalence induced by normalizing extensional Taylor expansions of lambda-terms is exactly H*, the greatest consistent sensible lambda-theory.
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
The equivalence induced on lambda-terms by the normalization of extensional Taylor-expansion is nothing but H*, the greatest consistent sensible lambda-theory -- which is also the theory induced by Nakajima trees. Extensional resource terms contain a language of finite approximants of Nakajima trees, and the dynamics of the extensional resource calculus simulate both beta-reduction and eta-reduction of the source lambda-terms.
What carries the argument
The extensional resource calculus of terms in infinitely eta-long form, whose normalization dynamics induce the H* equivalence via extensional Taylor expansion.
If this is right
- Modeling the extensional resource calculus and its dynamics is sufficient to exhibit models of H*.
- Extensional resource terms serve as finite approximants of Nakajima trees.
- The normalization equivalence matches the theory induced by Nakajima trees.
- Normal extensional resource terms stand in bijection with isomorphism classes of augmentations in the universal arena.
Where Pith is reading between the lines
- The untyped bijection with augmentations may support new proofs of properties of H* by importing techniques from game semantics.
- The finite syntax could enable algorithmic approximations or decision procedures for membership in H* via resource-term normalization.
- Similar extensional expansions might be definable for other lambda-theories to obtain resource-based characterizations.
Load-bearing premise
The dynamics of the extensional resource calculus correctly simulate both beta-reduction and eta-reduction of the source lambda terms.
What would settle it
A pair of lambda-terms equivalent under H* whose extensional Taylor expansions normalize to distinct normal forms, or a pair inequivalent under H* whose expansions normalize to the same form.
Figures
read the original abstract
We introduce a calculus of extensional resource terms. These are resource terms \`a la Ehrhard-Regnier, but in infinitely eta-long form. The calculus still retains a finite syntax and dynamics: in particular, we prove strong confluence and normalization. Then we define an extensional version of Taylor expansion, mapping ordinary lambda-terms to (possibly infinite) linear combinations of extensional resource terms: like in the ordinary case, the dynamics of our resource calculus allows us to simulate the beta-reduction of lambda-terms; the extensional nature of this expansion shows in the fact that we are also able to simulate eta-reduction. In a sense, extensional resource terms contain a language of finite approximants of Nakajima trees, much like ordinary resource terms can be seen as a richer version of finite B\"ohm trees. We show that the equivalence induced on lambda-terms by the normalization of extensional Taylor-expansion is nothing but H*, the greatest consistent sensible lambda-theory -- which is also the theory induced by Nakajima trees. This characterization provides a new, simple way to exhibit models of H*: it becomes sufficient to model the extensional resource calculus and its dynamics. The extensional resource calculus moreover allows us to recover, in an untyped setting, a connection between Taylor expansion and game semantics that was previously limited to the typed setting. Indeed, simply typed, eta-long, beta-normal resource terms are known to be in bijective correspondence with plays in the sense of Hyland-Ong game semantics, up to Melli\`es' homotopy equivalence. Extensional resource terms are the appropriate counterpart of eta-long resource terms in an untyped setting: we spell out the bijection between normal extensional resource terms and isomorphism classes of augmentations (a canonical presentation of plays up to homotopy) in the universal arena.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The manuscript introduces an extensional resource calculus extending Ehrhard-Regnier resource terms to infinitely eta-long forms while retaining finite syntax. It proves strong confluence and normalization. An extensional Taylor expansion maps lambda-terms to (possibly infinite) linear combinations of these terms and is shown to simulate both beta- and eta-reduction. The central result is that the equivalence on lambda-terms induced by normalization of the extensional Taylor expansions coincides with H*, the greatest consistent sensible lambda-theory (equivalently, the theory induced by Nakajima trees). Normal extensional resource terms serve as finite approximants of Nakajima trees. The paper also recovers a bijection, in the untyped setting, between normal extensional resource terms and isomorphism classes of augmentations (plays up to homotopy) in the universal arena of Hyland-Ong game semantics.
Significance. If the central claims hold, the work supplies a new, direct characterization of H* via models of the extensional resource calculus and its dynamics, simplifying the construction of models of this theory. It extends the Taylor-expansion/game-semantics connection from the typed to the untyped setting via the explicit bijection with augmentations. The finite syntax for eta-long terms and the explicit simulation of eta-reduction are technical strengths. The manuscript does not mention machine-checked proofs or reproducible code, but the finite-approximant view of Nakajima trees is a clear contribution.
minor comments (2)
- [Abstract / extensional Taylor expansion section] The abstract states the simulation of eta-reduction but the corresponding theorem statement (likely in the section on extensional Taylor expansion) would benefit from an explicit reference to the eta-rule being simulated, to make the claim easier to locate.
- [Preliminaries] Notation for the linear combinations in the extensional Taylor expansion (e.g., the coefficients and the infinite sums) should be introduced with a short example in the preliminaries to aid readability for readers unfamiliar with the ordinary Taylor expansion.
Simulated Author's Rebuttal
We thank the referee for their positive summary of the manuscript, recognition of its contributions, and recommendation to accept.
Circularity Check
No significant circularity
full rationale
The derivation introduces definitions for the extensional resource calculus and its dynamics, proves strong confluence and normalization internally, defines the extensional Taylor expansion mapping, and establishes via stated theorems that normalization induces exactly H* (equivalently Nakajima-tree equality). These steps are self-contained within the paper's own constructions and proofs; no central claim reduces by the paper's equations to a fitted parameter, a self-citation chain, or an ansatz imported from prior author work. External references (Ehrhard-Regnier, Hyland-Ong, Melliès) supply background but are not load-bearing for the equivalence result.
Axiom & Free-Parameter Ledger
axioms (2)
- standard math Standard confluence and normalization properties of beta- and eta-reduction in lambda calculus
- domain assumption Existence and uniqueness properties of Nakajima trees and the theory H*
Reference graph
Works this paper leans on
-
[1]
url: https://doi.org/10.4230/LIPIcs.FSCD.2024.29. [Ehr05] Thomas Ehrhard. “Finiteness spaces”. In: Math. Struct. Comput. Sci. 15.4 (2005), pp. 615–646.doi: 10.1017/S0960129504004645. url: https://doi.org/10.1017/ S0960129504004645. [Eil74] Samuel Eilenberg. Automata, Languages, and Machines. Vol. 59-A. Pure and Applied Mathematics. Academic Press, 1974. [...
-
[2]
Infinite normal forms for the lambda-calculus
Ed. by Bart Jacobs, Alexandra Silva, and Sam Staton. Vol. 308. Electronic Notes in Theoretical Computer Science. Elsevier, 2014, pp. 245–272.doi: 10.1016/J.ENTCS. 2014.10.014. url: https://doi.org/10.1016/j.entcs.2014.10.014. [Nak75] Reiji Nakajima. “Infinite normal forms for the lambda-calculus”. In:Lambda-Calculus and Computer Science Theory, Proceeding...
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.