pith. sign in

arxiv: 2605.26739 · v1 · pith:JSGZAWVWnew · submitted 2026-05-26 · 💻 cs.LO

From Actions to Obligations: A Deontic Action Model Logic

Giorgio Cignarale This is my paper

Pith reviewed 2026-07-01 16:16 UTC · model grok-4.3

classification 💻 cs.LO
keywords deontic logicaction model logicobligationsmulti-agent systemsmodal logicaxiomatizationsoundness and completenessdynamic logic
0
0 comments X

The pith

DAML derives obligations as actions maximizing expected deontic value in multi-agent action models.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces Deontic Action Model Logic to reason about obligations over actions in multi-agent systems with incomplete information. It extends epistemic action model logic by adding evaluation of actions according to the desirability and likelihood of their outcomes. Obligations attach to those actions that achieve the highest expected deontic value among the alternatives present at a given decision point. The resulting system yields a formal treatment of conditional and context-sensitive obligations while supplying an axiomatization together with soundness and completeness results.

Core claim

DAML enriches epistemic action model logic with deontic evaluation mechanisms that assess actions in terms of both desirability and likelihood of outcomes. An action is obligatory precisely when it maximizes the expected deontic value among an agent's available alternatives at a decision point. The logic supports reasoning about conditional obligations that arise in strategic interactions under incomplete information, is equipped with a sound and complete axiomatization relative to its semantics, and is illustrated on the Miners' Puzzle and related multi-agent deontic scenarios.

What carries the argument

The deontic evaluation mechanism that identifies obligatory actions by maximizing the product of desirability and likelihood within the action model update framework.

If this is right

  • Conditional obligations arise when the maximization is performed relative to the information available at the decision point.
  • The same mechanism applies to strategic interactions in which agents must reason about the possible actions of others.
  • Soundness and completeness guarantee that syntactic derivations match the semantic definition of obligation.
  • The framework directly supports formal verification of action selection in norm-governed multi-agent systems.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The combination of epistemic updates and deontic maximization offers a unified setting for studying how new information alters obligations.
  • The machinery could be used to implement obligation-driven policies inside automated agents that operate under uncertainty.
  • Restricting the models to single-agent or perfect-information cases would recover standard deontic reasoning as a special instance.

Load-bearing premise

Obligations are captured exactly by selecting the action that maximizes expected deontic value among alternatives at each decision point, and this selection rule extends naturally to conditional and multi-agent cases.

What would settle it

A concrete model of a multi-agent scenario in which the action that maximizes expected deontic value fails to match the obligations required by the semantics, or a derivation in the axiomatization that contradicts a semantically valid obligation.

Figures

Figures reproduced from arXiv: 2605.26739 by Giorgio Cignarale.

Figure 1
Figure 1. Figure 1: Initial pointed graded Kripke model M, v of the miner’s puzzle (left). The evaluation point v is underlined. Reflexive loops and agent-labels on relations are omitted. Numbers close to the worlds represents their desirability value. Action model U representing i’s possible actions (right). Updated model M ⊗ U (below) made of three action-generated submodels, one for each action, Mαi , Mβi , Mγi . Reflexive… view at source ↗
Figure 2
Figure 2. Figure 2: Initial graded Kripke model M (left). Transitive arrows are omitted. The evaluation point v is underlined. Gray rectangles represent agents’ submodels from v, the top being Mv b and the bottom is Mv a. Action model U (right top), representing b’s possible actions δb and γb. Action model U ′ (right bottom), representing a’s possible actions αa and βa. 6 Conclusion and future work This paper introduced Deont… view at source ↗
Figure 3
Figure 3. Figure 3: Updated model M⊗U, with corresponding action-generated submodels Mv,δb (left) and Mv,γb (right), representing the epistemic outcomes of b’s available actions. Transitive arrows are omitted. The evaluation points (v, γ) and (v, δ) are underlined. In each action-generated submodel, gray rectangles represent the agent-based submodels from v, δ, and respectively v, γ, the top one for agent b and the bottom for… view at source ↗
Figure 4
Figure 4. Figure 4: Updated model M ⊗ U ⊗ U ′ , representing all the epistemic outcomes of the actions of b followed by the actions of a, with four corresponding action-generated submodels, from left to right: Mw,(δb;αa) ,Mv,(δb;βa) ,Mw,(γb;αa) and Mv,(γb;β) . Transitive arrows are omitted. In each action-generated submodel, gray rectangles represent the agent-based submodels the top one for b and the bottom for a [PITH_FULL… view at source ↗
read the original abstract

We introduce the Deontic Action Model Logic (DAML), a dynamic modal framework for reasoning about obligations over actions in multi-agent systems. DAML extends the epistemic Action Model Logic by incorporating deontic evaluation mechanisms that assess agents' actions in terms of both the desirability and the likelihood of their outcomes. Obligations arise for those actions that maximize expected deontic value among an agent's available alternatives at a given decision point, yielding a formal account for reasoning about conditional and context-sensitive obligations in settings involving strategic interaction and incomplete information. DAML supports principled action selection in norm-governed multi-agent systems, and is the first such framework to derive these obligations using the action model logic machinery. We provide an axiomatization of the logic and prove soundness and completeness with respect to its semantics. Finally, we demonstrate the expressive power of our framework through applications to the Miners' Puzzle and other multi-agent deontic scenarios.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

0 major / 3 minor

Summary. The paper introduces Deontic Action Model Logic (DAML) as an extension of epistemic action model logic. Obligations are defined as those actions maximizing expected deontic value (desirability times likelihood) among an agent's alternatives at a decision point. The framework supports conditional and context-sensitive obligations in multi-agent settings with incomplete information and strategic interaction. It supplies an axiomatization, proves soundness and completeness with respect to the semantics, and illustrates the approach via the Miners' Puzzle and other scenarios. The work claims to be the first to derive such obligations using action model logic machinery.

Significance. If the semantics and proofs hold, DAML provides a technically coherent dynamic framework for deontic reasoning that layers deontic evaluation directly onto the standard product-update mechanism of action model logic. The direct interpretation of the obligation operator from maximization at decision points, together with the completeness result, supplies a rigorous foundation. This could support formal analysis of norm-governed multi-agent systems.

minor comments (3)
  1. [§1] §1 (Introduction): the novelty claim that DAML is 'the first such framework' would be strengthened by a short explicit comparison to prior deontic extensions of dynamic epistemic logic, even if only to note differences in how obligations are derived from action models.
  2. [Semantics section] The semantics section: the precise definition of 'expected deontic value' (product of desirability and likelihood) should include an explicit equation or clause number so that readers can trace how it interacts with the product update.
  3. [Axiomatization section] The axiomatization: while soundness and completeness are claimed, a brief remark on whether the obligation operator is treated as a primitive or derived would clarify the proof strategy for readers.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for the careful summary of our paper and the positive assessment of DAML as a coherent dynamic framework for deontic reasoning. The recommendation of minor revision is noted. No specific major comments were raised in the report.

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The paper introduces DAML as an extension of action model logic, defining obligations via maximization of expected deontic value (desirability times likelihood) at decision points, then supplies an independent axiomatization with soundness and completeness proofs relative to the new semantics. No load-bearing step reduces by construction to a fitted parameter, self-citation chain, or renamed input; the deontic layer is added on top of standard product-update semantics, and the completeness result is shown directly for the resulting models. The derivation is self-contained with no self-definitional or fitted-input patterns.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 1 invented entities

Review is abstract-only; the ledger is therefore limited to elements explicitly named in the abstract. No numerical free parameters are mentioned. The framework rests on standard modal logic background plus the novel deontic evaluation rule.

axioms (2)
  • standard math Standard semantics and axioms of epistemic Action Model Logic
    The paper states it extends this existing framework.
  • domain assumption Obligations arise precisely from actions that maximize expected deontic value (desirability and likelihood of outcomes)
    This is the core definitional step for deriving obligations.
invented entities (1)
  • Deontic Action Model Logic (DAML) no independent evidence
    purpose: Dynamic modal framework combining epistemic action models with deontic evaluation to derive obligations
    New logic introduced by the paper; no independent evidence supplied in abstract.

pith-pipeline@v0.9.1-grok · 5681 in / 1463 out tokens · 38629 ms · 2026-07-01T16:16:48.214747+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

21 extracted references · 8 canonical work pages

  1. [1]

    Moss, and Slawomir Solecki

    Alexandru Baltag, Lawrence S. Moss, and Slawomir Solecki. The logic of public announce- ments, common knowledge, and private suspicions. InProceedings of the 7th Conference on Theoretical Aspects of Rationality and Knowledge (TARK ’98), Evanston, Illinois, 1998. Morgan Kaufmann

    1998

  2. [2]

    Cambridge University Press, 2001

    Patrick Blackburn, Maarten de Rijke, and Yde Venema.Modal Logic, volume 53 of Cambridge Tracts in Theoretical Computer Science. Cambridge University Press, 2001. doi:10.1017/CBO9781107050884

    work page doi:10.1017/cbo9781107050884 2001

  3. [3]

    A priori belief updates as a method for agent self- recovery.Review of Analytic Philosophy, 4(1):1, 2024.doi:10.18494/sam.rap

    Giorgio Cignarale and Roman Kuznets. A priori belief updates as a method for agent self- recovery.Review of Analytic Philosophy, 4(1):1, 2024.doi:10.18494/sam.rap. 2024.0021

    work page doi:10.18494/sam.rap 2024

  4. [4]

    Gabbay, John F

    Dov M. Gabbay, John F. Horty, Xavier Parent, Ron van der Meyden, and Leendert van der Torre. Handbook of deontic logic and normative systems, volume 2. College Publications, 2021

    2021

  5. [5]

    Arbitrary action model logic and action model synthesis

    James Hales. Arbitrary action model logic and action model synthesis. In2013 28th Annual ACM/IEEE Symposium on Logic in Computer Science, pages 253–262, 2013.doi:10. 1109/LICS.2013.31

    2013

  6. [6]

    Ithaca: Cornell University Press, 1962

    Jaakko Hintikka.Knowledge and Belief: An Introduction to the Logic of the Two Notions. Ithaca: Cornell University Press, 1962

    1962

  7. [7]

    Horty.Agency and Deontic Logic

    John F. Horty.Agency and Deontic Logic. Oxford University Press, New York, 2001

    2001

  8. [8]

    Ifs and oughts.Journal of Philosophy, 107(3):115–143, 2010.doi:10.5840/jphil2010107310

    Niko Kolodny and John MacFarlane. Ifs and oughts.Journal of Philosophy, 107(3):115–143, 2010.doi:10.5840/jphil2010107310

    work page doi:10.5840/jphil2010107310 2010

  9. [9]

    Conflicting obligations in multi-agent deontic logic

    Barteld Kooi and Allard Tamminga. Conflicting obligations in multi-agent deontic logic. In Lou Goble and John-Jules Ch. Meyer, editors,Deontic Logic and Artificial Normative Systems, pages 175–186, Berlin, Heidelberg, 2006. Springer Berlin Heidelberg

    2006

  10. [10]

    Linguistic and philosophical considerations on bayesian.Deontic modality, pages 82–116, 2016

    Daniel Lassiter. Linguistic and philosophical considerations on bayesian.Deontic modality, pages 82–116, 2016

    2016

  11. [11]

    A Framework for Behavioural Cloning , booktitle =

    Daniel Lassiter.Graded Modality: Qualitative and quantitative perspectives. Oxford Uni- versity Press, United States, 2017.doi:10.1093/oso/9780198701347.001.0001

    work page doi:10.1093/oso/9780198701347.001.0001 2017

  12. [12]

    Deontic interpreted systems.Studia Logica, 75(1):63– 92, 2003

    Alessio Lomuscio and Marek Sergot. Deontic interpreted systems.Studia Logica, 75(1):63– 92, 2003

    2003

  13. [13]

    Deontic Logic

    Paul McNamara and Frederik Van De Putte. Deontic Logic. In Edward N. Zalta and Uri Nodelman, editors,The Stanford Encyclopedia of Philosophy. Metaphysics Research Lab, Stanford University, Fall 2022 edition, 2022

    2022

  14. [14]

    Ch Meyer

    J.-J. Ch Meyer. A different approach to deontic logic: Deontic logic viewed as a variant of dynamic logic.Notre Dame Journal of Formal Logic, 29(1):109–136, 1987

    1987

  15. [15]

    The logic of knowledge based obligation.Syn- these, 149(2):311–341, 2006.doi:10.1007/s11229-005-3877-6

    Eric Pacuit, Rohit Parikh, and Eva Cogan. The logic of knowledge based obligation.Syn- these, 149(2):311–341, 2006.doi:10.1007/s11229-005-3877-6

    work page doi:10.1007/s11229-005-3877-6 2006

  16. [16]

    Logics of public communications.Synthese, 158(2):165–179, September 2007

    Jan Plaza. Logics of public communications.Synthese, 158(2):165–179, September 2007. Originally published in 1989.doi:10.1007/s11229-007-9168-7

    work page doi:10.1007/s11229-007-9168-7 2007

  17. [17]

    Dδl: A dynamic deontic logic.Synthese, 185(S1):1–17, 2012.doi: 10.1007/s11229-011-9953-1

    Krister Segerberg. Dδl: A dynamic deontic logic.Synthese, 185(S1):1–17, 2012.doi: 10.1007/s11229-011-9953-1

    work page doi:10.1007/s11229-011-9953-1 2012

  18. [18]

    The Logic of Action

    Krister Segerberg, John-Jules Meyer, and Marcus Kracht. The Logic of Action. In Edward N. Zalta and Uri Nodelman, editors,The Stanford Encyclopedia of Philosophy. Metaphysics Research Lab, Stanford University, Summer 2024 edition, 2024

    2024

  19. [19]

    Propositional Dynamic Logic

    Nicolas Troquard and Philippe Balbiani. Propositional Dynamic Logic. In Edward N. Zalta and Uri Nodelman, editors,The Stanford Encyclopedia of Philosophy. Metaphysics Research Lab, Stanford University, Fall 2023 edition, 2023. 20 G. Cignarale

    2023

  20. [20]

    van Ditmarsch, J.Y

    H. van Ditmarsch, J.Y . Halpern, W. van der Hoek, and B.P. Kooi.Handbook of Epistemic Logic. College Publications, 2015

    2015

  21. [21]

    Springer, 2007.doi:10.1007/978-1-4020-5839-4

    Hans van Ditmarsch, Wiebe van der Hoek, and Barteld Kooi.Dynamic Epistemic Logic, vol- ume 337 ofSynthese Library. Springer, 2007.doi:10.1007/978-1-4020-5839-4

    work page doi:10.1007/978-1-4020-5839-4 2007