pith. sign in

arxiv: 2606.23055 · v1 · pith:BS5J7VM3new · submitted 2026-06-22 · 💻 cs.AI · cs.LO

Some Results about the Expressivity of Preference-Incomplete Structured Argumentation Frameworks

Pith reviewed 2026-06-26 08:47 UTC · model grok-4.3

classification 💻 cs.AI cs.LO
keywords argumentation frameworksASPIC+expressivityuncertain preferencesuncertain defeatsstructured argumentationabstract argumentation
0
0 comments X

The pith

ASPIC+ frameworks with uncertain preferences cannot match the expressivity of abstract argumentation systems with uncertain defeats.

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

The paper compares the power of ASPIC+ structured argumentation systems, where only partial information about argument preferences is available, against abstract systems that directly encode uncertainty about which arguments defeat others. Most of the comparisons yield negative results, showing that many defeat patterns possible in the abstract setting cannot be reproduced in the structured one. The authors conjecture that uncertain preferences nonetheless reach some non-trivial level of expressivity and prove initial steps toward identifying that level. A sympathetic reader would care because these limits affect which forms of uncertain reasoning can be captured inside structured argument-based AI systems.

Core claim

ASPIC+ argumentation frameworks with uncertain preference profiles are less expressive than several abstract formalisms with uncertain defeats, as established by mostly negative results. A positive non-trivial threshold for the expressivity of uncertain preferences is conjectured, and some essential preliminary steps toward confirming this conjecture are proved.

What carries the argument

Expressivity comparisons between ASPIC+ frameworks with uncertain preference profiles and abstract formalisms with uncertain defeats, performed via embeddings or simulations.

If this is right

  • Many patterns of uncertain defeat cannot be realized inside ASPIC+ even when preferences are allowed to be incomplete.
  • The choice between structured and abstract representations directly restricts which reasoning outcomes remain possible under uncertainty.
  • Any complete account of uncertain argumentation must treat the expressivity gap between preference uncertainty and defeat uncertainty as a design constraint.
  • The conjectured threshold, once located, would mark the precise point at which uncertain preferences begin to match abstract uncertain defeats.

Where Pith is reading between the lines

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

  • If the threshold exists, hybrid systems could be built that add just enough structure to preferences to reach it without moving fully to abstract defeats.
  • Small finite examples of argumentation graphs could be enumerated to locate the exact boundary conjectured in the paper.
  • The negative results may extend to other structured systems that encode uncertainty only through preferences rather than through direct defeat relations.
  • Practical implementations of argumentation in AI may need to fall back to abstract formalisms when full coverage of uncertain defeats is required.

Load-bearing premise

Standard embedding or simulation relations between structured and abstract argumentation frameworks correctly measure differences in what each can represent.

What would settle it

An explicit embedding that lets some ASPIC+ system with uncertain preferences fully reproduce one of the abstract formalisms previously shown to be strictly more expressive.

read the original abstract

This paper studies the expressive power of ASPIC$^+$ argumentation frameworks with uncertain preference profiles by comparing them with several abstract formalisms with uncertain defeats. Most of our results are negative (and some of them are theoretically unexpected). We also conjecture a positive, non-trivial threshold for the expressivity of uncertain preferences, and prove some essential preliminary steps toward the confirmation of this conjecture.

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 / 2 minor

Summary. The paper examines the expressive power of ASPIC+ structured argumentation frameworks that incorporate uncertain (incomplete) preference profiles. It compares these to several abstract argumentation formalisms allowing uncertain defeats, reporting mostly negative results on the ability of uncertain preferences to simulate or embed uncertain defeats. A positive non-trivial threshold for expressivity is conjectured, with some preliminary steps proved.

Significance. If the negative results are correct, they establish concrete limitations on the use of preference uncertainty within ASPIC+ to capture defeat uncertainty, which is relevant for understanding the boundaries between structured and abstract argumentation models. The conjecture, if substantiated, would delineate a precise expressivity threshold and could inform the design of hybrid argumentation systems.

minor comments (2)
  1. The abstract refers to 'embeddings or simulations' and 'reduction relations' for the comparisons but does not define these technical notions; a dedicated preliminary section spelling out the exact formal definitions of expressivity would improve clarity.
  2. The abstract states that proofs exist for the negative results and preliminary steps, yet the provided text contains no derivations; including at least the key proof sketches or references to lemmas in an appendix would strengthen verifiability.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for their summary of our paper, which correctly identifies the focus on negative expressivity results for ASPIC+ with uncertain preferences relative to abstract uncertain-defeat models, along with the conjecture and preliminary proofs. No specific major comments were provided in the report, so we have no point-by-point responses at this time. We remain available to address any questions or provide additional details during the review process.

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The paper's central claims consist of negative expressivity results via embeddings between ASPIC+ frameworks with uncertain preferences and abstract formalisms with uncertain defeats, plus a conjecture on a positive threshold with preliminary steps. No equations, definitions, or self-citations are visible in the abstract or context that reduce any result to its own inputs by construction, rename known patterns, or rely on load-bearing self-citations. The comparisons use standard notions of expressivity, making the derivation self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

No information on free parameters, axioms, or invented entities is available from the abstract.

pith-pipeline@v0.9.1-grok · 5573 in / 1001 out tokens · 25919 ms · 2026-06-26T08:47:47.421942+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

189 extracted references · 22 canonical work pages

  1. [1]

    Theory and practice of logic programming , volume=

    Defeasible logic programming: an argumentativeapproach , author=. Theory and practice of logic programming , volume=. 2004 , publisher=

  2. [2]

    CEUR WORKSHOP PROCEEDINGS , volume=

    On Polymorphic Attacks in the ASPIC+ and ASPICR Formalisms , author=. CEUR WORKSHOP PROCEEDINGS , volume=. 2025 , organization=

  3. [3]

    arXiv preprint arXiv:2510.18631 , year=

    Comparative Expressivity for Structured Argumentation Frameworks with Uncertain Rules and Premises , author=. arXiv preprint arXiv:2510.18631 , year=

  4. [4]

    Journal of Logic and Computation , volume=

    Qualitative uncertainty and dynamics of argumentation through dynamic logic , author=. Journal of Logic and Computation , volume=. 2023 , publisher=

  5. [5]

    Proceedings of the Fifteenth International Conference on Principles of Knowledge Representation and Reasoning , pages =

    ABA+: assumption-based argumentation with preferences , year =. Proceedings of the Fifteenth International Conference on Principles of Knowledge Representation and Reasoning , pages =

  6. [6]

    On the Instantiation of Argument-Incomplete Argumentation Frameworks , booktitle =

    Antonio Yuste. On the Instantiation of Argument-Incomplete Argumentation Frameworks , booktitle =

  7. [7]

    and Järvisalo, Matti , booktitle =

    Odekerken, Daphne and Lehtonen, Tuomo and Borg, AnneMarie and Wallner, Johannes P. and Järvisalo, Matti , booktitle =. 2023 , month =. doi:10.24963/kr.2023/52 , url =

  8. [8]

    Journal of Artificial Intelligence Research , volume=

    Argumentative reasoning in ASPIC+ under incomplete information , author=. Journal of Artificial Intelligence Research , volume=

  9. [9]

    Proceedings of the International Conference on Principles of Knowledge Representation and Reasoning , volume=

    Relating Abstract and Structured Accounts of Argumentation Dynamics: the Case of Expansions , author=. Proceedings of the International Conference on Principles of Knowledge Representation and Reasoning , volume=

  10. [10]

    CoRR , volume =

    Jean-Guy Mailly , title =. CoRR , volume =. 2020 , url =

  11. [11]

    International Journal of Approximate Reasoning , volume=

    A characterization of types of support between structured arguments and their relationship with support in abstract argumentation , author=. International Journal of Approximate Reasoning , volume=. 2018 , publisher=

  12. [12]

    Computational Models of Argument , pages=

    Abstraction in Argumentation: Necessary but Dangerous , author=. Computational Models of Argument , pages=. 2018 , publisher=

  13. [13]

    PloS one , volume=

    Reasoning on conflicting information: An empirical study of Formal Argumentation , author=. PloS one , volume=. 2022 , publisher=

  14. [14]

    The Thirty-Sixth

    Gianvincenzo Alfano and Sergio Greco and Francesco Parisi and Irina Trubitsyna , title =. The Thirty-Sixth. 2022 , doi=

  15. [15]

    Phan Minh Dung and Phan Minh Thang , title =

  16. [16]

    IfCoLog Journal of Logics and their Applications , volume=

    Foundations of implementations for formal argumentation , author=. IfCoLog Journal of Logics and their Applications , volume=. 2017 , publisher=

  17. [17]

    Cooper and Andreas Herzig and Faustine Maffre and Fr

    Martin C. Cooper and Andreas Herzig and Faustine Maffre and Fr. A lightweight epistemic logic and its application to planning , journal =. 2021 , url =. doi:10.1016/j.artint.2020.103437 , timestamp =

  18. [18]

    Basic beliefs and argument-based beliefs in awareness epistemic logic with structured arguments , editor=

    Burrieza, Alfredo and Yuste-Ginel, Antonio , booktitle=. Basic beliefs and argument-based beliefs in awareness epistemic logic with structured arguments , editor=. 2020 , organization=. doi:10.3233/FAIA200498 , series=

  19. [19]

    An Awareness Epistemic Framework for Belief, Argumentation and Their Dynamics , booktitle =

    Alfredo Burrieza and Antonio Yuste. An Awareness Epistemic Framework for Belief, Argumentation and Their Dynamics , booktitle =. 2021 , doi =

  20. [20]

    Erkenntnis , pages=

    Logic of Justified Beliefs Based on Argumentation , author=. Erkenntnis , pages=. 2021 , publisher=

  21. [21]

    Probabilistic argumentation: A survey , booktitle=

    Hunter, Anthony and Polberg, Sylwia and Potyka, Nico and Rienstra, Tjitze and Thimm, Matthias , volume=. Probabilistic argumentation: A survey , booktitle=. 2021 , publisher=

  22. [22]

    Recursion in Abstract Argumentation is Hard - On the Complexity of Semantics Based on Weak Admissibility , booktitle =

    Wolfgang Dvor. Recursion in Abstract Argumentation is Hard - On the Complexity of Semantics Based on Weak Admissibility , booktitle =

  23. [23]

    The Thirty-Fourth

    Ringo Baumann and Gerhard Brewka and Markus Ulbricht , title =. The Thirty-Fourth. 2020 , doi=

  24. [24]

    Synthese 169(2), pp

    Ditmarsch, Hans P. van and Andreas Herzig and J. Introspective forgetting , journal =. 2009 , url =. doi:10.1007/s11229-009-9554-4 , timestamp =

  25. [25]

    Gabbay and Odinaldo Rodrigues , title =

    Ringo Baumann and Dov M. Gabbay and Odinaldo Rodrigues , title =. The Thirty-Fourth. 2020 , doi=

  26. [26]

    Models of Persuasion Dialogue,

    Sanjay Modgil and Martin Caminada , editor =. Proof Theories and Algorithms for Abstract Argumentation Frameworks , booktitle =. 2009 , url =. doi:10.1007/978-0-387-98197-0\_6 , timestamp =

  27. [27]

    Journal of Applied Logics , volume=

    Enforcement in formal argumentation , author=. Journal of Applied Logics , volume=

  28. [28]

    Extension-Based Semantics for Incomplete Argumentation Frameworks

    Mailly, Jean-Guy. Extension-Based Semantics for Incomplete Argumentation Frameworks. Logic and Argumentation. 2021

  29. [29]

    Algorithmic Decision Theory: 4th International Conference, ADT 2015, Lexington, KY, USA, September 27--30, 2015, Proceedings 4 , pages=

    Verification in attack-incomplete argumentation frameworks , author=. Algorithmic Decision Theory: 4th International Conference, ADT 2015, Lexington, KY, USA, September 27--30, 2015, Proceedings 4 , pages=. 2015 , organization=

  30. [30]

    Algorithmic Decision Theory , pages=

    Verification in argument-incomplete argumentation frameworks , author=. Algorithmic Decision Theory , pages=. 2015 , organization=

  31. [31]

    2011 , url =

    Stefano Bistarelli and Francesco Santini , title =. 2011 , url =. doi:10.1109/ICTAI.2011.96 , timestamp =

  32. [32]

    Journal of Logic and Computation , volume=

    Semi-stable semantics , author=. Journal of Logic and Computation , volume=. 2012 , DOI=

  33. [33]

    doi:10.48550/arXiv.1908.08406 , year=

    Cramer, Marcos and van der Torre, Leon , booktitle=. doi:10.48550/arXiv.1908.08406 , year=

  34. [34]

    Abstract Argumentation with Qualitative Uncertainty: An Analysis in Dynamic Logic

    Herzig, Andreas and Yuste-Ginel, Antonio. Abstract Argumentation with Qualitative Uncertainty: An Analysis in Dynamic Logic. Logic and Argumentation. 2021

  35. [35]

    18th International Conference on Autonomous Agents and MultiAgent Systems

    Argumentation-based Negotiation with Incomplete Opponent Profiles , author=. 18th International Conference on Autonomous Agents and MultiAgent Systems. 2019 , pages=

  36. [36]

    Artificial Intelligence , volume=

    An abstract and structured account of dialectical argument strength , author=. Artificial Intelligence , volume=. 2024 , publisher=

  37. [37]

    Yes, no, maybe,

    Mailly, Jean-Guy , journal=. Yes, no, maybe,. 2022 , volume=

  38. [38]

    Reasoning over Argument-Incomplete

    Fazzinga, Bettina and Flesca, Sergio and Furfaro, Filippo , booktitle =. Reasoning over Argument-Incomplete. 2021 , doi =

  39. [39]

    2021 , month =

    Fazzinga, Bettina and Flesca, Sergio and Furfaro, Filippo , booktitle =. 2021 , month =. doi:10.24963/kr.2021/29 , url =

  40. [40]

    Constrained Incomplete Argumentation Frameworks

    Mailly, Jean-Guy. Constrained Incomplete Argumentation Frameworks. Symbolic and Quantitative Approaches to Reasoning with Uncertainty. 2021

  41. [41]

    AI Communications , volume=

    Constrained incomplete argumentation frameworks: Expressiveness, complexity and enforcement , author=. AI Communications , volume=. 2024 , publisher=

  42. [42]

    Proceedings of the Tenth International Conference on Principles of Knowledge Representation and Reasoning , pages =

    Coste-Marquis, Sylvie and Devred, Caroline and Marquis, Pierre , title =. Proceedings of the Tenth International Conference on Principles of Knowledge Representation and Reasoning , pages =. 2006 , isbn =

  43. [43]

    2021 , month =

    Herzig, Andreas and Yuste-Ginel, Antonio , booktitle =. 2021 , month =. doi:10.24963/kr.2021/69 , editor=

  44. [44]

    Argument & Computation , volume=

    Introduction to structured argumentation , author=. Argument & Computation , volume=. doi:10.1080/19462166.2013.869764 , year=

  45. [45]

    Abstract argumentation frameworks and their semantics , booktitle=

    Baroni, Pietro and Caminada, Martin and Giacomin, Massimiliano , pages=. Abstract argumentation frameworks and their semantics , booktitle=. 2018 , publisher=

  46. [46]

    2020 , school=

    Computational Approaches to Dynamics and Uncertainty in Abstract Argumentation , author=. 2020 , school=

  47. [47]

    European Conference on Symbolic and Quantitative Approaches to Reasoning and Uncertainty , pages=

    Encompassing attacks to attacks in abstract argumentation frameworks , author=. European Conference on Symbolic and Quantitative Approaches to Reasoning and Uncertainty , pages=. 2009 , organization=

  48. [48]

    arXiv preprint arXiv:2009.04869 , year=

    A Note on Rich Incomplete Argumentation Frameworks , author=. arXiv preprint arXiv:2009.04869 , year=

  49. [49]

    153--185

    Doutre, Sylvie and Herzig, Andreas and Perrussel, Laurent. Abstract Argumentation in Dynamic Logic: Representation, Reasoning and Change. Dynamics, Uncertainty and Reasoning. 2019. doi:10.1007/978-981-13-7791-4\_8 , pages="153--185", abstract="We provide a logical analysis of Dung's abstract argumentation frameworks and their dynamics. We express attack r...

  50. [50]

    Fourteenth International Conference on the Principles of Knowledge Representation and Reasoning , editor=

    A dynamic logic framework for abstract argumentation , author=. Fourteenth International Conference on the Principles of Knowledge Representation and Reasoning , editor=. 2014 , publisher=

  51. [51]

    Dynamic logic of propositional assignments: a well-behaved variant of

    Balbiani, Philippe and Herzig, Andreas and Troquard, Nicolas , booktitle=. Dynamic logic of propositional assignments: a well-behaved variant of. 2013 , organization=

  52. [52]

    Journal of Applied Logics , volume=

    Rationality postulates: applying argumentation theory for non-monotonic reasoning , author=. Journal of Applied Logics , volume=

  53. [53]

    2018 , publisher=

    Handbook of formal argumentation , author=. 2018 , publisher=

  54. [54]

    International Joint Conference in Artificial Intelligence , publisher =

    Multi-Agent Abstract Argumentation Frameworks With Incomplete Knowledge of Attacks , author =. International Joint Conference in Artificial Intelligence , publisher =. 2021 , doi =

  55. [55]

    Abstract Argumentation in Dynamic Logic: Representation, Reasoning and Change , booktitle =

    Sylvie Doutre and Andreas Herzig and Laurent Perrussel , editor =. Abstract Argumentation in Dynamic Logic: Representation, Reasoning and Change , booktitle =. 2018 , url =

  56. [56]

    , author=

    Knowledge and control. , author=. AAMAS , pages=

  57. [57]

    AI Magazine , volume=

    Towards artificial argumentation , author=. AI Magazine , volume=. 2017 , doi =

  58. [58]

    Synthese , volume=

    Dynamic epistemic logics for abstract argumentation , author=. Synthese , volume=. 2021 , publisher=

  59. [59]

    2004 , publisher=

    Reasoning about knowledge , author=. 2004 , publisher=

  60. [60]

    Artificial Intelligence , volume=

    Acceptance in incomplete argumentation frameworks , author=. Artificial Intelligence , volume=. 2021 , doi =

  61. [61]

    Artificial Intelligence , volume=

    A lightweight epistemic logic and its application to planning , author=. Artificial Intelligence , volume=. 2021 , publisher=

  62. [62]

    On the Acceptability of Arguments in Bipolar Argumentation Frameworks , booktitle =

    Claudette Cayrol and Marie. On the Acceptability of Arguments in Bipolar Argumentation Frameworks , booktitle =. 2005 , DOI =

  63. [63]

    International Joint Conference in Artificial Intelligence , publisher =

    Revisiting the Notion of Extension over Incomplete Abstract Argumentation Frameworks , author =. International Joint Conference in Artificial Intelligence , publisher =. 2020 , month =. doi:10.24963/ijcai.2020/237 , editor=

  64. [64]

    Argument & Computation , volume=

    Logical theories and abstract argumentation: A survey of existing works , author=. Argument & Computation , volume=. 2020 , doi =

  65. [65]

    2011 , publisher=

    A new approach for preference-based argumentation frameworks , author=. 2011 , publisher=. doi:10.1007/s10472-011-9271-9 , journal=

  66. [66]

    International Joint Conference in Artificial Intelligence , year=

    Opponent models with uncertainty for strategic argumentation , author=. International Joint Conference in Artificial Intelligence , year=

  67. [67]

    Proceedings of

    Mechanism design for abstract argumentation , author=. Proceedings of. 2008 , organization=

  68. [68]

    and Dvo r \'a k, Wolfgang and Falappa, Marcelo A

    Modgil, Sanjay and Toni, Francesca and Bex, Floris and Bratko, Ivan and Ches \ n evar, Carlos I. and Dvo r \'a k, Wolfgang and Falappa, Marcelo A. and Fan, Xiuyi and Gaggl, Sarah Alice and Garc \'i a, Alejandro J. and Gonz \'a lez, Mar \'i a P. and Gordon, Thomas F. and Leite, Jo \ a o and Mo z ina, Martin and Reed, Chris and Simari, Guillermo R. and Szei...

  69. [69]

    Artificial Intelligence , volume=

    Argumentation in artificial intelligence , author=. Artificial Intelligence , volume=. 2007 , publisher=

  70. [70]

    Artificial Intelligence Review , volume=

    A systematic review of argumentation techniques for multi-agent systems research , author=. Artificial Intelligence Review , volume=. 2015 , publisher=

  71. [71]

    Proceedings of ArgMAS , pages=

    Argumentation in multi-agent systems: Context and recent developments , author=. Proceedings of ArgMAS , pages=. 2006 , organization=

  72. [72]

    Complexity of Verification in Incomplete Argumentation Frameworks , booktitle =

    Dorothea Baumeister and Daniel Neugebauer and J. Complexity of Verification in Incomplete Argumentation Frameworks , booktitle =. 2018 , timestamp =

  73. [73]

    International Conference on Industrial, Engineering and Other Applications of Applied Intelligent Systems

    A dynamic logic framework for abstract argumentation: adding and removing arguments , author=. International Conference on Industrial, Engineering and Other Applications of Applied Intelligent Systems. , series=. 2017 , doi =

  74. [74]

    On the merging of

    Coste-Marquis, Sylvie and Devred, Caroline and Konieczny, S. On the merging of. Artificial Intelligence , volume=. 2007 , publisher=

  75. [75]

    Control Argumentation Frameworks , booktitle =

    Yannis Dimopoulos and Jean. Control Argumentation Frameworks , booktitle =. 2018 , url =

  76. [76]

    Proceedings of COMMA 2020 , volume=

    Possible Controllability of Control Argumentation Frameworks , author=. Proceedings of COMMA 2020 , volume=. 2020 , publisher=

  77. [77]

    International Joint Conference in Artificial Intelligence , year=

    Controllability of control argumentation frameworks , author=. International Joint Conference in Artificial Intelligence , year=. doi:10.24963/ijcai.2020/257 , organization=

  78. [78]

    Argumentation-based negotiation with incomplete opponent profiles , author=. 13

  79. [79]

    Artificial Intelligence , volume=

    Verification in incomplete argumentation frameworks , author=. Artificial Intelligence , volume=. doi:10.1016/j.artint.2018.08.001 , year=

  80. [80]

    Proceedings of TAFA , pages=

    Probabilistic argumentation frameworks , author=. Proceedings of TAFA , pages=. 2011 , series=. doi:10.1007/978-3-642-29184-5\_1 , organization=

Showing first 80 references.