Call for Papers

Normativity under Uncertainty

Normative systems are inescapably confronted with uncertainty. Agents must frequently deliberate about what ought to be done on the basis of information that is incomplete, imprecise, or conflicting, where the uncertainty may concern the facts of the case, the norms in force, the consequences of action, or the reliability of one's sources. Classical deontic logics, by contrast, typically assume consistent sets of norms and abstract away from the epistemic situation of agents.

The general theme of this edition of DEON, Normativity under Uncertainty, addresses the logical, philosophical, and computational foundations of normative reasoning once these idealizations are lifted. The aim is to investigate how central normative concepts, among them obligation, permission, responsibility, and blame, can be rigorously formalized, interpreted, and applied under diverse forms of uncertainty, including defeasible and non-monotonic inference as well as probabilistic and epistemic reasoning.

We invite submission of original and unpublished work. Parallel submission to journals or to other conferences or workshops with published proceedings is not permitted.

Papers should be written in English.

The reviewing process for DEON 2027 is double-blind. Submissions must be anonymized (see below).

Submitted papers should not be longer than 15 pages, excluding references and technical appendix. Reviewers are not required to read the technical appendix and may choose to do so at their discretion.

LaTeX and Word templates for preparing submissions: To be announced

Papers must be submitted in PDF format via EasyChair: To be announced

Upon acceptance, final papers must be submitted as a LaTeX file together with the corresponding PDF for the camera-ready version. Converting a Word file into the required LaTeX template is the sole responsibility of the authors of the accepted paper.

For authors. Papers containing text generated by large language models (LLMs) are not permitted. An exception is made where such text appears as the object of an experimental analysis, provided that the source model is explicitly cited. Authors may use LLMs to edit and polish text that they themselves have written. Generative AI models do not meet the criteria for authorship of scientific papers and may not be listed as authors, nor may their outputs be used as a citable source in a DEON submission.

Submissions that, in the judgement of the programme chairs, consist substantially of LLM-generated content will be desk-rejected. Where such a concern is raised, the submission will be examined independently by at least two programme chairs; a desk rejection requires their agreement. Authors will be informed of the decision and may respond to the chairs.

Any attempt at prompt injection — that is, inserting text into a paper with the intention of manipulating an LLM, for instance in order to obtain a favourable review — is strictly forbidden and will result in desk rejection.

For reviewers. Reviewers must not upload submissions, in whole or in part, to any LLM service that does not guarantee confidentiality; submissions are confidential material. Reviews must not be written by an LLM. Reviewers remain fully responsible for the entire content of their reviews, including any inaccuracies introduced by tools they may have used.

Responsibility. In submitting to DEON, authors are fully responsible for their submitted content, including avoiding plagiarism and complying with the policy above. Any allegation or violation of this policy will be investigated and, if substantiated, possible outcomes include rejection or retraction of the work.