The Principle of Total Evidence: Justification and Political Significance
Open accessAbstract
The principle of total evidence says that one should conditionalize one’s degrees of belief on one’s total evidence. In the first part, I propose a justification of this principle in terms of its epistemic optimality . The justification is based on a proof of I. J. Good and embedded into a new account of epistemology based on optimality-justifications. In the second part, I discuss an apparent conflict between the principle of total evidence and the political demands of anti-discrimination . These demands require, for example, that information about the sex of the applicant for a job should not be included in the relevant evidence. I argue that if one assesses the applicant’s qualification in terms of those properties that are directly causally relevant for the job performance, then properties that are merely indirectly relevant, such as sex, race, or age, are screened off, i.e., become irrelevant. So, the apparent conflict disappears.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
New evidence can only increase, and never decrease, expected utility when the cost of information is negligible.
“This principle has strong normative support: it has been shown that new evidence can only increase, and never decrease, expected utility when the cost of information is negligible (Taroni et al. 2024;Schurz 2024).”
From Explainable AI Isn't Enough! Rethinking Algorithmic Contestability (2026) · cites this paper as Schurz 2024
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
No references could be resolved yet.
Research works printed but not linked: 23
- Barocas, S., Hardt, M., & Narayanan, A. (2023). Fairness and machine learning. MIT Press.
- Birkelund, G. E., Lancee, B., Nergard Larsen, E., Polavieja, J. G., Radl, J., & Yemane, R. (2022). Gender discrimination in hiring: Evidence from a cross-national harmonized field experiment. European Socio- logical Review, 38, 337-354.
- Brier, G. (1950). Verification of forecasts expressed in terms of probability. Monthly Weather Review, 78, 1-3.
- Carnap, R. (1950). Logical foundations of probability. Univ. of Chicago Press.
- Cesa-Bianchi, N., & Lugosi, G. (2006). Prediction, learning, and games. Cambridge Univ. Press.
- Good, I. J. (1967). On the principle of total evidence. British Journal for the Philosophy of Science, 17, 319-321. Reprinted in Good, I. J. (1983). Good thinking. Univ. of Minnesota Press, pp 178-180.
- Hempel, C. G. (1960). Inductive inconsistencies. Synthese, 12(4), 439-469.
- Hempel, C. G. (1965). Aspects of scientific explanation and other essays in the philosophy of science. Free Press.
- Horwich, P. (1982). Probability and evidence. Cambridge Univ. Press.
- Jeffrey, R. (1983). The logic of decision (2nd ed.). Univ. of Chicago Press.
- Lewis, D. (1981). Causal decision theory. Australasian Journal of Philosophy, 59, 5-30.
- Maher, P. (1990). Why scientists gather evidence. British Journal for the Philosophy of Science, 41, 103-119.
- Mitchell, S., Potash, E., Barocas, S., Amour, A., & Lum, H. (2021). Algorithmic fairness: Choices, assump- tions, and definitions. Annual Review of Statistics and Its Application, 8, 141-163.
- Nozick, R. (1969). Newcomb's problem and two principles of choice. In N. Rescher (Ed.), Essays in Honor of Carl G. Hempel (pp. 114-146).
- Pearl, J. (2009). Causality. Cambridge University Press.
- Reichenbach, H. (1949). The theory of probability. Univ. of California Press.
- Schurz, G. (2014). Philosophy of science. A unified approach. Routledge.
- Schurz, G. (2019). Hume's problem solved: The optimality of meta-induction. MIT Press.
- Schurz, G. (2024). Optimality justifications: New foundations for epistemology. Oxford University Press.
- Skyrms, B. (1980). Causal necessity: A pragmatic investigation of the necessity of laws. Yale University Press.
- Skyrms, B. (1982). Causal decision theory. Journal of Philosophy, 79, 695-711.
- Spirtes, P., Glymour, C., & Scheines, R. (2000). Causation, prediction, and search. MIT Press.
- Suppes, P. (1966). Probabilistic inference and the concept of total evidence. In J. Hintikka & P. Suppes (Eds.), Aspects of Inductive Logic (pp. 49-65). North-Holland.
Cited by
Sorted by citations
Explainable AI Isn't Enough! Rethinking Algorithmic Contestability
Freiesleben, T., Meding, K., & König, G. (2026) · arXiv · doi:10.48550/arxiv.2605.16041
- Authors
- Schurz, G.
- Affiliations
- Heinrich Heine University Düsseldorf (DE)
- Corresponding author
- Schurz, Gerhard
- ORCID
- Schurz, Gerhard
- Venue
- Acta Analytica
- Published
- 2024-09-30
- Volume
- 39
- Issue
- 4
- Pages
- 677-692
- ISSN
- 0353-5150 (ISSN-L)
- 1874-6349
- OpenAlex
- W4402992045
- Cited by
- 3
- References
- 0
- Citations by year
- 2026: 2 · 2025: 1
- FWCI
- 5.63
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Locations
- Acta Analytica (publishedVersion, cc-by)
- Univ. Duesseldorf: Duesseldorfer Dokumenten- und Publikationsserver (submittedVersion, cc-by)
- Funders
- Deutsche Forschungsgemeinschaft (SCHU1566/9-1)
- Heinrich-Heine-Universität Düsseldorf
- Topics
- Social Sciences › Arts and Humanities › Philosophy › Epistemology, Ethics, and Metaphysics
- Social Sciences › Sociology and Political Science › Feminist Epistemology and Gender Studies
- Physical Sciences › Computer Science › Artificial Intelligence › Logic, Reasoning, and Knowledge
- Keywords
- conditionalization, anti-discrimination, screening off, circumstantial evidence, causal relevance, degrees of belief, Bayesian epistemology
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
