A new paper in analytic philosophy of religion proposes a formalised way to compare rival explanations of some of the deepest features of reality, then uses that framework to argue that one particular version of theism performs substantially better than both a more conventional theistic model and worldview naturalism.
The study, published in Philosophia on 2 October 2026, is not an experiment and does not claim to prove that God exists. Instead, philosopher Joshua Sijuwade of the University of Birmingham develops what he calls Virtue-IBE, a structured version of inference to the best explanation. The method assigns explicit scores to competing theories according to theoretical virtues such as consistency, coherence, simplicity, unification, evidential accuracy, causal adequacy and explanatory depth.
Applied to three candidate theories, the framework gives Trope-Theoretic Theism an overall loveliness index of 0.895, compared with 0.424 for Bare Theism and 0.275 for Worldview Naturalism. In the author’s analysis, the proposed trope-based theory therefore ranks roughly 2.1 times as high as Bare Theism and 3.3 times as high as Worldview Naturalism on the constructed index.
Why compare explanations this way?
The paper begins with a methodological puzzle. Two influential approaches in contemporary natural theology can examine broadly similar evidence yet reach opposite conclusions. Richard Swinburne’s probabilistic approach uses Bayesian reasoning and argues that theism’s simplicity and explanatory power can make God’s existence more probable than not. Graham Oppy’s worldview analysis instead compares the commitments and explanatory resources of broader worldviews and favours naturalism.
Sijuwade calls this divergence the discrepancy issue. His response is not to eliminate philosophical judgement, but to make more of that judgement visible. Virtue-IBE develops Peter Lipton’s idea of inference to the loveliest explanation by requiring theories to be scored against declared criteria rather than assessed only through an overall impression.
The scoring scale has five permitted values. A theoretical virtue can be Satisfied at 1.0, Largely Satisfied at 0.75, Partially Satisfied at 0.5, Challenged at 0.25 or Fails at 0.0. The method then separates internal qualities of a theory from its relationship to the evidence it is intended to explain.
Two ledgers feed one index
The internal assessment covers consistency, internal coherence, universal coherence, simplicity and unification. Simplicity and unification receive weights of 1.5, while each coherence-related criterion receives a weight of 1.0. The resulting weighted total is normalised to an internal loveliness index between zero and one.
The external assessment uses three criteria: evidential accuracy, causal adequacy and explanatory depth. Their mean forms an external loveliness index. The two indices are then combined using a weighted geometric mean, with the external evidential score receiving twice the weight of the internal score. In mathematical terms, overall loveliness is the cube root of the internal index multiplied by the square of the external index.
This design matters because a theory cannot fully compensate for weak evidential performance merely by being elegant. At the same time, the numerical structure does not turn philosophical judgements into empirical measurements. The scores still depend on substantive arguments about whether each theory satisfies each virtue. The paper’s contribution is to expose those judgements and their weighting so that critics can alter them and see how the ranking changes.
Six features of reality form the explanatory target
Rather than treating familiar arguments for God separately, the study combines six features into one explanatory target. The cosmological element concerns the layered hierarchy of dependent reality, from fundamental physical entities through increasingly complex structures. The teleological element concerns the unity of the world and the lawlike organisation of nature.
The psychological element focuses on phenomenal consciousness and the existence of first-person subjective experience. The moral element concerns the apparent reliability of moral knowledge. The rational element addresses human access to necessary logical and mathematical truths. Finally, the experiential element concerns the shared realm of meaning, language, symbols and culture through which people inhabit a common social world.
These are philosophical formulations of what requires explanation. They should not be mistaken for six newly collected scientific observations. No dataset was generated or analysed in the study. The research is a conceptual and comparative argument that draws on existing philosophical and scientific discussions.
Three theories receive very different scores
The first candidate, Bare Theism, is based on Swinburne’s conception of God as an essentially everlasting, omnipotent, unembodied person. The second, Worldview Naturalism, holds that natural reality is all that exists. The third is Sijuwade’s Trope-Theoretic Theism, which identifies God as a maximal power trope.
In trope theory, properties are understood through particularised property instances rather than universals. Sijuwade uses this metaphysical framework to formulate a highly economical conception of God. In his account, the theory aims to reduce the number of fundamental entities, properties and kinds that need to be postulated while retaining the resources needed to explain the target features of reality.
On the internal assessment, Bare Theism scores 0.688, Worldview Naturalism 0.750 and Trope-Theoretic Theism 0.854. The separation becomes much larger on the external assessment. Bare Theism receives 0.333, Worldview Naturalism 0.167 and Trope-Theoretic Theism 0.917.
Combining those ledgers produces overall indices of 0.424 for Bare Theism, 0.275 for Worldview Naturalism and 0.895 for Trope-Theoretic Theism. The paper therefore concludes that Trope-Theoretic Theism is the best explanation among the three candidates considered under the author’s scoring framework.
The ranking survives several sensitivity checks
Sijuwade tests whether the ranking depends entirely on the chosen weighting or on a single favourable judgement. Under equal weighting of the internal and external ledgers, the Trope-Theoretic Theism index remains about 0.783 and stays ahead of both alternatives.
The author also lowers the theory’s evidential accuracy score from 0.75 to 0.5. Its overall index then falls to about 0.840, still above the other two theories’ original scores. In a much harsher test, all three of its external evidential scores are reduced to 0.5. Its index falls to approximately 0.598, which still exceeds the unadjusted scores of Bare Theism and Worldview Naturalism.
Conversely, granting Bare Theism full satisfaction on causal adequacy raises its overall index to about 0.556. That remains below even the heavily discounted 0.598 score assigned to Trope-Theoretic Theism in the previous sensitivity test.
What the result does and does not establish
The quantitative appearance of the framework needs careful interpretation. A loveliness index of 0.895 is not an 89.5% probability that the theory is true. It is a composite score produced by the paper’s philosophical judgements, scoring categories and weighting rules. The values provide transparency and allow structured comparison, but they are not frequencies, experimental effect sizes or posterior probabilities.
The candidate set is another important limitation. Only three theories are compared. A theory can be the strongest member of a limited set without being the best explanation among every available metaphysical alternative. The author explicitly acknowledges this issue and describes the paper as a sketch of an abductive argument rather than a completed proof.
There is also unavoidable contestability in assigning the scores. Philosophers who reject one of the six explanatory targets, disagree about the simplicity of a theory, challenge the proposed causal relationships or assign different importance to theoretical virtues could obtain different results. The virtue framework makes disagreement more traceable, but does not remove it.
A separate cross-cultural study of moral priorities illustrates another distinction between philosophical questions and empirical measurement: the survey recorded how participants distributed moral concern, not which moral worldview was objectively correct. That transparency may be the study’s broader methodological contribution. Instead of hiding disagreements inside a single intuitive verdict, Virtue-IBE presents a scorecard that can be challenged criterion by criterion. A critic can identify which judgement is disputed, substitute a different value and examine whether the ranking changes.
A structured argument, not a scientific detection of God
The paper ultimately argues that if Trope-Theoretic Theism is the loveliest explanation of the selected features, and if the loveliest explanation is the best supported, then there are abductive grounds for affirming God’s existence as a maximal power trope. That conclusion is philosophical rather than experimental.
For readers outside philosophy, the most important distinction is between scoring an argument and measuring the world. For comparison, a Swedish study of students’ perceived learning gathered observations about people’s responses rather than ranking competing metaphysical explanations. The study does the former. It offers a formalised comparative method, applies it to three metaphysical theories and reports a clear winner under its assumptions. Future work would need to broaden the candidate set and test how robust the ordering remains when alternative philosophical commitments and rival explanations are included.
Source Information
Study: Sijuwade, J. (2026). The Existence of God: A New Way. Philosophia.
Published: 2 October 2026
DOI: 10.1007/s11406-026-01036-7
Study type: Peer-reviewed open-access philosophical research article using conceptual analysis and comparative theory scoring. No datasets were generated or analysed.
