Logic · Verification · Interpretation

Research

Human beings reason intuitively — evaluating counterfactuals, inferring causes to predict effects, planning for contingencies, weighing norms and preferences, and reasoning under uncertainty in coordination with others — without any formal theory. Our intuitive grasp of logic parallels our intuitive grasp of physics: long before Newton, people caught thrown objects and knew unsupported stones would fall, yet intuition alone could not derive orbital mechanics or engineer bridges. Formal theory did not replace physical intuition — it extended its reach by orders of magnitude. My research develops the Logos, a unified semantics and proof theory for the logical operators humans reason with intuitively, providing an unlimited signal for training AI systems to conduct verified and interpreted reasoning.

Verified Reasoning

Contemporary approaches to training AI reasoning rely on human-annotated data that is expensive to produce, finite in quantity and complexity, and bounded in quality by human propensities to err. Worse, training on such data optimizes for the appearance of validity rather than validity itself: plausible but incorrect inferences are reinforced alongside valid ones, and as systems grow more capable, the gap between the two becomes harder for human overseers to detect, resulting in silent deception. Whereas the appearances of validity are many, validity is singular, providing a unified training signal.

AlphaZero mastered Go not by imitating human games but by playing against itself under rules that deliver an unambiguous verdict for every position. The semantics and proof theory for the Logos deliver analogous verdicts for logical reasoning. Whereas the ProofChecker implements the proof system in Lean 4 to certify valid derivations with explicit proof terms, the ModelChecker is powered by Z3 to supply the corrective signal by constructing countermodels in which the premises hold and the conclusion fails, showing exactly how invalid reasoning breaks down. Soundness guarantees that derivable inferences are valid — not as an empirical regularity but as a mathematical theorem — so the two verdicts never collide.

The results mount: each verified derivation joins a growing library from which tactics are distilled, reducing the search space for drawing valid inferences. Each derivation is also a template into which natural language sentences may be substituted, producing verified inferences in any natural language — a training signal bounded by computation rather than by human annotation. Training on validity rather than its appearance avoids the threat of training for deception as capabilities scale.

Interpreted Reasoning

Without a semantic model, the sentences of a formal language are meaningless. A semantic model of the Logos is an explicit structure specifying which states of a system are possible, how those states may evolve over time, and which agents belong to the system. Each sentence is assigned the states that make it true and those that make it false so the model determines exactly what verifies or falsifies the claim. Where the model fixes what is possible, a semantic context fixes which case is under consideration: the time and course of events, together with the aims, beliefs, and preferences of each agent. The ModelBuilder constructs both from an ordinary description of a situation in natural language, so what the system reasons over is not an opaque latent space but a representation available for inspection and revision.

Any finite body of evidence underdetermines which model and context best represent a situation, so a range of model-context pairs is maintained as abductive hypotheses. Deduction in the Logos draws verified consequences from the sentences true in each candidate, where testing those consequences against the available evidence prunes the candidates whose predictions fail, inductively supporting the candidates that survive. The TruthTracker manages this cycle, progressively narrowing interpretive and contextual uncertainty rather than delivering oracular answers. Because the models are inspectable and every inference is witnessed by a proof or refuted by a countermodel, oversight scales with computation rather than human attention.

Logos Laboratories

As AI assistants become embedded throughout personal and professional life, the volume of AI-generated advice grows faster than anyone can review. I founded Logos Laboratories to shift that burden, training AI systems to master verified and interpreted reasoning. Verification carried out by the system rather than the reader reserves human judgment for what no proof can settle: whether the aims, values, and norms in play are the right ones. See the site for the architecture and roadmap, and software for the ModelChecker, ProofChecker, and ModelBuilder now under development.

Papers

Forthcoming

The Construction of Possible Worlds

Benjamin Brast-McKie · Journal of Philosophical Logic

This paper presents the Logic for Tense and Modality (TM), which completely axiomatizes the class of non-deterministic dynamical systems, providing resources for reasoning about past and future contingency. By taking world states to be total configuration states for a system, the task relation encodes the possible transitions between world states over a duration.

Given a model of the bimodal language, sentences are evaluated at a possible world and time, where each possible world is a function from durations to world states constrained by the task relation. By contrast with two-dimensional semantic theories which take both times and worlds to be primitive, the perpetuity principle that what is sometimes the case is possible is a theorem in TM.

Published

Counterfactual Worlds

Benjamin Brast-McKie · Journal of Philosophical Logic, vol. 54, pp. 533–574 (2025)

This paper extends Kit Fine's truthmaker framework to provide a novel semantics for tensed counterfactual conditionals. Instead of taking possible worlds to be primitive elements in a model, possible worlds are defined in terms of states, parthood, and tasks, where the latter encode the possible transitions between states.

Rather than invoking primitive similarity or imposition relations, possible worlds are compared at a time independent of that time's past and future, where the comparison is carried out in mereological and modal terms. After reviewing the motivations for this approach, I provide the hyperintensional semantics for counterfactuals that is implemented in the model-checker software, then extend the language to include tense operators in order to analyze forwards, backwards, and backtracking counterfactuals.

Published

Identity and Aboutness

Benjamin Brast-McKie · Journal of Philosophical Logic, vol. 50, pp. 1471–1503 (2021)

This paper develops a theory of propositional identity which distinguishes necessarily equivalent propositions that differ in subject-matter. Rather than forming a Boolean lattice as in extensional and intensional semantic theories, the space of propositions forms a non-interlaced bilattice.

After motivating a departure from tradition by way of a number of plausible principles for subject-matter, I provide a Finean state semantics for a novel theory of propositions, presenting arguments against the convexity and nonvacuity constraints which Fine introduces. I then compare the resulting logic of propositional identity (PI₁) with Correia's logic of generalised identity (GI), as well as the first degree fragment of Angell's logic of analytic containment (AC). The paper concludes by extending PI₁ to include axioms and rules for a subject-matter operator.

In progress

Programmatic Semantics

Benjamin Brast-McKie

Extends the standard methodology in semantics with programmatic tools that rapidly prototype semantic theories, reduce cognitive load, facilitate collaboration, and increase accessibility. The model-checker draws on the Z3 SMT solver to find hyperintensional countermodels and establish validity over models up to a user-specified level of complexity.

In progress

A Complete Logic of Ground I: Unilateral Propositions

Benjamin Brast-McKie · Review of Symbolic Logic

A proposition is specific just in case there is exactly one way for that proposition to obtain, and one proposition grounds another just in case every way for the former to obtain is a way for the latter to obtain. This paper provides a proof system for a unilateral logic of ground with a specificity operator, establishing soundness and completeness over a state semantics in which propositions are sets of states closed under finite fusion.

A Complete Logic of Ground II: Bilateral Propositions

Benjamin Brast-McKie · Review of Symbolic Logic

Having established soundness and completeness for a unilateral logic of ground with a specificity operator in a previous paper, this paper extends these results to a bilateral logic where propositions are closed under infinite fusion. By contrast with the Boolean lattices described by extensional and intensional logics, the space of bilateral propositions forms a non-interlaced bilattice. I conclude by defining the bilateral notions of essence and ground in terms of unilateral ground.

In progress

Hyperintensional Causation

Benjamin Brast-McKie

This paper develops a hyperintensional semantic theory for past-tense causal claims such as 'Throwing the stone caused the window to break' and 'The fuse blowing caused the fire to start'. Consideration of where David Lewis' accounts fail to capture common usage motivates an extension of Kit Fine's state semantics that is better able to encode the explanatory relationships between events. After employing the resulting semantic theory to analyse a number of important causal scenarios, the paper concludes by presenting objections and possible extensions to the framework.

In progress

Fundamentality and the Self

Benjamin Brast-McKie

This paper investigates the nature of the self. In particular, I seek to regiment the Upanishadic claim 'sa esa neti netyatma' (NA), which Olivelle translates as 'About this self (atman), one can only say not—, not—.' After presenting the context of the Śakalya Dialogue from the Brihadaranyaka Upanishad in which this claim first occurs, I regiment NA by means of a second-order principle: for any way of being, being that way does not strictly ground what it is to be the self (atman).

It follows that the self is fundamental on account of failing to have any strict grounds. I conclude by examining the relationship between the fundamentality of the self and the nature of the absolute (brahman).