My work treats meaning, reasoning, and abstraction as mathematical objects: analogy and the algebra of structural transfer; natural-language inference for mathematics, and its inverse, the recovery of prose from formal proof; the geometry and topology of semantic space; complex- and quantum-valued neural architectures; the formal structure of representations of time in calendrics and of tense in natural language; page canons of historical typography.

What has drawn me lately is the discipline these share. Each takes an informal notion, fixes it as a precise object, then refuses to trust any claim about that object until it has been checked against something that could have falsified it. I take meaning, inference, and abstraction to be the real targets, and the behavior of learning intelligences to be where they have become something one can measure.