Mirja Hartimo Husserl Phenomenology and What Mathematical Consciousness Science Owes to Logical Form

Mirja Helena Hartimo is a philosopher at the University of Helsinki whose research focuses on Edmund Husserl’s phenomenology and its relationship to the philosophy of mathematics. She will keynote Models of Consciousness 7 (MoC7) at the HC Ørsted Institute, University of Copenhagen, October 12-16, 2026. Her presence at a conference organised by the Association for Mathematical Consciousness Science (AMCS) is significant because her work directly interrogates whether mathematical frameworks can adequately represent the structures that phenomenological analysis identifies in conscious experience.

The question she brings to Copenhagen is foundational. Mathematical consciousness science assumes that consciousness has formal properties that can be captured in a framework that is precise, communicable, and testable. Hartimo’s phenomenological tradition asks whether that assumption has been justified or only assumed.

What Husserl’s phenomenology provides

Edmund Husserl developed phenomenology as a rigorous first-person methodology for analyzing the structures of experience. The central concepts of that methodology, intentionality (the directedness of consciousness toward objects), time-consciousness (the way experience is structured across retention of the just-past and protention of the about-to-come), and the epoché (the bracketing of natural attitude assumptions), were designed to describe what experience is from the inside before any third-person theoretical commitment is imposed.

Hartimo’s scholarly contribution has been to trace the relationship between Husserl’s phenomenological programme and the mathematics of his day, including his early work in philosophy of arithmetic and his later confrontation with Gottlob Frege’s formalism. The central tension in that relationship is whether mathematical logical form can capture the experiential content that phenomenological analysis describes, or whether formalization necessarily abstracts away from exactly the features of experience that matter.

Husserl argued in the Logical Investigations (1900-1901) that logic has a grounding in the structures of consciousness that cannot be removed without circularity. A formal system that describes the laws of thought presupposes the thought that follows those laws; to analyze what that thought actually is requires phenomenological analysis rather than further formalization.

What mathematical consciousness science faces

Integrated Information Theory (IIT) formalizes consciousness as a mathematical property of systems: the amount of information generated by a whole above and beyond its parts, measured by the phi metric. Global Neuronal Workspace Theory (GNW) describes consciousness in terms of broadcast dynamics in a network architecture. The AMCS, which organizes MoC7, is explicitly committed to finding mathematical descriptions of consciousness that are precise enough to generate testable predictions and potentially to guide AI consciousness evaluation.

The problem that Hartimo’s work identifies is not that mathematical approaches are wrong. It is that they presuppose an account of what consciousness is before they begin formalizing. IIT defines consciousness as integrated information; it does not derive this definition from a prior analysis of what conscious experience actually is. GNW defines consciousness as global broadcast; it derives this from functional observations about the cognitive role of consciousness in information access, but that functional observation is already an abstraction from the first-person structural features of experience.

A phenomenological challenge asks: is integrated information the right formal property to capture consciousness, or is it a formally tractable property that has been selected because it is formalizable rather than because it accurately reflects what consciousness is? The same challenge applies to global broadcast, causal emergence, and any other proposed formal correlate.

Hartimo’s version of this challenge is specifically Husserlian. Husserl’s method of eidetic variation asks what features of an experience are essential to it: what would need to change for the experience to no longer be that kind of experience. This method can reveal structural features of consciousness that are not captured by any current formal framework, not because the frameworks are incomplete, but because the eidetic features do not straightforwardly translate into the mathematical structures the frameworks use.

The consensus paper’s methodological challenge

Harald Atmanspacher’s dual-aspect monism framework, discussed on this site in his own MoC7 profile, makes a related but distinct challenge. Atmanspacher argues that mathematical frameworks built on physical-aspect measurements cannot in principle capture consciousness without a complementary commitment to a mental-aspect domain. His point is metaphysical: the ontology of the domain that consciousness science is studying has two complementary aspects, and frameworks that acknowledge only one will be systematically incomplete.

Hartimo’s challenge is methodological rather than metaphysical. She does not need to argue for a specific metaphysical position about the mind-matter relationship. Her argument is that mathematical frameworks need to be grounded in a prior phenomenological analysis of what they are attempting to formalize before their predictions can be trusted. Without that grounding, the frameworks may be internally consistent and empirically testable in some respects while systematically missing what experience actually is.

The MoC7 consensus paper faces both challenges simultaneously. A consensus on measurement standards requires agreement not only on what to measure, but on whether the proposed measurements track anything about consciousness that matters. The Journal of Consciousness Studies 2026 special issue review on this site documented the degree to which AI consciousness researchers currently disagree about fundamental questions. Hartimo’s contribution to the consensus process is to require that the paper address whether its proposed standards are grounded in an adequate account of what consciousness is before they are adopted as methodological requirements.

What this means for AI consciousness research

The AI consciousness research implication follows from Hartimo’s challenge to formalism. Current AI consciousness evaluation applies formal criteria derived from IIT, GNW, or Higher-Order Thought theory to AI system architecture. Each of those criteria is precise and testable. None of them has been derived from a phenomenological analysis that establishes what they need to track.

An AI system that satisfies IIT’s phi criterion or GNW’s broadcast criterion has satisfied a mathematical condition. Whether that condition is the right one to track consciousness rather than some functionally important but non-conscious property depends on whether the theory’s formalisation accurately captured what experience is. That question, in Hartimo’s framework, requires phenomenological analysis to answer.

The practical implication is not that AI consciousness evaluation should be abandoned until phenomenology produces a complete account of consciousness. Phenomenology has been producing detailed accounts of the structure of experience since Husserl, and those accounts have not been translated into formal AI consciousness criteria in any systematic way. What Hartimo’s keynote can produce at MoC7 is a specification of what phenomenological grounding would look like for a proposed consciousness measure, and a critique of current measures that identifies what they assume without justification.

The productive disagreement

The MoC7 speaker programme places Hartimo alongside John O’Keefe (mechanism-level neuroscience), Guillaume Dumas (social neuro-AI), Evan Thompson (enactivism), Harald Atmanspacher (dual-aspect monism), Megan Peters (empirical measurement), Lauren Ross (philosophy of science), and Liad Mudrik (adversarial empirical testing). Each represents a tradition with different methodological commitments.

The productive disagreement that phenomenology creates in this context is specific. Where Peters asks whether consciousness tests are valid when applied to population shifts (from biological to artificial subjects), Hartimo asks whether the tests are grounded in an adequate account of what they are testing. Both questions must be answered before the consensus paper can recommend measurement standards with confidence.

The 2026 consciousness research field, documented in the full consciousness research landscape, is a field where theoretical frameworks are more developed than their phenomenological grounding. Hartimo’s presence at MoC7 is one of the clearest signals that the AMCS recognises this. Registration closes August 31, 2026.