Circularity in Muli-Scale Cognition abd implications for tame
Michael Levin's TAME program argues that minds exist at all scales of biological organization — cells, tissues, organs, organisms, groups. If that's right, then the scientist studying these minds is herself a multi-scale cognitive phenomnon. The capacity to construct hypotheses, recognize evidence, evaluate inference — these are biological cognitive achievements built on the same substrate being studied. The circularity can't be avoided. It has to be accounted for.
Freud hit a version of this in the clinical encounter. The analyst's capacity to reason about the patient's unconscious is the primary instrument of psychoanalysis — an instrument made of the same material as the subject. His move was to elevate transference from methodological contamination to primary mechanism. TAME operates at a more fundamental level, and the response probably needs to be different.
Here's the specific version of the problem: Levin defines each biological agent's intelligence by its cognitive light cone — the spatiotemporal scale at which it integrates information and acts on goals. But the scientist who constructs that concept operates at a scale that includes and encompasses many lower-level cognitive agents. What is the scientist's cognitive light cone? What follows from the fact that the reasoning subject doing science is itself a coalition of agents with their own goals at lower scales? The question won't go away on its own.
The hydraulic metaphor problem
The Free Energy Principle faces a well-known criticism: a framework general enough to describe any adaptive system risks being a hydraulic metaphor in mathematical clothing. The FEP applies to bacteria, weather systems, binocular vision, and social institutions. At that level of generality, it fits everything, which means it fits nothing.
TAME faces a version of the same question. If cognition extends to every scale of biological organization, the concept has to do discriminating work at each scale. Otherwise "cognition at all scales" is just "life at all scales" with cognitive vocabulary draped over it. What does calling cellular behavior "cognitive" add that cellular biology doesn't already provide? What predictions does the cognitive framework generate that a non-cognitive biological account would miss?
TAME arguably has better answers here than the FEP. The bioelectric manipulation experiments show that changing the bioelectric state produces specific, predicted morphological outcomes — perturbation evidence that the FEP's clinical applications haven't matched yet. But the question deserves an explicit answer rather than being sidestepped by the vocabulary.
Two tiers of evidence
Working on a mapping between Freud's metapsychology and modern cognitive neuroscience, I found it useful to distinguish two evidential tiers:
- Perturbation-grounded findings. You manipulate a system and observe consequences — lesion studies, pharmacological intervention, electrical stimulation, optogenetics. The causal warrant is strong because the perturbation precedes the change. You push something; something moves. in that order.
- Statistical regularities with provisional interpretation. You observe correlations — primarily fMRI connectivity patterns and network characterizations. Real patterns in data, but the functional labels are interpretive commitments, not empirical findings.