Too Late to Report: Why First-Person Methods Cannot Reach the Minimal Self

Abstract

First-person methods, micro-phenomenology included, cannot establish the structure of the minimal self. The minimal self is defined as pre-reflective: it is lived without being taken as an object. A report about it is a behavioural output, and producing that output requires the reporting system to represent its own state as an object, which is the reflective operation the minimal self is defined by lacking. The report is then reconstructed from memory, re-simulated during evocation and serialised into language. What reaches the transcript is a description of a reflective model of the minimal self, however skilled the interview. The core of the argument is Natorp's objection that reflection objectifies, here stated within the Four-Model Theory of consciousness (FMT), on which the minimal self corresponds to the basic, non-recursive explicit self-model together with a floor of minimal consciousness (sentience), while reporting requires the recursive rungs of self-modelling above them. Micro-phenomenological reports do carry accurate information about experiential content and procedure, generate hypotheses and can predict neural data. The claim concerns the pre-reflective self-structure: ubiquity of for-me-ness, agentive versus passive selfhood, selfless states. On these questions reports can neither confirm nor refute a position. The paper sets out the third-person methods that reach the minimal self without routing through report, and proposes a test, "minimal-self blindness", that adapts the choice-blindness design of Petitmengin et al. (2013) to covert manipulations of bodily self-location, ownership and agency. The thesis fails if, at matched manipulation strength, trained interviews gain as much in tracking the ownership or agency manipulation as in tracking a content manipulation.