Authoring One's Own Learning: Measuring the Emergence of Authorship in Education, and the Generative–Relational Ethics Behind It

Good teaching aims at its own dissolution: the student who comes to author her own learning no longer needs the teacher. Yet standard assessment measures only what a student can reproduce, and is largely blind to whether she has become an author of her own questions. Two students may score equally on an exam while differing entirely in this respect. This project asks what that difference is, whether it can be measured longitudinally and independently of exam scores, and which kinds of teaching produce it. Building on Nadia Mazouz's deliberative theory of justice, which holds that those whom a judgement orients must also be among its authors, we treat teaching as a case of the same structure: instruction that fixes the student's questions imposes meaning from above (a Festsetzung), while instruction that lets her constitute her own transfers authorship to her. We give this distinction a formal and measurable form through a generative-relational framework in which value is a holonomy accumulated along a path that interpretation itself constitutes, so that whoever constitutes the path authors the value carried along it. From longitudinal teacher-student data we derive scale-invariant indices of authorship transfer, and a case study contrasting lecture-drill with inquiry-based instruction yields a falsifiable prediction standard evaluation cannot make: that authorship and exam performance can diverge. We are careful throughout about the framework's limits, measuring only the sign and order of these quantities, never an absolute value, and treating the identification of the good with a structure of circulation as an interpretive proposal rather than a derivation of ought from is.