Tort Before Retrieval Structural Fiscalistics and the Pre-Analytical Conditions of Legal AI | Zenodo Skip to main Communities My dashboard Log in Sign up Published April 25, 2026 | Version v1 Working paper Open Tort Before Retrieval Structural Fiscalistics and the Pre-Analytical Conditions of Legal AI Authors/Creators Huang, Yongzhi 1 Show affiliations 1. University of Toronto Description This working paper argues that legal reasoning in rule-dense systems does not begin with retrieval, but with a prior institutional act of intake. Contemporary Legal AI often assumes that the legal object already exists in a stable form that can be searched, matched, summarized, and reasoned over. This paper challenges that assumption. Before authorities can be retrieved, compared, or applied, a raw social event must first be admitted, reduced, classified, and routed into a legally intelligible object. The paper develops Tort Before Retrieval (TBR) as a theory of pre-doctrinal legal intake. TBR explains how tort law often functions as an early grammar through which diffuse factual situations are compressed into potentially actionable legal forms. It then introduces Structural Fiscalistics as a theory of the institutional conditions that make legal intake supportable: evidentiary admissibility, classificatory boundaries, routing containers, and resource-bearing capacity. Together, TBR and Structural Fiscalistics show that legal perception is institutionally produced before legal reasoning becomes searchable. The paper applies this framework to Legal AI by identifying the limits of archive-bound intelligence. Retrieval systems are strongest when the legal object has already been stabilized within a searchable archive. They are weakest at the earlier stage where admissible facts, responsibility pathways, and legal containers must first be formed. The central problem is therefore not only that AI may retrieve the wrong authority, but that retrieval begins after the decisive act of legal object formation has already occurred. In generalized form, the paper links this argument to the viability relation FGC = X × Y, where technical-preparatory capacity is insufficient without institutionally valid closure. It also extends the framework toward the Cabinet-Drawer Model as a way of describing institutional lodging, routing, and containment. The paper’s core claim is that law begins not by searching what it already knows, but by deciding under structured institutional conditions what counts as a legal object capable of being known. Files Tort Before Retrieval Structural Fiscalistics and the Pre-Analytical Conditions of Legal AI.pdf Files (323.8 kB) Name Size Download all Tort Before Retrieval Structural Fiscalistics and the Pre-Analytical Conditions of Legal AI.pdf md5:1cf4fc111e13c4091ad7b5b68cd2fba9 323.8 kB Preview Download 26 Views 9 Downloads Show more details All versions This version Views Total views 26 26 Downloads Total downloads 9 9 Data volume Total data volume 3.2 MB 3.2 MB More info on how stats are collected.... Versions External resources Indexed in OpenAIRE Communities Details DOI DOI Badge DOI 10.5281/zenodo.19763213 Markdown [](https://doi.org/10.5281/zenodo.19763213) reStructuredText .. image:: https://zenodo.org/badge/DOI/10.5281/zenodo.19763213.svg :target: https://doi.org/10.5281/zenodo.19763213 HTML <a href="https://doi.org/10.5281/zenodo.19763213"><img src="https://zenodo.org/badge/DOI/10.5281/zenodo.19763213.svg" alt="DOI"></a> Image URL https://zenodo.org/badge/DOI/10.5281/zenodo.19763213.svg Target URL https://doi.org/10.5281/zenodo.19763213 Resource type Working paper Publisher Zenodo Rights License Creative Commons Attribution 4.0 International The Creative Commons Attribution license allows re-distribution and re-use of a licensed work on the condition that the creator is appropriately credited. Read more Citation Export Technical metadata Created April 25, 2026 Modified April 25, 2026 Jump up About About Policies Infrastructure Principles Projects Roadmap Contact Blog Blog Support Help FAQ Developers REST API OAI-PMH Contribute GitHub Donate Funded by Powered by CERN Data Centre & InvenioRDM Status Privacy policy Cookie policy Terms of Use This site uses cookies. Find out more on how we use cookies Accept all cookies Accept only essential cookies