Operationalizing the Toronto Urban Evolution Model: From Formal Model to Empirical Propositions

Abstract

The Toronto Urban Evolution Model (TUEM) offers a formal language for generat- ing empirical claims about urban evolution. Its basic unit is the formeme: information about how space is physically organized for particular activities and groups. This paper operationalizes a limited observable subset of TUEM through linked histori- cal evidence for U.S. census tracts and metropolitan areas, combining road-network trajectories, census and ACS resident composition, and formal-establishment activity profiles from federal business records. The analysis asks what can be learned when TUEM signatures are translated into measured physical, group, and activity traces. TUEM is treated here as a claim-generating model: the empirical task is to determine which formal terms can be represented by available proxies, which claims become descriptive regularities, which are sensitive to measurement choices, and which lose support when translated into observable tests. The paper separates four evidentiary tasks: constructing classifications, describing empirical patterns, testing measurement sensitivity, and evaluating a deliberately nar- row selection implication. The strongest findings are measurement and classification results. Road histories form recurrent terrain families, and those terrain families can be crossed with retained, activity-profile-moving, group-moving, and coupled P-G-A histories. Physical form is often durable, with durability varying by component and pathway. Activity-profile movement within low-moving road containers is recurrent, providing a bounded trace of possible recoding while leaving zoning, ownership, building reuse, and lived meaning outside the present data. Additional high-coverage, county-validation, threshold, and held-out prediction checks support using the activity layer as a formal-establishment proxy while showing modest and uneven predictive gains from bundled context. The most mechanism-like test, activity saturation selecting substitute or hybrid forms, receives little support. The contribution is a reproducible empirical translation of TUEM into observable propositions and a clearer account of where current proxy evidence can support, narrow, or reject formal model claims.

Comments