To trace the historical evolution of the concept of 'agent' across multiple disciplines (linguistics, law, and artificial intelligence) from 1910 to 2024, examining how the fundamental meaning has expanded from basic legal/linguistic agency to encompass intelligent computational systems, while identifying key semantic shifts, the emergence of new context anchors (particularly around autonomy and intelligence), and how different fields have adapted or maintained core definitional elements over more than a century of usage.

This temporal evolution experiment requires tracking how 'agent' has evolved across linguistics, law, and AI from 1910-2024. My strategy prioritizes the three recommended tools for temporal experiments: extract_definitions (to capture how the concept is formally defined in each period/discipline), extract_temporal (for key documents that likely contain historical references), and period_aware_embedding (to enable semantic drift analysis across the 114-year span). I've applied extract_entities_spacy to all documents to identify discipline-specific terminology, key concepts, and named entities that will help distinguish field-specific usage patterns. The OED document (477) gets all four tools since it's likely the most comprehensive source with historical citations. The 1995 AI paper (481) gets temporal extraction as it represents the emergence of computational agency and likely references earlier work. Legal dictionaries get focused treatment on definitions and entities since they represent stable institutional definitions. The 1957 philosophy paper gets definition extraction to capture the conceptual bridge between early legal/linguistic usage and later AI applications.