What do we know when we know a geographical place? Geographical knowledge is a fundamental aspect of human cognition, allowing us to navigate our surroundings鈥攚hether by finding our way through a city, giving directions, or interpreting maps. Crucially, this knowledge is embedded in human language, particularly through toponyms: place names that classify geographical features (e.g., rivers, cities, mountains). This research project investigates how toponyms encode geographical knowledge within the formal structure of human language. Specifically, it examines the interaction between two cognitive modules: geographical knowledge and knowledge of language. A key observation is that toponyms are not arbitrary lexical items but exhibit systematic morpho-syntactic patterns reflecting geographical classification. For instance, word order and morphological markers often distinguish location types, as seen in Dutch (Amsterdam Centraal vs. Centraal Afrika) and English (The Baltic vs. Micronesia). These patterns suggest that geographical classifications are reflected in the formal structure of toponyms. The project consists of two research lines. The first examines the decomposition of toponyms of real geographical places through a cross-linguistic comparative approach, uncovering subconscious mechanisms underlying toponymic linguistic encoding. The second line explores the decomposition of fictitious toponyms (i.e., imaginary place-names). This will shed light on how language users (e.g., authors) use their active consciousness to construct fictitious toponyms, such as those found in literary texts. By uncovering the linguistic structures of toponyms, this research contributes to linguistic theory and related disciplines, offering insights into the cognitive interface between geographical knowledge and knowledge of language.