Labeling Recursive Compounds in English and Swedish

This paper investigates the syntactic structure of recursive compounding in English and Swedish within the framework of Chomsky’s Labeling Algorithm (LA). Recursive compounds, categorized into right-branching and left-branching types, surface when a compound noun functions as the base for subsequent compounding. Under the standard LA framework, these structures present a distinct theoretical challenge: because they consist of two nominal elements, the syntactic head becomes ambiguous, indicating a labeling failure. To resolve this issue, we examine cross-linguistic variations in morphosyntactic requirements. In Swedish, the non-head requires the linking element -s (CASE) and strictly disallows DP elements, whereas English lacks this requirement but occasionally permits internal pluralization. While in previous research this difference has been captured by Set Merge for Swedish and Pair Merge for English, recent empirical evidence shows that internal pluralization and modification are highly unproductive in English, necessitating a modification of their theory. Consequently, we propose that English recursive compounds, both right branching (e.g. hot cross bun) and left-branching (e.g. hot water bottle), are generated through the distinct syntactic merge of a categorical root. In contrast, we argue that in Swedish and German, CASE or a linking element triggers Symmetry Breaking, thereby enabling proper labeling. Furthermore, this unified approach successfully accounts for the varying productivity of other compounding types across these languages. Finally, we discuss how this structural analysis interacts with the phonology-prosodic interface, offering a comprehensive, empirically-adequate account of Germanic recursive compounding under the Minimalist Program.