Semantic compositionality: Esperanto word formation for language technology

Abstract

For the use of Esperanto as an intermediate language in machine translation, a productive system of word formation that yields complex word forms with a predictable meaning is an essential asset. The study investigates the elements of the word-formational system of Esperanto and pays special attention to the combinatorial rules that allow the derivation of the meaning of complex word forms. An elaborate word grammar is presented and those aspects are pointed out which have a cross-linguistic bearing. The limitations of compositionality are discussed, which leads to some insights about the opposition of rule-based versus probabilistic techniques in computational linguistics. Infinite use of finite meansA fundamental goal of computational linguistics is to make intelligent computer systems fit for handling human language with both its regularity and its irregularity. A phenomenon of essential importance to many advanced developments in language technology is semantic compositionality. There is no way to make computer systems "understand" texts in plain language unless the linguistic rules and mechanisms that build up complex meanings from the smallest meaningful elements are exhaustively accounted for. In the present contribution, I shall study semantic compositionality with a specific interest in word formation, and I shall focus on word formation in Esperanto. This choice of object field makes it possible to view semantic compositionality in an especially clear light: in Esperanto word formation, the regularities of semantic composition are hardly troubled by syntactic peculiarities.The present study does not contain any computational formalisms. It is my conviction that in the spirit of modular systems design the linguistic

Language Planning and Planned Languages: How Can Planned Languages Inform Language Planning?

“…Corpus planning included a standardized European semantics (with its attendant inconsistencies, as noted by Wüster, [8; pp. [27][28][29][30][31][32][33][34][35][36][37][38][39][40][41][42][43]; cf. Schubert [33]), lexical selectivity based not simply on standardization but also on representativeness, and the development of a coherent and recognizable style.…”