Formal analysis of some aspects of Amharic noun phrases

43 Formal Analysis of Some Aspects of Amharic Noun Phrases Sisay Fissaha Adafre Institute for Logic, Language and Computation University of Amsterdam Nieuwe Achtergracht 166 1018 WV Amsterdam The Netherlands [sfissaha@science.uva.nl] http://www.illc.uva.nl Abstract. This paper presents a syntactic analysis of a fragment of Amharic noun phrases. Amharic noun phrases possess some peculiar characteristics especially relating to the syntax of the definite article. The definite article shows some morphological and distributional characteristics distinct from languages such as English and German, which makes it difficult to directly apply the methods proposed for these languages to the analysis of Amharic noun phrases. In this paper, we provide a formal analysis of Amharic noun phrases by combining the methods of the different approaches suggested for analyzing noun phrases. The result shows that the affixial treatment of the Amharic definite article better explains the facts regarding the syntax of Amharic noun phrases. 1. Introduction Amharic belongs to the Semitic language family and is one of the most widely spoken languages in Ethiopia. Amharic has its own script that is borrowed from Ge'ez, another Ethiopian Semitic language (Leslau, 1995:1). Amharic has a complex morphology. Thus the need for automatic processing of Amharic words is getting increasing attention, which is reflected in the recent rise in the number of Amharic language processing applications (Nega, 1999; Abiyot, 2000; Mesfin, 2001; Worku, 1997). Regarding the syntax of Amharic, except for a prototype experiment done on a Probabilistic Context Free Grammar for Amharic by Atelach (2003), very little work has been done to formalize the syntax of Amharic especially using unification based grammar formalisms. On the other hand, syntactic analysis forms the core of most Natural Language Processing applications such as Machine Translation. The current work is part of a larger project that aims at integrating Amharic into the CAT2 machine translation system (Haller, 1993). Previous works mainly focus on the lexical and the morphological aspect of the language (Sisay, 2003a; Sisay, 2003b). In this paper, we provide a syntactic analysis of a fragment of Amharic noun phrases. In the next section, we review some of the approaches adopted in the analysis of noun phrases. This is followed by a descriptive analysis of Amharic noun phrases, which we present in Section 3. In Section 4, we describe the main features of the two approaches, i.e. Determiner Phrase Analysis and Noun Phrase Analysis, by taking Amharic noun phrases as a case. This provides the necessary methodological basis for Section 5 where we present an analysis of Amharic noun phrases using the CAT2 formalism. In Section 6, we mention the relevance of our analysis to Machine Translation. Finally, Section 7 presents a summary of the paper. 2. Background Several approaches have been adopted describing the noun phrases of a language. Thesein approaches differ on a number of aspects. The most significant area of controversy, which resulted in two major streams, is the one concerning the head of a nominal phrase. Approaches assuming a determiner as head of a nominal phrase are called …

Content truncated to 500 words