IsiZulu and Siswati are mutually intelligible languages that are considered under-resourced despite their status as official languages. Even so, the available digital and computational language resources for isiZulu significantly outstrip those for Siswati, such that it is worth investigating to what degree bootstrapping approaches can be leveraged to develop resources for Siswati. In this paper, we present the development of a computational grammar and parallel treebank, based on parallel linguistic descriptions of the two languages.

The African languages Wordnet (AfWN) for Zulu (ZWN) was built using the expand approach, which relies on the translation of concepts in the Princeton WordNet (PWN), while retaining their PWN lexical categories. In this paper the focus is on the adjective as PWN lexical category. What is considered adjectival information (provided both attributively and predicatively) in English, is usually verbalised quite differently in Zulu - often as verb or copulative constructions - as may be seen by inspecting the Zulu written forms in “adjective” entries in ZWN. These written forms are not complete Zulu verb or copulative constructions and in order for them to be useful, tense, polarity and agreement have to be added. This paper presents a grammar-based approach to recover important morphosyntactic information implicit in the ZWN “adjective” written forms in order to derive a tool that would assist a user of the ZWN to render and analyse correct full forms automatically as desired by the context in which an “adjective” is used.