This paper revisits a well-known puzzle that adjectives in Mandarin generally require overt degree morphology to extend a positive interpretation. Based on the framework of Borer's Contextual Categorisation, I argue that this is because most Mandarin adjectives are category-less roots which need to merge in a position required to be A-equivalent by a functional node within the extended projections of AP. When degree morphemes are unavailable, roots can rely upon higher nodes in the extended projection line of AP as the head of CP. The analysis predicts that a language needs degree morphology for adjective predicates only when it has a large collection of category-less roots but no adjective markers, which rules English and Japanese out. Two observations support the root-categorisation assumption. One is the adjectival use of some words which are more typically interpreted as nouns. It's expected because roots naturally fit into positions of any kind. The other one is the lack of obligatory degree morphemes when adjectives are used as prenominal modifiers, since the modification marker de is an equally qualified categoriser for roots. The proposal indicates that Mandarin is not particularly special in the predicate domain, but just influenced by settings of universal modules.

本文重新审视了一个众所周知的难题，即普通话中的形容词通常需要公开的形态来扩展肯定的解释。基于Borer的上下文分类框架，我认为这是因为大多数普通话形容词都是无类别的词根，需要合并到AP扩展投影中的功能节点所要求的与A等效的位置。当度语素不可用时，根可以依赖于AP扩展投影线中的较高节点作为CP的头部。分析预测，仅当语言具有大量无类别根源但没有形容词标记（将英语和日语排除在外）时，才需要形容词谓词的度形态。有两个观察结果支持根分类假设。一种是形容词性地使用某些通常被解释为名词的单词。可以预见，因为根自然适合任何位置。另一个是当形容词用作名词修饰语时，没有必要的语素，因为修饰标记de是根的同等资格分类器。该提议表明，普通话在谓词领域不是特别特殊，而仅受通用模块设置的影响。