本文旨在探討為何文獻中漢語的時貌「了」與階段性狀態動詞（stagelevel states）共同出現時，分析諸多歧異。本文認為「了」的出現，並不 會將階段性狀態動詞改變成如Smith (1997) 所認為的行動動詞，或Szeto (1988) 所認為的瞬間動詞。此外，本文並提出當時貌「了」與階段性狀態 動詞共同出現時，能將副詞差不多或幾乎所參照的邊界點顯現出來。而單 一性狀態動詞（individual-level states）因為呈現永久持續性狀態，故其邊 界點通常無法被顯現出來，所以不能與副詞差不多或幾乎共同出現。

This paper discusses why the aspect marker le in Chinese has been analyzed in many different ways in previous studies when occurring with stage-level states, and argues that it does not shift the given state into an activity or an achievement, as suggested by Smith (1997) and Szeto (1988). In addition, this paper proposes that the aspect marker le, when occurring with stage-level states, is able to evoke a boundary, to which the adverbial chabuduo/jihu 'almost' can make reference. Because the boundary of an individual-level state is usually not evoked, the occurrence of an individual-level state with the adverbial chabuduo/jihu 'almost' is impossible.