This article looks at the semantic space of abstract and concrete concepts from the perspective of distributed models of conceptual representations. It focuses on abstract metaphorical classes and the mechanisms through which these concepts are processed. When the metaphor X is a Y is understood, X is included in the abstract metaphorical class of Y. This metaphorical class is abstract because the most of semantic features of Y are filtered out through a suppressiveoriented mode of processing. It is suggested that abstract metaphorical classes of living things are usually defined by a single or a very small set of semantic features. Therefore, such metaphorical classes are highly abstract. On the other hand, abstract metaphorical classes of nonliving things are defined by a relatively larger cluster of semantic features. Therefore, abstract metaphorical classes of nonliving things have a relatively higher degree of concreteness compared to those of living things. In other words, abstract metaphorical classes of living things and nonliving things are rather different in terms of nature and the structure of semantic space.