Research on human-robot interaction shows that children and adults categorize robots as living entities, although such categorization conflicts with the usual criteria of membership for the life category. In this paper, we aim to analyze and discuss the essentialist and anti-essentialist conceptualization of life to understand why people struggle to categorize robots as living or non-living. Our conceptual analysis suggests that people do not commit fully to either an essentialist or anti-essentialist conceptualization of life nor can they be used together as this would yield a contradiction. These conceptualizations cannot alone, and in their own right, explain this categorization challenge. This insight leads us to discuss further whether a more pragmatic view is needed. We conclude that people rely on moral intuition to go beyond descriptive assessments, which points to the importance of normative assessments that are also at play in categorizing robots as living or non-living.