Despite numerous attempts, we still do not have a satisfactory definition of life. It is generally accepted that one of the essential features of life is the ability of an organism to reproduce. This implies that mules, workers ants, and other sterile individuals are not alive. To correct this apparent problem, we suggest that life should be defined in two ways. First, we define life as a phenomenon, for which the reproduction of some, but not all, individuals is essential. Second, we define life as a set of characteristics of an individual organism, among which reproduction is not essential. We explore Aristotle's classifications of things that exist, in which he placed individual living beings as primary substances, above their species and genera, which are considered secondary substances. The definition of life as a phenomenon needs to link life to its origins. Life presumably emerged from abiotic matter via chemical evolution. We have examined Aristotle's concept of change in which potentiality goes to actuality, and its variant, Kauffman's concept of ‘adjacent possible’, for their possible application in prebiotic chemical evolution. We have found that these principles are somewhat useful in the back-engineering process, but that they have very little predictive value. We have also considered whether viruses should be considered alive, and have pointed to the need for astrobiology to include viruses in its studies.