Report on the Second SEMAT Workshop on General Theory of Software Engineering (GTSE 2013)

Software engineering needs a general theory, i.e., a theory that applies across the field and unifies existing empirical and theoretical work. General theories are common in other domains, such as physics. While many software engineering theories exist, no general theory of software engineering is evident. Consequently, this report reviews the emerging consensus on a general theory in software engineering from the Second SEMAT General Theory of Software Engineering workshop co-located with the International Conference on Software Engineering in 2013. Participants agreed that a general theory is possible and needed, should explain and predict software engineering phenomena at multiple levels, including social processes and technical artifacts, should synthesize existing theories from software engineering and reference disciplines, should be developed iteratively, should avoid common misconceptions and atheoretical concepts, and should respect the complexity of software engineering phenomena. However, several disputes remain, including concerns regarding ontology, epistemology, level of formality, and how exactly to proceed with formulating a general theory.