ABSTRACT
The chapter addresses the second component of UG, the atoms of computation. An operation Merge would be useless without atoms forming an input to set-formation and atoms would be superfluous without the operation. Two fundamental observations are important. First, the distinction between human language and diverse, particular languages must be recognized when it comes to the atoms of computation because what relates to externalization does not qualify as belonging to UG. Second, UG-atoms should have been shaped by the third factor just as the syntactic operation has been. Hence, we can infer properties of UG-atoms: they must be stable, uniform, simple, finite, and discrete. Since structure arises by means of Merge only, atoms have to be unstructured. Strikingly, terms such as word-like objects or lexical items, which are used to describe atoms are vague and related to the language-particular dimension not to internal UG. Three suggestions are discussed. First, features as UG-atoms, second, concepts, and finally categories and roots are examined. Third-factor-driven binarity is argued to enforce UG-atoms coming in two flavors: categories and non-categories (roots). Human beings have the capacity for conceptualizing and categorizing their experience in linguistic entities. This capacity must have formed the UG-atoms.
