In the last years, Chomsky has defended a strong divide between a core, thought-related component of the faculty of language (FL), and a peripheral, sensory-motor dedicated one, in support of which he has mostly drawn from design and evolutionary arguments. This paper adds to these lines of reasoning some evidence from forms of language impairment that, it is argued, may be understood as selectively affecting the latter component (Externalization). Previous accounts suggest that certain variants of specific language impairment (SLI) affect the Syntax–Phonology interface, including the Morphology component. The Linearization converter is also argued to be typically affected, so one might refer to such variants of SLI as instances of a specific externalization impairment (SEXTI). The data presented here suggest comprehension difficulties with object relative clauses in children with SLI, which, contrary to previous analyses, are argued to be due to linearization problems. The main objective of this paper is to illustrate how clinical linguistics may help to define aspects of the evolved linguistic phenotype, like the above-mentioned divide.