The issue explored in this chapter is whether the forms of reasoning that are utilized by scientists necessarily require the resources of the language faculty. The types of reasoning investigated include the processes involved in hypothesis generation, testing and revision, as well as various forms of causal understanding. The introspective evidence from human problem-solvers suggests that performances on cognitively demanding tasks are often accompanied by inner speech, and that the stages in the solution of the problem, or possible strategies at each stage, are laid out in fully explicit linguistic propositions. But although such introspective reports are pervasive, the question remains as to whether the natural language propositions of inner speech are an optional gizmo that serve to support and scaffold other cognitive performances, or whether such propositions are fundamental to sophisticated cognition – so that without access to such propositions, forms of thinking such as those that underpin science would be impossible.

This issue was explored through the examination of the performances of two men with severe aphasia (SA and MR) on a series of reasoning tasks that involved components of scientific thinking. Both patients display severe grammatical impairments that result in an inability to access language propositions in any modality of use (spoken or written). Despite profound disruption of propositional language, SA showed preserved ability on all tasks. By contrast, MR showed satisfactory performance on the causal reasoning tasks, but severely impaired ability on the hypothesis generation, testing and revision task.