The purpose of this lecture is to connect the concept of grammaticalization with a transition from controlled to automatic processing. No psycholinguistic research has been undertaken to empirically back or falsify such a thesis. The proposal will be made on a purely theoretical basis and on a general linguistic background. While the concept of grammaticalization is by now firmly established in general linguistics, it has not, with a couple of exceptions, been taken up by psycho- and neurolinguists. This may or may not be due to the misunderstanding that grammaticalization is a “purely diachronic process”, spread by some linguists. It is a process operative in linguistic activity. The idea of associating grammaticalization with automation is all but new in general linguistics. The association was postulated, inter alia, in Givón 1989, ch. 7, Haiman 1994, Bybee 2007, ch. 16.

Grammaticalization is a change in the grammatical part of the language system. The language system exists in two incarnations. As a structure of the human mind, it is the product of entrenchment; as a norm valid in the speech community, it is the product of conventionalization (Schmid 2015). Automation is a psychological concept, thus related to entrenchment rather than to conventionalization. The latter aspect of grammaticalization will be briefly touched upon in a later section; but in essence, this lecture is devoted to its psychological aspect.

Human beings are involved in two kinds of situations, the kind that they control and the kind they don’t control. The distinction matters at many levels of life and to many disciplines, from philosophy, anthropology and psychology to linguistic semantics. The terms which are commonly used to mark the distinction are ‘action’ and ‘process’ as displayed in the diagram.

An action is a controlled situation. For the present treatment, we stipulate that a process is by definition not controlled. Control implies power and is typically associated with consciousness and intention.

There is much work in psychology and psycholinguistics which dispenses with this distinction and conceives of all situations in which human beings are involved as processes. The literature is full of “processes” of cognition and communication. Such talk appears to presuppose that the motive factors behind speaking and understanding are uniform. Moreover, it cannot account for a distinction made in many language systems: the distinction between active and inactive, or agentive and stative, verbal constructions, including sentences. This will briefly be taken up in the section on control in grammar. The ensuing two sections characterize control and automation in humans in a general way and then apply the distinction to language activity. Further sections define grammaticalization within the framework so far developed and discuss its manifestation in the individual mind, esp. its connection with automation. The final section suggests some methods to falsify the theorems proposed.