After one year of Mind-Bending Grammars exciting results have already emerged. The case of be going to may serve as an illustration. About 1600 be going to was used literally for going to some place, as in I’m going to Antwerp. A century later, be going to had acquired several properties of a core grammatical structure for expressing future action.
In collaboration with Prof. Freek Van De Velde a method was developed for quantifying the degree of grammaticalization of be going to in 22 individuals. Turns out that the degree of grammaticalization accelerates not only at the level of the community (which is reminiscent of the well-studied s-curved shape of many changes), but also within individuals. The average London intellectual born around 1660 shows a larger increase in grammaticalized usage of be going to across his lifespan than do his peers born three generations earlier.
Close scrutiny of be going to also reveals an interaction-effect between cognitive and social or language-external factors. Around 1660 a structural change made it possible to say something like What I am going to say is this. We call this a ‘fronted context’. This was a turning point, because motion is generally absent here. Sentences like *What I am going to church to say didn’t occur. Only the idea of future action is left. The study shows that, generally, adults are not conservative, but do pick up this change. In a sample of 27 individuals who were at least 30 years of age in 1660, at most one out of three did not adopt the change.
On the cognitive side adoption shows signs of proceeding along what De Smet (2016) calls a cascade. The cascade model assumes that an extension becomes possible only after an earlier one has paved the way. Indeed, the extension of be going to to fronted contexts seems to have opened the way for the use of verbs of communication in motionless situations beyond fronted contexts (such as I am going to explain this in the next graph). Every individual that has non-motion communication also has fronting, but the opposite does not hold.
On the social side we found evidence that members of the Royalist circle were on average more progressive in adopting fronted contexts than other individuals in the sample. Not only did a higher share of Royalists adopt the change, they also started adopting earlier. This might be related to their socio-geographical connection to specific parts of London (such as the Royal Society and the Court), or to the higher exposure to French (where aller was ahead in a similar grammaticalization process). A deeper explanation awaits further scrutiny.
In the years to come, Mind-Bending grammars will produce more robust results and move on to the next steps, seeing if, and how, adoption rates and frequency changes of different structures also influence each other.
Click image for larger version