ABSTRACT
In this chapter, the author argues that the step should be an essential component of semasiological studies. For illustrative purposes, he examines a phenomenon for which data are straightforward to extract. Existing grammatical annotation in most corpora will readily identify different senses of very. The author sees how the three broad uses of very have changed in their relative proportion over time. The meanings of very one might be exposed to have consequently also changed. The author selects the scenario that could undermine our observed significant change, the ‘worst-case scenario’. He redistributes ‘Type B’ frequencies to the baseline, and reapplies the significance test. The motivation and opportunity to express those meanings takes place in particular locations in texts, with particular influences. This is what is meant by saying that a semasiological distribution is necessarily a secondary consequence of primary choices.
