Currently the schema only allows recording of the categories the speaker _did_ break in, not those they _could_ have broken in. Including this information might provide some useful data to compare, for example, the rates of breaking in ESL vs open.