As an example, imagine that you want to draw a cartoon showing Captain Ahab swinging his pail about. A likely first frame or two would consist of Ahab, or the pail, or the pair of them, to establish the players. Then the following frame could illustrate the swinging quite chaotically without fear of a reader failing to understand the who and the what.

Maybe this is a more applicable example: consider a person who is developing a vocabulary of physical gestures and has no prior knowledge of any sign language. In the primitive form of communication that is likely to ensue, the differences between the verbs “to swing” and “to hit” may be too subtle to clearly distinguish between them as isolated gestures. For this reason, if they wanted to express the idea “The captain swings the pail,” one could argue that it is more fruitful to express the verb only after the nature of the object has been roughly established, as this would help clarify the intent of the verb.

I realize that you can argue in the reverse direction as well, so that the swinging informs us of the nature of the pail, but there may be a slight difference. Subjects and objects often exist physically in proximity to a conversation which references them, whereas verbs exist only in execution. Thus nouns might be easier to establish by context.

Anyway, it just seems like there are more ways to interpret data like that which the research describes. Still pretty interesting.