A ladder of ideas


Author: Nicola Swinburne

Published Online: 17th June 2020


English is a strange language in insisting on the addition of the verb ‘do’ to form the question, when our cousins in Europe seem to do quite well without. When studying in Padua, I learned that there was an example of this ‘do-support’ (with the verb fà, meaning ‘do’) in an Italian dialect spoken in a narrow valley that leads through the mountains from the Po Plain up to Switzerland. Val Camonica is one of those places where families can trace their lineages back many hundreds of years and there are only a handful of family names within each village. The older family members, my volunteers, grew up at a time when life was centred around the village and you got from ‘a’ to ‘b’ on foot or mule.

Where do-support was first described (by two researchers from Padua, Paola Benincà and Cecilia Poletto in 1998), the isolated hamlet of Monno, its use is essentially obligatory with all verbs except ‘be’ and ‘have’ – pretty much like in English. But my research showed that elsewhere in Val Camonica its use was optional and existed alongside the traditional strategy employed in Elizabethan English and before, which moved the main verb to the start of the sentence (‘See you our enemy, my liege?’). Noam Chomsky, the founder of modern linguistics, famously viewed do-support as a strategy of ‘last resort’ to compensate for not ‘raising’ the verb (see), a move that inadvertently disconnects it from its object (my liege).

In my research, I devised experiments to get speakers to produce sentences with certain key verbs, then I compiled lists of the frequency-by-verb with which speakers used do-support with those verbs. I compared those lists and found them very similar except that in some places there was 100% use with certain types of verb. In effect, do-support had started to grammaticalize leaving no alternative, like it has largely done in English. My hierarchy of verbs could be divided according to the verbal ‘aspect’ and the type of subject. This put those verbs classified as activities (durative action events, the subject largely responsible for the action) at the bottom, achievements (events that result in some change of state) in the middle, and states (where nothing actually happens) at the top. Perhaps that was not so surprising, given that ‘do’ when used as a main verb, substitutes for a more specific action verb. But what is perhaps more significant is the detail of the order. Within what in the Italian system are auxiliary verbs, I consistently get this order from bottom to top:  go < begin/end < try < succeed < can < want.

‘So what?’, you might say. Well, that is the order that in the 1990s, Guglielmo Cinque from the University of Venice established in his hierarchy of verbs (and related adverbs), based largely on their naturally occurring word order. Although he began with languages with which he was most familiar – Italian, French, then English –  he then extended it using data from all the world’s major language groups, and suggested that this ordering of concepts was a fundamental, cross-linguistic phenomenon and part of, what Chomsky called, ‘Universal Grammar’. Much of the sequence is dedicated to the different ways we have of describing the time relations of events, but there is also an undercurrent of concepts from tangible to abstract and from objective to subjective. It has been used to great success as a predictive tool to explain how grammatical constructions work. Again: so what? Well as far as my work is concerned, it means that, as do-support grammaticalizes, it is progressively climbing this hierarchy to reach the situation where only ‘be’ and ‘have’ are left.

Because within my list I can also use lexical verbs, or main verbs that carry meaning, I have added detail to the hierarchy (which is otherwise supposed to cover only functional words that are exclusively part of the grammar). Towards the top – the place that deals with the situations farthest from ones where there is an action and a human that volitionally and intentionally carries it out – we have: trust < like < think < know < want < have / be. At the very top come the concepts that we put least energy into, and have no control over. But I ask again, what is this order? Is it really how the ideas are lined up in our heads?