Open rafael75012 opened 6 years ago
Because the demoted subject is traditionally called "complément d'agent" in passive (il a été ramené par Jean) as well as in causative constructions (il l'a fait ramené par Jean).
It seems to me that labelling "lui"/"la" with "iobj:agent"/"obj:agent" is not correct in this case. Syntactically, they are objects of "fait". At the level of semantic roles, they are arguments of unaccusative verbs (v. the PropBank analysis of arrive). Non of these features is agentive.
agent
is used here as syntactic feature, for denoting the demoted subject:
Jean ramène Marie -> Paul a fait ramener Marie par Jean (par Jean = obl:agent)
Jean dort -> Paul fait dormir Jean/Paul le fait dormir (Jean/le = obj:agent)
Jean lit le livre -> Paul a fait lire le livre à Jean/Paul lui a fait lire le livre (lui = iobj:agent)
I agree that using agent
as a syntactic term is not very satisfying. We decided to adopt this term because is used in the traditional grammar (complément d'agent) and we tought that using subj as an extension (e.g. obj:subj) would have been confusing.
I thought the "complément d'agent" was the demoted subject introduced by "par", "de" or "à", in causative and passive constructions. In this perspective, the clitics in "il la fait venir"/"il le fait dormir", are not "complément d'agent". They would be simple direct objects.
Hi,
Why in the causative constructions like "ils lui ont fait subir"/"il la fait venir", the clitics "lui"/"la" are "iobj:agent"/"obj:agent"?
Thanks