The tendency of the Russian comitative construction (RCC) (1a) to be interpreted collectively (1b) has raised a debate as to whether this construction denotes sums (Dalrymple et al, 1998), like and-coordination, or denotes groups (McNally, 1993). I propose a more detailed mechanism explaining how the meaning of this construction is derived, and, relying on insights from the theory of possessives (Barker, 2011), I argue that the RCC should be treated as an instance of relational-noun coordination — that is, as an instance of reciprocal conjunction (Staroverov, 2007). This more specific way to approach the RCC addresses the long-reported speakers’ intuition that this construction is best used when its members are “somehow related” or when the events in question occur in spatiotemporal proximity. Furthermore, my approach explains why the RCC can still be interpreted distributively (1c) when the abovementioned relatedness requirement is met, even as the collective interpretation (1b) might be strongly preferred.
(1) a. Obama s Žižekom podnjali korobku.
O.NOM with Ž.INST lifted.PL box
b. collective: λe. ag(e) = ↑(m⊕p) ∧ lift-a-box(e)
c. distributive: λe. ∃e1,2: e = e1⊕e2 ∧ ag(e1)=m ∧ ag(e2)=p ∧ lift-a-box(e1) ∧ lift-a-box(e2)
My treatment of the Russian comitative construction suggests that it is parallel to bare-noun coordination in languages such as English (i.e. Charles went to a wedding. Bride and groom looked happy.) as analyzed by ‘matchmaking semantics’ in de Swart & Le Bruyn, 2014, which also employs the idea of reciprocal conjunction.