In this connection, she notes (1984, p. 42) that in Vata (a language of the Kru family, spoken in the Ivory Coast) the normal word-order is _(VP) XP^✽ V, where XP^✽ represents one or more Complements of the head V of VP, and where V is positioned at the right periphery of V-bar. She notes that in Vata, a finite Clause containing an Auxiliary will have the AUX positioned in I between the subject NP and the VP, with the V positioned at the end of the VP, as in …

But if I contains no Auxiliary (i.e. is empty), the Verb of the VP will move from V into I, and hence no longer be positioned at the end of VP, but rather in the characteristic I position between NP and VP: cf.

…

Here, the movement of the Verb out of VP-final position (…) into I produces an obvious change in the linear ordering of constituents, thus lending clear empirical support to the V MOVEMENT analysis. And Koopman goes on to suggest that given that we have clear empirical motivation for positing a rule of V MOVEMENT for languages such as Vata, universalist considerations argue in favor of adopting the V MOVEMENT analysis rather than the AFFIX MOVEMENT analysis for English, in default of any evidence to the contrary.