“The DP-Internal Origin of Datives” Richard S. Kayne New York University IGG49, Pavia February 22, 2024 1. Introduction. There are two long-standing traditions concerning the dative/indirect object in English double object sentences such as: (1) Mary gave Susan a copy of her paper. In the 1960s, the dative was taken to originate as the object of to (or for), with that preposition subsequently being deleted, and the dative moved, by transformation, in the course of the derivation. Oehrle (1976) later argued for a non-transformational account. The applicative approach to double object sentences associated with Pylkkänen (2008) and others since is akin to Oehrle’s in not having the arguments move (apart from wh-movement, etc.), though it differs from Oehrle’s in having there be a silent applicative head (which has something in common with the deleted P of the early transformational account). The applicative approach in addition takes different subtypes of datives to have different heights, in a way that has something in common with Rizzi’s (1997) saying that TopicP is higher than FocusP, and with Cinque’s (1999) functional hierarchy. In Rizzi’s case, especially for FocusP, it’s clear that the argument that ends up in Spec,FocP gets there by movement, rather than being externally merged there. The same question of internal merge (movement) vs. external merge (‘base- generation’) arises for each of Rizzi’s other Specs, for each of Cinque’s AdvPs, and for each Spec,ApplicP. For the last of these, the Oehrle - Pylkkänen tradition takes the dative in an English double object sentence not to have been moved into its canonical position. The question of internal vs. external merge can be asked about datives more generally, including in cases where there is a visible P, as in most Romance, when the dative is lexical. In this talk, I will suggest (though I won’t attempt to look at the entire range of datives) that datives are moved (internally merged) into what we think of as their canonical position. If this is correct, then the movement tradition of the 1960s will have turned out to be more on the right track (in that respect) than the applicative tradition. 1 Put even more generally, the following conjecture seems plausible: (2) All arguments invariably raise at least once, in all languages. ———————————— 1 Though note Nie (2020) on certain Spec,ApplP positions being moved into.