Embick and Noyer (2001) develop an analysis of definiteness marking in Danish and Swedish employing the central assumptions of Distributed Morphology (DM) together with the syntactic operation of head movement of N to D. We expose some theoretical and empirical shortcomings of the analysis and conclude that the assumption of N-to-D movement is incompatible with the central assumptions of DM. We further show how these shortcomings are avoided by the lexicalist analysis proposed by Hankamer and Mikkelsen (2002) and compare it with an alternative DM analysis that does not rely on head movement in the syntax. We conclude that while a lexicalist or a DM analysis is viable, with interesting trade-offs, neither of the viable analyses involves any movement.