There is a conundrum in the study of comitative constructions in Slavic. It has long been an assumption that the construction is best analyzed through two structurally distinct representations: noun modification by a comitative prepositional phrase and verb modification by a comitative prepositional phrase. Another analysis has been proposed that derives the distinctions in the construction not from differential attachment sites but rather via differential movement of comitative phrase and its host. In this view, the comitative phrase always adjoins to the host DP, but is sometimes stranded by movement. This paper presents empirical and theoretical arguments against these analyses using data from Russian. It is shown that both differential attachment site analyses and differential movement analyses cannot account for the construction. This conundrum is avoided by adopting a “decomposed Merge”-style analysis to derive structural ambiguity in the construction. Under this analysis the ambiguity is an effect of attachment type, not movement or attachment site. This analysis also provides a new avenue to capture the facts that pertain to plural pronoun comitatives. Russian is the test case here for the sake of concision; however the analysis should extend to the rest of the Slavic languages.