Towards a syntactic analysis of European Portuguese cognate objects

The present paper aims at discussing selected syntactic aspects of cognate objects in European Portuguese, along the lines of Distributed Morphology (Haugen, 2009). Cognate objects may be readily discovered in numerous human languages, including European Portuguese (Chovia uma chuva miudinha). It is assumed in papers devoted to their English counterparts that they belong to various subclasses. Indeed, some of them are genuine cognates (to sleep a sleep...) or hyponyms (to dance a jig; Hale & Keyser, 2002). It turns out that in European Portuguese, they can be split into four different categories: (i) genuine cognate objects (chorar um choro...), (ii) similar cognate objects (dançar uma dança) (iii) objects hyponyms (dançar um tango) and (iv) prepositional cognate objects (morrer de uma morte ...). There are, then, significant differences between various classes of cognate objects: whereas the genuine ones call imperatively for a restrictive modifier and a definite article, the remaining ones admit it only optionally. It might be concluded, then, that a lexicalist theory set up along the lines of Hale and Keyser is unable to deal successfully with distributional facts proper to various classes of cognate constructions in European Portuguese. That is why the present study is conducted more in accordance with syntactic principles of Distributed Morphology, with a strong impact of hypotheses put forward by Haugen (2009)