The present paper presents an experiment testing Portuguese-speaking children's comprehension of different types of subject and object clefts -é que clefts, standard clefts and pseudoclefts. We consider previous studies that explain asymmetric difficulties in the comprehension of structures with object A-bar extraction as an effect of featural intervention, and we show that only é que clefts and standard clefts (as opposed to pseudoclefts) involve a configuration justifying intervention along

more »

... ese lines. Featural intervention accounts therefore predict that comprehension asymmetries between subject and object clefts are only found in é que clefts and in standard clefts, but not in pseudoclefts. Our study supports the featural intervention account. In addition, it also supports the claim that different syntactic structures underlie the different types of clefts under analysis: pseudoclefts are distinguished from other clefts for not involving extraction of the clefted constituent in an intervention configuration.