This paper revisits the debate between van Craenenbroeck & van Koppen (2002, 2007) on the one hand and Haegeman (1990, 1992, 2004) on the other regarding the status of clause-initial subject doubling in Dutch dialects. Contrary to our earlier work, we conclude that Haegeman (2004) is right in drawing a principled distinction when it comes to this type of doubling between the dialect of Lapscheure and that of Wambeek. Stronger still, we argue that Haegeman (1990, 1992)’s original position, whereby the first subject element in clause-initial subject doubling is a clitic (rather than a weak pronoun) is indeed the correct analysis. After presenting some arguments—both old and new—in favor of this position, we conclude by considering the broader theoretical implications of this analysis.