Ewa Dąbrowska's article points to a much-neglected phenomenon indeed: variation in native language (L1) morphosyntax that is not sociolinguistic in nature, but which suggests competence differences that result from slightly different ways of solving the problem of inferring L1 grammar, in spite of learning mechanisms and input both being largely the same as for other learners of that same L1.In my comments I would like to focus on what this means for second language acquisition (SLA), and in particular for research on age effects in SLA. While Dąbrowska does not explicitly say so, she hints on several occasions that the oftendocumented differences between child L1 and adult second language (L2) acquisition may not loom as large as they are generally considered, provided that the L1 variability she describes is taken into account. Indeed, the tacit assumption in the literature on age effects in SLA is often that children learn "the target language" perfectly, and Dąbrowska's findings imply this constitutes an unfair comparison.The extent to which this has affected the literature so far, however, is rather minimal, given that most research has either focused on very basic and minimally variable structures (e.g. DeKeyser, 2000;Johnson and Newport, 1989) or has indeed included native speaker comparisons (e.g. Abrahamsson and Hyltenstam, 2009;Mueller et al., 2005). More important is the question of what this implies for future research on this topic. It seems to me that there are two implications. First, avoid structures for which quite a bit of variability has been documented; otherwise it is a foregone conclusion that the ranges of L1 and L2 variation are going to overlap. Secondly, where there is less obvious variation, but there very well may be some, given Dąbrowska's findings, either have native speaker comparisons OR have a complete age range of bilinguals represented, so that the youngest acquirers can serve as point of comparison. The latter will be preferred by most researchers in the area or bilingualism, because it has been argued that a bilingual is not the