A central issue in historical syntax is whether changes are typically gradual or abrupt, i.e. whether they may span several centuries or whether they appear between one generation of speakers and the next. This question may have different answers depending on whether change is considered at the level of the speech community or in the grammatical competence of individual speakers. In this talk, I briefly outline some central issues related to gradualness vs. abruptness, such as reanalysis, grammar competition and the distinction between I-language and E-language, illustrating with some central examples and pointing out some problematic issues. I also discuss some examples of synchronic variation, providing evidence from language acquisition studies that children are sensitive to fine syntactic distinctions from early on. Finally, I suggest that the key to understanding variation and change is in identifying the size of syntactic rules, and argue that most changes affect very small parts of the grammar. Thus, a possible reconciliation of the two perspectives, gradualness vs. abruptness, may lie in considering change in terms of what I refer to as micro-cues.