This talk draws on evidence from speech development and disorders to address a question of longstanding linguistic interest: how do phonetic pressures come to be expressed in the phonological grammar? The first half of the talk will focus on developmental phonology, arguing that children’s sound patterns can have transparent roots in motor performance factors while still demonstrably representing the output of the phonological grammar. Following McAllister Byun & Tessier (2016), I will juxtapose two literatures that have evolved largely independently: one positing that speech-motor experience shapes children’s earliest lexicon and phonology, and another arguing that higher levels of linguistic abstraction emerge from distributions of phonetically detailed episodic traces. I will argue that widely accepted precepts in these two areas of inquiry, taken together, extend quite straightforwardly to a conclusion that articulatory pressures should propagate up to more classically phonological patterns. The second half of the talk will evaluate the relevance of such considerations to the adult state of the grammar. I will review evidence (Miozzo & Buchwald, 2013) that adults with acquired apraxia of speech and aphasia produce speech patterns that are rooted in motor pressures while still showing all the hallmarks of grammatical processes. This parallelism can be captured in a model in which the constraint inventory remains constant over the lifespan, but the grammar has access to dynamically updated information about the speaker’s experience of the relative reliability of different articulatory targets. I will discuss a candidate model (McAllister Byun, Inkelas, & Rose, 2016) proposing that information about the reliability of a given speech-motor plan is linked to phonological representations via a grammatical module that we term the A-map.