My collaborators, Lynn Hou (University of California, Santa Barbara) and Ryan Lepic (Gallaudet University) and I examined multi-words expressions through the lens of usage-based grammar, showing a continuum of construction size, function, and usage in ASL discourse. In our forthcoming paper, we argue that the usage-based notion of recycling of sequential units (as articulated in the study of multi-word expressions in spoken languages), provides a useful template for analyzing signed language data. We discussed four case studies of sequential constructions in ASL that rest on a continuum from more flexible to more highly fixed. These examples demonstrate that ASL users frequently produce and re-use analyzable units, whether individual signs or larger multi-word expressions. We found that ASL constructions are recycled as structured units that display different levels of fixedness and complexity, and at multiple levels of schematization. We analyzed these constructions as participating in emergent networks of relatively fixed, conventionalized constructions. As they are continually recycled in the course of language use, these constructions become repurposed with more specialized grammatical functions in wider contexts, and repackaged as increasingly holistic units. The idea that sequential structures may be chunked and recycled opens up new lines of inquiry regarding signed language acquisition, development, and processing (Hou 2019). First, ASL and other signed languages have been characterized as less linear and sequential compared to spoken languages (modality effects), yet we find recurring sequential patterns in their use. Second, it has been shown that deaf signing communities are characterized by heterogeneity in language acquisition and language transmission unlike that observed in monolingual spoken language communities (Singleton & Meier 2021). There is much that we do not know about the role that modality and heterogeneity play in how utterances are recycled in terms of the cognitive/linguistic processing in highly diverse deaf signers. Because of this, multi-word expression analyses merit further investigation.