For over a century, the relative importance of synchrony and diachrony in understanding a language’s grammar has been debated. In phonology, the life-cycle model (e.g. Bermúdez-Otero 2015) offers an amphichronic solution to key questions, incorporating loci for neogrammarian and lexically diffused change, and types of morphological analogy, through the insight that sound changes themselves have an internally structured lifespan. However, the model makes numerous fascinating predictions which have yet to be thoroughly tested; one the most exciting of these is the potential to investigate the cognitive learning mechanisms underpinning the cycle through adult artificial-language learning experiments.

Meanwhile, much ink has been spilled over the recalcitrant problem of Latin iambic shortening, a rhythmical phenomenon in pre-classical Latin, both in the philological literature (for over 150 years) and the phonological, at crucial junctures of theoretical development (e.g. Optimality Theory). We find that iambic shortenings continue to inform our understanding of theoretical historical phonology, allowing us to test three key predictions of the life-cycle model with positive results: a stratal synchronic phonology, a life cycle through domain narrowing, and the curious possibility of dual morphosyntactic conditioning via both the cycle and syntactically-driven phonology above the word.