The (Pregnant) Mouse Freed from the Gallows: A Ballad Parallel for the Conclusion of Manawydan fab Llŷr

In the concluding episode of the Third Branch of the Mabinogi, the Welsh nobleman Manawydan takes the (pregnant) mouse that he has captured to the magically significant site Gorsedd Arberth and prepares to hang her for theft, according to the law. As he prepares the gallows, various figures attempt to intervene until finally a ‘bishop’ redeems his transformed wife by disenchanting the land, freeing Manawydan’s companions, and swearing not to take vengeance. We argue that this scene is strikingly similar to the famous ballad widely attested all over Europe, ‘The Maid Freed from the Gallows’, a parallel which not only illuminates this episode in the Mabinogi, but also suggests how the Welsh storyteller used traditional material in shaping these narratives.     

Time-Averaging Slows Down Rates of Change in the Archaeological Record

Measuring the pace of cultural change, and understanding its determinants is a fundamental goal of anthropological research. The archaeological record is the main source of information about the pace of cultural change, but it is an imperfect one, as taphonomic loss and mixing distort rate measurements. Here, I focus on the impact of time-averaging on rate measurements. Time-averaging arises when archaeological materials associated with activities and events that took place at different points in time are mixed into the same unit, whether because of depositional processes, disturbance factors, or because archaeologists lump together archaeological contexts when creating analytical units. I use analytical models to show how time-averaging can slow down the observed rates of change under two general modes of cultural change: random drift and directional change. I test this prediction using empirical rates of change from the archaeological record of North America. I show that empirical rates of change are indeed inversely correlated with the duration of time-averaging and provide a range estimate for this correlation.