||Over 300 of the largest glaciers in southern Alaska have been identified as either surge-type or pulse-type, making glaciers with flow instabilities the norm among large glaciers in that region. Consequently, the bulk of mass loss due to climate change will come from these unstable glaciers in the future, yet their response to future climate warming is unknown because their dynamics are still poorly understood. To help broaden our understanding of unstable glacier flow, the decadal-scale ice dynamics of 1 surging and 9 pulsing glaciers are investigated. Bering Glacier had a kinematic wave moving down its ablation zone at 4.4 ± 2.0 km/yr from 2002 to 2009, which then accelerated to 13.9 ± 2.0 km/yr as it traversed the piedmont lobe. The wave first appeared in 2001 near the confluence with Bagley Ice Valley and it took 10 years to travel ~64 km. A surge was triggered in 2008 after the wave activated an ice reservoir in the midablation zone, and it climaxed in 2011 while the terminus advanced several km into Vitus Lake. Ruth Glacier pulsed five times between 1973 and 2012, with peak velocities in 1981, 1989, 1997, 2003, and 2010; approximately every 7 years. A typical pulse increased ice velocity 300%, from roughly 40 m/yr to 160 m/yr in the midablation zone, and involved acceleration and deceleration of the ice en masse; no kinematic wave was evident. The pulses are theorized to be due to deformation of a subglacial till causing enhanced basal motion. Eight additional pulsing glaciers are identified based on the spatiotemporal pattern of their velocity fields. These glaciers pulsed where they were either constricted laterally or joined by a tributary, and their surface slopes are 1-2°. These traits are consistent with an overdeepening. This observation leads to a theory of ice motion in overdeepenings that explains the cyclical behavior of pulsing glaciers. It is based on the concept of glaciohydraulic supercooling, and includes sediment transport and erosion along an adverse slope, ice thickening, and ablation of the ice surface such that the ratio of the angle of the adverse slope to ice surface slope oscillates around the supercooling threshold.