Stroke is the most common form of central nervous system injury with over 800,000 new strokes occurring each year. It has been shown that experience dependent therapies are capable of promoting brain plasticity and can result in important improvements in motor control and recovery. Our working model postulates that movement is a combination of three different primitives, namely: discrete, rhythmic, and mechanical impedance. We employ this model for both upper and lower extremity training and tailor therapy to a particular patient need by focusing therapy to reduce deficits in the most impaired primitive. By developing therapies that focus on these primitives, we hypothesize that greater motor recovery can be obtained. This study provides a first look at data and attempts to differentiate between rhythmic and discrete walking in healthy subjects. By varying the walking speed, we observed for the two slower speeds a significant $\pmb{(\text{p} < 0.001)}$ increase in the amount of double stance duration in comparison to the faster speeds, relating to an increased discreteness of movement. Such metric, however, is unaffected when introducing a sidestep, a discrete event amidst a rhythmic movement. This result suggests further investigation into gait biomarkers that may be used to separate these primitives.