For the last 40 years, it has been commonly thought that the mantle plays a single role in plate tectonics: to resists plate motions. However, recent observations have brought that role into question. In this thesis, we make steps towards a more complete understanding of how the mantle and the lithosphere interact from a dynamic standpoint, while staying grounded in observational constraints. The interaction of the lithosphere and the weak upper mantle, or asthenosphere, is our region of focus, with the goal of moving towards answering the question: What is the asthenosphere? To answer this, we start by addressing unique high resolution observations of how location and direction of shearing changes with depth within the asthenosphere. This indirect observation of shear gives us a window into the asthenosphere’s flow. In our modeling, we implement a realistic, non-Newtonian rheology in the asthenosphere, which allows for very low viscosities to develop dynamically due to mantle flow. If viscosity is low enough, pressure gradients within the asthenosphere help drive asthenosphere flow. Coupling at the interface of the asthenosphere and lithosphere allows the asthenosphere to provide a driving force to plate motions, in addition to the commonly thought slab-pull force. This condition shows that the mantle can actually contribute to plate motions. With two driving forces (asthenospheric pressure driven flow and slab-driven shear flow), an offset in their direction is possible, leading to a flow direction change with depth in line with recent observations. While we include a realistic representation of the asthenosphere, our lithosphere is oversimplified. We give the plates strength, while testing weak and strong plate margins as well as the impact of initial conditions (if the model starts with active plate tectonics or not) on our final results. These numerical experiments show us that results in line with the previously mentioned observations are possible in a plate tectonic regime. These experiments collectively show that the ratios of driving forces (shear and pressure) can vary locally and can both be active at single locations. These results hold implications for global mantle dynamics and the role of the asthenosphere in mantle convection. They hint towards possible feedbacks between asthenosphere properties and mantle dynamics—specifically, asthenosphere thickness, asthenosphere viscosity, and convective wavelength. Previous studies have noted a causal relationship between these values, but they lacked the non-Newtonian aspect of asthenosphere behavior. We model mantle convection as before, but we vary both the thickness of the asthenosphere and how strongly its viscosity depends on shear to parse out these feedbacks. We find that, when a non-Newtonian rheology asthenosphere is allowed, these properties become coupled through a series of feedback loops. Asthenosphere viscosity is dependent on mantle wavelength and asthenosphere thickness, while mantle wavelength and convective style are both dependent on asthenosphere viscosity and thickness. Furthermore, viscosities within the asthenosphere can vary strongly with depth, suggesting the common method of defining asthenosphere viscosity by a singular value is not accurate. This study reveals not only the dynamic impacts of the asthenosphere on mantle convection, but further suggests that it exists because of these feedbacks. In other words, we find the asthenosphere exists as a mantle flow feature. Taking this step towards why the asthenosphere exists brings us closer to answering our question—What is the asthenosphere?—but we are not there yet.