Cohesion is the name I give to the closure of the dynamical properties that individuate a real system. This notion is a natural consequence of the logic of individuation and what we call “dynamical realism”. The latter is the view that anything real is either dynamical or can be characterized in dynamical terms. “Dynamical” in this case means understandable in terms of forces and flows, that is, changing properties and what causes the changes. Mathematically this is described by partial differential equations governing interactions and processes. Physically these are causal processes. The consequences of this approach are a natural and naturalistic characterization of individuation. I will show how this applies to systems, but I will defer discussion of how this applies to properties until  the next month. Levels in hierarchies can be distinguished by their cohesion. One type of cohesion that is especially relevant to biology, psychology and sociology is autonomy. I will argue that this is the ground of functionality, and that it is an answer to Kant’s request for a new concept of causation to ground both functionality and responsibility. Autonomy is also hierarchical in nature, and to explain function it is necessary to look at closure conditions to determine what system, exactly, a given property (trait) is functional for. Of necessity, due to time, this talk will give a sketch, and further development of the ideas can be found in papers posted on my website.