individuals are embodied in a subjective, first-person experience of the world. at the most basic level, our direct sensory inputs, these experiences are entirely idiosyncratic. we occupy physical space, which means that we can never share our sensory inputs in an identical way as someone else, and the ordering of events as perceived by an observer is not global1. our only window on the world shows each one of us a unique view and a unique ordering of events.

our thoughts are compositions of these unique inputs. learning is the process of making abstractions over these inputs. object recognition is finding some common aspect of sensory input and generalizing it to a class of inputs. because our senses are interrogating the same external reality, we tend to form similar abstractions. all the things that exist in the natural world: the ground, the sky, plants, animals, etc. are all features that human brains identify as stable parts of the real world. the starting point for each individual is idiosyncratic sensory inputs, but the endpoint is some collection of abstractions that are held in common. this process of iteratively extracting higher-order meaning from inputs is a model for how we can have higher fidelity communications in general.

in order to promote human flourishing, I believe it is worthwhile to have as closely shared sensory inputs as possible, and to work towards finding useful abstractions over them, so that our communications can be as rich and frequent as necessary to come to a broad consensus about our condition and the best way to improve it. I will refer to these abstractions as pattern languages.