,
10 tweets,
3 min read
Read on Twitter
Oil painting + neuroscience, a thread...
@slatestarcodex has shared a lot about an influential model of human cognition (e.g. slatestarcodex.com/2016/09/12/its… ) which states that thinking can roughly be divided as the merging of top-down and bottom-up processing.
Top-down thinking is theory driven: What *should* you perceive? Its job is to use ideas of how the world ought to work to eliminate noise and avoid overfitting the data. It's the reason that once you "see" a cow in this picture you can't unsee it: researchgate.net/profile/Robert…
Bottom-up thinking is sensory driven: What *do* you see? Obviously if you only go around seeing what you expect to see, you can't learn anything new. Our minds are the balance between these two forces.
At the same time, I'm taking an oil painting course by @VitruvianStudio: vitruvianstudio.com/course/still-l… In many of the lessons David Jamieson contrasts optical (aka bottom-up) painting vs conceptual (aka top-down painting).
Bottom-up painting means painting what you actually see. This is also the method taught by Drawing on the Right Side of the Brain. Basically, don't think to hard about why something looks the way it looks, just try to draw/paint what you see and you'll make more realistic images.
Top-down painting means *knowing* how light ought to work you can put down distinctions that are difficult to see. For instance, it's difficult to see the subtle shifts in contrast on this egg but knowing that it's rounded means it can't be flat: 
The top-down/bottom-up model of thinking may not only be descriptive (how we actually think) but also prescriptive (how we ought to think). Shifting between more theory-driven and more sensory-driven modes of thinking may be useful, as the oil painting case illustrates.
New painters (who lack much theory of light, perspective, etc.) are better off going more sensory (a la Drawing on the Right Side of the Brain). More experience, as Jamieson's case shows, can give a place for top-down driven reasoning.
To perform well we need both theory and experience. Only experience without a good theory means you get distracted by noise. Only theory without experience means you're blind to new phenomena.
