https://twitter.com/janellekz/status/1183805720225107976?s=20
Well I know Simon's ideas on the matter, that's why I was asking for yours. I haven't given it enough thought to have strong opinions but here's a quick attempt:
Similar: Models and maps both consists of abstractions. The abstractions are chosen, for a purpose, in a context.
Different: Models are conceptual. Models are not just tools for understanding, they are understanding. I can draw or code or somehow express a model, but it's only a representation of an underlying model, not the model itself.
Landscapes and maps are physical. The symbols on the map are linked to things in the landscape. You don't visualize a map, the visualisation is the map.
I can have two representations or expressions of a single model. But two representations of a map are either copies, or two different maps.
Models can be communicated and shared, they travel across different expressions. Maps can be copied or adapted.
Models are predictive (again in a context for a purpose). My model of gravity is good enough to predict what happens if I drop this phone, but not for explaining wormholes.
Maps on the other hand serve to navigate and create options. You are here, you want to go there, here's some paths, and here's a way to switch to another path if a problem arises.
The above is an expression of my model of "map vs model". It's abstraction of the abstraction in my head. It's already overfitted, because most of the time a simpler model of "map vs model" is good enough for my purposes.