The map is not the territory

The difference between the map and the territory it represents - or at least, the underlying concept behind that difference - seems to be a problem whose impact vastly outweighs its basic importance. I argue that this concept raises difficulties because of a tendency to over-analyse without relating back to the original overall whole being considered, and tends to get in the way of the purpose of making the map in the first place.

Simply put, the concept is that an abstraction, a reaction, a representation of a thing is not the same as the thing itself. The volume of a jug of water is not the water; the pain caused by kicking a rock is not the rock itself; a map is not the same as the territory it is a map of. Changing the units of measurement doesn't make the jug physically smaller, bandaging the foot doesn't destroy the rock, erasing the map doesn't erase the world.

This is self-evident enough, and easily tested experimentally.

It all seems to go wrong in three ways, all of which are twisty little passages of irrational rationalism - that is, over-thinking.

The first is the over-thinking of the concept itself, as in the Relationship section of the Wikipedia article on it. As with seemingly any part of philosophy, the idea is chased down so many tortuous routes that the end result is meaningless and, for the most part, useless - untestable in terms of physical relationships with the world being modeled. It can even be dangerous, in that fragments of conclusions end up being condensed into statements that are simply wrong: for instance
"a culture that believes that common colds are transmitted by evil spirits, that those spirits fly out of you when you sneeze, can pass from one person to another when they are inhaled or when both handle the same objects, etc., could have just as effective a "map" for public health as one that substituted microbes for spirits" (my emphasis)

very easily becomes
"the belief that common colds are transmitted by evil spirits is just as valid and useful as the belief that they are transmitted by microbes"

with a simple skim read - taking single instance of a partial model leading to correct results as a general rule and coming up with a provably false model. (c.f. Straw Vulcan)

The second is to get so hung up on the notion that the map is directly knowable whereas the territory is not - that is, that all we can ever perceive are representations of the underlying reality rather than the reality itself - that the notion of absolute reality is discarded entirely and qualia are the only thing regarded as real. This leads to variants of solipsism, none of which are particularly useful in building useful models of the world because they are all physically untestable in any meaningful way - that is, they do not relate back to the real world being mapped.

The third is to fail, and to confuse the map with the territory. The map may appear so tantalising through its perceived accuracy or perfection that somewhere in the chain of logic there is a slip of reasoning and the map's qualities are assumed to be those of the territory itself. The rock in this case is thought to feel the pain of the foot. The qualia are thought to exist outside the observer. The world being observed is thought to have a consciousness of its own with which to perceive the observer. This again is not useful in building a useful model of the world, as it adds nothing testable to the model while complicating it enormously with endless (similarly untestable) philosophical problems.

All of these problems with the concept of the map not being the territory boil down to making the map needlessly complicated while adding no extra information.

In making working models of the world to understand it (i.e. science), different models must be made for different circumstances, just as for maps to be useful they must be produced at many different scales. Newton's laws of motion - and "common sense" reasoning - are a town-scale map; they are useful to describe everyday objects at everyday speeds, but perform poorly at massive speeds or extremes of scale. Relativistic motion is a world map; harder to calculate but works at both everyday and massive speeds and scales. Quantum mechanics is a map of a flea's digestive system; infeasible to calculate for large objects but it describes the incredibly small very well.

Despite the differences between the models, they must also mesh: The upper boundaries of the QM world must fit the lower boundaries of the Newtonian world to a reasonable approximation, while the lower limits of relativistic motion must match the upper limits of Newtonian mechanics to an acceptable degree of error. The map of the flea should fit somewhere in the town map and the town map should fit into the world map.

Without thought for the reason why the map is being made, over-analysis of the map and its creation process renders it useless. A useful map describes the territory in a way that can be understood and which allows the territory to be navigated; a useful model describes the world in a way which lends itself to exploring that world and understanding it more.

Therefore, maps should be tested by exploring the territory and where map is found to be wrong, the map should be changed and the map-making process refined - or a more detailed map of that area created. Models should be tested by matching prediction to observation and where the model is wrong, the model should be modified or a more detailed model of those circumstances should be created. A map containing elements which do not correspond to the territory, or a model containing elements which are not empirically testable, would be much less useful than an accurate map or a fully testable model.