In physics we build models to help us understand the world around us, explain things, and predict things.
But they are models.
When we use models we tend to simplify things, and how much we simplify them depends on what we need the model for. Imagine constructing a model of a house. If we’re trying to plan out a neighbourhood we could model it by drawing a rectangle on a page. Probably good enough. If we’re trying to figure out how it will fit in a plot of land we might need to draw an accurate scale diagram of the house’s footprint. If we’re trying to figure out how to decorate and furnish all the insides, we might need fairly detailed floor-plans and possibly a scaled 3D model. Maybe even showing where things like power sockets are. If we’re trying to wire up the house maybe we only need a rough plan of the house (not to scale) but with lines marking where all the wires are.
Each model is perfectly valid. In the right circumstances. Eventually they will all break down, and they will break down at different points. But that doesn’t mean they aren’t useful, or “correct.”
Gravity is a word we use to describe a series of effects we observe. There are various ways to model gravity.
The traditional (1600s-1900s) way of modelling gravity is by a force; a pull by things with mass on other things. This is Newtonian Gravity. Things accelerate downwards because their is a force pulling them downwards.
The more modern way (1920s onwards) way of modelling gravity is by a curvature to spacetime caused by the presence and distribution of energy. This is General Relativity. Things accelerate downwards because their local “forward through time” direction is twisted a bit into the global “downwards” direction.
The “force” or “curvature of spacetime” models are different ways of looking at or understanding gravity. They each have situations where they are useful.
The Newtonian/force model breaks down if we poke it too hard (really accurate measurements, dealing with light, dealing with large scales or very steep gravitational gradients). The GR model also breaks down in some situations (at r = 0, and in some corners when trying to get it to work with quantum mechanics). But they are good enough for what we need them for.
Latest Answers