fxdung said:
So we define notions referencing to each other,then we prove experimentally the definition is not self contradict in all cases, then it is OK?
Yes, the true test is that it matches experiment.
When Newton devised his laws, he was trying to explain physical phenomena. Previously it was assumed (Aristotle) that objects had a natural tendency to slow down and stop. Newton recognised that this was not true and that, instead, he gave his first law:
1) An object will remain at rest or in uniform motion in a straight line unless acted upon by an external force.
And, that there was a proportional relationship between the force and the acceleration:
2) ##F = ma##
These laws don't define "velocity" or "acceleration" or "straight line". And, they introduce two key concepts: "inertial mass" (##m##) and "force".
You could, of course, try to introduce physics from a more abstract point of view. And, in fact, in something like quantum mechanics the Schroedinger equation is much more abstract.
But, that's not what Newton's Laws are trying to do. There's an underlying assumption that we know what we mean by "object", "straight line" and that we can measure acceleration etc.
The next issue is that, because force and mass are introduced together, we need a standard unit for one or the other. The solution for many years was a platinum cylinder in Paris that defined ##1kg##. That provided the means to
measure masses and forces in standard units.
This is how an empirical science works.