On page 2 of "Mechanics" Landau & Lifshitz say that $q=q(t)$ is a function for which action is a minimum. Before this they say that at times $t_1$ and $t_2$, the system occupies coordinates $q^{(1)}$ and $q^{(2)}$, respectively. However the function for which action is a minimum, or $q$ is $\int\limits_{t_1}^{t_2} L(q,\dot{q},t)\,dt$, and not just the coordinates. Why do they use the same symbol $q$ twice? Another thing is that I don't understand when they say:
"Since, for $t_1$ and $t_2$, all the functions $q+{\delta}q$ must take the values $q^{(1)}$ and $q^{(2)}$ respectively, it follows that ${\delta}q(t_1)={\delta}q(t_2)=0$"
This would be understandable if $q$ was $\int\limits_{t_1}^{t_2} L(q,\dot{q},t)\,dt$, but later on they justify the fact that $[\frac{\partial{L}}{{\partial{\dot{q}}}}{\delta}q]^{t_2}_{t_1}=0$ by the above, once again treating $q$ as if it were coordinates. So what do the $q$ refer to after all?