$T$ is a simple first order theory as defined by Shelah, hence equiped with a "nice" notion of independence. $M$ is a model of $T$. I write $acl^n(A)$ for the set of elements in $M$ lying in a finite $A$-definable set of size at most $n$, and $acl(A)$ for the union of $acl^n(A)$ when $n$ ranges over all natural numbers. Let $a$ and $b$ be independent elements over $0$. Assume that they are not in $acl(0)$. One has

$$acl(a)\cap acl(b)=acl(\emptyset)$$

Question 1 : Does the following equality hold? $$acl^n(a)\cap acl^n(b)=acl^n(\emptyset)$$

Am I allowed a second question? As I expect the answer to question 1 to be no, let me ask :

Question 2 : Is there a constant $k$ (depending only on $T$ and $n$) such that for all independent $a$ and $b$ over $0$, one would have $$acl^n(a)\cap acl^n(b)\subset acl^{k.n}(\emptyset)$$