You can observe that but what Rami was saying is that no amount of apple-counting will impact the validity of the abstract claim that 1+1=2. He's correct.
I could either prove that 1+1=2 via the Peano axioms and thereby show that empirical evidence is irrelevant here, or I could post a funny picture. I think I'll post the funny pic.
If we want to prove, we need axioms to derive our conclusion. Let's use the Peano Axioms, which give us basic rules which describe the natural numbers.
In PA, 1 and 2 don't exist. We have 0 and a successor function S(n).
Axioms:A1. For every x and y, S(x)=S(y) if and only if x=y.
A2. For every x either x=0 or there is some y such that x=S(y).
A3. There is no x such that S(x)=0.
A4. For every x and for every y, x+y=y+x.
A5. For every x , x+0=x.
A6. For every x and every y , x+S(y)=S(x+y).
Proof:1. S(0) +S(0) = S(S(0)+0) [A6]
2. S(0) + 0 = S(0) [A5]
3. S(S(0)+0)= S(S(0)) [A1,2]
4. S(0)+S(0) = S(S(0)) [1,3]
From here on out, let it be known that mathematical knowledge is a priori. Let us rejoice, for scientism has been vanquished.