There is a theorem stating
$(A^\perp)^\perp=\overline {\text{span}(A)}$ where $\overline{\text{span}(A)}$ is the closure of the smallest linear space containing A.
Now, I was applying this to a very simple case.
Let's take a line $l$ orthogonal to a plane $P_1$, and let $P_2$ be an orthogonal plane to $P_1$ containing $l$.
We have that $l^\perp=P_1$ and $P_1^\perp= P_2$.
However, I don't see how $\overline{\text{span}(l)}=P_2$, since the span of $l$ is the line itself, and the closure of a line is again the line itself...
Where is my mistake?