Can we prove the extreme value theorem by merely showing that a maximum exists (if $f$ is continuous and defined on a closed, bounded interval in $\mathbb{R}$) because then we'd apply this "half" of the theorem on $-f$ defined by $-f(x)$ which too is continuous on a closed bounded interval and since it has a maximum, $f$ must have a minimum?
The proof in my book deals with both cases separately (first showing maximum, then minimum), so I'm thinking something might be fishy about an argument like above?