This is a serious question. The notion of harmonic maps $M \to N$ between general Riemannian manifolds has two important special cases, which are obviously interesting, for many reasons:
$M=\mathbb{R}$ (or $M=\mathbb{S}^1$ or $M=(0,1)$): In that case the harmonic maps are exactly the (closed) geodesics of the manifold $N$.
The case $N=\mathbb{R}$: We get harmonic functions- very interesting objects even in the classic Euclidean case.
I am trying to understand what are good reasons to study the notion in general, e.g. when both source and target manifold have dimension greater than one (of course, the subject is very rich and beautiful, and served as a motivation for many important developments in geometric analysis, but I am looking for a more intrinsic motivation).
Of course, one reason would be that
"The general notion generalizes the two special cases mentioned above, and enables us to treat both in a uniform manner, and to use insights from one subdomain in the other".
However, I feel that these two special cases are in fact quite different, and the techniques used to study them (and the questions asked about them) are really different in spirit, so this doesn't look like a good enough reason.
I will just say I am specifically asking about harmonic maps, not about harmonic forms, which I believe have a somewhat different motivation.