Why the hadronic multiplicity distributions in e+e− annihilations are so narrow☆