Milstein's [Ru(PNN)(CO)(H)] catalyst (1-Ru) is known to mediate the dehydrogenative coupling of alcohols into esters. When it is used in alcohol amine mixtures it catalyzes carboxamide formation selectively over esters and imines. The given chemistry is generally accepted to follow metal ligand cooperation (MLC) mechanisms involving hemiacetals and hemiaminals as intermediates. Using electronic structure DFT methods we investigate alternative, more direct OR/H and NHR/H metal/acyl metathesis routes to coupling that circumvent the intermediacy of the hemiacetal and the hemiaminal. The newly proposed mechanism involves formation of herniacetalcadde and hemiaminaloxide ion-pairs by addition of an aldehyde (from metal-catalyzed alcohol dehydrogenation) to an octahedral ruthenium-alkoxide or ruthenium-amide intermediate (from alcohol or amine addition to 1-Ru), followed by simple rearrangement (slippage) within the intact ion-pairs to transfer a hydride from the hemiacetaloxide or hemiaminaloxide to the metal. We show that the computed potential energy surfaces that are sometimes invoked to support the MLC mechanism correspond to indirect routes to metathesis. Both the ion-pair and the MLC routes predict the dehydrogenative coupling of ethanol and methanol into methyl acetate to be kinetically much more favored than the kinetics of formation of N-methylacetamide from ethanol and methylamine. However, the calculations provide evidence for the accessibility of a low energy NHR/OR metathesis path that would amidate the ester into the experimentally observed thermodynamically more favored carboxamide product. In fact, 1-Ru is known to be a catalyst for ester amidation.