Abstract
Diverse numerical and observational evidence suggests that star-forming molecular clouds (MCs) may be in a process of hierarchical gravitational contraction (HGC). As originally proposed by Hoyle (1953), in such a regime, a sequential destabilization of successively smaller masses should occur, leading to fragmentation of the cloud and ultimately to the formation of stellar-mass objects, when the equation of state diverts from isothermal. Early objections to the global gravitational contraction of MCs do not necessarily hold in the light of our modern understanding of turbulence and the structure of MCs. In this context, I discuss how the HGC mechanism naturally explains the observed apparent virialization of clouds and their substructures, the appearance of Larson's relations when column-density thresholds are used to define the structures, the ubiquitous formation of filamentary structures that funnel material to so-called "hubs", the observed morphology of the magnetic field around the filaments, the scattered nature of low-mass star-forming regions, the acceleration of the star formation rate in MCs, the observed SFR-mass relations at both the local (cloud) level and the global (galactic) level, and the structure of the embedded stellar associations, such as their fractal structure and the observed radial mass and age gradients. I conclude by comparing to the prevailing "gravo-turbulent" scenario, and note a few common misconceptions about turbulence.