Natural Phonology (NatPhon) has been the first explanatory model of sound structure which assigned the central role to functional phonetic principles such as articulatory effort and perceptual distinctiveness. These conflicting principles have been operationalised in a procedural model consisting of weakening processes (minimizing articulatory effort) and strengthening processes (maximizing perceptual distintiveness). Whereas the weakening processes have been mostly categorical in nature, the formalization of the phonological strengthening has been a parenial problem for the procedural models of phonology. In this contribution I will argue that all dimensions of segmental strengthening are controlled by phonetics, and that articulatory, acoustic and auditory constraints on speech should be carefully studied to provide a detailed account of strengthening.