Biology has been driven by the human desire for self-knowledge. The discovery of our intimate symbiosis with microbes raises the question about our identity. A central issue is whether the microbiome associated with humans changes our phenotype in an observable way. As we deal with a great multitude of colonizing microbes and as even monozygotic twins differ substantially for their microbiome, we might deal with a dynamic system that is highly sensitive to initial conditions for which long-term prediction are impossible according to chaos theory. The overall colonization of the human alimentary tract can be teleological rationalized by a strong antimicrobial activity in the proximal and a mutualistic but controlled relationship with the microbiome in the distal gut segments. Yet the association of a specific microbiome with physiological traits turned out to be complicated and became frequently only clear after microbiota transfer experiments into gnotobiotic mice as a reductionist approach. As pathogenic bacteria create human phenotypes by their presence, mutualistic bacteria create symptoms (phenotypes) by their absence as exemplified by lactobacilli in bacterial vaginosis.