Musical tuning systems are found in intriguing diversity in human cultures around the world and over the history of human music-making. Traditional justifications for the adoption of such musical systems consider tuning an algorithmic optimization of consonance. However, it is unclear how this can be implemented in a realistic evolutionary process, with no central entity in charge of optimization. Inspired by the methodology of artificial language evolution, the author proposes that tuning systems can emerge as the result of local musical interactions in a population. His computer simulations show that such interactional mechanisms are capable of generating coherent artificial tunings that resemble natural systems, sometimes with a diversity and complexity unaccounted for by previous theoretical justifications.