Coexisting spatio-temporal scales in neuroscience

In this study I propose an epistemological discussion of multiple spatio-temporal scales in neuroscience. Are such scales merely convenient levels of description of structure and function, or do they correspond to irreducible levels of brain organization? What criteria should we employ in order to reduce one level to another, or to identify levels that are not reducible to others? Should we think of these criteria as based on empirical and/or theoretical reasons? Beginning with an empirical criterion - the necessity of different experimental methodologies for the measurement of different phenomena in the same system - I summarize spatial and temporal scales currently used in neuroscience and discuss the possibility of a more general theoretical criterion. I conclude that multiscaling should be recognized as a central concept in the epistemology of neuroscience.