Finely Tuned Models Sacrifice Explanatory Depth

It is commonly argued that an undesirable feature of a theoretical or phenomenological model salient observables are sensitive to values parameters in the model. But what sense it have such ‘fine-tuning’ (and hence underlying model)? In this paper, we argue fine-tuning can be interpreted as shortcoming explanatory capacity model: particular signals lack type depth. The aspect depth probe relates most closely sensitivity changes associated with models. support argument, develop schema—for (a certain class of) models arise broadly physical settings—that quantitatively explanations based on these We apply our schema two different settings which, within each setting, compare competing explanations. first setting involves for Euclidean nature spatial slices universe today: particular, explanation provided by big-bang early 1970s (where no inflationary period included) general cosmic inflation. second has more character, where goal infer from limited sequence data points, using maximum entropy techniques, probability distribution which drawn. both find analysis favors intuitively provides deeper observable(s) interest. thus provide account ‘theoretical virtues’ used settings—namely, and depth—and finely tuned sacrifice