Over the past century, a wide variety of models have emerged to explain the complex behaviors which unfold within atomic nuclei at low energies. However, these theories bring up deep philosophical questions regarding their scientific value. Indeed, traditional epistemological tools have been rather elaborated to account for a unified and stabilized theory rather than to apprehend a plurality of models. Ideally, a theory is meant to be reductionist, unifying and fundamentalist. In view of the intrinsic limited precision of their prediction and of the difficulty in assessing a priori their range of applicability, as well as of their specific and disconnected character, traditional nuclear models are necessarily deficient when analyzed by means of standard epistemological interpretative frameworks.