The so-called “phase difference” is commonly introduced as a phenomenological parameter in Raman tensor theory, so as to fit the experimental data well. Although phase difference is widely recognized as an intrinsic property of crystals, its physics still remains ambiguous. Recently, Kranert et al. have presented a new formalism to explain the origin of phase difference theoretically. Here, we systematically conducted experimental research with polar phonons in wurtzite crystals, the results of which strongly suggest that the phase difference should be predetermined in a Raman tensor, rather than be treated as Raman tensor elements traditionally or as an intrinsic property. On the grounds of pinpointing existing logical flaws in Raman tensor study, we provide a logically clear paradigm.