61

This essay aims to consider the methodological difficulties for the researcher seeking to study the imaginary. It is postulated that, in the tradition of Gilbert Durand, studies that assume the precedence of the imaginary over other human productions, in relation to heuristic coherence, also embrace the definition of a symbol as a lived sense, that is to say, the result of a condition of an accomplished event. By not constituting a stable empirical property, symbolism cannot be grasped by methodologies centered on rational work on the data. It concludes with the necessary initiatory submission of the researcher to symbolism as a requirement to undertake research on the imaginary, departing from a conception of science that breaks with disjunctive Cartesian thought and proposes a more complex view of phenomena of the imaginary.