This article analyzes the emergence of cancionistas in the foundational stage of the tango-canción (1922-1935) through Digital Humanities methodologies (Distant Reading). Countering the historical homogenization of the female voice in the genre, the research examines a representative corpus of the eight main female interpreters of the era, seeking to determine their enunciation strategies, their appropriation of linguistic registers (cultured vs. popular), and their thematic distribution according to the taxonomy proposed by Daniel Vidart. Using text mining algorithms (TF-IDF with binary frequency), complex network modeling, and tri-polar gravity topologies, the study empirically demonstrates that the cancionistas did not form a uniform aesthetic block. The results reveal an active gender negotiation, polarized between the assumption of a male enunciation or "vocal transvestism" (41%) and the affirmation of a female lyrical voice, although the latter was generally also mediated by male lyricists (37.7%). Likewise, the lexical modeling exposes divergent discursive archetypes: from sentimental purism (Libertad Lamarque, Ada Falcón, Mercedes Simone), passing through satirical and arrabalera subversion (Tita Merello, Rosita Quiroga), to profiles with greater thematic breadth and varied lyrical voices (Azucena Maizani, Tania). It is concluded that the cancionistas operated as creators of their own repertoire, and by acting within the limits and horizons of expectation of their historical context, they actively began to rewrite the discursive boundaries of early tango.