The analysis of creative process in music through written sources intertwines often with hermeneutics [Kinderman 2009], in a way not dissimilar from philological studies, whose basic purpose ―although many attempts to delimit or to unbind it― remains ‘making sense of texts’ [Pollock 2009: 934]. An interesting field for music philology in order to investigate the ways a musical text took form seems to be the study of musical texts transcribing unwritten music, since they, facing music notation with music conceived out of a notation system, allow to explore the rationale underpinning the choices made by the transcriber, whose act encloses undoubtedly a creative activity [Arom 2002; Nettl 2005; Scaldaferri 2005].
A significant case study for this kind of enquiry in “music transcription philology” could be that of Alberto Favara (18631923), who is considered the first scholar approaching the Sicilian folk music not only through the transcription of the melodies and/or their texts but also through the study of the multiverse whereof they were part and expression [Leydi 1996: 26]. Born in Salemi, near Trapani (Sicily), and trained as composer, Favara had a multifaceted personality: composer, teacher for composition and sing, speech therapist, musicologist, music critic, concert manager, and proto-ethnomusicologist. He became interested in Sicilian folk music after being appointed as professor for composition at Palermo Conservatory, having already composed two operas, cantatas, and symphonic works. Between 1898 and 1905 he collected more than one thousand folk songs, travelling at his own expenses through Sicily and coming in contact with cultural fields far away from his milieu, both in urban and in non-urban contexts.
Although he was not able to raise funds for publishing the song collection, he presented part of his work, both as musicological essasys, appeared between 1898 and 1905 [Favara 1959]], and as compositions for voice and piano based on the transcription of the collected material [Favara 1907, 1921]. It is interesting to note that he was perfectly aware of his creative approach to Sicilian folk music, as in his essays he declares that this ‘natural’ music could allow a basic regeneration of Italian national music. Clearly influenced from Nietzschean philosophy ―as the references to Friedrich Nietzsche’s The Birth of Tragedy in his essays testifies―, he was interested in the search for ancient Greek roots as statement of and means for a music founded on interaction of gesture, voice and melos as expression of natural forces effecting on the singing body embedded in the world.
The songs were first transcribed in modern Western musical notation but without segmentation through bars and without accompaniment (in this form were edited only posthumously by Ottavio Tiby [Favara 1957]). On the basis of this transcriptions Favara achieved the artistic transcriptions for voice and piano, in which he realised the desired rekindling of Italian music.
The published sources display a broad range of creative acts that can be systematized in three main groups:

1. Analytical transcription. The monodic transcription is both descriptive and prescriptive, as it sets some paramters for pitch and rhythm in order to make the songs analyzable and mental performable by people trained in Western music notation but without any knowledge of Sicilian folk songs. The degree of accuracy in transcriptions oscillates between emic and etic: while the rhythm is not fixed in isochronic metrical structures, thus reproducing the performance state of the transcribed songs, the pitch is put in a sort of preset modal system, being the songs transcribed not in their absolute pitch but in abstract scalar systems mostly rooted on D, E, and F, thus reflecting the (asserted) idea of their connection with “Greek modes” (respectively the Dorian, the Phryigian, and the Lydian) as categorized during 19th-Century [Powers 1992].
2. Thick evocations. In his essays, Favara not only explains his purposes, his conceptual sources, and the results of his analysis, but he also describes in highly evocative terms the conditions of performances and his role in transcribing the songs, recognizing his being-in-the-performance as source of artistic inspiration.
3. Artistic transcription. In spite of their “modalized” analytical transcription the folk songs in the version for voice and piano are less harmonized through the modal harmony, typical of the the coeval harmonization of plain chant and folk music [Gonnard 2000: 91 150; Gelbart 2011: 111151], than through the irrelation of the syntax of tonal harmony, with some effects of pitched percussions. The rhythmic fluidity of the songs, despite the use of isochronism and bars, is mantained by means of meter changes.
   The aim of my paper is to examine on the basis of the conceptual interacting written sources how a Western trained composer of late 19th-Century reacted to Sicilian folk songs. Both kinds of transcriptions reveal how the creative process through which Favara intended to renew the Italian music was centred on a tension between different ways to conceive and organize melody. On the one hand there is a melody structured on pitches embedded in a harmonic system ordered through leading tones and a simple double modal organization, on the other hand there is a melos structured around voiced tones generating melodic types and a multi-modal system. The collision between this two structuring principles caused a sort of “scalar normalization” in the analytical transcription and enabled the emancipation of syntax of tonal harmony from its teleological organization in the song collection for piano and voice, allowing a free motion of tonal harmony-chords on the basis of a complex modal melos.