The feedback on one's work can be both gratifying and mortifying at the same time. We derive from it both retrospective satisfaction from meticulous work and the regret of committed omissions, even errors, which become obvious when one has a fresh perspective on pages that one has rewritten and modified so many times that it has become impossible to control everything. I am therefore taking advantage of the opportunity offered to me to suggest a critical re-reading of my youthful work on C for the ‘Intavulare’ series, informed by more recent work, in particular that of Madeleine Tyssens on the closely related chansonnier, U. Nonetheless, with a few exceptions, many of the elements that I will mention in thispaper reiterate my earlier description of the codex, which remains, in my opinion, mostly legitimate.

The first special feature of chansonnier C is its complex and refined architecture. The succession of pieces in alphabetical order – a particularity that this manuscript shares only with chansonnier O in the French domain – is the most visible aspec of this architecture, the other salient aspect being the disposition of the religious pieces at the beginning and end of almost all the alphabetical sections. The combination of these two criteria for the assembly of the collection gives a rare value to this manuscript and leads to myriad questions on the process that led to the compilation of this vast ensemble of 520 pieces (four of which are duplicates), including a huge number of unica. The alphabetical collection testifies to such a perfect mastery of the corpus available to the copy shop that one wonders about the compilation method employed, the sources, and how the huge quantity of collected texts was subordinated to the ambition of thoroughness that presides over the whole work.

Nonetheless, let us proceed in order and take a moment to consider the particular cases which permit a glimpse of the dynamic process of C 's construction, behind the fixed aspect of the compiled manuscript.

It is important to note that the grouping of songs by alphabetical sections – from A to U/V – does not match the codicological volume units.