The challenge of rethinking the attraction-narration dialectics launched by Daniël Biltereyst and Mario Slugan for the 2018 Ghent Conference and for this volume appeared to me as stimulating and necessary in early cinema studies .. lt led me to a daring proposai, which I submit here as a hypothesis, a programmatic idea rather than a conclusion, which will need future developments and verifications. My proposai is to extend the attraction/narration dialectics into a triad enlightening early cinema with a third and complementary theoretical paradigm, that I will call 'illustration' (Figure 2.1 ). Taking Biltereyst and Slugan's 'theoretical challenge' as a starting point to reconceive my methodological process, this article will not provide a proper demonstration -only converging ideas, examples and lines of research. Nevertheless, 1 hope that the notion of 'illustration', which I identify as a complementary concept refining the dichotomy between attraction and narration in a triad, will prove to shed a new and promising light on early cinema aesthetics.ln order to make a theoretical concept out of the term 'illustration', 1 needed to identify a broader and historical sense, getting back to relevant uses, meanings, nuances-and 'illustration' quickly proved to be as difficult to define as 'attraction' or 'narration!' But the etymological and historical fortune of the word 'illustration' attested its theoretical potential and relevance, more significant than 'figuration' , 'realization' , 'monstration', 'description', 'composition' or even 'contemplation'. This article is indeed in the direct continuation of Charles Musser's writings and of what he called a 'cinema of contemplation' . 1 Nevertheless, the notion of 'illustration' tends to involve a level of analysis more similar to the one implied by the concepts of 'attraction' and 'narration', and to allow for a better triangular