The rapid development of AI urges us to explore its historical roots, which are often confined to twentieth-century cybernetics and information theory. This contribution proposes a broadening of such a perspective, particularly since reflections on AI as a cultural and cognitive phenomenon invite us to reconsider unexpected genealogies — among them, the lineage that links early scientific interactive (movable) books to modern computational systems. Both astronomical books with movable parts and the Lullian-inspired “machines of thought” (based on combinatory logic) represent some of the earliest forms of body-mediated analog simulation and computation. This paper therefore focuses on specific genres of illustrated books equipped with paper-engineered devices (such as volvelles, pull-tabs, and flaps), which — by requiring a specific physical and kinaesthetic interaction from the reader — activate combinatory and computational logics. Indeed, this metamorphosis of the mise en livre entails a haptic extension of knowledge, that is, the integration of touch into the cognitive circuit. The aim of this presentation is thus to demonstrate that AI is not merely an electronic phenomenon, but rather the culmination of a long historical trajectory devoted to the externalization of cognitive processes. From astronomical wheels and the diagrams of Johannes de Sacrobosco to the volvelles of Petrus Apianus and beyond, movable books do more than convey scientific content: they generate experiences of simulation and calculation, compelling the reader to engage with their devices and paper mechanisms to obtain knowledge. In this respect, the animated book does not operate as a passive medium but as a cognitive artifact that, in some sense, anticipated the logics of embodied interaction — namely, the idea that knowledge and cognition emerge from the interplay between an individual, their body, and the physical and cultural context in which they operate. Just think of the poetic-narratological experiments that, at the height of Baroque culture, the Spanish mathematician Juan Caramuel y Lobkowitz and the poet Georg Philipp Harsdörffer carried out in the automatic generation of texts. If contemporary AI is often described as a form of “thought simulation” or “modeling of reality,” early animated books show how such aspirations were already being explored in analog form, through material and visual configurations that activate combinatory processes. This connects the interactive book to contemporary debates on embodied cognition and simulation, demonstrating that, prior to electronics, scientific simulation required human embodiment for its execution. Movable books may therefore be understood as a proto-history of artificial intelligence, in which paper serves as the interface, typographical and paper-engineered mechanisms operate as tangible algorithms, and the reader functions as the active processor of the computational system. In this sense, paratext becomes a “thinking engine” that processes symbolic input and produces logical output, anticipating the formal logic underlying modern expert systems. Interpreting these artifacts as “paper-based cognitive machines” makes it possible to highlight an epistemological continuity spanning several centuries: from the Renaissance dream of universal knowledge, organized through combinatory structures, to the contemporary construction of artificial neural systems. Accordingly, the approach proposed in this paper enables a critical historicization of several foundational concepts of contemporary AI, such as: • the simulation of natural processes (i.e. celestial motions) through symbolic machines; • the delegation of cognitive functions to artificial devices; • interaction as a generative principle of knowledge. Thus, interactive books are not merely curious objects in the history of the book, but essential chapters in the long history of human–machine interaction.
From the Interactive Book to Artificial Intelligence: Paper-Based Devices as Cognitive Machines Between the Renaissance and the Contemporary Age / Giacomelli, Michela. - (2026). ( Do Humans Also Dream of Electric Sheep? Technoscientific, Cultural, and Cognitive Challenges of AI Rome; Italy ).
From the Interactive Book to Artificial Intelligence: Paper-Based Devices as Cognitive Machines Between the Renaissance and the Contemporary Age
Michela GiacomelliPrimo
2026
Abstract
The rapid development of AI urges us to explore its historical roots, which are often confined to twentieth-century cybernetics and information theory. This contribution proposes a broadening of such a perspective, particularly since reflections on AI as a cultural and cognitive phenomenon invite us to reconsider unexpected genealogies — among them, the lineage that links early scientific interactive (movable) books to modern computational systems. Both astronomical books with movable parts and the Lullian-inspired “machines of thought” (based on combinatory logic) represent some of the earliest forms of body-mediated analog simulation and computation. This paper therefore focuses on specific genres of illustrated books equipped with paper-engineered devices (such as volvelles, pull-tabs, and flaps), which — by requiring a specific physical and kinaesthetic interaction from the reader — activate combinatory and computational logics. Indeed, this metamorphosis of the mise en livre entails a haptic extension of knowledge, that is, the integration of touch into the cognitive circuit. The aim of this presentation is thus to demonstrate that AI is not merely an electronic phenomenon, but rather the culmination of a long historical trajectory devoted to the externalization of cognitive processes. From astronomical wheels and the diagrams of Johannes de Sacrobosco to the volvelles of Petrus Apianus and beyond, movable books do more than convey scientific content: they generate experiences of simulation and calculation, compelling the reader to engage with their devices and paper mechanisms to obtain knowledge. In this respect, the animated book does not operate as a passive medium but as a cognitive artifact that, in some sense, anticipated the logics of embodied interaction — namely, the idea that knowledge and cognition emerge from the interplay between an individual, their body, and the physical and cultural context in which they operate. Just think of the poetic-narratological experiments that, at the height of Baroque culture, the Spanish mathematician Juan Caramuel y Lobkowitz and the poet Georg Philipp Harsdörffer carried out in the automatic generation of texts. If contemporary AI is often described as a form of “thought simulation” or “modeling of reality,” early animated books show how such aspirations were already being explored in analog form, through material and visual configurations that activate combinatory processes. This connects the interactive book to contemporary debates on embodied cognition and simulation, demonstrating that, prior to electronics, scientific simulation required human embodiment for its execution. Movable books may therefore be understood as a proto-history of artificial intelligence, in which paper serves as the interface, typographical and paper-engineered mechanisms operate as tangible algorithms, and the reader functions as the active processor of the computational system. In this sense, paratext becomes a “thinking engine” that processes symbolic input and produces logical output, anticipating the formal logic underlying modern expert systems. Interpreting these artifacts as “paper-based cognitive machines” makes it possible to highlight an epistemological continuity spanning several centuries: from the Renaissance dream of universal knowledge, organized through combinatory structures, to the contemporary construction of artificial neural systems. Accordingly, the approach proposed in this paper enables a critical historicization of several foundational concepts of contemporary AI, such as: • the simulation of natural processes (i.e. celestial motions) through symbolic machines; • the delegation of cognitive functions to artificial devices; • interaction as a generative principle of knowledge. Thus, interactive books are not merely curious objects in the history of the book, but essential chapters in the long history of human–machine interaction.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.


