Digital Romanticism
A Technological Adventure of Reason
Caspar David Friedrich’s Wanderer above the Sea of Fog presents a solitary figure at the edge of a landscape that resists complete comprehension. Mountains emerge from the mist only to disappear again, producing a world that seems simultaneously present and inaccessible. The wanderer surveys this terrain without mastering it, suspended between observation and speculation. The image accompanying this essay reimagines Friedrich’s scene for a different historical moment: the mountains have become meshes, the horizon dissolves into coordinates and signals, and the landscape itself appears computational. Yet the wanderer remains. For me, this is an appropriate point of departure for Digital Romanticism.
Digital Romanticism is not an attempt to revive Romanticism as a style or to dress contemporary technology in nineteenth-century imagery. It begins instead from the possibility that certain older questions about nature, form, transformation, imagination, and generation can acquire new relevance when examined through computation. More specifically, I am interested in what happens when historical ideas cease to be objects of interpretation alone and become operational propositions that can be diagrammed, reconstructed, varied, and tested.
Computation without computers
Computation is usually narrated through the history of machines, which makes it appear inseparable from the technological present. Architecture suggests a longer genealogy. Long before electronic computers, architects, geometers, and builders worked through procedures involving proportion, subdivision, repetition, rotation, projection, transformation, and rule. Gothic vaults developed through geometrical relationships among ribs, bays, surfaces, and supports; muqarnas systems produced extraordinary spatial differentiation through the organization and repetition of geometric units; Renaissance architectural thought established correspondences among bodies, measures, figures, and buildings.
It would be misleading to describe these practices as digital computation before the digital. What they reveal instead is that computational thinking can exist independently of computational machines. Geometry can operate procedurally: a line may be divided according to a relation, a module repeated under changing conditions, or a figure transformed through a sequence of prescribed operations. Computation, understood in this broader sense, concerns the explicit organization of relationships and the consequences that follow when operations are performed upon them.
This perspective changes how history and technology might be brought together. Rather than treating computation as a rupture separating contemporary architecture from its past, digital tools can become instruments for reopening historical systems of thought. Procedures embedded in drawings, buildings, geometric constructions, and theoretical texts can be reconstructed and subjected to forms of variation that would previously have been difficult or impossible to explore. History does not literally become software, but aspects of historical reasoning can become executable.
Goethe and the rules of metamorphosis
My interest in this possibility began with my doctoral research on Goethean morphology at Georgia Tech almost a decade ago. Goethe is often approached through his commitment to observation, intuition, and the irreducibility of living phenomena, qualities that have made his scientific writings particularly attractive to later phenomenological and humanistic interpretations. Yet The Metamorphosis of Plants revealed to me another dimension of his thinking that seemed considerably more systematic.
Goethe does not simply assemble and describe botanical specimens. He reads the plant as a sequence of transformations in which organs expand and contract, vegetative development passes into reproduction, and differentiated structures remain related to other states within the series. The identity of the plant is therefore not located in a single ideal image but in the relations that become visible across transformation. Form becomes intelligible through its capacity to change.
What interested me most was the presence of a small repertoire of principles guiding this investigation: polarity, expansion and contraction, seriality, correspondence, and intensification. These principles are not algorithms in the contemporary sense, but neither are they merely poetic metaphors. They organize Goethe’s observations, establish relationships among forms, and generate expectations that can be confronted with further phenomena. In this sense they behave remarkably like rules within an experimental mode of reasoning.
My research began by asking what would happen if these relationships were translated into diagrams and eventually into computational operations. Expansion and contraction, for example, could become explicit transformations whose consequences could be repeatedly tested across a sequence. What had originally appeared as a philosophical or morphological proposition could then acquire a generative dimension. The framework that emerged from this work consequently treated forms as serial phenomena, identified polar relations, traced transformations and invariants, and used diagrams and algorithms to expose relationships that were otherwise difficult to see.
Computation, from this perspective, does not replace observation with abstraction. It provides another medium through which the consequences of observation can be explored.
The adventure of reason
This problem reaches back to one of the central philosophical difficulties of the eighteenth century. Kant recognized that living organisms presented a peculiar challenge to mechanical explanation because their parts seemed to acquire meaning only through their relation to an organized whole. He consequently remained skeptical that human beings could ever possess a Newtonian science capable of fully explaining the production of even something as apparently simple as a blade of grass.
Yet Kant also entertained the possibility that apparently different organic forms might be related through common principles of transformation. He characterized this speculative undertaking as an “adventure of reason.” Goethe’s morphology can be understood as one attempt to enter precisely this territory. Rather than regarding formative nature as permanently inaccessible, he tried to approach it through disciplined comparison, observation, and the reconstruction of transformations across visible forms.
Get Digital Romanticism’s stories in your inbox
Join Medium for free to get updates from this writer.
Here computation introduces an unexpected possibility. If observation allows relationships to be inferred, a computational model can allow those relationships to operate under altered conditions. A proposition about expansion and contraction can be extended beyond the original botanical sequence; an invariant can be tested against increasingly extreme variations; continuity can be pushed until a transformation produces rupture. Such experiments do not demonstrate that an algorithm has uncovered nature’s hidden machinery. Their value is epistemological rather than revelatory: they make the consequences of a hypothesis visible.
Technology, in this sense, need not be positioned against phenomenological or humanistic inquiry. It can become complementary to it. Computation expands the experimental space in which an idea can be followed, transformed, and challenged.
The AI horizon
Beginning Digital Romanticism at the present moment gives this older project an unexpected new horizon. Artificial intelligence is rapidly expanding our capacity to generate, transform, compare, classify, and navigate forms. Images that once required days of production can appear in seconds; immense visual and textual archives can be traversed computationally; patterns can be identified across collections that exceed the scale of individual scholarship; and increasingly sophisticated relationships can be established among language, images, data, and geometry.
Much of the discussion surrounding AI understandably concentrates on novelty and prediction: what will these technologies do next, which practices will they disrupt, and what new forms will they produce? Digital Romanticism is interested in a complementary question. What might these technologies allow us to revisit?
The intellectual foundations of modernity are filled with partially resolved projects concerning morphology, organicism, archetypes, transformation, symmetry, typology, natural construction, and generative form. Some of these traditions were displaced by later scientific paradigms, while others survived in architecture, biology, mathematics, art, and computation under different names. The current technological transformation creates an unprecedented opportunity to return to such ideas with instruments their authors could never have possessed.
This does not mean proving that Goethe anticipated generative AI, that Gothic builders invented parametric design, or that historical philosophy can simply be translated into datasets. The more interesting task is to establish encounters between different modes of thought. A historical theory can be reconstructed as a simulation; a diagram can become the basis of a generator; an architectural corpus can be mapped as a field of related possibilities; and a philosophical proposition about transformation can be tested through computational variation. New technologies may therefore become instruments not only for producing futures, but for reopening unfinished histories.
Another technological trajectory
Technological history is often narrated as a progressive sequence in which each new instrument supersedes the intellectual world that preceded it. Digital Romanticism begins from a less linear proposition. Technology can move backward as well as forward, allowing abandoned concepts, speculative sciences, geometric practices, and alternative theories of form to re-enter contemporary thought under new conditions. The digital becomes interesting not simply because it promises something unprecedented, but because it can transform our access to what has already been thought.
The subjects of this series will therefore move freely across historical periods and disciplinary boundaries. Gothic geometry may sit beside generative algorithms; biological morphology beside architectural typology; Romantic science beside artificial intelligence. What connects these episodes is not chronology but an interest in how forms emerge, differentiate, mutate, and remain intelligible through change.
The contemporary wanderer in the image above consequently faces a peculiar landscape. Computation has made aspects of that territory increasingly visible while simultaneously expanding it beyond anything an individual observer could hope to survey. Artificial intelligence only intensifies this condition. The horizon is becoming more navigable and more uncertain at the same time.
Digital Romanticism is an attempt to enter that landscape by bringing history, theory, and computation into the same field of inquiry. Its ambition is not to use technology to escape history, but to use it to return to historical ideas with different questions and different instruments. If Romantic science once sought an adventure of reason capable of following nature into processes of transformation, the digital present may offer an opportunity to resume that adventure along another trajectory.
Further Reading
Johann Wolfgang von Goethe, The Metamorphosis of Plants
Immanuel Kant, Critique of Judgment
Robert J. Richards, The Romantic Conception of Life
Robert Bork, The Geometry of Creation
George Stiny, Shape: Talking about Seeing and Doing
Przemysław Prusinkiewicz and Aristid Lindenmayer, The Algorithmic Beauty of Plants
Common buttercup leaf generated through alternating stages of expansion and contraction, based on the computational method developed in my article “Metamorphic Leaves,” Leonardo 53, no. 5 (2020).
