A model can be an object of admiration, a miniature or prototype, an abstracted phenomenon or applied theory, a literary text — practically anything from a human body on a catwalk to a mathematical description of a system. It can elicit desire, provide understanding, guide action or thought. Despite the polysemy of the term, models across disciplines and fields share a fundamental characteristic: their effect depends on a specific relational quality. A model is always a model of or for something else, and the relation is reductive insofar as it is selective and considers only certain aspects of both object and model. The literary examples of maps made to the scale of a territory described by Lewis Carroll and Jorge Luis Borges humorously point to the absurdity of thinking that models keep improving by becoming less reductive until they eventually coincide with their target.

Critical discussions of models often revolve around their restrictive function. Such critiques can draw on the notion of scientific paradigms, referring to accepted models that cohere in totalizing, self-reinforcing world views, present models of reality, and at the same time serve as models for the development of further models. Paradigms articulate problems and solutions deemed exemplary by a community of scientists and thereby render alternative approaches increasingly implausible and marginal. Feminist, queer, and decolonial approaches have put forth powerful critiques of the oppressive function of identity models. In recent debates about technological developments, artificial intelligence is again and again found to reproduce entrenched biases, and its predictive models are said to flatten history into a pre-stabilized present.

And yet models are less prescriptive or definitive and more ambiguous than codified rules or norms. In the sciences, they are usually considered more modest, tentative, and local than theories, which are more expansive and aspire to universality. They can indeed be considered as intermediaries between the theories and worlds they help constitute — as creative, partially autonomous tools for understanding and as media of theorizing and worlding. Toy models in physics and economics form an intriguing extreme in their deliberately simplistic and arguably non-representational character. On the other side of the spectrum, there are models so complex and messy that they cannot be understood, do not even aim at explanation, and instead promise accurate prediction through (computer) simulation. Somewhere in between there are models coordinating several heterogeneous, even mutually incompatible theories in a pragmatically efficacious manner.

The ICI’s focus project ‘Models’ will reflect upon modelling practices within and across different fields while focussing on the role of reduction and its critical potentials.

When artists emulate other artists, for example, they do not copy works so much as they model the aesthetic principle that generated these works. This process is by necessity selective and introduces a divergence in the artists’ own artistic output. At least since modernity, the tendency has been not to correct but to amplify this divergence in work after work, thus differentiating an ‘individual style’. In view of histories of camp or drag and associated theories of performativity, one could ask whether oppressive norms based on gender or race can similarly be seen as the basis for divergence — even if they also highlight the precarity of such balancing acts between ironic distancing and affirmation, and raise the question of how subversive or transformative they can ultimately be.

The reductiveness of models facilitates their travelling across historical, disciplinary, cultural, and other boundaries. This may encourage homogenization but can also generate new perspectives, unpredictable transformations, and complex entanglements. One can trace the transfer of evolutionary and morphological models between biology and art history and could even extend this analysis to the recent discourse on memes. If abstract painting has long been discredited for its universalist claims, this stance has recently been re-evaluated with regard to black abstraction, queer abstraction, and Indigenous abstraction, which may be seen to counter the pressure to represent otherness within a still largely Western-dominated art world. Technical media such as the phonograph and the telephone provided models of the ‘psychic apparatus’ and its functioning that were foundational to psychoanalysis. Donna Haraway’s cyborgs and feminist science fictions demonstrate how models change with their transposition to different domains and transform these fields in turn, giving rise to new questions and approaches.

Such transformations are far from uncontroversial. The digital humanities, for example, have often claimed to do away with the humanities’ reductive focus on canons by importing computational models that allow for scaling up the analysis to encompass discourses in their entire breadth. Conversely, this purely quantitative ‘fix’ to the problem of reduction has in turn been criticized as the ultimate reductionism — one replacing theoretical models, aesthetic categories, and deliberate prioritizations with mere statistics, providing excessively detailed answers to simplistic questions. Eschewing such reciprocal accusations of reduction, one may rather highlight the wide spectrum of possible reductions and their specific affordances and generative potentials. From big data to the moves toward inclusivity encapsulated in terms such as LGBTQIA+, attempts at countering the reductiveness of models through an accumulation of details, cases, and data abound. Debates over theoretical turns or the notion of conceptual personae in philosophy may provide examples on the deliberately reduced side of the spectrum, while diffractive or intersectional models that coordinate the cooperation of different or even mutually exclusive approaches lie somewhere in the middle.

What is the critical purchase of models and how does their generative potential relate to their constitutive reduction? What are the stakes in decreasing or increasing, altering or proliferating the reductiveness of models? How can one work with and on models in a creative, productive manner without disavowing power asymmetries and their exclusionary or limiting effects?