For the official version of record, see here:
Sirui, M. (2025). The Recursive Equation of the Post-Medium: Rosalind Krauss’s Critical Trajectory and Its Internal Tensions. Media Theory, 9(2), 29–54. https://doi.org/10.70064/mt.v9i2.1270

The Recursive Equation of the Post-Medium: Rosalind Krauss’s Critical Trajectory and Its Internal Tensions

MA SIRUI
Fudan University, CHINA

Abstract

This article systematically reconstructs Rosalind Krauss’s theory of the post-medium by redefining the concept of medium through a logic of double negation and reordering her discourse around the triadic structure of “medium–support–convention.” First, the medium is refigured as hypokeimenon (support): it is neither reducible to material properties (a negation of modernism), nor is it a purely signifying form devoid of reference (a negation of post-structuralism). Rather, the medium is dynamically constituted through conventions and recursive structures. Second, drawing from Stanley Cavell’s notion of “automatism,” Krauss conceptualizes convention as “semi-automatic”: the artist simultaneously inhabits and exceeds conventional procedures, and this internal tension propels the renewal of the medium. Finally, the post-medium operates through recursion – self-examination and redefinition – continually reconstructing its specificity. In doing so, it rejects the linear teleology of modernism while avoiding poststructuralism’s erasure of direction and standards.

Keywords

Rosalind Krauss, post-medium, medium, modernism, post-structuralism

1. Introduction

In Reinventing the Medium (1999b: 296), Rosalind Krauss initially borrows Guattari’s concept of the “post-medium” to describe the media of the contemporary art world. In Postmodern Deadlock and Post-Media Transition (2009: 300), Guattari imbues this concept with a post-utopian vision. Briefly, in Guattari’s view, the post-media era enables technologies to foster new forms of community and democracy. Through the miniaturization and personalization of technological devices, mechanized media are re-singularized, shifting from a unidirectional, single-centered “universe” to a polycentric and perpetually mutating one. Krauss similarly situates this concept after high modernism and postmodernism, redefining media within the context of contemporary art.

Furthermore, in Krauss’s framework, the concept of the post-medium undergoes a far richer elaboration in the art world than in previous discourses. She not only uses it to reflect on modernism and its medium-specificity but is also interpreted by many scholars as re-examining the widespread acceptance of post-structuralism (or, more broadly, postmodernism in art criticism) over the past two decades. As she states in the preface to Under Blue Cup (2011: 1-8), after experiencing the blankness left by a brain tumor, words appeared in her cognition as signifiers without signifieds, compelling her to reclaim language. She also recognizes that postmodernism, compared to modernism, fundamentally rejects not only specific standards but the very notion of standards. Thus, some scholars argue that the introduction of the post-medium concept reflects her shifting attitude toward postmodernism and her return to the fundamental characteristics of modernism (Butler, 2020: 2). From Reinventing the Medium (1999) to essays such as ‘Specific “Objects”’ (2005), ‘The Rock: William Kentridge’s Drawing for Projection’ (2000), and ‘Two Moments from the Post-Medium Condition’ (2006) collected in Perpetual Inventory (2010), and later works like A Voyage on the North Sea: Art in the Age of the Post-Medium Condition (1999) and Under Blue Cup, Krauss gradually outlines her theory of the post-medium.

2. Medium as support: The definition of medium after high modernism

Before Krauss adopted the concept of the post-medium in the 1990s, two definitions of media confronted her in art history. The first, championed by Clement Greenberg (1992: 755), advocates a linear reductive path, ultimately establishing clear boundaries for each medium. Greenberg’s modernism can be summarized as a pre-established standard and a teleological development. The second definition stems from conceptualism, particularly Joseph Kosuth (1991: 13-32), who proposes a tautological path. In this view, the goal of art lies in continuous self-reference, and the standard behind the medium is the art itself.

By the time Krauss began writing about the post-medium in the 1990s, her post-structuralist turn was widely acknowledged in academia influenced by David Carrier (2002: 1-16).[i] The question then arises: in her approach to media, does Krauss continue to adopt a post-structuralist stance that rejects modernism? This paper will explore this question by examining her definition of media in detail.

2.1 The path of reduction: Linear reduction or tautology

After transitioning from Artforum to October and employing structuralist, phenomenological, and psychoanalytic methods extensively in her writing from the 1970s to the 1990s, Krauss gradually began to reflect on the potential limitations of the methodologies she had adopted. She observed that Greenberg’s notion of medium-specificity had not been entirely negated, particularly its emphasis on materiality – each art form, according to its specific medium, is unique. In restoring the identity of art, the clarity of its medium must be emphasized. In the visual arts, the medium is inherently material. Its flatness, the shape of its support, and the properties of its pigment all constitute its physical constraints. This, in effect, prescribes the material form specific to modernist media.

Greenberg’s reductive approach, as outlined above, aligns with the historical schema of the avant-garde as a myth. It can be summarized as follows: the a priori prescription of art serves as the premise, embedded within the broader history of modernist art and subjected to continuous reduction. In other words, art strives for the purity of its medium, a developmental trajectory that implies a historicist homogeneity.

However, Krauss found herself in a different intellectual context from that of the avant-garde. After writing Terminal Iron Works: The Structure of David Smith, she was influenced by the waves of French structuralism and post-structuralism in the 1960s and 1970s. She applied these methodologies in works such as Passages in Modern Sculpture (1977) and The Originality of the Avant-Garde and Other Modernist Myths (1985). For Krauss, just as Hegelian historicism was criticized as a myth, Greenbergian modernism, whether applied to individual artists or the avant-garde as a whole, constituted a myth within the artistic field (private myths, due to their reductive nature, became public myths). Even in his later years, faced with the failure of modernist medium-specificity, Greenberg continued to call for the return of aesthetic standards. However, in Krauss’s view, such aesthetic standards merely masked arbitrariness with a discourse of homogeneity.

Both Kant and Greenberg, in this regard, are aligned. Kant (2000: 177) argued that in society, “since the latter (taste) indulges inclination, although this may be ever so refined, it also gladly allows itself to blend in with all the inclinations and passions that achieve their greatest variety and highest level in society”. When faced with the possibility that personal inclinations might influence judgments of taste, Kant (2000: 123-124) cautiously responded, “since the principle, though only subjective, is nevertheless assumed to be subjectively universal (an idea necessary for everyone), which, as far as the unanimity of different judges is concerned, could demand universal assent just like an objective one – if only one were certain of having correctly subsumed under it”. Greenberg, on the other hand, situated Kant’s description of societal groups within the context of the middle class since modernism and French modernism, confining modernist aesthetic standards to the pursuit of the medium (or form) itself (1979: 156)[i], as exemplified by Abstract Expressionism.

As for Greenberg’s self-negating reduction of flatness, Krauss sees it as inherently Hegelian (Danto, 1997: 101-107).[ii] This reductive process first defines the domains of various media as absolute origins, rendering it inadequate to explain minimalism. If minimalism is seen as a contingent event, the determinacy of the medium as an internal necessity should predict its emergence and provide an a priori explanation. However, in reality, beyond the framework prescribed by Greenbergian modernism, numerous aesthetic productions are overlooked. Thus, from Krauss’s Foucauldian perspective, Greenberg’s approach is essentially a guise of reductionism that masks the exercise of power.

To counter the hegemony of media described above, since the 1960s, various movements and exhibitions have critically reflected on the material determinacy of media. These include the “Immaterial” exhibition curated by Jean-François Lyotard, the conceptual art of Joseph Kosuth, and the “Formless” exhibition organized by Rosalind Krauss and Yve-Alain Bois. These initiatives marked the beginning of institutional critique. As Hal Foster (1996: 4) noted during this period, the so-called institution refers to “the dominant modernist model – a formalism focused on specific media[…] refined by Clement Greenberg and Michael Fried for the New York School and its followers”.

The motivation for critiquing the institution lies in modernism’s persistent maintenance and reflection of media and its updates, which ultimately solidified into formalism. Once it transformed into formalism, it lost its explanatory power for the contemporary art world. Lucy Lippard (1997: vii) succinctly captured the spirit of this era, stating, “The 1960s ushered in a thirst for imagination, centered on attempts to escape what Robert Smithson called ‘cultural confinement’ – the sacred, immutable ivory tower and the heroic, patriarchal myths”.

Furthermore, in response to the reductive path of medium-specificity that culminates in materiality, Joseph Kosuth, as a prominent voice of opposition, critiques modernism’s approach as “morphology.” He argues against this reductive trajectory, advocating instead for a tautological method. Specifically, in Art After Philosophy, Kosuth begins by separating art from aesthetics, thereby opposing formalism, which he views as a fusion of the two. Formalism, in Kosuth’s view, pursues a pure aesthetic judgment – one that strips away the use value (or function, as Kosuth terms it) and content of an object, resulting in a disinterested, purposive purposelessness.[iii] This pursuit, Kosuth contends, inadvertently degenerates into “morphology,” which he defines as habitual aesthetic judgments formed through extensive perceptual experience.

To transcend the condition where art is reduced to mere form or language and to seek its essence, it is necessary to treat art as an analytical proposition – a tautology. Kosuth references A.J. Ayer’s critique (1952: 78) of Kant’s distinction between “analytic” and “synthetic” propositions: “A proposition is analytic when its validity depends solely on the definitions of the symbols it contains; it is synthetic when its validity is determined by empirical facts”. Following Kosuth’s argument, if art relies on the “morphology” of aesthetics, our judgment of what constitutes art would naturally depend on empirical facts. However, when we seek the essence of art, it necessarily becomes an analytic proposition. This is because, without relying on empirical facts, we conceive the essence of art a priori and then declare that what we have conceived is art. Art is art because we have conceived it as such a priori.

Kosuth’s work seems sufficiently clear, but when he reduces the direction of art to tautology, he leaves room for criticism. Although he claims to eliminate the morphological legacy left by formalism, he does not oppose formalism’s core tenets – a priori essence and self-criticism. In formalism, self-criticism based on an a priori essence reduces each medium to its pure form; for painting, this means flatness and medium purity. For Kosuth, the true reason for separating art and aesthetics lies in the danger of aesthetics or aesthetic judgment devolving into morphology in practice, rather than a problem with its inherent mission. Additionally, Kosuth opposes the phenomenon brought about by formalism – emphasizing form while neglecting content, and focusing on the medium while disregarding concept or meaning. However, upon closer examination of Kosuth’s own solution, while it no longer incorporates form and focuses solely on meaning, the reflexive reduction method of formalism is still evident. That is, the reduction from the entanglement of empirical facts (where form, in his view, is merely based on critics’ personal aesthetic preferences masquerading as subjectively universal judgments) to the essence of art, operating within a closed loop of “a priori”-“empirical” tautology.

Thus, it can be argued that Kosuth and Greenberg share the same fundamental framework of formalism or pure modernism. Although they respectively ground the medium in concept and concrete material form, both acknowledge the reduction of the medium and its underlying a priori prescription. In essence, their methods exhibit a teleological character within an identical framework. Consequently, although the two aforementioned approaches differ in their paths – the former focusing on the material specificity of each medium, while the latter emphasizes absolute concepts opposed to materiality or form – they both share the belief that art must be grounded in an aesthetic essence, predetermined a priorily, and reduced through a teleological framework. Moreover, these two approaches stand in direct opposition to each other, embodying a binary logic of affirmation and negation.

2.2 Beyond binary logic: Medium as hypokeimenon

In Krauss’s critique of Greenberg, she avoids falling into the trap of binary logic. Instead, she employs the concept of the “post-medium condition” to transcend the essentialism and teleology of her predecessors, a perspective that finds its roots in Derrida’s thought. As Bojana Matejić notes in Rozalind Kraus (2018: 69), “The post-medium condition of contemporary art aligns with the post-structuralist impulse of the 1960s and 1970s (as seen in Derrida and others)”. In A Voyage on the North Sea (1999a: 32), Krauss begins her critique of the essentialist definitions of media by engaging with post-structuralist ideas about the frame, pure vision, and the “metaphysical fictions” behind them.

Both Derrida and Krauss use the frame as a representative element of metanarrative or essentialism. Specifically, the frame has historically signified discipline. As Krauss argues in Photography’s Discursive Spaces (1986: 138), a work, “once framed, bordered, and labeled, enters the space of historical reconstruction through the museum. (Thus) on the exhibition wall[…] it is ‘legitimized’”. Regarding the museum’s exhibition wall, Krauss highlights the operational logic of this aesthetic discursive space – the continuous wall itself serves as the fertile ground for criticism, as it situates artworks within a curated context. At the same time, the wall implies selection and exclusion. Krauss (1986: 133) views the wall as a signifier of inclusion, embedding works within a broader symbolic system, which she terms exhibitionality.

In Derrida’s framework (1987: 53-61), the frame (parergon) of a painting serves to separate its internal space from the external world, enabling self-referentiality. This is achieved through the use of the copula “is” or through the physical demarcation that distinguishes the artwork from its surrounding space. However, deconstruction reveals that no medium can be purely for-itself or self-contained. Every medium is inevitably influenced by external forces and can only constitute itself through the projection of the external onto itself. Consequently, a medium can no longer rely solely on self-reference, self-reduction, or tautology for its definition or existence.

The autonomous space of meaning traditionally established by the picture frame is further deconstructed in Paper Machine (2005: 42), where Derrida redefines the frame, paper, and other media as a form of support (hypokeimenon).[i] In conventional aesthetics, these elements have been regarded as external or secondary to the image  –  strictly speaking, not part of the artwork itself. However, for Derrida, these elements are not only the material bearers of the text or image, but also the very conditions of possibility for the text or image to become a “work.” This corresponds precisely to his understanding of the supplement: the supplement is not an external addition to something already complete, but rather exposes the internal precondition of that thing. Furthermore, the “supplement” reveals that the internal is never wholly independent from the external, nor entirely self-referential. As Derrida argues in Of Grammatology (1976: 141-157), writing is not a supplement to speech, but reveals instead the fundamental incompleteness and non-self-sufficiency of speech itself.

This resonates with the original definition of hypokeimenon as found in Aristotle: hypokeimenon, originally a Greek philosophical term, derives from the verb Hypokeimai, meaning “to lie beneath” or “to be placed underneath.” In Aristotle’s Categories (1a20) (1975: 4), it is understood as the subject in a subject-predicate structure, which can be predicated but cannot itself serve as a predicate. It functions as the support that carries the attributes of the predicate. Aristotle also discusses hypokeimenon in Metaphysics (1029a10–25) (2001: 785), where he considers matter as what remains after all prescriptions are stripped away. However, this matter itself is not an individual thing. In other words, after removing all universal prescriptions, what remains of the subject is not an undifferentiated pure matter but an absolute “individual” that cannot project meaning – a proper name that signifies both the “this” of an individual thing and an absolute empty term of meaning (since it refers to nothing, this absolute individual must be placed in quotation marks). Conversely, even pure matter can still function as a predicate, indicating that its meaning has not been entirely stripped away.

Thus, in Derrida’s view (2005: 42-47), paper transcends the traditional definition of a medium. He argues that a medium is generally understood as “a direct means of communication.” In the Aristotelian sense, a medium is a being that can be fully predicated. However, paper surpasses this definition, leading Derrida to conceptualize it as a “support.”

To illustrate this, Derrida (2005:45) provides an example of experimental works in which traditional linear writing is disrupted. Multiple trajectories or discourses can coexist on the same sheet of paper, where letters may belong to multiple words simultaneously, thereby transcending the conventional unidirectional referential relationship. This disruption also undermines the transparency of the flat, akin to the unidirectional reflective function of a mirror. The work remains visible, but this visibility can no longer be directly signified. The word “TAIN” is superimposed on “écriT,” “écrAN,” and “écrIN,” signifying “writing, encapsulating, and filtering.” Although it remains visible, the paper as a medium does not prevent mirror-like reflection, yet the things beneath the surface – unreflective and invisible – persists. This phenomenon can be likened to the “archive,” as suggested by the analogy of “writing, encapsulating, and filtering.” Because the archive possesses invisibility (while also having visibility), any medium derived from its corpus can be reshaped at any time.

The invisibility concealed beneath visibility is precisely what Derrida refers to as the “abyss.” Paper, as a medium, simultaneously embodies visibility and invisibility. It is not merely a signifying object but a base or support. Thus, it exists both as a technological invention independent of humans and as a humanized body-subject, with the two entangled to constitute the medium of paper. A medium cannot be reduced to a direct referent that declares itself through its material determinacy. Instead, it encompasses a base/support that enables direct reference.

In other words, the medium is not defined as a material with a specific referent but as a proper name that remains after all direct references are stripped away – a name that signifies nothing but itself. The medium thus transcends binary logic, moving beyond the logic of being (or not being) a specific material that refers to a determinate meaning, and beyond the logic of being (or not being) a specific concept that operates through tautology. Further, influenced by Derrida’s view of the medium and the field-like body-subject, Krauss also conceptualizes the medium as a field-like support.

Thus, Krauss’s definition of the medium is doubly negating: it can both signify (viewing modernist linear reduction as the predication of specific forms – i.e., extracting features – and treating tautology as self-reference, both of which imply that the medium itself is a referential “object”) and remain non-referential (a proper name).

3. A redefinition of medium-specificity and its reductive trajectory

Following the above line of thought, since the definition of the medium has already transcended the essentialist prescription of a specific material and is instead conceptualized as a foundational support, the definition of the medium will further lead to the following paths:

The Medium as a Proper Name: If, in Derrida’s sense, reduction refers to the ability to distill attributes into predicates, then the proper name is what remains after reduction – an unsignifiable entity that cannot be clearly defined. Since it can no longer be reduced, can this new definition of the medium still be considered essentialist?

Moving Beyond Essentialism: If we no longer understand the medium through essentialism – that is, the two paths outlined at the beginning of this chapter – how can we further explain it? If we regard the medium as a base/support-like proper name, it becomes challenging to find concrete evidence for this in actual artistic production.

In Reinventing the Medium (1999b: 296), Krauss provides a clear definition of the medium: The medium is “a set of conventions derived from (but not identical with) the material conditions of a given technical support, conventions out of which to develop a form of expressiveness that can be both projective and mnemonic”. This paper will elaborate on these two statements in the following two subsections, providing a detailed analysis of their respective definitions of medium. These two statements, taken together, can be seen to effect a replacement of the traditional formulation of medium-specificity and its corresponding trajectory of renewal – that is, its reductive path. The following two subsections will address each of these statements in turn, providing a detailed exposition of this redefinition of the medium.

3.1 The replacement of medium-specificity: Convention and production conditions

It is first necessary to note the first sentence of the quoted passage, which offers Krauss’s further elaboration of the concept of medium as support – namely, convention.

In Reinventing the Medium (1999b: 292), Krauss takes photography as a paradigmatic case through which she shifts the notion of medium from traditional, essentialist specificity toward a non-essentialist framework grounded in conventions. Briefly put, starting from a reflection on Walter Benjamin, she notes Benjamin’s assertion that photography no longer adheres to a fixed medium-specificity. As Benjamin (1999: 219-242) wrote in The Work of Art in the Age of Mechanical Reproduction, photography departs from stable aesthetic conventions: from the shock effects sought by Futurist and Dada photomontage, to the optical unconscious, and finally to montage, photography persistently exceeds its previously determined specificity. Furthermore, due to its structural reliance on textual titles, photography stands at odds with aesthetic autonomy or medium-specific purity. This is what Krauss (1999b: 294) calls the “hybridity” or heterogeneity of photography – an inseparable mixture of image and text. In this sense, photography cannot constitute its ontological identity through its material properties or medium-specific traits. Accordingly, it may be understood as post-medium.

Krauss (1999b: 300) further exemplifies this shift through an analysis of a photographic work by James Coleman[i], illustrating how photography, having surpassed its specificity, raises new questions regarding its own definitional boundaries. She argues that Coleman discovered an artistic convention: when using a medium, one not only relies on its physical properties as a support but also imbues this materiality with expressive power. This transcends the two approaches outlined at the beginning of this chapter and merges them. Krauss refers to this convention as “double face-out”. On the first level, this literally describes a scene in a photo series where two individuals in conversation are not focused on each other but are instead both facing forward. This defies cinematic or narrative logic, as films typically use shot-reverse-shot editing to alternate between speakers, simulating the exchange and reaction between them. This effect is known as “suture” – which was first proposed and discussed in Jean-Pierre Oudart’s paper (1969: 35-47). Further, through the viewer’s identification with the shifting bodies in the cinematic space, the viewer is integrated or “sutured” into the film’s structure.

However, in his photographic works, Coleman rejects the effect of “suture,” thereby detaching the stills from their narrative context. He neither subordinates the stills to their narrative function nor attempts to impose a sense of transparency, despite their flatness as a medium. Although the stills are flat, he does not adhere to the conventional logic of planar media, which typically demands transparency or completeness. In this way, he dissolves the symbolic real space behind flatness. Yet, on another level, he still respects flatness by ensuring that the planar stills no longer function as an extension of cinematic language, nor do they draw the viewer into the filmic structure of speaker and reactor. In this sense, the medium he employs no longer follows the medium-specificity that a mass medium like still photography is presumed to possess a priori. For Krauss, this demonstrates that medium-specificity appears to be the result of conventions, a product of socio-cultural contexts. The same medium may exhibit different specificities depending on the given context or user.

Benjamin Buchloh employs the concept of “administration aesthetics”[ii] to describe the encroachment of highly developed cultural industries on art. As the cultural industry grows increasingly powerful in late capitalism, large multinational corporations and the art institutions they lead, as the administrators of artworks, inevitably seize control over the definition of art’s essence. While the emerging middle class is deprived of the right to engage in artistic thought within this mode of production, they willingly comply with this situation, as it does not undermine their existence within this system. The power to define the essence of art, through tautological means, allows the cultural industry to ideologically manage the definition of art’s essence.

Krauss’s reflection lies in the idea that medium-specificity is rooted in conventions – the cultural industry’s regulations for each mass medium, which stem from the professional divisions within the modernist configuration of knowledge and institutions. Specifically, medium-specificity in modernism arises from: 1. Institutional configurations: The discourse of education and knowledge distribution under modernism (the first academies established and essentialized the differences between artistic disciplines); and 2. Standard systems constrained by institutions, ideologies, and economics. As Krauss traces the cultural inscriptions behind the images of the eagle and the donkey in A Voyage on the North Sea (1999a: 9), she argues that these standardized configurations overlook the issue of inscribed meanings and the generation and deconstruction of conventions arising from the individuals, discourses, and cultural modes that use the support (television, film, shovels, bottles, chairs, etc.). In other words, the support structure is constituted by the “possibility or convention of expression” (Kim, 2009: 114-123).

Thus, medium-specificity is first and foremost tied to conventions, which in turn are connected to the broader conditions of production. Krauss argues that post-medium art does not focus on the specific technical properties of its medium, but every production operates within a “general condition” based on process. This refers to a diverse world encompassing “discourses, institutions, physical supports, and their technical implications – in other words, both linguistic and non-linguistic elements.” It functions like a machine, but not a technological one. Here, what is scrutinized is not the medium itself but the factors that lead to its individuation. Economic forces integrate recording and consumption into production, and artworks become evidence of the shaping process, “just as footprints in soft ground are evidence that someone has walked there” (Hinderer, 2009).

In summary, the medium cannot be defined solely by its (im)materiality but is constructed through the conventions altered by the individuals who use the support. This ensures the legitimacy of individualized artistic practices in new media and transmedia art, shifting our focus on the medium beyond its specific material determinacy. In Reinventing the Medium, Krauss thus views the medium as an apparatus or environment that induces sensibility, with new media continuously shaped by elucidating their materiality or other characteristics. The medium thus transitions from an essentialist specificity to a differential specificity.

3.2 The replacement of the reductive trajectory of medium: A new mechanism for medium renewal

After eliminating the essentialist definition of medium based on linear reduction and replacing it with the notion of support and its underlying conventions, one must pay attention to Krauss’s formulation of the medium: that although the medium is grounded in established conventions, it does not remain confined to the conventional, or, as she puts it, is not identical with convention. More specifically, if the first sentence in Krauss’s definition of medium suggests that the aim of reduction is no longer to recover the medium’s specificity, but rather to treat the medium itself as a support – one that necessitates a search for the conventions behind it – then what is being replaced is the reductive aim itself. Consequently, the original reductive trajectory must also be replaced – that is, the very mechanism by which the medium renews itself. Put differently, on one hand, Krauss needs to provide a more systematic and robust account for cases such as that of James Coleman, in which the individual, through working with traditional media, subverts established conventions. On the other hand, she (2004: 221-224) aims to integrate works by artists like Edward Ruscha[i], who employ media not traditionally included in conventions, into the “medium → support → convention” structure. Both scenarios lead to the same question: a medium not incorporated into conventions, lacking an essentialist foundation, risks derailment, making it difficult to find appropriate evaluative criteria within the existing context. In response, it is necessary to begin with a close reading of the second sentence in Krauss’s definition (1999b: 296): “conventions out of which to develop a form of expressiveness that can be both projective and mnemonic”.

It is generally assumed that projectability implies a relation to the external, whereas singularity suggests something entirely independent of the outside world – something that is self-constituting and self-referential as its origin. However, as explored in the previous section, Krauss’s conception of the medium as a support challenges this opposition. From this perspective, projectability and singularity are not mutually exclusive. As a support, the medium encompasses all the attributes that allow for predication; in this sense, these universal attributes relate it to the external world. At the same time, the medium also functions as a proper name that cannot be fully exhausted by predicates – thus harboring meanings that cannot be transparently projected from the outside. In this way, it retains its singularity. From an art historical viewpoint, this projectability refers to the fact that the medium carries certain traits that can be found in multiple artworks – traits that inform the discourse through which we describe works executed in that medium. Singularity, on the other hand, points to the insufficiency of those shared traits to account for the full meaning of a work that employs the same medium; or rather, it is precisely through the emergence of new modes of using the medium that new specificity comes into being. This seemingly paradoxical yet actually complementary relationship is further developed in Krauss’s interpretation of Stanley Cavell’s notion of automatism.

3.2.1 The redemptive capacity of the medium and automatism

This analysis positions the medium within its own developmental trajectory, thereby setting it apart from traditional media defined by singularity. This conceptual shift is further elucidated in Krauss’s other writings – most notably, in her articulation of the medium’s redemptive capacity.

In Reinventing the Medium (1999b: 296), Krauss once again revisits Walter Benjamin’s diagnosis of photography: in contrast to traditional media that rely on medium-specificity, photography’s specificity constantly becomes obsolete and renews itself. Although one may observe within its historical development a repeated return to prior conventions, this does not negate its singularity. For example, many of Coleman’s works employ media he did not invent, such as timer-controlled slide sequences originally used in early commercial advertising, akin to those displayed on station screens. Coleman returns to this outdated medium because he recognizes its paradoxical effect of simultaneously capturing motion and stillness. Roland Barthes, in The Third Meaning (1985: 41-62), proposes the cinematic quality of film, where the horizontal movement of film corresponds to the narrative’s symbolic drive, enabling the depiction of plot, history, and psychology. In contrast, photography, as anti-narrative, disrupts this drive, rendering it seemingly meaningless. However, this does not mean photography is static. While it resists film’s illusion of real-time motion, its meaning emerges from the residual domain or base of “diegetic horizon”.

For example, Coleman’s stills cannot be seen as mere fragments extracted from a narrative sequence, hinting at preceding or subsequent events. Instead, they are vertical texts that disrupt existing narrative structures, derailing established codes and reorienting them into new arrangements. Living and Presumed Dead (1983-1985)[i], a slide sequence capturing a 45-minute parade performance, concludes with a random still from the event. This still resists narrative interpretation, creating a permutational play within what appears to be a narrative sequence. This effect parallels Derrida’s “abyss,” as it appears projectable but no longer reflects narrative directly. Instead, it generates new meaning through the derailment or permutation of codes. Thus, the still is static, anti-narrative, and non-cinematic, yet it is not meaningless.

For film, a medium that evolved from photography, adhering to the established path of medium-specificity would mean that film distinguishes itself through motion, utilizing its medium to articulate this specificity. However, as demonstrated above, this specificity is not exclusive to film. Photography, particularly stills, operates at the edges or folds of the diegetic horizon, generating meaning through the derailment of codes. Thus, media can be revisited or revived because they retain the capacity to renew existing media. Outdated media possess a redemptive quality (1999b: 296), reshaping themselves while also validating contemporary media.

The “redemptive capacity” of the medium manifests in the reinvention of current media through obsolete ones, thereby addressing the question of purpose – or meaning – after the teleology of prior essentialist reduction has been displaced. However, Krauss’s mechanism of medium renewal still faces a central challenge: within the transmission of convention, where does the impetus for medium-specific innovation arise?

In order to address this question, Krauss draws upon Stanley Cavell’s notion of “automatism” – let us first return to the context in which Cavell introduces this concept. Cavell (1979b: 101-108) argues that once established forms are no longer sufficient to ensure the communication of universal meaning, modernist masters must create new media or automatisms within their respective media to secure their works’ status as art. In short, for Cavell, automatism replaces the medium. Unlike modernist narratives, automatism encompasses both the formal traditions of art and the local events or themes that emerge around them. However, this does not stem from artistic autonomy but from existing traditions or conventions. Only by engaging with the series of automatic mechanisms that sustain tradition can artists ensure the meaning of their works.

While Cavell’s account already negotiates a balance between the inheritance of convention and the imperative of innovation, Krauss supplements this by introducing another concept to elucidate more precisely the generative force behind her own model of medium renewal. This can be seen in her analysis of William Kentridge’s artistic production process in The Rock (2010: 55-56). Kentridge’s Drawings for Projection (1989-2020)[ii] are a form of animation. Unlike traditional animation, where each frame depicts a unique pose and the frames are later synthesized during recording, Kentridge modifies a single drawing. By adding and erasing, he alters the image by a few millimeters each time, exposing it as a single frame. This method allows him to produce an 8-minute film by modifying only 20 drawings. However, this production process requires meticulous adjustments to individual images, forcing him to endlessly move between the camera on one side of the studio and the drawing on the other. Additionally, Kentridge does not rely on storyboards, overall recording plans, or film producers, as is typical in traditional animation.

Thus, Krauss identifies a new approach to the medium of animation in Kentridge’s work, which she terms Fortuna. Krauss describes this as a semi-automated phenomenon. In his production process, Kentridge engages in a dialogue with past conventions through analogy. Specifically, Krauss argues that Kentridge’s creative process resembles how we randomly construct sentences, which are largely influenced by the habits of drawing from a corpus and learned discursive patterns. In this sense, Kentridge’s actions follow an automatic mechanism. At the same time, however, automatism also carries the Surrealist connotation of “automatic writing” – the emergence of the unexpected from the unconscious, representing the liberation of spontaneity. This is akin to a sentence taking an unforeseen turn. In Krauss’s interpretation of Kentridge (1998: 52-57), Fortuna occupies an intermediate state, neither entirely pre-programmed nor purely accidental. Therefore, both automatic conventions and spontaneity can simultaneously describe automatism.

However, from the examples above – Coleman’s stills and Kentridge’s animations – it is evident that these artists have created unprecedented modes of aesthetic production. Although they use existing media, they do not (or do not entirely) adhere to established medium-specificity. This has sparked academic debate, primarily centered on the concept of automatism and its interpretation, which does not clearly explain how individual artistic production can transform into new automatisms. For instance, artists like Ruscha or Coleman, as discussed by Krauss, create or reshape media based on their personal reflections on past conventions. How, then, can such individual acts attain universality and be understood by others? In other words, the concept of automatism does not clearly elucidate how conventions are updated or how a production that deviates from past conventions can be integrated into the aesthetic automatism.

In Automat, Automatic, Automatism: Rosalind Krauss and Stanley Cavell on Photography and the Photographically Dependent Arts (2012: 819-854), Diarmuid Costello questions Krauss’s interpretation of Cavell. Although Cavell (1979b: 103-107) has stated that the challenge for modernist artists lies in discovering or creating a new medium or automatism, he emphasizes that all acts of reinventing media are conditional. For example, Pollock did not invent a new medium but expanded its possibilities, and before him, no one considered such possibilities valid. Before the emergence of individual artists, media gradually develop, containing numerous possibilities and generating a series of expectations and norms that must either be adhered to or defied. In Costello’s view, Krauss fails to recognize these norms. The cases she presents suggest that creating is out of nothing (ex nihilo), without any prior conventions.

Costello argues that, unlike Cavell’s view, where artists must extend and transform tradition by finding new ways to utilize existing sources, Krauss advocates for artists to completely shift to new forms of support, generating new meanings and sources – akin to inventing a new language. However, no language, not even Esperanto, is invented from scratch. As a medium for conveying meaning, language evolves through sedimentation, translation, and fusion. Therefore, for artistic media to be understood and judged, they must first be subject to public constraints derived from past media usage, or conform to established “Syntax.”

It becomes clear that Costello may have misunderstood Krauss. Krauss’s notion of “creation” or “reshaping” aligns with Cavell’s, as it operates within the framework of acknowledging established conventions. Her use of the concept of Fortuna encapsulates this. However, Costello interprets the other aspect of Fortuna – unconscious spontaneity – as a private language, reducing Krauss’s concept of “creation” to this single dimension and thus oversimplifying it.

In Rosalind Krauss: Between Modernism and Postmedium (2020: 1-21), Rex Butler offers a similar perspective. He argues that forms that have never appeared in the history of media usage, or even the creation of entirely new media, are entirely legitimate. Artists who “reshape” or “create” media are not detached from art history and its value systems. Moreover, because art history inherently lacks certain traits, their media are necessarily unique. This is exemplified by Coleman and Kentridge, who each extract one or a cluster of traits from the conventional use of film and celluloid, building upon them for spontaneous creation. This does not contradict their affirmation of film as a medium, nor does it prevent us from including their works in discussions of film and evaluating them accordingly. In other words, although Coleman and Kentridge share no obvious commonalities, the possibility of drawing comparative conclusions lies precisely in their shared underlying principles.

Furthermore, in Butler’s view, Costello’s critique of Krauss remains unresolved, as it seems to implicitly affirm a Greenbergian linear reductionism – that media evolve along a single, continuous path toward pure specificity. Under this premise, any artist deviating from this path would inevitably be dismissed. However, in the cases of Coleman and Kentridge, each has left behind distinct media traits, both of which fall within the scope of our discussion.

Thus, the mechanism of medium renewal has been redefined. By employing the concepts of “automatism” and Fortuna, she integrates two opposing elements: the static, established norms of the past and the dynamic, fully spontaneous creation of the new. This structure addresses the initial problem posed in this section: a medium that diverges from past conventions does not render its evaluative criteria or discourse invalid. On the contrary, this divergence is precisely what drives the renewal of medium-specificity. The medium no longer requires an a priori prescription to predetermine its path of renewal. Instead, it relies on the artist’s spontaneity, which operates within the unavoidable framework of conventions, to renew itself. In other words, the specificity of the original medium is continually updated through the examination of the post-medium condition. Krauss further elaborates this structure of media renewal as a recursive one, which will be discussed in detail below.

3.2.2 The legitimacy of the post-medium: A recursive model

Building on the previous two sections, which address Krauss’s transcendence of traditional medium-specificity and her redefinition of its reductive trajectory, her formulation of the post-medium emerges as a necessary theoretical gesture. As Butler (2020: 14) has observed, this underscores the true reason Krauss proposed the post-medium condition. In the contemporary era, a singular, totalizing discourse is insufficient to encapsulate a media ecology characterized by lateral comparisons and evaluations – or, as he puts it, there is no longer a “general medium”.

Thus, Krauss moves toward non-essentialism, viewing aesthetics or art as a form of production and tracing its conditions of production as the foundation of aesthetics. Medium-specificity is no longer sought through inward reduction but is instead replaced by differences in production conditions, which substitute for differences in artistic essence. She replaces linear reduction with a structuralist framework that transcends binary logic. As in Under Blue Cup (2011: 13-16), Krauss draws on Roland Barthes’ theory, arguing that meaning should be the sum of two opposing semes. Similarly, the medium should not be defined as material or immaterial, logical or illogical, but as the foundational possibility established by both logic and materiality. Therefore, negating the materiality of the medium does not negate its specificity.

Even though Krauss uses convention – a non-essentialist approach – to ground the support (i.e., the medium), she does not deny medium-specificity itself. For instance, in the case of Coleman, she aims to demonstrate how, while adhering to medium-specificity, one can reflect on the boundaries of conventional media usage. Thus, it can be said that her negation of past medium-specificity is ambiguous.

Here, it is necessary to clarify that the essentialism Krauss opposes, as defined by Derrida, implies that a subject can be fully predicated; for Barthes, it refers to directly referential symbolic codes; and for Greenbergian modernism, it entails linear reduction, an essentialist prescription of media based on universally communicable judgments of taste. However, Krauss’s lack of opposition to medium-specificity does not mean she is essentialist. As seen in the etymology of “support” (hypokeimenon), which she uses to describe the medium, a foundational base does not deny the existence of predicable attributes. However, after predication, there remains a residual meaning. This residual meaning, as Derrida describes it, is an “abyss” – something that cannot be directly signified or, in Greenbergian terms, preordained. Furthermore, the essentialist undertones of convention are also stripped away, as Krauss replaces the essentialist definition of the medium with the concept of spontaneity, which cannot be universalized. Thus, within the structure of “medium-support-convention,” essentialism is entirely eradicated.

However, the rejection of essentialism does not entail the rejection of legitimacy. For modernists, this is an inevitable outcome. Greenberg, in his later years, noted that the loss of aesthetic standards beginning in the 1960s signaled the impossibility of writing a new art history. Although Krauss opposes Greenberg’s essentialism, by the 21^st century – after her engagement with deconstruction and during the writing of Under Blue Cup – she undergoes a detour, seeking to reestablish a standard.

However, the principle behind this standard is not a Greenbergian linear reduction. Instead, Krauss revisits the concept of “double negation,” derived from the Klein group, which she first employed in Sculpture in the Expanded Field (1979: 30-44). In the 1970s, she used this concept to address the limitations of modernist methods in explaining minimalist sculpture. Unlike modernist sculpture, which maintains a dichotomy with real space through its pedestal, minimalist sculpture moves toward a neutral field – a double negation that is neither landscape (nature) nor architecture (culture). By the 21^st century, Krauss applies this concept to address the legitimacy of the post-medium condition. The standard she establishes is one of double negation. As Barthes suggests, meaning should be the sum of two opposing semes, and Krauss (2011: 67) adopts a recursive structure to articulate this.

In terms of the dual implications of automatism discussed earlier, the post-medium and traditional medium (i.e., convention) form a mutually validating field of opposing terms – or, as Krauss puts it, a recursion. Recursion, in mathematical terms, involves embedding the output of a function back into the function itself. While mathematically, recursion is strictly limited to finite cycles, Krauss extends this concept to a universal effect. Thus, in Krauss’s framework, recursion refers to the post-medium and traditional medium mutually validating themselves, akin to comparing a function’s output to its original input.

After her brain tumor, Krauss reexamined memory and language, driven by an urgent need to reconnect with the world. She sought to reclaim memory as an inscription of time and language as the means to describe it. Together, these constitute convention in art history. However, contemporary art’s transmedial and immaterial tendencies seem to discard convention, plunging into a void of meaninglessness and linguistic absence. This crisis, akin to the blankness she experienced during her illness, leads her to understand Greenberg’s concern: a complete detachment from convention results in blindness.

In summary, Krauss rejects the linear reduction of media. Her structuralist model – the recursion of the post-medium and medium – represents a non-essentialist path in which the post-medium and medium mutually validate themselves through negation. Further, the post-medium, as a double negation of the medium, affirms the specificity of prior media definitions (thus validating the original definition) while negating the linear reductionism used to interrogate that specificity (thus invalidating the original definition). The necessity of Krauss’s recursion lies in its doubly negating nature: media possess a validating structure that continually tests the boundaries of their specificity, yet they do not validate themselves from within but rather through their opposites. This makes it both modernist and non-modernist. At the same time, it neither prescribes a teleological path (thus aligning with post-structuralism) nor denies the existence of a validating framework (thus resisting post-structuralism).

4. Conclusion

In summary, Krauss opposes the rigid definition of media – viewing medium-specificity as either the result of linear reduction or as a conceptual tautology. She no longer confines the development of media to a predetermined teleological path dictated by a specific category or artistic standard. At the same time, she rejects the post-structuralist dismissal of conventions or standards, which leads to a stagnation in aesthetic evaluation. Instead, she redefines and expands the concept of media.

Krauss understands media as a support, which is first and foremost doubly negating. It transcends binary logic by encompassing two opposing terms: it can be directly referred to (whether through self-referential tautology or predication, both of which imply that media can be designated or that specific attributes and specificities can be extracted from it), yet it can also remain non-referential.

Secondly, within the post-medium definition, the essentialist framework behind the former is not grounded in an artistic essence but in convention. She interprets convention as automatism or Fortuna, concepts that are also doubly negating: they encompass both the fixed, established history of media and allow for entirely different creations or spontaneous acts arising from nothingness.

Finally, Krauss seeks to move away from writing the history of contemporary art and its media in a totalizing or teleological path. Instead, she proposes a recursive structure, which remains doubly negating: media possess a validating structure that continually tests and updates the boundaries of their specificity, yet they do not validate themselves from within but rather through their opposites. This is the direction or standard Krauss hopes to establish for art history after abandoning postmodernism – a standard that is neither modernist in its presuppositions nor a priori.

Thus, it can ultimately be said that Krauss’s media theory is doubly negating. It can be seen as modernist (providing a framework that legitimizes the renewal of medium-specificity) and post-structuralist (rejecting a priori determinism and teleological paths). At the same time, it is neither modernist (as it rejects self-validation through linear reduction) nor post-structuralist (as it affirms the necessity of adhering to established media standards).

References

Aristotle (1975) Categories and De Interpretatione, J. L. Ackrill (trans.) Oxford: Oxford University Press.

Aristotle (2001) The Basic Works of Aristotle, R. McKeon (ed.) New York: The Modern Library.

Ayer, A.J. (1952) Language, Truth, and Logic. New York: Dover.

Barthes, R. (1985) ‘The Third Meaning: Research Notes on Several Eisenstein Stills’, in R. Howard (trans.) The Responsibility of Forms: Critical Essays on Music, Art, and Representation. New York: Hill and Wang, pp. 41-62.

Benjamin, W. (1999) ‘The Work of Art in the Age of Mechanical Reproduction,’ in H. Arendt (ed.), H. Zohn (trans.) Illuminations. New York: Schocken Books, pp. 217-251.

Butler, R. (2020) ‘Rosalind Krauss: Between Modernism and Post-Medium’, Journal of Art Historiography, 23(21): 1-21.

Carrier, D. (2002) Rosalind Krauss and American Philosophical Art Criticism: From Formalism to Beyond Postmodernism. Westport: Praeger.

Cavell, S. (1979a) The Claim of Reason: Wittgenstein, Skepticism, Morality, and Tragedy. New York: Oxford University Press.

Cavell, S. (1979b) ‘‘Automatism’ and ‘Excursus’: Some Modernist Painting’, in The World Viewed: Reflections on the Ontology of Film. Cambridge: Harvard University Press, pp. 101-108.

Costello, D. (2012) ‘Automat, Automatic, Automatism: Rosalind Krauss and Stanley Cavell on Photography and the Photographically Dependent Arts’, Critical Inquiry, 38(4): 819-854.

Danto, A.C. (1997) After the End of Art: Contemporary Art and the Pale of History. Princeton: Princeton University Press.

Derrida, J. (1976) Of Grammatology, G. Spivak (trans.) Baltimore: Johns Hopkins University Press.

Derrida, J. (1987) The Truth in Painting, G. Bennington and I. McLeod (trans.) Chicago: University of Chicago Press.

Derrida, J. (2005) Paper Machine, R. Bowlby (trans.) Stanford: Stanford University Press.

Foster, H. (1996) The Return of the Real: The Avant-Garde at the End of the Century. Cambridge: MIT Press.

Greenberg, C. (1992) ‘Modernist Painting’, in C. Harrison and P. Wood (eds.) Art in Theory 1900–1990: An Anthology of Changing Ideas. Oxford: Blackwell, pp. 754-760.

Guattari, F. (2009) ‘Postmodern Deadlock and Post-Media Transition’, in Soft Subversions: Texts and Interviews 1977–1985. Los Angeles: Semiotext(e), pp. 291-300.

Hinderer, M. (2009) ‘Unwanted Positivism: On the Term ‘Post-Medium Condition’ in the Work of Rosalind Krauss’, Springerin, 1. Available at: https://www.springerin.‌at/‌en/2009/1/ungewollter-positivismus/ (Accessed: 10 July 2025).

Kant, I. (2000) Critique of the Power of Judgment, P. Guyer and E. Matthews (trans.) Cambridge: Cambridge University Press.

Kentridge, W. (1998) ‘Fortuna: Neither Programme nor Chance in the Making of Images’, De Arte, 33(57): 52-57.

Kim, J. (2009) ‘The Post-Medium Condition and the Explosion of Cinema’, Screen, 50(1): 114-123.

Kosuth, J. (1991) ‘Art After Philosophy’, in G. Guercio (ed.) Art After Philosophy and After: Collected Writings, 1966–1990. Cambridge: MIT Press, pp. 13-32.

Krauss, R.E. (1979) ‘Sculpture in the Expanded Field’, October, 8, Spring, 30–44.

Krauss, R.E. (1986) The Originality of the Avant-Garde and Other Modernist Myths. Cambridge: MIT Press.

Krauss, R.E. (1999a) A Voyage on the North Sea: Art in the Age of the Post-Medium Condition. New York: Thames & Hudson.

Krauss, R.E. (1999b) ‘Reinventing the Medium’, Critical Inquiry, 25(2): 289–305.

Krauss, R.E. (2004) ‘Specific ‘Objects’’, RES: Anthropology and Aesthetics, 46(1): 221–224.

Krauss, R.E. (2010) ‘‘The Rock’: William Kentridge’s Drawings for Projection’, in Perpetual Inventory. Cambridge: MIT Press, pp. 55-88.

Krauss, R.E. (2011) Under Blue Cup. Cambridge: MIT Press.

Lippard, L.R. (1997) Six Years: The Dematerialization of the Art Object from 1966 to 1972. Berkeley: University of California Press.

Matejić, B. (2018) Rozalind Kraus. Beograd: Orion Art.

Oudart, J.P. (1969) ‘Cinema and Suture’, Screen 18(4): 35–47.

Notes

[1] David Carrier was among the first to address this shift. In Rosalind Krauss and American Philosophical Art Criticism (2002), he argues for a periodization of Krauss’s thought.
[2] Greenberg’s emphasis on “form” and formalism stemmed from his aversion to expressionism. The formalism he supported (which he equated with modernism) originated from his admiration for Cézanne and the Francocentric tradition of the Renaissance, which he regarded as the primal source of all excellent modernist art. He believed that any modern art that could not trace its influence back to them was inferior. As Kuspit observed, Van Gogh, Soutine, and all other expressionists – who, according to Greenberg, sacrificed the medium of their work for excessive life content – were condemned by him for lacking a Cubist foundation. The subject of Cubist art, after all, was nothing other than composition itself. In contrast, Greenberg argued that expressionist art served emotion or sensation, a subject unrelated to art, and was thus an impurity.[3] As Arthur Danto said, the exclusivity of Greenberg’s modernist narrative is problematic. Danto argues that this narrative, while developmental, is too narrow: it derives definitions from specific works and thus cannot accommodate a sufficiently universal definition of art. Later, Danto also critiques Krauss’s inability to address works suited for psychoanalytic interpretation within Greenberg’s framework.
[4] Kant’s four moments of aesthetic judgment are characterized by disinterest, subjective universality without concepts, purposiveness without purpose, and a common sense. These moments elevate the individual to the universal, precede conceptual understanding, and differ from cognitive judgments, producing a universally pleasurable judgment. However, in practice, Kosuth argues that such aesthetic judgments, which exclude value and external attributes, inevitably become influenced by external factors. Consequently, aesthetics becomes entirely biased toward external attributes, which is why he rejects art based on empirical facts, as it leads to aesthetic judgments deviating from their original intent and eroding the essence of art.
[5] In the English translation of Derrida’s Paper Machine, hypokeimenon is equated with “support.”
[6] This refers to James Coleman’s Seeing Oneself (1987-1988).
[7] The author suggests that “aesthetics as administration” might be a more accurate translation. Adorno’s theory of “executive administration” is relevant here, as “aesthetics as administration” refers to the ideological management of the masses through the aesthetic norms imposed by those who dominate the art market.
[8] For example, Standard Station (1966); see https://www.moma.org/collection/works/‌76637 (Accessed‌‌:‌‌‌ 5 December 2025).
[9] See https://www.mariangoodman.com/artists/36-james-coleman/works/43652-james-coleman-liv‌ing‌-and-presumed-dead/ (Accessed: 5 December 2025).
[10] See https://www.kentridge.studio/william-kentridge-projects/drawings-for-projection/ (Accessed: 5 December 2025).

Ma Sirui is a PhD student in the Department of the Philosophy of Art at Fudan University. His research focuses on the art criticism of the American October School and French theory.

Email: msrartphi@163.com

Conflicts of interest
None declared

Funding
None declared

Article history
Article submitted: 01/05/2025
Date of original decision: 12/06/2025
Revised article submitted: 15/07/2025
Article accepted: 03/08/2025