Remediation, demediation and transmediation

Draft 25 October 2005, critique on Bolter & Grusin

3.1 Remediation, hypermediacy and immediacy

In what respect we see the new in terms of the old or the familiar had of course always been an urgent question for historians and futurologists, but with the emergence of media studies as an academic discipline the question is also pressing this field, especially regarding new media. And here McLuhan seems to be back in town, after a period of absence during the analytical hegemony of poststructuralism in the eighties of the 20th century.
The first attempts to design a general new media theory in the light of digitization are Remediation: Understanding new media by Jay David Bolter and Richard Grusin (2000) and The language of new media by Lev Manovich (2001). While Manovich can be seen as unconsciously performing McLuhan's rear-view mirror – Manovich's new media analysis is in fact an elaboration of film theory, firmly framed by Vertov's cinematic avant-garde piece Man with a movie camera ? Bolter and Grusin explicitly take McLuhan's rear-view mirror as the founding principle of media dynamics.
They call this the principle of remediation, the formal logic by which new media refashion and improve prior media forms, and prior media refashion themselves to answer the challenges of new media (ibid, 15). Remediation, according to Bolter and Grusin, always consists of two seemingly contradictory strategies, the strategy of transparent immediacy (making the medium invisible, suggesting an unmediated experience) and the strategy of hypermediacy (emphasizing the medium, celebrating or even exaggerating its presence). Remediation thus can accomplish an experience of immediacy by borrowing from media so familiar and taken up in our environment that we just don't see the medium anymore, just the 'message'. But the same movement of borrowing from other media can be very explicit and exuberant, thus enacting hypermediacy.
For instance, Massive Multi-player Online Role Playing Games (MMORPG's, such as World of Warcraft and Ultima Online) borrow modalities and genres from film, theatrical role playing and video games to evoke a sense of liveness and immediacy. The hypermediacy of the textual display of the names of the characters (never occurring in film, theatre or reality) does not spoil the sense of immediacy ? on the contrary, by this hypermediacy you recognize the other characters, once more enhancing the experience of immediacy.
A less spectacular example is ordinary chatting on the Internet. This mode of interaction borrows from typographic print, telex and mail but we don't see these media shimmering through, we experience the immediate presence of the other. Even in instances of hypermediacy beyond the strictly typographical, when pictural avatars or animated smileys are used, this does not spoil the immediacy.
In short, immediacy depends on hypermediacy and vice versa.
Though Bolter and Grusin's primary concern are visual representations, their claims about remediation, immediacy and hypermediacy are more general: 'Our culture wants both to multiply its media and to erase all traces of mediation: ideally, it wants to erase its media in the very act of multiplying them.' (Bolter & Grusin 2000, 5) The authors show with numerous examples how the history and development of all media is constituted by processes of remediation, yet they claim this is especially and more fundamentally the case in new digital media: 'remediation is a defining characteristic of new digital media'. (ibid 45)

3.2 Remediation and demediation

Bolter and Grusin did make a good point with their notion of remediation, their work can be seen as effectively elaborating and updating McLuhan's rear-view mirror for the age of the Internet. The principal iterability and modularity of digital mediation indeed invokes remediation after remediation, recycling juxtaposed and prior media forms, and challenging these media forms to accommodate themselves on their turn. Yet, Bolter and Grusin's vocabulary implies some problems which need to be addressed.
While their notion of 'remediation' focuses explicitly and grammatically on processes going on in mediation practices their definition of the two strategies involved as 'hypermediacy' and 'immediacy' narrows this down to states. After all, hypermediacy and immediacy are states, not processes. The authors do not talk of 'hypermediation' or 'immediation', they talk consequently about 'hypermediacy' and 'immediacy'. These terms tend to occlude the dynamic processes of remediation, the labor by designers, users and machinery from which immediacy and hypermediacy can occur as a result. I see immediacy and hypermediacy as the epistemological and psychological results of processes to be explained; not as the explanation themselves, as Bolter and Grusin are inclined to.
This is all the more strange, since the erasure of human agency and machinated agency is omnipresent as an undertow in Bolter's and Grusin's analyses. When the authors talk for instance about the linear perspective in painting as a form of remediation they note: 'The irony is that is was hard work to make the surface disappear in this fashion, and in fact the artist's success at effacing this process, and thereby himself, became for trained viewers a mark of his skill and therefore his presence.' (25) This intricate combination of hard work, depresenting this work, and representing the object and the maker in a specific configuration is what is at stake in any situation of remediation. Yet, by insisting on using ontologized terms as 'hypermediacy' and 'immediacy' the authors themselves contribute to the analytical disappearance of the hard work underlying these processes.
In other words, the consequent use of a grammar in terms of 'hypermediacy' and 'immediacy' is not neutral. In my opinion this is not just a grammatical coincidence. It reveals that Bolter's and Grusin's analyses are framed by ontologizing movements: reducing processes into states (or things, or places) and subsequently taking these states as fundamental constituting elements in an analysis. I argue that an analysis of mediation processes should give an explicit account of the remediating and ontologizing processes involved; it should not repeat and reinforce the icontology, as I called this before.11

So, while keeping the authors' basic notion of remediation as fundamental for any process of mediation, I propose a slight reformulation of the strategies involved. Mediation in my view always entails two strategies:
1.the strategy of remediation, that is representing, refashioning and re-using parts from other media forms. This strategy is aimed primarily at producing hypermediacy and icontology (by ontologizing the borrowed parts), and secondarily at producing immediacy-by-familiarity (parasiting on the taken-for-grantedness and familiarity of the re-used media forms)
2.the strategy of demediation, that is depresenting the machinery, the labor and the ontologizing acts involved in remediation, making invisible the underlying material mechanisms. This strategy is aimed primarily at producing immediacy-by-erasure.

This reformulation enables a more specific analysis of remediation processes. I extend the domain of hypermediacy with the concept of icontology, to indicate that the act of borrowing from other media domains does not just exaggerate or highlight the other media form, but also tends to ontologize the borrowed part in the new context (as happened with the mail and the telephone icon in the case of my friend). Further I distinguish between two kinds of immediacy, according to their different production mechanisms: one based on familiarity with the other media form (located in human cognition, memory and social convention) and one based on the deliberate labor of erasure (located in the material apparatus and the interface).
In this reformulation the involved strategies of mediation appear as processes of labor and transformation, and not as ontological states. In other words, it is framed by the discourse metaphor of the toolmakers paradigm instead of the conduit metaphor. In my view this enables a more specific analysis of what is going in the black boxes of mediation and media. The involved processes should be analyzed in terms of tools and labor (by humans and by machines), not in terms of transcendental transfer of ontologized states.

3.3 Two definitions of remediation

Yet, there is another problem with Bolter's and Grusin's conception of remediation. This is not a matter of grammaticality, but of ambivalence in definitions. The definition of remediation is in the first part of the book explicitly bound to challenging and transforming other 'media as such'. When the authors talk for example about making a movie based on a novel, they suggest this is in fact not real remediation: 'The content has been borrowed, but the medium has not been appropriated or quoted.' (44) They seem to suggest that this is a kind of pseudo-remediation, ubiquitous in history but not characteristic of new digital media: 'Again, we call the representation of one medium into another remediation, and we will argue that remediation is a defining characteristic of new media.' (45)
Though they acknowledge that forms of this pseudo-remediation can also occur in new media – 'Much of the World Wide Web also remediates older forms without challenging them' ? the authors hold that digital media enable a 'more aggressive remediation' which 'throws into relief both the source and the target media' (46). And this aggressive remediation seems to be the real remediation they are after.
But here we encounter some serious analytical problems. Firstly, as we have seen, it is impossible to distinguish between 'just the content' and a medium. There is no free floating content of ideas without tool or medium specific constraints, as we did learn from Reddy's toolmakers paradigm. There is no way to isolate the content from the medium, to represent content without wrapping.
Even when creating a movie from the storyline of a novel, the content will be adapted to enable the modal translation. And vice versa, the modality will be adapted with all kinds of tools and technologies available to transfer the specificities which come with the content. For instance, a first person voice in a novel can be translated in an auditive voice over or a subjective camera. Isn't that challenging and appropriating the medium, and thus remediation?
It might be that these adaptations are considered too small scale to transform and challenge the medium as a whole, but then the question arises: what scale is large enough to cover true remediation? And what timespan is granted to accomplish a transformative remediation? Creating a movie from the content of a novel does indeed not immediately and completely change the whole medium, but in the long run of history the narrative fiction movie, remediating the novel, developed as as dominant genre in cinema. Should we then say that somewhere in history the initial pseudo-remediation switched to real remediation?
These problems, stemming from a strict definition of remediation as total media transformation, can not be solved easily. And the set of such problems is endless. Is borrowing a visual representation of a mailbox to create a desktop icon for an email application 'aggressive' enough to be called real remediation? Which and how many borrowed parts of another medium are sufficient to invoke real remediation?

As said, these questions arise from the suggested distinction between total, aggressive remediation and partial pseudo-remediation. On top of that, the strict definition of remediation also implies a restriction of what can be remediated: only (complete) media, and by implication nothing but media. Nevertheless, the authors speak unproblematically about the remediation of social arrangements and material practices, and argue convincingly how remediation can also be done with reality, bodies, and selves ? yet things not commonly accepted as media. This suggests the opposite of the strict definition: remediation is not confined to media, but can be done with anything.
A possible reconciliation between the two definitions is to extend the notion of media so far that it can contain anything. The authors22 seem to take this course when they stipulate that 'the formal characteristics of media, their ?content?, and their economic and social functions [...] can never be entirely separated; a medium is a hybrid in Latour's sense.' (67) But when media are considered Latourian hybrids, mixtures of 'human subjects, language and the external world of things' (57) it is rather meaningless to restrict remediation to remediating 'media as such'.

So, what do we have by now? We can extract two definitions of remediation from Bolter and Grusin. One is a very strict definition: remediation is the representation of one medium into another, inducing a significant transformation of the source medium and the target media. The other definition is very broad: remediation is the recycling and re-use of anything at hand, which can be represented in a medium or as a medium.
How to solve this ambivalence? Assuming that remediation only works with prior existing media which should subsequently be challenged and transformed, can be qualified as problematic. The criterion of total transformation of a medium is not only almost infeasible in practice (little remediation would be left over…) it is also impossible in theory. As we have seen, the very concept of medium is a floating signifier with ontologizing tendencies, inappropriate as analytical category.33 The concept of remediation as a dynamic material-semiotic process was supposed to solve this conceptual problem, not to reinforce it. So, exit the strict definition.

3.3.1 Remediation and emergence

The broad definition offers better prospects. This conception of remediation is not based on problematic pre-existing origins, neither does it condition the scope or scale of the after effects. It just states that processes of remediation, re-appropriation and re-use are going on, reworking upon whatever material is at hand. And then something may happen at the level of media constellations. Or not. Or only partial, limited, on a very small scale. This open range of scales is a strong point of the broad conception. The scalability enables an analysis on whatever scale is appropriate, and it enables to connect small scale events with large scale transformations, and vice versa.
An objection might be that this conception is too broad, too open ended. It looks like anything goes. What could be clarified by such a free floating concept? I argue that this can direct us to the ontology of becoming we were looking for, especially the becoming of a new medium. The broad definition states in fact that remediation precedes media – media, ontologized or not, are the result of remediation, not the other way around. This remediation can be done with recognized media forms, with partial elements from media hybrids but also with things perceived as non-media. In short, anything can become a medium (or part of a medium), that's what remediation does: it turns things into media, it creates media.
This is even compliant with Bolter and Grusin's statement that 'a medium is that which remediates' (ibid. 65), since they include situations where it was not apparent that a device was a medium at all, until it started remediating other media. Only then could it emerge as a medium, as for example the computer did:

'As long as computers remained expensive and rare [...] their remediating functions were limited. In the 1970's, the first word processors appeared, and in the 1980's the first desktop computer. The computer could then become a medium because it could enter into the social and economic fabric of business culture and remediate the typewriter almost out of existence.' (ibid. 66)

The great advantage of such a broad conception is that it can analytically break out of what can be called the media prison, the mirror hall of existing media, where no external signal44 can come in, where the mirrors are just remediating and refracting each other. In this closed system there is only one way to give an account of the emergence of a new medium: newness emerges from recombinations of oldness. In fact, it is assumed in this model that all possible media must be already there, as virtual media, ready to be taken up by remediation.
It should be noted that this can be very productive; combinations and reconfigurations of existing stuff can indeed create fundamentally new phenomena. This kind of newness does indeed occur frequently in history. And perhaps this can even be analyzed as the dominant format by which new media emerge in history. The emergence of complexity on a higher level, induced by just combining simple existing elements on a lower level, is a well known mechanism in system theory, physics and biology. There is no reason why this should not be at work in the history of new media. Yet.
Yet there is also no reason why this should be the only way to create newness. Juxtaposed to the productiveness of recombinations in a closed system we can have the productiveness of recombinations with elements from the outside. Creativity often lies in thinking and linking out of the box, connecting with something not (yet) embedded in the system.
As we have seen, it is analytically quite problematic to delineate precisely a domain of media as a closed system. For that reason we might as well decide that the media domain is an open system, with permeable boundaries to cross. This makes it possible to capture forms of remediation of non-medial things and boundary objects on the border of medium and non-medium – remediations which would otherwise remain unseen and untheorized.

3.3.2 Remediation and representation

This remediation of non-medial and boundary objects seems to be especially at work in the digital domain. Perhaps this can even been seen as characteristic of digital media, at least of the Internet.55 After all, the Internet not only remediates classic media forms such as newspapers, television, radio and telex; it also remediates boundary objects,which can be situated somewhere between 'private media' and 'real life management'. Think for instance of address books, telephone calls, diaries, banking accounts, insurances, and to do lists. But what might be considered really new is that the Internet also remediates things usually not considered media at all, such as sex, friends' networks, local communities, and work.
It might be objected that calling all these different things 'remediation' undermines the distinctive power of the concept. What for instance is the difference between the remediation of sex by a webcam on the Internet, and pornographic magazines ? both are just a specific medium form with a specific content, representing something from reality. Remediation so conceived is in fact just the same as representation. And that is what any medium, old or new, digital or analog, always does.
I argue against this objection that, though remediation includes representation, these concepts should not be confused. Remediating sex, friendship or communities on the Internet can not be reduced to representation. It is not just 'about' sex, friendship or community – these things are actually enacted in the medium. And this remediated enactment challenges and transforms the practices and meanings of sex, friendship and community. That's what makes it remediation and not just representation.
While representation can be seen as projecting something from one domain (be it 'reality' or consciousness) into a another domain (usually called a medium); remediation is more like inserting it in a medium domain, thereby changing that domain as such. A change in what is represented in a medium may change the medium (as genre) but not necessarily. A change in what is remediated changes the medium, small scale or large scale.

3.4 Transmediation revisited

As has been noted before, the extension of the networks of electricity and digitality has enormously extended the amount and scale of the domains from which remediation can tap. No wonder we experience an explosion of media and mediatization; anything can be turned into a medium by inserting it in the network, this hybrid of machine and human (re)mediation.
As we have seen in the case of my friend, this networked situation enables transferences on several levels: between locations, machines, carriers, modalities, applications and media forms. I proposed to call these acts transmediations. These acts can be conducted by machines or by humans, but mostly they are enacted in collaboration between human users and machines, by intricate combinations of representation/depresentation, and remediation/demediation.
Yet, these kinds of material-semiotic collaboration (or conflict, or juxtaposition) can not adequately or completely be covered by the concepts of remediation and demediation. No matter how broad we define remediation, it covers primarily processes invoked by tools, machines and media apparatuses, with the spectators, audience, or public lagging behind rather passively. They experience the effects of remediation as the epistemological and psychological experience of hypermediacy and immediacy, but their role is more as re-actors, than as actors or interactors.
To emphasize the numerous acts of transference and translation going in between human and machine processing in contemporary networked media, I propose the term transmediation as a third concept to cover contemporary mediation mechanisms.
So, now we have a vocabulary consisting of remediation, demediation and transmediation.


  1. Though here no literal icons are involved, the movement is the same: taking things (remediating processes) at interface value (immediacy and hypermediacy). (↩ back to text ↩)
  2. Another possibility is that the two different definitions stem from two different authors. If so, then one of the authors is more into 'media as such' and the other more into 'hybrids'. (↩ back to text ↩)
  3. Though it is quite possible that an ontologized medium or part of a medium can be remediated productively, this is not what the narrow definition of remediation was supposed to include. (↩ back to text ↩)
  4. It is no coincidence that I slip into the more acoustic word 'signal', breaking out of the visual metaphorical frame of the mirror hall. Western philosophy, and especially new media philosophy, can be accused of having a visual bias, deaf ears for other perceptual modalities. (↩ back to text ↩)
  5. In that sense it is remarkable that Bolter and Grusin reserve their strict definition for digital media, and not the broad definition. (↩ back to text ↩)

Comments · ·

You must be logged in to post a comment.

Bad Behavior has blocked 1124 access attempts in the last 7 days.