Accueil
https://en.wikipedia.org/wiki/Yuk_Hui
"Matter is considered [by moderns] to be something to be mastered, dominated and controlled." (note 1 p.132)
"In contrast to the vague demand of the memory of the machine, as a component
of the digital cosmology advocated by philosophers such as Fredkin and Chaitin,
media theories argue more precisely for a ‘digital materiality’. While being
different from the digital philosophers’ approaches, these theorists tend not to
differentiate the ‘immateriality’ of the digital and the materiality of its support,
hence its condition of being material becomes synonym of its materiality.
Among these proposals, I find Matthew Kirschenbaum’s analysis deserving
of careful consideration. Instead of claiming that certain matter is the foundation of the digital universe, he starts from the analysis of technical objects;
like peeling an onion, he starts by asking what their constituents are. In his
book Mechanisms − New Media and the Forensic Imagination, Kirschenbaum
(2008) describes two types of materiality. One he calls forensic materiality and
the other formal materiality. With forensic materiality, he refers to the method
of analysing traces in a computer, going beyond what is visible on the screen.
He takes the disk image of the online game “Mystery House” as an example,
and demonstrates what he calls a “forensic walk through”. Kirschenbaum shows
the hexadecimal and ASCII representations of data inside the file with the open
source software FishWings and analyses the structure of the disk from track to
track and sector to sector. In doing so, he finds traces that are not visible on the
screen, for example the disk image also contains remnants of Bob Bishop’s Dung
Beetles game and a ground-to-air shooter Blitzkrieg. Kirschenbaum concludes
that the Mystery_house.dsk:
“becomes a multivalent forensic environment, one where all of these different levels of
engagement – player, pirate/cracker, postmortem investigator – find their correspondences in the multiple layers of textual events that both drive the game as code and are explicitly thematized within its forensically charged spaces.” (Kirschenbaum 2008: 109)
The researcher who performs this forensic analysis, is like a detective who
examines the traces of a criminal scene, “every contact leaves traces” as he
claimed in talk given ten years ago in the History of Material Texts workshop
at the University of Pennsylvania (Kirschenbaum 2005). With the notion of
‘formal materiality’ he refers to the “normative condition of working in a digital
environment” (ibid.). He suggests speaking of two types of formal materiality.
The first type is the explicit form of digital writing. Kirschenbaum compares
this with Nelson Goodman’s notion of allographic objects; like written text,
they fulfil “their ontology in reproduction” (2008: 133). In contrast to the allographic object stands the autographic object; similarly to a painting, its meaning
cannot be explicitly repeated. The difference between Goodman’s example of
the written text and digital writing is that digital writing has the “state of the art error detection and correction” which allow it to sustain an ideal allographic
environment.
The second type of formal materiality, according to Kirschenbaum, can
be seen in standardised formats such as JPEG, MPEG and predefined logic of
the application. In the footnote of the introduction, Kirschenbaum (2008: 9)
refers to Johanna Drucker’s definition of materiality in The Visible Word: Experimental Typography and Modern Art, 1909–1923. For Drucker, materiality is
considered as “two major intertwined strands: that of a relational, insubstantial, and non-transcendent difference and that of a phenomenological, apprehendable, immanent substance” (ibid.). In other words, materiality is defined
by the dialectics of the objective reality and the subjective experience (varying
from subjective to intersubjective). Kirschenbaum (ibid.) however, criticises that
Katherine Hayles and Drucker’s notion of materiality is only limited to what he
calls ‘forensic materiality’.
Even though Kirschenbaum’s analytic method of layering is very relevant
and the methodology of grounding his theory in objects is plausible, his critique
of Drucker and Hayles’ concept of materiality still has to be reconsidered. The
feminist theorists attempted to extend the notion of materiality from technical
objects to embodiment, which exceeds the material scheme of technical objects
upon the point of their invention (Hayles 1999). Kirschenbaum’s approach rather
emerged from the tradition of textual studies. In contrast, Drucker and Hayles
have suggested a larger scope in order to look at the question of materiality.
While Kirschenbaum is probably right to say that Hayles misses “the computationally specific phenomenon of formal materiality, the simulation or modelling
of materiality via programmed software processes” (2008: 9), a similar critique
cannot be applied to Drucker’s analysis of typography, since ‘formal materiality’
is exactly the condition of any phenomenological experience.
We can also observe that what characterises Kirschenbaum’s analysis is a
regression from the abstract concept of the digital to the endless layers of concrete
matters. This can be read in great contrast to digital physics. Rather than starting
from the One, Kirschenbaum starts with the Many. Unlike digital physics or
digital philosophy which addresses the composition of bits, Kirschenbaum’s
approach seeks a material base on which the digital can be recognized and
embedded, through the decomposition of objects. ‘Regression’ however does not
imply that this is a ‘bad’ development; instead, it indirectly avoids addressing the
digital by addressing its conditions. For example, one can go from one condition
to an outer condition layer by layer, and finally one will end up at the level of
silicones, and probably also further to the sources of energy, etc. However, the
speculative question remains, until when and to which level shall the forensic
operation proceed?
Furthermore, one will find that this analysis can actually be effectively
applied to any type of technical object, and does not necessarily contribute to the
clarification of the digital. Gilbert Simondon mentions a similar insight: with
regards to the manufacture of a needle in Great Britain, he writes that, without
exaggeration, the quality of the needle expresses the degree of perfection of the
nation’s industry (Simondon 2012: 90). However, in the thought of Simondon, there are two important concepts that lead to a third inquiry into digital materiality. The first concept is relation; the second concept is concretisation, which
effectively sublates the difference between the digital and the support of the
digital, which characterise the above approaches. As a point of departure from
digital physics and digital textuality, I would like to outline a third and progressive approach of analysis based on a particular reading of Simondon, which I
am tempted to call a ‘relational materialism’. The last example on XML that
Kirschenbaum gave in order to explain formal materiality would be a perfect
example to understand relational materiality." (pp.135-137)
"While we can see that the two abovementioned approaches either start with or end
up with substance, this approach is an attempt to move away from substance to
relations. Or more precisely, the aim of such relational materiality is to overcome
the hylomorphism proposed by Aristotle, and to see how the development of
technicity distances itself from this analysis.
Hylomorphism is a substantialist thinking: in this context, being can be
comprehended in terms of matter and form. The problem of the substantialist
view is that it limits the question of becoming to the realm of predicates; in
other words, there is only change in quality and quantity but not substance.
The concept of substance as essence [ousia] refers back to Aristotle. In Categories, Aristotle (350 BCE a) calls this the support [hypokeimenon] and later in
Metaphysics (350 BCE b) it is called form [eidos]. I see relations as the possibility to overcome this substantialist view, since a relational analysis will displace
substance from the centre of being (however substance as a concept is hardly
eliminated and it is still central to some construction of formal ontologies). This
possibility can already be seen in Aristotle’s own writings. Aristotle in Categories describes Relative, ‘towards something’ (τὰ πρός τι), as one of the 9 accidents
of the substance. By the end the section, the Stagirite doubted if substance is
itself relative – he tried to show that primary substance is self-sufficient (e.g.
man, horse), but he was not quite sure about secondary substance (e.g. hand,
head): “[…] it may be questioned whether it is true that no substance is relative,
as seems to be the case, or whether exception is to be made in the case of certain
secondary substances” (Aristotle 350 BCE a). In Metaphysics Book V, Aristotle
continues his analysis of three types of relations: identical relations (e.g. whiter,
twice, longer), causal relations (heating and being heated) and psychological
relations (knowing and the object to be known). Here it is worth repeating the
questions posed by the medieval metaphysicians resulting from the interpretation of Aristotle’s Metaphysics: what kind of being are relations? The peculiarity of
relation (in comparison with other accidents) provoked his interpreters, notably
Thomas Aquinas and Duns Scotus to name relation the weakest being (ens debilissimum). We can reiterate the question posed by the realists and reductionists
of the medieval theologians: is relation ‘real being’ in the sense of substantial
beings (which is a res), or just like other accidental beings (which have esse), or even weaker, meaning that they exist only as ratio or modus essendi (reason,
mode of being)." (pp.137-138)
"It was not possible to demonstrate the materiality of relation (as res) in the
time of these medieval thinkers, and it remained largely a speculation and an
attempt to solve the trinitarian problem [...] If we can talk about a relational materialism, it is because relation as a
real being that is made possible by a general tendency of technology, consisting
in the materialisation of all sorts of relations by rendering the invisibles visible
and in measurable forms. For example, writing puts thoughts and perceptions on paper; pulleys, wheels and chains concretise imaginary movements
in mechanical terms; the vapour engine instantiates flows of energy in the
relations between water, fuels, pipes and gears; one could give similar examples
for electricity, nuclear energy, etc. While in the digital environment, we can
observe a more intensive process of materialisation of relation, in terms of data." (p.138)
"Materialism in general did not pay much attention to relational materiality. Among the contemporary materialists, Karen Barad is probably closest to this conceptualisation." (p.138)
"Russell criticised the fact that mathematics has inherited the philosophical
error that an object has to be thought of in terms of subject-predicate propositions. Instead Russell proposed to move relation out of the Aristotelian ontology:
“This view is derived, I think, probably unconsciously, from a philosophical error: it has
always been customary to suppose relational propositions less ultimate than class-propositions (or subject-predicate propositions, with which class-propositions are habitually
confounded), and this has led to a desire to treat relations as a kind of classes.” (Russell
1937: §24)
Let us consider a simple example: “Heidegger knows Bertrand Russell” or “I
am taller than you” – it is impossible to think of these statements in terms
of subject-predicate class-proposition (‘I’ and ‘Russell’ cannot be reduced to a
class-proposition; besides, both of us belong to the class “human being”), but
there is still a need for an independent mathematical treatment of such statements. As Russell suggests, they could be expressed in the form ‘xRy’, in which
‘x’ is understood as the referent, ‘y’ as the relatum and ‘R’ as the relata (ibid.:
§29). It seems worth pointing out the implications of this regarding the development of a relational database. Such thinking gave rise to a relational calculus,
which was further developed in modern mathematics and computer science
into two branches: Tuple Relational Calculus and Domain Relational Calculus.
The Tuple Relational Calculus was introduced by the mathematician and information scientist Edgar F. Codd in the 1960s. It is part of the relational model,
which in turn is the foundation of the Relational Database.
Relation still consists of one of the core philosophical questions today;
furthermore it emerges from a pure metaphysical concept to a concrete and
material concept. In fact, Barad is critical of Russell’s notion of relation, since
she announced that: “I present a relational ontology that rejects the metaphysics
of relata, of “words” and “things” (Barad 2003: 812). However Barad is a science
scholar, but not a technology scholar, and this is the limit of her thinking when
she reproaches the linguistic turn and overlooks that even language is taking
a new form of materialisation, especially in databases, in artificial intelligence,
and, in the semantic web. If an operational and modulative metaphysics is
grounded in relations, as Simondon and Gaston Bachelard have shown, then
we will have to confront immediately the media technologies and the political
economy of such relations. I further propose to understand relations in terms
of what Simondon calls concretisation, by which what is non-material becomes
material, notably causalities." (p.139)
"Simondon lived at a time when
the principles of quantum physics were used for the development of electronic
devices, hence his examples are often focused around diodes, triodes, tetrodes,
pentodes, etc., whereas today this is generally taken for granted. His understanding of technical object is also limited by the working principles of these
devices, which depend mainly on physical contacts. Let us consider a diode: the
operation of a diode depends on the transmission of electrons from the anode
to the cathode, so is a triode, which in addition to the anode and cathode, puts a
gate in between them to amplify the current. The physical contact as the foundation of the reciprocal causality is displaced by a causality operated through data.
I use the word “displace” instead of “replace”, since it is not really replaced, we
can never replace a causality based on physical contacts, but rather within the
dynamic of the technical development, it becomes less and less the core part.
Each epoch has its own media technologies of concretisation. The epoch
of the digital is the epoch of the concretisation of relations in terms of data and
metadata (i.e. data about data). Data (which for me is essential to the understanding of the digital) becomes the new material medium of operation. It is
for this reason, that I separate digital objects from technical objects according
to Simondon, though my reading of digital objects inherits the spirit of Simondon’s analysis. When we consider the development of the relational database and
the technical lineage of mark-up languages from GML, SGML, HTML, XML,
and Web ontologies as proposed by the semantic web, we see that the question of
relations stands out above other concerns." (p.140)
"The core idea of the semantic web is to build a system in which all the data can be
structured according to predefined schemes or web ontologies. As I have shown
with the abovementioned example, the ontologies or schemes define already
“publisher”, “publication date”, “role” and hence it will be possible “[…] to automatise the search of information and to ease the navigation of data” (Tim BernersLee 1998). Everything on the web can be regarded as a resource, and is denoted
with an URI (Universal Resource Indicator). In other words, we could say that
it is a web of logic; however, this does not mean that the web should become a
consistent logical system (since it also produces problems of ambiguity). Instead
it is a web of materialised relations, which can be the URI or the comparison
between any two attributes, even when the relation only says “different from”." (p.141)
"A relational materialism is only explicit under digital conditions, since relation has been always considered to be mediated –it is grasped by a subject that comprehends it. In contrast, for digital materiality, the subjective grasp of relation is no longer the condition and this materialism is rendered visible through digital concretisation.
Datum, means given ; the French word donnée retains this meaning. However, what is now data is no longer given, but has to be mediated, as if this data is already considered as relations." (p.141)
"To invent a new concept of matter, is also to reinvent a new metaphysics
and probably also a technological (post-)humanism in the spirit of Simondon
(though there is no trace that Lyotard has read Simondon). Hence Lyotard
preferred to conceptualise it as interaction rather than creation, as he writes
“[…] if you say creation, that means that you prohibit the other metaphysics that I evoked
earlier: a metaphysics in which, precisely, man is not a subject facing the world of objects,
but only – and this ‘only’ seems to me to be very important – only a sort of synapse, a
sort of interactive clicking together of the complicated interface between fields wherein
particle elements flow via channels of waves” (Lyotard 1984: 9).
What does Lyotard mean by interaction here? Interaction signifies an ontology
of transmission of message without end, in which
“man is not the origin of messages, but rather sometimes the receiver, sometimes the
referent, sometimes a code, sometimes a support for the message, or sometimes the
message itself and the plasticity of human means that this famous communicational
structure looks like not something stable but instead something on which the identities
can no longer be fixed” (ibid.: 10).
The question of language was fundamental to the conceptualisation of Lyotard’s
formulation of new matter, especially since telecommunication technology had
created a new materiality of language between senders and receivers, and more
fundamentally served as the foundation of the postmodern turn. Furthermore,
one can speculate that the concept of interaction exists far before the digital. For
what reason does it re-emerge? My postulation is that the digital renders visible
and makes explicit a relational materialism: that the reflection on language
allows Lyotard to develop an ontology of the material or immaterial according to
the model of telecommunication. The new materiality, as we can see, has to be
mapped in the telecommunication model according to these 5 categories.
1. Matériau (support): by what medium speaks the message;
2. Matériel (receiver): to which destination speaks the message;
3. Maternité (sender): in which name speaks the message;
4. Matière (referent): of what speaks the message;
5. Matrice (code): in what way speaks the message." (p.142)
"What underlies these 5 categories of Mât is relational. It demonstrates a frame
work to understand the abstract concept of the digital in concrete and material
terms. There are several important points one should keep in mind. Firstly,
the maternity is no longer taken as the pure receptacle, which has to be shaped
by the idea, but rather it takes the form of a sender. Secondly the new material
is distributed throughout different components which cannot be separated;
among them there are two relations: one is from the sender (maternité) to the
receiver (matériel), the other is from the message to its referent (matière). Such a
relationality is carried by the support (matériau) according to the coding or rules
of coding (matrice). Thirdly, I would like to reflect on the referent, since it is also
probably something we may want to update after Lyotard.
Gottlob Frege in his famous article Sense and Reference [Sinn und Bedeutung],
distinguishes words according to ‘sense’ and ‘reference’. For example, morning
star and evening star have different senses, but they all have the same reference
which is the Venus (Frege 1948: 211-212). In this conception, the referent is
always something outside, it is not carried by the sense, but rather the sense
only points it to the reference. For Frege, Sinn and Bedeutung merely operate on
the level of signification, but it is only in the new material that Lyotard sees the
transformation of language. Namely, the most systematic medium of signification is turned into materialised computational operations. In a documentary
dedicated to Les Immatéraux, towards the very end of the film, Lyotard proposed
that “language is the most immaterial system that matter has succeeded in
forming” [le langage est le système le plus immatériel que la matière ait réussi à
former] (Lyotard 1986). We may suspect that Lyotard wrote this in the spirit of
the “linguistic turn” that Barad criticised, however, when re-contextualising it
in the relational materiality that I have described, language takes a different
form. We can probably simply replace the word “language” in Lyotard’s quote
with the word “digital”, since underlying the abstract and immaterial concept of
the digital is the most concrete and material system.
Considering the example of data technology, we can see that the referent
is materialised and rendered explicit. To a certain extent, it is similar to what
Kirschenbaum writes about Nelson Goodman’s distinction between the allographic and the autographic object which was mentioned before, but here the
referent is even more explicit and systematic. For example, the semantic web is
a constant process of integrating all materials, no matter what granularity, into
resources, and each resource is given an URI (Universal Resource Indicator). So
for example, now the morning star and evening star can all become resources
such as their referent Venus, and these relations can be named by predicates, for
example _is or _refers_to.
If this logic is well followed, then we can see that such a relational materialism does not only follow a progressive development of technicity, but also
embeds a resistance against the modernist and substantialist view of matter.
If human beings can also be interpreted through the Mât-system, on the one
hand we see that they become unmasterable since they are no longer created,
but rather emerge through interaction. On the other hand, when human beings
become part of this system, they are within it and part of it, and therefore human beings are no longer able to elevate themselves to any ‘transcendent’ plane. In
order to generalise this connection without losing specificity, one could say that
a relational materialism is only possible, if 1) nature is progressively overcome
and transformed in material terms (this is another aspect of the maternity of
matter – the materialisation of relations which turns significations into material
connections: matériau, matière, matrice), and 2) the receptacle itself becomes
relational, in the sense that it can be analysed in terms of relations which are
real, and such relations find their common medium with the embodied experience, which in turn affects its own structure.
A relational analysis is close to what Kirschenbaum calls formal materiality,
since most of these relations emerge from formal structures. However, we have
to consider not only how structures determine relations, but also how relations
determine structures: the inter/intra-action between 1) and 2) that traverse two
orders of magnitude. This recursive relationality was firstly foregrounded by the
theory of feedback in cybernetics; however, such feedback should not be understood as general term, but rather according to the scale, order and magnitude of
the investigated object (e.g., social networks, nanotechnology, synthetic biology,
etc.). It would have been impossible to analyse the same type of relations in the
time of steam engines, but it is possible with regards to digital writing which
possesses a totally different order of granularity. It is also due to this reason, that
Kirschenbaum’s reading of Drucker’s materiality as mere forensic materiality
is not justified – since with this ‘organic form’ materiality has become richer
than just data records. By organic form, I mean the reciprocal relation between
relations and the structure. Moreover, for Drucker, the materiality of embodiment is relevant (which this is not the case for Kirschenbaum). If we acknowledge this rapport between relations and technologies, it seems possible to derive
an analytic tool based on this formulation (between relations and matter), which
goes beyond any dogmatic account of digital materiality." (pp.143-144)
-Yuk Hui, "Towards A Relational Materialism", Digital Culture and Society, Vol. 1, Issue 1, 2015
"Matter is considered [by moderns] to be something to be mastered, dominated and controlled." (note 1 p.132)
"In contrast to the vague demand of the memory of the machine, as a component
of the digital cosmology advocated by philosophers such as Fredkin and Chaitin,
media theories argue more precisely for a ‘digital materiality’. While being
different from the digital philosophers’ approaches, these theorists tend not to
differentiate the ‘immateriality’ of the digital and the materiality of its support,
hence its condition of being material becomes synonym of its materiality.
Among these proposals, I find Matthew Kirschenbaum’s analysis deserving
of careful consideration. Instead of claiming that certain matter is the foundation of the digital universe, he starts from the analysis of technical objects;
like peeling an onion, he starts by asking what their constituents are. In his
book Mechanisms − New Media and the Forensic Imagination, Kirschenbaum
(2008) describes two types of materiality. One he calls forensic materiality and
the other formal materiality. With forensic materiality, he refers to the method
of analysing traces in a computer, going beyond what is visible on the screen.
He takes the disk image of the online game “Mystery House” as an example,
and demonstrates what he calls a “forensic walk through”. Kirschenbaum shows
the hexadecimal and ASCII representations of data inside the file with the open
source software FishWings and analyses the structure of the disk from track to
track and sector to sector. In doing so, he finds traces that are not visible on the
screen, for example the disk image also contains remnants of Bob Bishop’s Dung
Beetles game and a ground-to-air shooter Blitzkrieg. Kirschenbaum concludes
that the Mystery_house.dsk:
“becomes a multivalent forensic environment, one where all of these different levels of
engagement – player, pirate/cracker, postmortem investigator – find their correspondences in the multiple layers of textual events that both drive the game as code and are explicitly thematized within its forensically charged spaces.” (Kirschenbaum 2008: 109)
The researcher who performs this forensic analysis, is like a detective who
examines the traces of a criminal scene, “every contact leaves traces” as he
claimed in talk given ten years ago in the History of Material Texts workshop
at the University of Pennsylvania (Kirschenbaum 2005). With the notion of
‘formal materiality’ he refers to the “normative condition of working in a digital
environment” (ibid.). He suggests speaking of two types of formal materiality.
The first type is the explicit form of digital writing. Kirschenbaum compares
this with Nelson Goodman’s notion of allographic objects; like written text,
they fulfil “their ontology in reproduction” (2008: 133). In contrast to the allographic object stands the autographic object; similarly to a painting, its meaning
cannot be explicitly repeated. The difference between Goodman’s example of
the written text and digital writing is that digital writing has the “state of the art error detection and correction” which allow it to sustain an ideal allographic
environment.
The second type of formal materiality, according to Kirschenbaum, can
be seen in standardised formats such as JPEG, MPEG and predefined logic of
the application. In the footnote of the introduction, Kirschenbaum (2008: 9)
refers to Johanna Drucker’s definition of materiality in The Visible Word: Experimental Typography and Modern Art, 1909–1923. For Drucker, materiality is
considered as “two major intertwined strands: that of a relational, insubstantial, and non-transcendent difference and that of a phenomenological, apprehendable, immanent substance” (ibid.). In other words, materiality is defined
by the dialectics of the objective reality and the subjective experience (varying
from subjective to intersubjective). Kirschenbaum (ibid.) however, criticises that
Katherine Hayles and Drucker’s notion of materiality is only limited to what he
calls ‘forensic materiality’.
Even though Kirschenbaum’s analytic method of layering is very relevant
and the methodology of grounding his theory in objects is plausible, his critique
of Drucker and Hayles’ concept of materiality still has to be reconsidered. The
feminist theorists attempted to extend the notion of materiality from technical
objects to embodiment, which exceeds the material scheme of technical objects
upon the point of their invention (Hayles 1999). Kirschenbaum’s approach rather
emerged from the tradition of textual studies. In contrast, Drucker and Hayles
have suggested a larger scope in order to look at the question of materiality.
While Kirschenbaum is probably right to say that Hayles misses “the computationally specific phenomenon of formal materiality, the simulation or modelling
of materiality via programmed software processes” (2008: 9), a similar critique
cannot be applied to Drucker’s analysis of typography, since ‘formal materiality’
is exactly the condition of any phenomenological experience.
We can also observe that what characterises Kirschenbaum’s analysis is a
regression from the abstract concept of the digital to the endless layers of concrete
matters. This can be read in great contrast to digital physics. Rather than starting
from the One, Kirschenbaum starts with the Many. Unlike digital physics or
digital philosophy which addresses the composition of bits, Kirschenbaum’s
approach seeks a material base on which the digital can be recognized and
embedded, through the decomposition of objects. ‘Regression’ however does not
imply that this is a ‘bad’ development; instead, it indirectly avoids addressing the
digital by addressing its conditions. For example, one can go from one condition
to an outer condition layer by layer, and finally one will end up at the level of
silicones, and probably also further to the sources of energy, etc. However, the
speculative question remains, until when and to which level shall the forensic
operation proceed?
Furthermore, one will find that this analysis can actually be effectively
applied to any type of technical object, and does not necessarily contribute to the
clarification of the digital. Gilbert Simondon mentions a similar insight: with
regards to the manufacture of a needle in Great Britain, he writes that, without
exaggeration, the quality of the needle expresses the degree of perfection of the
nation’s industry (Simondon 2012: 90). However, in the thought of Simondon, there are two important concepts that lead to a third inquiry into digital materiality. The first concept is relation; the second concept is concretisation, which
effectively sublates the difference between the digital and the support of the
digital, which characterise the above approaches. As a point of departure from
digital physics and digital textuality, I would like to outline a third and progressive approach of analysis based on a particular reading of Simondon, which I
am tempted to call a ‘relational materialism’. The last example on XML that
Kirschenbaum gave in order to explain formal materiality would be a perfect
example to understand relational materiality." (pp.135-137)
"While we can see that the two abovementioned approaches either start with or end
up with substance, this approach is an attempt to move away from substance to
relations. Or more precisely, the aim of such relational materiality is to overcome
the hylomorphism proposed by Aristotle, and to see how the development of
technicity distances itself from this analysis.
Hylomorphism is a substantialist thinking: in this context, being can be
comprehended in terms of matter and form. The problem of the substantialist
view is that it limits the question of becoming to the realm of predicates; in
other words, there is only change in quality and quantity but not substance.
The concept of substance as essence [ousia] refers back to Aristotle. In Categories, Aristotle (350 BCE a) calls this the support [hypokeimenon] and later in
Metaphysics (350 BCE b) it is called form [eidos]. I see relations as the possibility to overcome this substantialist view, since a relational analysis will displace
substance from the centre of being (however substance as a concept is hardly
eliminated and it is still central to some construction of formal ontologies). This
possibility can already be seen in Aristotle’s own writings. Aristotle in Categories describes Relative, ‘towards something’ (τὰ πρός τι), as one of the 9 accidents
of the substance. By the end the section, the Stagirite doubted if substance is
itself relative – he tried to show that primary substance is self-sufficient (e.g.
man, horse), but he was not quite sure about secondary substance (e.g. hand,
head): “[…] it may be questioned whether it is true that no substance is relative,
as seems to be the case, or whether exception is to be made in the case of certain
secondary substances” (Aristotle 350 BCE a). In Metaphysics Book V, Aristotle
continues his analysis of three types of relations: identical relations (e.g. whiter,
twice, longer), causal relations (heating and being heated) and psychological
relations (knowing and the object to be known). Here it is worth repeating the
questions posed by the medieval metaphysicians resulting from the interpretation of Aristotle’s Metaphysics: what kind of being are relations? The peculiarity of
relation (in comparison with other accidents) provoked his interpreters, notably
Thomas Aquinas and Duns Scotus to name relation the weakest being (ens debilissimum). We can reiterate the question posed by the realists and reductionists
of the medieval theologians: is relation ‘real being’ in the sense of substantial
beings (which is a res), or just like other accidental beings (which have esse), or even weaker, meaning that they exist only as ratio or modus essendi (reason,
mode of being)." (pp.137-138)
"It was not possible to demonstrate the materiality of relation (as res) in the
time of these medieval thinkers, and it remained largely a speculation and an
attempt to solve the trinitarian problem [...] If we can talk about a relational materialism, it is because relation as a
real being that is made possible by a general tendency of technology, consisting
in the materialisation of all sorts of relations by rendering the invisibles visible
and in measurable forms. For example, writing puts thoughts and perceptions on paper; pulleys, wheels and chains concretise imaginary movements
in mechanical terms; the vapour engine instantiates flows of energy in the
relations between water, fuels, pipes and gears; one could give similar examples
for electricity, nuclear energy, etc. While in the digital environment, we can
observe a more intensive process of materialisation of relation, in terms of data." (p.138)
"Materialism in general did not pay much attention to relational materiality. Among the contemporary materialists, Karen Barad is probably closest to this conceptualisation." (p.138)
"Russell criticised the fact that mathematics has inherited the philosophical
error that an object has to be thought of in terms of subject-predicate propositions. Instead Russell proposed to move relation out of the Aristotelian ontology:
“This view is derived, I think, probably unconsciously, from a philosophical error: it has
always been customary to suppose relational propositions less ultimate than class-propositions (or subject-predicate propositions, with which class-propositions are habitually
confounded), and this has led to a desire to treat relations as a kind of classes.” (Russell
1937: §24)
Let us consider a simple example: “Heidegger knows Bertrand Russell” or “I
am taller than you” – it is impossible to think of these statements in terms
of subject-predicate class-proposition (‘I’ and ‘Russell’ cannot be reduced to a
class-proposition; besides, both of us belong to the class “human being”), but
there is still a need for an independent mathematical treatment of such statements. As Russell suggests, they could be expressed in the form ‘xRy’, in which
‘x’ is understood as the referent, ‘y’ as the relatum and ‘R’ as the relata (ibid.:
§29). It seems worth pointing out the implications of this regarding the development of a relational database. Such thinking gave rise to a relational calculus,
which was further developed in modern mathematics and computer science
into two branches: Tuple Relational Calculus and Domain Relational Calculus.
The Tuple Relational Calculus was introduced by the mathematician and information scientist Edgar F. Codd in the 1960s. It is part of the relational model,
which in turn is the foundation of the Relational Database.
Relation still consists of one of the core philosophical questions today;
furthermore it emerges from a pure metaphysical concept to a concrete and
material concept. In fact, Barad is critical of Russell’s notion of relation, since
she announced that: “I present a relational ontology that rejects the metaphysics
of relata, of “words” and “things” (Barad 2003: 812). However Barad is a science
scholar, but not a technology scholar, and this is the limit of her thinking when
she reproaches the linguistic turn and overlooks that even language is taking
a new form of materialisation, especially in databases, in artificial intelligence,
and, in the semantic web. If an operational and modulative metaphysics is
grounded in relations, as Simondon and Gaston Bachelard have shown, then
we will have to confront immediately the media technologies and the political
economy of such relations. I further propose to understand relations in terms
of what Simondon calls concretisation, by which what is non-material becomes
material, notably causalities." (p.139)
"Simondon lived at a time when
the principles of quantum physics were used for the development of electronic
devices, hence his examples are often focused around diodes, triodes, tetrodes,
pentodes, etc., whereas today this is generally taken for granted. His understanding of technical object is also limited by the working principles of these
devices, which depend mainly on physical contacts. Let us consider a diode: the
operation of a diode depends on the transmission of electrons from the anode
to the cathode, so is a triode, which in addition to the anode and cathode, puts a
gate in between them to amplify the current. The physical contact as the foundation of the reciprocal causality is displaced by a causality operated through data.
I use the word “displace” instead of “replace”, since it is not really replaced, we
can never replace a causality based on physical contacts, but rather within the
dynamic of the technical development, it becomes less and less the core part.
Each epoch has its own media technologies of concretisation. The epoch
of the digital is the epoch of the concretisation of relations in terms of data and
metadata (i.e. data about data). Data (which for me is essential to the understanding of the digital) becomes the new material medium of operation. It is
for this reason, that I separate digital objects from technical objects according
to Simondon, though my reading of digital objects inherits the spirit of Simondon’s analysis. When we consider the development of the relational database and
the technical lineage of mark-up languages from GML, SGML, HTML, XML,
and Web ontologies as proposed by the semantic web, we see that the question of
relations stands out above other concerns." (p.140)
"The core idea of the semantic web is to build a system in which all the data can be
structured according to predefined schemes or web ontologies. As I have shown
with the abovementioned example, the ontologies or schemes define already
“publisher”, “publication date”, “role” and hence it will be possible “[…] to automatise the search of information and to ease the navigation of data” (Tim BernersLee 1998). Everything on the web can be regarded as a resource, and is denoted
with an URI (Universal Resource Indicator). In other words, we could say that
it is a web of logic; however, this does not mean that the web should become a
consistent logical system (since it also produces problems of ambiguity). Instead
it is a web of materialised relations, which can be the URI or the comparison
between any two attributes, even when the relation only says “different from”." (p.141)
"A relational materialism is only explicit under digital conditions, since relation has been always considered to be mediated –it is grasped by a subject that comprehends it. In contrast, for digital materiality, the subjective grasp of relation is no longer the condition and this materialism is rendered visible through digital concretisation.
Datum, means given ; the French word donnée retains this meaning. However, what is now data is no longer given, but has to be mediated, as if this data is already considered as relations." (p.141)
"To invent a new concept of matter, is also to reinvent a new metaphysics
and probably also a technological (post-)humanism in the spirit of Simondon
(though there is no trace that Lyotard has read Simondon). Hence Lyotard
preferred to conceptualise it as interaction rather than creation, as he writes
“[…] if you say creation, that means that you prohibit the other metaphysics that I evoked
earlier: a metaphysics in which, precisely, man is not a subject facing the world of objects,
but only – and this ‘only’ seems to me to be very important – only a sort of synapse, a
sort of interactive clicking together of the complicated interface between fields wherein
particle elements flow via channels of waves” (Lyotard 1984: 9).
What does Lyotard mean by interaction here? Interaction signifies an ontology
of transmission of message without end, in which
“man is not the origin of messages, but rather sometimes the receiver, sometimes the
referent, sometimes a code, sometimes a support for the message, or sometimes the
message itself and the plasticity of human means that this famous communicational
structure looks like not something stable but instead something on which the identities
can no longer be fixed” (ibid.: 10).
The question of language was fundamental to the conceptualisation of Lyotard’s
formulation of new matter, especially since telecommunication technology had
created a new materiality of language between senders and receivers, and more
fundamentally served as the foundation of the postmodern turn. Furthermore,
one can speculate that the concept of interaction exists far before the digital. For
what reason does it re-emerge? My postulation is that the digital renders visible
and makes explicit a relational materialism: that the reflection on language
allows Lyotard to develop an ontology of the material or immaterial according to
the model of telecommunication. The new materiality, as we can see, has to be
mapped in the telecommunication model according to these 5 categories.
1. Matériau (support): by what medium speaks the message;
2. Matériel (receiver): to which destination speaks the message;
3. Maternité (sender): in which name speaks the message;
4. Matière (referent): of what speaks the message;
5. Matrice (code): in what way speaks the message." (p.142)
"What underlies these 5 categories of Mât is relational. It demonstrates a frame
work to understand the abstract concept of the digital in concrete and material
terms. There are several important points one should keep in mind. Firstly,
the maternity is no longer taken as the pure receptacle, which has to be shaped
by the idea, but rather it takes the form of a sender. Secondly the new material
is distributed throughout different components which cannot be separated;
among them there are two relations: one is from the sender (maternité) to the
receiver (matériel), the other is from the message to its referent (matière). Such a
relationality is carried by the support (matériau) according to the coding or rules
of coding (matrice). Thirdly, I would like to reflect on the referent, since it is also
probably something we may want to update after Lyotard.
Gottlob Frege in his famous article Sense and Reference [Sinn und Bedeutung],
distinguishes words according to ‘sense’ and ‘reference’. For example, morning
star and evening star have different senses, but they all have the same reference
which is the Venus (Frege 1948: 211-212). In this conception, the referent is
always something outside, it is not carried by the sense, but rather the sense
only points it to the reference. For Frege, Sinn and Bedeutung merely operate on
the level of signification, but it is only in the new material that Lyotard sees the
transformation of language. Namely, the most systematic medium of signification is turned into materialised computational operations. In a documentary
dedicated to Les Immatéraux, towards the very end of the film, Lyotard proposed
that “language is the most immaterial system that matter has succeeded in
forming” [le langage est le système le plus immatériel que la matière ait réussi à
former] (Lyotard 1986). We may suspect that Lyotard wrote this in the spirit of
the “linguistic turn” that Barad criticised, however, when re-contextualising it
in the relational materiality that I have described, language takes a different
form. We can probably simply replace the word “language” in Lyotard’s quote
with the word “digital”, since underlying the abstract and immaterial concept of
the digital is the most concrete and material system.
Considering the example of data technology, we can see that the referent
is materialised and rendered explicit. To a certain extent, it is similar to what
Kirschenbaum writes about Nelson Goodman’s distinction between the allographic and the autographic object which was mentioned before, but here the
referent is even more explicit and systematic. For example, the semantic web is
a constant process of integrating all materials, no matter what granularity, into
resources, and each resource is given an URI (Universal Resource Indicator). So
for example, now the morning star and evening star can all become resources
such as their referent Venus, and these relations can be named by predicates, for
example _is or _refers_to.
If this logic is well followed, then we can see that such a relational materialism does not only follow a progressive development of technicity, but also
embeds a resistance against the modernist and substantialist view of matter.
If human beings can also be interpreted through the Mât-system, on the one
hand we see that they become unmasterable since they are no longer created,
but rather emerge through interaction. On the other hand, when human beings
become part of this system, they are within it and part of it, and therefore human beings are no longer able to elevate themselves to any ‘transcendent’ plane. In
order to generalise this connection without losing specificity, one could say that
a relational materialism is only possible, if 1) nature is progressively overcome
and transformed in material terms (this is another aspect of the maternity of
matter – the materialisation of relations which turns significations into material
connections: matériau, matière, matrice), and 2) the receptacle itself becomes
relational, in the sense that it can be analysed in terms of relations which are
real, and such relations find their common medium with the embodied experience, which in turn affects its own structure.
A relational analysis is close to what Kirschenbaum calls formal materiality,
since most of these relations emerge from formal structures. However, we have
to consider not only how structures determine relations, but also how relations
determine structures: the inter/intra-action between 1) and 2) that traverse two
orders of magnitude. This recursive relationality was firstly foregrounded by the
theory of feedback in cybernetics; however, such feedback should not be understood as general term, but rather according to the scale, order and magnitude of
the investigated object (e.g., social networks, nanotechnology, synthetic biology,
etc.). It would have been impossible to analyse the same type of relations in the
time of steam engines, but it is possible with regards to digital writing which
possesses a totally different order of granularity. It is also due to this reason, that
Kirschenbaum’s reading of Drucker’s materiality as mere forensic materiality
is not justified – since with this ‘organic form’ materiality has become richer
than just data records. By organic form, I mean the reciprocal relation between
relations and the structure. Moreover, for Drucker, the materiality of embodiment is relevant (which this is not the case for Kirschenbaum). If we acknowledge this rapport between relations and technologies, it seems possible to derive
an analytic tool based on this formulation (between relations and matter), which
goes beyond any dogmatic account of digital materiality." (pp.143-144)
-Yuk Hui, "Towards A Relational Materialism", Digital Culture and Society, Vol. 1, Issue 1, 2015
