The digital and critical theory

The challenge to critical theory explored throughout this book is the prescient warning given by the cyberneticist Norbert Wiener in 1960:

It may well be that in principle we cannot make any machine the elements of whose behaviour we cannot comprehend sooner or later.

Today we live in a world of technical beings, whose function and operation are becoming increasingly interconnected and critical to supporting the lifeworld that we inhabit. Crucially though, this is combined with an increased invisibility or opaqueness of the underlying technologies, and an inability to understand how these systems work, either individually or in concert. This digital world is one of complex, process-oriented computational systems that take on an increasingly complex cognitive heavy-lifting role in society. Without these technologies in place our postmodern financialized economies would doubtlessly collapse - resulting in a crisis of immense proportions. Indeed, our over-reliance on digital technology to manage, control, monitor and support many of the aspects of society we now take for granted is predicated on avoiding the kinds of systematic failure and breakdown that occur routinely in computer systems.

Our societies are increasingly relying on digital technologies of the form that incorporate computational and therefore calculative and computational rationalities which therefore raise important questions for critical theory.

As our reliance on these technical systems grows the technical groups responsible for these systems grow in importance - such that their rationalities, expressed through particular logics, embedded in the interface and the code become internalized within the user as a particular habitus, or way of doing, appropriate to certain social activities. As Marcuse argued, ‘the objective and impersonal character of technological rationality bestows upon the bureaucratic groups the universal dignity of reason' and likewise with its application in computer programming source code. Indeed, the ‘rationality embodied of giant enterprises makes it appear as if men, in obeying them, obey the dictum of an objective rationality' (Marcuse 1978: 154). This is true too of algorithms, perhaps even more so.

As Held writes, ‘as a result, the conditions are created for a decline in the susceptibility of society to critical thinking' (Held 1997: 68). Critical theory has long been accustomed to questioning where ‘thinking objectifies itself', and applying critical approaches to the specific problems raised by instrumental rationality. Combining both close and distant readings of technologies with critical political economy enables the penetration of these technologies into the culture, ecology and even subjectivity of the postmodern actor to be critically appraised and analysed. Indeed, part of this discussion has to be the relationship between instrumental reason and what I want to call computational reason. But, as Held (1997: 69) argues ‘the decline of critical thought is also furthered by the incorporation of opposition. Opposition... have all too often become mimics of the dominant apparatus'. Hence, the dominant apparatus serves as a means of communication creating and legitimating accepted discursive strategies and rhetorics which are, to some degree, shaped not only by rationalities of computation, but also by the affordances of the medium of software itself - something we will return to later in the book.

Indeed, as Kitchin (2011) argues, software can be extremely amenable to certain forms of governmentality, an invisible web that wraps itself around society in increasingly dense weaves and patterns,

Over the past two centuries a mode of governmentality has developed in Western society that is heavily reliant on generating and monitoring systematic information about individuals by institutions.

It is at this point we can begin to materialize the digital and ask about the specific mediations that facilitate these changes. Here, we need to be cognizant of software and digital computers connected through powerful network protocols and technologies. These systems are generally opaque to us, and we rely on them in many cases without questioning their efficacy. Think, for example, of the number of poorly designed website forms that we are increasingly required to fill in, whether for subscriptions, job applications or college classes. These are becoming an obligatory passage point which cannot be avoided, there is no going around these computational gatekeep­ers, and they are the only way certain systems can even be accessed at all.

They are also built of computational logics which are themselves mate­rializations of assumptions, values and norms, often taken for granted, by the designers and programmers of the systems (e.g. gender, race, class, etc.). We need to develop methods, metaphors, concepts and theories in relation to this software and code to enable us to think through and about these systems, as they will increasingly have important critical and political consequences. That is why being able to read these code-based protocols is an important starting point (Galloway 2006). Indeed, what we might call protocol teardowns will be important for seeing the limits of reading code by breaking code, selectively removing selections from a protocol to see if it works and what happens.

The programmer himself state[s]... “legal” principles which permit... “appeals,” he may have only a very incomplete understanding of when and where in the course of the program's operation these procedures will call on each other. And for a particular “court,” he has only a sketchy idea of only some of the circumstances that will cause it to be called upon. In short, once past the beginner level... programmers write - not “sequences” [of computer code instructions] - but specifications for the individuals of little societies. Try as he may he will often be unable fully to envision in advance all the details of their interactions. For that, after all, is why he needs the computer. (Weizenbaum 1984: 234)

From the founding of the computational ‘polities', the interconnections between them are outlined and the interfaces and boundaries between them are detailed and circumscribed in computer code. These systems are then put into the management and oversight of production, distribution and consumption. This automation and control of the production ‘value chain' enables dramatic cost reductions and flexibility. In addition, there is a remarkable feedback loop such that these computational management systems are also applied back into computational technology itself. This leads to an acceleration in the shrinkage of hardware but the expansion of raw computing power for these computational systems and platforms.

In the wake of these new digital technologies facilitated by low-cost fabrication and breakthroughs in software and hardware production, the world has seen major changes in the ways in which we store, access and create information and knowledge, not to mention the profound changes in terms of communication, mobile media and instantaneous connection (Berry 2011a).

The internet has had a lot to do with the growth of this computational ‘digital' world, as a gigantic network of networks through the TCP/IP protocol. Built onto this foundation was the original notion of the web, which was designed around ‘pages' and a client-server architecture by Tim Berners-Lee. The way we have traditionally thought about the internet as pages or websites (and which was always only a subset of internet technologies) is now changing to the newer industrial internet and the concept of ‘real-time streams'.

So too in my own daily experience, I now check my Twitter or Facebook stream first thing every morning, as this aggregates the streams of other people and organizations, giving highlighted links, articles and key issues for the day. This is a crucial resource for keeping up with the latest news and scholarly literature, debates and ideas being circulated, and also connects me to people and things that are being discussed in conferences across the world. Then, if I have extra time, as the Twitter stream is often over-filled with material, I will look at other websites to see what is happening in the technology area and the key talking points and issues being raised - although these are increasingly destination front-ends to real-time back-end systems. These sites are updated in real-time by a process of aggregation and curation, both algorithmic and human, which gives them an excellent overview of their respective areas. Finally, I might check into the Guardian and the BBC to check the daily news. It is this reversal from reading the traditional daily newspaper of record, usually print or broadcast type news and information, that epitomizes the changes I want to argue as being paradigmatic in relation to the reconfiguration of everyday life by the digital.

Here, we might also consider the fragility of the infrastructural systems of the internet, sustained as they are by technical practices that enable incredible computational and communicational feats, while nonetheless being subject to possible break-downs and disruptions. This is demonstrated by the case of a lorry crashing into the power supply at Rackspace, a leading cloud computing provider, ‘knocking out power to the company's Dallas facility. The Rackspace data center switched over to generator power, but two chillers failed to start back up again, compromising the cooling system and forcing Rackspace to take customer servers offline to protect the equipment' (Miller 2007), widely reported as ‘Quick, Plug The Internet Back In: Major Rackspace Outage' (Arrington 2007) and ‘Failure Happens: An SLA is just a contract & Data Centers are single points of failure too' (Robbins 2007), it is nonetheless amazing how much of the internet traffic was taken down by this incident, but also how quickly it was brought back up again - in 3 hours in total. These problems of massive failure of systems that support the internet, especially linked to a single point of failure, are an ongoing research area related to recovery-oriented computing, but also business continuity and internet infrastructure engineering (Radlab 2009).

It should be noted at this point that Morozov's claim that the ‘internet does not exist' and referring to it using scare quotes as ‘internet' is unhelpful (2013: 15, 61-2, 114).² Of course, in an epistemological sense the internet is unrealizable; certainly instantiations of it in empirical, concrete moments clearly are deviations from this sociological ideal-type, however, ‘concepts [when] given ideal typical status, are held to be “logically coherent',' notwithstanding the fact that [ideal types] are unrealisable in practice. Any empirical deviations are held not to call into question their analytical truths' (Holmwood 2013). In the case of the internet, there are certainly certain claims made about its qualities and its force, which may be distinguished from its empirical reality; however, this is not to diminish the importance of the internet as an important signifier and for its conceptual force, as Holmwood further argues in relation to neoliberalism,

The “hegemony” of neo-liberalism... is achieved, in part, by its utilisation of a widely accepted epistemology of social science (that of ideal types, etc.). We could instead take the unrealisable nature of the neo-liberal construct of the market as an index of its irrationality (rather than claim it to be rational but unreal). That we do not is because we remain committed to the social sciences it makes possible. (Holmwood 2013)

With the growth of computationalism and the internet as, perhaps, a legiti­mating epistemological framework, many sectors of society and the economy are increasingly subject to computational processes and the challenge is to use critical theory to help think through the consequences of this. Rejecting the 'internet' as an irrationality as Morovoz aims to do does not help us to think critically about the way in which concepts can 'provide the basis for pol­icy based upon its putative rationality' (Holmwood 2013). This may be played out as an empty signifier and the idealist prescriptions it suggests, but it is also deployed through the proliferation of material black boxes, cables, routers and opaque computational devices, which due to their opacity we are seem­ingly unable to read or fathom. Indeed, as Weizenbaum warned in 1984:

most of us have thought enough about the progressively increasing dependence upon computers in commerce and industry to project a picture in which the very functioning of society depends on an orderly and meaningful execution of billions of electronic instructions every second.... In mastering the programming and control of computers, we [in the university] especially could play a critical role. It may well be that no other organization is able to play this role as we are, yet no more important role may exist in science and technology today. The importance of the role stems, as has been noted, from the fact that the computer has been incorporating itself, and will surely continue to incorporate itself, into most of the functions that are fundamental to the support, protection, and development of our society. (Weizenbaum 1984: 242)

The unique means of reading and writing in a computational society is clearly a crucial question in relation to these issues. Addressing the specific problems raised by a particular literacy connected to the digital is a pressing issue. How should we read the digital - and to what extent should and can we be expected to write the digital?

One way of thinking about this is through how the digital (or computa­tional) presents us with a number of seemingly theoretical and empirical contradictions which we can understand within this commonly used set of binaries outlined by Liu (2011) and which I want to extend and discuss here: (1) linearity versus hypertextuality; (2) narrative versus database; (3) perma­nent versus ephemeral; (4) bound versus unbound; (5) individual versus social; (6) deep versus shallow; (7) focused versus distracted; (8) close reading versus distant reading; (9) fixed versus processual; and finally (10) digital (virtual) versus real (physical). Here, I am not interested in critiquing the use of binaries per se, and which of course remains an important task, rather I think the binaries are interesting for the indicative light they cast on draw­ing analytical distinctions between categories and collections related to the digital itself. These binaries can be useful means of thinking through many of the positions and debates that take place within both theoretical and empirical work on mapping the digital and will play out in the discussions throughout this book. Indeed, the collection of binaries can be thought of as part of a mapping of the digital through a form of constellation of concepts, some of which I now outline in more detail,

Linear versus Hypertextuality: The notion of a linear text, usually fixed within a paper form, is one that has often been taken for granted within the humanities. Computational systems, however, have challenged this model of reading linearly due to the ease with which linked data can be incorporated into digital text. This means that experimentation with textual form and the way in which a reader might negotiate a text can be explored. Of course, the primary model for hypertextual systems is today strongly associated with the worldwide web and HTML, although other systems have been developed. There are, of course, examples of non-linear paperbound systems but it is argued that the digital manifests something new and interesting in this regard.

Narrative versus Database: The importance of narrative as an epistemological frame for understanding has been immensely important in the humanities. Whether as a starting point for beginning an analysis or through attempts to undermine or problematize narratives within texts, humanities scholars have usually sought to use narrative as an explanatory means of exploring both the literary and history. Computer technology, however, has offered scholars an alternative way of understanding how knowledge might be structured through the notion of the database (see Manovich 2001). Although as with all these binaries there are clearly a spectrum of normative forms, nonetheless these distinctions are often pointed towards as indicative of digital media (see Bassett 2007).

Permanent versus Ephemeral: One of the hallmarks of much 'traditional' or 'basic' humanities scholarship has been concerned with objects and artefacts that have been understood as relatively stable in relation to digital works. This is especially in disciplines that have internalized the medium specificity of a form, for example, the book in English Literature, which shifts attention to the content of the medium. In contrast, digital works are notoriously ephemeral in their form, both in the materiality of the substrates (e.g. computer memory chips, magnetic tape/disks, plastic disks, etc.) and in the plasticity of the form. This also bears upon born-digital from which derivative copies are made. Indeed, it could be argued that in the digital world there is only the copy (although recent moves in Cloud computing and digital rights management are partial attempts to re-institute the original through technical means). The ephemerality of the digital must also be thought about by reference to the counterfactual resilience and permanence of some digital objects such as hard disk drives whose contents can sometimes survive extreme conditions such as fire, water and even physical damage.

Bound versus Unbound: A notable feature of digital artefacts is that they tend to be unbound in character. Unlike books, which have clear boundary points marked by the cardboard that makes up the covers, digital objects boundaries are drawn by the file format in which they are encoded. This makes it an extremely permeable border, and one that is made of the same digital code that marks the content. Additionally, digital objects are easily networked and aggregated, processed and transcoded into other forms further problematizing boundary points between them. In terms of reading practices, it can be seen that the permeability of boundaries can radically change the reading experience. To some extent the boundlessness of the digital form has been constrained by digital rights management and related technical protection measures. However, these techniques remain open to hacking techniques and once the locks are broken the digital content is easily distributed, edited and recombined.

Individual versus Social: Traditional humanities has focused strongly on approaches to texts that are broadly individualistic in as much as the reader is understood to undertake certain bodily practices (e.g. sitting in a chair, book on knees, concentration on the linear flow of text). Digital technologies, particularly when networked, open up these practices to a much more social experience of reading, with e-readers like the Amazon Kindle encouraging the sharing of highlighted passages, and Tumblr-type blogs and Twitter enabling social reading and discussion around and within the digital text. There are also new forms of curating and collaborative writing methods mediated by digital tools such as in practices such as 'booksprints' which are highly creative and collaborative modes of writing text (Booksprints 2013). Relatedly new digital media are also multi-modal in terms of trans-media forms and enable new social reading practices.

Deep versus Shallow. Deep reading is the presumed mode of understanding that requires time and attention to develop a hermeneutic reading of a text, this form requires humanistic reading skills to be carefully learnt and applied. In contrast a shallow mode is a skimming or surface reading of a text, more akin to gathering a general overview or precis of the text. This deep form of reading that is closely associated with humanities scholarship and the appreciation of a well-read book, in contrast to the internet trend towards 'tl;dr' or ‘too long; didn't read'. Deep reading as a practice requires training over long periods of time which may be challenged by the real-time streaming forms of always-on short-form or micro-form media such as Twitter and email.

Focused versus Distracted: Relatedly, the notion of focused reading is also implicitly understood as an important aspect of humanities scholarship. This is the focus on a particular text, set of texts or canon, and the space and time to give full attention to them. By contrast, in a world of real-time information and multiple windows on computer screens, reading practices are increasingly distracted, partial and fragmented (hyperattention). However, there have been some interesting attempts to allay this problem with either focal technologies which remove clutter and distraction while working at the computer. Additionally augmented or extended cognition tools that offer reconciliation at a higher level have been developed; however, it does appear that digital media as currently constituted under late capitalism tends towards a distracted form of reading ideal for consumption practices.

Close Reading versus Distant Reading: Close reading is the practice associated with the careful, slow, explication and interpretation of textual sources. It is usually associated with a heteronomous collection of practices, but generally they are concentrated on a small number, or just a single text. Distant reading is the application of technologies to enable a great number of texts to be incorporated into an analysis through the ability of computers to process large quantities of text relatively quickly. Moretti (2007) has argued that this approach allows us to see social and cultural forces at work through collective cultural systems. Again, this is related to a notion of extended cognition through the use of digital technologies to precis, rework, summarize or visualize large quantities of data, whether textual or otherwise. This threat to close reading practices is often manifested or linked to a notion of liberal subjectivity and the enlightenment project of modernity. It is considered highly significant for all humanities disciplines.

Fixed versus Processual: The digital medium facilitates new ways of presenting media that are highly computational. This raises new challenges for scholarship in understanding digital media and the methods for approaching these media forms. It also raises questions for older humanities that are increasingly accessing their research object through the mediation of processual computational systems, and more particularly through software and computer code. The issue of processual media, in relation to the fixed or time-based media of the twentieth century, is that they incorporate feedback into their media forms, such as interactivity, but also this could be reading speed, colour preferences, etc. in order to change the way a media is perceived or understood. Digital media are also able to adapt to the reader/ viewer in real-time, changing the content, narrative, structure, presentation and so forth on the fly, as a kind of surveillance literature which is reading the reader as she reads.

Real (physical) versus Digital (virtual): This is a common dichotomy that draws some form of dividing line between the so-called real and the so- called digital. Increasingly, with the collapse in our experience of computation from its fixed desktop location, we are using computation in everyday spaces, and which infuses the everyday environment with a computational overlay or layer, such that the computational is not distinct from but deeply embedded in everyday life and hence is post-digital. It is still crucial to historicize the notion of the digital, though, particularly in relation to our changing experience of the digital as previously 'online' and todays 'always online', such that being offline increasingly becomes an experience of only the very rich (as choice) or the very poor (by necessity).

Above, we gestured towards the softwarization of 'close reading' and the changing structure of a 'preferred reader' or subject position, towards one that is increasingly algorithmic (of course, this could be a human or non-human reader). Indeed, it is suggestive that as a result of these moves to digital real-time streams we may see the move from a linear model of narrative, exemplified by print books, to a 'dashboard of a calculation interface' and 'navigational platforms', exemplified by new forms of software platforms. Indeed, these platforms, and here I am thinking of a screenic interface such as the iPad, allow the 'reader' to use the hand-and-eye in haptic interfaces to develop interactive exploratory approaches towards knowledge/information and 'discovery'. This could, of course, still enable humanistic notions of 'close reading' but the preferred reading style would increasingly be 'distant' partially, or completely, mediated through computational code-based devices.

At this point in the book these issues are necessarily abstract. However, these important aspects will be reflected upon in the chapters that follow, and in the particular critical problematics raised by the digital. Indeed, the use of computational systems creates a highly computationally mediated lifeworld which raises challenging questions that Horkheimer envisioned already in 1947 when he talked about the prevalence of science as the arbiter of knowledge and truth:

Justice, equality, happiness, tolerance, all the concepts that.... Were in preceding centuries supposed to be inherent in or sanctioned by reason, have lost their intellectual roots. They are still aims and ends, but there is no rational agency authorized to appraise and link them to objective reality. Endorsed by venerable historical documents, they may still enjoy a certain prestige, and some are contained in the supreme law of the greatest countries. Nevertheless, they lack any confirmation by reason in its modern sense. Who can say that any one of these ideals is more closely related to truth than its opposite? According to the philosophy of the average modern intellectual, there is only one authority, namely, science, conceived as the classification of facts and the calculation of possibilities. The statement that justice and freedom are better in themselves than injustice and oppression is scientifically unverifiable and useless. It has come to sound as meaningless in itself as would the statement that red is more beautiful than blue, or that an egg is better than milk. (Horkheimer, quote in Weizenbaum 1984: 252)

This dislocation of important emancipatory concepts from reason raises challenging questions for a world that increasingly encodes similar principles within the computational logics of software and algorithms. While it is clear that not all judgements and decision points can be delegated to code, nonetheless a substantial number can be, and it is not yet clear where that line is.The question of articulation of these emancipatory concepts within code, and the danger this might pose to critical thinking is an important question. As computation becomes increasingly powerful, there is a temptation to allow algorithms and code to 'judge', as indeed currently happens with, for example, some parking ticket systems. So rationality encoded within the limits of computation is a form of instrumental rationality and becomes automated and prescriptive, an issue that needs to be made manifest for democratic societies.

For this book, this includes a critical engagement with digital code (software), and I seek to reveal how the digital is embedded with a particular instrumental rationality, and how by the development of counter-institutions and 'counter­code', new possibilities for praxis are available within the digital. Indeed, the political potential of code signalled by certain civil society groups such as free and open source software, Wikileaks, hackers, cypherpunks and the Electronic Freedom Foundation (EFF), certainly raise intriguing possibilities, provided they are not limited by instrumentalized computational rationality. However, for the researcher or political actor, even just beginning to understand digital code is horrendously difficult due to its opacity. Indeed, it is still relatively rare for a researcher, let alone a political activist, to perform any form of reading of digital code at all. This again raises key questions about the kinds of literacies that are appropriate for active citizenship today.

I therefore argue, following Adorno, that we need to perform a critique of knowledge (Erkenntniskritik) in order that we can make a social critique (Kritik an der Gesellschaft) of contemporary social formations. In other words, just as theorists have discussed the ‘linguistic turn' in social and political theory, here I argue that we need to question the ‘computational turn'. For Horkheimer, Adorno and Marcuse, sociology and critique are inseparable, they must explore a cultural artefact's formation and reception. As Held (1997) explains, ‘such an inquiry seeks to understand given works in terms of their social origins, form, content and function - in terms of the social totality.... But a sociology of culture cannot rest with an analysis of the general relations between types of cultural products.... It must also explore in detail the internal structure of cultural forms (the way in which the organization of society is crystallized in cultural phenomenon) and the mechanisms which determine their reception' Held (1997: 77). That is, it should account for the processes of production, reproduction, distribution, exchange and consumption. With software we are presented with a unique location for just this kind of critique.

But, it is also no longer surprising that code is becoming an important source of profit and as such will be a key resource and source of conflict in the twenty-first century. Today, the digital has become the key medium in our societies and it is critical that we develop our critical attention on this little understood but hugely important subject. The situation has changed due to the quantitative shift in digital usage and reach, and as its functions change so does the way in which the digital interacts with our everyday lives. So, in the following chapters, I want to consider the quotidian experience of the digital and where it lies in our everyday relationships with each other and new media technology. I explore how the digital is increasingly mediating our experiences of the world, through computational devices like the smart phone and mobile tablets. I also start drawing links with critical theorists' concern with the notion of instrumental reason and domination - especially considering that a history of the digital is in many cases also a history of a military-industrial complex that has become a key aspect of the entertainment industry.

However, this is not merely a claim that a magnified or excess of instrumental reason is delegated into machines such that the computational and the instrumental become synonymous. Heidegger criticized this as a poor understanding of technology, and it holds equally true for computational technology. As Dreyfus argues,

In his final analysis of technology, Heidegger was critical of those who, still caught in the subject/object picture, thought that technology was dangerous because it embodied instrumental reason. Modern technology, he insists, is “something completely different and therefore new” The goal of technology Heidegger then tells us, is the more and more flexible and efficient ordering of resources, not as objects to satisfy our desires, but simply for the sake of ordering. (Dreyfus 1997)

This Heideggerian insight needs to be combined with critical theory to provide an understanding of the extent to which we live in a world full of technologies and devices that also serve to make possible the fulfilment of our desires - although usually without giving any indication of how digital technologies mediate this satisfaction. Digital devices have for many people become akin to magic. As David Parnas argued, ‘technology is the black magic of our time. Engineers are seen as wizards; their knowledge of arcane rituals and obscure terminology seems to endow them with an understanding not shared by the laity' (Berry 2011: 35-6). Indeed, ‘hiddenness is a powerful... force of mystification.... Marx writes that truly to understand the workings of exchange, one must leave the marketplace and venture “into the hidden abode of production”' (Galloway 2006: 99). The digital is simultaneously technical and social, material and symbolic, but it is also a historically located concept, as are its instantiations in concrete computational devices. This is not a new problem, but it does make the digital challenging to investigate and, I believe, it puts critical theory in a unique position to problematize.

Further, the social dimension of language production and usage is crucially important both to appreciate the way in which the digital is a social practice that aids interpretation, and the extent to which recent innovations in digital technologies like Twitter, Facebook and Google+ lie at the intersection of technology, language and social practice. As Gadamer argues, ‘in fact history does not belong to us, but we belong to it. Long before we understand ourselves through the process of self-examination, we understand ourselves in a self-evident way in the family, society and state in which we live' (Winograd and Flores 1987: 29).

Thus, history and the historicity of our experience of the world are critical to our ability to know and understand ourselves. Digital technologies form a greater part of the technical and media ecology of the environment that surrounds us and records our memories and in some cases is our memories. So, increasingly, digital media becomes part of our cultural background, and this contributes to our very way of experiencing the world and living and using language.³ Again, as Gadamer explains:

To acquire an awareness of a situation is, however, always a task of particular difficulty. The very idea of a situation means that we are not standing outside it and hence are unable to have any objective knowledge of it. We are always within the situation, and to throw light on it is a task that is never entirely completed. This is true also of the hermeneutic situation, i.e., the situation in which we find ourselves with regard to the tradition that we are trying to understand. The illumination of this situation - effective-historical reflection - can never be completely achieved, but this is not due to a lack in the reflection, but lies in the essence of the historical being which is ours. To exist historically means that knowledge of oneself can never be complete. (Winograd and Flores 1987: 29)

The task, then, is not merely to undertake better 'science' or more empirical work. Rather, this is an affirmation of the particular research programmes exemplified by the humanities disciplines and thus a call to them to explore the challenges raised by the digital, to uncover, not only the 'politics of the database' (Poster 1996: 93) but also the politics of the stream. Knowledge, and humanities knowledge in particular:

Have a greater familiarity with an ambiguous, intractable, sometimes unreachable [moral] world that won't reduce itself to any correspondence with the symbols by means of which one might try to measure it. There is a world that stands apart from all efforts of historians to reduce [it] to the laws of history, a world which defies all efforts of artists to understand its basic laws of beauty. [Man's] practice should involve itself with softer than scientific knowledge... that is not a retreat but an advance. (Weizenbaum 1984: 279)

In the next chapter, we turn to the questions raised here, particularly in relation to unpicking the problematic raised by the digital, to carefully think through the implications of a computational society. This, of course, necessitates a more thorough examination of the conditions of possibility of a digital or informational society, through an exploration and thinking through of the code, software and algorithms that act as the crucial infrastructure of the computational.

52