Data Visualisation as Transcription

If you stay you can change the world, you become a yes, 1.

(Mr Robot: Season 1. Episode 2. Time code: 25:00)

The capturing of screenshots during my ‘experiment in living’ does not collect data, only an image of the results. Yet saving ‘complete webpages’ of my interaction with algorithms captures code that can be processed afterwards. In order to ‘make sense’ of the small data sets I collected

from my search ‘experiment in living’ I needed to visualise them. In his text Are some things Unrepresentable? media theorist Alexander Galloway addresses the visualisation of information and data through the dilemma of ‘unrepresentability’.

Either data offer zero help as to how they ought to be aestheticized, or they eclipse all available possibilities under a single way of seeing. One might assign a name to this curious contradiction. One might call it the dilemma of unrepresentability lurking within information aesthetics (Galloway 2011:91 emphasis mine).

Furthermore, Galloway pinpoints the problematics of ‘unrepresentability’ in ‘information aesthetics’ [data visualisation] through the lens of discriminating between 0’s and 1’s by

applying a ‘dialectical logic’. He first maps out the etymology of data (that which is given) with a double entendre on the word ‘données’. Bundled up within this ‘gifting’ is the ‘ontological rawness’ of data, as they are not just measurements or recorded facts (ibid:87). Yet there is confusion between the term ‘data’, which is considered raw and numerical, and ‘information’––

that which has been ‘given’ form and he discriminates between them (ibid:81). The first thesis is that ‘data have no necessary visual form and are represented by a ‘0’. The second thesis is that only one visualisation has ever been made of an information network’ because all

‘visualizations look the same’ (ibid:90) and this is represented by a ‘1’.

Figure 29: Afghan Stability/COIN Dynamics used by McCrystal (2009).

He employs US General Stanley McCrystal’s power point slide (2009) depicting the American military strategy in Afghanistan as an example of his ‘1’.⁷⁹ (Figure 29) Yet according to Galloway ‘1’s are not sufficient; they are actually zeros: ‘when there is only one, there is

79 There was critique in the media that Power Point software simplifies information, ‘edits ideas’ and does not link these ideas or facts to any kind of human narrative (Borger 2010). Or, ‘Is there another way to present the

information that doesn't look like it has been put together by a kitten with a ball of string?’ (Rogers 2010).

nothing. For a representation of the one, is, in fact, a representation of nothing’ (2011:90 emphasis mine). Therefore, he argues that ‘information aesthetics’ actually facilitates a decline in the transparency of informatics.

The point of unrepresentability is the point of power. And the point of power today is not the image. The point of power today resides in networks, computers, information, and data (ibid:95).⁸⁰

Galloway then continues by developing an argument between ontology and aesthetics:

Thus if data open a door into the realm of the empirical and ultimately the ontological (the level of being), information by contrast opens a door into the realm of the aesthetic (Galloway 2011:96).

With aesthetics, the countable alphabet of 0’s and 1’s becomes ‘operable and develops

organizational powers’ yet the data needs to be processed somehow (Beverungen et al. 2019).

‘Any visualization of data must invent an artificial set of translation rules that convert [an]

abstract number to semiotic sign’ (Galloway 2011:96). Yet certain ‘critical’ cartographies remain incomplete. In this sense, information aesthetics has not been able to represent all that needs to be imaged and, referencing Gilles Deleuze (1992), Galloway proclaims that ‘adequate visualizations of control society have not happened. Representation has not happened. At least not yet’ (Galloway 2011:95).

Speculative visualisations, cognitive mappings (Jameson 1988), or diagrams of ‘work flows’

have already happened in artistic practices, such as the work of Mark Lombardi (2001). Within the realm of ‘digital aesthetics’ this provocation is precarious and tenuous. ‘Many information structures have graphical analogies and can be understood as diagrams that organise the relations of elements within the whole’ (Drucker 2009:16). However, part of the reason why these visualisations are inadequate is because algorithmic formulae, the code of proprietary algorithms, is unknown and their interaction with humans has not been able to be depicted visually ––what is now often called the ‘black box’ dilemma or ‘unrepresentability’. Galloway’s provocation incited me to invent a method of converting my search results, making what is not visible or ‘representable’ legible by transcribing data into information.

In 1997 Ellen Ullman elucidated how users not only interface with software tools and

techniques such as the spreadsheet, which are ‘maps’ of information, but how they demonstrate agency in the process. Even in an era when searching the web was a nascent technology, Ullman keenly articulated how the transcriptive process unfolds with a spreadsheet of ‘coding’ data, with the human actor giving data form, thereby ‘informing’ the tools of technology.

The user gives data its shape––places it in columns and rows, expresses the complex relationships among those columns and rows––and eventually turns data into more knowledge. It is the end user who creates information, who gives form to data, who informs the spreadsheet (1997:78).⁸¹

80 In his article The Total Archive, Andreas Bernhard concurs that making this ‘distinction is productive’ and that it could be applied as a means to ‘isolate the rift that exists between mathematically calculated and visualized knowledge’ (2015:22).

81 Perhaps this is what inspired Galloway’s distinction between data and information.

During my ‘experiment in living’ I carried out search queries with Google and Tor on two computers and compiled ‘small’ sets of data. I provided the input–– keywords, or ‘terms of art’

in order to test out their “currency” in an era of cognitive capital (Appendix A). The data from these qualitative ‘interviews’––my output, URLs and webpages of data––can be seen as ‘notes’

from my fieldwork that I then ‘transcribed.’ This entails not only collecting, extrapolating, cleaning and (re)structuring data but ‘coding’ the data. What used to be transcription in

ethnographic coding of data becomes algorithmic data visualisation in the era of digital research methods.

Figure 30: Data visualisation as transcription, Re:search - Terms of Art (Performativity, Contemporaneity)

Re:search - Terms of Art, produced together with the interactive graphic designer, Richard Vijgen, is the result of this conversion process (Appendix E). (Figure 30) These visualisations of my small data set show how the outputs (hyperlinks) are ranked, along with the similarities and differences between my two search methods: Google Search and the Tor browser.⁸² These ‘data visualisations as transcription’ are only representations of what was captured in a certain time frame and this presence also incites absence, in that they are incomplete. They offer only a glimpse––a peek into the mysterious black box––with the results as URLs. However, this data visualisation process enabled me to compare these two types of online querying as a base for analysis. Therefore, as much as I endeavoured to answer Galloway’s call for ‘a poetics as such for this mysterious new machinic space’ (2011), whether my data visualisations are adequate representations remains to be seen.

82 I was invited to take part at the exhibition Hacking Habitat, Utrecht, NL where I exhibited these data visualisations made in collaboration with interactive designer Richard Vijgen (see Appendix E).