Naturing With Big Data: How Writing Creation Myths Can Matter

The New Naturalism(s)

The object of ethical concern from a posthumanist perspective—what it regards as unethical—is research that claims to innocently represent the world “as it really is.” (Mauthner, 2019, p. 671)

On the one hand, a “renewed naturalism” (Tönberg & Tönberg, 2018, p. 2) prevails in the epistemological practices with big data, epitomized by “the end of theory” (Anderson, 2008) and “dataism” (Van Dijck, 2014). Behind the positivist, bottom-up epistemology, however, lies a strong “anti-scientific computationalism” (Longo, 2019a, 2019b) that claims “science can advance even without coherent models, unified theories” (Anderson, 2008). On the other hand, entirely external to epistemology, developing ethics of big data as a designation of top-down principles that emphasizes individual moral agency seems to be a key focus of concern. Although efforts have been made to advance ethics that correspond to the current conditions (see, for example, Floridi, 2009; Mayer-Schönberger & Cukier, 2013; Zwitter, 2014), developing a set of rules and guidelines for data production, privacy, access, and ownership and upholding accountability for the potential harm to individuals (Mills, 2018) remain primary, including in research ethics. Indeed, the term “ethics” is charged with morals, particularly humanist morals. The deontological, consequentialist ethics is deeply rooted in humanist philosophies that are quick to naturalize any given morality into existing normative frames. Similarly, ethics of AI and ethics of algorithms—where the issues of big data are inevitably involved—focus on building a responsibility model from data organization to fairness assessment, as challenges arise in governance and algorithmic decision-making for prediction, optimization, and control. The Fair Machine Learning (Fair ML) movement is a recent manifestation of a “responsible” ethics. ¹

From a posthumanist perspective, it is important to rethink the ethics of big data that treats knowledge-making itself as a matter of ethical concern (Mauthner, 2019). Since worlding (Barad, 2007; Haraway, 2008; Stewart, 2014) and knowledge-making are intrinsically ethical matters, it follows that ethics and epistemology are not separate. What seems missing to me is an affirmative mode of knowledge production (Braidotti, 2009, 2019) for qualitative research with big data that takes ethical concerns as an integral part of research. An affirmative mode of knowledge-making would involve multiple curiosity-driven practices whose “driving force” is the active desire to actualize unprecedented modes of epistemic relations (Braidotti, 2019, p. 44). Creativity—not as “the logic of technocratic solutionism” but as “a generate problem in its own right” (Harris & Rousell, 2022, p. 429)—would be the chief posthumanist ethical affect, an “ability to constitute the very nature of their objects/subjects of study” (Mauthner, 2019, p. 671).

In this article, I mainly draw on Barad’s posthumanist ethics of mattering as an alternative methodology to qualitative research with big data while foregrounding writing as mattering. Big data is outlined as linguistic matter to be understood in relational materiality and subjectivity. I then discuss writing in light of mattering that employs multiple strategies. To illustrate, I present an example of a course in which students wrote creation myths in an affirmative manner together with eighty existing ones. The final section sketches what the ethics of naturing with big data might be about.

Big Data as Linguistic Matter

Big data has been variously defined and is not simply denoted by volume (boyd & Crawford, 2012; Kitchin, 2014). This section explains big data in terms of relational materiality, that is, big data as matter, instead of seeking defining characteristics ² . The focus is on shifting the current mode of knowledge production and mode of inquiry when we work with digitally encoded data on a potentially large scale. The argument is that whereas analytics, interpretation, and reading ³ are the dominant modes of inquiry—in which big data is treated as neutral objects or texts—Barad’s notion of mattering can provide a richer perspective in facilitating an affirmative mode of knowledge production.

I define mattering as an ethico-onto-epistemological process of reconfigurating in/stabilities (Barad, 2003, 2007, 2012). At the core of mattering are “ontological indeterminacy, a radical openness, an infinity of possibilities” (Barad, 2012, p. 16). Moreover, by positioning big data as a linguistic matter I intend to make it explicit that language and matter are not in opposition; therefore, I break from Barad’s (2003, p. 801) lament on language and its representationalism regarding this issue. To be sure, I am not the first one to raise the concern. Palmer (2020), in particular, foregrounds language for mattering while queering the spatiotemporal axes of language that brings us far beyond representationalism. Chen (2012, p. 51) also writes,

I refute the recent moves to evacuate substance from language, for instance, the notion that language is simply dematerialized; one of the outcomes of this belief, it seems to me, is that language discussions seem to disappear in the theorizing of new materialisms. ⁴

I will return to the point about why language matters in the next section. For now, to proceed with the assumption that big data is linguistic matter, it is important to address two particular aspects of digital data, in that they call for a rethink in relation to subjectivity and mattering. These aspects are discreteness and alphanumericness, as I explain below.

First, in looking at discreteness, it is useful to consider Longo’s account of the nature of digital data and its consequences on knowledge construction. Digital data, in principle, is produced by discrete mathematics as an “imitation or reality” or an “image of the world” and is thus far from neutral (Longo, 2019b). In other words, digital data is from the beginning, materially manipulated in discrete, exact structures, which allows, in turn, arithmetic certainty and countability (Longo, 2019b). But what does this mean? In discrete structures, one can only count but not measure, because of the arithmetic certainty and because the metric is intrinsic. With counting, one gets meaningless—but not neutral, not objective—sequences of signs; with measuring, meaning emerges and knowledge may begin. So, following Longo, there is “the need for a knowing subject, with his or her choice of (pertinent) observables and form of access and measurement, his or her concrete friction to reality and knowledge proposal, in order to construct objectivity” (p. 9). The point is that understanding cannot automatically happen only by counting the discrete matter; no pure objectivity is pre-given regardless of the size of the data set.

As boyd and Crawford (2012, p. 663) also suggest, big data has more to do with a capacity to search, aggregate, and cross-reference than the big volume itself. Nevertheless, following Calude and Longo’s (2017) mathematical and information-theoretical analysis, the scientific inventiveness and intelligibility have been neglected in “a deluge of spurious correlations” which actually do not allow either prediction or action. Measuring and modeling, then, have a lot to do with our “own invention of the real,” which “in the process profoundly change our reality” (Longo, 2019a, p. 91). Big data as discretely encoded matter demands processes of invention, measurement, and reconfiguration.

Next, I would like to call attention to the alphanumericness of big data, among other kinds of multimodality. When we work with big data, we not only count and measure, but also navigate, search, guess, colligate, associate, make sense, project, abstract, and imagine; we work in a “computational space of imagination” (Finn, 2017), which involves a highly imaginative and speculative process of reasoning. All these processes are possible in large part because of the alphanumericness of digital data. Being alphanumeric means it has qualities of both numbers and letters; it can be technically measured and speculatively negotiable at the same time; and because it is double-articulated, it allows reasoning in a highly malleable spacetime both materially and symbolically.

In sum, one can navigate big data in a manner that operates between the subjective and the objective, as one projects his or her search for meaning and relation on the symbolically accessible level of alphaness, while on the meaningless but quantitative level of numericness data articulates itself materially. The dichotomies of human-data, human-machine, and subjective-objective are blurred. This might be what Braidotti (2013) calls “non-unitary subjectivity” that is “embodied and embedded” in the spectrum of big data, along with the idea of “becoming-machine” in a specific sense that “actualizes the relational powers of a subject that is no longer cast in a dualistic frame (. . .) the merger of the human with the technological results in a new transversal compound, a new kind of eco-sophical unity” (p. 92).

Writing and Mattering

Writing is a powerful “practice/action/doing” (Barad, 2003, p. 802) in which we can engage with language—and with linguistic matter. In this section, I explicate writing through the idea of mattering, an ethico-onto-epistemological process of reconfiguration. This allows us to elaborate the role of writing in posthumanist knowledge creation, including in the complex relationship with the materiality of the world.

Let us think of writing, first within the broad context of mattering, which I outline as the material-logical level. Writing, from alphabetical writing to computer programming, is a material technology ⁵ that can affect material reality. It works by rewriting, coding, or trans-figuring what is actually or virtually out there. Thus, it is from the outset, I argue, a process of reconfiguration. That is, writing on a material-logical level can bring things into “beings” and thereby enable things to operate. Clustering, parsing, capturing, figuring, encoding, encapsulating, and abstracting things into discernible units and rules, terms and contracts are the techniques of (re)writing that activate logical materiality. Here, I add an emphasis on the verb activate, too, because the creative activation of symbolic systems and intelligible/logical channels has a long history of invention in mathematics, science, and technology. For example, the gnomon, is one of the oldest “writing-machines”; the sun-dial casts a vertical shadow by intercepting the sun, thereby writing the time of a day geo-materially. Or, writing with new symbols (e.g., with “0” or “i” in mathematics) can also matter in that it allows “logistical kindling”(Peters, 2015, p. 176) that can affect reality on an infrastructural level.

Turning to the material-semiotic level of writing, I redefend the position that language and matter are not in opposition. As aforementioned Palmer (2020) is particularly helpful here as she is one of the few materialist scholars who foregrounds language for mattering. In her book Queer Defamiliarisation: Writing, Mattering, Making Strange, the deliberate mattering of the main statements like “defamiliarization can be queering can be mattering” and “rewriting can be refolding can be refleshing” articulate an example of what writing on a material-semiotic level can mean, that is to “conjure ‘a beyond’ by making language strange” (Palmer, 2020, p. 39). She claims,

The syntactical repetition beyond the regular analogy of two (A can be B) takes us out of our automatic perception of “transparent” words and reminds us of their thickness; their realness; their materiality. (p. 2)

This new materialist stance on language to “extract a radium of the word” ⁶ distinguishes it from other strands of posthumanism such as speculative materialism and object-oriented ontology in that it does not seek transcendental topoi outside the materiality of language; “(language’s) materiality itself is part of the signifying process” (Palmer, 2020, p. 41). Thus, joining Barad again at this point, “matter and meaning do not pre-exist, but rather are co-constituted via measurement intra-actions” (Barad, 2012, p. 6). One of Palmer’s material-discursive, material-semiotic strategies, then, is “queer defamiliarization” as a reconfiguration of linguistic spatiotemporal axes—the syntagmatic-paradigmatic axes—by stuttering, repetition, interruption, equivocation, neologism and diffracted or spliced morphemes (Palmer, 2020, p. 35). She also writes on the “speculative taxonomy” of senses (p. 144), which gestures toward an “imaginative labour that new neural pathways are forged and new concepts are created” (p. 43). So, writing on a material-semiotic level consistently requires us to project symbolic supports and tune ourselves in co-natured spatiotemporalities with the linguistic matter. Palmer rightly calls this “wor(l)ding” (p. 106); worlding through words.

Finally, I would like to discuss writing on the level of zoe, the “the non-human, vital force of life, the transversal entity that allows us to think across previously segregated species, categories and domains” (Braidotti, 2018, p. 42). How might we understand writing in a vitalist position and in relation to mattering? I associate it with how certain modes of being come into vivid configuration by having fictitious “affairs” (An, 2016) or by rhizomatic composition through multiple “outward-bound interconnections,” as Braidotti (2013) writes on how a text comes to matter:

The “truth” of a text requires an altogether different form of accountability and accuracy that resides in the transversal nature of the affects they engender, that is to say the outward-bound interconnections or relations they enable and sustain. (p. 165)

What she is discussing here is non-linearity, and I add, fictitiousness of posthuman subjectivity as the key to constructing a sustainable reality on a collective level. I recognize how Braidotti describes writing as also useful in this context: “a method for transcribing cosmic intensity into sustainable portions of being” (p. 165-166). In my reading, the “sustainable portions of being” is the alternative, transversal subject beyond anthropos, which is zoe, and it comes close to what Guattari and Deleuze (1987, p. 40) conceptualize as “body without organs” in a geophilosophical context. Writing on the level of zoe, then, can be translated as an “intra-activity” (Barad, 2007) of bringing sustainable portions of vitality into existence.

Writing Creation Myths

In spring 2020, I taught a course titled “Creation Myths in the Digital World.” This section describes the course and examines the processes and outcomes as examples for further discussion of the theoretical approaches discussed above. A dozen Master’s-level students from the Architectural Department gathered to invent their own creation myths that articulated a sense of who we are in the world now. Instead of writing from scratch we began by collecting eighty creation myths from the internet. ⁷ While conventional computational approaches would then incline to classify or naturalize the myths around specific parameters of interest, we created maps of word-clusters using a machine-learning algorithm called Self Organizing Maps (SOMs), thus without any external grammar, ⁸ which acted as a basic material-logical infrastructure for mattering with the eighty myths.

Naturing With Big Data

In summary, I have sought to practice a posthumanist ethics of mattering by treating big data as linguistic matter while foregrounding writing as a powerful act of mattering. The stakes were on pursuing an affirmative mode of inquiry to transform the predominant naturalisms and dualistic frames in qualitative research with big data. In writing their own creation myths, the students demonstrated their own modes of mattering, which help us to see how the suggested approach could work out.

In closing, I explore what an ethics of naturing with big data might be about. This is only a short sketch of a future investigation, but I believe it addresses important findings of the article. I propose that the notion of nature—both as a noun and a verb, just as matter works as both—can stretch the idea of mattering on a more sustainable level. What I recognize most strongly after discussing the final creation myths with the students is that mattering itself does not mean that knowledge is produced; the performativity of complex entangling has to be, in certain manners too, taken great care of. Mattering does not always end up with safe and sound reconfiguration: it can also be non-figuration, para-figuration, mal-figuration, and, I add, non-communication. As Savransky (2016) puts it, what we might need, in order to avoid the “double loneliness” and “relational reductionism,” is “careful crafting of delicate contacts and marvelous adjustments”:

. . . not just an affirmation of relationality as what may enact the on-going transformation of reality, but also a demanding, yet productive, care for the textures and patterns of the many modalities of relating that may become available in the coming-to-terms that characterizes and encounter. (p. 20)

To be sure, an ethics of mattering that is committed to new incorporeal becoming and knowledge-making would need “care for the textures and patterns” of how it comes to matter and by what modes of mattering. In addition, my intuition is that the notion of nature, following Emerson (1940/1836), can offer an adequate spatiotemporality to embrace many issues of care, location, vitality, inhabitance, and “endurance” (Savransky, 2016) on a sustainable level. That is, when nature is seen as “a larger reality,” an ongoing material and symbolic world, and “a training ground” (Emerson, 1940/1836), naturing could be about accommodating different modes of mattering and knowledge-making—not only materially but also symbolically and rhetorically—so as to proliferate meaning and matter in abundance. Finally, to cultivate an ethics of naturing with big data, while standing on the Baradian non-essentialist ground of “ontological indeterminacy, a radical openness, an infinity of possibilities,” I contend that researchers should carry out more investigations using rhetorical experiments. Rhetoric, that is, an art of transformation and communication, could cherish careful crafting and a plethora of “naturing technics” for many a figuration with linguistic matter.