Are digital tools different?

Is the difference between analog and digital tools one of degree, or one of kind? In theory, any and all tasks done by a computer or a network of servers could be accomplished, given enough time and resources, by a human. However, many of these tasks would take an impractical or impossible length of time, which would prohibit these tasks from ever being accomplished. Computers enable such functions to be implemented and, by doing so, enable processes, open new lines of inquiry, and allow us to work at new scales. Digital tools displace memory and some cognitive processes outside of the human brain; they represent a step on what Katherine Hayles dubs biotechnoevolution: “a hybrid process in which information, interpretations, and meanings circulate through flexible interactive human-computational collectivities” (Hayles 2019, 32). Intermediation—the flow of information from one media to another, and from humans to digital media—allows for the emergence of complexity and new meanings (Hayles 2007; Hayles 2019). Hayles describes the paradigm-shifting potential of digital tools in these words:

The new wrinkle is the power of computers to perform cognitively sophisticated acts. Compared, say, to a hammer or stone ax, a computer has much more flexibility, interactivity, and cognitive power. In addition, computers are able to handle both natural language and programming code, capabilities that allow them to function in complex human-computer networks. (Hayles 2007, 102)

Research in the humanities is not simply supplemented by computers, it is fundamentally changed (Ingvarsson 2021; Hayles 2012); our contact with digital architectures informs, forms, enables, and limits us; “space is the context of the action: it makes it possible, and it shapes it” (Vitali-Rosati 2016, 96). However, these effects can be hard to detect when digital tools are normalized, sometimes to the point of becoming invisible.

The naturalization of digital tools

Science and media become transparent when scientists and society at large forget many of the norms and standards they are heeding, and then forget that they are heeding norms and standards at all. (Gitelman 2008, 7)

A tool is naturalized or essentialized when it stops being thought of as a tool, but as part of the normal order of things or as extensions of its user (Gitelman 2008). In terms of digital tools, naturalization takes two forms that both involve invisibilization:

1. A tool becomes invisible when its presence goes unnoticed or is taken for granted (Underwood 2014).

2. A set of design principles is naturalized when it becomes the default and all other possible design choices are thus invisibilized (Vitali-Rosati 2024).

In many cases, the naturalization process of digital tools is the result of design as much as use: the current trend is to produce hardware and software that are specifically ready-at-hand (Heidegger 2002) and follow the functional imperative (Vitali-Rosati 2024). Digital tools tend to blend in a larger digital ecosystem and sometimes even reinforce themselves as monopolies (Smyrnaios 2016). This is most easily seen in everyday life where we, for instance, Google things; Zoom call people; are proficient in Office; Photoshop pictures; Venmo cash; and Uber from one place to the next. The genericization process does not seem to be as pervasive within humanist research practices, although many assumptions can be made on which tools—hardware, software, internet services, databases—are used when co-editing (Google Docs or Microsoft 365), searching for secondary sources (Google Scholar or WorldCat), redacting a paper (Word), putting together a presentation (Powerpoint or Google Slides), reading an article (Acrobat Reader), etc. More important, for most research activities, one can safely assume the use of a screen, a keyboard, a mouse or a trackpad, and internet access. This naturalization process covers and invisibilize a set of design principles that underlie most editorial choices in digital tools. This set of principles that constitutes the affordances of digital tools is based on the orientation of a few companies that cater to corporations first and consumers second (Smyrnaios 2016).

Orientation and affordances of digital tools

Every system is infused with choices, formed by technical challenges, and directed towards precise usage: digital tools bear the traces of a developer’s, a corporation’s, or a community’s intents, interests, and limits (Vitali-Rosati 2016). For instance, there is a common set of design principles—such as ease of use, streamlining, and uniformity—that characterizes the GAFAM (Google, Amazon, Facebook [now Meta], Apple, and Microsoft), which makes them quite attractive to new and established users alike, leading a larger user pool to adopt hegemonic methods and technologies (Norman 2002). In Éloge du Bug, Marcello Vitali-Rosati focuses on the functional imperative, intuitivity, and the rhetoric of immateriality:

1. The functional imperative is the design principle according to which technology should “simply work.” It prioritizes efficiency and functionality over other considerations such as ethics or culture. This imperative leads developers and designers to limit freedom, autonomy, and agency, and it normalizes the invisibilization of the protocols and algorithms that support a seemingly “flat” technological infrastructure (Vitali-Rosati 2024).

2. The rhetoric of immateriality is a common aspect of the discourse around technology that sets the material and immaterial apart and establishes one as valuable and pure (the realm of idea), and the other as trivial and dull (the material) (Vitali-Rosati 2024). This discourse affects digital technology by encouraging designers and developers to obscure or bury the reality of materiality and labour under a product that seems seamless. Digital tools try to subdue themselves by asking of their users natural gestures and as little interaction as possible with the back end of their hardware and software.

3. Intuitivity is a design principle that pushes for the alignment of function with natural or previously established behaviours (Norman 2002; Vitali-Rosati 2024). Its main goal is to create systems that can be learned and used effortlessly by any user who is familiar with a given network of affordances. Intuitivity limits the scope of what applications and hardware can do to functions that fit in that previously established network.

The homogenization of digital tools outside of academia has an impact on research practices: software and networks developed for the corporate world or the consumer market now shape social sciences and the humanities, since many research tasks have been digitized. The design principles that direct search functions (Underwood 2014), text editors (Fauchié 2018), and editorial chains (Vitali-Rosati 2016), among others, exemplify how digital tools that were not designed with scholarly work in mind have been adopted by universities and the academic community. In many cases, the naturalization process has been manufactured: educational pricing leads to the introduction of corporate tools as soon as first grade, and some technologies like recommendation algorithms are so opaque that they are not forgotten, but rather unknown (and, in certain cases where machine learning is involved, unknowable). The GAFAM strategies are only some of many possible strategies in developing a digital tool, but they reveal their orientation, and thus the orientation of their tools. However, these values do not constitute the only possible set of design principles: other guiding principles can lead to wholly different approaches to digital tools with their own epistemic framework and implications (Fauchié 2018). But technological resistance, the use of tools that follow alternative digital approaches, is difficult and can require going against established practices or institutional decisions.

Case study: The theorization and naturalization of the search bar

In Theorizing Research Practices We Forgot to Theorize Twenty Years Ago, Ted Underwood discusses the “search” tools, a “deceptively modest name for a complex technology” (Underwood 2014, 64). In this article, Underwood describes how the search function has been normalized and naturalized within academia to the point where its results are almost never challenged. However, the principles that underlie a simple search function are numerous and far from neutral; from the editorial choices behind a database structure to the care with which keywords were chosen by authors and publishers, to the display interface of the recommendation algorithms that filter and order results, most search functions are black boxes with a hidden agenda. More importantly, these same recommendation algorithms are invisibilized and the search bar naturalized: results are unquestioned, the shape of the tool is taken for granted, and other forms of search are unexplored. Underwood’s article exemplifies the need for a general theory of scholars–digital tools interaction: a conceptual framework researchers can use to understand how their relation to their computers, software, and databases affect their work. If a tool as simple, hegemonic, and ubiquitous hides such complexities and orientations, every digital tool warrants similar scholarly scrutiny.

Variables of a heuristic toy model

Our “toy model” characterizes scholars along seven axes with values between 0 and 1:

• Usage: Represents how frequently a scholar uses digital tools compared to analog ones. It can be modelled as the percentage (between 0 and 100%, the later being represented by the value 1) of academic work conducted with digital tools rather than traditional, analog methods.

• Breadth: Reflects the variety of digital tools a scholar employs in their research. At one end, a scholar might use only a few digital tools like email and a word processor while relying heavily on printed material. At the other end, a digital humanist might write with a diverse array of tools such as HedgeDoc, Google Docs, Obsidian, Stylo, Visual Studio Code, GitHub Editor, and Notepad, switching between them based on specific needs or preference. This scale is more arbitrary than the first one; the following function takes the number of digital tools used on a weekly basis as its independent variable x, and yields a value close to 0 for 1 or 2 tools, and gradually increases, reaching 0.5 for 15 tools, and approaching 1 around 30 different tools:

  $f\left(x\right)\hspace{0.17em}\hspace{0.17em}=\hspace{0.17em}\hspace{0.17em}50tanh\left(\frac{x-15}{5}\right)\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}+\hspace{0.17em}\hspace{0.17em}50$

• Competency: Describes the level at which a scholar uses their digital tools. Most digital tools have functionalities that a casual user might not even be aware of. Competency is the ability to navigate specific tools beyond normative usage, or to master practices that either require advanced technical skills or have a high learning curve. This scale, also between 0 to 1, as well as the next ones, might require scholars to self-assess, or could be based on an extensive test of their general competencies with, knowledge about, and general attitudes towards digital tools.

• Knowledge: Measures a scholar’s understanding of the underlying principles, protocols, and algorithms behind the software and hardware that they use. For instance, a user with a very good knowledge of statistical methods could deploy these methods outside of the digital media and without specialized software.

• Criticality: Assesses the scholar’s ability and willingness to engage critically with digital tools and methodologies. This axis involves questioning the biases, limits, and assumptions, in other words, the affordances and orientations of digital tools and how they influence their work as well as themselves and others.

• Resistance: Captures the scholar’s skepticism or reluctance to use and adopt digital tools within their work. Resistance covers one’s preference for analog tools, biases against new technologies, and concerns about privacy, technological determinism, agency, etc.

• Bias: Represents the scholar’s posture towards digital tools and the results they produce. If they are more likely to reject information and knowledge produced through digital technologies, they are said to have an oppositional posture, and if they are more likely to accept it at face value, they have a hegemonic posture (we borrow these terms from Stuart Hall [Hall 2001]). For the sake of the visualization, we set a bias of 0 to be the hegemonic posture, a bias of 1 to be the oppositional position, and 0.5 to be a negotiated posture where some information is taken at face value and other is rejected depending on the context.

This framework enables us to distinguish axes that might otherwise be conflated, especially in the humanities where, for instance, the distinction between knowledge and competency is often blurred. In more technical fields, to know and to do are well separated: one can use a tool without knowing how it works, and one can know how a tool is supposed to work and be unable to operate it. Furthermore, to be biased against a tool is different from being critical of a tool, which in turn does not always correlate with rejecting a tool. Also, a hyper-user could only use a very small selection of digital tools, leading to very little breadth, but have high competency, and vice versa.

This model doesn’t capture more subtle aspects of a scholar’s relation to digital tools such as personal experience, cultural relation to technology, and scholarly goals, among others. It is a heuristic tool to enable the formation of a varied, multidisciplinary team that covers many different relations to technology in humanistic research. To overcome the limits of their epistemic horizon, one first must see its edges, and be willing to consider practices and épistèmes beyond their own. The development of a general theory of digital tools in humanistic research has to take into account the varied experiences and expertise of humanists; doing otherwise would be an epistemic mistake.

Four archetypes

To exemplify the model, we deploy four stereotypes—the vigilant, the tinkerer, the classicist and the adopter—and showcase their profiles using a radar chart with the seven variables of our toy model. The vigilant and the tinkerer can be found in Figure 1, while the classicist and the adopter can be found in Figure 2.

Figure 1

Radar charts for the vigilant and the tinkerer according to the seven axes (scale of 0 to 1).

Figure 2

Radar charts for the classicist and the adopter according to the seven axes (scale of 0 to 1).

• The vigilant tries to use only technologies they fully understand. This limits greatly the breadth of tools they are willing to use, but they have mastery over a few select software. They are motivated by data security, but also attachment to the computer they bought 15 years ago and fixed themselves too many times to count. The vigilant openly criticizes their institution’s choice of handing over their digital infrastructure to Silicon Valley, and they do not carry a smartphone.

• The tinkerer is a technological anarchist with little regard for proprietary software or copyright laws. They launch open-source applications from the command line on their Linux laptop, and when they cannot find the right software for the task at hand, they develop it themselves. The tinkerer is likely to have a pile of old, broken computers at home to scavenge from according to their needs, and only an angry editor can get them to sign up in Office 365.

• The classicist’s expertise does not lie in digital tools, which they use sparsely and would gladly replace by a typewriter. The classicist knows the most useful functionalities of Word (the 1997 version) and Outlook and has no intention of ever learning anything else beyond that. Their favourite piece of hardware is the printer, and they are strongly biased against e-readers.

• The adopter wants to use the best tool on the market, although it sometimes means paying for a subscription service. They are willing to not know what is in the black box, and they put the emphasis on results rather than process, which does lead to impressive feats of analysis on large datasets. They are not afraid of software with steep learning curves but would prefer not to code anything from scratch themselves.

The vigilant, tinkerer, classicist, and adopter are extreme cases in terms of their relation to digital tools—most humanists fall somewhere between these caricatures—but they try to capture different forms of expertise and, more importantly, different experiences of digital tools.

Limits and affordances of the model

This heuristic model focuses on measurable—or at least, comparable—characteristics to showcase the plurality of postures a scholar might hold towards digital tools. The main point is that these differences extend beyond a scholar’s horizon and which processes they entrust to technology. Indeed, this model ignores cultural patterns, which “are primarily found in three areas: within a people’s values, their patterns, and their institutions [and] establish the parameters within which a people think and speak” (Simonton 2015, 31). The model also assumes access to digital tools, which cannot be taken for granted, as well as the freedom of choice in which tools they use. (Institutional rules, publishing practices, material situations and other more or less authoritative structures can direct or limit these choices.) Hence, the model is only a steppingstone in understanding one’s own technological, cultural and cognitive horizon, acknowledging other practices and épistèmes, and in gathering a diverse team in the context of a multidisciplinary discussion—a plurilogue—which we describe in the next section.

Structure for a plurilogue

Editorial