Digital humanities: Mission accomplished? An analysis of scholarly literature

Abstract

The field of digital humanities (DH) has evolved throughout the parallel evolution of computers, software and networking techniques, as well as the different attitudes of interested scholars. Since the earliest historical phases of this research field, scholars have been debating whether it can be considered as a new academic discipline and whether it is revolutionary in nature. About 20 years ago, the early denotation of ‘humanities computing’ evolved to the present label of DH, and deep changes occurred in digital information technologies, as well as in their humanities applications. Meanwhile, dedicated academic curricula were launched, thus adding an argument in favor of the debated disciplinarity of DH. This paper gives an account of the relevant scholarly debate, distinguishing between the early period and the most recent years; it then tries to frame this process in a model of scientific revolution.

1. Introduction

Digital humanities (DH), ‘a broad designation that many might know, but fewer understand’ (Pacheco, 2022), is a relatively recent field of research. Most historical accounts find its origin in 1949, the same year when the first stored-program computer was realized, in the collaboration between Father Roberto Busa and IBM for a computer-assisted compilation of word indices and concordances in St Thomas Aquinas's works. Electronic computing thus immediately became a working tool for humanities.

The linguistic computing centers and the joint projects funded in subsequent years led the humanists who were collaborating with computer scientists to develop quantitative approaches to their problems and originate what was then called humanities computing (HC) (Berry, 2012). Applying quantitative methods in fields in which the qualitative approach was the almost exclusive tradition raised discussions within the new community, as well as between them and traditional humanists; this produced a wave of scepticism motivated by the fear that the new methods would contaminate the ‘genuine’ humanistic disciplines, or that humanities may ‘sell their essence to technology’ (Pacheco, 2022). The growing HC community also launched academic and professional associations and, in 1966, the first dedicated journal, Computers and the Humanities (CHum) (Raben, 1966); this was followed in 1986 by Literary and Linguistic Computing (LLC).

The Digital Humanities Manifesto (Schnapp and Presner, 2009) describes the quantitative approach as the characterizing feature of the so-called first wave of DH. A second wave is more ‘qualitative, interpretive, experiential, emotive, generative’. According to Berry (2012), a third wave would be needed, centered on the notion of computational literacy and dealing with potentially new forms of literature and the media that support them. Indeed, the scope of DH has been progressively broadening, first by extending its concerns to disciplines other than linguistics and textual analysis, and then to including problems such as the production of new humanistic objects (e-books, web pages, thematic maps, etc.) and the study of other media, such as images and video and audio recordings, in addition to the classical scholarly subjects. This process was greatly accelerated by the availability of personal computers in the early 1980s and subsequently of the internet and the World Wide Web from the late 1990s.

The terminological switch from HC to DH occurred as an effect of this evolution, ¹ emphasizing the fact that this research is no longer viewed as a branch of computing but as the subject of a new academic discipline (Sula and Hill, 2019). In fact, scholars have long debated whether DH is a true academic discipline (Brink, 1990; McCarty, 1998; Burnard, 1999; Papadopoulos and Reilly, 2020; Drucker, 2021; Orlandi, 2021; Pacheco, 2022). A related long-standing debate is on interdisciplinarity, multidisciplinarity, transdisciplinarity (McCarty, 1999; Schreibman et al., 2004, 2016; Berry, 2012; Darbellay, 2019; Jacobs, 2021), or even postdisciplinarity (Schnapp and Presner, 2009). Also relevant to our purposes is the question on whether HC/DH represents some kind of scientific revolution. Many authors agree on a positive response, such as Brink (1990), Schreibman et al. (2004), Ess (2004) and some scholars interviewed by Nyhan and Flinn (2016), to name just a few. Berry (2012) even states, albeit without conducting a detailed analysis, that DH ‘could’ represent a science revolution in the Kuhnian sense. Some recent contributions treat this issue more problematically (see Keeler, 2002; Pierazzo, 2016).

It can be said that the scholars involved in HC were concerned about whether their mission was to found a new discipline, and whether this would have a revolutionary nature with respect to traditional humanities. To seek an answer, Salerno (2002) presented an account of the evolution of HC based on early scholarly communication, particularly on the 1966–2000 papers in CHum, exposing the attitudes of (computing) humanists towards the mentioned issues. He also tried to frame the alleged revolution within the Kuhnian model. The conclusion at that time was that a new discipline was probably about to be launched, but a Kuhnian revolutionary process was not complete. A further 20 years have passed now, however, and many things have changed.

First of all, the mentioned predominance of textual analysis among the research subjects has now reduced. For example, audio and visual data, cinema, music, dance, the production and study of born-digital content and computer games, biometry, geography, cartography and geographic information systems have become part of DH (Zeng et al., 2022; Dibeltulo et al., 2020; Hong and Wu, 2022; Escobar Varela and Hernández-Barraza, 2020; Bailey-Ross et al., 2017; Salah et al., 2021; Wei et al., 2022). Even recently, however, some privilege is reserved for text: Frabetti (2012) notes that DH is generally considered to embrace all the activities that draw their methods from computer science, such as image processing, data visualization and network analysis, ‘to produce new ways of understanding and approaching humanities texts’. Another aspect of the evolution involves the commercial exploitation of software and data originally devoted to academic purposes that are provided to either the same academic community or the public at large.

Steps towards humanistic disciplines, especially archaeology and cultural heritage, were also originated from different communities, often leading to new technologies and ‘publishable computer science research’ (Kirschenbaum, 2002). For example, the Journal of Cultural Heritage (Guarino, 2000) and Heritage Science (launched in 2013) were born within the chemistry community, the International Journal of Document Analysis and Recognition (Doerman et al., 1998) originated in the pattern-recognition community, and the Journal on Computing and Cultural Heritage (Arnold, 2008) came from the computer-graphics community. This variety of applications gave rise to a number of specialized subfields, each with its reference subcommunity.

A list of DH journals compiled in 2019 (Spinaci et al., 2019) counts 19 journals ‘exclusively’ and 17 journals ‘significantly’ dedicated to DH, out of a total of 104 titles. Meanwhile, between 2002 and 2005, numerous HC associations joined in the Alliance of Digital Humanities Organizations. ² In 2005 and 2015, respectively, the two main journals in DH, CHum and LLC, changed their titles. Curiously, the one that suggested a more specific scope, LLC, became Digital Scholarship in the Humanities (DSH), ‘in an effort to rebrand to a wider audience’ (Sula and Hill, 2019). Conversely, the one that appeared more general in scope, CHum, became the specialized Language Resources and Evaluation (LRE) (Ide and Calzolari, 2005). Schreibman et al. (2004) edited a volume describing the evolution of DH in its different applications and founding principles, gathering the contributions of many scholars that were ‘brought together to consider digital humanities as a discipline in its own right’. A second book from the same editors (Schreibman et al., 2016) gives an account of more recent developments, treating the different aspects of the discipline and concluding with a section on the past, present and future of DH. A new journal, Umanistica Digitale, was launched in 2017 by the Italian Association of Digital Humanities ‘as a discussion venue for topics pertaining to the Digital Humanities, ranging from the theoretical and methodological foundations of computational models in social science to the development and application of computational systems and digital tools in the humanities’. ³

Importantly, the early 2000s were also marked by the birth of the first academic curricula in DH. An assay of the difficulties encountered in this development is given by Rockwell (2003). In Italy, the first undergraduate curriculum in ‘informatica umanistica’ ⁴ was started in Pisa in 2002, and this was soon followed by a corresponding master curriculum (Salvatori et al., 2023). Whether this is the conclusive episode of a scientific revolution is one of the questions addressed in this paper.

First, the objectives and methods of this research are stated based on a close reading of the scholars’ opinions in the literature. Then, results briefly analysing the changes in scholarly communication and the debate on the two cultures are reported. Finally, the two basic questions raised here—about disciplinarity and revolution—are examined, and an attempt is made to find a synthesis among the different positions.

2. Objective

By comparing opinions in the literature in the past 20 years, this paper seeks to complement the analysis presented by Salerno (2002) on how the scholars consider the issues of the (inter/multi/trans/post)disciplinarity of DH and its revolutionary aspects. Their opinions emerge in what they write in journals, either in opinion/position papers or between the lines of more technical contributions. Interviews, as in the work of Nyhan and Flinn (2016), and seminar talks, such as those of McCarty (1999) and Burnard (1999), when available, can help with completing the picture.

Regarding the questionable disciplinary status of DH, this paper summarizes the ongoing debate, emphasizing the most recent opinions and trying to draw some conclusions. As far as a digital revolution in humanities is concerned, an attempt is made to fit the state of the art into the Kuhnian model in the light of the most recent events. In particular, it is perhaps time to ask whether the ‘renewed’ humanistic disciplines are now in a phase of Kuhnian ‘normal science’.

3. Methods

To find a significant sample of the opinions of scholars who reflect upon themselves and the peculiarities of their community, we first observe that these opinions are often stated explicitly in CHum and DSH. Other journals, including those launched by non-humanities communities mentioned above, are not so rich in this sense. Conversely, contributions in books such as those of Schreibman et al. (2004, 2016), Berry (2012) or Salvatori et al. (2023) often contain useful material. To update what was reported by Salerno (2002), this research first examined all the papers published in CHum from 2001 to the end of that publication in 2004. Consideration of more recent contributions relied on the papers in DSH—the journal that seems to continue the spirit of CHum—from its foundation in 2015 to 2022. During the analysis of these papers, several further contributions were included in the reference set, including books, conferences and seminars, which contain useful indicators for our purposes. As a result, some 150 contributions were selected to be examined more thoroughly. These works were further scrutinized to identify the most relevant opinions of digital humanists about the problems of interest. Only these sources have been analysed carefully and are cited in this paper.

During the past two decades, as has been noted, a number of DH-related journals were founded by non-humanities communities. It is now worthwhile to take them into account to complement our information on the current status of DH. A complete panorama is out of our present scope, but considering even a small sample of published papers can be useful. A shallow analysis was performed on the Journal on Computing and Cultural Heritage (JOCCH), appearing in the list of journals ‘exclusively’ dedicated to DH mentioned in Spinaci et al. (2019). Generally, the papers are not dedicated specifically to the themes dealt with here, so extracting the authors’ opinions is not easy. However, as the relevant topics of this journal are considered to be part of DH, some statistics considering the number of authors per paper, their main disciplines and the most appropriate index terms can be used to highlight possible specificities.

4. Results

In this section, the early period of DH is recalled to put the issues treated in the right frame. The first and second subsections are largely based on early materials. The reader can also find useful information and details regarding the years before 2000 in other works (see Brink, 1990; McCarty, 1998, 1999; Burnard, 1999; Kirschenbaum, 2002; Sula and Hill, 2019). The most recent contributions are instead used to pursue the main goal of this paper; that is, in the third and fourth subsections, to analyse the debate on disciplinarity and revolution that has occurred since the early 2000s and to try to find an adequate synthesis thereof.

4.1 Evolving scholarly communication

While analysing the early contributions in HC literature, their differences with respect to the papers published in scientific and technical journals are apparent. The pioneer journal CHum was originally conceived as a newsletter, featuring scholarly papers as well as initiatives such as conferences and projects, book reviews and software for humanities applications. During its first five or six years, say, until 1972, the format of the papers was different from that characterizing the papers in scientific journals: an abstract was not always provided, and the text was not always divided into standard sections; furthermore, the bibliographic references did not follow a fixed standard. These features are typical of humanities production: a humanistic text needs to be read carefully and understood thoroughly; the typical reader of this kind of paper normally does not need an abstract and does not need to quickly retrieve technical or quantitative information. Furthermore, typical humanities scholars used to work alone, or at least outside large collaborations. Thus, most papers in the early volumes of CHum feature single authors. Things, however, were gradually changing.

Lessard and Levison (1998) reported that, in the first five volumes of CHum, the proportion of single-author papers was 92%, whereas in volumes 26–30 (1992–1997), it was 63%. Compared to literary journals, in which single-author papers accounted for more than 95%, this difference was remarkable; however, it was also huge when compared to a sample of technical/scientific journals, in which that ratio tended to range between 5% and 35%. ‘The sociology of our research’, they conclude, ‘is closer to that of the humanities than that of the sciences.’ They continued, ‘If we do insist on using the scientific approach, perhaps we should consider developing more research teams, including specialists from statistics, computing science, linguistics, and whatever other field might be of benefit.’

The statistics provided by Sula and Hill (2019) confirm an increase in the number of authors per paper until 2004, and this trend is further confirmed for the subsequent years by the literature analysed here. Sula and Hill also provided information about the disciplines to which the authors belonged, thus confirming the wishes of Lessard and Levison (1998): in both CHum and LLC, the number of first authors in computing and computer science is just slightly smaller than the number of first authors in English language and literature, while the total number of authors and co-authors in computing and computer science is remarkably larger than all the others. Furthermore, a significant number of authors are from mathematics, statistics, engineering and science backgrounds. This result is highly relevant, since Sula and Hill (2019) considered all the papers published in CHum from 1966 to 2004. It is likely that, if the most recent years were disentangled from the timeline, the result would have been even more impressive. In the view of Pacheco (2022), supported by other recent contributions, the field of DH is now fully collaborative. Although the quantity of non-text media considered has been increasing with time, Sula and Hill (2019) reported that, in 2004, papers dealing with text made up 59% of those in CHum and 72% of those in LLC. In the present work, looking at the papers published in DSH during 2021 and 2022, it is estimated that about 63% of papers dealt with text; this means that the situation has not changed much in the past two decades.

As far as JOCCH is concerned, analysing the 27 papers with no fewer than 27 citations, ⁵ we find that the influence of humanities in computer science has probably not been as deep as the influence of computers in the humanities. These papers show 124 unique authors and co-authors, with a median of four per paper in a range of one to 20. The median of the citations received is 42, in a range of 27 to 333, and the dominant index terms are ‘computer graphics’, ‘human–computer interaction’ and ‘information systems’, making up a total of 14 papers; this is followed by ‘archaeology’ and ‘arts and humanities’, with just two papers each. The departments of provenance of the authors are mostly scientific: computer science and the various branches of engineering occur in 95 cases. Only 24 authors belong to departments such as archaeology, technologies for cultural heritage and humanities, maritime civilizations, paleontology, anthropology, art, creative and cognitive technologies, history, linguistics and psychology.

In summary, computers in the humanities have modified scholarly communication, in that new journals have appeared, and the format and authorship of papers have evolved towards a more ‘scientific’ attitude. An analysis of authorship in the journals launched by the humanities community revealed an increasing collaboration between humanists and computer scientists, including all the related or implied disciplines, such as statistics, mathematics and engineering, even though first authors in humanities are still slightly dominant. As far as the journals born in the technological communities are concerned, the analysis should be deepened, and more journals and other sources should be considered.

From a limited analysis of the most visible papers in one of the most representative journals dealing exclusively with DH, the situation seems to be nearly the opposite. The dominant disciplines remain computer science and engineering, and the collaboration between humanists and computer scientists does not seem so pronounced as in the former case: most papers feature first authors coming from computer science, and specialists in the target humanities applications are rarely included in the author lists. This could mean that the collaboration between humanities and informatics in those cases is rather limited.

4.2 The two cultures

As soon as Charles Snow gave his famous Rede Lecture at Cambridge (Snow, 1959), HC was seen as a promise to bridge the gap between the two cultures, in that HC encourages the combination of the two cultures ‘to bring the rigor and systematic unambiguous procedural methodologies characteristic of the sciences to address problems within the humanities’ (Hockey, 2004). Some of the early reactions to this conjecture are recorded in a survey titled ‘The Two Cultures’, in which some Italian humanistic personalities express their opinions about whether HC actually helps to reconcile science and the humanities (Morando, 1961). Aside from a few positive reactions, most of the responses exhibit the sense of scepticism outlined above.

According to Burnard (1999), HC began as an ‘empirical reply’ to the debate on the two cultures. More recent positions, with the new label of DH definitely in use, include Rommel (2004), who says that literary computing has brought the two cultures into contact, and Porsdam (2013), according to whom DH is a ‘hybrid meeting point between the two cultures’, but the assumption that ‘quantitative research can handle everything which the humanist must take into account’ must be rejected, thus agreeing to her interpretation of Schlesinger (1962) and the criticism to Snow made by Leavis (1962). In the opinion of McCarty (2016), what DH inherits goes far beyond the debate about the two cultures: ‘It inherits many centuries of now relevant work that has been foreign to the humanities since Galileo.’ To Montfort (2016), closing the gap between the two cultures would entail that ‘programmers’ should learn something about the humanities as well. Edmond and Lehmann (2021) treat DH as an interdisciplinary research area, particularly in relation to ‘big data’; they argue that, in practical experience, communication across different epistemic cultures is neither easy nor smooth: ‘computer scientists showed a reluctance to discussing what certain key terms might mean or imply, a lack of precision that would surely draw criticism in a purely humanities context’. However, ‘when reaching across the boundaries of disciplinary norms and epistemic cultures become central rather than peripheral to progress, creative and productive compromises can be found’.

There is no doubt that ‘creative and productive compromises’ have been found. However, despite the fact that many authors maintain that the gap between the two cultures has definitely been filled, it seems to still exist in many respects. More specifically, considering the evolution in scholarly communication sketched above, it does not seem that the ‘programmers’ have learned much about the humanities, although the HC community ‘has had a hand in some of the most important developments in information technology’ (Flanders and Unsworth, 2002).

4.3 Is DH an academic discipline?

The debate about DH as a discipline is not as old as the one about the two cultures. In a series of contributions in Morando (1961) titled ‘Electronics and literature’, the problem is not addressed explicitly. Even years later, when the launch of CHum represented an essential step towards the establishment of a new discipline, neither Raben (1966) nor Milic (1966) spoke explicitly of HC as a discipline, rather preferring to name it a ‘community’. Apparently, even though a research community was already established through some 200 literary computing centers worldwide and a dozen conferences on the subject (Nyhan and Flinn, 2016), a consciousness of the formation of a new discipline was not yet mature. Indeed, this question has been debated for years, and it only recently seems to have been settled with a positive response (see Salvatori et al., 2023).

Ten years after his introductory editorial in CHum, Raben (1976) finally declared that this ‘scholarly area’ was a well-established discipline, although he warned against the hazards of scientism. As summarized by Salerno (2002), three factors enabled by the availability of computers were expected to change methods in the humanities: high speed, logic processing and storage capacity. To leverage the logic-processing capabilities offered by computers, the problems in the humanities needed to be formalized, which was not common at that time. Schreibman et al. (2004: xxiii–xxvii) stated that computing had provided the disciplines not only with tools, but also with ‘methodological focal points’. This change in methods could be considered as a basis for a new discipline, but the emergence of a discipline also depends on social aspects.

According to McCarty (1999), as soon as a research area is recognized as a discipline, an institutional change must occur, such that the interested scholars can be recognized academically and their community is allowed to grow. He presents HC as an ‘interdiscipline’, meaning that it exists ‘in the interstices of the existing fields’. As such, it is not ‘just another administrative entity’. What is first needed is a model for what is really interdisciplinary and then a method to draw from the disciplines whatever can be useful to HC. Since HC at that time did not entail any new administrative structure, the conclusion was that it was not an academic discipline. Rather than expecting the birth of a dedicated department, the ultimate response depended on fundamental changes in how academia as a whole is conceived.

Years later, in a thorough examination of its implied meanings, McCarty (2016) defined interdisciplinarity as an ‘abstraction’, saying that much more energy has been spent to decide what it is (‘a Glasperlenspiel’, in his view) than to investigate the how; that is, how the DH discipline—and the DH scholar—can really become interdisciplinary. Contrary to the view of Klaassen (2020), to whom interdisciplinarity is basically ‘the ability to listen to one another, [and the] willingness to learn from one another’, in the eyes of McCarty, it is something that should be earned individually by the scholar. Klaassen's basic definition entails that the humanities have an essential role in the provision of communication ‘that is key to capturing knowledge production and the dissemination of insights towards relevant fields’. DH, extending this concept, needs its fully humanistic nature to find a motivating common ground, be it pertaining to any form of interdisciplinarity or leading to ‘new disciplinary boundaries’.

Burnard (1999) agreed with McCarty that an academic discipline is ‘an organizational, bureaucratic concept’, determined by sociopolitical considerations (see also Turbanti, 2023). Even though many scholars believe that an underlying theoretical framework is necessary, there are many examples of successful and established theory-free disciplines, as well as research fields endowed with strong theoretical bases, that have never been recognized as disciplines. Therefore, Burnard concluded that HC is an academic discipline as many others. One of the arguments he brought in support of his thesis was, again, interdisciplinarity: HC is intrinsically interdisciplinary, as it encourages a ‘holistic’ vision of the visual, aural and linguistic aspects of artefacts. With no harm to the traditional introspection characterizing humanities scholars, HC is methodologically focused. Edmond and Lehmann's (2021) perspective does not agree with this account, rejecting the vision of a discipline without a theory: ‘interdisciplinary co-operations often failed because of the lack of a shared theoretical framework’, and ‘data without theory is as problematic as theories without evidence’. Bradley (2019) reported complaints about ‘the lack of theoretical underpinning behind toolmaking for humanists’.

A recent account of the developments that have occurred in modern universities is provided by Wernli and Darbellay (2016), in which the presence of institutional difficulties in recognizing interdisciplinarity is still highlighted, since the system is primarily built for disciplinarity. Nyhan and Flinn (2016) noted that the very definition of a discipline is not agreed upon by everybody and, somehow resembling Burnard's position, said that DH could be considered a discipline because it has the characteristics provided by such a social construct; that is, the presence of university courses, academic journals and scholarly societies. This notwithstanding, many authors still reject the idea on grounds related to proper accreditation, the existence of very few autonomous departments, the lack of senior scholars agreeing to collaborate in interdisciplinary research, and the very sparse variety of subjects and personalities allegedly belonging to the community.

An attempt to define HC as a discipline by identifying the common features of the research work was made by Unsworth (2000) using the concept of scholarly primitives: a number of functions that are shared among disciplines and are independent of any theoretical orientation. An extended and hierarchized set of primitives was recently proposed by Pacheco (2022) to theorize the fundamentals of DH from a mixed qualitative/quantitative literature review. Accounting for the evolution from HC to DH, he recognized that only recently has the latter entered the academic organization through new courses and graduations, and that it will probably take further time to see its influence on academia and society at large. Regarding interdisciplinarity, Pacheco evokes a ‘dilution of frontiers’ between classical studies, literature, languages, computer science and data management.

Even today, however, while discussing whether DH should be considered a science, an interdiscipline, a frontier science or whatever else, the lack of a theoretical debate can lead to an uncritical acceptance of the reality or an unjustified fear of the new technologies. Hughes et al. (2016) used the scholarly primitives to identify the common methods in DH. Put simply, these should be computational—that is, either based or critically depending on information technology—and located at a ‘key point of intersection’ between disciplines, thus enabling otherwise impossible research to be conducted. From the scholarly primitives, Palmer et al. (2009) identified what characterizes interdisciplinary work through a graph including the fields of ‘humanities’, ‘sciences’ and ‘interdisciplinary’, showing their mutual intersections, similar to the concept of humanities methodological commons proposed by Hughes et al.

‘Interdisciplinarity’ is the most-used term referring to DH, meaning that this field is placed between disciplines. Two other terms found in this study, ‘multidisciplinarity’ and ‘transdisciplinarity’, respectively, seem to have the subtly different meanings of including many disciplines and going beyond one discipline to steal something from another one. The term ‘postdisciplinarity’, as used in the Digital Humanities Manifesto (Schnapp and Presner, 2009), seems to mean something more; namely, that the production of knowledge does not derive any longer from disciplinary approaches. Darbellay (2019), aside from providing a deeper insight into the meanings of the different prefixes to ‘disciplinarity’, attaches a high revolutionary potential to postdisciplinarity, since the other definitions are all based on the central notion of disciplinarity; that is, on the existence of more or less rigid boundaries between different epistemic, methodological and bureaucratic communities. He contests the notion of discipline as given by nature, since it has been developing only during the last one or two centuries (see also Tavosanis, 2023; Turbanti, 2023). However, the Manifesto states that postdisciplinarity can foster ‘disciplinary cross-fertilization’; that is, as we understand, the disciplines evolve through mutual exchanges of paradigms and methods but, anyway, are still alive and necessary to the advancement of knowledge. The Manifesto also maintains that DH is ‘not a unified field, but an array of convergent practices’.

Looking at the variety of new scientific journals more or less relevant to DH, it can be conjectured that the initial convergence among different humanistic disciplines fostered by the use of computers has evolved in the sense that they have gained new approaches and insights, thus renovating their statuses. This can explain the multiplication of specialized communication sources, and also the fact that, as observed, even DSH, a journal with a supposedly general attitude, features papers in a few privileged fields. The increased attention to quantitative aspects has shown that many problems, even in different disciplines, can be treated using the same approaches, and DH could thus be considered a ‘cross-discipline’ rather than an ‘inter-’, ‘multi-’, ‘trans-’ or ‘post-’ discipline (Palmer et al., 2009).

After the flourishing of HC centers, DH also started to establish administrative structures by launching specific academic curricula, with chairs, courses and faculties. The DH Course Registry ⁶ lists 128 bachelor's, master's and PhD programs worldwide, excluding individual courses, modules and summer schools offered by several universities. Whereas the success of these initiatives was ‘obviously unknown’ in their early lives (Salerno, 2002), the growing number of academic curricula and their survival over the years demonstrate that, first, the attempt to provide an intrinsically multidisciplinary academic initiative with adequate administrative structures has been successful; second, the curricula in DH have attracted a relevant number of students, which is also an index of their actual career perspectives (Young-Powell, 2023). Thus, the shared bases of DH are now accompanied by administrative structures and academic curricula, but the debate on its disciplinary status is still open. The evolution of HC into DH convinced many humanists to value the digital approaches to their research, but some of the original scepticism persists. Nothing has been simple, and a number of issues are still open. Regarding his experience in Pisa, Lenci (2023) says that humanities informatics had to find its narrow room between the two classical strongholds of the humanities and computer science, but its success has demonstrated that its initial motivation was well devised.

According to Orlandi (2021), the humanities should not include social sciences, ⁷ even though linguistics—which is ‘indispensable’ to him—is often considered a social science. Apparently, Spinaci et al. (2019) do not agree with him: the journals dealing ‘significantly’ or ‘marginally’ with DH include titles referring to linguistics and sociolinguistics, literature and literary criticism, philology, librarianship and information science, philosophy, ethics, sociology, history, art history, music and musicology, and other fields. Applying methods from information technology to all these disciplines is now quite common. Did these disciplines merge in DH or, rather, did they partially change their methods and objects of study, including digital tools in their daily practice? In the latter case, is this a revolutionary change or does it just parallel what happened in physics, chemistry or mathematics, for example, in which no one has thought to add the adjective ‘digital’ to mean that informatics has become a usual research tool?

In the preface to Schreibman et al. (2016), the authors foresee a time when the modifier ‘digital’ will have become pleonastic in humanities. Pacheco (2022) agrees with them and believes that, at present, that modifier is necessary to mark a methodological transition, and it will eventually be dropped. Robertson (2016) maintains that several disciplines are contained in DH, and they have not erased their differences; he looks at DH as a house with many rooms—‘entry points to central spaces where those from different disciplines working with particular tools and media can gather’. This position can be compared to that maintained by Turbanti (2023). Papadopoulos and Reilly (2020) say that DH has become a collage of disciplines, including several fields pertaining to social sciences.

From these views, in summary, it can be said that many different disciplines now share digital methods to pursue their aims, but each maintains its specific research objects and problems, and their evolution is conditioned by the new technological possibilities. ⁸ A pragmatic synthesis of the different positions could resume the above example of physics and other scientific disciplines and extend it to their relationships with engineering. Just as engineering includes mathematics, physics and other disciplines in its methods to provide something useful to society at large but also to mathematics and physics, so too does DH leverage information technology and other sciences to provide something that can be useful to society at large as well as to humanities research itself. If engineering is a discipline in its own right, so too is DH.

4.4 Did humanistic disciplines experience a revolution in recent decades?

In Morando (1961), a qualitative change in humanities caused by the use of computers was envisaged by several authors, including Busa (1961) and Pacifico (1961). Nyhan and Flinn (2016), in accounting for the early debate about the revolutionary nature of DH, say that, although the term ‘revolution’ has been used largely in the literature, detailed discussions about it are seldom found. Many authors in the works of Schreibman et al. (2004, 2016) refer to DH as a revolution.

Salerno (2002) tries to include this process in the model of scientific revolution proposed by Kuhn (1962). In summary, this model assumes that the development of science is neither cumulative nor linear: successive historical phases of scientific knowledge do not entail a progress towards something (e.g., a deeper knowledge of ‘the truth’); rather, they mark some progress from something (Kuhn, 1962); that is, from a phase in which a certain scientific paradigm is in force to a phase in which a new paradigm comes into effect. To Kuhn, a scientific paradigm is a set of beliefs, conceptual tools and permitted problems shared by all the members of a scientific community during a historical phase called ‘normal science’. The shift from one paradigm to the next is triggered by some crisis—not necessarily related, for example, to some failure in experimental validation, but rather to the emerging impossibility of continuing the ‘puzzle-solving’ activity that characterizes normal science. Different paradigms are incommensurable: apparently similar concepts referred to different paradigms cannot be compared, and the most recent cannot be considered an extension or a refinement of the previous one. A dialectical phase accompanies the transition between two normal-science periods, during which more paradigms coexist until only one of them predominates.

In reality, studies in the humanities do not proceed as those in the sciences: it is difficult to identify a period of normal science in humanistic studies. ⁹ The humanities community is trained historically, through the original sources, and not through ahistorical manuals, which instead dominate education in the sciences and are completely rewritten at each change of paradigm. Kuhn (1962) excludes his model from being appropriate for humanities. However, the concept of paradigm is not so rigid as to prevent an extension thereof from being applied to humanities; ¹⁰ furthermore, many authors in DH, as well as in science, value the Kuhnian model as being representative of their vision.

Brink (1990) has no doubt about the revolutionary nature of HC, adding some aspects of what Nyhan and Flinn (2016: 259–270) call ‘the motif of the underdog’; that is, the diffuse feeling among scholars of being misunderstood and marginalized. Nyhan and Flinn also treat ‘the motif of revolutionary’, which is very popular among the opinions of digital humanists. These two motifs are often put forward together and, in some sense, they could be considered as foundational myths: evoking a revolutionary present can be useful for providing the discipline with connections to previous successful revolutions, for example when comparing the use of digital tools and concepts in humanities to the Gutenbergian print revolution. Presner (2010) makes explicit reference to the Kuhnian model, also examining its premises and consequences: ‘Digital Humanities 2.0 introduces entirely new disciplinary paradigms, convergent fields, hybrid methodologies, and even new publication models that are often not derived from or limited to print culture.’ Suggestively, he also foresees the emergence of a new ‘normal humanities’, apparently not wondering about the existence of any old one. We are, says Presner, ‘at the beginning of a shift in standards governing permissible problems, concepts, and explanations, and also in the midst of a transformation of the institutional and conceptual conditions’. That is to say: we are experiencing a Kuhnian paradigm shift.

Hughes et al. (2016) examine the digital techniques as used in the diverse disciplines and talk about changes in paradigms that create new knowledge, in particular, ‘asking new research questions’; that is, being part of the shift characterizing a Kuhnian revolution. McCarty (2004) speaks explicitly of revolutionary changes, sharing the Kuhnian view with his colleagues. He also treats the supposed revolution in the context of his vision of interdisciplinarity, which is perhaps the main transformative aspect in the recent history of the humanities (McCarty, 2016). Papadopoulos and Reilly (2020) identify the old paradigm as ‘the time-honoured heroic paradigm in which curiosity-driven, professional singleton scholars are privileged but securely tethered by tenure and scholarly tradition to an academy’, and, ‘At the far end of DH practice, we imagine an environment of knowledge pluralism engendered by promiscuous crowds of independent knowledge workers operating under an open, and overtly socially inclusive, ethos.’ Regarding the new paradigm, however, they complain about ‘knowledge workers … more or less unaffiliated, and thus unfettered by traditional career paths and ties to specific institutions, working on short-term contracts with no expectation of career progression within the contracting organization’.

Papadopoulos and Reilly are not the only authors to fear exclusively optimistic views. Porsdam (2013) states: ‘the many new developments within DH must be discussed with a view not only to their potential, but also to their limits. Whether or not it may properly be classified as a Kuhnian paradigm shift—and the vote is still out on this—the digital turn and the involvement with computer-based approaches will cause substantial changes for both the teaching and the research in the humanities’. To Schnapp and Presner (2009), the revolution is about expanding the quality and the impact of knowledge in human sciences and the ‘direct engagement in design and development processes’. At the same time, they warn against the passive acceptance of technology—part of the so-called traditionalists’ response to the newly available tools offered by the internet. Critical remarks are also made by Pacheco (2022), who discusses different opinions on DH and complains about the lack of a thorough debate.

In the light of these opinions, we now try to frame the evolution of DH into the Kuhnian scheme. Its first feature is the crisis of a paradigm. The steps to verify this condition have been summarized by Salerno (2002), who found some cues in musicology (Bel and Vecchione, 1993) and lexicography (Spinosa, 1990). More recent contributions speak explicitly of a crisis in the humanities, not just in music and literary studies; they include, for example, Pacheco (2022), Thomas (2004), Ryan (2004), Van Zundert (2016) and Warwick (2016). To Presner (2010), a crisis was still present at the time of his writing, but he also foresaw ‘the imminent disappearance of one paradigm and the emergence of another’. For Claus Huitfeldt, interviewed by Julianne Nyhan (Nyhan and Flinn, 2016), indeed, the crisis is not even a crisis, as it has been going on for decades. There is a recent change, however, consisting of a new emphasis on ethical aspects of the use of technology rather than the evolution of methods and tools.

The second feature of the Kuhnian model is the paradigm shift. The problem remains in deciding whether such a thing as a ‘paradigm of the humanities’ really exists. ¹¹ We would find it rather easier to admit that each of the diverse humanistic disciplines has its individual paradigm, not excluding that they can also be grouped by paradigmatic similarities. Salerno (2002) identifies common traits in ‘intuition and trained mind’: mechanical analysis was refused by traditional humanists, and also by the early computing humanists, who accepted the use of computers in their research but did not even think about changing the foundations of their disciplines. Formalization can surely be considered one of the characterizing traits of the new paradigm, and its effects have surely contributed to some change of viewpoint in humanistic disciplines. Furthermore, quantitative and statistical analysis have found in computers the enabling technology needed to extend their use and make their results senseful.

In any case, considering the recent contributions testifying that a lively debate is still in place, it seems that the dialectical phase postulated by Kuhn is continuing among humanities scholars: a ‘normal humanities’ (fortunately) seems to be far from being realized. As noted, humanists are only marginally trained through manuals, so we cannot expect a fundamental rewriting and a subsequent abandonment of the current educational materials. Perhaps the dedicated university curricula will contribute to changing something in this respect: the modern digital humanists are more likely to have been trained through manuals than their predecessors. Although Salerno (2002) eventually declared that the paradigm shift was not complete, many things in the past 20 years have contributed to making our story fit better into the Kuhnian model.

5. Conclusion

This paper is an account of opinions regarding the disciplinary status and revolutionary nature of DH. The evolution of scholarly communication is also synthesized to allow each contribution to be placed in the right context.

The disciplinary status of DH has always been the object of lively debate, giving rise to a host of specific terms, such as ‘interdiscipline’ or ‘postdiscipline’, on the meanings of which not all the proposers agree. Compare, for example, the positions expressed by McCarty (1999) and Burnard (1999) to the recent analyses by Darbellay (2019) and Klaassen (2020). The launch of official curricula in DH contributes to corroborating the position, shared by many, that DH is rightfully an established discipline. Conversely, the number of recently founded academic journals dealing with applications of information technology to humanistic disciplines could be considered a sign that, in fact, different disciplines in the humanities have been absorbing new tools and methods, thus emerging as paradigmatically renewed but maintaining their specificities. However, the concept of methodological commons presented by Hughes et al. (2016) could be used to delimit DH as a unique discipline, if we accept that it is a brand-new discipline, exchanging themes and methods with human sciences just like engineering exchanges themes and methods with physics and mathematics without being either physics or mathematics.

Salerno (2002) tried to look thoroughly into the supposed revolution in the humanities, considering the aspects of the Kuhnian model that he deemed a sufficient fit to the humanities. The new facts as evolved since 2002 contribute to filling some of the entries in that model, except perhaps the notion of ‘normal science’, which seems by no means appropriate to the humanities. We could say that, even though new paradigms are being introduced, a Kuhnian phase of normal humanities will perhaps never start. As Keeler (2002) recalls, ‘asking how to capture the essence of what works well in the present in order to improve the future’ is more interesting than declaring a revolution.

Today, nearly all scholars make use of information technology; this does not, however, mean that all of them have become digital scholars. For example, a historian who uses informatics for the quantitative aspects of their research probably does not feel themself to be a ‘digital historian’ or a ‘digital humanist’. Despite the current aspirations to reach postdisciplinarity, individual disciplines are still alive and effective, even though their mutual boundaries are becoming less rigid, and they keep absorbing relevant technological innovations as they always did. DH, in turn, can be considered as a newly introduced discipline that is distinct from literary studies, linguistics, history and whatever else, but it is perhaps the best place for the whole of the humanities to effectively exercise the ‘ability to listen to one another, willingness to learn from one another’ mentioned above, and it is suitable, in my opinion, to form the most appropriate definition of ‘transdisciplinarity’.

Is the ‘mission’ of digital humanists accomplished? As has been shown, the fact that DH is an independent discipline is now almost unquestioned. Academic DH curricula seem to be quite healthy and growing, dedicated journals and communities are active and productive, and the influence of this new field is now well experienced by society (also in the form of new businesses). Is DH revolutionary in itself, or has it rather caused specific revolutions in the traditional humanistic disciplines? This is a more theoretical question, and we have seen here a variety of responses. Unlike Presner (2010), I think that, rather than a ‘normal humanities’, we could expect to see a ‘normal DH’ in the near future. ¹² Maybe this brings us to a question that is easier to answer; however, this would require a preliminary analysis encompassing the now very broad field of DH, not one that is limited to the scholarly contributions published in a few journals.