In the previous post, I mentioned the issue of co-research between the humanities and digital technologies quite in passing. Today I will go deeper into this issue.
Usually, the digital humanities of the first kind use information technologies as an auxiliary collection of techniques to help solve humanistic problems. This is not unlike archaeology using radiocarbon dating or art history using analytical chemistry. In other words, the research questions that we often see in digital humanities are about humanistic objects of study and of humanistic relevance, rather than being related to information technologies.
This does not constitute a problem by itself, but leaves much room for improvement in two different but connected aspects. First of all, humanities benefits from information technologies (IT) by adopting and using them, but information technologies rarely obtain much in exchange. In other words, the relationship between humanities and IT usually takes the form of a mere service-providing scenario. However, we can envision a situation where IT, while providing support to the humanities, are also inspired and extended by them, thus achieving a symmetrical relationship where the results provided to research questions of one discipline boost the other, and the other way around. This would need that the posed research questions incorporate elements from both the humanities and IT, and that the objects of study are also combined in meaningful manners.
Secondly, the humanities usually adopt existing information technologies as they are, barely scratching the surface and rarely “seeing through” into the underlying theories of information or computation. This barely critical adoption means that information technologies are often embraced or rejected depending on hype, commercial success, availability or perceived ease of use rather than true value for the task at hand. This is not much of a surprise, since digital humanists are not (and are not expected to become) software engineers or computer scientists, and nobody is free from commercial and peer pressures. The impact of the theoretical implications of IT in archaeology, for example, has been studied by a number of authors in relation to areas such as geographical information systems (GIS) [Hacιgüzeller 2012, Huggett 2012] or 3D reconstructions [Dell’Unto 2014]. Despite the adjective “critical” popping up everywhere these days, adoption of technology is rarely critical but driven by fashion and market drive. This has two consequences. First, it is likely that the technologies that we select for a particular project are not the best, and that more valuable ones are left aside unknown to us. Second, we are missing a great opportunity to challenge existing technologies and push for their improvement. Digital humanities projects can be an ideal arena to assess existing approaches, suggest new ones, or develop techniques that go beyond what’s in the market.
There is an additional issue in relation to this. Most specialists in the humanities (very much like specialists in economy, biology or linguistics) usually possess a very superficial awareness of the technologies that they employ. This would not be a problem if technologies were properly packaged into products oriented towards end users; in fact, we can successfully drive a car without needing to know much about its insides. However, and as described in a previous post, most conventional semantic technologies lack good modularity and layering properties, so that a relatively deep knowledge about their internals is needed to successfully operate them. For example, only someone with proper training in conceptual modelling may develop a complete, robust, extensible and overall high-quality ontology of a humanities problem. Of course, knowledge about humanities is needed too, but knowledge about conceptual modelling is essential. Unfortunately, it is not always present. As an analogy, consider the fact that oceanographic ships are not designed and built by biologists or other future users of the ship, but naval engineers. Biologists do provide the functional requirements and needs for the ship, but it is engineers who actually design and construct it. However, large projects are tackled in digital humanities with no or very little concern for the involvement of software engineers or other specialists with the proper training to carry out the necessary IT work, leading to outcomes of inferior quality. A notorious example is the INSPIRE (Infrastructure for Spatial Information in the European Community) directive, an effort by the European Union to homogenise the treatment of geographical information in European countries. INSPIRE has adopted OMG’s UML to develop a complex information model which, unfortunately, contains numerous defects such as stereotype misattribution [Henderson-Sellers & Gonzalez-Perez 2006]. This makes the outcome of a very costly effort paid with public money of a much lower quality than if expert engineers had been involved. Similarly, large multi-million research projects in the digital humanities, usually funded by the H2020 programme or similar sources, also aim to produce highly complex IT outputs but lack the necessary expertise in the associated teams. And the academic literature is no better; many class diagrams published in journals or conference proceedings, such as those in [Sauerbier et al. 2008, Belussi et al. 2014] contain glaring semantic and syntactic defects that seriously hinder their applicability.
- We are missing an opportunity to improve the state of the art in IT as well as humanities in digital humanities projects.
- Specific information technologies in digital humanities projects are usually selected through hype and commercial pressure rather than technical adequacy.
- Many digital humanities projects are carried out without the contribution of IT specialists, producing very poor results.
How can this situation be improved? In my opinion, through trans-disciplinar research organised as co-research programmes. Let me define these two terms.
A trans-disciplinary effort is one that involves multiple disciplines, and where the outcomes obtained through each of them benefit the others. For example, imagine a project involving anthropology and knowledge engineering, where the results in anthropology help to advance the state of the art in knowledge engineering, and the results in knowledge engineering contribute to anthropology. We often see multi-disciplinary projects (involving multiple disciplines) or even inter-disciplinary ones (where research questions and objects of study take place in the crossing between disciplines). Trans-disciplinarity implies all this and goes beyond it, by requiring that the project outcomes benefit every discipline involved.
Co-research, in turn, is a strategy to organise research projects by which work in different disciplines is carried out in parallel and in an interconnected manner. For example, imagine a research institute where a team of archaeologists excavate a mound to reveal how the burial was performed and constructed, while a team of astrophysicists study what the sky looked like back in 3000 BCE. Each team draws on the advances of the other to situate and contrast their work.
Co-research and trans-disciplinarity are closely related. Given the fact that the digital humanities are a young hybrid field, I suggest that work here is organised as co-research in order to obtain trans-disciplinarity and, in turn, advance the state of the art in both IT and the humanities. The goal would be to obtain something between digital humanities and the “human digitalities” that Pierson demanded. This is not easy, because the overall research management and policy-making structures in the world have evolved towards heavily uni-disciplinary work and discipline-oriented organisations. For example, most research organisations are conceived around a specific discipline, and most research funding bodies evaluate proposals through panels specialising in specific disciplines. To be successful in trans-disciplinarity through co-research, we would need research centres that are conceived around specific research problems or objects of study, regardless of what disciplines may study them. A good example is the National Cancer Institute in the US, where researchers in disciplines as diverse as doctors, biologists, geographers, economists or political scientists work together to understand, prevent and treat cancer. The Institute of Heritage Sciences where I work is also of this kind, employing anthropologists, archaeologists, architects, astrophysicists, engineers, geographers, historians, psychologists and sociologists to study the phenomenon of cultural heritage. In addition to problem-centred research organisations, we would also need funding bodies that are capable of assessing proposals that do not clearly fall into a conventional disciplinary area.
Some areas have been successful in establishing strong trans-disciplinary research programmes, such as bioinformatics. In this field, research outcomes often benefit both biology and IT (see for example brane calculus), and research is often organised as co-research. Not only this; some subfield in bioinformatics have evolved into distinct research areas themselves, with objects of study and research questions that are not exclusively linked to either biology or IT. What would it take to move digital humanities to a situation similar to that of bioinformatics? This question was suggested to me by Christof Schöch and, to be honest, I don’t have an answer. I do have an observation to make, though: we should not assume that every combination of two or more disciplines must make sense as a trans-disciplinary hybrid. Some may not. For this reason, we should be critical of the digital humanities as such a hybrid. And, still, keep trying to move it forward in this direction.
- Belussi, A., Migliorini, S. & Grossi, P. 2014. Managing Time Dimension in the Archaeological Urban Information System of the Historical Heritage of Rome and Verona. In 21st Century Archaeology: Concepts, Methods and Tools – Proceedings of the 42nd Annual Conference on Computer Applications and Quantitative Methods in Archaeology. Giligny, F., Djindjian, F., Costa, L., Moscati, P. & Robert, S. (eds.). Archaeopress. [Link]
- Dell’Unto, N. 2014. The Use of 3D Models for Intra-Site Investigation in Archaeology. In 3D Recording and Modeling in Archaeology and Cultural Heritage. Theory and Best Practices. Campana, S. & Remondino, F. (eds.). Archaeopress. [Link]
- Hacιgüzeller, P. 2012. GIS, Critique, Representation and Beyond. Journal of Social Archaeology, 12 (2). [Link]
- Henderson-Sellers, B. & Gonzalez-Perez, C. 2006. Uses and Abuses of the Stereotype Mechanism in UML 1.x and 2.0. In Model Driven Engineering Languages and System: 9th International Conference MoDELS 2006. Nierstrasz, O., Whittle, J., Harel, D. & Reggio, G. (eds.). Springer. [Link]
- Huggett, J. 2012. Thinking through Time and Space: The Implications of GIS in Archaeology. At “Les Dynamiques Spatio-temporelles en Archéologie”, Tours, France. [Link]
- Sauerbier, M., Fux, P., Kersten, T. P. & Lindstaedt, M. 2008. Integration of 3D Data, Texture and Archaeological Information in a Database Management System for Petroglyph Documentation And Interpretation. ISPRS Archives, volume XXXVII, part B2. [Link]