Iza Romanowska ^[1]

Barcelona Supercomputing Center

Abstract

Archaeology has been one of the disciplines in humanities most pressed for multidisciplinary approaches. Partially due to the nature of the data and partially because of the general paucity of it, archaeologists have always worked very closely with researchers from outside of their profession to squeeze the most of the little information there is about ancient societies. We get fellow geomorphologists to look at the soil profiles, osteoarchaeologists to analyse the bones and physicists to date the finds, etc. However, this kind of collaboration often does not extend beyond a “customer – service provider” type of relationship, in which members of one discipline use their skills and knowledge to provide a service to another, for example, to establish the most probable age of a sample using a particular dating technique. Although useful and necessary this is hardly a model for ‘transdisciplinarity’.

Throughout this piece the word “multidisciplinarity” is understood as this kind of collaboration – using the expertise of other disciplines to perform a specific service such as an analysis of the particular type of data. In contrast, the term “inter-” and “transdisciplinary” is used here to mean a much closer and mutually dependent type of working together towards a common goal. An interdisciplinary team aims to work out a problem together by applying, developing and adapting tools of different disciplines in a synthetic manner. This type of collaboration formed the spine of a recently concluded ERC-funded EPNet project: “Production and Distribution of Food during the Roman Empire: Economic and Political Dynamics” (ERC-2013-ADG 340828). Led by prof. J. Remesal Rodríguez – the head of CEIPAC group (CEIPAC 2019) at the Department of History and Archaeology at the University of Barcelona, Spain, it united historians, archaeologists, network scientists and specialist in computer simulation to try to understand how commerce of foodstuff shaped the economy of the Roman Empire. The ambitious goal was to use the existing archaeological datasets, and in particular, the epigraphic dataset collected over the last two decades by CEIPAC (romanopendata, 2019), to validate or reject existing hypotheses regarding the functioning and organisation of Roman trade. The project ran for five years between 2014-2019 and employed over 15 researchers at different stages of their academic career based at three research institutions (University of Barcelona, Barcelona Supercomputing Center and Siris Academics) (Remesal et al., 2015). Here, we present a few ‘lessons learnt’ in the five years of this collaborative interdisciplinary project.

1. The team needs to share a baseline epistemological framework

No collaboration can happen if there is no common framework between researchers specifying how data, models and tools are collected, manipulated and used. Although differences in the language used across different disciplines are commonly raised as a major challenge for interdisciplinary research, diverging views of what the scientific method entails and what are the basic requirements of data collection, analysis and modelling are much more fundamental obstacles for a fruitful collaboration between disciplines. There are three major axes of scientific research where definitions, methods and interpretation need to be synchronised between partners prior to the beginning of work: theory, data and models. Progress is unlikely to be achieved if partners disagree about such fundamental questions as to what constitutes evidence, what is the function of data and theory or whether the research questions (hypotheses) at the center of the project need to be testable. This common challenge can be overcome with a solid amount of goodwill and openness between researchers involved in an interdisciplinary project, but it requires time, effort and appreciation that different disciplines may have different takes on the fundamentals of how we do science.

2. An interdisciplinary researcher to ‘bridge’ between the disciplines is a worthwhile investment

One may hope that putting together top experts in their disciplines in one room is enough to spur research excellence. In practice, this has proven numerous times to be an almost infallible recipe for a disaster. Different disciplines operate in different semantic realms meaning that their representatives speak almost foreign languages but also in different epistemological worlds where things get done differently. Similarly, it is easy to under -or over- estimate the level of knowledge and understanding of almost any topic by a partner from a different discipline since their educational trajectories differ significantly from ours. A common result of putting two experts together without someone acting as a ‘bridge’ or a ‘glue’ between them are high levels of frustration, researchers accusing each other of incompetence and eventually, having everyone retreat into the safety of their disciplinary boundaries. A team member who is neither the top specialist in discipline A nor in discipline B but has experience in working in both can mean that the full potential of all team members is achieved. Where two people may be working at 20% of their capacity because they struggle to communicate, adding an interdisciplinary researcher is likely to decrease the risk of catastrophic failure due to miscommunication but also to ensure that the expertise of all team members is used to the full.

3. One of the legacies of an interdisciplinary project needs to be cross-disciplinary training

When developing a new method, paradigm or research area within one discipline is often coupled with strong dissemination and training of other members of the community. The same should apply to any interdisciplinary work. If method “a” from discipline A can be successfully applied to problem “b” in discipline B then part of the work should focus on laying out the groundwork for a wider recognition of the new method among practitioners. This serves two purposes – first of all, it allows other researchers to critically engage with the study and its results; second, it enables them to pick up the knowledge and skills and ensures that what was originally ‘interdisciplinary’ work becomes part of the standard toolkit of the discipline.

4. The systemic barriers are standing strong despite the common calls from higher education management for more interdisciplinarity

High level of competitiveness and strict ‘selection’ criteria are the reality of academia across most of the world. Yet despite the common calls for ‘interdisciplinarity’ coming from all ranks of the Higher Education establishment the footsoldiers of science – PhD students, postdocs, early-career faculty need to carefully navigate a system where transgression across disciplines’ boundaries are commonly punished. What is the value for a physics PhD student of a paper published in an archaeological journal? Will a postdoctoral researcher be awarded a fellowship in computer science if their first degree was in history? Which department should an interdisciplinary researcher teach at if their expertise stretches over two distinct disciplines? Despite the seemingly unending ‘hype’ of interdisciplinarity career paths and progression for interdisciplinary researchers are poorly defined, leaving them in a difficult situation in which only part of their academic output ‘counts’. Although having well defined disciplinary boundaries is in many cases useful and necessary it is also important to counteract the disadvantage that they place on certain type of research. Without establishing a clear professional development path for interdisciplinary researchers this kind of work will be considered ‘risky’ in the long perspective even if short term incentives (such as project funding) are presented. The work necessary to make systemic changes, such as the way in which tenure is granted, cannot be left entirely for early career researchers to shoulder.

5. Summary

The EPNet project proposed a research agenda focused on using formal modelling techniques to gain insight into the inner dynamics of trade in the Roman Empire. At the time, this was a revolutionary proposition and almost entirely terra incognita on the map of archaeological and historical research. Throughout the lifetime of the project physicists, computer scientists, historians and archaeologists had to come together, iron out their differences and find a common way forward. It would be naive to suggest that this is an easy process but some of the challenges described above have also opened up new ways of looking at old problems. For example, one of the most common ‘sins’ of humanities research is the non-formal formulation of hypotheses (models) meaning that any attempts at operationalizing them in order to test them against existing data must necessarily include a solid amount of ‘interpretation’ of what the original author had in mind. This turned out to be an interesting exercise for all researchers involved and generated some great insights. Some of the issues (e.g., the systemic disadvantage of interdisciplinary research in academia) could not be solved within one specific project but it did help to identify where the disciplinary boundaries are most strictly enforced. However, the major impact of the EPNet project lies in demarcating the basic framework of this kind of interdisciplinary research and adopting tools and methods from other disciplines to the humanities context. This would not be possible without engaging with scholars across disciplinary boundaries.

References

CEIPAC. (2019). “Centro para el estudio de la interdependencia provincial en la antigüedad clásica.” «http://ceipac.ub.edu/ Accessed 20/05/2019.

Remesal, J., Díaz-Guilera, A., Rondelli, B., Rubio, X., Aguilera, A., Martín-Arroyo, D., Mosca, A., Rull, G. (2015). “The EPNet Project. Production and Distribution of Food during the Roman Empire: Economics and Political Dynamics.” In Information Technologies for Epigraphy and Cultural Heritage. Proceedings of the First EAGLE International Conference, pp. 455-464. Rome: Sapienza Universita Editrice. http://ceipac.ub.edu/biblio/Data/A/0771.pdf.

romanopendata. (2019). “Roman Open Data. A Data Visualization & Explratory interface.” www.romanopendata.eu Accessed 20/05/2019.