Daniel’s response to question 6: ML Research and Process

Does “collaboration,” the word most commonly used to describe working with others, reveal anxieties inherent to any division of labor? “Collaboration” opens up problems from the start, given that it means both “united labor” and “traitorous cooperation with the enemy.” To what extent is one sharing or abandoning one’s own principles through collaboration? Yet to even raise the question of collaboration’s connotations requires one to engage in a humanist practice of concentrating on ambivalences, rather than on productive meanings. To ask about what “collaboration” means may simply be a way to hold up the process of collaboration. Is this query merely a form of epistemological quibbling or does it acknowledge institutional disparities in joint research? Anyone with a dictionary can recognize the ambivalence in “collaboration,” but just how serious is this concern?

The abundant commentary on the collaboration concentrates on both the ethical and methodological decisions required in setting up digital projects, as well as in interpreting the data they generate. The moral implications of collaboration usually involve making sure that all the participants are acknowledged appropriately. Humanities scholars like to presume that the natural and social sciences have worked out the conventions of crediting authorship for jointly written articles, however the dynamics and specializations of digital projects often do not align with what are imagined to be clear-cut protocols. The “Collaborator’s Bill of Rights” may suggest an orderly assignment of roles, but its reliance on Enlightenment universalist categories would lead anyone to doubt their effectiveness: http://mcpress.media-commons.org/offthetracks/part-one-models-for-collaboration-career-paths-acquiring-institutional-support-and-transformation-in-the-field/a-collaboration/collaborators’-bill-of-rights/

As for methodology, the first point introduced is that although humanities scholars tend to work alone, they now need to learn collaborative methods of research and institutional evaluation if they are going to participate in digital research. Asymmetrical relationships are almost a requirement in the most innovative digital projects. Collaboration ideally includes very different kinds of partners. So that it does not look like a bunch of professors all discussing the same problem, nor should it follow a corporate IT model whereby a team is assembled of different experts to produce one result. And if the collaboration is meant to be even-handed, then it cannot be structured along a simple service arrangement, whereby data is brought forward from an archive for the humanist to interpret.

The most exciting digital scholarship argues, of course, that we are not faced with a choice between either humanities questions or scientific methods. Ben Schmidt’s work on this point is exemplary: http://sappingattention.blogspot.com/2012/11/reading-digital-sources-case-study-in.html Likewise, there a plenty of people arguing against the imperative to work in teams. Presumably, individual scholarship will revive as digital scholarship refines its tools: http://blogs.lse.ac.uk/impactofsocialsciences/2014/09/10/joint-authorship-digital-humanities-collaboration/ Before we formulate a utopian model of even-handedness in research, we should note that interactions between disciplines rarely operate at the same speed or with a calm sense of parity. Interdisciplinary work inevitably entails appropriating methods and information from other fields so that they can redeployed in unfamiliar contexts. As Ted Underwood states, humanists who require convincing of the virtues of distant reading should not hesitate for too long, lest researchers from outside their discipline start explaining their own field to them. http://tedunderwood.com/2015/10/03/can-we-date-revolutions-in-the-history-of-literature-and-music/ Collaboration thus can be motivated by a certain urgency to join a discussion that is already well under way.



One thought on “Daniel’s response to question 6: ML Research and Process

  1. Daniel Purdy’s response to the question on Modern Language Research and Process opens up some interesting questions about the meaning of ‘collaboration’ in the humanities. For those of us carrying out research in the digital humanities, collaboration is usually not even an option – the vast range of skillsets required on a typical digital project is beyond the scope of a single person, and in any case such a wide focus would not lead to good research. My experience on digital Modern Languages research projects is that there is a sense of inevitability that roles will (and need to) change, but also concern (much of it justified) about what that will mean to core humanities values and how researchers can learn new skills without diluting existing expertise.

    We certainly should not introduce false symmetries, but equally we do need to move beyond over-simplistic dichotomies represented by scholar/technician and research/service role divisions when talking about digital research and recognise new actors and roles in an increasingly dynamic and connected research ecosystem, a point analogous to arguments made for the role of translators in the recent ‘Translation as research’ manifesto, published by MLO Open http://www.modernlanguagesopen.org/index.php/mlo/article/view/80.

    Humanists often treat claims about the value of opening up research tools and methods using digital infrastructure with some suspicion, although the picture is far less clear than sometimes thought, as demonstrated by recent research by Smiljana Antonijević in Amongst Digital Humanists: An Ethnographic Study of Digital Knowledge Production (forthcoming) which examines “shifts in research practice, knowledge and legitimacy claims” in some detail and finds a surprising degree of divergence across subfields within the humanities, and across different aspects of the whole research cycle from finding and collecting, through analysing, visualising and interpreting, to publishing and archiving.

    The key question, then perhaps, is ‘why should we reveal or share our underlying research tools and methods?’ One the one hand, inhibitions are imposed by current humanities research culture (which rarely provides academic incentives for collaboration or sharing) whereas on the other, calls for open scholarship often ignore the risks or the costs – the extra resources required to make a research object interpretable or reusable by others. Some research domains have developed models to explore these issues: federated research environments like NINES (Networked Infrastructure for Nineteenth-Century Electronic Scholarship) http://www.nines.org/about/ or integrated research frameworks like Papyri.info (http://papyri.info/), which enables collaborative online authorship, is open to anyone but subject to a peer-reviewed framework, and enables open examination of decision trails for approved/published papyrological editions. How might this look in the Modern Languages, and which subfields within ML would most benefit? And to repeat a question I asked earlier in the year, ‘How useful is the ‘Commons’ model in thinking about the future of Modern Languages research?’ (http://www.paulspence.org/mlr-and-dh/).

    In part, the answer to these questions can be found in thinking about the new classes of research object which are emerging from digitally mediated Modern Languages research, in considering what specific skills we need to develop (which may involve interpreting deep or human/machine translated data as much as big data) and in providing a vision for a new scholarly reputation economy which concords more closely with transformations elsewhere in human culture and society, without losing the critical focus which make the humanities a crucial “part of any vision of a future society” (http://4humanities.org/).

    And finally, since we are contributing to this writing sprint during Academic Book Week (http://acbookweek.com/), I’d like to pose some questions about the implications for the relationship between research and publishing. Are the old wrappers for content such as ‘book’, ‘journal article’ still valid? Are they enough? If we step back from print-era assumptions and think of publication more abstractly as “a range of modelling activities that aim to develop and communicate interpretation” (Blanke et al., 2014: 17), how should we define what we wish to model within the Modern Languages?

    Blanke, T. et al. (2014) Digital Publishing Seen from the Digital Humanities. Logos. [Online] 25 (2), 16–27.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s