University of Chicago and UC Berkley have initiated a planning process codenamed Bamboo for developing a platform for digital humanities. The Bamboo initiative is quite broad, its manifesto starting with this statement of purpose:
Bamboo is an multi-institutional, interdisciplinary, and inter-organizational effort to bring together researchers in arts and humanities, computer and information scientists, librarians, and campus information technologists to collectively tackle the question: How can I enhance arts and humanities research through the development of shared technology services? Source
The planning process intends to be quite open ended and is structured as a series of workshops. An initial series, currently underway, looks at the user requirements of the humanities scholar. A second series of workshops will look at services and tools. Finally, a third group will collect insights on achieving the most important goal stated in the Bamboo planning document, namely “to create shared technology services that will support arts and humanities research.” Source The final idea, at least so far, is to create a services oriented architecture (SOA 1, 2, 3, 4) of information gathering, organization and dissemination.
I find this initiative fascinating and extremely productive. It converges with efforts invested by several major academic projects, among them our own Visible Past and Thought Ark initiatives at Purdue University. In what follows I would like to draw on our experience to invite the Bamboo community to take a look at what Bamboo wants to be at a more abstract level.
In order to provoke, rather than close the discussion, I would like to propose a specific question and a tentative answer that could help move the discussion along. This is not to preclude future discussions that might take the Bamboo process in a different direction that the one suggested here, but rather to invite our colleagues to challenge, amend or improve on both of them. Without further ado, my question is: should Bamboo be more than a collection of tools? And my answer is that while Bamboo could and by necessity should try to accommodate as many perspectives and needs as feasible, it should at the very least propose a set of standards, interoperability norms, design principles, and communication algorithms that allow present and future initiative funded or supported under the Bamboo umbrella to collaborate and share data and capabilities. By this I mean that the Services Oriented Architecture focus of the project should be more than a theoretical framework, it should have an immediate impact on everyday life scholarly practices. It should be something on a par with the wiki paradigm, which is more than an idea of collaboration, it is a specific collaboration platform.
My suggestion is rooted in an all too human insight and in two more theoretical major premises. The insight is that the web’s main advantage, ability to congregate vast amount of information, is offset by its main vice (at least partially derived from this virtue). It is easy, actually too easy and thus messy, to add new information to the collection of what is increasingly improper, since most web content is dynamically created from various databases, to call “web pages.” Although most content is structured, structure is dictated more by the superficial requirements of format and display, rather than by the strictures of (inexistent) semantic ontologies. Web content, say a text, is usually tagged with markers that indicate where the title, body or images are to be located. Yet, it is rare the case where conceptual content is annotated with subject terms or thematic synonyms. It is even rarer the case where semantic tagging goes all the way down to the level of the idea of a specific paragraph. In brief, web content is organized to facilitate display . Ability to easily sort or combine information with respect to its substantive characteristics and meaning is minimal.
The two theoretical premises that buttress my suggestion that Bamboo should work toward a specific platform are that:
- digital content is vastly enhanced via networking both of users and platforms,
- networking is and will be for some time tied in with the distributed architecture of Internet/like networks.
The corollary is that our thinking process should proceed from the assumption a Bamboo-fostered application platform ought
- at the very least to aim to structure and standardize data in such a manner as to make it accessible to any present or future user of such a platform (think RSS or OMPL feeds)
- create interoperable (pluggable) tools by which various digital objects and artifacts can be embedded into more traditional publishing and research practices (Think WordPress and its profusion of community built plugins)
- all objects, text, artifacts and even user behaviors that do not threaten privacy should be made available for data mining and knowledge discovery by the other users (Think Digg.com and Slashdot.org metafiltering methods)
- a system of universal access should rest on a system of personal authentication and identification that reduces the need to use multiple digital identities, passwords, etc. (Think OpenID for humanities).
Switching to what I believe to be the more specific aims of Bamboo, a future digital environment for humanities should meet a number of requirements. I will enumerate a number of them, mostly for staking out a field for discussion, but I will only focus on the last one, which seems to me to be the most important.
- It should be accessible to all users, including those with limited or low interest in digital technologies. This level of simplicity should at the same time be able to serve the technological demands of users who although endowed with advanced technical skills are primarily employed to deal with content, not with the manner in which this content should be manipulated or disseminated, and who would need power tools delivered in a simple manner.
- It should provide easy and indexable access to the existing store of knowledge and evidence about any specific aspect of human life. Specifically, it should connect information residing in different digital and analog domains (web, specialized databases, mobile applications, microfilms, recordings, films, etc.) through indexing and meta documentation that is common, simple, and open.
- It should allow wide and facile interaction with other scholars and users of humanities information while interacting with information that interests those scholars. This can be realized via social networking affordances embedded in the platform.
- And finally and most importantly, it should mesh up with the most fundamental activities of this scholarly guild: identifying sources, selecting them, learning and producing learning from them via a process of annotating, writing and rewriting text and/or other types of content (images, digital artifacts, numeric representations of social realities). In a word, it should center around an application that facilitates the creation process that is native to the humanities scholar, namely that of writing and composing tangible content objects.
Of these requirements the one I find most important, and the one I would like to spend more time on, is the last one. Humanities scholars are interested in organizing, distributing, connecting and publishing information found on diverse knowledge domains. Thus, their first and most fundamental need is to connect information and to make inference or bring up warrants for such connections. Vannevar Bush’s 1945 Memex, counted as the theoretical precursor of all modern information/content manipulation computers and networking technologies, is in effect not very far from this vision. Bush’s vision of a network of links between topics and ideas, used to support arguments and preserved for eternity in mechanical form, is another way to say that we have been looking for a social and intellectual networking device for the better part of the 20th and 21st centuries. In fact, the quest is even older. As Socrates famously put it in Plato’s dialogue Phaedrus (Phaedrus 265e1-3; cf. Statesman 262a8-b1), the role of the true dialectician (humanist) is to find the natural joints between everything there is, not merely to describe the “bones” (structure) of the world. Thus, a platform that will be connection-centric, is not just a simple hyperlinked or footnoted environment, but one where arguments can be supported in a substantive way. It will be a space where statements such as Weber’s famous “Economic acquisition is no longer […] the means for the satisfaction of [man’s] material needs [but] is closely connected with certain religious ideas” or Marx’s “The hand-mill gives you society with the feudal lord; the steam-mill, society with the industrial capitalist” can be constructed, tested, and validated (or not). These connections can only be warranted when enough evidence can be marshaled in their support. The evidence does not speak for itself, instead it needs to be argued in a narrative format. In other words, a platform for humanities scholarship should support seamless source annotation, writing, editing and reviewing.
Otherwise put, your regular historian would not see the advantage of a digital tool for identifying and utilizing information until this tool will be integrated in her daily work routines of thinking and writing.
For example, a colleague who teaches history classes has dabbled for a while in the newest methods of displaying geographic information. He has invested a considerable amount of time in publishing a special website, containing data about an archaeological survey he has conducted for the past several years. When asked, however, if he were interested in investigating the possibility of adopting a more open, collaborative platform, a wiki, that would allow mapping, publishing and collaborating with his colleagues, he rejected the offer. The new website, although offering a wider array of data publication options and social networking enhancements, did not meet his requirements for what he called “publishability.” Namely, the content could not be easily turned in an “article” or “book”. He believes that such a site could not be counted as a “publication”, thus the ability to add true scholarly content to his geographically annotated website was of minimal importance for him, despite the great potential advances in terms of flexibility and true scholarly interaction the platform presented.
He even confessed he is not at all sure that digitizing and mapping his data on a server was a good idea in the first place since that effort would duplicate his more “analog” effort to keep track of data using more primitive, but more affordable and simpler tools, such as notebooks, maps and word processing files. He finally added that even if the effort was that of reorganizing the information into a different genre, this was too much trouble for him at this point in his career. He is a full professor.
However, if the platform I proposed could’ve turned any and all of his contributions into annotated, vetted and reviewable “articles” or monographic book chapter, I am quite sure he would’ve embraced the medium.
This brief example brings us to our first tentative conclusion, which I want to make into an invitation for further discussion. The fundamental idea of the Bamboo process, namely to identify the basic scholarly practices primitives is absolutely essential and I support it wholly. In trying to identify these practices, however, it is essential to figure out what is truly essential and common across disciplinary boundaries. This ur-primitive, in my view should aim at deciding if it is not the most fundamental process, that of writing and of figuring out through writing and expression the boundaries of a problem and its potential solutions, that is the holy grail of the whole process. A Bamboo fostered platform could thus start from a fundamentally user-centric assumption. Namely, that in the center of the digital humanities universe is a scholar whose life has become overly complicated and difficult by a slew of tools, methods of designing and integrating scholarship in his daily life. He or she would adopt a new platform only if it made what he or she is already doing simpler, not more complicated. If it simplified and reduced the number of tools and objects he or she needs to manipulate. The role of the Bamboo platform would be to simplify this task by making access to tools, by enhancing our ability to connect digital objects and artifacts, our ability to connect with colleagues and students via simple, directly intuitive and universally available interfaces that all converge on the scholars’ desktop, preferably in the format of a word processor. Moreover, the platform should integrate in the most straightforward manner the learning and writing processes with those dedicated to publishing. This should be done in such a manner that dedicated genres and modus operandi (articles, book monographs, peer review, scientific validity checks, etc.) would survive, flourish even, under the new digital regime.
I stop here, rather abruptly, waiting for reactions. I am planning, however, to release a sketch of such a platform, including essential services and affordances. It will also try to leverage the idea of the mashup editor as basic architecture strategy, which could be used to support the infrastructure of the system. To be continued.