I received quite a few tweets after my last post about locating the theoretical base of eResearch. One tweet particularly stated that Library and Information science is a very active research area; which of course it is, but I am not sure if this is what I had in mind by attempting to locate eResearch’s theoretical base (and I perhaps needed to define my terms a little better). Library and Information science rarely use the term ‘eResearch’ and are perhaps a lot closer to what we understand as the Digital Humanities than eResearch (again these terms have meaning to those of us that understand that computing is not one thing and exists in various institutional setting and research concerns and isn’t simply about words with an ‘e’ in front of them).
I found an interesting article from Bill Appelbe and David Bannon of the Victorian Partnership of Advanced computing here in Melbourne that also attempts to define and critique ‘eResearch’. It is science focused (of course) but it is a good attempt at explaining the eResearch agenda. It is a refreshingly critical article, but it does shift the responsibility for research away from ‘eResearch’ to somewhere else (ie. they partly argue that eResearch is a support service and not research thus the research produced by eResearch must be the responsibility of the researchers themselves). There is nothing surprising in this claim, but I still hold that eResearch is too distant from where research happens and especially much of humanities research (and there needs to be greater understanding of how to make this happen otherwise eResearch can’t move from service to research).
WHAT IS eRESEARCH?
eScience, or the more generic eResearch, has come into vogue recently, following on the heels of the more well-established term eCommerce. Like eCommerce, which can include anything from supply-chain integration to CRM (Customer Relationship Management), the definition of eResearch is very much dependent upon an individual or organization’s perspective, and to confuse matters further it is called Cyberinfrastructure (Cyberinfrastructure, 2006) in the USA. So any group of researchers will have differing, and often vocal, opinions on what eResearch is. For example, for users of large data sets, such as climate modeling, it is all about having large data sets readily accessible, without them having to worry or waste time about sharing, data formats, backup, or security. To a big compute user, such as modeling cell membranes, its having massive compute capacity available on demand, without having to know anything about underlying details of the computers, operating systems, or file systems. Yet another group, such as the International Virtual Observatory in Astronomy, will tell you eResearch is a matter of breaking down the barriers between researchers, be they geographical, cultural or technical. So there is no quantitative definition possible, or even desirable of what is, and what is not eResearch. Instead, there are useful characteristics of eResearch projects that can distinguish the degree to which a particular project might be promoted as eResearch, or “traditional” research as shown in Table 1 (see attached article).
A further point of confusion in the use of the term eResearch relates to whether eResearch is actually the “research” conducted this way, or the infrastructure that enables the research conducted this way. We adopt the view that eResearch strictly means research conducted relying on supporting infrastructure that should properly be called either eResearch Infrastructure or Cyberinfrastructure. From the table above, it is clear that the supporting infrastructure can includes hardware, software, networking, and human resources. But eResearch is not just about using new IT tools, such as teleconferencing or web publications, to support research projects. Use of such tools is a common mischaracterization of eResearch. eResearch projects do not just use IT technology, rather they are reliant on IT technology and organizational changes such as online collaboration to achieve the research outcomes. It is also important to note that eResearch adoption is highly discipline dependent. Scientific disciplines such as observational Astronomy or High-Energy Physics have arguably been using eScience for close on decades. Such disciplines intrinsically have some of the characteristics of eResearch above; large, expensive shared instruments and the need to share data internationally using agreed standards (e.g., astronomical coordinates and reference frames). By contrast, some disciplines such as Pure Mathematics or Linguistics intrinsically have few of the characteristics of eResearch. Yet even here, the trend is towards eResearch. For example the case of Mathematics it is the growing use of computers for proofs and proof checking, and a repository of known theorems
So eResearch support is not a “one size fits all” – it is discipline and project dependent. There is no such thing as “eResearch Support In a Box”. The very nature of research, which is constantly testing and pushing the boundaries of knowledge, means that eResearch support itself must be constantly pushing the boundaries of networking, data, computational, and collaboration support. eResearch infrastructure is a large system that is made up of a number of organic components software and hardware and organizational components, where each researcher should be able to readily find the components they want and need and not worry about the remainder. But they will have to be able to reach out and grab additional components when the need arises as it invariably will. For example, a scientist may find that they need access to a statistical analysis or visualization tool to interpret their data, or import data from a new instrument source that has just become available. Clearly, the individual components must work seamlessly together and the researcher who is widening his use must find the additional components working exactly as he expects, no surprises! (link to article).