SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Background to e-Science
  5. The National and Global Shaping of Research Programs
  6. Approaches to e-Science
  7. Discussion
  8. Conclusions
  9. Acknowledgments
  10. References

The use of advanced high performance computing tools across the sciences, social sciences, and humanities has rapidly expanded in recent years. Several programs have been initiated under the umbrella of e-science with large-scale funding and a wide range of projects. The vision underlying these programs is that new tools will greatly enhance research and enable new forms of global collaboration. This article maps out the different social science approaches to e-science and provides illustrations of how they have been deployed. The aim is to highlight the diversity of these approaches, show complementarities among them, and point to how they may shape the e-science enterprise in years to come.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Background to e-Science
  5. The National and Global Shaping of Research Programs
  6. Approaches to e-Science
  7. Discussion
  8. Conclusions
  9. Acknowledgments
  10. References

The use of advanced computing tools in the sciences, social sciences, and humanities has accelerated in recent years. Although the use of computing in various research disciplines is not new, the Internet and high-performance computing have given a new impetus to these efforts. Several programs have been initiated under the umbrella of e-science, with large-scale funding and a wide range of projects. The vision is that these new tools will greatly enhance research and enable new forms of global collaboration.

To a large extent, these new large-scale and complex research efforts have been driven by technological developments, but they also raise many non-technological issues, including legal, ethical, institutional, and disciplinary ones. Indeed, it is increasingly argued that the obstacles to the effectiveness of e-science are not so much technical, as social. In the social sciences, however, there is a variety of approaches to tackling the shaping of new forms of research, governance, communication, and collaboration—such as the sociology of science, the economics of innovation, and research policy.

In this essay, we categorize different approaches to e-science and provide illustrations of how they have been deployed. The aim is to highlight the diversity of these approaches, show complementarities among them, and point to how they might shape the e-science enterprise. Before we go into different social science approaches, it will be useful to provide some background to how e-science came about and how it has been organized. We then categorize different social science approaches to e-science and give examples of each. The conclusion weighs the different approaches and assesses their potential influence on the development of e-science.

Background to e-Science

  1. Top of page
  2. Abstract
  3. Introduction
  4. Background to e-Science
  5. The National and Global Shaping of Research Programs
  6. Approaches to e-Science
  7. Discussion
  8. Conclusions
  9. Acknowledgments
  10. References

E-science can be defined in different ways. We use e-science in the singular to indicate both natural and social sciences (the humanities will only be dealt with in passing), and we adopt a narrow definition that includes only those efforts that enhance scientific research by means of high-performance computing coupled with high bandwidth networks. This is obviously a somewhat arbitrary definition, since what is “high-performance” and which networks are “powerful” is bound to change over time. Nevertheless, in the technical community, this use of the term is widely shared and taken to indicate tools that are being extended beyond their existing capabilities.1 On the social side, this coincides with the programs of national research funding bodies to create a “cyberinfrastructure” (in the United States) and an “e-infrastructure” (in the European Union) or to promote e-science (in the United Kingdom). All of these are aimed at supporting efforts in e-science.2

The definition thus contains both a scientific and a social element. Sharing access or services over a network is central to both because it enables the use of computational resources (GRID, Semantic Web) or research resources (instruments, data) that would not otherwise be available. It is important to note immediately that e-science tools are thus by nature not a stand-alone technology; the tools that are developed are such that they are used in networked mode and can be used or accessed in different locations. The distributed nature of e-science is a point that we return to in the conclusion.

The development of e-science has been driven by a number of factors, both scientific and social. On the scientific side, researchers have pointed to the possibilities of using new tools to foster global scientific collaboration (Atkins, et al., 2003; Taylor, 2001). Hey and Trefethen (2003) talk of the need to cope with “the data deluge,” the increase in the sheer volume of data that needs to be stored and made accessible in different ways, and how this serves both as a driver and a challenge for e-science. These and other calls for funding e-science programs (see the U.K. National e-Science Centre website http://www.nesc.ac.uk/ and the National Centre for e-Social Science website http://www.ncess.org/ as well as the U.S. National Science Foundation’s Office of the Cyberinfrastructure funding opportunities website http://www.nsf.gov/funding/pgm_list.jsp?org=OCI) have identified the need for more effective communication, putting materials online, sharing instruments, and other needs of the sciences.

This call for research has gone hand in hand with several funding initiatives by national research councils. The main programs have so far been in the U.S. and the U.K. In the U.S., the so-called Atkins report (2003) is often referred to as a seminal document. The report presented a “vision” and talked in terms of a “revolution” in science and engineering. It called for large-scale funding (one billion U.S. dollars per year) for a wide-ranging program consisting of a number of linked large-scale facilities and some national repositories and libraries that would fall under the umbrella of providing an “infrastructure.” A number of reports (such as those on the publications website of the National Science Foundation’s Office of the Cyberinfrastructure http://www.nsf.gov/publications/index.jsp?org=OCI) have since been issued for different areas of science and engineering and several rafts of funding instigated.

The call for cyberinfrastructure was subsequently extended to the social sciences in the report by Berman and Brady (2005). The report makes the case for a “cyberinfrastructure” for the social sciences on pragmatic grounds: “It goes to the heart of helping America make a successful transition into the 21st century” (2005, p. 11), the authors say, citing issues such as attacks on the security of electronic computer networks and online identity theft as urgent issues that e-social science can address. They also stress new tools for analyzing social behavior as well as the study of the implications of using the new tools. The U.S. initiatives are now housed in the Office of Cyberinfrastructure within the National Science Foundation, which continues to develop programs. In 2005, a report on “Cyberinfrastructure for Humanities and Social Sciences” was added by the American Council of Learned Societies (http://www.acls.org/cyberinfrastructure/cyber_report.htm).

In the U.K., the main impetus came in 2000 with the establishment of a program of funding and of the National e-Science centre by the Engineering and Physical Science Research Council (EPSRC) and the government Office of Science and Technology. Since then, the program has come to comprise many projects in a variety of science and engineering areas. In the U.K. context there has similarly been a subsequent initiative in the social sciences by the Economic and Social Research Council (ESRC) that also now has a national center that coordinates research projects. The ESRC program initially called for a two-sided approach: Both the development of new tools and their “social shaping” were to be promoted. In practice, however, the former have been supported much more than the latter.3

The U.K. initiative in science and engineering is already winding down as a distinct initiative that is separate from mainstream programs of research funding. This will not mean the end of e-science, but rather the integration of e-science within other initiatives. It also means, however, that unless there are new national programs, the development of e-science as a separate research initiative will take on a different guise. In the U.S., there is likewise increasing concern about how to extend and ensure the future uptake of the cyberinfrastructure initiatives. In both the U.S. and U.K., much of the emphasis is thus turning to how the various tools can be put to more widespread use and how the e-sciences can be sustained in the future. To give just one example, a recent project of the European Commission to enhance uptake of e-science, drawing particularly on the U.S. and U.K. experiences, is entitled “A study on requirements and options for accelerating the transition from traditional research to virtual research organisations through e-infrastructures” (Barjak, 2006).

In addition to the U.S. and the U.K., there has been a successive internationalization whereby the efforts that began in the U.K. and the U.S. have subsequently led to initiatives in Europe (at the EU level via the e-Infrastructures Reflection Group, http://www.e-irg.org/), in Asia and Australia, and elsewhere. China, for example, was a major presence at the third “All Hands Meeting” in Nottingham in September 2005, an annual conference that brings together e-science projects in the U.K. It remains to be seen whether e-science, having been developed mainly in the form of national research initiatives, will continue as nationally coordinated efforts or as individual research programs targeted at particular areas and with limited timeframes.

In the same vein, it also remains to be seen if the label e-science continues to be used for research in this area or if e-science becomes subsumed under new and different forms of research. As mentioned previously, there are many earlier and parallel efforts to integrate computing tools in research, and it is an open question whether those that can be labelled e-science and have to do with the uses of high-end computers via networks retain their distinctiveness. Another question is whether there will be a set of tools that can be identified as an e-science “infrastructure,” that is, greater than the sum of individual tools and the outcomes of particular research efforts. In other words, the question is whether, as envisioned, there will be an “infrastructure” that is similar to other infrastructures (electricity, transportation) that provide a common system that supports many different users.

The National and Global Shaping of Research Programs

  1. Top of page
  2. Abstract
  3. Introduction
  4. Background to e-Science
  5. The National and Global Shaping of Research Programs
  6. Approaches to e-Science
  7. Discussion
  8. Conclusions
  9. Acknowledgments
  10. References

The vision of e-science is a global one, but the different national and supranational (EU) e-science initiatives reflect past legacies of their respective systems of innovation. In the U.S. “cyberinfrastructure” initiative, the direction has been to enhance the national research base in the tradition of “big science” (Galison & Hevly, 1992), “national labs” (Westwick, 2003), and so-called grand challenge pushes on particular large-scale problems (spaceflight, the war on cancer, the human genome, etc). The U.K. is partly in this mold but has also focused on initiating a series of research programs targeted at topical areas of interest, such as generic GRID middleware and interdisciplinary research collaborations. At the same time, the EU’s effort is following a pattern of successive framework programs under different headings, where the major aim is to create an infrastructure that integrates the research institutions from among the member countries. One question that these various initiatives raise is whether e-science will continue in these molds rather than transcending them.

Despite the national differences, there are also similarities. One example is setting up central offices as part of the national (or, in the EU case, supranational) research council structure to foster and coordinate e-science efforts. This is part of the global character of science, whereby all nation-states have adopted similar institutions in order to promote science and research (Drori, Meyer, Ramirez, & Schofer, 2003). Also common across the various efforts is the technology push: that creating the tools will be a stimulus to research, as well as the visions guiding this push, whereby these tools will accelerate and bring major benefits to researchers and, via innovation, to society at large.

Disciplines and Boundaries

Apart from describing e-science in terms of national efforts, it will be useful to examine the disciplines involved and the extent to which they are being transformed. The first feature to notice is that, as with many other technology-led initiatives, there is an imbalance that has meant technologies have been developed without taking social aspects into account. Overlooking the social side is typical of the early phase of large-scale and complex science and engineering projects (Hughes, 1998). Thus, David (2004) has pointed out that “engineering breakthroughs alone will not be enough to achieve the outcomes envisaged for these undertakings. Success in realizing the potential of e-science and ’cyberinfrastructure’ will more likely be the resultant of a nexus of interrelated social, legal and technical transformations” (p. 3). These social issues in e-science are now moving into the foreground, but they also relate to wider changes such as those in electronic publishing, distributed collaboration, and large-scale networked computing. Still, on the whole, the discipline of computer science and its development of new technologies have dominated the early phase of e-science.

Another point to note is that it is not necessarily the case that natural sciences are “earlier adopters” of Internet-related tools more than the social sciences and humanities. In fact, Nentwich argues (2003) that there are some disciplines within these broad disciplinary categories that are “ahead” of others. Nevertheless, as we saw earlier, the national programs have broadly followed the pattern of initiating programs in natural science first and then following up with social science and humanities initiatives. Thus, in the U.K., the e-science program was followed by e-social science and now also an Arts and Humanities Research Board program.4 The same applies to the U.S., where the Berman and Brady report (2004) was published two years after the Atkins report. In any case, the initiatives have rippled outward across more and more fields and disciplines, although perhaps the argument that Kling and McKim (2000, p. 1306) have made for “electronic media supporting scientific communication,” that field differences are likely to persist, also applies to e-science.

If we consider the scale of funding for the different projects and disciplines involved, clearly funding in the natural sciences outweighs funding in the social sciences. For example, in the U.K. it is estimated that £250 million has been spent on natural science (Jeffreys, 2005), whereas the social sciences have committed far less. This must be put in the context of the historical competition between social and natural sciences for resources, whereby the social sciences have traditionally been funded on a much smaller scale. This current drive towards more technology-based research with its funding for tools could, on the other hand, equally be seen as a boost for social science if the emphasis is placed on the novel competition for funding of these tools with the natural sciences.

Approaches to e-Science

  1. Top of page
  2. Abstract
  3. Introduction
  4. Background to e-Science
  5. The National and Global Shaping of Research Programs
  6. Approaches to e-Science
  7. Discussion
  8. Conclusions
  9. Acknowledgments
  10. References

Against this background, we can now look at how social science is involved in shaping e-science efforts. To do this, we can separate four social science approaches arranged along two continua, depending on whether they are aimed more at research or at shaping e-science in a practical way, and whether they adopt a stance of detachment from, or of normative engagement with, e-science (see Table 1 below):

Table 1.  Four main social science approaches to e-science
 PracticeResearch
DetachmentUsability/practical (How appropriation can be enhanced through refining understanding of practice, user representations, and human computer interaction)Value free/attempted neutrality (Measuring dimensions of distributed communication and collaboration)
Normative EngagementAdvocacy/steering and aligning structures (Fostering institutional, economic, and legal structures that enable distributed communication and collaboration. Promoting a particular type of open and accessible e-science)Critique/reflexive or prospective (Social implications of e-science; ability to deliver on claims; policy)

As we will see, despite this analytical separation into four main types of approach, there are synergies at the intersection of the practice/research continuum and the detachment/normative engagement continuum, which we will discuss in a later section. We shall return to this typology after presenting a brief illustration of each approach.

Usability/Practical

This perspective is concerned with how the effectiveness and uptake of e-science can be enhanced through, for example, refining understanding of practice, user representations, and human computer interfaces. Research councils developing e-science programs are increasingly acknowledging that usability issues are a key barrier to the uptake and use of advanced digital technologies. Funding mechanisms to support usability research have been implemented in programs such as the EPSRC call for “Research in the Usability Challenges to Emerge From e-Science.” This focus is also reflected in the emergence of communication fora such as the “Designing for Usability in e-Science” Workshop sponsored by EPSRC and NeSC (see http://www.nesc.ac.uk/action/esi/contribution.cfm?Title=613,).

An illustration of this approach is the eDiaMoND project (Jirotka, et al., 2005), which was a flagship U.K. e-science project. The aim of this project was to create a federated database of digitized mammograms. This would enable x-rays for breast cancer screening (mammograms), which are currently done on film, to be stored and annotated digitally. This digitization would also allow the images, which are now kept at individual breast cancer screening units, to be shared across a network. Finally, such a system could also allow remote experts to be accessed to handle difficult cases, and enable the creation of a repository which could be used to analyze the images with the help of computer-assisted techniques. The readings of these images could then also be done in distributed mode, which would reduce the need for expert readers at each site (each image is currently “double-read” to ensure accuracy).

In this case, many usability issues were found by means of ethnographic and quasi-naturalistic evaluations of work practices in the creation of prototype systems. One of the most important was “trust,” both in the new technology, and between radiologists. Interpersonal trust is developed in particular ways in the 100 or so Breast Care Units across the U.K. The co-location of technicians, radiologists, and other experts in traditional Breast Care Units allows personal histories to be developed, such as where a person trained, who they trained with, how they calibrate the machinery, their particular competencies, and knowledge gaps. The introduction of a distributed system for reading mammograms reconfigured the ways in which trust could be built, and new practices were necessary in order to know the other’s capabilities and to come to rely on the other in new ways. Another set of issues concerned the sharing of data, whereby it was deemed generally unethical to share mammograms between clinics and outside of the clinicians and readers who had had direct involvement with the patient and with each other.

The usability findings in this case could be used to try to address these issues by technical means, such as developing interfaces that afford the communication of contextual information, or by social means, such as reconfiguring practice. In fact, the two were difficult to separate into distinct issues. An important outcome of the usability research on the eDiaMoND project has been to highlight the immense obstacles that would be faced in translating the “proof of concept” into a workable system that could be adopted on a large scale across the U.K. National Health Service, as originally envisioned (http://www.ediamond.ox.ac.uk/whatis.html). In this respect, the usability aspect of this project was typical of usability research: identifying practical obstacles on the ground that they can potentially be overcome by taking into account the needs of diverse sets of users and social settings.

Seen from a greater distance, such case-bound usability research is typically and increasingly a tool for making laboratory or prototype technologies robust enough to withstand the vagaries of real world settings. Note also that from a disciplinary viewpoint, this social science approach is closest, as we shall see, to the computer science end of e-science, even though, in the case of this illustration, the researchers responsible for the usability aspect of e-DiaMoND employed naturalistic and “interpretive” techniques, such as ethnography and participant observation.

Value Free/Attempted Neutrality

There are of course debates in the social sciences about whether value-free research and neutrality are possible. At the same time, many researchers aspire to such a goal, even when they are aware of its potential limits. One area of the study of e-science where this approach is often in evidence is in the study of communication and collaboration. Research that falls into this category includes, for example, the analysis and evaluation of collaboratories (Finholt, 2002; http://www.scienceofcollaboratories.org/). Further examples of work in this area include identifying patterns of new modes of scientific communication and collaboration using methods and measures such as social network analysis, co-citation practices, co-authorship, and hyperlink networks. This work is sometimes policy related and often involves large-scale analysis or comparison of many cases, for example, when measuring scientific productivity or incentives among scientists for collaboration. However, one of the aims is typically an objective assessment of the inter-organizational relations in science.

One illustration of this approach is a study by Cummings and Kiesler (2005) of collaborative scientific projects in the U.S. They examined 62 collaborative projects that were funded by the National Science Foundation, most with partners distributed across a number of locations, and most of which were multi-disciplinary. One of their findings was that collaborative research across institutions is more difficult than collaborative research across disciplines in terms of project coordination and project outcomes. They also noticed that communication in distributed projects tended to drop off over time. It seems, then, that there may be a “trade-off” in the advantages that these projects should bring between “innovation opportunities versus coordination costs” (p. 720). It also seems that technology does not overcome distance (p. 718).

These findings, even if they did not arise directly from e-science projects, are clearly relevant to e-science. Many e-science projects are still at too early a stage of developing the capabilities for collaboration-at-a-distance for one to be able to tell if they are effective and able to bridge distances in collaboration. It is clear, however, that coordination problems of the type identified by Cummings and Kiesler (2005)—which are almost always multi-disciplinary and distributed—will also affect e-science. Sonnenwald (2006), for example, who studied one e-science multi-institutional collaboration (although with limited multi-disciplinarity) over the course of several months, noticed a host of problems of coordination across distributed institutions. Since e-science typically involves both inter-institutional collaboration and communication, these issues are bound to play a major part in the success or failure of e-science. What we see in Cummings and Kiesler’s work (2005) and Sonnenwald’s study (2006) is an attempt, in the first instance, to measure and gauge the extent of collaboration and communication across the projects, even if, in the second instance, they also identify problems that have implications for research policy.

Critical/Reflexive

The critical/reflexive approach is concerned with the social implications of e-science and tends to highlight the discrepancy between visions and practice. The focus is often on the analysis of high-level discourses around e-science, such as policy documents or future visions. The discussion typically revolves around the “values” and “expectations” embedded in technologies as well the evolving meaning of terminology, definitions, and boundaries. This perspective tends to couple its argument with the need for researcher immersion in ethnographic case studies (Vann & Bowker, 2006; Woolgar & Coopmans, 2006).5

Vann and Bowker (2006), for example, explore “prospective texts” around e-science and argue that in order to understand the impact of e-science on knowledge-producing practices, social scientists need to “consider decisions that get made about how the skill, commitment, performance, and product demand of scientists can be coordinated and stabilized” (p. 71). In other words, visions lock in particular trajectories of the way in which the work of the research is done. This approach thus looks to concrete instantiations of local scientific practices—such as how the “labor” of research of different forms of research is talked about (Hine, 2005; Vann & Bowker, 2006)—or at new disciplinary identities (Hine, 2006) and forms of accountability in scientific practice.

Hine’s work (2005) is a good illustration of this approach. She explores the relationship between the visions for global online cataloguing and digitization of species and the cultural specificity of the systematics field in biology. To do this, she examines the discourse of the House of Lords Select Committee on Science and Technology review of the state of systematics in Britain, combining this with ethnographic observation of the problems encountered in practice. In the process, she challenges the inevitability of the digital solutions perspective: “if you build it they will come.”

Systematics is also an interesting case study for exploring what Hine (2005, p. 4) describes as the “political geography” of a discipline, whereby the geopolitical history of a discipline shapes collective notions of social justice in the present. For example, human geographers try to distance themselves from the colonial geographies of the 18th century, and anthropologists are also keen to rectify and distance themselves from the cultural imperialism of the same period. In the case of systematics, access to the world’s most critical taxonomic collections have historically been held by national institutions of developed countries, often with a history of colonization. This centralization of resources and exclusion of developing countries in the past (which are typically the source of the data held in national repositories of specimens) make the rhetoric of open access and data sharing compelling to a sense of social justice within the disciplinary community. The “political geography” of systematics also converges with the “archaic image of systematic biology” (Select Committee on Science and Technology, 2002, in Hine 2005, p. 6) to ensure that digital technologies are perceived within systematics to hold both “practical purchase” and “symbolic qualities.” Indeed, Hine (2005) argues that “particular sets of expectations around digital technologies have played an important part in shaping the discipline’s response to the internet, playing out both on a level of individual practices and in high level policy forums where the discipline’s activities are evaluated” (p. 5).

Researchers who adopt the critical/reflexive perspective argue that the compelling rhetoric around the transformative nature of digital technologies will preclude appropriate discussions about the real needs for digitizing resources and for evaluating the needs of users. The aim in this case is to critique the rhetoric to make it less techno-centric and to open up the debate to a wide set of stakeholders, some of whom might otherwise be marginalized.

Advocacy/Steering and Aligning Structures

Advocacy is mainly aimed at general issues affecting e-science, rather than specific research projects or agendas. These include fostering structures that enable communication and collaboration across disciplinary, institutional, and geographic frontiers. For example, in his evaluation of the legal and economic dynamics of scholarly publishing, Nentwich (2006) seeks a socio-technical solution to current quality control mechanisms that are seen as a threat to open science. He argues that “what is needed is hardly more technology, but organization, management procedures and legal as well as economic knowledge at the interface of technology and the social environment” (Nentwich, 2006, p. 201).

There are parallels between Nentwich’s (2006) discussion of digital publishing and Hine’s (2005) case study of systematics in biology. It may be, for example, that the drive for open digital access to information about species will undermine processes whereby cataloguing was traditionally peer reviewed, just as Nentwich’s advocacy of open access may undermine the traditional peer review structure of journals. The de-commodification of information that Nentwich advocates also raises issues (again, echoing Hine’s work) of how archives and collections will be maintained and disseminated effectively in an open-access system and who will fund such resources.

Such general issues of resources and organizational frameworks are typically at the forefront of the “advocacy” social science approach. This approach is illustrated by the work of David and Spence (2003), in their account of the institutional infrastructures for e-science. They seek out the economic advantages and pitfalls of the Internet’s open architecture and culture. In the process, they discuss, among other issues, the question of how such an infrastructure could be funded, including whether the traditional models of journal publisher revenues can be maintained in open publishing, and how intellectual property rights can be implemented when data are made openly accessible or shared. From the point of view of the structural/advocacy position of David and Spence, it is primarily the legal, political, and administrative structures that have been experienced as constraints to GRID-based collaboration by the first generations of Internet and Grid users.

Along these lines, David and Spence propose that the legal barriers relating to intellectual property rights and other ownership issues, such as confidentiality of data, can be overcome and potentially have the capability to open up frontiers to communication and collaboration. To address these issues, they outline a framework for the governance of e-science via a new organization that would remove some of the current uncertainties via regulation and laws. David and Spence also want to put in place contractual arrangements for inter-institutional collaboration. The details of this institutional body need not concern us here; the main point is that they are concerned with remedying the institutional shortcomings in e-science.

A specific area-related example of David and Spence’s work that they touch on is the movement towards open access publishing. This movement aims to make research publications that are currently available only in the form of expensive journal subscriptions freely available. A number of initiatives are currently ongoing to try to put this idea into practice, with varying degrees of success. A major question in this case is whether, and in what form, open access will prevail over the established fee-paying models (Armbruster, 2005). Regardless of the outcome, however, in this case social science can point to the advantages (and disadvantages) of an open publishing model, make policy recommendations accordingly, and promote putting these into practice.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Background to e-Science
  5. The National and Global Shaping of Research Programs
  6. Approaches to e-Science
  7. Discussion
  8. Conclusions
  9. Acknowledgments
  10. References

In this article, we have used a limited number of examples to illustrate each of the main types of social science approaches and how they have been applied to e-science. Although we have categorized these approaches into four discrete types, the two continua in Table 1 represent tensions and synergies between each approach. For some types of e-science problems it may be beneficial if one or more approaches were adopted. For example, a project concerned with the sustainability of e-science may feasibly employ a usability/practical approach as a complement to a value free/attempted neutrality approach. However, the various approaches—whether practical in contrast to critical, or neutral rather than advocacy—are likely to shape e-science in different ways.

Potential Impact

The critical/reflexive stance towards e-science priorities is likely to have an indirect effect on e-science, seeking to shift the debate about its overall aims and opening up a space for a wider debate about the role of science in society that includes more than just researchers. Advocacy may also contribute to steering e-science, and may do so in a direct way, but one question mark is bound to be the strength of the influence of social science on policymaking.6 In this case, it is important to consider the forums—policy organizations, scientific advisory boards, academic conferences, and the like—in which these policy debates take place.

The examples illustrate that the line between the social, technical, legal, and economic are becoming blurred. In usability approaches, the line between technical and social improvements may be hardest to draw, even if, as in the eDiaMoND case described earlier, there were many researchers with a social science background involved in this part of project. In a neutral approach, social science can learn much about how science works in the case of these large technological systems. E-science is a novel form of science and technology, raising questions about whether new concepts, such as the “collaboratory,” have become necessary for new constellations of research. Similarly, is it necessary to rethink traditional analytical categories in the sociology of science, such as the notion of “disciplinary cultures,” in view of the new combinations of disciplines involved in e-science research (Becher & Trowler, 2001; Klein, 1996)? Questions can thus be posed about whether collaborative research represents a fundamental shift in the nature of science or whether it reflects a continuum along a single dimension.

Potential Synergies

It will be obvious that, despite our efforts to separate the four social science approaches, there is much overlap between them. This applies not only to the thin line between critical and neutral approaches, but also to critical approaches and usability: For example, the usability issues that were described in the eDiaMoND study led the researchers to say that their findings aimed to “respecify” (Jirotka, et al., 2005, p. 395) the original “visions” of the project, just as Hine’s work wants to re-size the visions of systematics. Note, however, that in the eDiaMoND case, this respecification of visions was done in order to “make their accomplishment more achievable in the long run” (Jirotka, et al., 2005, p. 395), whereas Hine’s (2006) aim is to throw into doubt the very achievability of the visions.

There is also an age-old tension between theory and practice7 or, in this case, between becoming so involved in constructive engagement (the policy and usability approaches) that all the necessary distance of the critical or neutral approaches is forsaken, with the result that little of value is gained for social science understanding of the implications of new research technologies. Or the opposite: that the distance between the neutral or critical perspectives and practice is so large that no lessons can be derived for guiding the development of new research technologies. It will only be possible to gauge these tensions—and how they pull in one or other direction in the case of the e-science—once e-science projects and infrastructures have matured.

Another point is that different social science approaches may influence e-science at different stages. Thus, for example, critics of the visions may dominate in the early phases of the e-science enterprise and may fade as the visions—in whatever form—become translated into mundane practices or when they have faded into practical realities. Or again, usability concerns tend to be an “add-on” at the early stages of research, but if they are an afterthought, it may be too late to integrate usability issues to good effect.

Units of Analysis

A further question arises about appropriate levels of evaluation: How should e-science initiatives be analyzed, in the context of national or international programs, disciplines, clusters of collaborating projects, or individual projects? If the focus is on national or international research programs, this is most likely to be the province of advocacy and neutral approaches. One example here is assessments or evaluations of whether funding priorities have been deployed in the most effective and beneficial way. Another example is comparative studies of collaborative projects, of disciplinary cultures, or of communication patterns that are often carried out on a national level. The issues for the usability and critical approaches tend to be on a different level as these two approaches typically proceed from individual cases and use these as examples to raise larger issues.

Research (as opposed to funding) priorities are also typically addressed at the national level, although in this case the critical approach also comes into the foreground by raising the question of whether these programs are targeting the most urgent areas from the point of view of science and engineering, or about the urgency of social science issues and projects. The main purpose here will be to contribute to steering or reshaping the overall trajectory of research. From the perspective of research councils and science policy makers, e-science is a push toward the coordination of research efforts, creating resources that can be shared and therefore also avoiding duplication and making the best tools more widely available. Policy is further driven by the belief that these tools are crucial to competitive advantage. A key question, then, is to assess whether this objective is being realized; this will have an advocacy dimension but will also be subject to imperatives for neutral measurement.

Social Science and Computer Science

The range of disciplines currently represented by e-science has been called into question (Wouters & Beaulieu, 2006), with the concern that e-science has mainly been driven by computer science developments. A more general question than the instrumental assessment of e-science is: Has e-science lived up to the expectations of creating tangible benefits? Or, do we need to see e-science as simply another step in a “computerization movement” (Hine, 2006, drawing on Iacono & King, 2001). Hine implies that the impetus behind e-science is more fashion than substance. To answer this, we would need to establish whether e-science is “pushed” from computer science or “pulled” from domain applications. To date, many researchers including Hine (2006) and Wouters and Beaulieu (2006) would regard this as a computer science “push,” with the expectation that this push will diminish over time and become more of an applications “pull.” But it may be possible to answer this question only when much more e-science development has taken place, as part of a more general argument about how the advances in computer science have served to enhance a number of scientific disciplines—some more than others. In the meantime, there will be strong demands to improve the uptake of e-science tools, and these will fall especially on computer scientists, since this is where usability research has traditionally been located.

In terms of the question about the research disciplines involved in e-science, identifying the links between computer science, on the one hand, and the particular discipline supported by computer science, on the other, is an obvious starting point. This is because involvement with computer science applies to all e-science projects. One question that has been raised is whether it could be said that in some cases, computer science tools are being developed for their own sake, without any immediate regard for support of, or use by, another discipline. But the wider question of how well the disciplines collaborate is bound to be a question in the other three approaches as well.

Common Ground and Diversity in the Social Sciences

We would argue that the newness of developments in e-science should constitute a core question for all four approaches, critically informing them even if they are not centrally concerned with it. The reason for this is that this newness will pinpoint, in a neutral way, the element of e-science that sets it apart from other kinds of research. However, the social implications of science and technology in-the-making fall most closely within the social sciences into the subdiscipline of the sociology of science. One question that can therefore be raised is the extent to which this subdiscipline informs the four approaches that have been identified here, and how it is placed within the social sciences to contribute to the understanding and shaping of e-science.

One perspective within the sociology of science suggests “high-consensus rapid-discovery science” advances quickly because technical equipment can be cloned and used to replicate results (Collins, 1994, p. 155). E-science can be seen as an example of this, with technical apparatuses for analysing and storing data and sharing resources that equips researchers with the apparatus to extend and build on each others’ findings.8 But e-science does not just consist of such research apparatuses for manipulating data and manipulating the physical environment; e-science tools can also be means for scientific communication.9 With e-science, the scope and scale of collaboration have generally increased, which means that communication itself poses formidable organizational problems and yet extends the technological infrastructure of research. Put differently, e-science faces new challenges involving the coordination and control of research (Fry, 2006). Collaboration-at-a-distance could thus arguably also be seen as a way of underpinning consensus and replicability by using research instruments.10 The balance between the two—research instruments for manipulating the objects of research and instruments for communication—is not yet clear, but one interesting point of departure for all four approaches might be to ask: What is the more important part of e-science efforts: more distributed ways of doing research at a distance (communication) or more powerful instruments for doing research?

Another useful perspective within the sociology of science distinguishes between different disciplines and their degree of task certainty and mutual dependence (Whitley, 2000). Many tasks in e-science seem to be governed by a consensus or shared approach even at this early stage—a degree of task certainty—across natural and social e-science: putting data and other resources online, sharing expensive instruments, and sharing scarce computational resources. At the same time, there is considerable task uncertainty about the value and aims of other e-science projects, such as the idea of developing a semantic web to cut across and unify knowledge databases.

Apart from task certainty versus uncertainty, another characteristic varies in how scientific disciplines are organized, according to Whitley (2000): mutual dependence. In one sense, e-science “automatically” increases mutual dependence since it invariably ties researchers in different disciplines and also different scientific endeavors together through common goals. In this case it is possible to identify the extent that e-science efforts vary in their mutual dependence or how closely coupled the link between computer science and domain science is. Again, however, there is a flip side, inasmuch as this mutual dependence may not materialize between computer scientists and domain scientists or across other cross-disciplinary efforts. Nevertheless, as with task certainty and uncertainty, this is a useful way to categorize e-science efforts.

A final set of ideas that can be taken from the sociology of science is that across all e-science projects, one effect which seems mostly unintended is that e-science projects make visible or explicit many processes that were previously invisible or implicit.11 Examples abound, and include workflow organization, intellectual property issues, and inter-institutional commitments. E-science projects do this partly because computerization necessitates, for example, that data are put into a certain format or stored in a certain way or made accessible across a network. These e-science developments all require standardization of rules and procedures that may previously have been unstated or unformalized or do not apply to non-digital material. At the same time, e-science may also hide or marginalize other research efforts that are not amenable to the process of rendering research into digital formats.

The newness of e-science is perhaps least relevant to usability, because here the main concern is a pragmatic one: making tools work. Still, the concepts that have just been mentioned will locate the focus of where bottlenecks may lie, such as in the new forms of interdependence between researchers. And a critical approach also will not be concerned primarily with this novelty, but rather with the extent to which the vision lives up to or reifies the realities of the sciences. Nevertheless, a useful way of gauging whether e-science becomes translated into effective scientific practice is how e-science projects cope, for example, with the challenges of task uncertainty. Advocacy is concerned with whether e-science is enhancing research efforts, and so will focus on whether, using Whitley as a benchmark, e-science is providing a basis for more powerful ways of organizing research. A neutral approach will need to establish which of these concepts are adequate to the task of describing this new form of science and how well they fit different e-science developments.

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Background to e-Science
  5. The National and Global Shaping of Research Programs
  6. Approaches to e-Science
  7. Discussion
  8. Conclusions
  9. Acknowledgments
  10. References

These ideas about making research more visible (and invisible), combined with Whitley’s (2000) schema and Collins (1994), provide a good starting point for all four approaches, because in interlinked ways they pinpoint the novel features of e-science. How do new instruments further e-science research? By enhancing distributed collaboration or by providing more powerful instruments to build on and extend previous research? Are the structures of e-science becoming more transparent because e-science projects share novel research tools, or because the means of collaboration and dissemination are becoming more similar across different disciplines? Does e-science exemplify increasing task certainty in terms of building the new tools for these two purposes, and how much mutual dependence is there between those who build the new tools and those who need and use them? Finally, Collins and Whitley are also useful in that they adopt a long-term historical perspective on the sciences and an all-embracing social science perspective (as opposed to one confined to a particular discipline or subdiscipline within social science). Whether task certainty and mutual dependence go hand-in-hand for e-science is a question for future research. Still, it will be possible to identify these characteristics fairly easily because of the way in which computerization renders them more visible (Schroeder, 2006).

At the same time, these questions about e-science can also be turned back onto the sociology of science. Is the focus within the subdiscipline of sociology of science coupled to its object, such that it promotes the advance within this subdiscipline, or is the subdiscipline promoted by means of engaging pragmatically to help reshape the new technologies? Is the diversity of social science perspectives (we have identified four) a benefit, contributing different inputs to different levels of research policy and practice? Or is this diversity evidence of the fragmentary nature of the social sciences, without a close coupling to the object of their research? These questions must be posed with the understanding that the shift of resources and collaboration within online networks represents an entwining of a large technological system with the social environment of research institutions and settings. Can such a socio-technical system be reshaped by social scientists who analyze research in diverse ways, or will the social sciences themselves rather become engulfed by the new system of tools that are reconfiguring research across different arenas and institutions of knowledge production? In short, how is the object of research (e-science) coupled to the different social science perspectives that have been discussed? Categorizing the different social science perspectives in terms of their potential relation to the object of research would be a useful start and would provide us with a better purchase on e-science, just as e-science, in turn, re-shapes the sciences.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Background to e-Science
  5. The National and Global Shaping of Research Programs
  6. Approaches to e-Science
  7. Discussion
  8. Conclusions
  9. Acknowledgments
  10. References

We would like to thank Nick Jankowski, two anonymous referees, and the participants at the “Social Science Perspective on e-Sciences” workshop at the Second International Conference on e-Social Science, Manchester, June 28-30, 2006 for helpful comments and suggestions. The work for this article has been supported by ESRC grant RES-149-25-1022 and is part of the Oxford e-Social Science (OeSS) project.

Notes
  • 1

    David (2004) favors an interpretation that restricts e-science to the intersection of Grid and collaboratory research, but it remains to be seen if the term “Grid” retains its currency. And “collaboratory” can be taken to imply research that entails distributed work or when non-collocated researchers work together, which is an important subset of e-science research but not the whole (for example, single users accessing remote e-science instruments would fall outside of this definition).

  • 2

    We use e-science here not to indicate the U.K. program, but as a generic label for all of these programs.

  • 3

    In fact, the Oxford e-Social Science project is the only major research project (research node) that was funded under the “social shaping” strand, with the other six nodes developing tools. The same balance applies to the smaller grants and earlier pilot projects that were funded by the ESRC.

  • 4

    This is not to say that advanced computing developments in the arts and humanities need to follow the same path as the natural and social sciences; for example, there have been developments in mega-corpora and probabilistic techniques in linguistics and, more recently, Virtual Research Environments (VREs) in the humanities.

  • 5

    Woolgar and Coopmans (2006) use the term “virtual witnessing.”

  • 6

    One question here, for example, might be whether reflexive sociology of science might have more of an impact on steering e-science than innovation studies.

  • 7

    Max Weber’s “Objectivity” essay (1949) remains a key starting point in this debate; see also Rule (1997) for a recent overview.

  • 8

    One implication of this view is that the social sciences, which so far have not made much use of research apparatuses in this way, stand to benefit more than the natural sciences that have done this for some time.

  • 9

    Scientific communication here is taken in the broad sense to include both formal communication via publications and informal communication, which is the sharing of results that does not take place in the public domain. One potential implication of e-science is that the line between the two may be blurring.

  • 10

    Koku, Nazer, and Wellman (2000) argued some time ago that electronic communication does not do away with face-to-face communication among researchers; here, as elsewhere, the two are complementary. Nevertheless, Koku, et al. also noted that communication among researchers in a variety of different modalities has become denser—that is, more frequent and more multi-channel. E-science contributes to this trend, complementing rather than displacing existing ways of doing research.

  • 11

    This has been discussed by Fuchs (1992) outside of the context of e-science, whereby researchers make their findings presentable to their peers, or they standardize and make their instruments more robust, contributing to a process of enhancing visibility and explicitness in scientific practices.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Background to e-Science
  5. The National and Global Shaping of Research Programs
  6. Approaches to e-Science
  7. Discussion
  8. Conclusions
  9. Acknowledgments
  10. References
  • Armbruster, C. (2005). Open access in social and cultural science: Innovative moves to enhance access, inclusion and impact in scholarly communication. Social Science Research Network. Retrieved September 28, 2006 http://papers.ssrn.com/sol3/cf_dev/AbsByAuth.cfm?per_id=434782
  • Atkins, D. E, et al. (2003). Revolutionizing Science and Engineering through Cyberinfrastructure. Report of the National Science Foundation Blue-Ribbon Advisory Panel on Cyberinfrastructure. Arlington, VA: Directorate for Computer and Information Science and Engineering, National Science Foundation.
  • Barjak, F. (2006). Personal communication, December 18, 2006.
  • Becher, T., & Trowler, P. (2001). Academic Tribes and Territories: Intellectual Inquiry and the Culture of Disciplines (2nd ed.). Milton Keynes, UK: Open University Press.
  • Berman, F., & Brady, H. (2005). Final Report: NSF SBE-CISE Workshop on Cyberinfrastructure and the Social Sciences. Retrieved December 7, 2006 from http://vis.sdsc.edu/sbe/reports/SBE-CISE-FINAL.pdf
  • Collins, R. (1994). Why the social sciences won’t become high-consensus, rapid-discovery science. Sociological Forum, 9(2), 15577.
  • Cummings, J., & Kiesler, S. (2005). Collaborative research across disciplinary and institutional boundaries. Social Studies of Science, 35(5), 70322.
  • David, P. A. (2004). Towards a cyberinfrastructure for enhanced scientific collaboration. OII Research Report No. 4. Oxford: Oxford Internet Institute, University of Oxford. Retrieved September 28, 2006 from http://www.oii.ox.ac.uk/resources/publications/RR4.pdf
  • David, P. A., & Spence, M. (2003). Towards institutional infrastructures for e-science: The scope of the challenge. OII Research Report No. 2. Retrieved December 5, 2006 from http://www.oii.ox.ac.uk/resources/publications/RR2.pdfm
  • Drori, G., Meyer, J., Ramirez, F., & Schofer, E. (2003). Science in the Modern World Polity: Institutionalization and Globalization. Stanford, CA: Stanford University Press.
  • Finholt, T. A. (2002). Collaboratories: Science over the Internet. In S. J.Lita, S. D.Nelson, & A. H.Teich (Eds.), AAAS Science and Technology Policy Yearbook (pp. 339344). Washington, DC: American Association for the Advancement of Science.
  • Fry, J. (2006). Coordination and control of research practice across scientific fields: Implications for a differentiated e-science. In C.Hine (Ed.), New Infrastructures for Knowledge Production: Understanding e-Science (pp. 167187). Hershey, PA: Information Science Publishing.
  • Fuchs, S. (1992). The Professional Quest for Truth. Albany: State University of New York Press.
  • Galison, P., & Hevly, B. (Eds.) (1992). Big Science: The Growth of Large-Scale Research. Stanford, CA: Stanford University Press.
  • Hey, A., & Trefethen, A. (2003). The data deluge: An e-science perspective. In F.Berman, G. C.Fox, & AnthonyHey (Eds.), Grid Computing: Making the Global Infrastructure a Reality (pp. 809824). Chichester, UK: John Wiley & Sons, Ltd.
  • Hine, C. (2005, November). The politics and practice of accessibility in systematics. Past, Present and Future of Research in the Information Society. An official side event preceding Phase II of the World Summit on the Information Society (WSIS), Tunis.
  • Hine, C. (2006). Computerization movements and scientific disciplines: The reflexive potential of new technologies. In C.Hine (Ed.), New Infrastructures for Knowledge Production: Understanding e-Science (pp. 2647). Hershey, PA: Information Science Publishing.
  • Hughes, T. (1998). Rescuing Prometheus. New York: Pantheon Books.
  • Jeffreys, P. (2005). Presentation at Oxford Internet Institute workshop. December 2, 2005.
  • Jirotka, M., Procter, R., Rodden, T., & Bowker, G. C. (2005). Collaboration and trust in healthcare innovation: The eDiaMoND case study. Computer Supported Cooperative Work, 14(4), 369398.
  • Iacono, S., & Kling, R. (2001). Computerization movements: The rise of the Internet and distant forms of work. In J.Yates, & J.Van Maanen (Eds.), Information Technology and Organizational Transformation: History, Rhetoric and Practice (pp. 93135). Thousand Oaks, CA: Sage.
  • Klein, J. T. (1996). Crossing Boundaries: Knowledge, Disciplinarities, and Interdisciplinarities. Charlottesville, VA: University Press of Virginia.
  • Kling, R., & McKim, G. (2000). Not just a matter of time: Field differences and the shaping of electronic media in supporting scientific communication. Journal of the American Association for Information Science, 51(14), 13061320.
  • Koku, E., Nazer, N., & Wellman, B. (2000). Netting scholars: Online and offline. American Behavioral Scientist, 43(10), 17521774.
  • Nentwich, M. (2003). Cyberscience: Research in the Age of the Internet. Vienna: Austrian Academy of Sciences Press.
  • Nentwich, M. (2006). Cyberinfrastructure for next generation scholarly publishing. In C.Hine (Ed.), New Infrastructures for Knowledge Production: Understanding e-Science (pp. 189205). Hershey, PA: Information Science Publishing.
  • Rule, J. (1997). Theory and Progress in Social Science. Cambridge, UK: Cambridge University Press.
  • Schroeder, R. (2006, June). e-Sciences: Infrastructures that reshape the global contours of knowledge. Paper presented at the Second International Conference on e-Social Science, Manchester. Retrieved September 28, 2006 from http://www.ncess.ac.uk/events/conference/2006/papers/
  • Select Committee on Science and Technology. (2002). What on Earth? The Threat to the Science Underpinning Conservation. London: House of Lords.
  • Sonnenwald, D. (2006). Collaborative virtual environments for scientific collaboration: Technical and organizational design frameworks. In R.Schroeder R. & A.-S.Axelsson (Eds.), Avatars at Work and Play: Collaboration and Interaction in Shared Virtual Environments (pp. 6396). Dordrecht, Netherlands: Springer.
  • Taylor, J. (2001). Presentation given at U.K. e-science meeting, July, London. Retrieved September 28, 2006 from http://www.ncess.ac.uk/insight/tutorials/e-social_science/vision/
  • Vann, K., & Bowker, G. (2006). Interest in production: On the configuration of technology-bearing labours for epistemic IT. In C.Hine (Ed.), New Infrastructures for Knowledge Production: Understanding e-Science (pp. 7197). Hershey, PA: Information Science Publishing.
  • Weber, M. (1949). “Objectivity” in social science and social policy. In M. Weber, The Methodology of the Social Sciences (pp. 49112). New York: The Free Press.
  • Westwick, P. (2003). The National Labs: Science in an American System, 1947-1974. Cambridge, MA: Harvard University Press.
  • Whitley, R. (2000). The Intellectual and Social Organization of the Sciences (2nd ed.). Oxford: Oxford University Press.
  • Woolgar, S., & Coopmans, C. (2006). Virtual witnessing in a virtual age: A prospectus for social studies of e-science. In C.Hine (Ed.), New Infrastructures for Knowledge Production: Understanding e-Science (pp. 125). Hershey, PA: Information Science Publishing.
  • Wouters, P., & Beaulieu, A. (2006). Imagining e-science beyond computation. In C.Hine (Ed.), New Infrastructures for Knowledge Production: Understanding e-Science (pp. 4870). Hershey, PA: Information Science Publishing.
About the Authors
  1. Ralph Schroeder is James Martin Research Fellow at the Oxford Internet Institute at Oxford University. Before coming to the OII, he was Professor in the Department of Technology and Society at Chalmers University in Gothenburg. He is currently an investigator on the Oxford e-Social Science (OeSS) Project: Ethical, Legal, and Institutional Dynamics of e-Sciences. His publications include books and essays on the sociology of science and technology, Max Weber, and virtual environments.

    Address: Oxford Internet Institute, 1 St. Giles, Oxford OX1 3JS, UK

  2. Jenny Fry is currently researching the legal, ethical, and institutional barriers to e-science as part of the Oxford e-Social Science (OeSS) Project. She received her Ph.D. in Information Science in 2003 from the University of Brighton. She has held postdoctoral fellowships at the Royal Netherlands Academy of Arts and Sciences, Amsterdam and the School of Information and Library Science at the University of North Carolina, Chapel-Hill. Her research has been concerned with the disciplinary shaping of networked digital resources and information practices.

    Address: Oxford Internet Institute, 1 St. Giles, Oxford OX1 3JS, UK