Copyright Sociological Research Online, 2002

Christine Hine (2002) 'Cyberscience and Social Boundaries: the Implications of Laboratory Talk on the Internet'
Sociological Research Online, vol. 7, no. 2, <http://www.socresonline.org.uk/7/2/hine.html>

Abstract

This paper examines the use of an online forum for the discussion of laboratory science. It is argued that such forums are significant in the light of claims made for the impact of information and communications technologies (ICTs) on scientific research, and of broader debates about the role of ICTs in reconfiguring social boundaries. It appears that the impacts of ICTs on scientific research are likely to be diverse and unpredictable, in line with emerging findings in other application domains. In particular, the potential to break down the boundaries between science and lay persons, and between different areas of scientific research, is likely to be limited by the ways in which particular forums are preserved as bounded spaces for specific specialisms. In the case of the forum studied in this paper, discursive practices function to re-establish laboratory boundaries in the online setting. Laboratory talk on the Internet may help to break down barriers between individual laboratories, but is not, in itself, any more accessible to lay people than talk in the private spaces of the laboratory.

Keywords:

Boundary; Discourse; Internet; Laboratory; Science; Sociology

Introduction

1.1

This paper examines the use of an online forum for the discussion of laboratory science. The aim of the paper is to consider how the forum is made into an appropriate place to discuss the details of laboratory practice, thus providing a place for the exchange of advice that participants can use in their work. To a certain extent this is a straightforward description of the use of an electronic forum by a professional group to discuss practical work. However, the fact that the activity discussed within the forum is scientific research gives it an additional significance. The use of electronic discussion forums in science is topical in science policy contexts, where it has been suggested that information and communications technologies (ICTs) have the potential to bring together a spatially dispersed science community into a more efficient and effective body. The forum discussed here provides an opportunity for discussing in detail how this might happen. Online scientific forums also, however, raise important issues for the sociology of scientific knowledge, which has long held that the special cultural status of scientific knowledge is maintained by the careful distinction between formal public scientific communication and less formal repertoires used in private contexts. The public forum analysed here is used to discuss informal aspects of scientific practice, bringing together researchers working in different laboratories at the detailed level of daily practice, and apparently breaching the usually clear distinction between public and private science. These observations can therefore also be assessed in the light of their implications for understanding of the laboratory boundary in sociology of scientific knowledge. Addressing these concerns requires a discourse analytic approach to understanding the ways in which participants' contributions are formulated.

1.2

The first section of the paper reviews the recent policy and academic interest in cyberscience and proposes that these concerns suggest a focus on how social boundaries are shaped and reshaped in science in the face of ICTs. The following section then discusses work in the sociology of scientific knowledge on the role of formal and informal communication in science which establishes an interest in how embodied practice within the laboratory and communication beyond the laboratory boundary are linked. This sets up a series of questions concerning the use of online forums by scientists: how do scientists establish their credibility online in the absence of peer review or personal knowledge of one another; how are the local concerns of laboratory practice translated into a global forum; what sources of authority are drawn upon in doing so? Considering these questions in the light of observations of a particular forum then sets the scene for a discussion which reviews the potential of online forums to reconfigure scientific social boundaries.

Cyberscience

2.1

There has been considerable interest of late in how research in general, and scientific research in particular, might be affected by the use of ICTS. Much of this interest revolves around possibilities that ICTs might help to make research more efficient. By decreasing the costs of acquiring information and communicating with other researchers, the potential is that new technologies could reduce the chances of duplicated effort, increase collaboration and bring previously isolated groups of marginal researchers into a wider scientific community. The persuasiveness of this form of thinking is demonstrated by the level of attention it has received in policy forums: a European Technology Assessment working group considered the issue of how European science could be transformed through ICTs (ETAN, 1999); a series of three OECD conferences and reports in 1996, 1998 and 2000 discussed 'The Global Research Village' focusing on the potential of ICTs within the science system (OECD, 2000); and, going back to 1990, Robert Maxwell told the audience at the British Library Dainton Lecture how IT was going to make science more efficient (Maxwell, 1990). In the UK, information technology for science has received considerable recent investment under the guise of e-science (see <http://www.research-councils.ac.uk/escience/> ). It seems reasonable to conclude that there is a belief in science policy circles that ICTs have the potential to make science more efficient, and possibly to promote the creation of better scientific knowledge.

2.2

Outside the immediate policy context, groups of researchers with an interest in science communication and science practice have begun to talk of cyberscience: a new form of scientific research in which ICTs are integrated with and transform research practice (Nentwich, 1999; Wouters, 2000; Beaulieu, 2001). The term cyberscience suggests the generation of new forms of knowledge through collaboration and interaction between information technologists and scientific researchers and through the development of digital resources. Communication may play a large part in bringing together previously separated groups of researchers and hence altering the social dynamics of knowledge production. As Matzat (1998) comments, computer-mediated communication has been seen as the means of broadening or breaking down altogether the "invisible colleges" that exclude marginal scientists. It is to this communicative aspect of cyberscience that the current paper contributes empirical observations, by considering just how an online forum might become a site for communication between geographically dispersed researchers. The paper looks at a well-established form of ICT use in science, which is available even to scientists working in fields where large scale IT resources are not implicated. If such electronic forums can be used to share the detail of laboratory work, rather than simply reproducing aspects of formal scientific communication, then the knowledge production work of laboratory scientists may indeed come to take on a new form we could describe as cyberscience.

2.3

Both the hopes for improvement of scientific research efficiency in policy circles and the sociological thoughts of new ways of doing science revolve around the perceived abilities of ICTs to break down social boundaries. We are familiar with the rhetoric surrounding the ability of ICTs to promote social re-organization across time and space. We also know that, appealing as this rhetoric may be, in practice the upshot of introduction of ICTs is often unevenly socially distributed and unpredictable (for recent general discussion see Castells, 2001 and for a critical approach, Woolgar, 2002). It seems likely that the impacts of ICTs on scientific research will be similarly diverse. In particular, different disciplines will probably utilise online communication forums and electronic publishing differently (Kling and McKim, 2000), just as they have other information technologies such as bibliographic searching tools (David and Zeitlyn 1996, Zeitlyn et al 1999). It is therefore not appropriate to begin analysis on the basis that "effects" of online communication will be observed. As Hine (2000) argues, the uses to which particular online spaces are put depend both on the expectations which participants bring to these spaces, and the social dynamics which they encounter once there. The observations in this paper are therefore aimed at considering the process through which an online space comes to be used in particular ways, such that social boundaries are either preserved or redrawn and such that it is, or is not, rendered as a continuation of laboratory space. In order to contextualise the analysis contained in this paper, it is necessary to review work in the sociology of scientific knowledge which concerns the role of social boundaries in the production of scientific knowledge. The next section of the paper will consider the role of local context and tacit knowledge in laboratory practice and the ways in which these informal practices are withheld from formal scientific communication.

Formal and Informal Communication in Science

3.1

It has been argued that science historically has a specific spatiality, and that it matters for the form of scientific knowledge where the work of scientific research is practised (Ophir and Shapin, 1991; Shapin, 1995; Livingstone, 1995; Galison, 1999). The key scientific space in sociology of scientific knowledge is the laboratory. Laboratory ethnographies were the key tool for establishing the locality of scientific research practices, and hence examining the social practices of scientific knowledge construction (e.g. Knorr-Cetina 1981; Zenzen and Restivo, 1982; Lynch 1985; Latour and Woolgar 1986). Such studies emphasised the continuity of scientific practice with ordinary everyday practice, thus challenging the privileged epistemic status of scientific knowledge. Shapin (1998) points out that the 'localist' focus in the sociology of scientific knowledge on practices within the laboratory has left it with the not inconsiderable problem of explaining how it is, that given scientific knowledge is actually so like everyday knowledge in its production, it should be able to travel and should appear to be universal. Some answers to this issue focus on the way in which relations between the laboratory and the outside world, and between different laboratories, are carefully managed. The laboratory is a private place, with its own epistemic culture (Knorr- Cetina 1999), where public knowledge is produced. Indeed, laboratories were quite often deliberately built as private places to permit the exclusion of outsiders (Galison, 1999). The laboratory communicates knowledge to the outside and to other laboratories through documents and demonstrations, produced via processes of inscription, standardisation and translation which make scientific facts into stable entities that will travel outside the laboratory, stripping away the human contingencies and uncertainties that prevail inside the laboratory (Latour and Woolgar, 1986). Work is coordinated across different laboratories and disciplines via standardised packages (Fujimura, 1996) or boundary objects (Star and Griesemer, 1989), which are sufficiently stable to travel, yet sufficiently flexible to be interpreted differently in their different contexts of use.

3.2

The upshot of these communication practices is to leave the details of what happens in the laboratory largely unspoken outside the face-to-face encounters of those directly involved. This is far from saying that informal communication is therefore not important. Tacit knowledge is often held by researchers to be crucial to success. Collins (1975, 1985) paid particular attention to the circumstances under which a scientific experiment was considered to have been replicated. In describing attempts to replicate the experiments of others, Collins' scientist interviewees held that formally published descriptions of method were unlikely to be sufficient for a successful replication. Jordan and Lynch (1998) also showed how important informal communication was held to be in learning a new technique. Gilbert and Mulkay (1980: 284) found that laboratory skills were considered particularly sensitive, scientists who were interviewed holding that such skills could "only be understood satisfactorily by close contact with someone who is already proficient". The understandings involved could be far removed from formal scientific rationales (Cambrosio and Keating, 1988). Lynch (1988: 266) describes how laboratory concerns take a quite different form to more formally expressed scientific knowledge:

Other understandings about animals, rarely communicated in accounts of laboratory methods, pertained to rats as holistic living creatures. These understandings were communicated informally, and were not validated through rigorous testing. They were part of the everyday life of the laboratory, consisting of various sorts of tacit 'know-how', recipe knowledge and experimental craft that enabled practitioners to deal with the contingency of 'handling' laboratory rats.

3.3

Sociology of scientific knowledge therefore holds that the production of scientific work depends on mundane practices within the local contexts of laboratories, and that scientists themselves believe this to be so. Communicating science beyond the laboratory then involves the erasure of local contingencies from accounts of knowledge production. Through analysing scientific discourse in different contexts it is possible to see how the informal and mundane laboratory work gradually vanishes as accounts of scientific knowledge become more public (Latour and Woolgar, 1986). Knorr Cetina (1981) analyses the transformations that a research paper goes through in successive rewrites, as the traces of laboratory practice are gradually removed. The operation of this process at the discursive level is demonstrated by Gilbert and Mulkay (1980, 1984). Their study showed that scientists had different ways of describing their work, depending on the context in which descriptions were formulated. The empiricist repertoire, relying on the conventional understanding of scientific knowledge as empirically grounded and objective, was typically produced in formal contexts, or where an account was being given of the establishment of accepted scientific fact. By contrast a contingent repertoire, stressing external factors or qualities of individual scientists, was often advanced in discussing the reasons for errors. This repertoire was characteristic of less formal contexts. The deployment of these two repertoires in different contexts has the consequence that:

scientists themselves create that appearance of impersonality, detachment and universality which sociologists have customarily regarded as literally descriptive of social action and technical belief in science. Gilbert and Mulkay (1980: 270).

3.4

Following the link between laboratory practice and scientific communication it therefore appears that the more public the communication context, the less we might expect to see the contingencies of local context appear, and the more we might expect to see science presented as objective, detached and grounded in data alone. It is therefore interesting to consider a forum for discussion of laboratory practice which is publicly accessible, and to consider the consequences for the public perception of science which we might predict from the sociology of scientific knowledge's understanding of the role of social boundaries. This question of the publicness of the forum will be reconsidered in the concluding section of this paper. First, however, it is necessary to consider just how communication between scientists in different laboratories is organised, given that, as we have seen, discussion of everyday practices is usually confined to face to face encounters between scientists who are co-located. Communication in academic newsgroups can take a form somewhere between the oral and the written (Hert, 1997), and it is therefore of interest to see just how the participants interact and what aspects of laboratory talk they manage to transfer to the online setting. Specifically, the empirical section of this paper will address the following questions:

• How do scientists establish their credibility online, in the absence of both the peer review processes of the formal communication system and the ability to be directly observed as a successful worker within the laboratory?
• How are the local concerns of laboratory practice translated into the potentially global forum of a newsgroup discussion? How do participants make their contributions authoritative, and does this involve reference to the kind of contingencies that are usually erased from formal scientific communication?

3.5

Answering these sets of questions will enable a consideration of an over-arching question, which makes reference both to debates on cyberscience and to sociology of scientific knowledge's understanding of the laboratory:

• To what extent are laboratory boundaries reconfigured in laboratory talk on the Internet?

Laboratory Talk on the Internet

4.1

The newsgroup chosen for this study was one of the most active (several hundred messages per week) of the "bionet" newsgroups devoted to researchers in the biological sciences. Each newsgroup has a designated topic for discussion, which is referred to in the title of the group, and laid down in the group's founding charter. Some groups are moderated: the messages posted to it are sifted for relevance by a nominated individual before being allowed to appear on the newsgroup. The chosen newsgroup was not moderated, and hence anyone with the appropriate technology was free to post messages. A bionet World Wide Web home page gives access to archives of messages, and more general information on how to use the newsgroups.

4.2

The chosen group is referred to here by the name of "protocols"^[1]. The topic of the protocols newsgroup is described on the bionet World Wide Web site as "Requests for information and lab reagents". Since the group is unmoderated, this remit is policed by consensus. In practice, the majority of the postings in the period observed focus on techniques associated with molecular biology, particularly genetics research. The messages used in examples are drawn from a sample taken during the summer of 1996. As Figure 1 shows, the group peaked in popularity in 1995 and 1996, and has since experienced a steady decline. It is not clear from observing the newsgroup whether this decline is due to dwindling use of the techniques discussed in the group, or decreasing numbers of researchers experiencing problems. The decreasing frequency of postings could, indeed, point to a loss of faith in the newsgroup as a source of advice, but there is nothing in the content of postings in later years to suggest that this is so.

Figure 1 Number of postings in protocols archive per year, 1993-2001

4.3

In order to carry out a more detailed analysis of the group's discourse, and in particular the way in which participants identify themselves and frame their contributions, a sample of messages comprising one month's archived postings from 1996 have been chosen. The month chosen was August. As Figure 2 shows, in 1996 there is a monthly variation in the frequency of postings, with relatively low numbers in the summer months and in December. This cycle is repeated in other years, suggesting that the newsgroup may be reacting to underlying variations in the intensity and focus of laboratory work with seasonal holidays. The chosen sample is thus one of the less active months in one of the more active years in the group's history, comprising 1123 messages. The content of the messages in the sample appears not to depart from the usual style for the group, although the particular techniques and products under discussion vary over time. The spelling and punctuation in quotations from postings have been kept as originally posted, in order to preserve the 'feel' of the originals. Numbers allocated to postings used as examples relate to their index number within the sample corpus.

Figure 2 Number of postings in protocols archive per month in 1996

4.4

As is standard for newsgroups, the topics of discussion are shown in the header of the message, in the "Subject" field. These are in most cases quite specific, detailing the topic of the post down to a specific problem with a specific technique. The relatively high level of posting on protocols may render participants particularly aware of the benefits of clear subject lines. For the purposes of analysis, the messages in the sample of protocols postings were divided into five main groups: query; response; capping response; announcement; and irrelevance.

• Query. These messages raise a new topic for discussion, usually by introducing a problem which the poster is having with their work, or a product or resource which they are hoping to locate.
• Response. These are identified by a subject field which contains "Re:" and often include portions of a previous posting, marked out by ">" at the beginning of each line of the quotation from a previous posting. This practice is automatic with many news readers, and assists in following the sequence of query and response. A query often prompts more than one response. Due to the delay in receiving messages, responders may not have seen the other responses to the query when they respond themselves. This delay generates a certain amount of repetition or contradiction.
• Capping response. Some responses receive responses themselves, either to reinforce or dispute the statements made in a previous response. Sometime, previous postings are quoted to produce a cascade: the initial query is quoted with a leading ">>", and the initial response with a leading ">".
• Announcement. Some messages appear to neither prompt for nor receive response. These are relatively few in the protocols group, but sometimes include information on conferences, courses or job vacancies.
• Irrelevance. Messages in this category contribute nothing relating to the substantive topic of the newsgroup. Some appear to have been sent by mistake, such as a message with an empty subject field containing only one word, "subscribe". Other messages in this category are declared as such by their authors, for example with the subject line "Test - do not read".

4.5

The remainder of this paper will focus on queries, responses and capping responses, as the messages of most interest in considering the issues raised in the introduction to this paper. A broadly discourse analytic approach, as advocated by Potter (1996) and applied to newsgroup postings by Hine (2000), will be employed to examine the ways in which postings to the group are constructed to appear relevant and convincing. A consideration of the ways in which the identity of participants is made available is followed by an examination of the ways in which statements about laboratory practice are made authoritative.

Becoming a Credible Online Scientist

4.6

The first section of this analysis focuses on the means by which participants in the group establish identities and address one another. In this forum few of the participants appear to know each other personally. The group has been in general characterised by a large and changing group of contributors. In the sample month, a total of 644 different authors were responsible for 1123 messages, an average of 1.74 messages per author. Of this 644, 234 or 36.3% posted just one query. There is therefore little chance for many participants to build up a reputation within the group. Participants are also without the usual resources that might be drawn on in a laboratory context to establish the trustworthiness of potential advice givers. In the laboratory one might be influenced by a colleague's obvious success with a particular protocol, or impressed by a reputation for dexterity or knowledge and hence be inclined to seek and act on that person's advice. In the online forum these resources are absent: it is therefore of interest to see in what ways one's trustworthiness might be established. As Shapin (1995: 302) points out, the problem of establishing trustworthiness in other is not new to science, although the means by which it is established may vary:

The recognition of trustworthy persons is a necessary component in building and maintaining systems of knowledge, while the bases of that trustworthiness are historically and contextually variable.

4.7

In this newsgroup, the identity of participants, and hence their trustworthiness as advice-givers, is open for decoding by other participants in a number of ways. The first of these is the message header, which contains the "Subject" field discussed above. The header also includes (minimally) the date of receipt by the news feeder, the originator's electronic mail address ("From:" field), their organization, and the newsgroups to which the message has been sent. A few "From:" and "Organization:" fields contain cryptic information or nicknames, but the vast majority appear to be the names of real people and real organizations. Of the original 644 authors, only 103 or 16% failed to be identified by name and surname in the message header. It therefore appears that in this forum it is normal that, rather than assuming false identities or obscuring who they are, participants will reveal their offline identities.

4.8

Posters of messages may also offer clues to their identity in addition to the automatic message headers. Message 0936, which contains the single line of text:

Simply chopping the top off a blue 1ml pipette works.

in addition to the automatic headers, is unusual in this newsgroup. The majority of messages are signed off with at least a first name, and many authors (199 out of 644 or 30.9% in the sample) include some kind of signature in their postings. These range from a name alone, to the use of a full signature at the end of the message giving contact details, affiliations and qualifications. If we consider authors not identified by name and surname either in header information or in a signature, we are left with only 7.6% of the sample or 49 out of 644 authors remaining anonymous.

4.9

Despite the oft-cited tendency for users of the Internet to indulge in identity experiments of the kind Turkle (1995) describes, it appears that use of apparently real life identifiers is quite common in newsgroups. For many groups where serious discussions are held or factual information is exchanged participants will not appear to indulge in identity play, nor seriously to think that others are doing so (Baym, 2000). In this respect, the protocols newsgroup is on a par with other groups, such as the soap opera discussion newsgroup studied by Baym (2000). Where the protocols group does differ from this group, may be in the extent to which contact details and institutional affiliations are given out. Also, Baym points out that identity play is effectively ruled out in the group that she studied, since many of the frequent participants do meet one another offline. This does not seem to be the case in the protocols group, and it seems likely that here it is the emphasis on giving and receiving scientifically credible advice that lends weight to the crafting of an appropriate identity. Of course, it is quite likely that the header and signature file settings that participants use in this particular group are their default settings for all their email and newsgroup communications: in that case, what we are seeing here is participants establishing what they deem to be an appropriate electronic identity for a professional scientist.

4.10

The majority of the signatures used are formal in nature: some, however, are personalised with quotations or jokes (45 authors out of 644 in the sample, or 7%). A selection of these are shown below:

1. You can give a PC to a Homo habilis and he'll use it, but he'll use it to crack nuts. Message 0745
2. My opinions are my own, but I tend to give them away to anyone who doesn't flee fast enough. Message 0706
3. If it ain't broke yet it doesn't have enough features yet. Message 0819
4. "You! Out of the pool!"---the Gene Police. Message 0820
5. You can tell me.......I'm a doctor. Message 099

4.11

Some of these are "in-jokes", specific to the community of molecular biologists (4), academics (5), or people using computers (1, 3). Others are oblique reference to the need for users of the Internet to make clear whether they are speaking as individuals, or as representative of the organization through which they have access to the Internet, and whose details may well appear in the message header (2). In this kind of joke signature, participants could be seen as making a statement about the community which they see themselves as addressing: the implication is that the audience will be part of the same community and hence share the joke. In this way, they identify both themselves and their audience.

4.12

The information which participants provide about themselves to others, and the ways in which they address one another, contain a mixture of formal and informal modes of communication. Initial addresses to the group are often informal and widely inclusive, evoking a sense of a familiar and friendly (although not intimately known) social setting, in which one can expect to receive favourable treatment. This sense of informality is enhanced by the practice of picking up on a first name in order to reply to a query message, and also by the use of joke signatures which construct community. However, this informality is juxtaposed with a far more formal mode of discourse, in which institutional affiliations and scientific credentials are displayed as a matter of course.

4.13

Observing modes of identification therefore suggests that, while formal scientific credentials and organizational membership are held to be important in this forum, a less formal concern with inclusiveness, community membership and mutual assistance also prevails. Trustworthiness as a giver of scientific advice, and positioning as an appropriate seeker of such advice, is achieved through a combination of both formal credentials and informal devices such as jokes that display appropriate community membership. Such resources display to other participants that a contributor to the group has an appropriate level of contact with laboratory work to make an intervention. Explicit display of credentials in particular may help to overcome any concerns due to lack of direct observational knowledge of the participant as a successful laboratory scientist. However, it seems unlikely that the mere statement of scientific credentials or sharing of a joke should be enough to motivate someone to invest the amount of time, effort and money that may be involved in acting upon advice given in the online forum. It is therefore important to look at the content of queries and responses, in order to examine the ways in which participants construct them to be convincing. Without further ethnographic study in laboratory settings it is, of course, not possible to say simply from observing the newsgroup whether participants do find advice sufficiently convincing to act upon it. That advice is indeed considered plausible can only in this analysis be deduced from the continued popularity of the group and the willingness of participants to contribute. It is the argument here that advice is constructed in particular ways which are aimed at achieving a plausible and trustworthy impression, just as Potter (1996) suggests is the case for factually oriented discourse more generally.

Problem Identification and Resolution.

4.14

Before moving on to consider how advice is rendered convincing, it is first worth considering how queries are positioned. Queries are frequently justified by reference to the circumstances which make this a reasonable question to ask, in a relatively apologetic manner which solicits assistance. For example, message 0266 reads:

Hi

Could someone help me out with the reaction conditions for Mung bean nuclease? I need to convert a NcoI digested site to blunt ends, but I seem to be losing most of my plasmid...

Some info on reaction times and units of enzyme / ug of DNA would be very helpful!

Thanks in advance...

4.15

The message sets a framework, both of the circumstances within the author's experience which made this a reasonable question to ask, and of the kind of reply that would be deemed appropriate. Such contextualisation is common where advice on a technique is sought, but much less frequently found where the author is looking for a discrete resource such as a published protocol or commercially available product. These resource-based questions tend to be both asked and answered in a brief and matter-of-fact fashion as compared to technique-based queries like 0266.

4.16

In the case of this technique-based query, the response in 0676 is short and to the point, tacitly accepting the framework set by the initial query:

I can't remember the units I use, but I incubate @ 30 deg. for 35 min, and then 95 deg. for 10 min to incactivate.

4.17

Although brief, this response does contain a narrative to contextualise the answer in the author's experience. Message 0390, also a response to 0266, also accepts the framework in the original query and provides its own contextualising narrative, but is qualified by providing a description of the expectations one would need to accept in order for its advice might apply:

Hi Francois

Mung Bean Nuclease (like most other DNA modifying enzymzes it seems) generally does what it wants when it wants. I have had the same problem as you but generally succeed in the end. Routinely I use approx. 5U Mung Bean/ug DNA, leave at RT (25?C) for 1 h, add EDTA to 10mM and heat to 68?C for 10 min. I don't know if this is the correct method (!!) but it has worked for me in the past.

Good luck

Don

4.18

The relevant circumstances in this instance include a picture of laboratory work as relying on perseverance and the repetition of routines that "work", in the face of self-willed enzymes and without relying on scientific rationales for all stages of the procedure. The justification of advice can therefore involve a picture of laboratory work much closer to the contingent repertoire identified by Gilbert and Mulkay (1984) than the formalised version of scientific papers. Message 0423 also provides experience-based advice, again with a somewhat anthropomorphised view of the behaviour of enzymes:

It is good idea to titer the amount of mung Bean nuclease one uses in the reaction since it has a tendancy to be overzealous in its activity when excess enzyme is around. I used it to make a blunt end and it resulted in clones with a variety of different termini when sequenced.

I just transformed with ligations from several rxns with diffrent amounts of nuclease and sequenced several clones, many were diffrent but I got the one I wanted!

Anthony Williams Ph.D

4.19

The final statement of success provides the justification for the response: experience-based responses are required here, but more specifically they should be responses that can point the author of the original query in the direction of some action, based on scientific rationale or not, that someone else has witnessed as successful. This resonates strongly with the beliefs within the laboratory in the power of learning from someone already skilled at the technique, rather than relying on published protocols alone. This is not to say that scientific rationales are never used, or that published information is always disregarded: rather, that these formal sources are filtered through the less formal notions of experience and successful operation, and by a belief in the likelihood of techniques and resources not to behave as scientifically expected. Self-disclosure can help to make a statement more convincing (Potter, 1996). Where Baym (2000) found that participants in a soap opera newsgroup used self-disclosure in creating an open and friendly atmosphere, here the disclosure of personal experience is more restricted, and is directed to bolstering the authority of particular statements. The atmosphere is not one of generalised friendliness, but of a very specific shared set of understandings.

4.20

In some cases, a conflict of authorities arises, where the suggestions made in multiple responses appear to be mutually exclusive. Such sequences are unusual, and should not be taken as characteristic of the group: the majority of messages receive either no public responses, or a very few. It is worth examining an unusually lengthy sequence of messages, however, to the extent that the extended discussion involves participants in elaborating on appropriate forms of reasoning in the group. One of the longest sequence of responses in the corpus of messages examined here occurred following the posting of a particularly exasperated query in message 0829:

I've been using [brand name] DNA preps for years. The quality is great: as everyone who's used them knows, though the yield is at best inconsistent. At random intervals, and for no apparent reason, there will be absolutely no DNA produced. I've been willing to deal with that when I was getting 5 out of 6, or whatever, with good yields; with important DNA, I'd always run duplicates to be reasonably sure of getting something.

In the last two days, I've run 10 [brand name] DNA preps - 8 midi-preps, two maxi-preps - and got one (1) with detectable DNA. Meanwhile two identical cultures, run on a different company's resin, gave 160 ug and 175 ug of DNA from 200 ml bugs (low copy-number plasmid.

One out of ten is not acceptable. Over the years I've gone around and around with [brand name] reps. I've tried dozens of variations on the protocol, I've tried everything they suggest, I follow the directions exactly, and in spite of it I just spent two days pouring DNA down the sink.

The hell with it. Until I hear from users that the [brand name] preps are reliable, I'm not going to use them any more.

Blowing off steam,

John

4.21

The poster sets himself up as having already tried everything possible, so rather than technical advice, he wishes to hear that the technique is working for others. This message generated a range of different responses. Among these was message 0842, which began as follows:

It your right to be pissed but it might be more productive to find a source of the problem in your protocol. Unless no one else uses your reagents I'd start with finding a dumbass with hands growing out his/her rear end that used the kit before - every place has at least one. Like last week suddenly [brand name] kit suddenly stopped sequencing while newly purchased fmol kept on rolling with the same template and primers. Until that is the same schmuck got his hands on it too. Now, if you can't find one.....

4.22

There then follows a series of technical factors which could affect the outcome of the technique, and a description of the good experience of the author of the response with the same technique. The message ends with the line:

Now you can resume being pissed.

4.23

The author of this response at first implies that the author of the query is acting inappropriately as a scientist, then suggests that he might instead examine specific local conditions, such as an inept person in the laboratory who may have contaminated the reagents. He then challenges the assertion in the query that every possible technical reason why the protocol might not be working has been tried and justifies his accusation with a performance of his own technical competence and experience.

4.24

Another type of response altogether was the "me too" response offered in message 0941, which appears much closer to the experience-based answers which the original query requests:

I have had low or no yields with [brand name] niniprep spin columns!

4.25

This message accepts the parameters set in the query, and offers a confirmation, based on experience, that the exasperation of the originator of the query is justified. No advice is offered, unlike the next "me too" in message 0995

Hi

Yep, me too.
We went back to the old [brand name] kit and then started to make our own version (from the original PNAS reference. We can't call it by the original name, but it is the same I reckon. Yields are consistently good, honest.
Full reference at [URL for World Wide Web site]
courtesy of Dr James Waters who has trouble shooted this. P.S. Use TBE in the gel for best results.
Good luck,
Alan

4.26

This response offers advice directed at bypassing whatever problems there might have been with the technique being used, by using another one altogether. No diagnosis of the reason for the original problem is offered, nor is there a rationale for why the alternative should work any better. The support for the solution offered is couched in terms of the experience of the author, rather than more conventional scientific criteria. Message 0846 also offered a change of tack to a different technique:

Switch to CsCl...it's really not as labor intensive as everyone thinks. Given your futile preps, it will no doubt save you time and money. And you will never have to worry about the quality of your DNA...CsCl banding is the gold standard, (ever notice what [brand name] compares their preps to?) Brad Staten

4.27

Here the alternative technique is justified in terms of its general acceptance (the "gold standard"), but again there is no scientific rationale offered for the comparison. Both message 0995 and message 0846 accepted the authenticity of the original query, on the basis of their own local experience. Message 0972 was a "me too" post which challenged the premise of the query, on the basis that the factors which had been taken into account were not the relevant ones to consider:

The [brand name] maxiprep is definitely inconsistent both with regard to yield and purity. I have the impression that some of the problems relate to the plasmid; certain plasmids (even high copy numbers) don't give high yields with the [brand name] maxiprep. I have also the impression that one should not use the filter tips but should stick to the centrifuge step. Two times now I had no yield using the filter tips, whereas the same prep was fine when using the centrifuge step instead. Lately I am having a lot of trouble with the purity; even multiple phenol/chloroform extractions does not increase purity.

Cheers, Robert

4.28

This series of responses therefore offered a range of different rationales for explaining the problem, and different ways in which the problem could be either solved or circumvented. Also at stake is the appropriate frame of relevance for defining the problem: some participants accept the framework set by the original query, while others dispute or extend it. Whilst all base the authority of their statements in experience, a variety of other sources of authority are brought in, including "what everyone knows about laboratory work", attribution of actions and motivations to the company producing the problematic kit, references to published literature and comparison with known standards. Some, but not all, of these allusions are deemed acceptable by the author of the original query, in one of the few instances where a reaction to advice is made public. The poster of the original query posted a capping response to message 0842, which had offered a range of extra factors to take into consideration. Message 0841 reads as follows:

Nope. I've gone through debugging of the [brand name] preps a dozen times; in no case have I ever found a clear source of a problem. All the points you mention I've ruled out already - in the past five years or so I've got scores of tips and suggestions and warning from the [brand name] tech representatives and other users, and I'm careful with my reagents and all that.

The bottom line is that if the procedure is so fragile that it can't tolerate different phases of the moon, I'm not interested in using it. It's not productive to spend my time fixing [brand name]'s problems. It's productive to get my DNA and do my experiments.

Thanks to everyone who's sent me suggestions, commiserations, and "Hey-that-happened-to-me-too!" comments.

John

4.29

Having excluded the fixes suggested on grounds of technical rationale, local circumstances (such as carelessness with reagents), and having asserted his own experience and technical competence, the poster of the original query resorts to an explanation which characterises the technique as defying rational explanation: "it can't tolerate different phases of the moon".

4.30

Taken together, this series of messages suggest the coexistence of conventional scientific explanations together with the more contingent explanations which draw on experience, location, the ineptitude of other researchers and sheer bad luck. The focus is on giving practical advice, although, as we saw in the capping response from the original poster in message 0841, moral support from validation of bad experiences can be valued too. It is apparent both from normal times and from this episode of controversy that participants draw on a variety of representations of laboratory work in discussing problems and rendering their contributions convincing. Rationales of the type characterised by universalist science (if it works here it should work anywhere) co-exist with references to the local contingencies and variations, and the characteristics of the individual researchers which may influence success. Reference is frequently made to factors which the contributor 'knows' but which are not available in published protocols. The discourse therefore pays tribute both to explicit, formal scientific communication and experience-based knowledge. The discourse in the protocols newsgroup has a hybrid form, somewhere between the formalised discourse of a scientific paper and the chat in the laboratory. The hybrid is not, however, entirely a comfortable compromise: in periods of controversy, such as that detailed above, the two perspectives can come into direct conflict. We cannot know from analysis of the newsgroup alone what the ultimate resolution of conflicts might be or what participants will, in the end, believe: the conventional structure of newsgroup discussions leaves agreement and consensus largely invisible. However, we can say that since participants continue to use the newsgroup to post their queries and their responses, they appear to find this form of discussion meaningful and valuable, and to that extent it is likely that at least some of the advice given is deemed plausible.

Conclusion

5.1

It has been argued elsewhere that it is not the particular technical nature of newsgroups themselves that promote or determine a specific style of interaction (Hine, 2000). Different newsgroups, even when discussing similar topics, can produce very different interactional styles. Each newsgroup is sustained as a distinctive space through the contributions of participants. The newsgroup is simultaneously performative space and performed space. The newsgroup provides the space within which participants aim to produce appropriate contributions. At the same time, through their contributions participants affirm and sustain the newsgroup as a particular kind of social space. In particular, there is often a specific concern to display 'situational relevance' (Hine, 2000: 109), that is, to demonstrate that a contribution is an appropriate one to this particular newsgroup.

5.2

The protocols newsgroup is sustained as a performative space for scientific discussion, within which a message will count as situationally relevant if it displays a knowledge of the appropriate scientific field, either by drawing on established sources of scientific authority, or by demonstrating an experience of scientific practice. Participants have to be able to display themselves as credible scientists through appropriate affiliations and identification, and also to "talk the talk" of laboratory science. In providing a hybrid form of communication, incorporating elements of both formal scientific communication and informal experience-based understanding, the protocols group could be seen to provide a means of over-coming geographic distance as the proponents of cyberscience suggest. It could also to a certain extent provide a bridge across the social distance between scientists working in different laboratories, although the emphasis on display of affiliations means that participants can still decode one anothers' place in the scientific community and react accordingly. The group does provides a conduit for informal information flows to accompany the formality of the published scientific papers, protocols and kits which link laboratories. For researchers who can display the right kinds of expertise, the protocols newsgroup does, then, have some capacity to overcome the boundaries of individual laboratories, and to assist in stabilising techniques across different laboratories (Jordan and Lynch, 1998). The newsgroup is sustained as an extension of laboratory space by the focus on experience as a form of authority.

5.3

Making oneself a credible scientists online involves a mixture of formal and informal means including: drawing on institutional affiliations and addresses; appropriately addressing a community; having access to formal resources for communicating science; and demonstrating awareness of the messiness and contingency of laboratory procedures using personal experience. These formal and informal ways of communicating contain elements of the empiricist and contingent repertoires identified in scientific discourse by Gilbert and Mulkay (1984). The focus of the protocols group on giving advice on problems may encourage the use of the contingent repertoire. Rather than the formal scientific communication outlets where reporting findings and procedures that work is the norm, the protocols newsgroup revolves around things not working. It is in not working that a technique becomes an appropriate topic for a query, and in finding reasons for things not having worked that the contingent repertoire often prevails. Some participants draw on the resources offered by their scientific credentials and institutional location to provide authority for their statements. Others adopt a far less formal style, and draw on the authority of experience and involvement in "hands on" scientific work to lend credibility to their observations. Scientists from laboratories across the world are able to access the group, and are provided with a forum to ask questions and debate issues which have no place in the formal scientific communication system. However, the overall focus on formal identification and credentials, through extensive use of apparently real names and institutional affiliations, helps to lend a credible air to even informally framed advice.

5.4

As science policy makers hope, it does appear that the protocols newsgroup and others like it could contribute to the efficiency of scientific research. Through becoming a space for the exchange of experience-based advice, the group is able to supplement the formal scientific communication process and provide for ways of working to be shared across widely geographically dispersed laboratories. It is not clear, however, that this represents "cyberscience", in the sense of a major transformation in scientific practice. Rather, it seems to be an extension of an existing focus on the value of experience as a complement to formal information, as described in Section 2. Where the spread of a new technique might previously have occurred slowly, via the movement of staff between institutions and through meetings at conferences and training courses, the protocols newsgroup enables a faster and more immediate exchange of experience, albeit one that has to "tell" rather than "show" experience. This is not, in itself, indicative of a whole new approach to scientific knowledge, although it may point to a weakening of the social boundaries that prevail between laboratories.

5.5

Thus far, the implications for scientific knowledge of the protocols group appear not to be very radical. As discussed earlier, however, the boundaries of the laboratory have been held in sociology of scientific knowledge to be crucial in protecting the special epistemic status of science. The boundary of the laboratory is also the boundary between science as a messy, contingent everyday practice and science as a universalised and privileged form of knowledge. If the protocols newsgroup weakens this boundary, might it not in a small way contribute to a weakening of science's special status? The protocols newsgroup does contain large elements of the side of scientific research not usually made publicly available. The conventional public notion of the laboratory does not include enzymes that behave as they wish or techniques that fail with "the phases of the moon".

5.6

Within some frameworks of understanding, the kind of public display of everyday reasoning found in the protocols newsgroup would be seen as a threat to the respect of non-scientists for scientific knowledge. Collins and Pinch (1998) suggest that public participation in scientific decision making might be promoted by giving people a less idealised version of science to work with: in effect by giving the public access to the contingent repertoires of science. Collins and Evans (2002) propose that incidents where the everyday nature of scientific work is exposed act to overcome the social distance between scientists and everyone else. They suggest that "Suddenly, the conclusions formally wrought by science alone are the property of everyone, and each has a right to contribute their opinion along with that of the no-longer-so-special scientists" (Collins and Evans, 2002: 248). Such suggestions are far from new: in 1988, Lynch (1988: 282) suggested that "If scientists wish to neutralise 'anti-scientific' public attitudes they might consider acknowledging more openly in their pedagogy and public presentations that 'objective' results are rooted in commonsense knowledge and subject to social contingency".

5.7

To suggest that a newsgroup could act as the kind of "breaching experiment" that Collins and Evans propose would be in line with much of the general thinking about the social consequences of ICTs. Meyrowitz (1985) suggests that new media for communication can disrupt the previously take for granted distinctions between frontstage and backstage behaviour (Goffman, 1971). He suggests that new electronic media can play a part in the reassessment of previously taken for granted social roles precisely because they allow for a disruption of the boundary between frontstage and backstage behaviour. Meyrowitz (1985) is therefore suggesting that electronic media might render previously secure social boundaries somewhat more "leaky", using the example of parents finding it harder to perform being "adult" in front of their children once those children have access to the backstage behaviour of parents via television shows. In a similar way, Collins and Evans (2002) could be taken as suggesting that exposure to the backstage behaviour of scientists could make the public less likely to accord their statements special status. For such an effect to occur, however, the everydayness of science has to be not only potentially available but also discursively available in the sense that people feel they are part of a discourse and thus able to contribute. In the protocols newsgroup, this would appear not to be the case.

5.8

The discursive availability of the protocols newsgroup is limited to practising scientists on a number of levels. Firstly, while the bionet World Wide Web site is in theory open to anyone to access, it does explicitly directs lay persons elsewhere for their discussions:

BIOSCI promotes communication between professionals in the biological sciences. All postings to the newsgroups should be made in that spirit. While the general public may "listen in" to the discussions, these newsgroups are intended primarily for communications between researchers. There are other forums on Usenet such as sci.bio.misc for the asking and answering of biological questions from lay persons. (<http://www.bio.net>)

5.9

This statement may also serve to remind the participants in bionet newsgroups that they may not be backstage. A second level of exclusion depends on the language used in the group. Without a background in the specific techniques concerned one can still detect the use of everyday reasoning in discussions, as this paper has demonstrated. One cannot, however, necessarily comprehend the significance of problems discussed for scientific outcomes that might have some meaning in the everyday world. The specialised language is partly responsible for this, so too is the focus on the detail of technique without the sharing of the overall goals of the scientific projects in which they occur. It is possible that participants deliberately refrain from discussing their overall projects in order to maintain competitive advantage: in this regard the group certainly does not break down the social boundaries between laboratories. In Lewenstein's (1995) terms, the focus of the group on "little ideas" may limit both the radical consequences of the group for scientific knowledge formation and the implications for the breaching of the science/non-science boundary. Finally, the emphasis on credibility both through formal location and through discourses of experience would preclude any meaningful intervention by non-practising scientists. The entire discursive framework of the group is based upon active participation in the scientific process. While the accessibility of informal laboratory talk in online forums might make the laboratory seem to be less of a private bounded space, this does not mean that social (as opposed to merely perceptual) boundaries therefore disappear automatically.

5.10

The protocols group, therefore, while to some extent breaking down social boundaries between geographically separated laboratories, does little in any practical sense to breach the boundary between science and the public. This reinforces a view that ICTs on their own do nothing to challenge social boundaries: rather it is the social and discursive practices within which they are made meaningful that can in some circumstances cause boundaries to be redrawn. In this case, the discourse of the newsgroup keeps the scientific boundary largely intact, but reinstates it in practical engagement with the work of science practice rather than in the presentation of a mystified and depersonalised scientific knowledge. While the raw material might be here for an altered public perception of science, the postings on the newsgroup are devoid of the mediating representations with Gieryn (1999) argues are so crucial for the formation of science as a cultural object. Readers without scientific backgrounds may lack the necessary narratives to make sense of what they see: one might view the postings differently were one to know that the scientists involved were carrying out a paternity test in which one had a stake, or searching for the gene for a disease from which one suffered.

Notes

¹.In an original version of the paper the name of the newsgroup was altered, as were the names of the contributors to discussions quoted in the analysis and any brand names or institutional references. An anonymous referee subsequently pointed out that, since newsgroup discussions are routinely archived, it would be very easy for a sufficiently astute reader to locate the original data source, and that he or she had in fact done so. It was suggested that the mere illusion of protecting informant confidentiality had been presented, and that the decision to change names should be reconsidered. One might, it was suggested, simply publish a link to the archive and offer readers the possibility to conduct their own analysis. Considering my response to this suggestion has led to considerable reflection on the ethics of analysing public online interactions, and since these issues are far from resolved as yet, I present my response in this lengthy footnote.

There appears to be a clear tension between the allure of a "democratic" analysis which would allow readers to formulate their own understandings of the raw data, and an ethical commitment to protecting the subjects of study from unnecessary exposure. The words of informants may have been presented in a public forum, but it is unlikely that they expected their words to subsequently become fodder for a sociological analysis of scientific discourse online such as presented here. On the other hand, the talk presented here is, I would suggest, not in itself terribly shocking, and its (more?) public exposure in this paper is unlikely to cause particular distress or harm to those concerned. But should I then promote even more public exposure, by directing readers to the original source to examine the phenomenon for themselves?

This discussion is somewhat ironic, since the paper focuses precisely on the implications of the "publicness" of online talk about laboratory practices. My eventual conclusion suggests that no more harm is likely to accrue to participants on the basis of this analysis than from the initial fact of the publicness of their discourse. In this sense, the decision not to consult participants on the use of their data is situated within the context of the analysis developed in the paper, where I conclude that particular discursive practices, rather than the mere fact of publicness, are likely to keep this form of public science talk from threatening the special status of science. My analysis is unlikely to matter sufficiently to the participants in the discussion. The ethical decision taken here is, then, wholly contextual: I have suggested elsewhere (Hine, 2000) that this may be a more useful and sensitive means of making ethical decisions on the use of online data than the application of absolute rules. Finally, many ethical discussions also revolve around the reasonable expectations of speakers as to the uses to which their words might be put. In this case, I can reasonably expect that anyone who reads thus far in the current paper is sociologically inclined, and hence alive to the ethical issues which surround the sociological analysis of public discourse. Consequently, I have decided to leave the question unresolved and abdicate some ethical responsibility to readers: I may or may not have altered identifying details of individuals and the group in this version of the article, and it is for readers to decide for themselves whether to test which I have done so and to consider whether this would make a difference to their reading of the analysis. This decision is made as a contribution to ongoing discussions about the appropriate ethical stance for Internet researchers, within the context of an over-arching belief that nothing in the analysis is likely to be of harm to the participants. Were the subject of discussion to be more personally sensitive to the participants, or my form of analysis more directed at understanding individuals rather than broad patterns of discourse, I might have taken the decision to discuss my analysis more fully with participants, or to withhold direct quotations from the analysis.

References

BAYM, N. (2000) Tune In, Log On: Soaps, Fandom and Online Community. Thousand Oaks, CA: Sage.

BEAULIEU, A. (2001) 'Voxels in the Brain: Neuroscience, Informatics and Changing Notions of Objectivity', Social Studies of Science Vol. 31, No. 5, pp. 635-680.

CAMBROSIO, A. and KEATING, P. (1988) 'Going Monoclonal: Art, Science and Magic in the Day-to-Day use of Hybridoma Technology', Social Problems Vol. 35, No. 3 pp. 244- 260.

CASTELLS, M. (2001) The Internet Galaxy: Reflections on the Internet, Business and Society. Oxford: Oxford University Press.

COLLINS, H.M. (1975) 'The Seven Sexes: a Study in the Sociology of a Phenomenon, or the Replication of Experiments in Physics', Sociology Vol. 9, No. 2. pp. 205-224.,p> COLLINS, H.M. (1985) Changing Order: Replication and Induction in Scientific Practice. London: Sage. COLLINS, H.M. and PINCH, T. (1998) The Golem: What you Should Know about Science, Second edition, Cambridge: Cambridge University Press.

COLLINS, H.M. and EVANS, R. (2002) 'The Third Wave of Science Studies', Social Studies of Science Vol. 32, No. 2 pp. 235-296.

DAVID, M. and ZEITLYN, D. (1996) 'What are they Doing? Dilemmas in Analyzing Bibliographic Searching: Cultural and Technical Networks in Academic Life' Sociological Research Online, Vol. 1, No. 4, <http://www.socresonline.org.uk/1/ 4/2.html>

ETAN Expert Working Group (1999) Transforming European Science through Information and Communication Technologies: Challenges and Opportunities of the Digital Age. Prepared for the European Commission. Directorate General for Research, Directorate AP-RTD Actions: Policy Co- ordination and Strategy.

FUJIMURA, J. (1996) Crafting Science: a Sociohistory of the Quest for the Genetics of Cancer. Cambridge, MA: Harvard University Press.

GALISON, P. (1999) 'Buildings and the Subject of Science', in P. Galison and E. Thompson (editors) The Architecture of Science. Cambridge, MA: MIT Press. pp. 1-15.

GIERYN, T. (1999) Cultural Boundaries of Science: Credibility on the Line. Chicago: University of Chicago Press.

GILBERT, G.N. and MULKAY, M. (1980) 'Contexts of Scientific Discourse: Social Accounting in Experimental Papers', in K.D. Knorr, R. Krohn, and R. Whitley (editors) The Social Process of Scientific Investigation. Dordrecht: Reidel, Sociology of the Sciences Yearbook Vol. 4, pp. 269-284.

GILBERT, G.N. and MULKAY, M. (1984) Opening Pandora's Box: a Sociological Analysis of Scientists' Discourse. Cambridge: Cambridge University Press.

GOFFMAN, E. (1971) The Presentation of Self in Everyday Life. (1956) Harmondsworth: Penguin.

HERT, P. (1997) 'The Dynamics of On-Line Interaction in a Scholarly Debate', The Information Society Vol. 13, No. 4, pp. 329-360.

HINE, C. (2000) Virtual Ethnography. London: Sage.

JORDAN, K. and LYNCH, M. (1998) 'The Dissemination, Standardisation and Routinisation of a Molecular Biological Technique', Social Studies of Science Vol. 28, Nos. 5/6, pp. 773-800.

KLING, R. and McKIM, G. (2000) 'Not Just a Matter of Time: Field Differences in the Shaping of Electronic Media in Supporting Scientific Communication', Journal of the American Society for Information Science Vol. 51, No. 14, <http://www.slis.indiana.edu/kling/publications.html >

KNORR-CETINA, K. (1981) The Manufacture of Knowledge: an Essay on the Constructivist and Contextual Nature of Science. Oxford: Pergamon Press.

KNORR-CETINA, K. (1999) Epistemic Cultures: how the Sciences make Knowledge. Cambridge, MA: Harvard University Press.

LATOUR, B. and WOOLGAR, S. (1986) Laboratory Life: the Construction of Scientific Facts. Princeton: Princeton University Press.

LEWENSTEIN, B.V. (1995) 'Do Public Electronic Bulletin Boards Help Create Scientific Knowledge? The Cold Fusion Case', Science, Technology and Human Values Vol. 20, No. 2, pp. 123-149.

LIVINGSTONE, D.N. (1995) 'The Spaces of Knowledge: Contributions towards a Historical Geography of Science', Environment and Planning D: Society and Space. Vol. 13, No. 1, pp. 5-34.

LYNCH, M. (1985) Art and Artifact in Laboratory Science: a Study of Shop Work and Shop Talk in a Research Laboratory. London: Routledge and Kegan Paul.

LYNCH, M. (1988) 'Sacrifice and the Transformation of the Animal Body into a Scientific Object: Laboratory Culture and Ritual Practice in the Neurosciences', Social Studies of Science Vol. 18, No. 2 pp. 265-289.

MATZAT, U. (1998) 'Informal Academic Communication and Scientific Usage of Internet Discussion Groups', paper given at IRISS '98 International Conference, 25-27 March 1998, Bristol, UK. <http://www.sosig.ac.uk/iriss/papers/paper19.htm>.

MAXWELL, R. (1990) Information Technology as a Way of Reducing the Costs and Time in the Dissemination of Scientific and Technological Information. Fourth Annual Dainton lecture. Boston Spa: British Library.

MEYROWITZ, J. (1985) No Sense of Place: the Impact of Electronic Media on Social Behaviour. New York: Oxford University Press.

NENTWICH, M. (1999) Cyberscience: The Future of Research in the Age of Information and Communication Technologies (1998-2002) <http://www.oeaw.ac.at/ita/cyberscience.htm>.

OECD (2000) The Global Research Village Conference 2000: Access to Publicly Financed Research Organised jointly by the Netherlands Ministry of Education, Culture and Science, Directorate Research and Science Policy and the Organisation for Economic Co-operation and Development (OECD), Directorate for Science, Technology and Industry (DSTI), Committee for Scientific and Technological Policy (CSTP) OPHIR, A. and SHAPIN, S. (1991) 'The Place of Knowledge: a Methodological Survey', Science in Context, Vol. 4, No. 1 pp. 3-21.

POTTER, J. (1996) Representing Reality. London: Sage.

SHAPIN, S. (1995) 'Here and Everywhere: Sociology of Scientific Knowledge', Annual Review of Sociology Vol. 21, pp. 289-321.

SHAPIN, S. (1998) 'Placing the View from Nowhere: Historical and Sociological Problems in the Location of Science', Transactions of the Institute of British Geographers, Vol. 23, No. 1, pp. 5-12.

STAR, S.L. and GRIESEMER, J.R. (1989) 'Institutional Ecology, Translations and Boundary Objects: Amateurs and Professionals in Berkeley's Museum of Vertebrate Zoology, 1907-39', Social Studies of Science Vol. 19, pp. 387-320.

TURKLE, S. (1995) Life on the Screen: Identity in the Age of the Internet. New York: Simon and Schuster.

WOOLGAR, S. editor. (2002) Virtual Society? Technology, Cyberbole, Reality. Oxford: Oxford University Press.

WOUTERS, P. (2000) 'The impact of the Internet on the sciences'. Paper presented at 4S/EASST conference Worlds in Transition: Technoscience, Citizenship and Culture in the 21st Century. September 27-39, 2000, University of Vienna, Austria.

ZEITLYN, D, DAVID, M. and BEX, J. 1999 Knowledge Lost in Information: Patterns of Use and Non-Use of Networked Bibliographic Resources. British Library Research and Innovation Centre Research Report no. RIC/G/313.

ZENZEN, M. and RESTIVO, S. (1982) 'The Mysterious Morphology of Immiscible Liquids: A Study of Scientific Practice' Social Science Information Vol. 21 No. 3, pp. 447-473.

Copyright Sociological Research Online, 2002