The peripherals at the core of androcentric knowledge production: an analysis of the managing editor’s knowledge work in The International Encyclopedia of Education (1985)

ABSTRACT Building on the approach that knowledge is socially constructed, this study aims to deepen the understanding of knowledge production processes by adapting the concept of a laboratory on an historical example of editorial collaboration. We use the editorial process of The International Encyclopedia of Education (IEE) (1985) as an example. Considering knowledge as situated in contrast to the illusion of objectivity, the paper pursues identifying a perspective on the editing process that is located in its periphery. Rooted in constructivist assumptions of Science and Technology Studies (STS) andFeminist Standpoint Theory (FST), the paper builds on the potential of understanding the social construction of knowledge by taking marginalised perspectives into account. Including the concept of gendered division of labour, the approach offers a deeper understanding of the editorial process, beyond the presented picture of relevant actors and actions. The source material consists of correspondence concerning the editorial process of the encyclopaedia, which is filed in the archive of Torsten Husén. There, we encountered the publisher’s managing editor Barbara Barrett, the only woman in the editorial board of the IEE’s first edition. By analysing almost 4000 pages of correspondence by coding according to qualitative content analysis it is possible to show that the kind of work that Barrett performed was much more central than presented. This study contributes to a deeper understanding of knowledge production processes and of related marginalisation of certain groups in the past. Hereby, it offers a departure point for reflecting upon present practices.


Introduction
Recent decades have . . . witnessed increased global interdependence which has prompted education to become far more internationalized. The exchange of research results and experiences among scholars and practitioners has become much more frequent, making comparative education an important field of scholarship in education. In view of these conditions, it is convenient for educators, social scientists, and the inquiring public to find, in one set of volumes, summaries of the state of all major aspects of education, gathered from worldwide sources. 1 Building on the approach that knowledge is socially constructed, this article aims to deepen the understanding of knowledge production processes. We use the editorial process of The International Encyclopaedia of Education (IEE) (1985) as the object of study. The IEE -first edited 1985 by Torsten Husén and T. Neville Postlethwaite and published by Pergamon Press had as its ambition to be the first true international education encyclopaedia. 2 This meant breaking with the ethnocentrism the editors identified in earlier encyclopaedias of educationespecially those produced by the American Educational Research Association (AERA) in 1969 as well as reaching out to educational systems in, as they put it, "the Third World". 3 As shown in the quote, the editors aimed at offering the public an appropriate worldwide status of education and, in doing so, enhancing the recognition of comparative education, as well as contribute to the field by collecting and publishing the status quo. Torsten Husén 4 , who called himself an educational generalist, 5 and T. Neville Postlethwaite 6 (1933-2009) considered themselves and their network of actors able to give that comprehensive overview. They worked closely with a larger editorial board and an honorary editorial advisory board. 7 The IEE was an extensive project and contracted authors from 100 different countries, resulting, after five years of work, in 10 volumes of the first edition, with an index of more than 45,000 entries. 8 The extent of this project strengthens the argument to analyse the editorial process of an encyclopaedia as an example of collaborative academic work in education and therefore as an example of joint knowledge production processes in the field.
The editors' claim to offer "summaries of the state of all major aspects of education, gathered from worldwide sources" 9 creates the impression of a collection of given, unbiased, and unlocated facts. 10 From a constructivist perspective, producing an encyclopaedia is not an independent and objective editorial process. It is culturally, socially, and institutionally embedded. By choosing topics and authors and by structuring the compilation along sections, the encyclopaedia is not only formally shaped; doing so influences, in various ways, the social process and practices of the encyclopaedia's formation and therefore the social construction of knowledge. The editorial process underlies constant changes and is very much influenced by its editors' perspective. To maintain the illusion of objectivity, facts are presented as given and unbiased, a "god trick" as Haraway 11 put it. In following notes we simply use the short SE/RA/to refer to this archive. 6  Routed in constructivist assumptions of Science and Technology Studies (STS) and in Feminist Standpoint Theory (FST), 12 the paper builds on the potential of understanding the social construction of knowledge by taking marginalised perspectives into account. Considering knowledge as situated in contrast to the illusion of objectivity 13 the paper pursues identifying a perspective on the editing process that is located in the periphery of the IEE production process. This approach offers a deeper understanding of the editorial process, beyond the presented picture of relevant actors and their responsibilities and influences. 14 Whether the editors for example could meet their aim to avoid ethnocentrism is questionable not only considering the fact that most of the 15 experts in the editorial board next to Husén and Postlethwaite were employed in the United States, some in the UK, and only two in non-Anglo-Saxon countries. 15 Furthermore, reviewers of the IEE's first and second edition (1994) criticise the entries for being biased towards the US and UK. 16 Apart from an ethnocentric perspective, also the dominance of an androcentric perspective in the knowledge represented in and produced for the IEE attracts attention. It happens for example that the experts in the editorial board were only male scholars from Western Europe and the US. 17 Furthermore, the editors received during the process the critique of a "lack of female representation". 18

Points of departure
There already exist several guidelines on how to edit encyclopaedias, 19 but it is difficult to find less normative and more analytical literature or research on editing encyclopaedias. Most of what can be found instead consists of biographical notes and reflections. 20 Investigating how an encyclopaedia comes into being aligns with the sociology of knowledge and the history of knowledge. Therefore, knowledge is regarded as produced under historical circumstances and reflecting society as well as some of its needs. 21 Hence, not only "truth criteria" and the pure preservation of knowledge can be expected as reasons for producing an encyclopaedia. Encyclopaedias represent knowledge, which is geographically, socially, and chronologically embedded. 22 In the following paragraph, we introduce the paper's theoretical framework. First, we summarise general constructivist assumptions and a concept of STS as it builds the foundation of the study. Furthermore, we draw on a rather marginalised perspective on knowledge production for its potential and legitimacy in FST. By that, the study ties in with findings of research on gendered division of labour in academia in the past and today.

Science and Technology Studies (STS)
In the influential book Laboratory Life: The Construction of Scientific Facts, 23 Bruno Latour and Steve Woolgar describe the production of scientific facts as a complex fabric of everyday hard work in the laboratory. Reference cultures, scientific prestige, curriculum vitae, publications, group dynamics, and many other aspects influence it. Research is to a large extent a socio-material activity.
If we look at the editorial work of IEE as a kind of laboratory, we have to consider it, given its international character, as a global network that contributes to the development of social science research beyond fixed time and space dimensions. That extends science beyond individual ideas, researchers, environments, institutions, universities, and even national borders. These are practices, which create, mobilise, sustain, and challenge relations between actors in knowledge creation and various social activities. 24 Using IEE as an historic empirical example contributes to the understanding of the everyday practices through which, eventually, ideas and fact-making are established and stabilised in the field of comparative education.
The origins of the powerful contemporary international research organisations, such as for example the IEA, 25 in education, and their increasing influence in the shaping of national and global education governance lie in the construction of new research networks, associations, and publications in postwar Europe. 26 Approaches contributing to understanding the history of educational research since then emphasise for example the role of research institutes 27 and international conferences 28 or international networks in general. 29 Access to the unique archive of Professor Husén gives the opportunity to reconstruct circumstances and the production process of the IEE as an example of a publication in and cause for international networks and associations in the second half of the twentieth century. The archive somewhat documents a social science laboratory and therefore makes it possible to explore knowledge production in retrospect. The digitalisation of correspondence changes the way of collecting and storing written material of knowledge production processes of the present and therefore the access to it for research.

Feminist Standpoint Theory (FST)
Drawing on FST, 30 this paper follows a peripheral, less visible, and somewhat less powerful position to support a more in-depth understanding of the process that goes beyond the presented picture of responsibilities and influences (presented, for example, in the preface of IEE). However, the danger of this approach of focusing on a marginalised perspective is to reproduce unlocated claims of objectivity. According to Haraway, "'Subjugated' standpoints are preferred because they seem to promise more adequate, sustained, objective, transforming accounts of the world". 31 The innocent appearance of subjugated positions may again lead to inappropriate "claims of scientific authority", 32 which the project tries to avoid by offering transparency and constant re-examination along the analysing process. Analysing and arguing from a marginalised standpoint constantly requires reflection. This paper tries to reveal the illusion of self-evident and comprehensive knowledge claims without romanticising the perspective of subjugated positions and without simply condemning the work of powerful scholars. Hence, this work offers a certain viewpoint on the IEE editing process as an example that only represents one way of deepening the understanding of social knowledge production. Furthermore, considering the changes and discussion of approaches within standpoint theory, 33 also in alignment with the waves of feminism, it is important to reflect upon the chosen category for identifying peripheral perspectives. To focus on the dimension of gender and in this study on women's perspectives in particular is based on the material itself. As mentioned in the introduction, the archive material shows that the editorial board was criticised for a "lack of female representation". 34 Even if in general the reduction of the marginalised group as women risks overlooking intersectional dimensions of discrimination and processes of reproducing differences, we draw on the potential of strategic essentialism, 35 which builds on "certain essential features . . . to give power back to historically marginalized and disadvantaged groups". 36 Even if the focus comes from the material, it is important to emphasise here that this study only covers one possible dimension of discriminative practices in academia; further reflections and studies may consider continuing the discussion on dimensions like class, race, ethnicity, sexual orientation, and more, and how they intersect.
Apart from that, the introduced project attempts to make the analysing process as far as possible transparent by describing the theoretical sampling as well as the coding process to avoid the illusion of self-evident truth.

Gendered division of labour in academia
A gendered division of labour is proven to have a long tradition in academia. 37 Considering its stability over time and the impact research has on public opinion and social practices, it remains important to investigate the phenomena historically and currently. This applies especially when it comes to knowledge production per se. 38 In contemporary research about gendered division of labour in academia, various ways are revealed on how the division is institutionally and socially reproduced, and performed. Women predominantly fulfil services like administrative tasks, care tasks, and even teaching, to the benefit of men in the system. This phenomena is recently also summarised as "academic housekeeping". 39 These tasks limit academic careers, even though they are necessary for the system to function. 40 This article shows with its historic case how a woman beyond care work and administration performs knowledge work without being portrayed as an expert with equal merits.

Material and the coding process
Husén's archive offers 3852 pages of correspondence concerning the editorial processes of the first two IEE editions, between himself and Postlethwaite, as well as the publisher's staff like the managing editor, section editors, and commissioned authors. 41 Thus, it allows detailed insights into several perspectives and practices of actors involved in the social knowledge production processes of both IEE editions. 42 There, we encountered the IEE publisher's managing editor Barbara Barrett (b. 1945). She was the only woman on the editorial board of the IEE's first edition. Barrett in many ways was especially involved in the first edition's (1984) formation process and thus contributed to the IEE's knowledge production. Even if she is mentioned in the preface, by not being part of the list of references, which support knowledge claims in academia, 43 she seems to be publicly excluded from the group, which is framed as expert. The potential of deepening the understanding of the IEE knowledge production process by following her is justified through her alleged peripheral position. Moreover, it is even strengthened because by analysing almost 4000 pages of correspondence by coding according to the qualitative content analysis 44 it is possible to show that the kind of work that Barrett performed was much more central than presented.
To approach the described aim, the analysis was divided into several steps, which allowed an appropriate theoretical sampling for the analysis. At first, an overview was generated by looking into every single page of the archive's material related to correspondence about the first two editions of the IEE (3852 pages) and into the preface of the encyclopaedia. The material was coded by roughly summarising the subject matters of the respective pages with a rather general interest in the complexity of the knowledge production process and possible peripheral perspectives within it. In the material we encounter external critique to the editorial board about lacking "female representation" 45 in the encyclopaedia. The critique combined with the fact of only male scholars on the editorial board, motivates the study to focus on the gender dimension when identifying perspectives in the periphery of the editorial process.We encounter Barrett as alleged representative of this periphery. Especially in the editing process of the first edition, she emerges as a more central figure than the preface implies.. 46 Hence, we follow her trace to investigate how she was involved. To have a lasting reasonable understanding and documentation of the content, striking details of the coded paragraphs were recorded and first ideas about possible main categories that structure Barrett's case evolved. Following the overview, special attention was given to relevant material involving Barrett. As the publisher's managing director, not only does she generally appear very often in correspondence, her letters are also archived in a separate file of 447 pages. These were included completely 47 in the analysis, which was at first 41 In the archive, there is also one volume, file 194, containing documents from the initial editorial meetings, of approximately 300 pages. 42 The productions of both editions are not clearly distinctive. Close to the launch of the first IEE, the editors agreed on supplement editions, which were supposed to update the first parent book. oriented on rather abstract main categories that divide her modes of involvement into administrative work, care work, and content work. Whenever Barrett was contacted or involved in organising authors or section editors in various administrative matters (e.g. contracts, transmitting information about IEE, expenses) or received copies of letters to be informed about progress status or how things need to be handled, her role was classified as mainly administrative (administrative work). Paragraphs coded as care work concerned different incidents of personal exchange, greetings, and additional support, which showed that the correspondence went beyond strict administrative or content-related matters. Whenever Barrett is directly contacted about or involved in discussing the IEE content (i.e. discussing entry or section), is giving feedback on entries, or is content-wise involved in a discussion that also includes correspondence, where the author or editor explains decisions with content-related arguments specifically to her, the paragraph is classified as content work.
During the coding process, it became clear that these categories overlap and in at least two ways cannot be understood as distinctive forms of involvement. On the one hand, certain letters deal with matters that can be assigned to one of the categories, but their response(s) can extend or turn into different subject matters. If the initial sender, for example, addresses administrative questions, the second person can extend the matter to further issues or turn it into a different topic. On the other hand, it became clear that the categories are interdependent. Administrative work and care work influence content work and vice versa in various ways significantly, which will be further elaborated in the following paragraph concerning the findings.
Furthermore, Barrett's position in the publisher's hierarchy changes over time. While she rises career-wise, her involvement in the IEE decreases from the first to the second edition. In the second edition, letters from her or addressed to her do not occur as many times as they do in the first edition. Of the 447 pages of letters in the archive's file that are labelled with her name, 385 are dated between 1980 and 1985, whereas only 61 pages are dated after 1985 and therefore after the publication of the first edition. 48 Thus, the final detailed coding and analysis focuses on the material considering the first edition.
Through theoretical sampling of relevant paragraphs and an inductive analysis, an indepth, subject-related interpretation was possible. The coding frame was constructed, adapted, and detailed along the coding process, which also means that not all categories existed from the beginning, even if relevant paragraphs had maybe already occurred. Therefore, quantities of coded references need to be interpreted with caution and play a minor role in this analysis.

Empirical findings
The International Encyclopedia of Education, as an idea by Robert Maxwell , the publisher at Pergamon Press, starts to materialise in 1979 when he approaches Husén, Professor at Stockholm University, to edit a "work that would draw together information on internationally significant, recent work in education". 49 The material shows that they start the project by commissioning suitable partners or section editors in 1980. The publisher and Husén invite "international specialists" to a first meeting in June in Oxford for in-depth exploring of the project's editorial feasibility and organisation. At the meeting, the participants lay the groundwork for the vast project, such as defining it as an encyclopaedia. 50

Barrett's peripheral perspective
It happens that those experts invited to make that decision were only male scholars from Western Europe and the US. 51 Barrett first appears in the material in the beginning of 1981 regarding rather in-depth planning of the editing process, which implies that she is from the beginning quite a fundamental part of the project. 52 Husén emphasises in a letter to her (in March 1981) that Maxwell should ensure any assistance necessary to her in order to manage the project. 53 Furthermore, her extensive involvement is shown, as mentioned, by numbers in the archive material. She is the publisher's main contact, engaged in editing, and therefore plays a vital part in the process of the first edition. Particularly the editors-in-chief's acknowledgement of her in the preface prove that: We would like to express our deep gratitude to Pergamon Press for assigning such an excellent editorial unit to the task of guiding the Encyclopedia through the steps of production. Barbara Barrett, the Managing Editor, combined personal sensitivity with a high level of professional skill in facilitating the publication process at all its stages. 54 By being described as "facilitating the publication process", Barrett's contribution is mainly framed as editorial support rather than as editorial responsibility. Additionally remarkable is that they emphasise her "soft skills", which are traditionally constructed as feminine. 55 That already implies a dualistic and gendered division of relevant labour for the editing process: on the one hand, content-related responsibility by man experts of the field, and on the other hand care work and administrative editorial support, performed by a woman, Barbara Barrett. This relates to the general unequal involvement of women and men in the IEE, which is not only observable but also directly addressed within the archive material.
As mentioned at the beginning, Barrett as the managing editor receives critiques about the "lack of female representation" 56  the board of the second edition are women. They are responsible for the topics "Education of Children with Special Needs", "Preschool Education", and "Girls and Women in Education". The latter section is added to the new edition after revising the first edition, in which they handled the critique by inviting women experts to the honorary editorial advisory board. 57 Adding women experts to the advisory board can be understood in at least two directions. On the one hand, the critique is absorbed and processed; the invited woman expert can influence the subsequent editing process. On the other hand, the inclusion of only women's advice and not women's expertise in matters of content-related responsibility is a rather superficial solution and framing strategy. The actors may have prioritised presenting a picture of the involvement of women's perspectives to legitimise the IEE in the field. Furthermore, considering the already ongoing production, including a woman in the honorary editorial advisory board instead of changing section editors or adding new sections is a practical approach with low costs as the board of section editors is already compiled. This instance allows an insight in the importance of framing the IEE according to presumed culturally accepted quality norms.
Considering the predominant male composition of the editorial board and the critique of it within the material, Barrett's alleged peripheral perspective on the editing process as the only woman member in the first edition becomes particularly convincing. Considering Barrett as a tracer offers great potential for a broader understanding of the knowledge production process in the wake of the IEE.

Barrett's modes of involvement: main findings of the analysis
At first glance, Barrett's involvement in the IEE's editorial process seems to be limited to administrative matters and editorial support, as you might expect from the idiosyncrasy of the position of managing editor, whereas Postlethwaite and Husén are considered professional experts. By analysing the archive material, her complex involvement not just as a support, but rather as an actor involved in the IEE's content production becomes apparent. Her indirect and direct impact on content-related matters is strongly visible. Anyway, acknowledging the messiness of practices, her administrative work, content work, and care work cannot be considered as being clearly distinctive, but rather as being interdependent or related, which will be explained further on.
At the centre of Barrett's responsibilities is the overview. It seemed to be vital having her as a central node in the production's network. This concerned her being informed about the status and progress of the project on several levels. Her obviously administratively relevant tasks can be summed up to her central position of having the overview. She was responsible for keeping track of the process, informing all actors, and controlling the progress along the time schedule. By that she was able to have the overview and in turn having the overview enabled her to perform these relevant tasks. This indirectly mattered for the concrete content production and required her soft skills (care work), which are even emphasised by the editors-in-chief. In the following paragraphs, Barrett's administrative work, care work, and finally content work as well as their interrelation will 57 Progress Report by Husén, Postlethwaite and Barrett to editorial board, June 12, 1981, SE/RA/720869/Vol. 2:276. be described and explained in more detail by giving also verbatim examples of the vast empirical data corpus.
Editorial support At first glance, Barrett's task of having the overview seems to match with the idea of managing and administrative duties of her position as managing editor and also with a traditionally gendered division of labour. She keeps track of the agreed time schedule and the involved section editors and authors. Therefore, the section editors are addressed with progress reports, which include overviews of the author and entry commission. They are explicitly required to keep Barrett informed. 58 Barrett was not only informed about the status of commissioning authors or of their success to hand in the entries in time, but also was in charge of sending out the official invitations and contracts as well as chasing the authors or section editors if delay appeared.
As many of you have recently raised various queries relating to the proofs we felt it would be a good time to bring you up to date on a number of matters . . . As you know we are working to a very tight schedule to try and make up some of the lost time. Therefore could you please encourage authors to return their proofs to us within 14 days as specified in the contract. 59 She was responsible for the overview and was not just supposed to handle matters of progress but also to control them. This went beyond apparent administrative matters like contracts, relevant payments, or securing the time schedule.
Here, the interdependence of the editorial support and direct work on content as well as Barrett's central role become clear. She keeps track, informs the actors, and provides the conditions for quality control. Even if she might not be involved in discussing the quality of specific content or structure of the whole bookwhat one could certainly conclude from the pronoun "we" in the quote aboveshe at least provides the conditions and therefore influences the process.

Connecting and mediating
Thus, Barrett, in the centre of production, not only was in charge of the progress overview, but also functioned as a central mediator. 61 She was informed and, in turn, informed the actors about incoming and outgoing entries by, for example, sending out progress reports and guidelines, which she prepared together with Husén and Postlethwaite. This meant that her communication with all actors involved played a vital part in the production process. Apart from mentioning her several times in information letters, and highlighting her position and duties in the process to the section editors, the section editors were specifically required to keep her, next to Postlethwaite, informed about progress.  61 One could argue she functioned as a pure messenger. We instead make an argument that Barrett's involvement clearly is not limited to messaging.
We would very much appreciate it if Section Editors would keep Barbara and Neville informed (say once a month) of their progress. 62 This relates to her most important responsibility: to have the process overview and control. She needs to be informed about progress to fulfil this position. Therefore, next to Husén and Postlethwaite, she was emphasised as the most important contact to the section editors.
Please be sure to be in contact with Barbara, and do not hesitate to write to her about help you might need. 63 Here, Barrett's unique position is stressed. Even if, as the material shows, the editors-inchief are also responsible for support, Barrett is positioned in the centre of the process by being the main contact and support. Moreover, she is presented as a gatekeeper who decides and organises meetings for possible exchange between the actors. 64 As the material in the archive shows, the editing process of the first edition leads very fast into editing process of first two supplements and finally the second edition. 65 Considering that the editing processes of the first and second edition nearly overlap, it is worth it to take correspondence into account, which occurs after the first edition is published. In informational letters addressed to the section editors about for example the second edition, the "trio" of Husén, Postlethwaite, and Barrett is presented as a team equally in charge of the editing process: We three met on 4 June 1991 and reviewed queries sent in by Section Editors. We also reviewed the "final" lists which have come in. In several cases, we wrote to individual Section Editors, and, in other cases, to groups of Section Editors. 66 The framing of the three editors as a team taking decisions together does not of course necessarily mean that all three share responsibilities on every level. For the central point here, it is not even necessary to assume that. The remarkable aspect rather is that this framing of the editorial team in the end deviates from the picture portrayed in the publications by naming only Husén and Postlethwaite as editors-in-chief on the cover and acknowledging Barrett only as a "peripheral" support instead of an equivalent member on the editorial board. 67

Soft skills
In favour of keeping track of progress and having the overview, Barrett was not only informed about conditions on a technical level, she was also approached about personal circumstances like illnesses which influenced the editing process.
My virus troubles have not yielded and I am almost in the same shape as when we met in London, something that has affected my capacity to get on with my work. Hope that there will be an improvement soon. 68 Therefore, Barrett was also on a personal level involved in the process. These unpredictable circumstances influenced required flexibility in the time schedule and consequently had to be communicated to Barrett. It also required sensitivity in communication, which she, according to the paragraphs in the preface, successfully performed. Herewith, another main field of her involvement shows, which is categorised as care work. By administratively supporting the section editors or editors-in-chief and being informed about personal circumstances, the personal level is indicated. It further shows in several occasions when Barrett is required to mediate between actors with tensions or handle cases with editors where the section progress seemed to be failing. Therefore, Barrett, for example, protocols a meeting with Husén and Postlethwaite and its issues in a contact report as follows: [Section editor A] was labouring under a disadvantage because of his late appointment following [author A's] resignation. Mr Maxwell agreed that Barbara Barrett should visit [section editor A] to assess his requirements and advice as necessary; should he require any financial support for an editorial/secretarial assistance up to 2000 USD would be available for this purpose. 69 Incidents like this and the way of handling it by sending Barrett onsite justify the conclusion that her responsibilities went far beyond technical matters and required soft skills. This also shows in the case of Robert Kidd's death during the editing process of the first edition.
We will, as agreed, be writing to all Adult Education authors informing them of Roby's death. 70 J. Roby Kidd (1915Kidd ( -1982, section editor of Adult Education, unexpectedly dies in 1982. Barrett therefore takes the responsibility to inform the authors, who belong to the section, about the loss. This again relates to her soft skills, as the editors-in-chief emphasise in the preface. Barrett explains the further procedure and discusses relevant editing steps after Kidd's passing with Elizabeth Burge 71 (n.d.), who as an associate temporarily assumes the editing of the section on Kidd's behalf. Because of the sudden lack, only Burge's engagement keeps the editing running. Her case is another strong example of gendered public acknowledgement of labour in the production of the IEE. The material namely shows that she took on nearly full editorial responsibility for the section before Colin Titmus (from Goldsmiths' College, University of London) takes up editorship. But contrary to Burge's effort and full commitment, she is in the preface only mentioned as an editing assistant 72 and not as a member of the editorial board. The decision on whether or how to present Burge as a vital part of the section's editing or at all to publicly emphasise the organisational struggle after Kidd's death was certainly dependent on the editor's strategy to frame the IEE as a legitimate contribution to the profession. According to Beede, it is necessary to include and present recognised contributors on a high career level to construct authority for a reference work like this. 73 Burge did not seem to fulfil these criteria even if her actual contribution was on an editor's level.
Barrett not only handled care work on behalf of the editors-in-chief by supporting section editors, she even addressed both directly in that sense.
I thought you might be interested to hear that an American Library has nominated the Encyclopedia for the Dartmouth Medal . . . I'm sure you will agree that even if the Encyclopedia is not the final winner it is something of an honour to have been nominated and is yet another testimony to the enormous effort that you and Neville put into the venture. 74 She herewith takes responsibility for motivating and praising Husén's and Postlethwaite's effort, without including herself as part of the team even if that is, as mentioned above, presented as a "trio" to the section editors. She therefore also seems to incorporate the distinction between her kind of knowledge work, apparently only editorial support, in contrast to their knowledge work of editorial contribution and content-related input. As it indicates in this example, the tone between Barrett and Husén becomes personal over time by involving personal information and, on Husén's side, personal greetings from his wife Ingrid to Barrett, 75 who neither includes much personal information about her nor mentions a partner. That Husén and Barrett did not equally include personal information implies that it did not merit a mutual personal approach, but rather a professional way of interaction between co-workers on implicitly different hierarchy levels.
Indirect and direct influence on content/knowledge Beyond the modes of Barrett's involvement described above, which prove her rather indirect influence on content by facilitating (as the editors-in-chief name it) the editing process, she even seemed to become in some way an expert and actor of the field herself. By managing the stream of entries, for example, Barrett also handles, transmits, and translates content in the network. This function requires her to communicate content in a comprehensive way to the actors involved to avoid misunderstandings or misinterpretations of necessary action to be taken. Therefore, it was vital that Barrett herself was able to understand what she was talking about. She thus protocols the meeting's issues for distribution to the editors-in-chief and publisher's management: Because of the culture-bound nature of specific vocational programs [section editor B] had structured the entries in his Section according to generic principles. He was however 72 Husén and Postlethwaite, "Preface," xviii. 73 Beede, "Editing a Specialized Encyclopedia", 5. 74 Letter from Barrett to Husén, 31 January 1986, SE/RA/720869/Vol. 2:276. 75 Husén includes greetings from his wife Ingrid in his correspondence to several colleagues, but it is especially striking in his letters to Barrett. considering including coverage of the five principal areas of vocational and industrial education e.g. agricultural education and business and office education and he would let me have his final decision by the end of November. 76 Barrett's handling, transmitting, and translating of information represents her indirect influence on content. But she is not just responsible for passing on information and therefore to understand concrete content; she also seems to be engaged in approving content and therefore is directly involved. Explanations of content to her in the correspondence imply that entries needed to be justified to convince her about their appropriateness. 77 This kind of involvement (being consulted concerning content) does not just prove her direct influence on content, it even shows that she was provided with influential agency regarding the actual content. Eventually, this mode of involvement culminates in specific discussions about concrete entries. The most remarkable and condensed example is Barrett's debate with a section editor in the editing process of the first edition. She wrote the following to him after he complained about changes in his entry: To take your first pointwe have, as requested, re-instated the term "aversive consequences". I am, however, somewhat puzzled by your definition of negative reinforcement, particularly your statement that a negative reinforcer is "not punishment at all". It was always my understanding that a punishment, such as giving a rat an electric shock every time it touched a lever, was correctly termed negativesee, for example, the enclosed copy of a passage from the Open University course unit on socialization (this was incidentally based on Hilgard and Atkinson's well-known Introduction to Psychology). You will see from this passage that negative reinforcement is actually defined as punishment leading to the suppression or inhibition of the response as opposed to increasing the future incidence of the behaviour. I also noted that L J Keil in his article on ATTITUDE DEVELOPMENT equated negative reinforcement with punishment (line 344) while Daniel Duke has used the term in this sense in his article on PUNISHMENT, although you have in this instance changed it to "aversive consequences". I hope this goes some way to explaining why the editorial change was made. 78 This example shows that Barrett performs as an actor in the process power and direct influence on content in several ways. 79 At first, she or her team seemed to have edited the entry in a significant way without consulting the author. This requires a certain level of content-related confidence and legitimate power. Secondly, she dares to further discuss the matter after he complains and requests to delete the changes. And finally, she does that by precise references to professional literature and resources. These last examples clearly show how Barrett had the confidence and legitimacy to directly affect or interfere with content.

Discussion
When Latour and Woolgar worked on Laboratory Life 80 they based their findings on realtime observations of laboratory work, studying the social processes which belong to the construction of scientific knowledge. In social science, knowledge production processes are not accessible or locatable like an actual laboratory is. Therefore, they are even more dependent on the activities allegedly outside the laboratory: readings, translations, meetings, dialogues, correspondence, thinking, and writing over vast time and space. We show with the example of the IEE that especially editorial processes of collaborative publications offer great potential to be informative by treating it as an equivalent to the laboratory. Here working with historical, archived material allowed a deeper understanding of the knowledge production process by analysing the correspondence in the context of the completed publication. Therefore, it was possible to put the correspondence in relation to the preface of the IEE, which offered the picture of the process that was portrayed to the public by its editors. The study shows that Barrett's actions and influences are more complex, fundamental, and central than publicly framed and expected from her alleged peripheral position. In contrast to the preface, in which the editors-in-chief present a rather traditionally gendered division of labour into the relevant content-related and administrative responsibilities involved in the process, Barrett clearly not only supports the knowledge production process, but also influences the content level indirectly and directly. At first, it becomes clear that her overview and control over the IEE production are central to its success. She provides on different levels the conditions for all project steps and therefore the content production. She keeps the flow of information running by being a central mediator. She connects all actors involved by keeping actors informed in correspondence, by organising meetings and basically by providing the necessary paperwork. She also undertakes care work, which is emphasised by the editors-in-chief by praising her "personal sensitivity". 81 This traditionally constitutes a task rather performed by women; the performance by Barrett therefore might not be surprising. The presented analysis thus stresses the lack of acknowledging Barrett's involvement content-wise.
As shown above, by performing care work and therefore providing at least necessary conditions to proceed she is a central actor in the knowledge production process. Furthermore, having the overview, she is able to consider flexibility in the time schedule and adjust the process, if unpredicted and even personal obstacles occur. She mediates or ensures support if needed by section editors and even seems to feel responsible for motivational incentives. Beyond her indirect influence on content level it was possible to prove even her direct involvement in the knowledge production. She was not only included in correspondence about content in matters of being informed and updated, she was also directly involved in discussions. The most remarkable proof is her discussion with a section editor, which is presented in the end of our findings. It shows that she not only edited content without consultation of authors, but also dares to discuss that further with references to professional sources and literature. This requires a certain confidence in her and the acceptance by other actors of her interfering. This incident proves the legitimacy of her actions and therefore her direct influence on the IEE, even though it is presented differently. 80 Latour and Woolgar, Laboratory Life. 81 Husén and Postlethwaite, "Preface," xxii.
By following Barrett through the laboratory, through material of this historic example of an encyclopaedia's editing process, we are able to show beyond a gendered division of labour a gender related disbalance in acknowledging knowledge work. We encounter the "lack of female representation" 82 and the example of Elizabeth Burge, who temporarily takes over a section after the former editor's death and similarly only gets mentioned as support. Therefore, not only the exclusion of women becomes visible but also the underestimation of their contributions and performances, when they are included.
For further studies, these findings encourage to follow even more apparently peripheral perspectives in the laboratories of social science knowledge production. With its approach, this paper contributes to a deeper understanding of knowledge production processes and of related marginalisation of certain groups at the end of the last century. This historical example offers a departure point for reflecting upon present practices, and not only those of editing but in social sciences at large.