FFF proceeds from the fractured ontological nature of social life. Feminism as a politics centres a radical social ethics; for FFF, this necessitates producing feminist knowledge in an open, accountable and defensible way. Doing this starts with the research labour process. In particular, FFF is concerned with analytical processes concerning how knowledge is produced and the claims made for it. For FFF, method in the narrow sense is unimportant and what matters is why and how they are used and for what purpose. FFF is instead concerned with analytical reflexivity and using retrievable data: providing evidence as well as the interpretation based on this means readers can reach their own conclusions.
2.1 Britain provoked the South African War (1899-1902) against the two Boer Republics of the Orange Free State and South African Republic and annexed these as the Orange River (ORC) and Transvaal Colonies. The war, however, did not end there. Boer commandos continued fighting; the British 'scorched earth' policy was a response, as was the institution of direct British rule in the former Republics. Our camp children project was designed with two inter-related components, concerned with exploring the two main ways that experiences of children were represented following the removals of Boer people to the concentration camps during the war. One is the official record, the other involves photograph collections.

2.2 The official records were kept by British administrators at local level in the approximately forty Boer concentration camps, and also in the headquarters of each Colony.[17] These were records of an organisational apparatus concerned with maintaining and regulating the 'removed' populations, with accountability ultimately held by the British government in Whitehall, but beneath this by the colony governments[18] and their chief administrators of the camps, and then beneath this by the superintendents or administrators of each individual camp. There are many extant records – 'returns' of information and also summarising registers.[19]

2.3 The first component of our research involved picking one element of these records, in one of the annexed colonies, for detailed investigation: this was the 'minutes' or files of inquiries between the Chief Superintendent of Refugee Camps (SRC), and the superintendents in charge of the Orange River Colony's (ORC) eleven camps. In the SRC archive collection, there are nearly 11,000 such inquiries, some of which are specifically concerned with children's lives, illnesses or deaths.[20] Accountability for the concentration system was grounded in bureaucratic record-keeping and meticulousness in following agreed procedures; and the fullness of the official records, originating in many different local sites, is striking. So too is the emphasis on following laid-down procedures, like those instituted to bring the epidemics under control, with the inquiries indicating that officials could be so concerned with following procedures that they were unable to 'see' the emotional and political impact that these were likely to have on camp inhabitants. In addition, there was a high level of official inquiry and intervention concerning safeguarding the welfare of children:[21] there are many examples of cutting-edge good professional practice, for instance regarding likely cases of infanticide and child maltreatment by children's carers.

2.4 The second component of the research involves our investigation of the archive collections of photographs of the South African War, specifically a sub-set of the concentration camps in one particular collection, that in the Free State Archives Depot.[22] These were taken primarily by professional photographers, who travelled to war locations and routinely visited the camps, selling their photographs to newspapers world-wide, and to British soldiers and Boer commandos and people living in the camps. A minority were taken by people employed in the camps, as administrators, doctors, teachers, nurses and so on, using one of the Kodak/Eastman cameras available from 1898 on. Many of the professionally-taken camp photographs were 'views' and general pictures sold to newspapers or in local shops; others were taken 'to order' and are 'scenes with people' or posed individuals and groups, which would have been bought by camp inhabitants as mementoes or posted to absent relatives on commando.[23]

2.5 There are three main collections of South African War photographs in South Africa holding around 80,000 photographs.[24] The collection held by the Free State Archives Depot in Bloemfontein (VAB) covers the same camps as the SRC records of inquiries and therefore became the focus of our investigation because of this. These photographs throw a different light on the concentration camps from the official records, the inquiries concerning children in particular, because they feature a much wider range of children's presence and activities. This is absolutely not to suggest that the photographs are more 'real' than the official records (or vice versa). But it is to emphasise that the two sources have different origins and purposes, and that the photographs involve camp inhabitants being more active in choosing how they were represented and also in keeping some of these as mementos post-war.

2.6 In what follows, 'retrievable data'[25] are used to tie the discussion as closely as possible to key documents (both written and photographic) that 'stand for' or exemplify wider aspects of our research and interpretations of it, so that readers can evaluate arguments against their own. We do not claim referentiality for these texts, including the photographic ones. They are not directly reflective of 'the real happenings' that gave rise to them, and nor are they the total evidence drawn on. However, providing them shows something of the evidential base being worked from and enables a more informed readerly evaluation of interpretations and conclusions. This claim is returned to and discussed in the last section of the article, where we comment on the FFF process as a whole and our use of retrievable data within it.

(ii) Grounding feminist knowledge

While FFF may advance preferential knowledge-claims on various grounds, it rejects epistemological privilege, proposing that all knowledge-claims should be evaluated on their specific merits. Consequently the knowledge-claims made by FFF should concern specific examples and be grounded in particular evidence and interpretations. And whether the feminist researcher 'knows' in an authoritative sense depends on the sufficiency of evidence for the conclusions drawn, the plausibility of interpretations and conclusions, and the reception by readers.

2.7 We start with some more information about the research process and why the particular inquiries and photographs that were selected were decided upon, beginning with the SRC inquiries. Many of the total set of SRC inquiries cover matters including or of relevance to children, who composed the large majority of the populations of all the camps. However, our interest was in those where children were the specific focus, so we worked through detailed notes on all the SRC inquiries to identify this smaller group, using Liz's previous research on the collection as a basis. Then the inquiry papers for each of the children's inquiries were read in detail (using other SRC records to generate additional information where available[26]), with Table 1 summarising the concerns of these.

Table 1: SRC inquiries concerned with children

Area of concernNumber
Policy & practice regarding orphans[27]15
Child mortality issues8
Formal inquiries & P-Ms7
Specific children in/out camps5
Administrative issues & mis-records4
Illegitimacy 3
Children's rations2
Miscellaneous 5
Total 49

These 49 inquiries were examined in detail, including in relation to wider aspects of the war and the camps. From this, seven inquiries were selected, discussion of which would enable covering the broad distribution of inquiries in Table 1.[28] The retrievable data for each is provided in Appendices 1 to 7, and will be referred to in the discussion which follows through hypertext links in the text at appropriate points.

2.8 A similar procedure was adopted regarding the VAB photograph collection, which neither of us had worked on before. The result was identification of 116 concentration camp photographs, in 61 of which children appeared, with classification of them shown in Table 2.

Table 2: VAB photographs representing children

SubjectsNumbers Totals
Views 14 14
Adults 41
- employees & others 20
- camp inhabitants 12
- Relief Committee[29] members 9
Children 61
- posed (formal, informal) 14
- playing 2
- working 9
- school/Sunday school 3
- soup distribution[30] 2
- 'relief' activities 12
- Peace Day[31] 3
- sick & hospital 2
- views with people, including children 14
Total 116116

The decision to use retrievable data meant considerable discussion of the photographs with children in them, including in relationship to the SRC inquiries being worked on. As with the inquiries, photographs were selected to broadly reflect the distribution of subjects across the whole set, resulting in the eight photographs shown in Appendices 8 to 15.[32] These too will be referred to in the discussion which follows through hypertext links.

2.9 Finally, back in the UK and during the initial work on writing up the research, we focused down on three inquiries and three photographs around which to discuss putting FFF into practice, a small enough number we thought at the time to provide readers with specific examples and also the interpretations based on these. However, we rapidly realised this was too many when working within constrained word limits.[33] The result in our very first published account of the research was a decision to focus mainly on one inquiry and one photograph, but still to provide the retrievable data for the more representative group of seven inquiries and eight photographs, enabling readers to place the two cases discussed in a broader context. The inquiry was selected first, on the basis that it raised the widest range of issues. This was the Swart[34] case, with the retrievable data here a letter from Harrismith's superintendent, Arthur Bradley, written on 25 December 1901 to the ORC's Chief Superintendent, Captain Trollope, and is provided as Appendix 3. Then the photograph was selected, to raise some different issues about a different camp. This is a photograph from Kroonstad, provided as Appendix 10. The wider group of retrievable documents are referred to through hyperlinks in the following discussion which direct readers to the original documents, and also notes containing summary interpretations of these. Active readership of these by readers is required in two respects. Firstly, the whole set of retrievable documents provides a broader informational context for readers to locate the Swart case and the Kroonstad photograph in; and so all of these need reading (the inquiries) and looking at (the photographs) in detail. And secondly, the accompanying summarising interpretations of these need to be evaluated against reading the retrievable documents themselves.

(iii) Unalienated knowledge

Bracketing or cancelling out the act of knowing is highly consequential in feminist terms: it renders invisible the indexical properties of knowledge within a false 'universalism', and by denying the labour involved it alienates this as a devalued commodity. Important dimensions of an unalienated feminist knowledge include grounding the feminist researcher and her research as an actual person at work in a concrete setting; recognising that understanding and theorising are material activities which can be accounted for; and linking the 'act of knowing' (research process) with claims about 'what is known' (research product).

2.10 We start with a methodological conundrum. Superintendent Bradley's letter in Appendix 3 provides a retrievable document which sets out many of 'the events' concerning the death of a young girl, Christina Swart, that came under inquiry in late December 1901 and it can be read and understood as a self-contained text. At the same time, the inquiry went from SRC to Colonial Secretary levels and generated two linked inquiries involving 27 individual records and more words than this entire article, and fully understanding any one of these documents would necessarily draw on wider research on the concentration records as well as 'what came after' in the history of South Africa. Some of the inquiries are even more fully documented; and it is also worth noting that even those which are composed solely by the retrievable document provided, such as Superintendent Shutte's letter about purchasing things for the people affected by the marquee burning in Appendix 1, still require such background knowledge to make full interpretive sense of them. Three examples of interpretive work on Bradley's letter requiring this background knowledge are as follows.

2.11 Firstly, it is highly likely that the heart-wrenching death of Christina Swart was the origin of the repeated refrain of 'sick children seemingly routinely torn from of their mothers' arms to die en route to hospital' that appears in the (proto-) nationalist-orchestrated women's testimonies which were published post-war.[35] Secondly, this is absolutely not to deny that what happened to Christina was appalling and inhumane, but instead to emphasise that the fierce political divisions of the war impacted on people's understandings of 'the facts' and the intentions and meanings they ascribed to these, the source of the disputes commented on in Bradley's letter and across all the Swart inquiry documents. It is also important to contemplate what the seven retrievable documents in Appendices 1 to 7 and the photographs in Appendices 8 to 15 demonstrate concerning what has been 'forgotten' from received or canonical knowledge about the South African War concentration camps: the many accidental deaths brought about by parental tiredness or carelessness (see Appendix 1), cases of infanticide and also murders (see Appendices 2 and 4), the presence of large numbers of sometimes powerful men in the camps (see Appendices 2 and 3), internal disputes and differences among officials (see Appendix 5), organisational requirements and reprimands, the detailed recording of black people's lives and deaths and overall presence (see Appendix 6), and the need to 'do something' about orphans along with the political context this was perceived within (see Appendix 7). And thirdly, some of these complexities are brought into sight by reading Bradley's letter from different viewpoints. Thus the dispute 'on the day' involved Superintendent Bradley, doctors Rossiter and Ralston, Christina's mother Martina Johanna Swart, Resident Magistrate Arthur Leary, and also the civil surgeon carrying out the post-mortem, William Beor,[36] with statements of their various positions appearing in the full set of inquiry documents; and we take 'recovering' and attempting to understand the grounding of these disputing viewpoints to be crucial to the FFF labour process.

(iv) The knowing subject

At an epistemological level, FFF involves a double-take on what 'knowledge' consists of: knowledge is what is constituted as such within everyday knowledge practices, and it is also what feminist researchers make of this (and these may conflict, of course). All society members are engaged in such activities, because all of social life revolves around practical knowledge matters. However, while researchers engaged in social inquiry will usually do this in a more structured and formal way, FFF does not see these activities as different in kind from everyday practices. FFF resists seeing people as immersed in the local and unable to discern the wider relations and structures of ruling, and rejects epistemological privilege for feminist research.

2.12 From Superintendent Bradley's letter, it is clear that Trollope, the doctors and the superintendent were all aware of wider 'non-local' issues of different kinds. Similarly, Parry Edwards, as the Orange River Station medical officer, was clearly aware of overlaps but also differences between overlaying, infanticide and murder, in Appendix 2 ; while Major Wilkins – who replaced Captain Trollope as ORC Chief Superintendent in early 1902 - in the adoption inquiry in Appendix 7 has an eye on the legalities as well as on finances and wanting orphans removed from 'the books' where possible. Returning to Bradley's letter, in it Trollope is positioned around notions of accountability within an organisational hierarchy, the doctors as focused narrowly on the epidemics and deaths and their status in this hierarchy, while the superintendent presents himself as concerned with both humanitarian matters and with issues concerning the Swart 'clan'. In addition, Bradley's letter clearly implies that the Swart men were capable of acting as a power bloc in the camp at the present-time (not just concerning the future 'settlement of the colony') and could influence people back into their 'old discontented state', with one of its sub-texts being how the Swart men would react to Christina's death and its possible political reverberations. The 'absent presence' of Klass Saaiman's father in Appendix 2 raises similar questions about the presence and power of men in the Orange River Station camp, too.

2.13 However, while the Swart men are presented as wealthy and influential, there is no document on file providing their actual viewpoint: this is assumed and invoked, but not presented in its own right. By contrast, Mrs Swart was the determined and active agent whose response to her daughter's illness, and then the circumstances of Christina's death, forced the inquiry to happen; and she and others testify under oath about this. Without her insistence that a post-mortem by an independent person should be carried out, this would have remained an unseemly dispute between adults around a dying child, rather than erupting into organisational sight at a high level. Mrs Swart's insistence then required the Resident Magistrate to write a report concerning the post-mortem and the inquest following, at which the key parties all made sworn statements; and in this, the Magistrate went out of his way to punitively conclude that Christina's death was due to 'improper mutrition [sic] and bad treatment by parents' (because Christina's intestines were full of food, whereas typhoid patients had to be starved to have a chance of survival).

2.14 Another factor requiring contemplation here is that Christina's parents failed to report her typhoid for two to three weeks, with what they thought were her best interests at heart.[37] As other filed documentation shows, they had deliberately hidden Christina from the daily 'sick call' carried out by Boer probationer nurses[38] and secretly buried her lethally-infected faeces. Also, while the post-mortem shows that Christina had been fed on solids, her mother's sworn statement mentions feeding liquids but not solid food:[39] it seems Mrs Swart was aware of the strict advice to starve typhoid patients, but her wish to do her best for her daughter had over-ridden this and she thought that the doctors attempt to take Christina to hospital, rather than her feeding solids to her daughter, had killed Christina.

(v) The fractured ontological base

FFF is predicated on the ontological position that social life is inter-subjectively constructed around ideas and practices concerning structures understood as 'social facts' which are external to and constraining upon society members. Social life is at one and the same time experienced as independent of social construction, but is actually constituted by this. Knowledge necessarily has an ontological basis: there is always a knower situated in time and place with others, going about the business of knowing the social world. Because different collectivities of people understand realities and facts from where (geographically, socially, politically) they are situated, everyday fractures of understanding and meaning - reality disjunctures - frequently arise; however, these are negotiated (sometimes successfully, sometimes with remaining disagreements) around the shared premise that there is real meaning, facts and truth – a social reality - to be arrived at.

2.15 There are complications surrounding 'the facts' in the Swart case, something which also comes across concerning Parry Edward's attempt 'to know' about the death of Klass Saaiman in Appendix 2, with his classification of possible causes of death, reasons why he made the post-mortem, and his careful conclusion. Superintendent Bradley's letter conveys that his response to the doctors was a humanitarian one about Christina Swart, echoed in the comment written in an accompanying document by the ORC Governor, that 'undoubtedly the mother is to blame; yet it seems strange that the child would be removed when actually dying'. However, the superintendent had other motives too, for the implication of his remarks (corroborated in other documents) is that he actually intervened because of the importance of Christina's uncles in the camp. Also, while his letter stresses the 'future settlement of the colony' (a political and imperialist reason he perhaps thought would 'go down well' with Trollope and the latter's political superiors), it also indicates that his pressing practical concern was 'managing' the men in the camp and preventing their 'discontent'. In addition, other filed documents show him 'covering his back' by masking the fact that Mrs Swart had reported Christina's sickness to him before the doctors found out about it.

2.16 Further complexity can be added to this 'mixed motives' reading of the superintendent's role by taking the best possible view of why doctors Ralston and Rossiter acted as they did (it is easy to think the worst because of the effects on Christina Swart, but it should be recognised that they were working in an extreme situation). There are various letters and sworn statements from both doctors on file and, among other matters, these propose the superintendent's sympathies had led him to neglect his duties in the extreme medical emergency that Harrismith was experiencing, while the Resident Magistrate, arriving in camp for the post-mortem, put on record that he had witnessed Bradley appeasing another family who were resisting their child ill with typhoid going into hospital. There had been a large rise in the camp's death rate from November on. The doctors also comment that the Swarts had buried Christina's typhoid-infected faeces in the immediate vicinity; not explicitly stated, because for the doctors self-evident, was that consequently the Swarts were likely to have been responsible for part of Harrismith's hugely-increased death rate in December and January, shown with devastating clarity in Table 3.

Table 3: Harrismith deaths and death rates, May 1901 – March 1902[40]

Camp Population Deaths Death Rates[41]
May 1901---
June 656118
July 927338
August 1134552
September 1304218
October 1596537
November 16501394
December 162363761
January 1902147032261

(vi) Knowers and competing knowledge-claims

The response of FFF to the question of who can be a 'knower' (in the 'having authority' sense) is that this all depends on where people are situated within the relations of ruling[42] and the operations of power/knowledge in particular contexts. Certainly subjugated knowledges can be given greater or even privileged status, as feminism has accorded to the category 'women'. However, while FFF involves re-evaluating the perspectives and knowledges of subjugated and dominant groups, those of particular women or men are not necessarily viewed as preferential or devalued: this will depend on persons and circumstances. FFF considers competing knowledge-claims around the context or situation, the people involved, how 'the facts' are represented, and who sees these as convincing or flawed.

2.17 Reading the full documentation of the Swart inquiry, what comes across is that all the adults involved had good (but different) moral reasons for acting as they did; however, their same actions can also all be seen as morally wrong when looked at from a different angle. The superintendent had good humanitarian reasons for resisting the doctors' decision – but he only acted on his sympathetic response because Christina's uncles were powerful in the camp. The doctors were battling with rampant typhoid and bringing the death rates down – but they wanted to be top dogs in the official hierarchy, despised the peasant medical beliefs of Boer farming people, and forced the removal of Christina even though she was dying. Mrs Swart comes across as a woman of considerable determination and it is impossible not to feel sympathy for both Swart parents in wanting to protect their daughter – but they probably killed dozens of other people in the process, while Christina might have survived had she gone into hospital (political rumour had it that children were deliberately killed in the hospitals, while the actuality is that survival rates in hospitals were many times better than when people stayed in their tents[43]).

2.18 We have found it impossible (and nor do we want) to adjudicate in the Swart case, in the sense of deciding 'who knew' and whose 'the facts' should prevail: we recognise the different viewpoints involved and we agree and disagree with them all and at the same time. And something similar arises concerning Magdalena Saaiman in Orange River Station camp in Appendix 2: for the doctor in the case, Parry Edwards, there was a difference between 'genuine' overlaying and overlaying linked to infanticide, although in practice he knew that no 'charge of criminal intention' would be brought in relation to either. Its challenging complexity is precisely the reason we selected the Swart case from among the range of possibilities: it raises not only the intricacies of moral or ethical positioning, but also the difficulties in getting a firm purchase on 'the facts'. Another example of this is that, while in one sense 'the facts' are very clear concerning Katit's death and her Death Notice in Appendix 6, the moral dispute here arises concerning 'canonical knowledge' about the concentration camps of the South African War, which 'forgot' the presence of black people in all the 'white' camps, and also 'forgot' that the organisational record system applied as fully to black people as to whites. In a sense, the facts in the Swart case are also completely certain; but also it has to be recognised that the disjunctures and disputes arose because what was important for the adults concerned were their competing interpretations of intention and meaning, which rewrote how they understood these facts.[44]

2.19 One of the features of historical research is that in a sense such disputes are "over" now, and so the temptation to take sides is less than in present-time research. That is, there is no possibility here of the researcher 'intervening' other than by suggesting interpretational possibilities. Also, in the South African case, this 'after' involved rapid political control by the Afrikaners and the institution of segregation and then later apartheid. Taking E.P. Thompson's (1963) injunction for historical research to avoid 'the enormous condescension of posterity' seriously, it has to be recognised that the Boer Republics before annexation, then South Africa after Union and Afrikaner political control, were predicated on a highly racialised form of (proto-) nationalism, and the Swarts and many other people like them were a part of this. This is certainly not to deny the sufferings of Boer people during the South African War – but it is to insist that politics greatly impacted on how wartime events were remembered, and to emphasise that politically-motivated interpretations of intention and meaning really did imbricate 'the facts' both during and post-war (Stanley, 2006).

(vii) Moral epistemology

Epistemology always has a 'moral' or ethical dimension: claims to know are made against or over others, not everyone is deemed a competent knower and so on. 'Moral knowledge' is knowledge that is transparent, produced through non-exploitative means, makes defensible knowledge-claims, produces open accounts of 'findings' and conclusions, and is fundamental to FFF. Also FFF involves a feminist, rather than women's, standpoint, organised around ensuring transparent and accountable good practice for feminist social research concerning the activities involved and the knowledge-claims and the written or other knowledge products that result.

2.20 Evidence from other camps where records about where people died – their own tents or hospital - are extant shows that children survived much better in the hospitals.[45] Clearly, there was something else at work that led to powerful fears about the camp hospitals from women like Mrs Swart. Certainly implicit in the Swart case is a 'clash of medical cultures', the peasant one of Boer farming people and the European-educated doctors, nurses and administrators.[46] In addition, even commentators most sympathetic to the Boer cause and way of life, like Emily Hobhouse,[47] noted the preoccupation with fatty meat and fat in other forms, with rations of lean meat interpreted as deliberate starvation. Alongside this, the standard contemporary treatment for typhoid and all forms of diarrhoea was no food at all. Rations of lower-fat meat, and starvation as standard medical treatment of diarrhoea conditions, underpinned the fears of even well-off farming people like the Swarts that the intention was to starve them to death, so that like Mrs Swart they resisted hospitalisation for what seemed good knowledge-based reasons.

2.21 Does considering 'moral knowledge' in a more direct sense add to comprehension of the Swart case? This might, for instance, involve exploring more about Mrs Swart's agentic and precipitating role in events, by looking closer at how she appears in the range of inquiry documents, and also exploring other SRC records for Harrismith to find further information about her.[48] And a somewhat different take on moral knowledge in this project concerns our determination not to repeat the 'forgetting' of black people that marks canonical knowledge about the South African War concentration camps, and thus our determination to include in our inquiry selections Katit's death notice in Appendix 6, and also the photograph of the smiling young girl in Appendix 11.

2.22 However, FFF is a political and analytical position, rather than a representation of the world according to women, and we are interested in the dimensions of 'moral knowledge' around the most vulnerable and least powerful person involved: the dying Christina Swart, who appears in the inquiry documents only in extremis, with other children in extremis in the retrievable documents being Jacob du Preez (Appendix 1), Klass Saaiman (Appendix 2) and the baby daughter of Miss De Klerk (Appendix 4), for none of them have or can be provided with a narrative voice within the confines of the existing documents and existing ideas of what 'narrative' is composed by.[49] Christina is a cipher in the Swart inquiry documents, an 'absent presence' at the heart of events - that is, she appears only insofar as adults, particularly her parents, attribute feelings and motives to her and not 'in her own right', a clear case of 'adultism'.[50] So, might it be possible to read the inquiry documents to bring Christina into view? We concluded it was not and that this requires consideration of other situations where children became the focus to understand more of the dynamics at work.

2.23 A less fully-documented but more notorious case of a child whose death was disputed between adults exists and involves Lizzie van Zyl, a young girl from Bloemfontein camp. There is a much re-published photograph of a skeletal Lizzie, with her appearance almost automatically assumed to 'show starvation' by adults on all sides.[51] Initially, her mother removed her from hospital following typhoid, because of the starvation diet she was placed on. Then people in neighbouring tents suspected that none of the food visibly present in her family tent was being given to Lizzie. And once the photograph was sent by Hobhouse to the international press, 'the British' were blamed for starving Lizzie, while the British response claimed that she had come into the camp looking as she did and her mother must be responsible. Thus Lizzie van Zyl's illness and death became caught up in a two-sided propaganda battle, as well as occasioning major disputes about 'the facts' during the events of her long illness and then death in Bloemfontein camp itself.

2.24 Like Christina Swart, Lizzie van Zyl has 'vanished' in these disputes between adults, except that her pitiful photograph remains (and is still used, in both main senses of the word).[52] Even when children were the focus, we conclude this was still overwhelmingly as they were seen and positioned (and 'voiced') by adults. Across the entirety of the concentration records, there are very few exceptions to this. One partial exception involves official records in which the clerks or other adults concerned have written things on behalf of orphaned children. The other exception is equally but differently problematic and is provided by a collection of reminiscences by elderly adults of their experiences as kampkinders, camp children, testimonies which were solicited, edited and published, as well as told, through a nationalist political lens some seventy years after the event.[53]

(viii) Analytical reflexivity

The knowledge practices and products of FFF reflect a specifically feminist politics and ethics, with analytical reflexivity key to ensuring transparency and accountability. Analytical reflexivity focuses on the acts of knowing and what goes into this, looking in detail at the analytical processes involved and the evidences supporting these. Analytical reflexivity also entails writing an open research text that: adduces evidence in retrievable form that is appropriate and sufficient for the argument being made and it provides sufficient detail for readers to be able to make their own interpretations and so evaluate conclusions and claims.

2.25 Unless the focus of a research project is based entirely on retrievable documentary sources, it is really not possible to provide readers with all of the relevant evidence.[54] Even a small and limited archive project such as ours, and even focusing on just one case within it as we have done, generates more documentation than can be made available; and anyway many of its arguments and interpretations depend upon knowledge gained from a wider research context, as we have repeatedly commented. It is therefore useful to explore the process of 'knowing' around an archive source which we know no more about than readers. This is a photograph which archive information states was taken in Kroonstad camp and is provided in Appendix 10.[55] Our research project combined researching all three of the main archival collections of photographs from the South African War as well as the concentration camp written records, and an earlier version of this article discussed a group of photographs selected by similar means to the inquiries we have focused on. However, because of space reasons, we have abandoned discussing all of these at equal length, although we have provided all the initially selected photographs as appendices to which our discussion of the particular photograph selected is hypertext-linked at appropriate points.

2.26 As a letter written by a young woman, Lottie Roussouw, to her friend Lottie Theron in Harrismith camp, suggests, professional photographers visited and photographed the inhabitants of Kroonstad camp:

"Thank you too for going to send us the photo you mention. Having ourselves taken in 'Camp Style' is a great wish of mine, but owing to shortness of cash I do not like to plague ^Papa^.[56] If I had a small income of my own I would think the money well spent."[57]
Lottie Roussouw's letter to Lottie Theron links Harrismith, the camp the Swarts were living in, with the very different Kroonstad camp. This was much bigger and had many extremely poor Boer people living in it;[58] consequently, using a retrievable document concerning Kroonstad provides a means of recognising the experiences of Boer people living very differently from the Swarts (and see the photograph in Appendix 14, showing a family of comparable status to the Roussouws and the Swarts, and also those in Appendixes 13 and 15, featuring people like most of Kroonstad's inhabitants).

2.27 The mundane content of this photograph from Kroonstad (Appendix 10) - distributing wood and people queuing for this[59] - provides a direct contrast to the dramatic and upsetting events of the Swart inquiry and enables some of the more everyday and routine aspects of camp life to be – literally – seen. We start with what can be seen when looking closely at this photograph; then consider some of the things not actually 'in it' but which can be 'seen' when it is looked at with knowledge of the wider context.

2.28 Amongst other things, this Kroonstad photograph illustrates the orderliness and routinisation which characterised much, indeed most, of camp life. Many of the 'views with people' across the three archive collections are of this kind and suggest, firstly, that 'ordinary life' continued, because cooking, eating and, in South Africa's extremes of weather, keeping warm and dry, remained priority activities; and secondly, that concentration introduced new levels of regulation (e.g. rationing food and firewood) and routinisation (e.g. organising set times and places for their distribution) to ensure this.[60] Also like other 'views with people' in the collections, this photograph witnesses the significant presence of able-bodied men, whereas received wisdom is that the camps were composed of women and children and very elderly people, with all Boer men supposedly loyalists on commando. It also shows the highly gendered division of labour at an everyday level - men and older boys in the foreground are sorting and apportioning the firewood, while women, girls and younger children in the background are queuing and waiting to be given the appropriate ration.

2.29 This Kroonstad photograph was taken by a professional photographer who would have sold the image on – a printed caption appears at the bottom, as does its number in a sequence. It is a 'new photograph', in the sense that its subject departs significantly from both studio portraiture and the more usual 'views' without people on sale in this period…. And it was taken for sale to people, in South Africa and elsewhere, wanting informative photographs showing 'what was going on' in the war and the camps (indeed, apart from the photograph of the Venter children in Appendix 8, all the photographs in the Appendices come under this heading). Other archived photographs in this particular Kroonstad sequence are also 'views with people', including more firewood distribution photographs, the camp mortuary with coffins outside, and burning infected cattle in an area close to the camp, with both the latter being 'sensitive' subjects because raised as part of contemporary criticisms of conditions in this camp.[61]

2.30 There are two further matters worth commenting on, neither 'there' in the photograph itself but important to interpreting it. The first concerns the women and children standing in line and queuing. This of course can be seen - but what is not visible is what this meant for the camp inhabitants. Queuing was a raced activity in the Boer Republics, with such 'lowly' tasks done by black people, not whites, and was a highly resented aspect of life in all the camps because it indicated Boers not being treated as 'superior' in race terms. Queuing was understood as part of the inexplicable and unforgivable 'topsy-turvy' features of the war in overturning the assumed natural order, with other resonant examples of this being the children fetching water in Appendix 9, Boer women washing clothes in Appendix 11, and cooking and cleaning in Appendix 15.

2.31 The second is what this and other photographs by professional photographers indicate about the openness of the camps to people from 'outside'. As noted above, many 'sensitive' subjects were photographed. However, there is little sense that photographing in the camps was subject to any very stringent controls. Thus, looking at photographs in this Kroonstad sequence (some of which are numbered), it is clear the photographer in question was walking about the camp, taking photographs as different activities and 'scenes' met his eye (all the known professional photographers who worked in the ORC camps were male). The travelling photographers would have needed permission from the camp superintendent to carry out their activities. However, many of them would have been English-speakers, perhaps the source of their 'freedom to roam', as well as of the averted gaze of the little girl shown in Appendix 9. And the photographers were by no means the only visitors: most camps were routinely visited by journalists, 'do-gooders' from abroad, shop-keepers and peddlers, as well as members of the local Relief Committees noted earlier.

2.32 The photographs we have alluded to, the mundane everyday activity of queuing for firewood in Kroonstad and the equally mundane photographs from the other camps, and also our research on the wider number from which these were selected, are absolutely not to be seen as optional extras to our research on the concentration camp records. Inevitably, given their part in establishing and confirming bureaucratic and governmental notions of responsibility and accountability, the written organisational records focus on specific areas of camp life – population size and turn-over in relation to rations-provision, illnesses and hospital attendances, deaths and burials, school attendances and similar matters. Equally so, the extant photographs bear the marks of their ad hoc origins and equally ad hoc survival in the post-war period and concern a diverse array of subjects, from views to persons to peak events like sports days and confirmation classes and including deaths and burials in the camp cemeteries. Each throws interesting and important light upon the others and both are needed to gain some perspective on 'what it was like'.

In Conclusion - A Modest 'Internalist' Approach to Feminist Knowledge Production[62]

While many feminist approaches see the grounding for feminist knowledge-claims as possession of 'the facts', indeed better or best facts, for FFF the grounding lies in 'moral knowledge', that is, accountable knowledge produced by the 'knowing subject'. Rather than macro-level societal change 'out there' and how to change oppressive circumstances by producing better or hegemonic feminist facts, which is the 'externalist' concern of much feminist research, FFF has a modest 'internalist' concern with academic feminism's prime task of crafting knowledge in a feminist form.