Whose Information? What Knowledge? Collaborative Work and a Plea for Referenced Collection Databases

Abstract

In this article, I attempt to trace and reflect on how and what knowledge emerged when I discussed historical artifacts with members of the community of origin. The information coming up during our collaborative work sessions does not easily fit into existing standard categories of collection databases. Rather than issues such as protagonists, dates, locations, etc. of photo shoots, or materials, methods of production and uses, etc. of objects, our conversations were about relations between people, life circumstances, uncertainties, legitimate informants, and social meanings. The emergent knowledge, as I argue, did not simply reflect different concerns or perspectives of community members as compared to those of scholars and provided for in database fields but was co-produced from all participants’ prior knowledges through concerted doings. Without presenting a ready-made solution, I further argue that we need to find ways of accommodating references into our collection databases revealing authors and ways of knowledge production.

Introduction

The community concerned are the Khwe, a group of “San,” ”Bushmen,” or “former hunter-gatherers” in Northeastern Namibia. The historical artifacts in question are photos, ethnographic objects, and samples of dried plants currently housed in the Oswin Köhler Archive at the Institute of African Studies of the Goethe University Frankfurt/Germany. The late Professor of African Studies, Dr. Oswin Köhler (1911–1996), compiled those artifacts between 1959 and 1992 during 22 extended field trips to the Okavango in Namibia. The Khwe collection in the Oswin Köhler Archive also contains texts on all kinds of subjects for a vernacular encyclopedia on Khwe culture (Köhler 1989, 1991, 1997, 2018, 2021a, 2021b, 2021c) as well as audio files, films, and drawings. ¹ Together, they make up the world’s largest inventory of documents on Khwe language and culture outside the Khwe communities in Southern Africa. So far, the Oswin Köhler Archive has no online collection database but just EXCEL item-level spreadsheets for items in each individual object class (ethnographic objects, audio-files, photographs, etc.) unable to represent the numerous relations between objects, photos, audio and video files, texts, Khwe terms or concepts, people, places, publication venues, etc. Although this is a deplorable state of affairs in itself, I feel that a data model allowing to embed and requiring references, which reveal the very way how a certain piece of information came to be known, ² is a just as if not more important a desideratum as an online relational database.

In order to truly guarantee onto-epistemological ³ multiplicity, structures, and categories by which members of the communities of origin know cultural artifacts in ethnographic collections should not only involve representative activities such as exhibitions but also touch a collection’s “heart,” that is the permanent information record or catalog (e.g., Krmpotich 2019; Srinivasan et al. 2009; Turner 2015). At the same time, digital databases have been praised for being able to allow for the inclusion of alternative or multiple voices (Boast, Bravo, and Srinivasan 2007), while pre-digital catalogs have been criticized for coming along as authoritative and dominant voices (Macdonald 1998). However, pre-digital catalogs often carried at least traces of knowledge production processes such as individual handwritings, deletions, and additions with different pens on different materials, if only rarely explicit backgrounds of producing and revising entries. In digital catalogs, such traces are lost if not tagged somehow, or even worse, pieces of information from diverse sources are simply collated (e.g., Opp 2008). My point here is that, whether digital or analog, multiple or dominant, entries in catalogs, and digital databases need to be referenced and reveal how the knowledge contained therein was produced.

Digital asset and collection management software, at least the software commonly used for ethnographic and other academic collections in Germany such as easyDB or digiCULT ⁴ , although able to represent relations between items within and across object classes, do not, by default, plan for including references. Different from practices in academic writing, the authorship or forebears of entries in databases, thus, most often remain hidden and the entries themselves appear detached from their originating context. Indicating a date and the name of the person who typed the entry might count as a minimum improvement but authorship, as I understand it, is much more complex and encompasses all aspects and factors making up the process of knowledge production. Among these are the personal and onto-epistemological environments of the producers, their interests and (imagined) audiences, their relations, as well as the diverse back-stories of communicative processes. By describing what this meant in a particular case, I hope to deconstruct the unsound appearance of objectivity still adhering to database entries and demonstrate the need to somehow trace and indicate their originating contexts.

To describe the process of knowledge production requires self-reflexivity or self-inquiry, the particular relevance of which for participatory methodology has often been stressed in the literature (Chilisa 2012; Doudaki and Carpentier 2021) as an attempt to decolonize the research from within (Denzin and Lincoln 2008; Gaudet 2014). According to Bergold and Thomas (2012), such reflexivity should account for personal, social, political, and epistemological reflexivity, and Dahinden, Fischer, and Menet (2021) call for also including the reflexivity of the research participants. Furthermore, users who are searching collection databases, just like visitors of museum displays, will bring their own backgrounds and biases and make their own choices and interpretations. In my description of two collaborative work settings, I adhere to some aspects of these dimensions knowing that it is impossible to account for or even be aware of all aspects and factors contributing to and interacting in such a process (cf. Moser 2010; Rose 1997). In doing so, I acknowledge that it is me who remembers, selects, and stresses the aspects of the collaborative experience, and that this description is entirely non-participative or non-collaborative. That is also the reason why I talk about myself but use initials when talking about my collaborators.

The notion that knowledge is situational (Haraway 1988), produced in and shaped by specific circumstances (Rose 1997) and that there is “no view from nowhere” (Levy 1998: 168) is, of course, not a new insight. Indeed, it is somehow trivial that differently situated people will have different experiences and convictions, as well as that coming to know or becoming familiar with something is an open and never-ending process, which keeps knowledge inchoate or even ephemeral. I am dealing here with two aspects of this notion. The first one is the specific situatedness of joint knowledge production in the collaborative settings in which I participated. Here I argue that the emergent knowledge is not so much a product of negotiations between different and separable pre-existing knowledges of the researcher and her collaborators, but a product of sharing time, space, practices, and ideas, that is, of the collaborative process itself. The second one is the question, on which grounds information gets to be considered duly substantiated, evident, legitimate, relevant, reliable, sufficient, trusted, or valid to become a database entry and how such processes can be captured.

Some Background on My Engagement With the Khwe Materials

I started to work with the Khwe materials in the Oswin Köhler Archive in 2015. The main aim was to complete a vernacular encyclopedia on Khwe culture of which Köhler himself had prepared four out of twelve parts, published in three volumes (Köhler 1989, 1991, 1997) and of which Anne-Maria Fehn and I, with the assistance of native speaker Thaddeus Chedau, have recently edited the missing volumes (Köhler 2018, 2021a, 2021b, 2021c). This was, however, not my first engagement with the Khwe. I had already done research on Khwe material culture and social change during my PhD phase (1998–2003), and occasionally later during a comparative project on Khoisan kinship, and as contract researcher for the Legal Assistance Centre, a Namibian human rights NGO.

From the beginning of my work in the archive, I have been trying out ways of organizing Khwe access to their cultural heritage and of researching it collaboratively with them. This corresponds to demands by Khwe individuals who have repeatedly asked me for access to Köhler’s materials. It also corresponds to my own understanding of ownership of cultural heritage, and the current zeitgeist in anthropology, concerned with collaboration and decolonization of ethnographic collections (e.g., Ames 1999; Clifford 1997; Krmpotich and Peers 2013; Lonetree 2012; Peers and Brown 2003; Philips 2011; Turner 2015) or with indigenizing collections (Zawadski 2018). It also relates to ideas that access and participation generate and stimulate memories, contribute to creating richer cultural heritage, and will sustain or revitalize cultural traditions (e.g., Fienup Riordan 2003; Fluehr-Lobban 2008; Howarth and Knight 2015; Zawadski 2018).

In 2016, I brought the digitized films home to Namibian Khwe communities as a first experimental step. I selected the films for this first effort, because they are few, and were, therefore, comparatively easy to manage. Köhler shot the films in the village of Mutc’iku in 1962 and 1965. I arranged public film shows in schools or community halls in several Khwe villages, and then recorded Khwe comments on the film contents with a work group of six elders, three men, and three women, who had lived in Mutc’iku at the time when Köhler took the films. The final product (Boden 2019 [2016]) is a DVD, fifty five minutes long, with images from the 1960s and Khwe comments from 2016. At the request of Khwe, we produced a second edition with English subtitles in 2019.

The present article describes two further endeavors in collaborative work, one dealing with digitized photographs, and the other with ethnographic objects and plant samples. I took the digitized photos to Namibia, while Khwe came to inspect the ethnographic objects and plant samples in Frankfurt, resulting in quite different work settings. The aim in each case was both to add information on the historical artifacts to the collection record and enable Khwe access to their cultural heritage.

The Artifacts

Original metadata or records from Köhler’s hand on photos, objects, and plant samples (as well as audio recordings and films) are mostly missing in the archive. One of the reasons for this is certainly that Köhler’s main interest lay in the elicitation of linguistic data for a dictionary and grammar, and in the recording of descriptive texts for his vernacular encyclopedia on Khwe culture, ⁵ and that he collected the other materials as illustrations to the text contents in the first place. Another reason might be his idea of a consistent traditional culture, independent of individuals’ subjectivities. In this context, it is remarkable that even early text versions, that is, text versions recorded before the late 1970s, do not bear author information. Why Köhler started to add author information in the course of the text revisions from the late 1970s onwards is not evident from the archival material. Likewise, it is unclear, how he remembered author contributions to earlier versions without written notes. The lack of such notes might also be an indication of his confidence in his memory capacity that died with him before he had the time to document it for posterity. After his retirement in 1977, Köhler devoted himself to completing the encyclopedia and stopped collecting other documents in a systematic way. The majority of photos, objects, and plant samples in the archive were thus collected at a time when their individuality was not yet one of Köhler’s concerns.

The Photographs

There are about 20,000 photographic objects ⁶ in the legacy of Oswin Köhler, mostly 6x6cm mounted slides, as well as a small number of 35-mm-slides, negatives, and paper prints. They show all kinds of contents, namely portraits of people and group pictures, cultural practices and techniques, the scenery of Köhler’s research camp, landscapes, animals, plants, and objects. One challenge and one of the reasons why collaboration with Khwe is indispensable even from a purely archival perspective, is that the photos have virtually no metadata attached to them by Köhler himself. What makes things even more difficult is that the photographs are mostly disordered.

Only rarely the photos in one box or container belong to the same series. Instead, one can find photos from the Khwe together with photos from other African groups and photos from private holidays in the same box. Although the collection had to move between university buildings several times after it had been given to the Institute of African Studies, I do not believe that the contents of boxes were moved in the process, not the least because nobody has worked on the photos. During the digitization process, the original (dis-)order was maintained and can be reconstructed from the file names. In some cases, Köhler seems to have composed picture collages for lectures covering comparative topics such as physical types of people or house types of different ethnic groups. However, I could not (yet?) identify one-to-one correspondences between lecture manuscripts and photo collages.

Only occasionally, a paper slip tells a name, a place, or a year. Often such specifications only refer to a part of the photos in the same container or cannot possibly be right. For example, a photo showing one of Köhler’s main Khwe informants, a man named Kafuro (†1982; cf. Köhler 1989: 33–38), while he is starting to fill the excavated sand into the grave hole after a burial, and which was obviously taken parallel to the respective film session, was found in a box labeled “Uganda/Rest.” Since I have seen the film, I am able to recognize the scene as well as Kafuro by his clothes and body posture, even though he is turning his back to the camera. In other cases, the location where a certain picture was taken is obvious to me, either because I know the surroundings from my own stays in Southern Africa, or from my growing acquaintance with the photos and other materials in the archive.

It is not even clear who shot the individual photos as no mention is ever made of a photographer. Potential candidates are Oswin Köhler himself, his wife Ruth as well as their various visitors, among them people like Köhler’s successor in office as government ethnologist ⁷ Kuno Budack or professional photographer Anneliese Scherz. The latter, together with her husband Ernst Rudolf Scherz, an archeologist known for his work on Namibian rock art, were friends with the Köhler couple and also visited them in their research camp in Dikundu. According to the younger brother of Köhler’s field assistant Ndo, Köhler also used the self-timer or asked a Khwe to operate the camera when he himself wanted to be in the picture. However, he did not let his Khwe collaborators take pictures independently in the same stance as he had some of them do drawings or conduct interviews and note texts (HM, pers. comm. 28.2.2022). The quality of the photos varies strongly from brilliant to overtly poor in terms of exposure, definition, and image details, making me wonder whether Anneliese Scherz made at least some of the brilliant ones. ⁸

The Ethnographic Objects

The number of ethnographic objects amounts to about 500. They include items in the collection categories of clothing, jewellery, dancing gear, household items, containers, tools, hunting equipment, musical instruments, medical tools, toys, raw materials, and miscellaneous. From my own engagement with the material culture of different San groups (Boden 1997), I assume that most of the ethnographic objects in Oswin Köhler’s academic legacy are from the two San groups he worked with most, that is, the Khwe and the !Xun, plus some objects from other African communities. It was, however, and remained so during the collaborative work, not always possible to determine the ethnic group from which Köhler acquired a certain object as the Khwe visitors sometimes felt incapable of telling whether a certain object once was part of their cultural repertoire.

Köhler’s research reports and other archival files contain very few information about the acquisition of the ethnographic objects as is typical for Köhler’s academic generation, yet has changed more recently in the context of debates about ownership, control, and restitution of cultural heritages. Khwe reported that Köhler used to buy objects from people or asked them to make objects for him (cf. Boden 2018). However, in spite of his long-standing personal relationships with Khwe, he did not note who sold or made for him an object. It seems that he considered the objects he collected as representative of object types and replaceable rather than as individual specimens made by individual people, much in the same way as he was assuming cultural consensus when recording the texts for the encyclopedia (OKW 323-1: 5). Köhler’s encyclopedic ambition also with respect to the ethnographic objects speaks from one of his reports, stating that for “completing the photo documentation” he also acquired ”relatively rare objects” or “disposable goods that hardly ever come into the researcher’s sight” (OKW 329-1: 3). ⁹ Future research may still enable the identification of individual producers or former owners.

There are no metadata from Köhler’s hand directly attached to the objects, except for an occasional paper slip with a Khwe term noted in Köhler’s handwriting. Nevertheless, the documentation as a whole contains a lot of information on the objects. The texts in part III of the encyclopedia deal with material culture (Köhler 1997) ¹⁰ and describe how certain types of objects were made and used and by which gender, whether they were adopted from other ethnic groups, and whether they were still in use at the time of text recording or text revision. ¹¹ Recurrent information of this type suggests that Köhler was particularly concerned with such details. In his 1983 report, Köhler states that he “re-discussed and unified” the texts on material culture items recorded in different years up to 1979. As detailed there, he took the following points into account: name or term, whether the term is a loan term, materials, technology, uses, ritual uses, and taboos, known since when or where seen first by informants, who makes an item for whom, who is the owner, still in use or not (OKW 333-1: 1–2). In 1987, Köhler recorded additional information on cultural change based on the question what the Khwe were using before they started to use a certain type of object (OKW 336-1: 2). Köhler’s system of collecting field data, his encyclopedic ambition, his ideas about what information was relevant, and his belief in an originally pristine traditional culture and subsequent cultural loss, are thus inscribed into the collection and the encyclopedia.

The Plant Samples

The academic legacy of Oswin Köhler also contains 163 samples of dried plants only 115 of which had been located before the workshop with the Khwe. The plant parts are mounted with adhesive tape on DIN A4 paper sheets with a cover sheet stapled to it in order to protect them. In 159 cases, Köhler wrote the Khwe names for the plants on the sheets. That is why I consider the plant samples as part of the Khwe cultural heritage although other groups in the area were using the plants as well. For 140 plants, Köhler (1989) also added the scientific names, identified with the aid of plant identification books as well as several botanical experts: John Mortimer of the Forestry Department in Grootfontein, Willi Giess, the curator of the national herbarium in Windhoek, and Hermann Merxmüller, professor for botany at the university in Munich (p. XIX).

As is the case with the objects and most of the images, no further information is attached to the specimens themselves. Again, we find a lot of information in the encyclopedia: on food plants as part of subsistence in the texts for part II (Köhler 1991), on plants as agents for illness and healing in part IX (Köhler 2021b), and on animals’ food plants and plant habitats in part XII about “World and Nature” (Köhler 2021c).

Köhler (1989) started to note local names of plants already in 1959 and 1962 (p. XIX) but the majority of the plant samples were compiled during the field stay in Mutc’iku in 1965. In the entry for the 6^th of June 1965, Köhler’s daughter noted in her diary: “The Bushmen have brought us a large collection of leaves from various trees, shrubs, and grasses. We press the leaves and carefully glue them onto white sheets” (Trester 1986: 35). She also mentions index cards, on which the plant species’ names, characteristics, size, edibility, toxicity, and uses were noted, and of which there are 129 in the archive, filled in in her father’s handwriting.

According to Köhler’s report on the 1965 trip, “the collection of 100 species of leaves and grasses, roots, and tubers served as a starting point for the botanical inquiry” (OKW 323-2: 3). The report on the 1968 trip speaks of the creation of a herbarium and the identification of about 100 !Xun names for trees and plants, a working procedure that also “had a benefical effect on the Khwe dictionary” (OKW 324-1: 8–9). In 1971 a herbarium of 221 plants was created with plant names documented in various !Xun dialects as well as in Mbukushu and Kwangali while “the Khwe names were mostly present” (OKW 325-1: 8–9). In 1982, Köhler sent out Khwe “to collect tubers and medicinal plants that have not yet been determined” (OKW 332-1: 9). So far, I do not have an explanation for the differences in number mentioned in the various reports as compared to the 163 plant samples actually present in the archive.

Work Settings

In January 2018, I took the digitized photographs, 8,000 at the time, to Namibia in order to explore how collaborative work on them could look like. At the time, the only available room with electricity was a tiny church room of the Dutch Reformed Church in Block D of the Mutc’iku resettlement scheme. ¹² By then, it was not in use for services. Instead, it contained heaps of unsorted books and other papers on old school desks and on the floor. Goats were using the room as a night shelter. I engaged people to clean the room and install a lock at the door and bought a prepaid electricity voucher. Even with the furniture pushed to the side, it did not have space for more than four people, including the interpreter and myself. Without any other viable alternative at hand, I thought of the room as sufficient for a first experiment in working with the photographs.

I asked one of the female elders who had been in the film workshop and had impressed me with her clear statements and openness. Her niece, who lives in short walking distance to the church, joined her. In the mornings, I went to fetch the older woman and the interpreter, who both live in other blocks of the resettlement scheme, by car, and installed the chairs, the computer, and the projector. Then we looked at the photos in the disorder I had found them. We usually worked for three hours in the morning, then had a lunch, which I had prepared at my camping place the evening before, and continued for another three hours in the afternoon before packing up, locking the door, and bringing the female elder and the interpreter back home. Occasionally, children or other curious individuals squeezed in and sat down on the floor, but soon left the room because of the heat in the darkened room with the projector operating, or perhaps also because they were not paid. However, usually we stayed among the four of us and worked like this for two weeks, Monday to Saturday.

What originally was meant to be a test, emerged into an established procedure which only ended when we had gone through all of the pictures. I had neither planned to work for such a long time with only two women. Instead, the improvised working procedures developed into expectations on the side of the women not to stop until they had seen all the photos as well as to be served lunch. Because of the logistic constraints to change the work setting, I myself was satisfied with a first complete run in a feasible and sufficiently comfortable setting, expecting that I would have time and opportunity to follow up questions and suggestions that arose during the process. At the time I did not know or expect that, because of the Covid pandemic, I would not be able to do a follow up in the near future and up until today (March 2022).

In September 2019, two Khwe men from Bwabwata National Park came to visit the Oswin Köhler Archive in Frankfurt for three weeks. They were TC, a sixty year old headman and resident of Mutc’iku village on the Okavango River, and SG, the forty five year old president of the Bwabwata Khwe Custodian Committee ¹³ and resident of Chetto village about 100 kilometres further east. With both men I had worked intensely before, during my PhD research and afterwards. Besides discussing issues of access and control of the Khwe documents in the archive, looking at the ethnographic objects and samples of dried plants was one of the main objectives of their trip. ¹⁴

For this purpose, I had a student assistant unpack and put all objects on tables in a room, and the specimens of dried plants in another (they would not have fit in the same room), except for some objects on display in showcases on the corridors of the institute. We worked on the objects together with JB, the coordinator of collections at the Goethe University (seven days), and on the specimens of dried plants with two botanists from the Goethe University, KH and BE, both specialists in African plants (one day). The student assistant joined the discussion and did the photo and video documentation.

The selection of the Khwe representatives to come to Frankfurt had been difficult and conflict-ridden. Initially my budget had allowed for four visitors. Since I had been accused of excluding people when working on the films, I left the decision on the participants of the trip to Frankfurt in the hands of the community. I only demanded that the selected people must have a valid passport for traveling, and that they had shown interest in working on Khwe cultural heritage. I further suggested that the group should be gender-balanced and include people from different villages. The selected people were the two men, who finally came to Frankfurt, and two women: the female elder, who had already been a member of the work groups for the film and photos, and a middle-aged woman from a different “fire place,” that is, another family. When I came to Namibia in March 2019 in order to organize the trip, it turned out that other people were quite unhappy with the selection and tried hard to convince me to make changes to it, not the least by sending the police, the secret service and the Regional Councilor after me. People accused me of wanting to misappropriate money that Oswin Köhler had allegedly provided for the Khwe, and make money out of the information that the female elder would share with me, in particular on the use of medicinal plants. Subsequently, the latter offered to step back and let others travel in her place, and the middle-aged woman said she could not go because, if she did, others would use witchcraft to kill her. After all this trouble, I was even less prepared to invite one of the troublemakers, even though this meant that no woman was part of the team. The lesson to be learned is that delegating the delicate issue of selecting participants to community members does not necessarily make things better or can even cause more damage than taking the responsibility oneself.

Talking About Historical Photos

Recognizing people in photos was a central topic of our discussions. It was a genuine concern of myself for completing the “protagonists” field in the record, and likewise turned out to be a genuine concern of my Khwe collaborators. Their way of identifying people, however, is not easy to accommodate in standard fields of collection databases, in this case name fields, without further ado. Instead of identifying people as individuals by their names, of which Khwe often have several (Boden 2005; Brenzinger 1999), they identified people by their (kin) relations and by specific or extraordinary life circumstances and events. Both strategies, I would argue, conceive of people as situated in a network of people and experiences for which standard fields of collection databases are not equipped. Despite or because of this misfit, the process was definitely mutually beneficial as my Khwe collaborators were able to reminiscence, provide documentation on their own terms, and I received information from their recollections together with a better understanding of their way of identifying people.

Examples for identifying people via their relationships are: “the child of Y’s father’s elder brother,” “M’s wife’s father’s elder brother,” “G’s younger sister,” “my mother’s elder brother’s daughter,” “she calls N távà” ¹⁵ or “he came from [the village of] Bwabwata to marry my mother’s elder sister,” etc. It was obvious that my Khwe collaborators tried to explain the relationships with reference to people known to everybody in the group, including the interpreter and myself, adding, in a way, another layer to their concern with relatedness or, perhaps better, with familiarity. In the examples quoted above, I knew and they knew that I knew Y, M, G, and N, as teacher, headmen, or one of my collaborators in previous fieldwork. With other people present or a researcher with other familiarities within the Khwe community, the statements would certainly have been different. One can also say that our communication took shortcuts depending on the known or expected ground of common knowledge. Not the least, this is evident from the almost annoyed reactions of the two women or the interpreter when I had forgotten about a relationship they had explained to me on a previous day. The names of people in the photos were mentioned only afterwards and possibly merely as a reaction to my concerns. Only when a person in a photo was or had become someone known to be known from previous photos or other contexts, her or his name was mentioned straight away.

My collaborators (and sometimes I as well) recognized people not only by their faces but also by their postures, clothes and jewellery, or the appearances of the homestead where a picture was taken. This shows that my collaborators connected people not only to other people but also to things and places. It further shows that knowledge emerges from a myriad of perceptions, experiences, and conclusions that, in principal, are all part of how people know and, thus, of the reference for a piece of information.

An instance exemplifying my own involvement in the process can serve as a case in point here. One man in a picture, which shows the slaughtering of a leopard, reminded me of a man I know, in fact of the brother of the younger woman and sister’s son of the older woman. The two women considered what I had said and concluded that the man in the picture was, indeed, that man’s late father whose name they said was Bara. I later found the name Bara in the diary of Köhler’s daughter, where she described her own observations of the slaughtering of a leopard in 1965 (Trester 1986: 32–33).

Knowledge production, thus, appears as a process in which ideas solidify, in this case, first, by my familiarity with the people, secondly, the women’s approval of my suggestion and, thirdly, the written testimony of Köhler’s daughter. Entering just the name Bara into the name field for protagonists in that photograph does not at all reflect this process. I would certainly not have entered the name only based on my subjective recognition of a facial similarity. After the approval by the two women, I might have entered it with a question mark, and after reading the entry in the diary, I would have probably removed the question mark. The respective entry would then be based on three types of evidence whose significance derives from the fact that they support one another.

In other cases, however, the information in the archival materials or their original documentation, the knowledge from community members, and a researcher’s knowledge deriving from her familiarity with both the people and the archival materials might either not fit as smoothly as in this example or simply not be available in such a mutually supportive triad way. The general question is then, who, under which circumstances and on which grounds recognizes an information to be reliable enough and worth to become an entry in a collection database and in which form. Revealing the evidence by referencing information in databases will also reveal the inchoate and subjective character or even randomness of knowledge claims.

Another way of describing or positioning people was by narrating details of their life circumstances and events. The women routinely provided the information whether a person in a photo was still alive and, if so, where she or he was currently living. I had asked for this information in some instances with the ulterior motive of being able to confront a person with her picture at a later point in time, and I feel unable to tell whether such information would have been as prominent in our conversations without me having asked for it.

Other talk referred to extraordinary life and death circumstances for which I definitely had not asked explicitly. Examples are, “She died in a car accident in Chetto” or “her firstborn died, then people gave her medicine not to become pregnant again.” Sometimes it also referred to current issues and contestations, for example, “she lived in Golf ¹⁶ at the place that Y is now trying to say is her place.” Sometimes it seemed to be what might be qualified as gossip. Such information was definitely not answering a question of mine, but it was obviously of interest for the people in the group, including the interpreter. Should it then enter the database, and who should decide?

When Khwe living in other villages were in the pictures, I was regularly told that I should go there and ask them as the right people to provide information about themselves or their relatives, raising the issue of who is legitimized to provide information, in the first place. Collaborating with any members of the community of origin does obviously not solve the issue.

Another of my concerns in completing the record was to identify, if not individual people, then at least whether a picture showed a Khwe context or whether my collaborators could identify a place. The people in the work group identified ethnic groups mainly by items of material culture, such as the hunting altar with its characteristic forked branch and old carrying poles or the characteristic dome-shaped mat houses as being Khwe, or the rich jewellery worn by women as !Xun.

The photos also stimulated cultural information that was not visible in the photos themselves but resulted from associations or perceived misfits. For example, in one photo of the series showing the slaughtering of a leopard, a man is sitting in a way that make his raised knees look like a child taken to a woman’s breast. The work group wondered how a woman could be present at the scene as it is strictly taboo for women and children to attend the slaughtering of a leopard. On another picture of the series one could see, that the person was, indeed, a man. This shows how fragile also the local expertise can be, at least if a photo is not properly exposed.

Talking About Ethnographic Objects

Any concerns that the Khwe visitors to the archive would not (have to) say much about the artifacts and envisaged methods to stimulate storytelling turned out to be obsolete as they immediately picked up objects and started to talk about them. Objects could stimulate statements and stories touching on a great number of issues, ranging from past cultural practices, ecological conditions, practical uses, and social roles to current political issues.

“These shoes [ǁàvòo; leather sandals] have given us life. All our parents, my parents have used them. [They protected them from] hot sand, thorns, everything. They did not feel the hot sand in summer time. If my father was chasing eland or giraffe, he used to wear them. These are shoes that I could only see when I was young. My father did not know shops or hospitals. All the food, all the medicine was from the bush. Everything was from the bush: hunting [equipment and prey], clothes, shoes, everything. So it was a good time when people were using this [i.e. the ǁàvòo]. Now, everything is in the government’s hands or government’s power. Now they stop us [from hunting and gathering in the bush]. You can only buy in the shop, and without money, you cannot go to the shop. So, thinking back of them, when I see these shoes, it was a free time, not like now.” (TC, 24.9.2019).

Another example for the wide-ranging meanings of objects that go beyond the scope of standard database categories is the statement “without it, you are not really a man and this might even destroy your marriage” (TC, 24.9.2019). It referred to a piece of python skin, strips of which were used to tie together the individual parts of a góndò (springhare catch stick) with which to arrest a springhare in its burrow or probe a beehive for honey. TC further went on to explain, how to go about in order to make a python disgorge its prey for people’s consumption, and how to kill a python.

As with the photos, conversations also turned to things which were actually not present, because they were made from the same material as or used together with other objects, or because they were found to be missing in the collection in spite of Köhler’s ambition to establish a complete inventory. An example is the container for hunting medicine made from a tortoise shell that is kept at the ǁgóé-ǁgàá (hunting altar, lit. “tortoise-forked branch”). In fact, the Khwe visitors were surprised not to find such an object in the archive and even wondered whether it might exist but be kept somewhere else:

“Köhler is talking a lot about the ǁgóé-ǁgàá [in the texts of the encyclopedia]. Then where is the small tortoise shell? Maybe elsewhere? [. . .] Or did he not take it with him and just wrote about it or what happened? I don’t know. Because to start talking about hunting, you must start with that thing. [. . .] That is why I am asking.” (TC, 24.9.2019)

By no means all objects invoked a story in the Khwe visitors, and, as was the case with people in pictures, the Khwe visitors did not want to talk and did not feel legitimized to talk about things they were not familiar with. On the other hand, legitimacy to talk was sometimes explicitly claimed as was the case with a khuúrú (calabash) about which to speak TC claimed to be legitimized on the grounds that his clan name was Thovo-khuuru, literally “bee’s wax-calabash,” given to his group because of the practice of gluing broken calabashes with wax (cf. Boden 2014b: 61–63).

SG, stated right from the beginning. “To be honest to the group. I am very weak in recognizing the objects. I am younger than him [i.e., TC].” (SG, 24.9.2019). When he nevertheless talked about some of the objects, he often turned to his older colleague with questions for details and confirmation of his statements in a similar way I did when translating the little Khwe I understood to my scholarly colleagues. What was explained, thus, also depended on the (imagined) audience. A lot of information was considered understandable without explanation, such as place names or certain historical or political backgrounds that the two visitors knew I was largely familiar with. At one point, TC stopped himself when complaining about political injustices the Khwe are facing by saying, “Sorry, we are not in the Namibian government. I think like I am talking to them” (TC, 25.9.2019).

The coordinator of collections and I had hoped that we would be able to design one showcase according to the ideas of our Khwe visitors. Instead, we ended up designing a complete exhibition (in the available nine showcases) for which they defined the subjects, selected the objects and composed the texts as the exhibits in the existing showcases did not “have friendship with each other” (TC, 24.9.2019). They had clear ideas about the messages they wanted to convey, came up with very practical solutions when initial ideas were difficult or impossible to realize, for example by substituting an object too big to fit into the available showcases by a picture. Nevertheless, they stressed several times that they would like to have the chance to discuss the design and messages of the exhibition with fellow Khwe, saying that the objects were not their property alone. Therefore, we decided to make the texts and photos of the exhibits into a mobile exhibition on roll-ups to be taken to Namibia and be discussed within the community.

Talking About Plant Samples

Köhler regarded the information on plants, in particular on medicinal plants, as sensitive knowledge. He wrote:

“To address this cardinal area in Khwe life, firm trust between the European and the five or six Bushman informants was the first prerequisite. I believe I may note without exaggeration that this trust existed and that it formed the basis for extensive and important information about Khwe healing knowledge and ultimately about their conception of disease in general. The discussion about the factual relationships of the words sometimes got out of hand to such an extent that the loss of time caused serious concern. However, I remained faithful to the principle of discussing all keywords only in the Khwe language. From the point of view of academic reliability, this procedure seemed to me to be the only viable way, since, according to my experience with the Mbukushu interpreter in 1962, misunderstandings and errors have crept in. In some cases, the Khwe had obviously given false testimony so as not to entrust the Mbukushu with everything of their own.” (OKW 323-2: 2-3).

During the introduction of participants, Khwe headman TC raised a similar concern by referring to the principle of Access and Benefit Sharing as specified in the Nagoya Protocol of 2010. ¹⁷ While Köhler addressed the issue as one on the local level of Khwe fears of Mbukushu domination and exploitation, ¹⁸ the context of TC’s statement was that of Indigenous Peoples’ Rights movements on the global level, in which Southern African San partake since the 1990s when the Working Group of Indigenous Minorities in Southern Africa (WIMSA) was created.

TC voiced worries that outsiders might take advantage of Khwe knowledge on the healing properties of certain plants without having the Khwe benefit from potential profits. In doing so, he certainly also referred to the accusations of exactly such plans in the run-up to the trip in the course of which people referred to the case of the Hoodia plant, used by the San to suppress hunger and exploited by the pharma industry as a weight loss drug. ¹⁹ We agreed that the Khwe visitors would only disclose knowledge on the plants not prone to the risk of exploitation, because either considered unsuitable or because the information was already public. ²⁰ Later, when talking about individual plants, TC stressed time and again that he was not “going deep” into revealing information as others might use it to become rich while the Khwe would remain poor.

The botanists were also interested in other than medicinal uses of the plants and asked several times for such information. They even concluded that plant knowledge was mainly on medicinal uses, while I think that the focus on medicinal uses might well be due to the back-story in Namibia and the concern that medicinal knowledge might be exploited. In any case, the knowledge produced was mainly on the effects of plants or plant substances on people, and occasionally also on animals. The effects were bad and good ones, on health, and fertility as well as success in hunting or love matters. To categorize these effects as medicinal, magical, or otherwise would mean to impose academic categories on them. The word in English used by the Khwe participants was nevertheless “medicine.” The term does not grasp the full meaning of the Khwe term tcóò which can be translated as “medicine,” “poison,” “charged with power,” “taboo,” “impure,” or “inedible” (Kilian-Hatz 2003; Köhler 2021a). One should have this in mind when interpreting the following statement: “I don’t think there are plants or trees without medicine. Even those which can be eaten are ‘medicine’.” (TC, 26.9.2019).

As with the photos and ethnographic objects, it turned out that being a member of the community of origin is not tantamount to being a knowledgeable expert. SG frankly stated that he recognized the plants only from the Khwe names written on the paper backings by Oswin Köhler. With very few exceptions, he left the explanations to TC, even when the latter wanted him to take over for a change or break for himself. Instead, SG confined himself to help with English vocabulary, asked questions himself, partly referring to information he remembered to have heard. During the round-up session, he stated that the workshop had been a great opportunity for him to learn about the plants and that he had only known some of the plants as food plants, but not as “medicine,” before. ²¹

Though not feeling expert in plant knowledge but certainly in Khwe orthography, SG immediately set about to correct the spelling of one of the plant names as written by Köhler. The conservator soul in me trained to defend the “original document” cringed and set off to forbid this intervention but realized a second later that this was, indeed, a desirable appropriation of the collection by a member of the community of origin. SG did not repeat the task and I cannot tell whether this was because he did not find further spelling errors or whether he had noticed my initial fright. I left it like that and actually forgot about the incident until listening to the recordings much later, but the question how to go about such interventions should certainly be on the agenda for future work.

I had given the plant samples consecutive numbers in the (dis-)order I had found them in their boxes. The Khwe visitors did not question the order of the plants while the botanist colleagues asked me how Köhler or I had sorted them. I had to answer, that, for me, a potential botanical grouping would not have been evident anyway and that I had recognized no other grouping either. When asking the Khwe for how they would order the plants, the answer was that all plants stay in the bush and that, therefore, they did not know how to put them in order since they would go and look for what they needed and not in order to collect and keep them. The lack of order was, so-to-speak the perfect bush simulation.

I questioned this, arguing that they would not just go anywhere to look for a certain plant but that they knew which plants were growing in which habitat. The answer was that they would go to a place where they had seen the plant before, that is, at a time when they did not need it. It foregrounded experience and familiarity with the area instead of abstracted knowledge of habitats. Such abstracted or categorical knowledge can nevertheless be queried as obvious from the texts in Köhler (2021c) on “World and Nature” or come forward otherwise as, for example, when I identified one plant as “grass,” which, according to botanical criteria is not a proper grass (German: Süßgras) but a sedge (German: Sauergras). Even though that conversation had been in German, TC confirmed that it was not a grass for the Khwe either as grasses have raised nodes, which the very plant did not. The raised nodes of proper grasses are also an identification feature for botanists, showing that people with different knowledge backgrounds can also share criteria and categories.

While the researchers agreed that the focus of the collection had most probably been on medicinal plants, the Khwe refrained from such speculations. Instead, they said that it was not them to tell as they did not know how and why the people who collected the plants decided as they did. This statement might count as an expression of their awareness of the subjectivity and the context-dependent or situational character of all knowledge, also stated by other indigenous researchers (Kovach 2009; Zawadski 2018).

When asking whether and which plant samples should be part of the exhibition, I had a selection in mind, even more so after the initial statement that only the information on some plants would be unproblematic to share publicly. However, both men argued that all plant samples were important to them as they had the Khwe names on them, that is, on the paper backings, and because the Khwe were not seeing these plants these days and thus would at least be able to know them by their names. SG then suggested that we should prepare a book including the photos of all plant samples together with their Khwe names in order to be able to show them to young people and do research on the plants with local experts.

Whose Information? What Knowledge?

The creation of catalogs or databases for ethnographic collections and the practice of cataloguing relate to methods of authoritative knowledge production and control. Why certain information about objects is recorded and made prominent while other information is not, depends on the specific context of documentation (Turner 2016). In this article, I have disclosed how knowledge in relation to Khwe historical artifacts in the Oswin Köhler archive was produced in particular collaborative work sessions rendering the questions brought up in the title quite complex, to say the least. If information is the outcome of concerted doings of particular researchers and selected community members in a collaborative situation, as whose contribution should it then enter a collection database, and how can entries in a collection database reflect the situational character of knowledge?

Since knowledge is situationally emergent and incomplete by nature, it is only logical to provide references as to how pieces of information came to be known. This should also apply to databases, which, more than academic writings, still have an aura of containing objective and firm data, not the least, as I think, because a simple and unreferenced piece of information in a database hides the delicate process of its production as well as its incomplete and instable nature. References, even in their most basic form, indicating just the name(s) of authors and a date of creation, would be an improvement in shaking the firmness of database entries since a reference always implies subjectivity and possibilities for alternative knowledges. However, as shown above, authorship goes far beyond such basic information and encompasses all aspects and factors making up the process of knowledge production and includes diverse back-stories, the relations between participants as well as their prior knowledges and (sometimes strategic) ignorances.

Collection and digital asset management software allows for creating relational databases which could also contain a sub-database on authorship. Such a sub-database could specify the details behind the reference. Beside the participants and date of the underlying communicative process it could further contain more detailed information and reflections on the knowledge production process. Ideally all participants in the process should either agree on the phrasing or have a chance to add possibly deviant voices. Such references would enable and oblige users, who take entries as basis for further interpretations, for example in cultural comparison, to track and reflect the significance of the very data. As users become part of related knowledge production processes themselves, they need to be aware of the boundaries and contingencies of the knowledges they draw from.

While the authorship of a database entry is already a complex issue, its authority is even more so, involving at least two questions, first, which knowledge is considered correct, reliable, or valid by whom and on which grounds, and secondly, who considers whom legitimized and responsible for the provision as well as for the use of information.

Speaking in very general terms, one might say that legitimacy and expertise rely on familiarity and experience with a subject matter and are attributed to both academics and community members, if relying to varying extents on different types of experiences such as teachings, observations, participations, lived experiences, sensations, writings, etc. Therefore, the information on authorship should give users also an idea of the subjective expertise of those who created the information behind a database entry.

The identification of the Khwe man Kafuro on a photo in the box labeled “Uganda/Rest” (see above) may serve as an example to illustrate my concern here. I am convinced that the man in the photo who is only to be seen from behind is Kafuro, due to my familiarity with other materials in the archive, in particular the film showing the same event. The information as such is, however, not necessarily replicable for someone else from looking at the photo alone, not even for a young member of Kafuro’s own family, who is not familiar with his stature and clothes at the time. She would have to rely on and trust me or another “expert,” unless I provide evidence for how I know that the man on the picture is, indeed, Kafuro. ²² By doing so, I would expose my responsibility for this interpretation and render my authority testable. In this case, the linkage is, of course, rather simple, and made up of just two sources of knowledge, that is, the film sequence and the photo taken at the same occasion. As soon as the young member of Kafuro’s family watches the film or asks an elder who has seen Kafuro in these clothes, she would probably trust that the entry is valid. In this case, scholarly expertise, based on familiarity with the archival material, and local expertise, based on personal experience, would probably match.

Another example is the identification of the man Bara in the picture series showing the slaughtering of a leopard (see above). For me, my personal perception of a recognized facial similarity, confirmed by Bara’s aunt and sister, and solidified by the entry in the diary of Köhler’s daughter were enough evidence to enter Bara’s name as a protagonist into the respective entry field. I have to admit, though, that first, I was only convinced when reading the diary entry, and, secondly, that I did not even think of asking my Khwe collaborators what they thought about the evidence until writing these lines. This is not because I wanted to ignore them but because it did not occur to me that it might be different for them. That is to say, that even entering the three pieces of evidence does not reveal what convinced them or how they know.

In other cases, academic knowledge and community knowledge may be downright inconsistent. An example for the latter actually occurred during the workshop in Frankfurt when TC stated that the Khwe had not been using iron in the past and that they only got iron when they met with White people (TC, 24.9.2019). This statement contradicts Khwe texts recorded by Köhler (1989: 383ff., 1997: 257ff.; for an archeological voice, see also Kose 2016) according to which the Khwe learned about iron and forging from their Bantu neighbors. Reading the texts as documented by Köhler would probably convince contemporary Khwe. At least, I think so because also in other cases I have experienced that Khwe think of the words of Köhler’s Khwe sources as being more reliable than even their own knowledge when speaking about the past, although Köhler’s sources definitely had their own biases (Boden 2014a, 2018) as has everyone else.

As it is impossible to escape the subjectivities and biases of any knowledge, no matter whether academic or local, and no matter how meticulously efforts at triangulation might be, it is necessary to reflect on the processes of knowledge production, to provide references and thereby take responsibility for one’s perspectives and writings (cf. Zawadski 2018). This is true no matter whether “experts” are academics or community members. For outsider users, any community member might be a better expert than she herself but insider users or community members will have more nuanced perspectives on this as well as other ways of tracing and understanding the origins of a piece of information, but only if it is referenced.

The critique of the academic dealings with ethnographic collections (and other data) has often been in terms of “proprietary rights” over “superior knowledge” as a “type of vestigial colonialism” while the critique of letting communities of origin take representative power holds that political correctness takes precedence over intellectual integrity (Ames 1999; Zimmerman 2010). Academic authority often comes along not only as objective but also anonymous or institutional where the agency and agenda of individuals remain hidden. Revealing the source of information also involves standing by academic contingency and uncertainty. Likewise, local knowledge can only be ephemeral since local people also always have new experiences and insights (Willerslev 2016). ²³ All of these considerations are certainly no new insights. I name them in order to stress the necessity for references in collection databases.

Anthropological or ethnographic knowledge has always been situational, co-produced, and relational as it cannot but emerge or grow from communicative processes and relations with others (e.g., Ingold 2014; White and Strohm 2014). The outcome of every observation, interview, or group discussion is a situation-dependent co-construction on the part of the participants. Situations are part of much larger lines of communicative processes with respect to time, place, and people since participants do not come to the situation as blank sheets but with their experiences, back-stories, intentions as well as prior knowledges and ignorances. A greater effort to acknowledge the co-authorship of indigenous participants in these processes is increasingly observable (e.g., Lekgoathi 2009; Sanjek 1993; Vawda 2016). This goes hand in hand with the developing insight, that scholars need such collaborators to make sense of the ethnographic collections in academic institutions as well as an intensified move toward granting them authority over the research processes and its results. ²⁴

The power dynamics between academic and non-academic partners in participatory knowledge production have become central to the effect that “one cannot engage in a discussion about participation, while disregarding the concept of power” (Doudaki and Carpentier 2021: 3). However, thinking with “power” means to take it for granted, acknowledge and act in it, just as researching and thinking with gender is “doing gender” as McDowell (1992) has put it. What, if we could, indeed, get beyond the concept of power as both, (negative) imposition and (positive) empowerment, yet always antagonistic, and, instead, think with the concept of “responsible ownership” where knowledge, authority, and legitimacy derive from being concerned about, being familiar with, and being involved in instead from a position within a power hierarchy? What if, instead of seeing participation as equalization of power relations to be achieved, we start by taking equality for granted while acknowledging historically specific contexts, which structure, constrain, and promote the doings of all partners in the collaborative process in particular ways? What if we take for granted or routinely think with responsibility and care, and commit ourselves to those values as persons who act with and relate to each other instead of more or less powerful position holders distanced in terms of race, nationality, age, gender, or social and economic status influencing the “data”? (cf. Rose 1997).

Not the least, I advance this as one of the lessons I learned from my Khwe partners. The Khwe term díxà with the respective person-gender-number-suffix denotes people who are concerned by or involved in something and, therefore, the ones to be addressed and legitimate to speak. In the Khwe dictionary, the meaning of díxà is given as both “possessor, owner” and “self” (Kilian-Hatz 2003: 36). The díxà-ǁ’áé-na (“owners” of the homestead) are the household or village heads who are responsible and care for their people and, therefore, are the ones to be addressed in all issues relating to their homestead. Likewise, the “owners” of a story are the people involved in an affair. Already during my previous research, Khwe used to refer me to the responsible and legitimate “owner” of information on personal issues (Boden 2005). While talking about the photos and artifacts, the Khwe participants also often referred me to the “owners,” that is, the people concerned. They were even unwilling to make suggestions in their place as was the case also when I asked the two visitors to Frankfurt to speculate about the selection of plant samples. Rather than notions of power and hierarchy, the term, thus, carries notions of affectedness and responsibility, from which (subjective) knowledge and legitimacy to speak derive.

If Khwe feel that one should always ask the people concerned, then one should certainly ask Khwe about historical artifacts from their community. One should also try to find out who is in the pictures, who made an object, who gave it to Köhler, who collected the plant samples, or, at the very least, who still knows how to make or use objects and plants in order to be able to talk with the legitimate and responsible people. Although much more complex and costly, this would also be an answer to the methodological question as to which persons, or groups of persons, should, or must, be involved when research is conducted together with the affected people (Bergold and Thomas 2012). In other words, relying on the utterances of some Khwe might be better than not involving community members at all but is not enough. Srinivasan et al. (2009) call for a debate within collecting institutions about how they can enable and support uses of databases in the sense of creating alternative data by members of source communities. Just like the data entries by academics, these will be subjective and context-dependent and should be referenced and reflected upon as a matter of responsibly taking ownership.

A last thought concerns the laurels of participatory or collaborative research as radical praxis through which marginalized people become aware of their expertise and acquire research capabilities that they use to transform their own lives (Hurtig 2008; Park 1993). SG’s statement in our evaluation session, “I learned a lot. Because before I did not know how to archive those things but now, I learned how to put them in an archive. Maybe in the future I will be an expert” (SG, 10.10.2019), can be read as a success story in this sense, while, on the other hand, one could also argue that such training erodes what could have been a distinctive Khwe point of view (cf. Jones 1970). I think, instead, that collaborating means becoming familiar with each other’s practices and learning from one another and that all participants’ knowledge cannot but be modified in the process. SG also did further research when back home in Namibia, namely on animal figures from wax, of which the collection has several and which he did not know at the time. Later he discovered that Khwe children still make and use them. He then documented this by means of pictures, which he sent to me. His idea to further work with local experts on the photos of the plants is another point in case. I would argue, however, that this is as much if not more due to his researcher personality in combination with his biography as one of the cultural brokers of his community, engaged and influenced by NGOs, churches and other foreigners, as it is an outcome of our collaborative practice.

Summarizing Conclusions

By describing collaborative work sessions with Khwe on historical artifacts of their cultural heritage, I have demonstrated the contingent and dynamic character of knowledge production in which knowledge emerged from sharing and hiding aspects of prior knowledges and ignorances, making conscious and unconscious choices, from relations and reservations between participants, as well as improvised or self-organized practices. All participants as often as not turned out to be ignorant and amateurs, instead of knowledgeable and experts.

Instead of separable knowledges of local experts, who know image contents and objects as embedded in their local cultural practices, and scholarly experts, who know image contents and objects as representative of a wider culture and embedded in scholarly practices of categorizing and ordering (Srinivasan et al. 2010), knowledge emerged situationally from shared interaction (Edelenbos, van Buuren, and van Schie 2011). Collection databases containing such contingent knowledge, yet mined for “objective” data stripped off from their originating context, like other outcomes of knowledge production, need to include references that allow tracing how the data were produced and by whom, if we subscribe to take responsibility for entries in databases just as for our writings. I advocate for finding ways to include such references into collection databases even though I cannot offer a ready-made solution.