Conservation notes: ‘Akosi’ figure from Fugar

‘Akosi’ shrine figure collected by Northcote Thomas in Fugar in present-day Edo State, Nigeria, in 1909.

In her latest blog post from UCL’s Conservation Lab, Carmen Vida discusses how Northcote Thomas’s historical field photographs inform the work of reassembling ‘composite’ objects from the collection and help conservators’ make decisions about appropriate conservation treatments.

It often comes as a bit of a surprise to people when they first get to know about museum conservation to learn that conservators do not necessarily always do everything that can be done to an object, or try to make it complete, new, or ‘like it was’. Out of many possibilities, conservators decide on what is an appropriate treatment for each object in dialogue with experts, curators and other stakeholders. As a conservator, I am very aware that every conservation intervention is a new event in the life of an object, an event that can be ‘life-changing’ – though, hopefully, a change for the better by extending the object’s life and making it more meaningful to others. It is the conservator’s job to ensure that the conservation intervention always fits with and helps reveal what the ‘life’ of the object was and is, and that it never obscures its significance, values and stories, but rather helps to reveal them.

For conservators, damage is not always bad. It can, rather, be an interesting thing: it can, for example, tell us about the way an object was used, help us to understand its ‘biography’, inform us about the conditions in which it has be stored, and so forth. For this reason, conservation always starts with research and investigation. We seek to get to know an object as closely as possible through documentation, through comparison with similar or related objects, and through the signs left on the object by its previous history. This helps us to design conservation treatments that fit with the object’s past history as well as its present and future use. In a way, conservation is a bit of a time machine, moving between the object’s past, present and future!  

As discussed in a previous post, Giving Objects a Voice, many objects appear to be ‘mute’. That is, they have no accompanying information, and conservators must rely entirely on what they can discover from their analysis of the object itself. But working with the collections assembled by Northcote Thomas is providing me with a unique opportunity because the archive itself is so rich and varied: not only objects, but written records, sound recordings and, very importantly for the conservator, historic photographs. These different elements in the archive can sometimes be brought together to shed light on each other. So just as the historic photographs of people have been affording their descendants in West Africa the possibility of reconnecting with their ancestors (see for instance the blog Ancestral Reconnections), the historic photographs of the objects are affording conservators the possibility of reconnecting with the earlier life of some of the objects we are treating. This information is vital to guide our conservation treatment choices because it allows us to compare two different moments in the life of the object, and it helps us decide what the treatment should achieve and how. It ultimately helps us make ethical treatment decisions.

Elements of the figure prior to conservation. NWT 2659. MAA Z 12292, Z 12293, Z 12294.

Some of the objects we have been working with in the UCL Conservation Lab illustrate this well. I have recently been revisiting the treatment of a figure that Thomas collected in Fugar in present-day Edo State, Nigeria, in 1909, which was conserved by one of our students last summer. In Thomas’s catalogue, the figure is labelled with the single word ‘akosi’, with no further information. The object is a ‘composite’ insofar as it consists of several elements and materials: (1) a carved wooden figurine with a feather, (2) a red glass bead ‘necklace’ or ‘bracelet’, and (3) a ‘headdress’ consisting of strings of cowrie shells threaded though cane and plant material. At the University of Cambridge Museum of Archaeology and Anthropology, these three elements have been accessioned and stored separately. Bringing together the [Re:]Entanglements project’s archival research, collections-based research and fieldwork, it has, however, been possible to re-associate the elements that make up this assemblage with reference to a photograph that Thomas made of the figure at the time of collection. This, we assume, shows the assemblage as Thomas initially encountered it in its original context.

Northcote Thomas's photograph of 'akosi' figure, taken in Fugar in 1909.
Northcote Thomas’s field photograph of the ‘akosi’ shrine figure, taken in Fugar, present-day Edo State, Nigeria, in 1909. NWT 1095. MAA P.29204.
Fieldwork in Fugar, examining Thomas's photograph of the 'akosi'.
Discussing Northcote Thomas’s photograph of the ‘akosi’ figure during fieldwork in Fugar, March 2020. Paul Basu notes that ‘Those we spoke with in Fugar did not recognise the type of figure or the word”akosi” recorded by Thomas. In some dialects of the Edo language, however, “akosie” refers to shrine figures moulded from mud. The Fugar carving resembles an “ikenga” or “okega” figure more typically found in Igbo- and Igala-speaking areas of Nigeria, where they are a form of ‘alusi’ (deity). Igala and Igbo influence can be discerned in masquerades and material culture in the area around Fugar. An example of an “okega” from Igala in the Smithsonian collection shares some features with that collected by Thomas in Fugar’. Photograph by Paul Basu.

This ‘akosi’ figure will form part of the forthcoming exhibitions at the SOAS Brunei Gallery and University of Cambridge Museum of Archaeology and Anthropology. As such, one of the conservation treatment aims was to ensure that, after 111 years, all three elements could once again be put back together for display.

'Akosi' assemblage collected by Northcote Thomas in Fugar in 1909.
Elements of the figure prior to conservation annotated to highlight areas of damage (see notes below).

An initial condition assessment of the object revealed several instances of damage:

  1. The wooden figure was covered in surface dirt and dust, and debris from insect activity.
  2. There was a through crack in the left wrist that severed it completely from the rest of the arm.
  3. The left foot had suffered extensive insect damage and was largely missing.
  4. The feather had also suffered extensive insect damage and was dirty, broken and misshapen. There is a corresponding hole on the right side of the figure where presumably a second feather used to be, but this is totally missing now.
  5. The cane and plant fibres in the cowrie shell headdress had become brittle, inflexible and unable to support the weight of the cowries. The headdress was made up of five tendrils. Of these, one was very short and another slightly shorter than the other three. There were also loose cowries bagged separately, indicating that perhaps originally there were only four strands and that when one had broken, the broken end had been reattached at the top.

We had to decide what level of treatment would be appropriate for each of these issues, and we were able to do so thanks to the possibility of referring to the historic photograph.

Annotated photograph of 'akosi' figure, taken in Fugar in 1909.
Annotations on Northcote Thomas’s 1909 photograph of the figure showing existing condition and damage at the time of collection (see notes below).

Close examination of the photograph revealed:

  1. That the surface dirt present was most likely museum dirt, although there were areas of dirt in the original figure the ghost of which could still be seen in the object. For this reason, it was decided to sensitively clean the wooden elements with only dry cleaning methods and to avoid the use of any solvent which may have removed more than simple surface dirt. In this way we could guarantee the cleaning of recent dirt but leave behind any deposits that may be related to the use or beliefs associated with the figure.
  2. The through crack on the wrist can already be seen in the historic photograph. Wood is anisotropic and moves in different directions in reaction to changes in humidity and temperature. Cracks of this type often occur through movement tensions in green wood that has not been properly allowed to dry before being carved. That the crack can already be seen in the historic photograph suggests that the object may have already been in existence for some time before Thomas acquired it. This information clarified that it would be totally inappropriate to fill that crack because it has been there for over 100 years and allowed the wood to move in response to environmental changes without further stresses, but also because a fill would have obscured important information about the object history.
  3. The insect damage caused to the left foot can also be seen in the historic photograph, although perhaps it was not as extensive then. Because of this, only minimal intervention fills were done to support any areas at risk. The fills were done with long fibre Japanese tissue paper (a very thin but strong paper) and a cellulose based adhesive that was sympathetic to the nature of wood. Watercolours were used to tint the Japanese tissue paper to blend the fills with the surrounding wood. The insect damage visible in the historic photograph again seems to indicate that the object was not new when it was acquired by Thomas.
  4. Two feathers are visible in the historic photo, one on either side. The remaining feather was repaired with fills done in the vane to strengthen it and realign it back to its original shape. The feather was also dry surface cleaned and also cleaned with solvents to restore its shape as much as possible.
  5. It was clear from the historic photos that only four strands of similar length were originally present as part of the cowries headdress, confirming that the two shorter strands were originally one and that the loose cowries were probably part of this broken strand. That information, together with the need to strengthen the cane so that it would be able to support the weight, allowed us to take quite an interventive approach: the fourth strand was lengthened with the lose cowries, and all the strands were stabilized by threading them with nylon fishing wire, to support the weight instead of the fragile cane threading. Tinted epoxy buttons were made matching the colour of the cowries to serve as stoppers for each of the cowrie strands. Three nylon lengths were braided to create a stronger wire that was used at the top of the object to connect the nylon fishing wire used on the cowrie strands, and to allow the headdress to sit again on top of the figure during exhibition.
Cowrie 'headdress', part of Aksoi assemblage photographed and collected by Northcote Thomas in Fugar, 1909.
The cowrie ‘headdress’ after conservation treatment. The nylon thread is nearly invisible, but can just be made out, particularly at the top.

The conservation treatment given to this object is a good illustration of the decision making processes we conservators go through as part of our work. Ethical treatment decisions were made in this case because we were seeking to stabilise the object and bring it to the condition that best reflected its values and affordances. This meant different approaches to different areas of the object:  minimal intervention was adopted for most elements whereas the headdress required a far more interventive solution to allow the object to be displayed back together and have its integrity restored.

Giving objects a voice: conservation and the N. W. Thomas collection

UCL Conservation student conducting visual inspect of headdress collected by N. W. Thomas for condition report
UCL students conducting an inspection of objects from the N. W. Thomas collection at the Cambridge University Museum of Archaeology and Anthropology prior to their transfer to the UCL Conservation Lab. Photograph: George Agbo.

Among the many collaborations involved in the [Re:]Entanglements project is a partnership with the Museum Conservation programme at UCL’s Institute of Archaeology. Under the supervision of Drs Dean Sully and Carmen Vida, students on the graduate programme have been working on a selection of objects collected by Northcote Thomas during his anthropological surveys of Nigeria and Sierra Leone. Project coordinator, Carmen Vida, introduces some of the contributions of conservation in this the first in a series of posts from the UCL Conservation team.

Starting in February 2019, as part of the [Re:]Entanglements project, a selection of objects belonging to the N. W. Thomas Collection at the University of Cambridge Museum of Archaeology and Anthropology (MAA) have come to UCL’s Institute of Archaeology for conservation. The project is allowing us to interact with the collection in new ways and to approach, explore and elicit some of the past and present meanings of the archive Thomas put together. As a result, in many cases the objects in the stores are being studied for the first time and the cobwebs are being dusted off, sometimes quite literally.

Conservation has a particular role in the [Re:]Entanglements project: not only does it ensure that the objects are stable enough to be used by anyone interacting with them, but as a distinct form of engagement, it provides intimate knowledge of the objects and can shed light on their make-up and their biography. Due to the lack of contextualising documentary evidence, museum objects often appear to be ‘mute’. But even where there is documentary evidence, the histories and narratives that have reached us provide one side – often someone’s side – of the story. Conservation can help provide a voice to the objects themselves, which may sometimes corroborate, but at other times question, the established histories associated with the objects. In doing so, conservation affords present day audiences the possibility to re-engage with the objects, whilst exploring the way in which the archive came to be both from past and contemporary perspectives.

UCL Conservation student conducting visual inspect of mask collected by N. W. Thomas for condition report
Detailed documentation is essential in conservation practice. Here a student annotates photographs of a mask collected in Ibillo in present-day Edo State, Nigeria prior to conservation treatment. Photograph: George Agbo.

By the time the project is finished more than 40 objects will have been worked on at the Institute of Archaeology’s conservation lab and at the MAA. The objects being treated vary in type and size, and are but a small part of the N. W. Thomas Collection, which includes over 3,000 objects. Some will form part of upcoming exhibitions in London and Cambridge, whilst others are being conserved because, after over a century of existence, they have become unstable even when packaged away in a museum store. The work is being carried out both by professional conservators and students, and is allowing these students to develop their skills and further their training, another affordance of the project. We are also providing running some workshops for the Art Assassins, the youth forum of the South London Gallery, which is another collaboration within the [Re:]Entanglements project.

Fish trap collected by N. W. Thomas being reconstructed in UCL Conservation Lab
A fish trap collected by N. W. Thomas in Awgbu in present-day Anambra State, Nigeria undergoing reconstruction in the UCL Conservation Lab. Made from strips of bamboo, tied with plant fibres, this fragile trap was little more than a pile of sticks prior to conservation. NWT (2) 2, MAA Z 13941. Photograph: Paul Basu.

There is a whole array of tools and techniques that we use in conservation to better understand the objects and inform our treatments. Documentation is essential, starting with any existing historical information to which we can correlate the object itself: documents, photographs, previous museum information in the form of labels or markings on the objects. Labels are interesting things: sometimes informative, sometimes deceptive and indicative of misidentification at some stage. Our conservation work on the N. W. Thomas Collection includes the preservation of labels and museum numbers marked on the object, as they have by now become part of it and of its history.

Historical labels attached to a fish trap collected by N. W. Thomas
Documentation of labels associated with the fish trap being conserved.

Visual analysis of the objects can be extremely revealing for a trained eye. We use microscopes to examine the surface of the objects, looking for evidence of damage and instability, but also for clues as to the history of the object: manufacture materials and techniques, damage and/or repairs (whether before or after collection), evidence of rituals or of use are all things we will be looking for in our work. All these tell stories and give the object a new voice. Some of the objects treated so far exhibit extensive pest damage, and microscopy has helped us find evidence of the pests themselves, which will be later used by expert entomologists to identify the species and give us a good idea of where the damage might have occurred. The insect damage in some of the objects reflects their places of origin in West Africa (termites) as well as their later history (wood boring insects that can be found in the UK).

Horned mask collected by N. W. Thomas exhibiting insect damage
Horned mask, described by Thomas as ofulu mpi, collected in Awgbu, present-day Anambra State, Nigeria in 1911. Conservators use microscopes to inspect the insect damage: top-right: insect carcass; bottom-right: insect eggs. These will be sent to an entomologist for identification. NWT (2) 375, MAA Z 14231.1.

But conservators do not rely on visual inspection alone and we are trained to use most of our senses as part of our work, not just our eyes and hands. Smell, for instance, can help us identify fungal infestation, or certain materials used in the past. And by listening to the sound when we very gently tap a surface, we can ascertain the extent to which it may be hollow under the surface as a result of pest activity. One sense we are definitely not encouraged to use is taste! – not least because of the toxic materials earlier conservators may have used to treat objects.

Other techniques we use help us to identify materials whether original or introduced by repairs done in the past. Ultraviolet fluorescence, for instance, can reveal the presence of modern adhesives and materials in objects. Chemical tests can help identify binders and pigments used in their decoration. Elemental analysis with portable technology such as p-XRF (portable X ray fluorescence, which detects inorganic elements such as metals and minerals) can help us identify not just what the objects are made of, but also, for instance, if they were treated with pest treatments no longer used such as arsenic or mercury salts. And we can use Fourier-transform infrared spectroscopy (FTIR) to identify organic compounds. In this way we were able to establish, for instance, that a white deposit on one of the masks (Z 13728) was paraffin wax. This identification helped us both clean it safely by using the right solvents, but also to speculate that the mask may have received some treatment, possibly soon after arrival at the Museum of Archaeology and Anthropology in Cambridge. Paraffin wax was often used as a filler and protective coating treatment for wooden objects from the end of the 19th century and into the first half of the 20th century. In this case, our findings support documentary evidence from the museum’s Annual Reports in the years shortly after acquisition that the objects were cleaned, mended and restored.

Horned mask collected by N. W. Thomas showing parafin deposits
Mask described by Thomas a onye kulie collected in Nibo, present day Anambra State, Nigeria in 1911. Using FTIR techniques, conservators were able to identify the white deposit evident on parts of the mask as paraffin wax, likely to have been used to treat the mask at the Museum of Archaeology and Anthropology soon after it was accessioned. NWT (2) 428, MAA Z 13728.
Excerpt from Cambridge Report of the Antiquarian Committee, 1913
Excerpt from the Annual Report of the University of Cambridge Antiquarian Committee, 1914. The report notes that much of the museum assistant’s time ‘has been devoted to cleaning, mending and restoring’ objects collected by Thomas, many of which had been damaged in transit from Nigeria. The report also notes that Thomas himself spent a week at the museum ‘classifying and labelling his collections’. This would have been during the period of leave between his third and fourth tours.

Our work so far has also allowed us to interrogate and qualify some of the other knowledge resulting from the [Re:]Entanglements project’s collections-based research. Thomas’s correspondence, for instance, indicates that – in stark contrast to the current practice in his time – he commissioned some of the objects he collected to be made or else bought them at markets. One of the objects conserved at UCL in 2019 is an Igbo maiden spirit mask. The initial visual examination of the mask identified a number of historical repairs, including an iron sheet stabilising one of the bird figure on the proper right of the headdress, an iron staple across a crack on the proper left side of the headdress, and a plant fibre tie on the same area as the staple but further down. In the case of the iron staple and the fibre ties, they are covered with the same pigment used on the rest of the mask, suggesting they are original repairs. Checking the current condition against Thomas’s field photographs showed that the iron sheet repair securing the bird was already there when the mask was photographed by Thomas, presumably at the time of acquisition. These repairs could both be indicative of use prior to collection, in which case the mask would had had a previous life and not been simply commissioned by Thomas, or else be repairs carried out during the process of manufacture, not an uncommon occurrence. Currently we have not been able to resolve the matter of previous use in relation to this object, but nevertheless this conservation work has raised the question, allowing us to continue looking for evidence of previous use in other objects. That is one of the research questions we will be looking to find evidence for when treating other objects.

Maiden spirit mask collected by N. W. Thomas showing historical repair
Maiden spirit mask collected by N. W. Thomas showing historical repair
Maiden spirit mask described by Thomas as isi abogefi collected in either Agukwu-Nri or Nibo, present-day Anambra State, Nigeria, in 1911. Thomas also photographed the mask in the field (top right), revealing that the repair to one of the bird decorations was made prior to acquisition. A close examination of the the iron staple and fibre tie used to repair a split on the proper left of the mask shows that they are covered in the same pigment used on the rest of the mask, suggesting that the repair was done during the making of the mask. NWT (2) 390, MAA Z 13689.

To date, nine objects have received treatment, and conservation of the remaining objects is underway. In this and subsequent conservation-themed blogs we will be sharing some of the stories that are coming to light as a result of the conservation work.

As noted above, many of the objects undergoing conservation and being discussed in this series of posts will be exhibited at the [Re:]Entanglements exhibition at SOAS’s Brunei Gallery, London (October to December 2020) and subsequently at the University of Cambridge Museum of Archaeology and Anthropology (2021, dates to be confirmed).