Navigation – Plan du site

AccueilThématiques16Reopening the archives of archaeo...

Reopening the archives of archaeology at Ile-Ife, Nigeria: Towards a history of the Osangangan Obamakin sacred grove

Rouvrir les archives de l’archéologie à Ile-Ife, Nigéria : vers une histoire du bosquet sacré d’Osangangan Obamakin
Tomos Llywelyn Evans

Résumés

En 1953, les archéologues Bernard Fagg et A.J.H. Goodwin ont fouillé des sites de bosquets sacrés à Ile-Ife, dans le sud-ouest du Nigeria, dans l'espoir de découvrir des preuves matérielles et stratigraphiques de l'histoire ancienne de la ville. Ils s'intéressaient particulièrement à l'élaboration de typologies et de chronologies des artefacts, dans le cadre de la systématisation de l'archéologie en Afrique pendant cette période. Mais loin d'être des lieux primordiaux conservant des archives intactes des pratiques humaines passées, ces bosquets sacrés étaient des espaces dynamiques avec des histoires complexes, où les communautés successives ont remodelé et réinterprété à plusieurs reprises les caractéristiques culturelles antérieures, ainsi que les paysages environnants. Cet article revisite les fouilles du bosquet d'Osangangan Obamakin, en s'appuyant sur les archives largement inédites pour poser de nouvelles questions sur le passé ancien et récent du bosquet, ainsi que sur le sens de son archéologie pour comprendre l'histoire plus large des paysages culturels d'Ife.

Haut de page

Texte intégral

Introduction: The elusive 1953 excavations at Ile-Ife

  • 1 F. Willett, 2004, I.2.
  • 2 A.T. Aderemi, 1937.
  • 3 Ciroma’s contributions as an archaeologist in Nigeria were especially important, he going on to wor (...)
  • 4 F. Willett, 2004, I.2; S.P. Blier, 2015.
  • 5 T.L. Evans, 2024a; 2024b.
  • 6 A.B. Babalola et al., 2018; G. Chouin, A. Ogunfolakan, 2018; A. Ogundiran, 2020; T.L. Evans, 2023.

1In 1953, in the sacred Yorùbá town of Ile-Ife, an ambitious, large-scale archaeological project was embarked upon, organised and directed by Bernard Fagg, Nigeria’s assistant surveyor of antiquities.1 Fagg was joined by the influential South African archaeologist A.J.H. Goodwin from the University of Cape Town in South Africa, and his brother William Fagg, the curator and art historian from the British Museum. Their work was facilitated by Ooni Adesoji Aderemi, the traditional ruler of Ife, who was active in conserving its cultural heritage.2 Furthermore, several Nigerian archaeologists, recently trained by B. Fagg at the newly founded Jos Museum, and well-diggers from Northern Nigeria, were recruited as part of the team. Notable among these were Liman Ciroma, Musa Gombe, Edward Thaddeus, and Haruna Rashid, who played vital, but largely unrecognised, roles in this work.3 The project was partly a response to the many chance artefactual finds made in Ife throughout the early and mid-20th century, often the result of urban development,4 as well as to the steady expropriation of its known “antiquities” by Western interests.5 Most of these finds derived from the era of Ife’s zenith as a major power in the region (c. 11th to 15th century AD) when industries of glass, copper alloy, and terracotta art flourished.6

  • 7 F. Willett, 2004, I.2.
  • 8 C.A. Folorunso, 2011; F. Kense, 1990; A. Ogundiran, 2015.
  • 9 G. Chouin, 2008; A. Ogundiran, A. Ogunfolakan, 2017; M.J. Sheridan, C. Nyamweru, 2008.

2The archaeologists sought to uncover “ancient works of art” in their archaeological context for the new museum being developed in Ife, obtain evidence of copper alloy casting, and understand the depth and nature of the occupation deposits.7 It was an important moment in the advent of “scientific” archaeology in West Africa,8 involving the implementation of rigorous survey and excavation techniques and the (intended) development of standardised typologies and chronologies. These archaeologists, seeking out both undisturbed stratigraphy from the early city and reburied examples of early arts by later communities, largely targeted Ife’s sacred groves for excavation. These places were dedicated to important deities (orisa/ebora) and ancestral figures, where old objects were often stored and curated, and in the words of the Annual Report of the Antiquities Service for the Year 1952-53 (p.8), locations “where ancient works of art are thought to have been buried”. Many of these places were protected from the urbanisation that had swept through other areas of the city, and so were considered to be some of the only possible sources of undisturbed stratigraphy from the early periods of Ife’s history. West African sacred groves have since been shown to be historically dynamic places rather than static, primordial reservoirs of traditional culture,9 and the archaeologists would soon learn the extent to which disturbance, rediscovery, and reuse of early deposits was the norm here. Indeed, although Ife’s sacred groves are often associated with old archaeological features, their relationships with these are not straightforward. Discerning the nature of the “chicken and egg” dimensions of what features and concepts came first versus what developed later, when and how these places came to be sacralised, and their evolving uses and meanings, can be complex and challenging.

  • 10 Four archives provided useful material, including the three mentioned in the body text. Some additi (...)
  • 11 Yorùbá orthography: Ọ̀sángangan Ọbámákin (R. Abiodun, 2014).

3Although these concerns were not at the forefront of B. Fagg and Goodwin’s research aims, revisiting their archaeological work at sacred groves, very little of which was ever published, can offer insights into new questions about their histories, significances, and relationships with the larger historical landscape. This article explores unpublished materials from this field season10 to reconstruct the excavations of Osangangan Obamakin,11 an important sacred grove site in Ile-Ife associated both with features of the early city as well as later “secondary contexts” as the space was continually used and reinterpreted (Fig. 1). The article draws upon the unpublished diaries, reports, and photographs of B. Fagg, Goodwin, and W. Fagg to reconstruct the grove’s archaeology. This includes four documents written by B. Fagg and A.J.H. Goodwin, now in Frank Willett’s archive at the University of Glasgow, and an unpublished book manuscript (written like a diary) by A.J.H. Goodwin in several parts recounting his time in Ife in 1953, located at his archive at the University of Cape Town:

4A) Papers of Professor Frank Willett, University of Glasgow Archive Services, Glasgow (UG), GB248 ACCN 3120, Box 1:

  • B. Fagg, Ife 1953 Season BEBF’s Diary.

  • B. Fagg, BEBF 1953 Notes, ‘Some Archaeological Problems at Ife’.

  • B. Fagg, BEBF 1953 Notes, Other.

  • J. H. Goodwin, File No. 27/A/S.1 Ife Antiquities, 1953.

5B) The Goodwin Papers, University of Cape Town Libraries: Special Collections, Cape Town (UCT), BC 290, J2.23 Unpublished book manuscript on Goodwin’s Nigerian expedition. Goodwin’s papers primarily pertain to his work in southern Africa, with only a small portion deriving from his work in Nigeria, which occurred towards the end of his life. This archive survived the wildfire that impacted the university in 2021.

6References to these collections are abbreviated for convenience later in the article as UG, ACCN 3120/1, and UCT, J2.23 respectively. Additional photographs (taken by W. Fagg during this field season in 1953) were obtained from the William Buller Fagg Collection, Royal Anthropological Institute of Great Britain and Ireland, London (RAI). These sources complement each other nicely, with Goodwin’s more expressive and introspective writing, and W. Fagg’s photographs offering an improved understanding of B. Fagg’s more straightforward and detailed record-keeping, as well as additional details.

7By integrating and synthesising this wide range of material, the article both lays out the specific “data” of these excavations (recorded artefacts, stratigraphy, features, etc.) in a way that, it is hoped, will be useful for other researchers, while also addressing a new set of questions. While the archaeologists of 1953 were largely interested in questions of typology and in “salvaging” artefacts, this article instead explores:

  1. What can be reconstructed of the evolving history of the sacred grove.

  2. The ways in which it came to be used and understood over time by successive generations.

  3. The relationship of the grove with the larger landscape of Ile-Ife, especially its system of earthen walls, and what this informs us about their uses and meanings.

Figure 1: Reconstruction of the walls of Ile-Ife

Figure 1: Reconstruction of the walls of Ile-Ife

The map shows the location of Osangangan Obamakin (marked in red) and other sacred grove sites (marked in black) that are in close proximity to the town walls (shown as brown lines and dotted lines where inferred).

Based on F. Willett (1967), P. Ozanne (1969), and D. Simmonds (2002) in F. Willett (2004). Tomos Llywelyn Evans, 2025.

Published work on the Osangangan Obamakin excavations

  • 12 Throughout this article, where possible, I identify finds from Osangangan Obamakin with Frank Wille (...)
  • 13 F. Willett, 2004, II.5.
  • 14 This is contained in a document entitled LIST OF PLACES IN IFE KNOWN OR BELIEVED TO CONTAIN ANTIQUI (...)

8Most of the information on the excavations at Osangangan Obamakin is restricted, confined to a small number of very short government reports. F. Willett’s (2004) comprehensive catalogue and database, The Art of Ife, provides a useful synthesis of some of this material.12 In his discussion of Osangangan Obamakin,13 Willett draws upon four main sources: a short unpublished report by Kenneth Murray of his brief reconnaissance and excavation there,14 The Annual Report of the Antiquities Service for the Year 1952–53 summarising the 1953 excavation, B. Fagg’s unpublished diary and notes from 1952–3, and a short description of excavations by Goodwin obtained from the files of the Department of Antiquities (A.J.H. Goodwin, “W. Nigeria Jan. Feb. 1953”, notes on the season’s fieldwork in the files of the Department of Antiquities, cited in Willett (2004: II.5)). The main information deriving from these sources is outlined below.

  • 15 This is presumably a reference to the Ibadan Road, north of the grove, or, alternatively, to a road (...)
  • 16 As discussed below, the site was close to a river, apparently separating Ife proper from Modakeke. (...)
  • 17 Such deep shafts are common in the archaeology of early Ife, seen at a number of sites, and probabl (...)

9In 1943, Kenneth Murray collected information on the site’s oral traditions and landscape, noting that it was associated with an early Ooni – Osangangan Obamakin – and located at a thickly wooded area a quarter of a mile from the road.15 The grove was surrounded by large peregun (Dracaena fragrans) trees, located “behind an ancient abandoned market” with “signs that it was once surrounded by a wall”. There was an upturned pot marking a shrine to Esu, an important liminal deity, near the grove’s entrance, and an overgrown pool not far from this.16 Within the grove, Murray observed a shrine accumulation (“a heap of small offering vases”) and some distance to the right of this was a spot that Murray identified as the main shrine, which he stated appeared to have been in a small (now collapsed) hut, about 2 yards by 2 yards in its dimensions (approx. 1.83m x 1.83m). Murray, in collaboration with the Ooni, removed several items from this: a large, upturned pot which covered several fragments of terracotta and stone, a stone carving of a pot, a decorated pot (FW T243), and several other pots and “offering vases”. The fragments included a small terracotta head (FW T236), probably broken off a body, pieces of terracotta figurines, and two small stone carvings (FW S43, S44). Murray undertook a small, shallow excavation at this feature, running from outside the shrine to the spot where the upturned pot had been. He found two terracottas: a “broken face” (FW T237) and “head of a dog” (FW T242), and at around 14″ deep (approx. 36cm), small fragments of pot that appeared to be of recent manufacture. Here, he also excavated through a “sill” – a feature of overlapping large pieces of pot towards the shrine’s edge – and some further fragments and a possible hole towards the centre of it, as the soil here became soft and yielding compared with the harder deposit around it. The priest of the grove had told Murray of a deep hole here.17 Murray thus interpreted some of the material excavated as thrown in to fill the hole. Few of the excavated terracotta finds fitted together, although the body, bit of leg, feet, and swollen testicles of a squatting figure were successfully refitted (FW T239).

  • 18 Unpublished evidence suggests that this feature was associated with Osangangan, only peripherally w (...)

10Descriptions of the 1953 excavations of B. Fagg and Goodwin outline their work at three shrines at the site. One of these was the first accumulation observed by Murray, which is referred to as ‘Esu Ejubo’ (Esu ojubo, “Esu shrine”).18 They also excavated the main shrine, as well as a third, smaller shrine to the east identified with Ogun, deity of iron. The “Esu” and “Ogun” shrines were “found to be superficial and of recent age, consisting largely of votive pots heaped into low mounds”. Their excavation of the “main shrine” not far from where Murray had dug revealed further terracotta finds, including a fine head (FW T255) near a complete food bowl, though as Murray had found, most of these were disordered pieces. Beneath this was a hardened gritty layer that they had assumed was subsoil, though below it they excavated a deep shaft – presumably the deep hole that Murray was told about – to a depth of almost 14 feet (approx. 4.3m). This contained older pottery (some finely decorated) and charcoal. The hole feature was interpreted as older than, rather than coeval with, the shrine above, and a possible burial chamber, though no bones were uncovered.

11Willett, drawing from sections of B. Fagg’s unpublished diary and notes, outlined the basic principles of the excavation grid system that the archaeologists established at the site. This additional information from Fagg’s diary has been invaluable as a means of reconstructing the spatial relationships of the features and excavations, and the grid system is described and illustrated in more detail below. Willett mentioned other unreported material from Fagg’s diary, such as basic information on the contents of the “Esu Ejubo” and Ogun shrines. Goodwin’s report also mentioned that the main shrine structure had likely been round, and he suggested that Murray’s “sill” of large pottery may have been a platform within it for sacred objects. A further published piece with limited, but significant, information on the excavations not cited by Willett is an article by Goodwin (1958):

  • 19 A.J.H. Goodwin, 1958, p. 46.

Far more impressive than the Ife city walls is the extraordinary pavement that at one time floored the entire city. I came across this in a broken condition at the overgrown Osongangan Obamakin shrine at one of the city gates, but without in the least appreciating its amazing significance.19

  • 20 A.J.H. Goodwin, 1958, p. 47.

At Osongangan Obamakin the paving has been overlaid by a foot or two of soil; roots have disturbed the floor, as this is now overgrown into a forest area, but the presence of the paving is readily detected from the vast numbers of “two-shilling pieces” or one-inch discs in the soil.20

  • 21 Later authors (e.g., P. Ozanne, 1969; F. Willett, 1967, 2004) do not represent this area as a gatew (...)
  • 22 P. Ozanne, 1969.
  • 23 See G. Chouin, A. Ogunfolakan, 2015, 2018 for the archaeological evidence. Making reference to oral (...)
  • 24 S.P. Blier, 2015, p. 20, 172.

12Thus, Goodwin’s article reveals that they may have uncovered a potsherd pavement at the site, and that they considered it to be a shrine associated with one of the gateways in Ife’s walls.21 Overall, these archaeologists interpreted the site as a place where old terracotta material – much of it perhaps found broken and perhaps from diverse sources given that most of it could not be refitted – was brought during comparatively recent periods, deposited on top of old features such as the “pavement” (in Goodwin’s words) and shaft. Despite the usefulness of Willett’s synthesis, his treatment of Fagg’s unpublished material is brief, and he made little use of Goodwin’s writings, having not utilised information from his papers at the University of Cape Town. Willett’s synthesis also lacks maps, sketches, or photographs of the site, thus missing important contextual detail. Partly due to this, Willett deals little with the archaeologists’ observations of the fortified nature of the grove, and its close association with the walls of Ile-Ife. Regarding these, although Ozanne suggested that a multi-phase construction between the “Medieval” period and the 19th century occurred,22 documented tradition and direct archaeological evidence so far appears to suggest construction only in the 18th and/or 19th century for the bank and ditch features (e.g., at Ita Yemoo and Oke Atan),23 though areas where there is said to have been a single earthen wall (and no ditch) remain unexplored. Lastly, another relevant feature at the grove that has been documented is a game (ayò) board carved into the living rock.24

Unpublished evidence on the layout and dimensions of the grove

  • 25 UG, ACCN 3120/1, B. Fagg Diary, 1/1/53, 3/1/53, p. 1–2; UCT, J2.23, Goodwin, Part IV, p. 2.

13Outside of Murray’s basic observations, no published material describes the dimensions and layout of the grove and the various cultural features within it. However, Goodwin’s writings give a more vivid description of these, providing an approximate image of what the grove would have looked like. The grove today is reduced due to urbanisation; thus, these materials offer us some valuable information since lost. B. Fagg and Goodwin’s diaries made some preliminary observations of the location and surrounding landscape of the grove during their first attempts to locate it on the 1st and 3rd of January 1953. The main approach was from the direction of Modakeke, north towards the Ibadan Road, the path crossing a stream overlooked by the site. It was in a relatively dense area of forest, where Goodwin described “thickets of bamboo and wild strelitzia with cotton trees towering above”, stretching “300 yards or more” in various directions around the site.25

  • 26 UCT, J2.23, Goodwin, Part IV, p. 6–7.
  • 27 P. Ozanne (1969, p. 29–30) seemingly implies that there was no ditch at this section of the wall (a (...)
  • 28 This sectional arrangement is typical for sacred groves in Ife (and elsewhere), as outlined in othe (...)

14It was not until the 5th of January that the team reached the grove, thanks to the guidance of the “chief priest”, Emmanuel “Oshungbekon” (probably Ọ̀súngbẹ́kún), who appears to have led them there through the forest “along the old city wall” (probably from north to south), they initially passing the shrine so that they could enter by the correct entry point.26 The city wall in this area likely comprised a linear bank or freestanding earthen wall and perhaps a ditch27 (see Fig. 2). Based on Goodwin’s observation that the grove was largely “L” shaped, the “crook” of the “L” being the main shrine area, on a SW–NE axis (based on the direction of their excavation grid), it can be tentatively extrapolated that the main path of approach was oriented, approximately, from the south-east to the north-west. Goodwin used ecclesiastical language to describe the grove’s layout, dividing it into a “chancel”, where the main shrines were located and where initiated men performed the main rituals (the “crook” of the “L”), and a “nave”, where the main entrance to the “chancel” was located and where women and uninitiated males were confined.28 Given that this is the best description we have of the shrine, which my reconstruction is largely based upon, I have opted to retain this language, albeit marked in quotation marks when used given its ethnocentrism (which Goodwin himself was aware of).

Figure 2: The ditch and bank of one of the Ife earthworks in 1953, with bamboo growing from the latter

Figure 2: The ditch and bank of one of the Ife earthworks in 1953, with bamboo growing from the latter

The location is not clear, but may be near the Iwinrin Grove to the west of Osangangan Obamakin, as B. Fagg mentions in his diary that he and his brother “explored the ancient wall” together in this area (UG, ACCN 3120/1, B. Fagg Diary, 2/3/53, p. 25).

Photograph by William Buller Fagg, 1953 (400_WBF.1953.05.05) © RAI.

  • 29 UCT, J2.23, Goodwin, Part IV, p. 7.
  • 30 UG, ACCN 3120/1, B. Fagg Diary, 17/1/53, p. 7; UCT, J2.23, Goodwin, Part IV, p. 7.

15Goodwin described the “nave” – the main area of approach – as “a hundred feet or more long and twenty feet wide, cleared between the euphorbias which stretched thirty feet above us”.29 Goodwin is referring here to peregun trees (Dracaena fragrans), apparently conflating them with some of the euphorbia plants of his homeland in South Africa, broadly similar in appearance (e.g., Euphorbia ingens) and also often associated with archaeological sites. Both B. Fagg and Goodwin, as well as Murray in 1943, attested to the former presence of a representation of Esu at a point between the “nave” and “chancel”. Murray described an upturned pot marking this, and Goodwin mentioned “a figure of crude schist” that had guarded the entrance to the “chancel” since removed by a man from Lagos living in Modakeke several years prior.30 This marked the spot beyond which women and uninitiated males could not pass, with a long palm midrib “lintel” stretched along an “arch” of peregun trees also marking this sacred area of transition. The Esu stone was one of two (or more) at the grove, the other being within the main “chancel”.

  • 31 R.E. Dennett, 1910, p. 139. For other examples of this type of tree being used as boundary markers (...)
  • 32 Bamboo patches have also been associated with anthropogenic features elsewhere in West Africa. For (...)
  • 33 UG, ACCN 3120/1, B. Fagg Diary, 5/1/53, p. 2.
  • 34 UCT, J2.23, Goodwin, Part IV, p. 2.
  • 35 A.J.H. Goodwin, 1958, p. 46.

16Both B. Fagg and Goodwin explicitly noted the close proximity of the wall of Ife to the sacred grove, and some of the peregun trees associated with this feature – an observation likely connected to the common use of such trees as boundary markers throughout Yorùbá history and that of other parts of West Africa.31 Goodwin (1958, p. 45) also mentioned that bamboo was often found along the bank (which he interpreted as the remnants of older wicker-work defences), which may explain some of the bamboo clumps that were observed during their earlier efforts to locate Osangangan Obamakin.32 B. Fagg viewed the grove as “a fortified place, possibly a gate in the old Ife perimeter wall”,33 and Goodwin stated, seemingly with more certainty, that the site was “at one of the city gates in the old days, overlooking the Modakeke stream”,34 as he did in his later publication.35 Goodwin also wrote:

  • 36 UCT, J2.23, Goodwin, Part V, p. 4.

Bernard followed the wall and found that the pattern of euphorbia trees (the sacred peragun [sic] of the groves) followed the old wall, which here came out into a bay or fortress. Curiously enough, although the wall faced Ife more or less, we were told that that aspect was the outside, so that some great recurving sweep must have brought this portion out to face on the holy city. The bay was on the inner face of the city wall.36

  • 37 P. Ozanne, 1969; F. Willett, 1967, 2004.
  • 38 T. Falola, 1989, p. 73–74.

17This supposed “bay” or “fortress” is not shown at the site of Osangangan Obamakin on Willett and Ozanne’s maps of the walls (both of which were based on Willett’s observations of the wall near the site). However, they do show a convexity followed by a concavity (south of it) in the wall at this point, which may be related to what Goodwin is referring to (see Fig. 1). It might be speculated that such a feature might appear similar to the loop between the walls observable near Oke Atan (Fig. 1). Willett’s maps suggest that the site was just within the city wall here.37 Beyond this, the purported presence of an “ancient abandoned market” (as documented by Murray in 1943) that the grove was behind also supports the possibility that this was an old, fortified entranceway, markets being often commonly associated with such spaces.38 That being said, caution is warranted as there is no specific mention in the archaeologists’ writings of a breach in the earthwork that would conclusively suggest an entranceway near the site. The liminal deities present at this grove, their shrines almost immediately next to the city wall, may also be associated with protection and transition across this boundary. These factors pertaining to the wall are all crucial for understanding the history of the sacred grove and its arts, yet were never before published.

18Within the “chancel” was a reportedly spectacular landscape of peregun trees, as well as larger iroko (Milicia excelsa) and silk-cotton trees (Ceiba pentandra) (see Fig. 3). On the 3rd of April 1953, the famous American anthropologist Melville Herskovits visited Osangangan Obamakin with Bernard Fagg during his travels through West Africa that year. He was evidently very impressed with this area of the grove, writing in his diary:

  • 39 Melville J. and Frances S. Herskovits Papers, Schomburg Center for Research in Black Culture, New Y (...)

The great trees and the peregun bushes, that rose to the greatest heights I have seen for them, were most impressive. It was a rather unkempt place, however, but I suppose that they clean it up for festivals.39

19Given Herskovits’s extensive experience as a researcher across the African continent, such a testimony suggests that, in spite of its apparently dishevelled appearance, the grove at Osangangan Obamakin was visually and aesthetically remarkable and, given the presence of these large trees (capable of living for centuries), home to perhaps quite old botanical features. Sitting amongst this arboreal grandeur were the main shrine features noted above. The precise dimensions of these features are not given, but some approximate estimates are present for the main shrine: the remains of a small earthen structure (approx. 2 yards (approx. 1.83m) each way); and the “Esu Ejubo” is said to have been the larger of the two accumulations of pots, the smaller one being the Ogun shrine to the east. Although the original report of this excavation refers to these as “subsidiary” “Esu Ejubo” and “Ogun” shrines, apparently relational to the main shrine of Osangangan Obamakin, Goodwin’s diary suggests that this is misleading. Following conversations with the priests, Goodwin noted that:

  • 40 UCT, J2.23, Goodwin, Part VI.

… the Eastern shrine was dedicated to the god of iron (?OGUN?). The other was dedicated to Osangangan and contained the Eshu (guardian) within the shrine as opposed to the gatekeeper eshu in the nave, which had been taken off to Lagos. This eshu within the shrine is the insignificant bit of schist. The central shrine is the spot where Osangangan met his wife and peopled the earth from Ife.40

Figure 3: The main area, or “chancel”, of Osangangan Obamakin in 1953

Figure 3: The main area, or “chancel”, of Osangangan Obamakin in 1953

Photograph by William Buller Fagg, 1953 (400_WBF.1953.04.14) © RAI.

  • 41 UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, p. 3–4.

20Further confirming this, he wrote that the western shrine “is reputed to be the Osangangan Shrine”, with only the “small schist stone” within it representing Esu.41 Based on this, then, the three main features present at the “chancel” of the grove were as follows:

    • 42 Ibid., p. 4.

    The north-eastern “main shrine” marking the spot where Osangangan peopled the earth. Referred to by Goodwin as the Central Shrine.42

    • 43 See T. Falola, 2013; A. Ogundipe, 2012; J.A. Pemberton, 1975.
    • 44 UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, p. 3.

    A larger, western accumulation of pots: a shrine dedicated to Osangangan Obamakin, with an Esu stone (yangi)43 within it. Referred to by Goodwin as the South-Western Votive Shrine.44

    • 45 UG, ACCN 3120/1, B. Fagg Diary, 17/1/53, p. 7.
    • 46 UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, p. 4.

    A smaller, eastern accumulation of pots: a shrine representing, in B. Fagg’s words, “… the Ogun of Os. Ob. himself”.45 Referred to by Goodwin as the Eastern Votive Shrine.46

21It was in this “chancel” area that the archaeologists then laid out their excavation grid, and undertook their work.

Reconstructing the excavation grid

  • 47 UG, ACCN 3120/1, B. Fagg Diary, 7/1/53, p. 4; 1 yard = 0.9144 metres or 3 feet.

22Here I explain and reconstruct the excavation grid laid out in the “chancel” as recorded by B. Fagg in his diary. This grid, cross-referenced with Murray, B. Fagg, and Goodwin’s descriptions of the grove’s landscape, is important for developing an approximate sketch map of the site and its shrines, excavations, and other features in the absence of one drawn by the archaeologists themselves (see Fig. 4). Note that this sketch map is based almost entirely on written descriptions and estimates and not on any pre-existing maps of the grove and excavations. All dashed lines represent approximate dimensions of features based on these descriptions. Greyed squares within the grid represent those that were either partially or wholly excavated. Approximate locations of the most prominent artefactual finds are marked. As B. Fagg wrote: “The grid set out was based on capital letters for the vertical strips and small letters for the horizontal, allowing more latitude for expansion to the S.W. and to the East.” Each square (e.g., Gc, Hd, We etc.) was 1 yard by 1 yard (approx. 91cm by 91cm).47 The capital-lettered vertical strips ran from G to W but with the letters I and O omitted, presumably to avoid conflation with the numbers 1 and 0 respectively. The small-lettered horizontal strips ran only from c to e in Fagg’s sketch of the grid letters, although his later notes mention ‘j’ squares (at the location of the Ogun shrine) suggesting expansion of the grid by several yards to the south-east. No information is given on the letters between e and j; in my reconstruction of the excavation, I have assumed that the letter i was similarly omitted from the small letters for consistency’s sake, although this may be incorrect. The grid was oriented on a SW–NW axis. The total length from G to W was 15 yards (45 feet; approx. 13.72m), and from c to e 3 yards (9 feet; approx. 2.74m), perhaps an additional 4 yards (approx. 3.66m) added to reach the j squares to the south-east (see Fig. 4). Each of these squares was subdivided into 9 squares, each one a square foot (approx. 30.5cm by 30.5cm). These increased in a descended and lateral direction (see Fig. 5).

Figure 4: A reconstructed sketch map of Osangangan Obamakin

Figure 4: A reconstructed sketch map of Osangangan Obamakin

Sketch map showing the location of the excavation grid, and important features and finds within it. It has not been georeferenced to a specific location within modern Ife: as of now, only the approximate location can be discerned.

Tomos Llywelyn Evans, 2024.

Figure 5: An example of how the 1x1 yard grid squares (see Fig. 4) were subdivided into square feet, and how the system of coding finds and features worked (in this case, square Gc)

Figure 5: An example of how the 1x1 yard grid squares (see Fig. 4) were subdivided into square feet, and how the system of coding finds and features worked (in this case, square Gc)

Finds in the greyed square would be coded ‘Gc9’ – G being the x-axis letter (capital letters), c being the y-axis letter (small letters), and 9 being the square-foot subdivision within the larger square.

Tomos Llywelyn Evans, 2024.

23These numbered subdivisions allow us to better pinpoint the locations of finds within each yard square. As an example of how the coding system worked in assigning contextual information to each excavated object, “Gc9” would mean that an artefact was found in the yard square corresponding to the horizontal letter G, vertical letter c, and subdivision 9 (the rightmost corner) of the square (Fig. 5). Using this grid system, cross-referenced with B. Fagg and Goodwin’s written field descriptions, we can pinpoint with relative precision the places that were excavated, although it is not always clear where within these grid squares the archaeologists opened excavations.

Survey observations prior to excavation

  • 48 F. Willett, 2004.
  • 49 UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, p. 4.
  • 50 Ibid., p. 4.
  • 51 Ibid., p. 5.
  • 52 Note that references to “maize-cob roulette design” may be misleading, given that certain strip and (...)

24The features observed by Murray in 1943 – the heap of offering pots (B. Fagg and Goodwin’s grid squares Hc and Jc – the Osangangan and Esu shrine) and an accumulation that was likely previously a small round hut (grid squares Sd, Se, Td, and Te) where the main sacred objects were kept (the “main shrine”; see above)48 – were observed by the archaeologists in 1953. They also identified further shrine accumulations – a smaller pile of pots to the east (grid squares Pj and Qj) later identified as a shrine to Ogun, and a heap of sherds which Goodwin and B. Fagg speculated could be another “forgotten eshu [sic] and shrine”.49 The main shrine was described by B. Fagg as “… apparently very recent, though it may have replaced earlier examples, seems to me to have been circular in plan, directly represented by the circular heap found there”, the round plan further suggested by their later excavations.50 On this shrine was a large stone sculpture, resembling a calabash and decorated with “rosette” designs (FW S110), referred to by the main priest as “the water-cooler”.51 Goodwin was unsure about its significance, but highlighted the importance of the two halves of the calabash as a cosmological symbol of the Yorùbá – signifying heaven and earth. Goodwin also commented on the overturned bowl collected by Murray, stating that it was a type “used for indigo-dyeing” and decorated with a maize-cob roulette design.52 Goodwin mentioned that they removed the “water-cooler” from the shrine. He described it as follows (see Fig. 6):

  • 53 UCT, J2.23, Goodwin, Part V.

Standing on the central shrine was a large carved stone, about eighteen inches high and weighing a hundredweight or so. It was shaped like a calabash with a carefully carved loop of rope about the neck between the two lobes. At the top could be seen a fracture suggesting some sort of neck that had disappeared. We found no trace of it. The upper lobe was patterned all over with small rosettes. Clearly the work had been done by peeling away at the surface of the schist. This was set aside and later removed from the shrine. The priest described it as “the water bottle”.53

Figure 6: Photographs of the “water cooler” from 1953 (FW S110)

Figure 6: Photographs of the “water cooler” from 1953 (FW S110)

Left: Photograph by William Buller Fagg, 1953 (400_WBF.1953.08.03) © RAI. Right: Photograph by A.J.H. Goodwin, 1953, UCT BC 290, Photographs, 1570–1612.

  • 54 F. Willett, 2004, S110.
  • 55 UG, ACCN 3120/1, B. Fagg Diary, 9/1/53, p. 5.
  • 56 UG, ACCN 3120/1, B. Fagg, Notes, Other, 3/3/53.
  • 57 Oral tradition suggested that this head was kept hidden under the inverted pot so that it would not (...)

25Despite the “calabash” interpretation, Willett added that “cooler” refers to “a porous water pot” and observed that the piece was decorated with concentric circles on the globular section at the tip, perhaps representing the foam of palm-wine as it ferments inside. He also observed that the top is broken away, the surface eroded, and that a triple cord is represented, knotted to produce a loop.54 A brief clue pertaining to its use at the shrine in 1953 is given by B. Fagg, who recorded that, on 9/1/53, the chief priest “tipped over” the object during a somewhat confrontational meeting with the archaeologists as they disturbed the shrine – perhaps an action of spiritual significance.55 B. Fagg was informed that previously at the main shrine structure, there had been “statues” of Osangangan Obamakin and his wife, Awunrin Olojo (Grid Squares Td6 and Sd6 respectively).56 A delegation of priests also informed B. Fagg of their identifications of the other objects taken from the shrine by Murray in 1943, notably that the terracotta head under the overturned bowl (FW T236) had represented Osangangan himself,57 that another terracotta head represented his wife (FW T246), and that other items represented his dog (FW T242), his ram (FW T245), and one of his messengers (FW S43 or S44). The seated figure with large testicles excavated from below the shrine by Murray (FW T239) was identified with Aroni, “spirit of whirlwind who used to be sent by Osanganga Obamakin at his pleasure”, though Willett (2004, T239) considered this a potential confusion between spirits, Aroni typically being depicted as a “one-legged wood spirit”. Willett believed that the image depicted elephantiasis of the scrotum, or possibly hydrocele, and thus may have been associated with “cults” specialising in the curing of afflictions.

  • 58 UCT, J2.23, Goodwin, Part IV, p. 7.
  • 59 UG, ACCN 3120/1, B. Fagg Diary, 6/1/53, p. 3.
  • 60 Ibid., 9/1/53, p. 4–5.
  • 61 Ibid., 6/1/53, p. 3; UCT, J2.23, Goodwin, Part V.
  • 62 UG, ACCN 3120/1, B. Fagg Diary, 6/1/53, p. 3.

26Goodwin also commented on the shrine features at the “chancel”, describing the western Osangangan Obamakin shrine as “a long heap of pots ranging in size from that of a golf ball to that of a child’s head, some round, some elongated, rising from a deeper, buried heap, mossed over and slowly filing with humus”.58 He identified the north-eastern main shrine (“debris of a clay wall left by the collapse of the little shrine”) to the right of these pots as “the exact spot where Osangangan came to earth to mate with his beloved and so to people the earth”. He described the eastern Ogun shrine a “little to one side” of these as “another heap of pots, like the first [Osangangan/Esu shrine] but smaller and more recent”. One other surface feature described was a “metal rod” which was “observed in situ at the North end of the ring of earth surrounding the main shrine”.59 This was later stated by the chief priest to be “the end of a hoe”, and he told the archaeologists that “nobody must move it”, suggesting it to be of sacred significance.60 The archaeologists cleared the whole site of leaves, laid out the string line of the grid, and went over the area with a mine detector.61 According to B. Fagg, “No sherds or other remains were observed at this point”, but the mine detector turned up two metal objects lying in the main shrine.62 Goodwin described these as follows:

  • 63 UCT, J2.23, Goodwin, Part V. As Blier (2015, p. 350–351) has documented, Osangangan Obamakin was as (...)

A scratch showed a small bell or rattle of iron, exactly like the socket of a spear-head, an inch wide at the mouth and perhaps four or five inches long, with a loose tongue. When we eventually unearthed the second it proved to be a second, less complete example.63

27Despite the lack of surface evidence of activity, B. Fagg learned that in both the recent past and present, this “chancel” area had been a centre of religious practice. As B. Fagg wrote in his field diary (based on the testimony of the priests):

  • 64 UG, ACCN 3120/1, B. Fagg Diary, 5/1/53, p. 2–3.

It seems to have been established by Olafin [Olofin] (who came down from heaven), the father of Obalufon after his son had usurped the throne. Orungbe is the annual festival. One votive pot is placed at the shrine each year at this time. It contains palm wine. Dried meat and matches (?) brought from Afin. Sheep sacrificed and the participants share the meat, the women remaining in the larger clearing on the main approach to the shrine. The ground is probably no longer really hallowed and there appears to be no objection to our excavations. Everything, they said, was so mixed up now that they could not possibly say if we are likely to find more things or not … The votive plots are placed on the shrine once a year. At first they affirmed that only one pot is placed there each year at the beginning of the festival, but they later stated that after the festival another pot is placed on the shrine.64

28Goodwin also noted the chief priest’s input pertaining to the recent use of the shrine, as well as how this shaped the archaeologists’ early hypotheses about the site:

  • 65 UCT, J2.23, Goodwin, Part IV, p. 7–8.

The old priest was willing to help, but clearly puzzled by our intentions. He could see no interest in the possibilities of old broken images buried under soil and humus. The spot itself was sacred, discarded figurines would have been replaced as the cult thought fit, and the old things might lie where they were. He told us that the accumulation of pots represented one offering of a pot of palm oil at the beginning of the annual sacrifice, and a second at the end of the sacrifice. If this is correct, the shrine goes back for well over 300 years.65

29According to Fagg, the grove had lost some of its significance to the religious practices of Ife:

  • 66 UG, ACCN 3120/1, B. Fagg, Notes, ‘Some Archaeological Problems at Ife’, p. 1.

Since 1943 no further sacrifices had been made in the grove, which had not been maintained in any way. The annual ceremony is now performed within the walls of the Oni’s palace.66

  • 67 UCT, J2.23, Goodwin, Part V; UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, p. 7.

30This highlights the transformational impact of Murray’s actions in collecting the objects from the site in 1943 and bringing them to the palace, the centralisation of storage of Ile-Ife’s sacred objects contributing to a corresponding centralisation of ritual practice in the afin (palace complex). Nevertheless, Goodwin learned that night-time services of song and dance occasionally occurred at the grove, led by the local church of the Nigerian Christian Order of Cherubim and Seraphim.67 Based on all of this input, the archaeologists appeared to be of the opinion that the grove had lost a lot of its spiritual significance, partly due to the removal of objects in 1943. This conclusion proved incorrect, with the chief priest objecting to the excavations, and evidence of ongoing offering practices observed later on during the fieldwork.

The excavations

31The archaeologists excavated the three aforementioned shrines, and test excavations and shafts were opened elsewhere on site (see Fig. 7). Below are details of the locations, artefactual and stratigraphic evidence, and interpretations of each excavation area. In these syntheses, I strive to represent the interpretations of the archaeologists at the time, occasionally giving my own additional input (largely in the footnotes, with a more detailed discussion in the next section).

Figure 7: The 1953 excavations at Osangangan Obamakin

Figure 7: The 1953 excavations at Osangangan Obamakin

Photographs by William Buller Fagg, 1953 (left: 400_WBF.1953.04.13; right: 400_WBF.1953.04.15) © RAI.

Osangangan Obamakin (and Esu Yangi) shrine excavations

32Feature: A shrine dedicated to Osangangan Obamakin (Obalufon I), with an Esu yangi (“schist stone”) within it.
Grid Squares: Hc, Jc.
Dates of Excavation: 7/1/1953 to 3/2/1953.
Excavators: A.J.H. Goodwin, Musa Gombe, Bernard Fagg, Liman Ciroma.
Location: Western area of the “chancel”, in front of the entrance from the “nave”.
Layers and Total Depth: Five defined layers mentioned – Layers I, II, III, IV, and V. Consisted largely of complete pottery vessels (hundreds) of various sizes. No clear depth measurements given.

Observations

  • 68 UG, ACCN 3120/1, B. Fagg Diary, 8/1/53, p. 4.
  • 69 UCT, J2.23, Goodwin, Part IV, p. 7.
  • 70 Ibid., Part V.
  • 71 UG, ACCN 3120/1, B. Fagg Diary, 9/1/53, p. 4–5.
  • 72 J.A. Pemberton, 1975.

33For the first few days (7/1/1953 to 9/1/1953), a mass of hundreds of complete pots was removed across both squares in four layers (I to IV).68 These varied in size and shape, from “a golf ball” to a “child’s head”, some elongated, others round.69 The stripping of the pottery revealed a “small schist stone at the end of the heap”70 which, based on their description and later identification by the priests, appears to have been an Esu stone (yangi). This is further supported by B. Fagg’s observation that “… a large proportion of pots appeared to have a spear thrust (cut 1″x1/4″) through side”71 given that pots associated with Esu often had holes bored into them.72

  • 73 UG, ACCN 3120/1, B. Fagg, Notes, ‘Some Archaeological Problems at Ife’, p. 1–2.
  • 74 UG, ACCN 3120/1, B. Fagg Diary, 8/1/53, p. 4.
  • 75 Ibid., 9/1/53, p. 4–5.
  • 76 UG, ACCN 3120/1, B. Fagg, Notes, ‘Some Archaeological Problems at Ife’, p. 1.
  • 77 UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, p. 3–4.
  • 78 UCT, J2.23, Goodwin, Part V.
  • 79 UG, ACCN 3120/1, B. Fagg Diary, 9/1/53, p. 4–5.
  • 80 UCT, J2.23, Goodwin, Part V.

34On 8/1/1953, an unusual find was made in Layer IV of square Hc “Deep in the heap of pots”73 – a large bowl (about 9 in diameter and 10 high (approx. 22.9cm and 25.4cm respectively)) within which was a small pot and a terracotta figurine “firmly embedded in the laterite which the bowl contains”.74 Further excavations on 9/1/1953 revealed this find to be an overturned bowl, within which were three small pots as well as the figurine, approximately 6″ (approx. 15.2cm) in height, which depicted a simian – interpreted as a young gorilla (“aged 6 years”)75 or chimpanzee76 (see Fig. 8 (FW T247)). The figure was apparently “pressed against the N.E. edge” of the overturned pot and “facing towards the Central [Main] Shrine” to the north-east.77 Goodwin also described the figure’s orientation: “… with his face turned to a small schist stone at the end of the heap, which had been uncovered by our removing the overlying pottery”.78 Thus, the figurine was facing both the nearby Esu yangi, as well as the main shrine structure to the north-east – a pair of descriptions that allows us to situate the Esu stone in the north-eastern edge of the pottery accumulation. It appears that the overturned pot was within Layer IV but with its mouth embedded in the deepest layer of pots below (Layer V), and B. Fagg speculated that two of the pots inside it might have fallen in from the layers above,79 the upward-facing base of the pot having been broken.80 Indeed, the overturned pot had been submerged with other ceramics on all sides. As Goodwin stated: “… the mouth of the pot under which it stood was completely wedged into the bottom level of pots, and pots lay all over this ten inch high beaker”.

Figure 8: Terracotta primate figurine (FW T247) excavated from the shrine of Osangangan Obamakin

Figure 8: Terracotta primate figurine (FW T247) excavated from the shrine of Osangangan Obamakin

See F. Willett (2004, T247), R. Abiodun (2014, p. 69–70) and S. P. Blier (2015, p. 325–328, 512) for interpretations of this piece.

Date: 1953; obtained from F. Willett, 1967; © NCMM, Nigeria.

  • 81 UCT, J2.23, Goodwin, Part VIII.
  • 82 UG, ACCN 3120/1, B. Fagg, Notes, ‘Some Archaeological Problems at Ife’, p. 1.
  • 83 UG, ACCN 3120/1, B. Fagg Diary, 31/1/53, p. 11.
  • 84 UG, ACCN 3120/1, B. Fagg, Notes, ‘Some Archaeological Problems at Ife’, p. 1.
  • 85 UCT, J2.23, Goodwin, Part VIII.
  • 86 UG, ACCN 3120/1, B. Fagg Diary, 31/1/53, p. 11.
  • 87 UCT, J2.23, Goodwin, Part VIII.
  • 88 Ibid., Part VIII.
  • 89 UG, ACCN 3120/1, B. Fagg Diary, 3/2/53, p. 12.
  • 90 Ibid., 3/2/53, p. 12.
  • 91 Ibid., 3/2/53, p. 12; UCT, J2.23, Goodwin, Part VIII.

35After the first three days of excavation, serious work at this feature appears to have resumed on 31/1/1953 when Musa Gombe continued digging at squares Hc and Jc. Gombe was responsible for the next major find made within the feature, this time in square Jc at the Esu stone (yangi). The yangi appeared to be “set almost on the surface, weighing only about ten pounds”,81 and was “less than 12 inches at its maximum length” (less than approx. 30.5cm).82 About 3 (approx. 7.6cm) below the yangi, Gombe found skull bones, and further excavation revealed what appeared to be a “crushed adult skull” facing east towards the Ogun shrine.83 B. Fagg believed that the condition (and even the existence of) the cranial fragments suggested it to be of recent date,84 with Goodwin describing the remains as “soft, like chalk, and thin”.85 The skull appears to have been small, with B. Fagg writing: “At first size of skull appeared to make it unlikely that it was human but further digging made it appear like a crushed adult skull (facing E.)”86 and Goodwin stated that the individual was “probably a youngster”.87 B. Fagg took over excavation at the skull,88 and on the 2nd and 3rd of February focused excavations on the area around it, carefully digging and sieving the soil but leaving the skull itself in situ.89 The excavations located the lower jaw and found fragments of teeth in deteriorated condition, but no trace of any post-cranial material was present.90 Rusted fragments of a small iron blade were associated with the skull, and no other artefacts were found here.91 Given that the terracotta simian was facing towards the yangi and the main shrine beyond, it would thus also have been facing towards the skull. While B. Fagg conducted the excavations around the skull in square Jc, Liman Ciroma undertook excavations in the adjacent square Hc.

  • 92 UG, ACCN 3120/1, B. Fagg Diary, 19/3/53, p. 28.
  • 93 F. Willett, 2004, T248.

36A final notable discovery, probably from this feature, was a small terracotta head found in an excavated pot during finds processing afterwards on 19/3/1953,92 with lips that “protrude” and eyes “set in deep sockets”.93 Overall, little description of the excavated pots in these squares is given besides the information quoted above, although Goodwin did make comparisons between one pot and examples from Benin City and Ogunlaadin – a shrine associated with Ogun, and the legendary blacksmith Laadin, within the Ife palace (also excavated in 1953):

  • 94 UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, p. 3.

One pot from the SW shrine suggests affinities with those found at Ogun Ladi – I refer to the examples with corrugated neck-decoration, the neck rising from a symmetrical daisy-like pattern at the top of the body of the pot. (Cf. W. Ling Roth Great Benin, p. 75 and 126, the latter in metal.). The OS. OB. example has no neck left.94

Interpretations

  • 95 UG, ACCN 3120/1, B. Fagg Diary, 6/1/53, p. 4.

37The mass of pots were initially thought of as possibly representing the gradual deposition of offerings south-west of the main shrine. As such, B. Fagg and Goodwin hoped that these deposits would help develop a typology of ceramics, as they changed over time across (presumed) hundreds of annual offerings. As B. Fagg stated early in the fieldwork: “It may be possible statistically to develop a key to the changing trends in pot shapes, as the shape, shoulders, neck, aperture decoration and base vary considerably.”95 Goodwin described their assumptions as follows:

  • 96 UCT, J2.23, Goodwin, Part V.

In stripping an accumulation of this sort it has to be presumed that the earliest “layer” was a small patch of pots set side by side. The next “layer” covered these and extended all round, the next did likewise and so on. This is all our assumption, as people’s actions are likely to vary. In one year everything may be piled at one end, in the next at the other, but as we were hoping to get some slight clue to the order in which pot-shapes were fashionable, we had to work on some approximate system. The lowest, and especially the buried pots, must be the earliest, the top one must be the later in order.96

  • 97 UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, p. 3.
  • 98 Annual Report of the Antiquities Service 1952–3, p. 8.
  • 99 UCT, J2.23, Goodwin, Part V.
  • 100 T. Euba, 1984, p. 3.
  • 101 UG, ACCN 3120/1, B. Fagg Diary, 31/1/53, p. 11.
  • 102 This interpretation may not be farfetched given the common occurrence of foundation sacrifices in t (...)

38However, Goodwin was unable “to recognize any consecutive trend or development in type”,97 and it came to be supposed that the accumulation developed over a relatively short period of time and not over hundreds of years as they had previously speculated – a feature that was “superficial and of recent age”.98 The accumulation then was probably assembled as part of the development of this specific shrine to Osangangan Obamakin and Esu. Consultation with the priests contributed significantly to the archaeologists’ understanding of the feature, it being revealed to them that the “schist stone” related to Esu (“Esu Ejubo”) and that the accumulation was a shrine to Osangangan himself. During an early encounter on 9/1/1953, both B. Fagg and Goodwin’s accounts suggest that the chief priest identified the feature as associated with Sango. No later mention of this attribution was made, however, including during later conversations with the priests. It may thus have been an erroneous conclusion reached due to problems of translation, Goodwin describing the priest’s interpreters on the 9th as “extremely inefficient”.99 On the other hand, although Sango is a thunder deity associated with Oyo / northern Yorùbá spirituality rather than that of Ife, it is possible that close conceptual overlaps between him and the figure of Osangangan Obamakin led to the syncretic worship of Sango at the grove. Osangangan Obamakin is closely connected to, and/or conflated with, Oramfe, a “solar-thunder god” and “the most senior of the humanized spirits” in Ife.100 Due perhaps to increasing influence from the northern Yorùbá, especially those who settled in adjacent Modakeke, throughout the 19th and early 20th centuries, an association of the grove and shrine of Osangangan with the Oyo deity Sango may have developed. The excavated skull was interpreted by W. Fagg as a “sacrifice of consecration”,101 with Goodwin writing that the individual was “… clearly sacrificed to act as the eternal watcher at this shrine, the EDINA [probably Yorùbá Idena/Edena] or guardian”.102 An alternative possibility not considered by the archaeologists is that these ancestral remains pre-dated the formation of the shrine. In this scenario, it is possible that these skeletal elements were found and interpreted as the remains of Osangangan, their presence thus sacralising this spot, with the shrine constructed here for this reason.

Ogun shrine excavations

39Feature: A shrine dedicated to Ogun.
Grid Squares: Pj, Qj.
Dates of Excavation: 13/1/1953 and 4/2/1953.
Excavators: A.J.H. Goodwin, Musa Gombe.
Location: Eastern area of the “chancel”, to the right of the “nave” entrance, some seven yards east of the Osangangan Obamakin shrine.
Layers and Total Depth: No details of layers or depth given.

Observations

  • 103 UCT, J2.23, Goodwin, Part VI.
  • 104 UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, p. 4.
  • 105 UCT, J2.23, Goodwin, Part V.

40This accumulation was smaller in size than that of the Osangangan Obamakin shrine and consisted of “possibly 200 pots”,103 “in cleaner and newer condition”.104 It was situated near one of the large trees of the grove.105 Goodwin began excavations here on 13/1/1953 during which most of the finds were made, documented as follows:

  • 106 Ibid., Part VI.

Under the pots at the southern end lay about 30 used and abraded river pebbles about two inches [approx. 5cm] in diameter. Bernard thinks they were pecking stones used for shaping figurines by endless tapping. The “water bottle” [the stone gourd sculpture – see above] was probably made with these.106

  • 107 One simple sketch by Goodwin of a design that he refers to as “basket surface” (UG, ACCN 3120/1, Go (...)
  • 108 Ibid., Part VI.
  • 109 UG, ACCN 3120/1, B. Fagg, Notes, ‘Some Archaeological Problems at Ife’, p. 2.
  • 110 UG, ACCN 3120/1, B. Fagg, Notes, ‘Some Archaeological Problems at Ife’, p. 2. Ceramic crucible frag (...)
  • 111 UG, ACCN 3120/1, B. Fagg Diary, 4/2/53, p. 12.

41According to B. Fagg, these were found “partly under [a] grindstone” within the deposition of pots. Among the cluster of pebbles was a stone axe which Goodwin recognised as resembling ones “of the small simply [sic] type found in Rhodesia and the Transvaal” as well as “a few larger stones, pieces of a large pot with impressed basketwork design107 (or possibly rolled on with a maize-cob) and an abundance of snail shells left as sacrifices”.108 B. Fagg mentioned that the shrine yielded “mortars, pounding stones”109 implying that the larger stones observed by Goodwin were grindstones and/or hammerstones. B. Fagg also mentioned that a single fragment of glass crucible was excavated here (see Fig. 9 for a comparable recent example).110 Some final work was undertaken by Musa Gombe at these squares (Pj and Qj) on 4/2/1953, but no mention is made of his finds.111

Figure 9: A glazed crucible fragment from early Ife, observed on the surface at Olokun Grove in 2017, with a blue surface where the glass has fused with the ceramic

Figure 9: A glazed crucible fragment from early Ife, observed on the surface at Olokun Grove in 2017, with a blue surface where the glass has fused with the ceramic

Such objects were often coveted, and incorporated into shrines.

Tomos Llywelyn Evans, 2017.

Interpretations

  • 112 Annual Report of the Antiquities Service 1952–3, p. 8.
  • 113 Such stone axes were associated with certain deities and frequently occur in shrine contexts in Ife (...)
  • 114 UG, ACCN 3120/1, B. Fagg Diary, 17/1/53, p. 7.
  • 115 E.g., see P. Garlake, 1974, 1977.

42As with the Osangangan Obamakin shrine accumulation, it was concluded that this feature was “superficial and of recent age”.112 Of course, despite this, the assemblage consisted of artefacts of perhaps various periods – a glass crucible likely of “Classical Ife” (c. 11th to 15th century AD), a stone axe that may be from the Late Stone Age (or an imitation of such)113 and more recent ceramics. Probably viewed as spiritually potent, such objects were composed to form the new shrine aggregation. Initially, after the identification of the Osangangan shrine to the west with Sango, Goodwin interpreted this shrine as also related to Sango, using the presence of the stone axe (known icons of Sango) as evidence of that. Later, after more thorough consultation of the priests, the shrine was revealed to be “… to the Ogun of Os. Ob. [Osangangan Obamakin] himself”.114 Besides the oral testimony, the close relationship between the deities is further suggested by the orientation of the skull from the Osangangan shrine, which faced the Ogun shrine (see above), as well as that of the purported entrance to the main (central) shrine structure, also likely being oriented towards the shrine of Ogun. The stones were interpreted as “used for grinding spices for sacrificial means”, while the pebbles were speculated to have been used in shaping stone sculptures. A possibility that the archaeologists did not consider is that the pebbles may conceivably have been obtained from old pavement features, which often incorporated small stone nodules either exclusively or in tandem with potsherds.115

Main shrine excavations

43Feature: A shrine on the spot where Osangangan is said to have met his wife (Awunrin Olojo) and peopled the earth.
Grid Squares: Sc, Sd, Se, Tc, Td, Te.
Dates of Excavation: Approx. 15/1/1953 to 23/3/1953.
Excavators: A.J.H. Goodwin, B. Fagg, W. B. F. Fagg, K. C. Murray, E. Thaddeus, M. Gombe.
Location: North-eastern area of the “chancel”, some seven yards north-east of the Osangangan Obamakin shrine.
Layers and Total Depth: Varies by square. Maximum depth of about 14 feet (approx. 4.3m) from the surface to the base of the deep shaft feature.

Observations

  • 116 UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, p. 4.
  • 117 Ibid., p. 5–6.
  • 118 UG, ACCN 3120/1, B. Fagg Diary, 15/1/53 and 17/1/53, p. 6–7.
  • 119 Ibid., 15/1/53, p. 6.
  • 120 Ibid., 16/1/53, p. 7.
  • 121 Ibid., 16/1/53, p. 7.
  • 122 UCT, J2.23, Goodwin, Part VI.
  • 123 UG, ACCN 3120/1, B. Fagg Diary, 16/1/53, p. 7.
  • 124 UCT, J2.23, Goodwin, Part VI.
  • 125 Ibid., Part VI.

44This shrine at the time consisted of a “circular heap” that the archaeologists believed represented the collapsed remains of a circular earthen structure.116 Excavations occurred within six grid squares, beginning with an “approach trench” in the north-west, on the opposite side from Murray’s small 1943 excavation.117 The first digging, by B. Fagg, appears to have begun at squares Sc and Tc, just north-west of the shrine, on or just before 15/1/1953.118 These squares were dug rapidly, with B. Fagg photographing a “stone” in Square Sc at 12″ (approx. 30.5cm) deep.119 On 16/1/1953, the approach trench was then expanded into squares Sd and Td, with Goodwin and W. Fagg taking charge of excavation here, and Murray also joining them on this day.120 The initial excavations in the approach trench on 16/1/1953 and 17/1/1953, just outside the main shrine, reached 2′6″ (approx. 76.2cm) in depth including the removal of an “18″ [approx. 45.7cm] shelf from Sd and Td”.121 This shelf was comprised of “old scrappy potsherds”,122 and notable finds included a “large piece of pot”123 and “one very small pot at 8″ [approx. 20.3cm]”.124 Goodwin figured that the small pot was “possibly one that had rolled off the nearest votive heap, down the root-hole of some old euphorbia [peregun] tree.”125 Goodwin’s interpretation of these approach trench deposits highlights the town wall’s (very) close proximity to this shrine area, and its potential impacts on site formation processes:

  • 126 Ibid., Part VI.

Several whole or crushed pots were found, big holes where trees have rotted may account for some of these, but quite clearly this was a lower level at one time and the pots may have been offered then and later covered over by wash from the city wall.126

  • 127 Ibid., Part VII.
  • 128 UG, ACCN 3120/1, B. Fagg Diary, 17/1/53, p. 7. As at the Ogun shrine, the “basketry design” may be (...)
  • 129 Ibid., 17/1/53, p. 7.
  • 130 Ibid., 21/1/53, p. 8.
  • 131 Ibid., 17/1/53, p. 7.

45Indeed, Goodwin later described that, as they excavated the main shrine, “the rise of the wall” was “above [them]”.127 Such a close association further implies the close “functional” relationship between the grove shrines and the town wall. Following these excavations just outside the shrine, the archaeologists began their initial collecting and digging at the main shrine itself on 17/1/1953. B. Fagg reported that they “Removed all furniture from [the] top of [the] main shrine” including a “Large central pot with impressed basketry design” which “appear[ed] to have been inserted into a similar pot with an incised line below rim and parallel to it” and an “iron bell”.128 Surface “humus”, “no more than 4 [approx. 10.2cm] deep” was then removed from the shrine mound beneath which was “reddish earth”,129 presumably representing the collapsed structural deposits. It is not clear exactly where these first shrine excavations were, but it may have been at the southern/eastern margin of squares Sd and Td where the approach trench met the collapsed shrine.130 A lid, three cowry shells, and a pot were excavated, apparently at a depth of 26″ (approx. 76.2cm).131

  • 132 Ibid., 19/1/53, p. 7; UCT, J2.23, Goodwin, Part VII.
  • 133 UCT, J2.23, Goodwin, Part VII.
  • 134 UG, ACCN 3120/1, B. Fagg Diary, 20/1/53 and 21/1/53, p. 8. “A.J.H.G. watched shrine wall (so carefu (...)
  • 135 Ibid., 21/1/53, p. 8. This appears to be a case of the difference in scope of experience in field a (...)
  • 136 Ibid., 21/1/53, p. 8.
  • 137 UCT, J2.23, Goodwin, Part VII.
  • 138 Ibid., Part VII.

46On the next day of excavation on 19/1/1953, Goodwin took over the digging of the main shrine, uncovering several pots and pottery fragments, an iron bell and fragment of iron, as well as a large circular pottery dish 9″ (approx. 22.9cm) in diameter.132 Notably, he also uncovered a hole – apparently “the top of an old pit”.133 During excavation work on 20/1/1953, alongside W. Fagg, Goodwin worked to delineate the position of the walls of the purported shrine hut, but this effort apparently failed when W. Fagg unwittingly demolished the feature on 21/1/1953.134 On the 21st, the team finished excavating to the edge of squares Sd and Td where, at 10″ (25.4cm) deep, they uncovered the edge of a terracotta head which, as B. Fagg lamented, was “also demolished by W.B.F.”.135 B. Fagg, excavating “further down, at 12″ [approx. 30.5cm] at [the] middle of [the] shrine”, uncovered “1 ear” from a “smaller [terracotta] head – possibly conical”.136 Goodwin also reported that they found “one eye of a life sized terracotta figure”.137 Goodwin described the deposit here as “almost as hard as soft burnt brick”.138

  • 139 UG, ACCN 3120/1, B. Fagg Diary, 22/1/53, p. 8–9; UCT, J2.23, Goodwin, Part VII.
  • 140 UCT, J2.23, Goodwin, Part VII.
  • 141 UG, ACCN 3120/1, B. Fagg Diary, 24/1/53, p. 9.

47On 22/1/1953, excavations on the southern/eastern edges of squares Sd and Td, as well as in square Te (all within the shrine structure) at 18 to 20″ (approx. 45.7cm to 50.8cm) deep uncovered a spectacular cluster of objects. A 4″ (approx. 10.2cm) terracotta head with a “gag” (tied rope) in its mouth (FW T255) was found at 18″ (approx. 45.7cm) in square Td3 closely associated with a flat bowl in square Te1 and a heap of fragmentary pots in square Sd9139 (see Fig. 10). B. Fagg provided a sketch of approximately where these were found within the squares, and a photograph of this feature is present in W. Fagg’s archive (Fig. 10). Goodwin described this as “our greatest find”.140 On 24/1/1953, they completed excavation of squares Sd and Td, finding some further pottery that appeared to have slipped down beneath the shrine. They interpreted this as perhaps the result of a fallen tree creating a cavity, with objects subsequently slipping in, while the head and dish came to be viewed as intentionally interred.141

Figure 10: The archaeological context of the gagged terracotta head and associated finds, 18″ to 20″ (approx. 45.7cm to 50.8cm) below surface

Figure 10: The archaeological context of the gagged terracotta head and associated finds, 18″ to 20″ (approx. 45.7cm to 50.8cm) below surface

Left: a reproduction of the finds’ location within the grid squares, as represented in UG, ACCN 3120/1, B. Fagg Diary, 22/1/53, p. 9 (1953). Right: a photograph of the finds in situ. Photograph by William Buller Fagg, 1953 (400_WBF.1953.001.08).

Left: Tomos Llywelyn Evans, 2024. Right: © RAI.

  • 142 Ibid., 26/1/53, p. 9.
  • 143 Ibid., 26/1/53, p. 9–10.
  • 144 Ibid., 27/1/53 and 28/1/53, p. 10.
  • 145 Ibid., 29/1/53, p. 10–11.
  • 146 Ibid., 29/1/53 and 30/1/53, p. 10–11.
  • 147 Ibid., 30/1/53, p. 11.
  • 148 Ibid., 30/1/53 and 31/1/53, p. 11.
  • 149 Ibid., 2/2/53 to 5/2/53, p. 11–12.
  • 150 UG, ACCN 3120/1, B. Fagg, Notes, ‘Some Archaeological Problems at Ife’, p. 2.

48On 26/1/1953, excavations occurred at lateral trenches opened on each side of the main shrine – at squares Se and Te 1, 2, 4, and 5 where Goodwin worked; and squares Te 7 and 8 where W. Fagg worked.142 The digging came down on “a fine grit about 18 – 24 [approx. 45.7cm to 61cm] from ground surface” and sloping towards the heap in the west (in Sd9 – see Fig. 10). They recovered further terracotta objects: an ear, knees, two possible bells, a foot, a “large piece of right side of Oni’s head with ear and part of hair and head-dress” and several “bits of pottery”.143 On 27/1/1953, W. Fagg uncovered some pottery fragments in Te8, and on 28/1/1953, Goodwin and W. Fagg had come down on Murray’s “sill” of pottery, measuring 4x2′ (approx. 1.22m by 0.61m), with a fragment of spherical pot half exposed underneath.144 On 29/1/1953, in square Te9, Goodwin uncovered “a small terra cotta face with bulging eyes and fat features about 2¼″ [approx. 5.7cm] in diameter” resembling “a bottle stopper”, and there was “a pottery ‘neck’ close by”.145 Gradually, the archaeologists removed the earth between the visible pottery until the “gritty layer” or underlying pottery was exposed, and on 29/1/1953 and 30/1/1953 the entirety of the “sill” was removed, together with the underlying pottery.146 Edward Thaddeus worked on straightening the edges of the trenches and washing and marking the excavated pottery.147 On 30/1/1953 and 31/1/1953, Goodwin continued clearing the “mass of pottery” in the S and T squares under the “sill”.148 Stripping of the main trench here appears to have taken several more days (2/2/1953 to 5/2/1953).149 B. Fagg later noted that these lower layers consisted of “a mass of broken votive pots similar to those in the other shrines, though noticeably better fired, but not a trace of figurine pottery”.150

  • 151 B. Fagg had assumed that the hard gritty layer below the shrine accumulation represented an undistu (...)
  • 152 UCT, J2.23, Goodwin, Part XI.
  • 153 UG, ACCN 3120/1, B. Fagg Diary, 24/2/53 and 25/2/53, p. 22.
  • 154 Ibid., 2/3/53, p. 25.
  • 155 Ibid., 3/3/53, p. 26.
  • 156 Ibid., 3/3/53, p. 26.
  • 157 Ibid., 19/3/53 to 23/3/53, p. 28–30.
  • 158 Annual Report of the Antiquities Service 1952–3, p. 8.
  • 159 UG, ACCN 3120/1, B. Fagg, Notes, ‘Some Archaeological Problems at Ife’, p. 2.

49Following this, it appears that B. Fagg had Thaddeus and Gombe, as well as locally recruited well-diggers, continue excavation, and they appear to have unexpectedly151 revealed the presence of pottery below the main shrine excavations.152 Most notably, Gombe uncovered “2 quite distinctive pots at a depth of 39 [approx. 1.14m] and 43″ [approx. 1.3m] below ground level”, apparently somewhere in squares Sd and Td.153 Gombe then found “a softer area below” this on 2/3/1953, “suggestive perhaps of a tomb or burial chamber”.154 On 3/3/1953, Gombe continued digging the shaft in 6″ [approx. 15.2cm] layers, uncovering pottery as he did so. B. Fagg commented that “Musa is making a good job of following its limits” and Gombe soon found that the “shaft bellied out somewhat like those of Olokun Grove”155 (see Fig. 11). B. Fagg wrote that the shaft showed “unmistakable” evidence of infilling, and considerable quantities of charcoal and pottery (often from the same vessels, and much of which could be reconstructed) was present.156 B. Fagg made a simple sketch of the shaft prior to its full excavation, which gives a sense of its shape (Fig. 11). The shaft excavation was completed on 23/3/1953, with B. Fagg noting the unlikelihood that the shaft was a disturbed site (based on the pottery finds).157 The shaft (roughly 3¼ feet [approx. 0.99m] wide for most of its body)158 ended “abruptly” at “nearly 14 feet [approx. 4.23m] from the surface” and “went down immediately below the shrine, constricted at the top and opening wider below, rather in the shape of a wide-necked pickle bottle”.159

Figure 11: A sketch of the old shaft feature beneath the main shrine

Figure 11: A sketch of the old shaft feature beneath the main shrine

Based on a drawing by B. Fagg (UG, ACCN 3120/1, B. Fagg Diary, 3/3/53, p. 26), following Fagg’s own representation. He sketched similar shaft features at Olokun Grove (UG, ACCN 3120/1, B. Fagg Diary, 23/2/53, p. 21). They speculated that these might be burial shafts.

Tomos Llywelyn Evans, 2025.

Interpretations

50Broadly, the terracotta figurative art was all found relatively close to the surface of the shrine, on top of the “floor of large sherds”, below which were the deposits of better-fired “votive pots” overlying the “gritty” layer, beneath which was the deep shaft. The main features interpreted are outlined below:

  • 160 UG, ACCN 3120/1, B. Fagg Diary, 28/1/53, p. 10.
  • 161 F. Willett, 2004, II.5.
  • 162 UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, p. 6.
  • 163 Ibid., p. 6.
  • 164 A.J.H. Goodwin, 1958, p. 46.
  • 165 UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, p. 6.
  • 166 See G. Chouin et al., 2023, p. 20–21.

51The pottery “sill” or “carapace”: Interpretations of this feature, consisting of large pieces of pot overlapping each other, were a little contradictory. W. Fagg at the time of excavation proposed that the paving was “either a ceiling or a floor” and B. Fagg added that it was perhaps “a protective covering for something underneath”.160 Murray had originally proposed that it may have protected the doorway of the hut,161 while Goodwin initially tentatively suggested that it “may have been a platform for placing various sacred objects”.162 Initially, Goodwin was sceptical of the idea that there were potsherd floors here, writing that he “could see no evidence of paving” and did not remember “finding the small discs of pottery which form the upper surface of Ife paving”.163 However, he later suggested that there were potsherd pavements at the site, although this may be an erroneous statement given the apparent lack of recognisable pavement material reported in the unpublished documents.164 According to him, the bulk of the sherds in the “sill” had “maize-design impressed on them”.165 If we turn to more recent archaeological excavations, a feature that appears to bear resemblance to this is a large concentration of hundreds of pottery fragments from several large jars (broken in situ) uncovered at the Ita Yemoo site in 2023. They were found on a surface some 10–15cm above a potsherd pavement, and some bore examples of a folded strip roulette design (commonly mistaken for “maize” designs) rather than the “ladder pattern” carved roulette design typical of “Medieval”/“Classical” Ife. They were thus interpreted as dating to a later (perhaps 19th-century) period at which point the site had become a sacred grove.166 Based on this comparative evidence, the “sill” at Osangangan Obamakin was more likely a collection of shattered shrine pots from a later era – bearing examples of folded strip roulette patterns – rather than a potsherd pavement from the “Classical” period of Ife.

  • 167 Ibid., p. 5–6.
  • 168 He added: “Same applies to other figurine fragments.” (UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, (...)
  • 169 UCT, J2.23, Goodwin, Part VII.
  • 170 S.P. Blier, 2015, p. 82, 148.

52The terracotta gagged head: Goodwin wrote that the terracotta gagged head, and the associated pottery dish, were likely to have been intentionally buried in a pit, the top of which was indicated by “a marked circular groove” observed when digging the approach trench.167 Thus, this, and perhaps the “eye and an ear from broken figures”, may have been buried there “or thrown in with the infilling”, suggesting the digging of pits at the main shrine for the (recent) deliberate interment of older sacred objects.168 Although Goodwin and the others initially interpreted this as “a victim gagged for sacrifice”, Goodwin added “but most of these figures depict diseased persons and the Seventh Day Adventist doctor at once said “Tetanus”. The rope was inserted to hinder the complete locking of the jaw.”169 These observations are interesting given the presence of other figurative arts depicting disease associated with the grove, and its recent association with healing.170

  • 171 Annual Report of the Antiquities Service 1952–3, p. 8.
  • 172 UG, ACCN 3120/1, B. Fagg, Notes, ‘Some Archaeological Problems at Ife’, p. 2.
  • 173 E.g., see L. Frobenius, 1913, p. 23–26; M. Herskovits, 1938, p. 359. See also J. Goody, 1962, p. 14 (...)
  • 174 C. Chapelain De Seréville-Niel et al., 2024.
  • 175 G. Chouin, Pers. Comm., 2025.
  • 176 UG, ACCN 3120/1, B. Fagg, Notes, ‘Some Archaeological Problems at Ife’, p. 4.
  • 177 H. L. Roth, 1903, p. 43.

53The deep shaft: This appears to have been an important focal point of the sacred grove – potentially a feature of the early city (“very much older than the overlying stratum”)171 which later (perhaps 19th-century or earlier) communities constructed a shrine on top of where terracotta arts would be kept. Although no bones were found, B. Fagg believed this to be an old burial.172 While examples of deep burial shafts, of various forms, abound across West African history, including bell-shaped chambers,173 it remains unclear whether the example at Osangangan Obamakin was originally, or ever, used for burial as opposed to having some other function(s). At Ife, only Willett’s discovery of human remains at the bottom of a deep pit at Ita Yemoo, near the north-eastern section of Ife’s walls (F. Willett, 2004, II.2), may lend credence to B. Fagg’s interpretation. However, this has since been reinterpreted as potentially an earlier burial disturbed by a later pit rather than an example of a chamber dug specifically for interment.174 Since then, excavations of various types of deep shaft at Ife have yet to recover any evidence of human remains.175 Therefore, the interpretation of this shaft as a potential burial pit remains speculative and based largely on ethnographic/ethnohistorical analogy rather than on archaeological evidence from Ife. The archaeologists’ main sources for their interpreting the pit at Osangangan Obamakin as well as similar features at Igbo Olokun (Olokun Grove) in the north of Ife as burial vaults are accounts of royal burial practices at Benin city recorded in the Dutch writer Olfert Dapper’s Description of Africa, published in 1668, and by the French naval officer Jean-François Landolphe in the late 18th century. In one of B. Fagg’s writings that discusses Olokun Grove, there is a penciled in note stating “Compare Benin graves. Roth p. 43”.176 This refers to Ling Roth’s translation of the aforementioned passages. For example, Roth’s translation of Dapper’s description of Benin royal burial states that, “a large hole is dug out in his palace, wide at the bottom and narrow at the top, and so deep that the diggers are even drowned in the water.”177 We can therefore trace this specific interpretation of the shafts of Ile-Ife, which has prevailed in the literature, to an Early Modern account pertaining to the Benin Kingdom. Evidently, they saw similarity in the form of these shafts and the narrow-topped, wide-bottomed burial pits described by Dapper and Landolphe.

  • 178 L. Frobenius, 1913, p. 93.
  • 179 “Heaven” here is likely a reference to orun where the deities reside.
  • 180 UG, ACCN 3120/1, B. Fagg, Notes, Other, 3/3/53. The association between old deep holes and stories (...)

54In any case, the shaft clearly held a sacred and legendary significance by the time of the excavations in 1953. The placing of the shrine on top of this shaft feature was unlikely to be coincidental given the frequent tendency of tradition in Ife to recall the locations of old sites178 and the local narratives that alluded to such holes at the site. In narratives recorded from the priests by B. Fagg in 1953, Osangangan Obamakin himself was considered the first Ooni of Ife, who came down from “heaven” and moved between heaven and earth,179 meeting his wife at the spot of the hole (where the main shrine was constructed).180 The grove, then, was considered a place where heaven and earth met, as well as a place where the world of Ife met the world beyond (on the wall). Early in the excavation, the chief priest had come to the site and, seemingly perturbed by what they were doing, questioned the archaeologists’ work. As B. Fagg wrote in his diary:

  • 181 UG, ACCN 3120/1, B. Fagg Diary, 28/1/53, p. 10.

He seemed worried by the hole we had dug because the early people made rather free with the means of getting to heaven. Whenever they got in a bad temper they crashed down holes which led to heaven. So someone blocked up the holes and we were trying to rediscover these holes!181

  • 182 S.P. Blier, 2015, p. 1.
  • 183 L. Frobenius, 1913, p. 307–308. Such narratives were likely often based on memory of (re)discovery (...)
  • 184 E.B. Idowu, 1962, p. 13.

55Folklore surrounding the presence of deep otherworldly holes, often connected with apotheosis, and the movement of souls, spirits, and deities to and through Ife, is common in the town’s history,182 and tales exist of such holes being associated with wealth183 and of caves being “sealed up” to prevent “greedy” people from constantly entering heaven to “beg for some favour or blessing”.184 As such, the main shrine and its dense archaeological record, as well as surface artefacts such as the stone “cooler”, may exemplify later “sealing” (entrance blocking) and further sacralisation of an old burial shaft associated with the otherworld. The rest of the sacred grove landscape may then have developed around this focal point. To many locals, a group of archaeologists digging here may have been perceived as threatening given the belief that such “tunnels” should be sealed up, and the diaries of B. Fagg and Goodwin speak of unusual offerings and sacrifices left near the grove throughout their work, perhaps in response to their disturbance of this place.

Other test excavations

56Features: Test pits and shafts opened near the shrines.
Grid Squares: Lc, Wc and/or Qc, and elsewhere.
Dates of Excavation: Lc: 6/1/1953 and 7/1/1953. Wc and/or Qc: 15/1/1953 and 16/1/1953.
Excavators: Bernard Fagg (Lc), Liman Ciroma (Wc and/or Qc), locally recruited well-diggers.
Locations: The “chancel”; one test pit was a yard north-east of the Osangangan Obamakin shrine (Lc), the other one (or two) a few yards from the main shrine (Wc and/or Qc) according to B. Fagg (Goodwin, however, claims that it was just north of this shrine). Excavation shafts opened by well-diggers elsewhere.
Layers and Total Depth: Lc: Layer I was 2″ (approx. 5.1cm), Layer II was 8″ (approx. 20.3cm), Layer III has no recorded depth. Wc and/or Qc: 15″ (approx. 38.1cm) depth.

Observations

  • 185 UG, ACCN 3120/1, B. Fagg Diary, 6/1/53, p. 3.
  • 186 Ibid., 6/1/53 and 7/1/53, p. 3.
  • 187 UCT, J2.23, Goodwin, Part V.

57On 6/1/1953, while Goodwin excavated the nearby Osangangan Obamakin shrine, B. Fagg opened a “trial pit”, “at a point least likely to yield positive results”,185 apparently in square Lc (Willett, 2004: II.5). It appears that this was undertaken as a means of better understanding the local stratigraphy outside of the shrine accumulations. The excavation ended up being relatively shallow, the stratigraphic profile consisting of three layers. The first was “2 inches [approx. 5.1cm]” of “humus containing a fairly high proportion of pottery” (Layer I) overlying a second 8″ (approx. 20.3cm) layer of “compacted earth resembling a “river sand” and water worn fragments of white quartz” (Layer II).186 Goodwin added that this was “a hard compacted reddish soil” that was “filled with mica and quartz sand” and reported a “constant appearance of potsherds” in this deposit.187 The presence of cultural material indicated to B. Fagg that

  • 188 UG, ACCN 3120/1, B. Fagg Diary, 6/1/53, p. 3.

it may be taken as probable that this earth consists of soil brought from the river, to build the Ife city wall or a fortified gateway of the shrine itself. Whatever the structure may have been it has collapsed and trodden down into a compact mass.188

58Goodwin wrote of this that

  • 189 UCT, J2.23, Goodwin, Part V.

the earth had been dug up, made its walls, had fallen back, been churned over again and again by successive builders, then inundated again by the forest, leaving no sign of any reconstructable stratification.189

59B. Fagg also mentions a third layer (Layer III), but with no stratigraphic or depth information, and describes the trial pit finds as follows:

  • 190 UG, ACCN 3120/1, B. Fagg Diary, 6/1/53, p. 3.

Layer 1 contained 8 fragments of votive pots, layer II one possible, and layer II [sic: likely means “layer III”] no fragments at all, but included one piece of modelled pottery suggestive of a figurine. Layer II contained 2 pieces of slag and a hammer stone.190

  • 191 UCT, J2.23, Goodwin, Part VI.
  • 192 F. Willett, 2004, II.5.
  • 193 UCT, J2.23, Goodwin, Part VI.
  • 194 UG, ACCN 3120/1, B. Fagg Diary, 16/1/53, p. 7.

60Another notable test pit mentioned was Liman Ciroma’s brief excavation at “a third, probably more recent, accumulation of pots a couple of yards to the north” of the main shrine,191 apparently square Wc.192 Ciroma appears to have opened this on 15/1/1953, and Goodwin notes that they “started on it” and found the soil to consist “of broken pots from an old shrine” but soon lost interest having, by that point, already recovered some 600 pots from the other shrines.193 B. Fagg writes that on 16/1/1953, Ciroma “dug square Qc to a depth of 15″ [approx. 38.1cm]194 although no further details are given and this may be an erroneous reference to the Wc excavation. Later rapid, deep, less systematic excavations were undertaken by local well-diggers. Of these, Goodwin wrote:

  • 195 UCT, J2.23, Goodwin, Part VI.

Meanwhile at OS-OB [Osangangan Obamakin] we had imported a pair of “well-diggers”. These are especially trained members of a Northern tribe who, with a pick-head and a head-pan or bucket, will dig a two-foot [approx. 61cm] diameter hole until stopped. They are capable of going down for hundreds of feet. I discovered later at Jos that they are perfect in a 2 ft [approx. 61cm] hole, but utterly devastated if the hole is to be 3 ft [approx. 91cm] wide and quite incapable of a decent job. To work in this space they develop amazingly supple backbones, place their chins on their chests, run the head down the front of the body and over the navel and work about their feet … The second man pulls up the bucket or a rope, while the one below unbuckles himself and squeezes aside to let it scrape past his body …195

  • 196 B. Fagg, 1953, p. 126.
  • 197 UG, ACCN 3120/1, File No. 27/A/S.1. Ife Antiquities, Shafts.
  • 198 UG, ACCN 3120/1, B. Fagg Diary, 25/2/53, p. 22.

61These men appear to have excavated multiple (more than 80), typically ~30 feet-deep (approx. 9.14m-deep)196 shafts across Ife during this field season, including Shafts 1, 41–49, 51, 52, and 53 at Osangangan Obamakin. Shaft 1 was recorded as being 16 (approx. 4.88m) in depth while others were only one or two feet deep.197 These helped the team better understand the site stratigraphy, and some yielded artefacts including Shaft 43 in which there were “rare sherds to depth of some 2 feet [approx. 61cm] from surface”.198

Interpretations

  • 199 Ibid., 9/1/53, p. 4–5.
  • 200 F. Willett, 2004, II.5.

62The Lc trial pit was significant in that it appeared to indicate to the archaeologists the remains of further structures at the sacred grove, perhaps connected to the town wall. Indeed, when asked by the chief priest what they were doing, B. Fagg recorded that: “I explained the Yoruba had been here a very long time, they had built walls and forts and houses which had fallen down and we wanted to find out how old they were.”199 Everyday pottery, slag, and a hammer stone tentatively suggested the occurrence of everyday (including craft) activities here, in a site otherwise characterised by sacred objects. These may, tentatively, have been associated with the operation of the purported earthwork gateway, and/or connected to the nearby old market mentioned by Murray. It is not clear if these finds predate the development of the grove. According to Goodwin, square Wc might have contained “an old and forgotten Eshu and shrine”.200

Artefact images

63Below are a series of photographs taken by W. Fagg of artefacts obtained from the grove both by Murray, in 1943, and the 1953 team (Figs. 12 to 17). Artefact processing largely occurred at the Atobatele House, a residence of the Ooni. W. Fagg’s photograph with the largest number of terracotta fragments is shown in Fig. 13. In this image, the two broken heads (furthest left, front; FW T237 and T238), the dog head (far right, back; excavated by Murray; FW T242), and the ram (far left, back; FW T245) are observable. According to a delegation of priests of the grove that met B. Fagg in 1953, the animal terracottas represented Osangangan’s ram and dog (UG, ACCN 3120/1, B. Fagg, Notes, Other, 18/1/53). Others were excavated in 1953: the larger gagged head (second from right, front; from Square Td3; FW T255), the “right side of Oni’s head” (in B. Fagg’s words; very front, centre-left), and the ear (very front, centre-right). Willett recorded that the smaller gagged head (far right, front; FW T249) was brought to the palace by Murray: no record of its location was available, but Willett speculated that it was from Osangangan Obamakin – a suggestion strongly supported by W. Fagg’s decision to photograph it with other items from the grove in 1953 as shown here. The broken face in the very middle is FW T740, likely from the Iwinrin Grove. See the figure text for descriptions and attributions of the terracotta objects shown in the other photographs (Figs. 12, 14, 15, 16, 17).

Figure 12: Partially reconstructed fragments of a terracotta figure of a diseased man taken from the main shrine

Figure 12: Partially reconstructed fragments of a terracotta figure of a diseased man taken from the main shrine

By K. C. Murray in 1943, claimed to have been Aroni, the spirit of whirlwind (FW T239).

Photograph by William Buller Fagg, 1953 (400_WBF.1953.07.1) © RAI.

Figure 13: Terracotta objects, largely from Osangangan Obamakin

Figure 13: Terracotta objects, largely from Osangangan Obamakin

Several were obtained from the main shrine by K. C. Murray in 1943. Photograph by William Buller Fagg, 1953 (400_WBF.1953.07.05).

© RAI.

Figure 14: A conical head

Figure 14: A conical head

Probably taken by K. C. Murray in 1943, though he does not mention it in his writing. According to the delegation of priests that met B. Fagg in 1953, this represented Awunrin Olojo, wife of Osangangan (UG, ACCN 3120/1, B. Fagg, Notes, Other, 18/1/53). FW T246. Photograph by William Buller Fagg, 1953 (400_WBF.1953.07.06).

© RAI.

Figure 15: Terracotta objects from the grove

Figure 15: Terracotta objects from the grove

The head (left; FW T236) was said to be a representation of Osangangan himself and to have been taken from the main shrine by K. C. Murray in 1943. The figurine depicts a kneeling man with an animal and was presumably collected by K. C. Murray in 1943 although he does not mention it in his writing (FW T244). Photograph by William Buller Fagg, 1953 (400_WBF.1953.07.09).

© RAI.

Figure 16: A decorated globular pot

Figure 16: A decorated globular pot

Taken from the main shrine by K. C. Murray in 1943, apparently used to hold water for sacrifice (FW T243). Photograph by William Buller Fagg, 1953 (400_WBF.1953.07.10).

© RAI.

Figure 17: Terracotta fragments

Figure 17: Terracotta fragments

Probably all from the grove. Back left: appears to be FW T253, part of a head and ear, excavated from the main shrine (Square Td9) in 1953. The other fragments (four of which appear to be feet) may be other items from the main shrine excavations in 1953. Photograph by William Buller Fagg, 1953 (400_WBF.1953.08.08).

© RAI.

Discussion: Of shafts, walls, and groves

  • 201 For example, see O. Myers, 1967, p. 67.
  • 202 Despite growing evidence that at least sections of the wall network are recent in date. It is worth (...)
  • 203 See B. Lawal, 2008.

64It can be concluded from this excavation data that Osangangan Obamakin represents two or more phases of occupation: a (probably) much earlier deep shaft, speculated to have been used for burial by B. Fagg, followed by later, perhaps 19th-century, (re)discovery of older features, and the development of the site into a sacred grove with shrines to multiple deities, and associated with the peopling of the world. The grove, its focal point apparently being the deep shaft associated with “heaven”, developed into a space where examples of the early art of Ife were kept, curated, and used as part of the grove’s larger role in the city, perhaps associated with health and healing. The diverse range of artistic fragments, coupled with its association with more recent pottery in these shallow deposits (see discussion of braided strip roulette above), suggests that, in spite of its largely being made in the earlier period, it was indeed brought to the site (perhaps from diverse sources) at a later date. This later assembling of older objects in sacred groves is a pattern observable in the archaeological records of other such sites in Ife, such as Olokun Walode and Igbo Obameri.201 From the evidence, it is not wholly clear the extent to which continuities did or did not exist in the use of the space between these early and late periods given the lack of radiometric dates obtained by these researchers; however, the shallow nature of the upper deposits and preponderance of recent ceramic forms and styles imply that the later phase was relatively short-term and recent in date rather than a continuous stratigraphy extending back to the early city. A further key to interpreting much of the archaeology of the grove may be its relationship with the wall, its purported gateway here, and the associated old market (all of which may have been 18th- and/or 19th-century features), notably the role it may have played as part of a regulatory, liminal boundary within this system of fortifications.202 The grove is a liminal site in more ways than one: where heaven met earth via the old shaft feature(s) – where Osangangan met his wife – and where Ife met the world beyond on the threshold of one of its walls. This series of connected dualities may not have been coincidental, part of the “dialectics of twoness” central to Yorùbá metaphysics,203 and especially associated with sacred places. Both B. Fagg and Goodwin suggested that the site was once a fortified place, perhaps a gateway of the town wall. As Goodwin wrote:

  • 204 UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, p. 3.

The relationship between the Os. Ob. Shrine and the old encircling wall of Ife is clearly important, and the casual way in which at least two informants spoke of it suggests that the relationship is habitual and expected. If possible, an attempt should be made by aerial survey and groundreccy [reconnaissance], to locate the whole wall, to isolate it from the Modakeke wall, and to locate the old gateways of both cities. This should allow us to convince ourselves whether or not gateways were always the foci of shrines.204

  • 205 E.g., see P.J. Darling, 1984, 1997; P.C. Lloyd, 1959; T.L. Evans, 2025b, p. 317, 318; C. Schepers e (...)
  • 206 S.P. Blier, 2015, p. 82, 148.
  • 207 R. Abiodun, 2014, p. 52–53, 241.
  • 208 A. Ogundiran, 2020, p. 501.
  • 209 See A.J.H. Goodwin, 1957, p. 71; J. Egharevba, 1960, p. 14–15; P.J. Darling, 1984, p. 40, 49–50; P. (...)
  • 210 K.C. Macdonald, 2012, p. 353–356.

65The archaeologists’ informants thus implied that shrines and sacred groves were commonly consciously associated with the town walls and gateways (and see Fig. 1). Such a close association was not only a typical feature of Ile-Ife’s 19th-century urban landscape: it is also a relationship seen across Yorùbá and other West African contexts205 – an interesting dimension that may help tentatively and partially explain why, in this case, Osangangan Obamakin’s speculated associations are with disease and healing206 and perhaps forms of sanction against criminals.207 Many sacred groves are associated with quarantine and curing of ill people, while town walls, especially the gateways where roads pass through them, have often been characterised as important barriers and filters regulating the spread of infectious diseases.208 Chouin (2022, p. 140–142) has argued that sacred groves, across West Africa, shielded communities from incoming disease, while Norman (2014, p. 14) has explored logics surrounding the regular placement of shrines on roads in the Fon region, which may have held people traversing them to oaths and pledges, at the risk of sanction. Warnier (2007, p. 140–150) recorded that linear earthwork entranceways at the Mankon Kingdom (Cameroon) were sites where medicines were placed, as a means of regulating movements of people and things through the thresholds: a protection of people within from hostile forces. Evidence from festivals and oral tradition in Benin City also suggests practices of inserting medicines and offerings at linear earthwork gateways to protect people within from dark forces.209 Olupona (2011, p. 34–35) has commented on how the walls of Ife, and associated shrines to deities such as Esu, helped regulate the dangerous powers of the bush beyond to protect the town. We might then term town walls and their entranceways “spiritual filters”, both protecting political and economic interests while also regulating darker, more esoteric, forces.210

  • 211 B. Agbaje-Williams, 2005; A.W. Gurstelle, 2013.
  • 212 UG, ACCN 3120/1, B. Fagg, Notes, Other, 3/3/53
  • 213 R. Abiodun, 2014, p. 52–53, 241.

66From this, it might be tentatively hypothesised that the close spatial association between the grove and wall was connected to trapping dangerous powers at a threshold entry point to prevent them from entering further into Ife and causing epidemics, natural disasters, and other such calamities. Shrines such as those to Osangangan Obamakin, Esu, and Ogun – all important liminal deities – and their associated art objects may have been technologies associated with these regulatory functions. Similarly, the sacred trees in the grove, such as the iroko and silk-cotton trees, may also have played a role in this, they often being associated with the protection of communities from dark forces.211 Osangangan himself was said to have inflicted leprosy on his son Obalufon and commanded a “spirit of the whirlwind”, further associating the grove with powers of disease and natural disaster (and perhaps protection against these).212 Another possibility is that art at the grove depicting afflictions, and perhaps criminality,213 were related to ensuring that those passing into Ife followed community laws and regulations, or else risked sanction by disease affliction and/or execution: a means of maintaining threshold covenants. Such boundary features might have played an important role during the 19th-century conflicts and tensions between Ife and Modakeke.

  • 214 F. Willett, 2004, II.2; G. Chouin, A. Ogunfolakan, 2015, 2018.
  • 215 E.g., P. Shinnie, 2005; M. Posnansky, 2015.

67Of course, the “chicken and egg” dimensions here are somewhat unclear. The deep shaft, probably a feature of the early city, likely predated the sacred grove developed around it, with the wall also postdating this feature. As such it is possible that, at the time of the wall’s construction, the site was already sacred ground, perhaps with an old road constructed to run past it so as to reap its benefits, which the wall incorporated when it was constructed (perhaps the reason for the “bay” created by the “recurving sweep” of the wall that Goodwin suggested was present). Thus, this may be a case of a wall and entryway being created in relation to an already sacred site and road, rather than their being constructed first with a sacred grove later developed to serve them. Perhaps some of the structural evidence suggested by the excavations was from structures continually added, including after the earthwork’s construction (as houses associated with operation of its entrance), or perhaps these structures predated the transformation of the site into a grove. This sequence may resemble that at Ita Yemoo, near Ife’s earthworks, where excavations have shown the presence of possible early burial shafts, potsherd pavements, and other early features that developed into a sacred grove landscape (perhaps based on sacralisation of old artefactual features), with part of the wall then constructed on top of these features and incorporating the grove.214 As noted above, it is certainly the case elsewhere in West Africa that deep anthropogenic shafts and their associated ancient artefacts have been interpreted as important sites of ancestral origin/genesis, and were developed into shrine and sacred grove features by subsequent communities. Therefore, although there are specific logics and narratives unique to Ife here, it is clearly part of a broader phenomenon of how various societies across West Africa reinterpreted, and reengaged with, early ancestral subterranean architecture and the old material culture within or around these features.215 Further study of archaeological data will be critical in helping to disentangle these complex sequences and relationships, and their significances.

Conclusions

68It is hoped that this analysis helps demonstrate the importance of combining the archives of different archaeologists to create more fully fleshed-out pictures of excavations, which can then help better clarify the nature of archaeological features and finds and their relevance within their larger landscapes. Revisiting these old unpublished materials is particularly rewarding as a means of addressing new questions, with more recent evidence across disciplines aiding new interpretations. It is further hoped that this work shows the merit of drawing upon archaeological archives to better understand the histories, roles, and nature of sacred groves in West Africa. Lastly, these archives help us tell a fuller, more accurate history of archaeology in the late colonial period that better appreciates the often significant role played by African participants both in undertaking fieldwork and shaping interpretations, from “Technical Assistants” and well-diggers to local priestly, chiefly, and royal authorities.

Haut de page

Bibliographie

List of archives consulted

Papers of Professor Frank Willett, University of Glasgow Archive Services, Glasgow, UK, GB248 ACCN 3120.

The Goodwin Papers, University of Cape Town Libraries: Special Collections, Cape Town, South Africa BC 290.

William Buller Fagg Collection, Royal Anthropological Institute of Great Britain and Ireland, London, UK.

Melville J. and Frances S. Herskovits Papers, Schomburg Center for Research in Black Culture, New York, USA, Sc MG 261.

Kenneth Murray, List Of Places In Ife Known Or Believed To Contain Antiquities: Compiled November, 1943, Revised October, 1948, the National Museum, Lagos, Nigeria.

Études

Abiodun, R., 2014, Yoruba art and language: Seeking the African in African art, Cambridge, Cambridge University Press. DOI: https://doi.org/10.1017/CBO9781107239074

Aderemi, A.T., 1937, “Notes on the city of Ife”, Nigeria Magazine, 12, p. 3–6.

Agbaje-williams, B., 2005, “Clothed ritual trees: An insight into Yoruba religious thought” in B. Agbaje-williams and E.P. Renne (eds.) Yoruba religious textiles: Essays in honour of Cornelius Oyeleke Adepedgba, Ibadan, Bookbuilders Ibadan, p. 157–187.

Aliyu, N., 2011, Katsina city-wall (Ganuwar Katsina): The history of a monument, its preservation and collapse c. 1436—2008, MA dissertation, Usmanu Danfodiyo University.

Anon, n. d., Annual report of the Antiquities Service for the year 1952-53, Lagos, Federal Government Printer.

Babalola, A.B., 2017, “Ancient history of technology in West Africa: The indigenous glass/glass bead industry and the society in early Ile-Ife, southwest Nigeria”, Journal of Black Studies, 48 (5), p. 501–527. DOI: 10.1177/0021934717701915

Babalola, A.B., Dussubieux, L., Mcintosh, S.K., Rehren, T., 2018, “Chemical analysis of glass beads from Igbo Olokun, Ile-Ife (SW Nigeria): New light on raw materials, production, and interregional interactions”, Journal of Archaeological Science, 90, p. 92–105. DOI: https://doi.org/10.1016/j.jas.2017.12.005

Babalola, A.B., Mcintosh, S.K., Dussubieux, L., Rehren, T., 2017, “Ile-Ife and Igbo Olokun in the history of glass in West Africa”, Antiquity, 91 (357), p. 732–750. DOI: https://doi.org/10.15184/aqy.2017.80

Blier, S.P., 2012, “Cosmic references in ancient Ife” in C.M. Kreamer (Ed.) African cosmos: Stellar arts, Washington, DC, Smithsonian National Museum of African Art.

Blier, S.P., 2015, Art and risk in ancient Yoruba: Ife history, power, and identity, c.1300, Cambridge, Cambridge University Press, p. 204–215. DOI: https://doi.org/10.1017/CBO9781139128872

Bowen, T.J., 1857, Adventures and missionary labors in several countries in the interior of Africa from 1849 to 1856. Charleston, Southern Baptist Publication Society.

Burton, R.F., 1864, A mission to Gelele: King of Dahome (Vol. 1), London: Tinsley Brothers.

Chapelain de Seréville-Niel C., Chouin, G., Ogunfolakan, A., 2024, “Retour sur l'archéologie mortuaire d'époque médiévale dans le sud-ouest du Nigéria: Approches forensiques d'une mortalité discrète”, Afrique: Archéologie & Arts, 20, p. 37–62. DOI: https://doi.org/10.4000/12meo

Chouin, G., 2002, “Sacred groves as historical and archaeological markers in southern Ghana”, Ghana Studies, 5, p. 177–196. DOI: https://doi.org/10.3368/gs.5.1.177

Chouin, G., 2008, “Archaeological perspectives on sacred groves in Ghana” in M.J. Sheridan, C. Nyamweru (eds.), African sacred groves: Ecological dynamics & social change, James Currey; Ohio University Press; Unisa Press, p. 178–194.

Chouin, G., 2009, Forests of power and memory: An archaeology of sacred groves in the Eguafo polity, southern Ghana (c. 500-1900 A.D.), Doctoral Dissertation, Syracuse University.

Chouin, G., 2022, “Environments of health and disease in tropical Africa before the colonial era” in L. Jones (ed.), Disease and the environment in the medieval and early modern worlds, Routledge, p. 127–158. DOI: https://doi.org/10.4324/9780429055478

Chouin, G., Ogunfolakan, A., 2015, Ife-Sungbo Archaeological Project: Preliminary report on excavations at Ita Yemoo, Ile-Ife, Osun State and on rapid assessment of earthwork sites at Eredo and Ilara-Epe, Lagos State, June-July 2015, William & Mary, p. 1–34.

Chouin, G., Ogunfolakan, A., 2018, Ife-Sungbo Archaeological Project: Quadriennal Report 2015-2018 including a new preliminary report on excavations at Oduduwa College II and at OAU International School, Ile-Ife, Osun State, May-June 2018, William & Mary.

Chouin, G., Ogunfolakan, A., Roth, L., Logan, A., Acquaviva, C., 2023, Mission archéologique d'Ife-Sungbo: Rapport de la saison VIII, 2023, Ministère de l'Europe et des Affaires étrangères.

Connah, G., 1964, Polished stone axes in Benin, Federal Department of Antiquities, Nigeria.

Connah, G., 1975, The archaeology of Benin: Excavations and other researches in and around Benin City, Nigeria, Clarendon Pr.

Darling, P.J., 1984, Archaeology and history in southern Nigeria, Parts i and ii: The ancient linear earthworks of Benin and Ishan. Part i: Fieldwork and background information, Part ii: Ceramic and other specialist studies, Cambridge Monographs in African Archaeology.

Darling, P.J., 1997, “Sungbo’s Eredo: Africa’s largest monument”, The Nigerian Field, 62 (3–4), p. 113–129.

David, N., 1992, “The archaeology of ideology: Mortuary practices in the Central Mandara Highlands, northern Cameroon” in J. Sterner and N. David (eds.), An African commitment: Papers in honour of Peter Lewis Shinnie, Calgary, University of Calgary Press, p. 181–210.

Davidson, J.M., 2012, “They laid planks ’crost the coffins’: The African origin of grave vaulting in the United States”, International Journal of Historical Archaeology, 16 (1), p.86–134. DOI: https://doi.org/10.1007/s10761-012-0170-5

Dennett, R.E., 1910, Nigerian studies or, The religious and political system of the Yoruba, Macmillan and Company, Limited.

Egharevba, J., 1960, A short history of Benin (3rd ed.), Ibadan, Ibadan University Press.

Egharevba, J., 1968, A short history of Benin (4th ed.), Ibadan, Ibadan University Press.

Euba, T., 1984, “The Oòni of Ifè's Arè Crown and the concept of divine head”, Nigeria Magazine, 53 (1), 1–18.

Evans, T.L., 2023, “A copper alloy head count: Contextualising and accounting for the Wúnmọníjẹ̀ compound discoveries of Ilé-Ifẹ̀, Nigeria”, Afrique: Archeologie et Arts, 19, p. 67–100. DOI: https://doi.org/10.4000/aaa.4918

Evans, T.L., 2024a, “Collecting copper alloy portrait heads”, Journal of the History of Collections, 36 (1), p. 101–116. DOI: https://doi.org/10.1093/jhc/fhad042

Evans, T.L., 2024b, “Debating the Bascom controversy: Revisiting the expropriation and repatriation of the Ifẹ̀ bronzes”, Africa: Journal of the International African Institute, 94 (1), p. 186–214. DOI: https://doi.org/10.1017/S0001972024000032

Evans, T.L., 2025a, Earthwork landscapes of protection and regulation: Chronology, construction, and meaning at Sungbo's Eredo, Africa's largest known structure, Doctoral Dissertation, College of William & Mary. DOI: https://dx.doi.org/10.21220/s2-x5gw-pq52

Evans, T.L., 2025b, “The walls of Benin reconsidered: Interpreting West African linear earthworks using A.J.H. Goodwin's unpublished excavation data”, African Archaeological Review, 42, p. 309–332. DOI: https://doi.org/10.1007/s10437-025-09627-7

Fagg, B., 1953, “Some archaeological problems at Ife”, Conférence Internationale Des Africanistes de l’Ouest, Ve Réunion, Abidjan, Compte Rendu, p. 125–126.

Falola, T., 1989, “The Yoruba toll system: Its operation and abolition”. The Journal of African History, 30 (1), 69–88.

Falola, T. (ed.), 2013, Èṣù: Yoruba god, power, and the imaginative frontiers, Carolina Academic Press.

Folorunso, C.A., 2011, “The practice of archaeology in Nigeria”, in L.R. Lozny (ed.), Comparative archaeologies: A sociological view of the science of the past, Springer, p. 807–826. DOI: https://doi.org/10.1007/978-1-4419-8225-4

Forbes, F.E., 1851, Dahomey and the Dahomans; being the journals of two missions to the king of Dahomey, and residence at his capital, in the year 1849 and 1850 (Vol. 1), Longman, Brown, Green, and Longmans.

Frobenius, L., 1913, The voice of Africa: Being an account of the travels of the German Inner African Exploration Expedition in the years 1910-1912 (Vol. 1), Hutchinson & Company.

Garlake, P., 1974, “Excavations at Obalara's Land, Ife: An interim report”, West African Journal of Archaeology, 4, p. 111–148.

Garlake, P., 1977, “Excavations on the Woye Asiri family land in Ife, western Nigeria”, West African Journal of Archaeology, 7, p. 57–96.

Girshick, P.D., 1995, The art of Benin (2. rev. ed), British Museum Press.

Goodwin, A.J.H., 1953, “The origin of maize”, The South African Archaeological Bulletin, 8 (29), p. 13–14.

Goodwin, A.J.H., 1957, “Archaeology and Benin architecture”, Journal of the Historical Society of Nigeria, 1 (2), p. 65–85.

Goodwin, A.J.H., 1958, “Walls, paving, water-paths and landmarks”, Odu, 6, p. 45–53.

Goody, J., 1962, Death, property and the ancestors: A study of the mortuary customs of the LoDagaa of West Africa. London, Tavistock Publications.

Gurstelle, A.W., 2013, “Sacred trees of the Savè Hills cultural landscape”, University of Michigan Working Papers in Museum Studies, p. 1–18.

Hambly, W.D., 1935, Culture areas of Nigeria: Vol. XXI, No.3, Field Museum of Natural History.

Haour, A., Keita, A., 2010, “Folded strip roulette / Roulette de fibre plate pilée”, in African Pottery roulettes past and present: Techniques, identification and distribution, Oxbow Books, p. 169–176.

Herskovits, M., 1938, Dahomey: An ancient West African kingdom. Vol. 1. New York, J.J. Augustin.

Idowu, E.B., 1962, Olodumare God in Yorùbá belief, Longman.

Jeffreys, M.D.W., 1953, “Pre-Columbian maize in Africa”, Nature, 172 (4386), p. 965–966.

Kense, F., 1990, “Archaeology in Anglophone West Africa” in P. Robertshaw (ed.), A history of African archaeology, James Currey.

Lander, R., lander, J., 1832, Journal of an expedition to explore the course and termination of the Niger with a narrative of a voyage down that river to its termination, John Murray.

Law, R., 1989, “‘My head belongs to the king’: On the political and ritual significance of decapitation in pre-colonial Dahomey”, The Journal of African History, 30 (3), p. 399–415.

Law, R., 2011, “West Africa’s discovery of the Atlantic”, International Journal of African Historical Studies, 44 (1), p. 1–25.

Lawal, B., 2008, “Èjìwàpò: The dialectics of twoness in Yoruba art and culture”, African Arts, 41 (1), p. 24–39. DOI: https://doi.org/10.1162/afar.2008.41.1.24

Lloyd, P.C., 1959, “Sungbo’s Eredo”, Odu, 7, p. 15–22.

Macdonald, K.C., 2012, ““The least of their inhabited villages are fortified”: The walled settlements of Segou”, Azania: Archaeological Research in Africa, 47 (3), p. 343–364. DOI: https://doi.org/10.1080/0067270X.2012.707478

Merkyte, I., randsborg, K., 2009, “Graves from Dahomey: Beliefs, ritual and society in ancient Bénin”, Journal of African Archaeology, 7 (1), p. 55–77. DOI: http://dx.doi.org/10.3213/1612-1651-10126

Moloney, L., 2004, June 30, “Obituary: Liman Ciroma”, The Guardian.

Moody, H.L.B., 1969, The walls and gates of Kano City, Department of Antiquities, Federal Republic of Nigeria.

Murray, K.C., Willett, F., 1958, “187. The Ore Grove at Ife, Western Nigeria”, Man, 58, p. 137.

Myers, O., 1967, “Excavations at Ife”, The West African archaeological newsletter, 6, p. 6–11.

Norman, N.L., 2014, “Liquid gods within cellular landscapes: Micro-monuments, ethnic pluralism and the political landscape of Hueda, Bénin, 1650–1727”, Azania: Archaeological Research in Africa, 49 (1), p. 3–20. DOI: https://doi.org/10.1080/0067270X.2013.868620

Ogundipe, A. (with Olupona, J.K.), 2012, Èșù Elegbára: Change, chance, uncertainty in Yorùbá mythology, Kwara State University Press.

Ogundiran, A., 2002, “Of small things remembered: Beads, cowries, and cultural translations of the Atlantic experience in Yorubaland”, The International Journal of African Historical Studies, 35 (2/3), p. 427–457. DOI: https://doi.org/10.2307/3097620

Ogundiran, A., 2015, “Pioneers of archaeological thought and practice in postcolonial Nigeria” in S. Wynne-jones, J.B. Fleisher (eds.), Theory in Africa, Africa in theory: Locating meaning in archaeology, Routledge, p. 90–110.

Ogundiran, A., 2020a, “Managing epidemics in ancestral Yorùbá towns and cities: “Sacred groves” as isolation sites”, African Archaeological Review, 37 (3), p. 497–502. DOI: 10.1007/s10437-020-09407-5

Ogundiran, A., 2020b, The Yoruba: A new history, Indiana University Press.

Ogundiran, A., Ogunfolakan, A., 2017, “Colonial modernity, rituals and feasting in Odùduwà Grove, Ilé-Ifẹ̀ (Nigeria)”, Journal of African Archaeology, 15 (1), p. 77–103. DOI: 10.1163/21915784-12340004

Olupona, J.K. (2011). City of 201 gods: Ilé-Ifè in time, space, and the imagination, University of California Press.

Ozanne, P., 1969, “A new archaeological survey of Ife”, Odu, 1, p. 28–46.

Polet, J., 1983, “Localisation d’une nécropole du deuxième millénaire de notre ère en zone lagunaire de basse Côte d’Ivoire grâce à un indicateur végétal : Dracaena Arborea L. Rennes”, Revue d’Archéométrie, 7, p. 1–10.

Parrinder, G., 1967, The story of Ketu, an ancient Yoruba kingdom, Ibadan, Ibadan University Press.

Pemberton, J.A., 1975, “Eshu-Elegba: The Yoruba trickster god”, African Arts, 9 (1), p. 20–27.

Posnansky, M., 2015, “Begho: Life and times”, Journal of West African History, 1 (2), p. 95–118. DOI: 10.14321/jwestafrihist.1.2.0095

Rattray, R.S., 1923, Ashanti, Oxford, Clarendon Press.

Roth, H.L., 1903, Great Benin: Its customs, art and horrors, Halifax, F. King & Sons.

Roth, L., Chouin, G., Ogunfolakan, A., 2021, “Lost in space? Reconstructing Frank Willett’s excavations at Ita Yemoo, Ile-Ife, Nigeria: Rescue excavations (1957–1958) and Trench XIV (1962–1963)”, Afrique: Archéologie et Arts, 17, p. 77–114. DOI: https://doi.org/10.4000/aaa.3328

Schepers, C., Hubert, J., Edward-Ekpu, U., Folorunso, C.A., Fonseca, S., Lequesne, C., Omogbai, O., Linstädter, J., 2025, “Current conditions of the Iya in Benin City, the gates and future preservation strategies”, African Archaeological Review, p. 1–19. DOI: https://doi.org/10.1007/s10437-025-09630-y

Sheridan, M.J., Nyamweru, C. (eds.), 2008, African sacred groves: Ecological dynamics & social change, James Currey; Ohio University Press; Unisa Press.

Shinnie, P., 2005, “Early Asante and European contacts”, Journal des africanistes, 75-2, p. 25–42. DOI: https://doi.org/10.4000/africanistes.113

Soper, R., 1983, “Roulette decoration on African pottery: Some technical distinctions and misconceptions and its significance in culture-historical reconstruction”, The Proceedings of the 9th Congress of the Pan-African Association of Prehistory and Related Studies, p. 28–35.

Soper, R., 1985, “Roulette decoration on African pottery: Technical considerations, dating and distributions”, The African Archaeological Review, 3, p. 29–51.

Stanton, W.R., Willett, F., 1963, “150. Archaeological evidence for changes in maize type in West Africa: An experiment in technique”, Man, 63, p. 117.

Warnier, J.-P., 2007, The pot-king: The body and technologies of power, Brill.

Willett, F., 1962, “The introduction of maize into West Africa: An assessment of recent evidence”, Africa, 32 (1), p. 1–13.

Willett, F., 1967, Ife in the history of West African sculpture, McGraw-Hill.

Willett, F., 2004, The art of Ife: A descriptive catalogue and database, The Hunterian Museum and Art Gallery, The University of Glasgow.

Haut de page

Notes

1 F. Willett, 2004, I.2.

2 A.T. Aderemi, 1937.

3 Ciroma’s contributions as an archaeologist in Nigeria were especially important, he going on to work at other key sites such as Benin City and Igbo Ukwu (L. Moloney, 2004).

4 F. Willett, 2004, I.2; S.P. Blier, 2015.

5 T.L. Evans, 2024a; 2024b.

6 A.B. Babalola et al., 2018; G. Chouin, A. Ogunfolakan, 2018; A. Ogundiran, 2020; T.L. Evans, 2023.

7 F. Willett, 2004, I.2.

8 C.A. Folorunso, 2011; F. Kense, 1990; A. Ogundiran, 2015.

9 G. Chouin, 2008; A. Ogundiran, A. Ogunfolakan, 2017; M.J. Sheridan, C. Nyamweru, 2008.

10 Four archives provided useful material, including the three mentioned in the body text. Some additional unpublished descriptive information about Osangangan Obamakin was obtained by the author from a fourth notable archive: the Melville J. and Frances S. Herskovits Papers, at the Schomburg Center for Research in Black Culture in New York. See below for further information.

11 Yorùbá orthography: Ọ̀sángangan Ọbámákin (R. Abiodun, 2014).

12 Throughout this article, where possible, I identify finds from Osangangan Obamakin with Frank Willett’s (2004) system of classifying the art of Ife, although some of the objects excavated in 1953 were not given attributions by him. References to this are abbreviated for convenience as, e.g., FW S110: his initials followed by S110 – his ID number for the object.

13 F. Willett, 2004, II.5.

14 This is contained in a document entitled LIST OF PLACES IN IFE KNOWN OR BELIEVED TO CONTAIN ANTIQUITIES: COMPILED NOVEMBER, 1943, REVISED OCTOBER, 1948. It is an invaluable record of Kenneth Murray’s reconnaissance of heritage sites in Ife and elsewhere housed at the National Museum in Lagos, and helped guide the work of B. Fagg and Goodwin in 1953. I am grateful to Dr. Léa Roth for sharing it with me.

15 This is presumably a reference to the Ibadan Road, north of the grove, or, alternatively, to a road running through Modakeke to the south of the grove. Most reports suggest it was about the same distance either way, through areas of forest.

16 As discussed below, the site was close to a river, apparently separating Ife proper from Modakeke. The pool that Murray mentions here is probably the same as that described by Goodwin, who observed “… a clump of bamboo with stagnant water at its foot” near “cross-paths”. This appears to have been a sacred location, sacred perhaps to Esu, as the 1953 team observed a “sacrifice” here consisting of one of their finds taken by a local: “… we found a new “sacrifice”, one of our own pots from the shrine, still with our number on it, but carefully bored through at the bottom. It had been filled with bamboo leaves and set alight, but had only half burnt.” (UCT, J2.23, Goodwin, Part VII).

17 Such deep shafts are common in the archaeology of early Ife, seen at a number of sites, and probably the source of a lot of the legends pertaining to underground tunnels and passageways in the city. See discussion below.

18 Unpublished evidence suggests that this feature was associated with Osangangan, only peripherally with Esu. This is an important detail, as subsequent publications citing this early work have tended to repeat the idea that the entire feature is associated with Esu. This is discussed further below. The correct spelling and grammar (excluding tone markers) is “Ojubo Esu” – the point of propitiation for Esu. This article retains the spelling of the archaeologists to be authentic to the archival record and does so with double quotation marks to emphasize this.

19 A.J.H. Goodwin, 1958, p. 46.

20 A.J.H. Goodwin, 1958, p. 47.

21 Later authors (e.g., P. Ozanne, 1969; F. Willett, 1967, 2004) do not represent this area as a gateway point in Ife’s walls on maps, nor do they even consider this idea in spite of its being present in writings pertaining to the 1953 fieldwork. Ozanne did not directly observe the wall at Osangangan Obamakin but recorded that the line of the wall was “strongly evidenced” there, based on survey by Willett. He speculated that the wall here was part of the earliest phases of (“Medieval”) construction (P. Ozanne, 1969, p. 29–30, 44), though the sources and basis for this attribution are not clear.

22 P. Ozanne, 1969.

23 See G. Chouin, A. Ogunfolakan, 2015, 2018 for the archaeological evidence. Making reference to oral tradition, Ozanne reported that “traditions of the town are adamant in referring all of the walls to very recent times in comparison with the ancient works of art–the first half of the nineteenth century, with some perhaps a couple of generations earlier” (1969, p. 28); however, these traditions referred only to “ruins from this later period”, while Frank Willett and Catherine Fagg observed sections of what they identified as ruined walls which were not regarded as walls by local people, who explained them in alternative “imaginative ways except the true one”. This led Ozanne to interpret them as belonging “to times beyond surviving reliable traditions”. The basis behind Ozanne’s formulations on this is difficult to gauge given the lack of reliable evidence from absolute dating, and the destruction of many of these walls which he considered to be “Medieval” in date. Beyond Ozanne’s claims, Ife’s early prominence as an urban centre, and the presence of town walls reported in contemporary documentary sources pertaining to a number of “Medieval” Sahelian West African towns, is perhaps circumstantial evidence that there was likely to be some sort of wall fortification in the early period.

24 S.P. Blier, 2015, p. 20, 172.

25 UG, ACCN 3120/1, B. Fagg Diary, 1/1/53, 3/1/53, p. 1–2; UCT, J2.23, Goodwin, Part IV, p. 2.

26 UCT, J2.23, Goodwin, Part IV, p. 6–7.

27 P. Ozanne (1969, p. 29–30) seemingly implies that there was no ditch at this section of the wall (apparently based on Willett), but this may be conjectural. A.J.H. Goodwin (1958, p. 45), however, appears to have been of the opinion that a ditch accompanied the entirety of the wall where it was surveyed and visited in 1953.

28 This sectional arrangement is typical for sacred groves in Ife (and elsewhere), as outlined in other early and mid-20th-century sources (e.g., R.E. Dennett, 1910, p. 19–23; W.D. Hambly, 1935, p. 466; K. Murray, F. Willett, 1958, p. 137). As Idowu (1962, p. 127) writes, they were divided into “apartments” with different levels of access depending on the person’s identity and level of initiation. The most sacred, and restricted, parts of these groves were the places that held the most sacred emblems: at Osangangan Obamakin, this would have been the main shrine within the “chancel” (see below), which, traditionally, only a small number of privileged people could access. Shrines (ojubo) were conceptualised as “faces” of deities, through which they could be accessed.

29 UCT, J2.23, Goodwin, Part IV, p. 7.

30 UG, ACCN 3120/1, B. Fagg Diary, 17/1/53, p. 7; UCT, J2.23, Goodwin, Part IV, p. 7.

31 R.E. Dennett, 1910, p. 139. For other examples of this type of tree being used as boundary markers elsewhere in Africa, see J. Polet, 1983 (Côte d’Ivoire) and G. Chouin, 2002 (Ghana).

32 Bamboo patches have also been associated with anthropogenic features elsewhere in West Africa. For example, in southern Ghana, Chouin was informed that bamboo groves were connected with cemeteries, marketplaces, yam farms, and other human spaces where bamboo sticks deposited by various human activities (e.g., farming or architecture) would germinate (G. Chouin, 2009, p. 199, 337, 348, 357–358, 365–366, 350, 448–452). This tentatively lends further credence to the idea that there was an old marketplace and/or gateway at Osangangan Obamakin in the past, when it was perhaps a more intensively inhabited area. Bamboo patches have also been observed in close association with linear earthworks elsewhere in Nigeria, such as at Sungbo’s Eredo in Ogun State, and in the Benin-Ishan region (see T.L. Evans, 2025a, p. 226, 594–595).

33 UG, ACCN 3120/1, B. Fagg Diary, 5/1/53, p. 2.

34 UCT, J2.23, Goodwin, Part IV, p. 2.

35 A.J.H. Goodwin, 1958, p. 46.

36 UCT, J2.23, Goodwin, Part V, p. 4.

37 P. Ozanne, 1969; F. Willett, 1967, 2004.

38 T. Falola, 1989, p. 73–74.

39 Melville J. and Frances S. Herskovits Papers, Schomburg Center for Research in Black Culture, New York, Sc MG 261, Box 28 “Research Material: Field Trip--Africa, 1953 Diary”, p. 228.

40 UCT, J2.23, Goodwin, Part VI.

41 UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, p. 3–4.

42 Ibid., p. 4.

43 See T. Falola, 2013; A. Ogundipe, 2012; J.A. Pemberton, 1975.

44 UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, p. 3.

45 UG, ACCN 3120/1, B. Fagg Diary, 17/1/53, p. 7.

46 UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, p. 4.

47 UG, ACCN 3120/1, B. Fagg Diary, 7/1/53, p. 4; 1 yard = 0.9144 metres or 3 feet.

48 F. Willett, 2004.

49 UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, p. 4.

50 Ibid., p. 4.

51 Ibid., p. 5.

52 Note that references to “maize-cob roulette design” may be misleading, given that certain strip and string roulette decoration types have been commonly mistaken for it (see A. Haour, D. Keita, 2010; R. Soper, 1983, 1985; see also: A.J.H. Goodwin, 1953; M.D.W. Jeffreys, 1953; F. Willett, 1962; W.R. Stanton, F. Willett, 1963).

53 UCT, J2.23, Goodwin, Part V.

54 F. Willett, 2004, S110.

55 UG, ACCN 3120/1, B. Fagg Diary, 9/1/53, p. 5.

56 UG, ACCN 3120/1, B. Fagg, Notes, Other, 3/3/53.

57 Oral tradition suggested that this head was kept hidden under the inverted pot so that it would not see the sun – a concept perhaps connected with Osangangan’s reputation as having brought light to the earth. See F. Willett, 1967, p. 47, Plate 21; S.P. Blier, 2012, p. 3. Ife’s associations with the powers of light (of the sun and moon) were old, with the American Baptist missionary T.J. Bowen writing in 1857 that the sun and moon were “worshipped” there and that large “brazen” (brass) images of these were said to exist in the town at that time (see T.J. Bowen, 1857, p. 316–317).

58 UCT, J2.23, Goodwin, Part IV, p. 7.

59 UG, ACCN 3120/1, B. Fagg Diary, 6/1/53, p. 3.

60 Ibid., 9/1/53, p. 4–5.

61 Ibid., 6/1/53, p. 3; UCT, J2.23, Goodwin, Part V.

62 UG, ACCN 3120/1, B. Fagg Diary, 6/1/53, p. 3.

63 UCT, J2.23, Goodwin, Part V. As Blier (2015, p. 350–351) has documented, Osangangan Obamakin was associated with elewo (crotal bells/rattles) – his “special prerogative” – and bringing light to the earth. The regular unearthing of “bell” objects at this grove likely relates to these associations.

64 UG, ACCN 3120/1, B. Fagg Diary, 5/1/53, p. 2–3.

65 UCT, J2.23, Goodwin, Part IV, p. 7–8.

66 UG, ACCN 3120/1, B. Fagg, Notes, ‘Some Archaeological Problems at Ife’, p. 1.

67 UCT, J2.23, Goodwin, Part V; UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, p. 7.

68 UG, ACCN 3120/1, B. Fagg Diary, 8/1/53, p. 4.

69 UCT, J2.23, Goodwin, Part IV, p. 7.

70 Ibid., Part V.

71 UG, ACCN 3120/1, B. Fagg Diary, 9/1/53, p. 4–5.

72 J.A. Pemberton, 1975.

73 UG, ACCN 3120/1, B. Fagg, Notes, ‘Some Archaeological Problems at Ife’, p. 1–2.

74 UG, ACCN 3120/1, B. Fagg Diary, 8/1/53, p. 4.

75 Ibid., 9/1/53, p. 4–5.

76 UG, ACCN 3120/1, B. Fagg, Notes, ‘Some Archaeological Problems at Ife’, p. 1.

77 UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, p. 3–4.

78 UCT, J2.23, Goodwin, Part V.

79 UG, ACCN 3120/1, B. Fagg Diary, 9/1/53, p. 4–5.

80 UCT, J2.23, Goodwin, Part V.

81 UCT, J2.23, Goodwin, Part VIII.

82 UG, ACCN 3120/1, B. Fagg, Notes, ‘Some Archaeological Problems at Ife’, p. 1.

83 UG, ACCN 3120/1, B. Fagg Diary, 31/1/53, p. 11.

84 UG, ACCN 3120/1, B. Fagg, Notes, ‘Some Archaeological Problems at Ife’, p. 1.

85 UCT, J2.23, Goodwin, Part VIII.

86 UG, ACCN 3120/1, B. Fagg Diary, 31/1/53, p. 11.

87 UCT, J2.23, Goodwin, Part VIII.

88 Ibid., Part VIII.

89 UG, ACCN 3120/1, B. Fagg Diary, 3/2/53, p. 12.

90 Ibid., 3/2/53, p. 12.

91 Ibid., 3/2/53, p. 12; UCT, J2.23, Goodwin, Part VIII.

92 UG, ACCN 3120/1, B. Fagg Diary, 19/3/53, p. 28.

93 F. Willett, 2004, T248.

94 UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, p. 3.

95 UG, ACCN 3120/1, B. Fagg Diary, 6/1/53, p. 4.

96 UCT, J2.23, Goodwin, Part V.

97 UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, p. 3.

98 Annual Report of the Antiquities Service 1952–3, p. 8.

99 UCT, J2.23, Goodwin, Part V.

100 T. Euba, 1984, p. 3.

101 UG, ACCN 3120/1, B. Fagg Diary, 31/1/53, p. 11.

102 This interpretation may not be farfetched given the common occurrence of foundation sacrifices in the Yorùbá past, including in association with the entranceways of houses, villages and towns, although Idowu notes that typically people would have been buried alive (rather than decapitated) and “armed like sentries” (E.B. Idowu, 1962, p. 125). These sacrifices were referred to as A-d’-ibodè – “That which bars the gates” – referencing their role in guarding entranceways and preventing evil from entering (E.B. Idowu, 1962, p. 125). Human sacrifice, human skulls, and other body parts are commonly associated with monumental wall entranceways across West African contexts, such as in the histories of Ketu (E.G. Parrinder, 1967, p. 21–22), Dahomey (F.E. Forbes, 1851, p. 68–69, R.F. Burton, 1864, p. 290; R. Law, 1989, p. 404), Segou (K.C. Macdonald, 2012, p. 354), and the Hausa towns (H.L.B. Moody, 1969, p. 61–62; N. Aliyu, 2011, p. 48–49. 52).

103 UCT, J2.23, Goodwin, Part VI.

104 UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, p. 4.

105 UCT, J2.23, Goodwin, Part V.

106 Ibid., Part VI.

107 One simple sketch by Goodwin of a design that he refers to as “basket surface” (UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, p. 9–10) represents the ‘ladder’ pattern (as defined by P. Garlake, 1977), common in 11th to 15th century Ife assemblages and created using a specific type of carved wooden roulette. However, Goodwin’s conflation here with maize-cob designs suggests that, in this instance, he is very likely not referring to the distinctive ‘ladder’ pattern and instead to braided strip roulette designs which more closely resemble maize-cob impressions (see footnote above) and are typical of 19th century contexts in Ife (see, for example, G. Chouin et al., 2023, p. 20–21).

108 Ibid., Part VI.

109 UG, ACCN 3120/1, B. Fagg, Notes, ‘Some Archaeological Problems at Ife’, p. 2.

110 UG, ACCN 3120/1, B. Fagg, Notes, ‘Some Archaeological Problems at Ife’, p. 2. Ceramic crucible fragments, with colourful glaze left from the glass production process (a key industry of early Ife), were popular objects among later West African communities, including those who rediscovered them (R. Lander, J. Lander, 1832, p. 180–181; L. Frobenius, 1913, p. 97; F. Willett, 2004). It is probable that, like the glass beads they helped produce, they were associated with powers such as ase and thus with activation of the power and authority of people and deities (see A. Ogundiran, 2002, p. 434–437; R. Law, 2011; A.B. Babalola et al., 2017, p. 733–734).

111 UG, ACCN 3120/1, B. Fagg Diary, 4/2/53, p. 12.

112 Annual Report of the Antiquities Service 1952–3, p. 8.

113 Such stone axes were associated with certain deities and frequently occur in shrine contexts in Ife and elsewhere in West Africa (see G. Connah, 1964; L. Roth et al., 2021, p. 42).

114 UG, ACCN 3120/1, B. Fagg Diary, 17/1/53, p. 7.

115 E.g., see P. Garlake, 1974, 1977.

116 UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, p. 4.

117 Ibid., p. 5–6.

118 UG, ACCN 3120/1, B. Fagg Diary, 15/1/53 and 17/1/53, p. 6–7.

119 Ibid., 15/1/53, p. 6.

120 Ibid., 16/1/53, p. 7.

121 Ibid., 16/1/53, p. 7.

122 UCT, J2.23, Goodwin, Part VI.

123 UG, ACCN 3120/1, B. Fagg Diary, 16/1/53, p. 7.

124 UCT, J2.23, Goodwin, Part VI.

125 Ibid., Part VI.

126 Ibid., Part VI.

127 Ibid., Part VII.

128 UG, ACCN 3120/1, B. Fagg Diary, 17/1/53, p. 7. As at the Ogun shrine, the “basketry design” may be a reference to braided strip roulette designs, indicative of a relatively recent (19th-century) date.

129 Ibid., 17/1/53, p. 7.

130 Ibid., 21/1/53, p. 8.

131 Ibid., 17/1/53, p. 7.

132 Ibid., 19/1/53, p. 7; UCT, J2.23, Goodwin, Part VII.

133 UCT, J2.23, Goodwin, Part VII.

134 UG, ACCN 3120/1, B. Fagg Diary, 20/1/53 and 21/1/53, p. 8. “A.J.H.G. watched shrine wall (so carefully uncovered) ‘demolished’ by W.B.F. who subsequently said that there were not any walls anyhow!”

135 Ibid., 21/1/53, p. 8. This appears to be a case of the difference in scope of experience in field archaeology between B. Fagg and Goodwin on the one hand and W. Fagg on the other.

136 Ibid., 21/1/53, p. 8.

137 UCT, J2.23, Goodwin, Part VII.

138 Ibid., Part VII.

139 UG, ACCN 3120/1, B. Fagg Diary, 22/1/53, p. 8–9; UCT, J2.23, Goodwin, Part VII.

140 UCT, J2.23, Goodwin, Part VII.

141 UG, ACCN 3120/1, B. Fagg Diary, 24/1/53, p. 9.

142 Ibid., 26/1/53, p. 9.

143 Ibid., 26/1/53, p. 9–10.

144 Ibid., 27/1/53 and 28/1/53, p. 10.

145 Ibid., 29/1/53, p. 10–11.

146 Ibid., 29/1/53 and 30/1/53, p. 10–11.

147 Ibid., 30/1/53, p. 11.

148 Ibid., 30/1/53 and 31/1/53, p. 11.

149 Ibid., 2/2/53 to 5/2/53, p. 11–12.

150 UG, ACCN 3120/1, B. Fagg, Notes, ‘Some Archaeological Problems at Ife’, p. 2.

151 B. Fagg had assumed that the hard gritty layer below the shrine accumulation represented an undisturbed subsoil with no further material beneath it (UG, ACCN 3120/1, B. Fagg, Notes, ‘Some Archaeological Problems at Ife’, p. 2.).

152 UCT, J2.23, Goodwin, Part XI.

153 UG, ACCN 3120/1, B. Fagg Diary, 24/2/53 and 25/2/53, p. 22.

154 Ibid., 2/3/53, p. 25.

155 Ibid., 3/3/53, p. 26.

156 Ibid., 3/3/53, p. 26.

157 Ibid., 19/3/53 to 23/3/53, p. 28–30.

158 Annual Report of the Antiquities Service 1952–3, p. 8.

159 UG, ACCN 3120/1, B. Fagg, Notes, ‘Some Archaeological Problems at Ife’, p. 2.

160 UG, ACCN 3120/1, B. Fagg Diary, 28/1/53, p. 10.

161 F. Willett, 2004, II.5.

162 UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, p. 6.

163 Ibid., p. 6.

164 A.J.H. Goodwin, 1958, p. 46.

165 UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, p. 6.

166 See G. Chouin et al., 2023, p. 20–21.

167 Ibid., p. 5–6.

168 He added: “Same applies to other figurine fragments.” (UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, p. 6).

169 UCT, J2.23, Goodwin, Part VII.

170 S.P. Blier, 2015, p. 82, 148.

171 Annual Report of the Antiquities Service 1952–3, p. 8.

172 UG, ACCN 3120/1, B. Fagg, Notes, ‘Some Archaeological Problems at Ife’, p. 2.

173 E.g., see L. Frobenius, 1913, p. 23–26; M. Herskovits, 1938, p. 359. See also J. Goody, 1962, p. 142–143; N. David, 1992, p. 88–89; I. Merkyte, K. Randsborg, 2009, p. 60, 66, for bell-shaped examples. See J.M. Davidson, 2012, for a useful summary of West and Central African shaft burials, including examples which bear some resemblance to Fagg’s sketch.

174 C. Chapelain De Seréville-Niel et al., 2024.

175 G. Chouin, Pers. Comm., 2025.

176 UG, ACCN 3120/1, B. Fagg, Notes, ‘Some Archaeological Problems at Ife’, p. 4.

177 H. L. Roth, 1903, p. 43.

178 L. Frobenius, 1913, p. 93.

179 “Heaven” here is likely a reference to orun where the deities reside.

180 UG, ACCN 3120/1, B. Fagg, Notes, Other, 3/3/53. The association between old deep holes and stories of ancestral origin, genesis, and the peopling of the world appear in other West African contexts – for example, see R. S. Rattray, 1923, p. 121-132, P. Shinnie, 2005, p. 10–12, for an example at Asantemanso, Ghana, where such a hole is also associated with a sacred forest; and M. Posnansky, 2015, p. 110–111, for an example near Hani, Ghana, an ancestral hole that appeared to be a bilega (water cistern) associated with old lithics (indicating ancient settlement).

181 UG, ACCN 3120/1, B. Fagg Diary, 28/1/53, p. 10.

182 S.P. Blier, 2015, p. 1.

183 L. Frobenius, 1913, p. 307–308. Such narratives were likely often based on memory of (re)discovery of old glass beads, crucible fragments, and figurative arts in deep shafts as successive communities found archaeological treasures from earlier centuries.

184 E.B. Idowu, 1962, p. 13.

185 UG, ACCN 3120/1, B. Fagg Diary, 6/1/53, p. 3.

186 Ibid., 6/1/53 and 7/1/53, p. 3.

187 UCT, J2.23, Goodwin, Part V.

188 UG, ACCN 3120/1, B. Fagg Diary, 6/1/53, p. 3.

189 UCT, J2.23, Goodwin, Part V.

190 UG, ACCN 3120/1, B. Fagg Diary, 6/1/53, p. 3.

191 UCT, J2.23, Goodwin, Part VI.

192 F. Willett, 2004, II.5.

193 UCT, J2.23, Goodwin, Part VI.

194 UG, ACCN 3120/1, B. Fagg Diary, 16/1/53, p. 7.

195 UCT, J2.23, Goodwin, Part VI.

196 B. Fagg, 1953, p. 126.

197 UG, ACCN 3120/1, File No. 27/A/S.1. Ife Antiquities, Shafts.

198 UG, ACCN 3120/1, B. Fagg Diary, 25/2/53, p. 22.

199 Ibid., 9/1/53, p. 4–5.

200 F. Willett, 2004, II.5.

201 For example, see O. Myers, 1967, p. 67.

202 Despite growing evidence that at least sections of the wall network are recent in date. It is worth repeating, however, that we do not have absolute dates for the section of wall here, which Ozanne speculated was “Medieval”, and which conceivably (as an alternative interpretation) could have thus been much older in date, and/or reused and reconstructed in later periods (on the earlier foundations). This ambiguity in date may also apply to the abandoned market area.

203 See B. Lawal, 2008.

204 UG, ACCN 3120/1, Goodwin, ‘Ife Antiquities’, p. 3.

205 E.g., see P.J. Darling, 1984, 1997; P.C. Lloyd, 1959; T.L. Evans, 2025b, p. 317, 318; C. Schepers et al., 2025.

206 S.P. Blier, 2015, p. 82, 148.

207 R. Abiodun, 2014, p. 52–53, 241.

208 A. Ogundiran, 2020, p. 501.

209 See A.J.H. Goodwin, 1957, p. 71; J. Egharevba, 1960, p. 14–15; P.J. Darling, 1984, p. 40, 49–50; P.D. Girshick, 1995.

210 K.C. Macdonald, 2012, p. 353–356.

211 B. Agbaje-Williams, 2005; A.W. Gurstelle, 2013.

212 UG, ACCN 3120/1, B. Fagg, Notes, Other, 3/3/53

213 R. Abiodun, 2014, p. 52–53, 241.

214 F. Willett, 2004, II.2; G. Chouin, A. Ogunfolakan, 2015, 2018.

215 E.g., P. Shinnie, 2005; M. Posnansky, 2015.

Haut de page

Table des illustrations

Titre Figure 1: Reconstruction of the walls of Ile-Ife
Légende The map shows the location of Osangangan Obamakin (marked in red) and other sacred grove sites (marked in black) that are in close proximity to the town walls (shown as brown lines and dotted lines where inferred).
Crédits Based on F. Willett (1967), P. Ozanne (1969), and D. Simmonds (2002) in F. Willett (2004). Tomos Llywelyn Evans, 2025.
URL http://journals.openedition.org/afriques/docannexe/image/5491/img-1.png
Fichier image/png, 269k
Titre Figure 2: The ditch and bank of one of the Ife earthworks in 1953, with bamboo growing from the latter
Légende The location is not clear, but may be near the Iwinrin Grove to the west of Osangangan Obamakin, as B. Fagg mentions in his diary that he and his brother “explored the ancient wall” together in this area (UG, ACCN 3120/1, B. Fagg Diary, 2/3/53, p. 25).
Crédits Photograph by William Buller Fagg, 1953 (400_WBF.1953.05.05) © RAI.
URL http://journals.openedition.org/afriques/docannexe/image/5491/img-2.jpg
Fichier image/jpeg, 681k
Titre Figure 3: The main area, or “chancel”, of Osangangan Obamakin in 1953
Crédits Photograph by William Buller Fagg, 1953 (400_WBF.1953.04.14) © RAI.
URL http://journals.openedition.org/afriques/docannexe/image/5491/img-3.jpg
Fichier image/jpeg, 457k
Titre Figure 4: A reconstructed sketch map of Osangangan Obamakin
Légende Sketch map showing the location of the excavation grid, and important features and finds within it. It has not been georeferenced to a specific location within modern Ife: as of now, only the approximate location can be discerned.
Crédits Tomos Llywelyn Evans, 2024.
URL http://journals.openedition.org/afriques/docannexe/image/5491/img-4.png
Fichier image/png, 138k
Titre Figure 5: An example of how the 1x1 yard grid squares (see Fig. 4) were subdivided into square feet, and how the system of coding finds and features worked (in this case, square Gc)
Légende Finds in the greyed square would be coded ‘Gc9’ – G being the x-axis letter (capital letters), c being the y-axis letter (small letters), and 9 being the square-foot subdivision within the larger square.
Crédits Tomos Llywelyn Evans, 2024.
URL http://journals.openedition.org/afriques/docannexe/image/5491/img-5.png
Fichier image/png, 26k
Titre Figure 6: Photographs of the “water cooler” from 1953 (FW S110)
Crédits Left: Photograph by William Buller Fagg, 1953 (400_WBF.1953.08.03) © RAI. Right: Photograph by A.J.H. Goodwin, 1953, UCT BC 290, Photographs, 1570–1612.
URL http://journals.openedition.org/afriques/docannexe/image/5491/img-6.png
Fichier image/png, 2,5M
Titre Figure 7: The 1953 excavations at Osangangan Obamakin
Crédits Photographs by William Buller Fagg, 1953 (left: 400_WBF.1953.04.13; right: 400_WBF.1953.04.15) © RAI.
URL http://journals.openedition.org/afriques/docannexe/image/5491/img-7.png
Fichier image/png, 822k
Titre Figure 8: Terracotta primate figurine (FW T247) excavated from the shrine of Osangangan Obamakin
Légende See F. Willett (2004, T247), R. Abiodun (2014, p. 69–70) and S. P. Blier (2015, p. 325–328, 512) for interpretations of this piece.
URL http://journals.openedition.org/afriques/docannexe/image/5491/img-8.png
Fichier image/png, 1,1M
Titre Figure 9: A glazed crucible fragment from early Ife, observed on the surface at Olokun Grove in 2017, with a blue surface where the glass has fused with the ceramic
Légende Such objects were often coveted, and incorporated into shrines.
Crédits Tomos Llywelyn Evans, 2017.
URL http://journals.openedition.org/afriques/docannexe/image/5491/img-9.jpg
Fichier image/jpeg, 377k
Titre Figure 10: The archaeological context of the gagged terracotta head and associated finds, 18″ to 20″ (approx. 45.7cm to 50.8cm) below surface
Légende Left: a reproduction of the finds’ location within the grid squares, as represented in UG, ACCN 3120/1, B. Fagg Diary, 22/1/53, p. 9 (1953). Right: a photograph of the finds in situ. Photograph by William Buller Fagg, 1953 (400_WBF.1953.001.08).
URL http://journals.openedition.org/afriques/docannexe/image/5491/img-10.png
Fichier image/png, 436k
Titre Figure 11: A sketch of the old shaft feature beneath the main shrine
Légende Based on a drawing by B. Fagg (UG, ACCN 3120/1, B. Fagg Diary, 3/3/53, p. 26), following Fagg’s own representation. He sketched similar shaft features at Olokun Grove (UG, ACCN 3120/1, B. Fagg Diary, 23/2/53, p. 21). They speculated that these might be burial shafts.
URL http://journals.openedition.org/afriques/docannexe/image/5491/img-11.png
Fichier image/png, 38k
Titre Figure 12: Partially reconstructed fragments of a terracotta figure of a diseased man taken from the main shrine
Légende By K. C. Murray in 1943, claimed to have been Aroni, the spirit of whirlwind (FW T239).
Crédits Photograph by William Buller Fagg, 1953 (400_WBF.1953.07.1) © RAI.
URL http://journals.openedition.org/afriques/docannexe/image/5491/img-12.png
Fichier image/png, 1,3M
Titre Figure 13: Terracotta objects, largely from Osangangan Obamakin
Légende Several were obtained from the main shrine by K. C. Murray in 1943. Photograph by William Buller Fagg, 1953 (400_WBF.1953.07.05).
Crédits © RAI.
URL http://journals.openedition.org/afriques/docannexe/image/5491/img-13.jpg
Fichier image/jpeg, 388k
Titre Figure 14: A conical head
Légende Probably taken by K. C. Murray in 1943, though he does not mention it in his writing. According to the delegation of priests that met B. Fagg in 1953, this represented Awunrin Olojo, wife of Osangangan (UG, ACCN 3120/1, B. Fagg, Notes, Other, 18/1/53). FW T246. Photograph by William Buller Fagg, 1953 (400_WBF.1953.07.06).
URL http://journals.openedition.org/afriques/docannexe/image/5491/img-14.png
Fichier image/png, 1,2M
Titre Figure 15: Terracotta objects from the grove
Légende The head (left; FW T236) was said to be a representation of Osangangan himself and to have been taken from the main shrine by K. C. Murray in 1943. The figurine depicts a kneeling man with an animal and was presumably collected by K. C. Murray in 1943 although he does not mention it in his writing (FW T244). Photograph by William Buller Fagg, 1953 (400_WBF.1953.07.09).
Crédits © RAI.
URL http://journals.openedition.org/afriques/docannexe/image/5491/img-15.png
Fichier image/png, 682k
Titre Figure 16: A decorated globular pot
Légende Taken from the main shrine by K. C. Murray in 1943, apparently used to hold water for sacrifice (FW T243). Photograph by William Buller Fagg, 1953 (400_WBF.1953.07.10).
Crédits © RAI.
URL http://journals.openedition.org/afriques/docannexe/image/5491/img-16.png
Fichier image/png, 388k
Titre Figure 17: Terracotta fragments
Légende Probably all from the grove. Back left: appears to be FW T253, part of a head and ear, excavated from the main shrine (Square Td9) in 1953. The other fragments (four of which appear to be feet) may be other items from the main shrine excavations in 1953. Photograph by William Buller Fagg, 1953 (400_WBF.1953.08.08).
Crédits © RAI.
URL http://journals.openedition.org/afriques/docannexe/image/5491/img-17.jpg
Fichier image/jpeg, 191k
Haut de page

Pour citer cet article

Référence électronique

Tomos Llywelyn Evans, « Reopening the archives of archaeology at Ile-Ife, Nigeria: Towards a history of the Osangangan Obamakin sacred grove »Afriques [En ligne], 16 | 2025, mis en ligne le 09 mars 2026, consulté le 15 mai 2026. URL : http://journals.openedition.org/afriques/5491 ; DOI : https://doi.org/10.4000/161vu

Haut de page

Auteur

Tomos Llywelyn Evans

Assistant Professor, Department of Art History and Archaeology, Department of African and African-American Studies, Washington University in St. Louis

Haut de page

Droits d’auteur

CC-BY-NC-ND-4.0

Le texte seul est utilisable sous licence CC BY-NC-ND 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont susceptibles d’être soumis à des autorisations d’usage spécifiques.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search