Current Archaeological Research in Ghana 9781407303345, 9781407333496

This work presents and evaluates internal perspectives on the profile of archaeology in the University of Ghana, Legon,

323 44 25MB

English Pages [159] Year 2008

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Front Cover
Title Page
Copyright
Table of Contents
Frontispiece
Acknowledgements
Archaeology in the University of Ghana, Legon. A Survey of Emic Perspectives on its Profile and Future, with an Etic Commentary
An Investigation of a Kintampo Complex Site at Boyasi Hill, near Kumasi, Ghana
Molluscs in Archaeological Reconstruction: The Kpone Coastline, Ghana, as a Case Study
Excavations at Fort Amsterdam, Abandze, Central Region, Ghana
Researching the Internal African Diaspora in Ghana
Current Archaeological Research at the Krobo Mountain Site, Ghana
Placing the Tongo Hills, Northern Ghana, in Archaeological Time and Space: Reflexivity and the Research Process
Rethinking the Stone Circles of Komaland. A Preliminary Report on the 2007/2008 Fieldwork at Yikpabongo, Northern Region, Ghana
The Archaeology of Slavery: A Study of Kasana, Upper West Region, Ghana
Clay Toys of the Grandchildren of a Potter in Salaga: Insights for Archaeology in Ghana
The Late Stone Age in Ghana: The Re-excavation of Bosumpra Cave in Context
Recommend Papers

Current Archaeological Research in Ghana
 9781407303345, 9781407333496

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

BAR S1847 2008

Cambridge Monographs in African Archaeology 74 Series Editors: John Alexander, Laurence Smith and Timothy Insoll

INSOLL (Ed)

Current Archaeological Research in Ghana Edited by

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

Timothy Insoll

BAR International Series 1847 2008 B A R

Cambridge Monographs in African Archaeology 74 Series Editors: John Alexander, Laurence Smith and Timothy Insoll

Current Archaeological Research in Ghana Edited by

Timothy Insoll

BAR International Series 1847 2008

Published in 2016 by BAR Publishing, Oxford BAR International Series 1847 Cambridge Monographs in African Archaeology 74 Current Archaeological Research in Ghana © The editors and contributors severally and the Publisher 2008 The authors' moral rights under the 1988 UK Copyright, Designs and Patents Act are hereby expressly asserted. All rights reserved. No part of this work may be copied, reproduced, stored, sold, distributed, scanned, saved in any form of digital format or transmitted in any form digitally, without the written permission of the Publisher.

ISBN 9781407303345 paperback ISBN 9781407333496 e-format DOI https://doi.org/10.30861/9781407303345 A catalogue record for this book is available from the British Library BAR Publishing is the trading name of British Archaeological Reports (Oxford) Ltd. British Archaeological Reports was first incorporated in 1974 to publish the BAR Series, International and British. In 1992 Hadrian Books Ltd became part of the BAR group. This volume was originally published by Archaeopress in conjunction with British Archaeological Reports (Oxford) Ltd / Hadrian Books Ltd, the Series principal publisher, in 2008. This present volume is published by BAR Publishing, 2016.

BAR

PUBLISHING BAR titles are available from:

E MAIL P HONE F AX

BAR Publishing 122 Banbury Rd, Oxford, OX2 7BP, UK [email protected] +44 (0)1865 310431 +44 (0)1865 316916 www.barpublishing.com

Contents Frontispiece ............................................................................................................................................................ iii Acknowledgements ................................................................................................................................................ iv Chapter 1. Archaeology in the University of Ghana, Legon. A Survey of Emic Perspectives on its Profile and Future, with an Etic Commentary. Timothy Insoll ...................................................................... 1 Chapter 2. An Investigation of a Kintampo Complex Site at Boyasi Hill, near Kumasi, Ghana. James Anquandah ..................................................................................................................................... 13 Chapter 3. Molluscs in Archaeological Reconstruction: The Kpone Coastline, Ghana, as a Case Study. Fritz Biveridge................................................................................................................................... 25 Chapter 4. Excavations at Fort Amsterdam, Abandze, Central Region, Ghana. J. Boachie-Ansah ................................................................................................................................................... 37 Chapter 5. Researching the Internal African Diaspora in Ghana. Kodzo Gavua ................................................... 63 Chapter 6. Current Archaeological Research at the Krobo Mountain Site, Ghana. William Narteh Gblerkpor..................................................................................................................................... 71 Chapter 7. Placing the Tongo Hills, Northern Ghana, in Archaeological Time and Space: Reflexivity and the Research Process. Timothy Insoll ........................................................................................... 85 Chapter 8. Rethinking the Stone Circles of Komaland. A Preliminary Report on the 2007/2008 Fieldwork at Yikpabongo, Northern Region, Ghana. Benjamin W. Kankpeyeng and Samuel Nilirmi Nkumbaan ............................................................................................................................. 95 Chapter 9. The Archaeology of Slavery: A Study of Kasana, Upper West Region, Ghana. Samuel Nilirmi Nkumbaan .................................................................................................................................. 103 Chapter 10. Clay Toys of the Grandchildren of a Potter in Salaga: Insights for Archaeology in Ghana. J. Ako Okoro ....................................................................................................................................... 117 Chapter 11. The Late Stone Age in Ghana: The Re-excavation of Bosumpra Cave in Context. Derek Watson........................................................................................................................................ 137

i

ii

Frontispiece. Top row (left to right). James Anquandah; J. Boachie-Ansah; Kodzo Gavua. Middle row (left to right). Timothy Insoll; William Narteh Gblerkpor; J. Ako Okoro. Bottom row (left to right). Benjamin W. Kankpeyeng; Samuel Nilirmi Nkumbaan; Derek Watson.

iii

Acknowledgements The inspiration for this volume was due to my being a visiting scholar in the Department of Archaeology, University of Ghana, Legon, and I am most grateful to Drs Kodzo Gavua and Benjamin Kankpeyeng for helping organise my placement. I am also grateful to all my colleagues in the Department of Archaeology for making my stay there so rewarding and pleasurable and for writing their contributions to this volume. I would also like to thank David Davison at BAR for agreeing to prepare the camera-ready copy, and Rachel MacLean for all her support and for preparing some of the maps and illustrations.

iv

Archaeology in the University of Ghana, Legon. A Survey of Emic Perspectives on its Profile and Future, with an Etic Commentary Timothy Insoll Archaeology, School of Arts, Histories, and Cultures, University of Manchester, and Visiting Lecturer, Department of Archaeology, University of Ghana, Legon

abroad, new collaborations and partnerships, and new staff members, all mean that the next five years both in the Department of Archaeology (as was), and in wider Ghanaian archaeology promise to be intellectually stimulating, and potentially beneficial to the nation as a whole, as is clearly indicated by the survey data included here.

Introduction It is useful to start this paper by stating what it will not do, and it will not provide a review of what has been done in Ghanaian archaeology for this is a subject that has been considered by others (e.g. Anquandah 1982; Stahl 1994). Equally, it will not provide a history of archaeology in Ghana which again has been reviewed elsewhere (e.g. Anquandah 1982; Kankpeyeng and DeCorse 2004; Posnansky 2006-2007; Kense 1990), nor, for the same reasons will a history of the Department of Archaeology, University of Ghana, Legon be provided (see Agbodeka 1998: 76-77; Posnansky 2006-2007:5354). Rather, focus is placed here upon presenting and evaluating emic (internal) perspectives on the profile of archaeology in the University of Ghana, Legon, internationally, and nationally, and also its future. That this author can presume to write such a contribution is because he was fortunate to have been a visiting scholar in the Department (February-May 2008), and was thus able to gather relevant material. The interpretation of this data is thus partly personal and perhaps subjective, but also hopefully objective in that it is truthful to state at the outset that the reputation of the Department of Archaeology, University of Ghana, is so well established that it has been one of the cornerstones of West African archaeological research almost since its inception (see Kense 1990), and indeed is now at the forefront of archaeological research and teaching in the continent of Africa.

These factors are not irrelevant in dictating future research directions and projects, be they foreign or indigenously conceived. Hence the data contained here is of significant relevance beyond the immediate context of the Department of Archaeology, University of Ghana, Legon, and is pertinent to all those involved or interested in Ghanaian archaeology, and indeed in the changing nature of archaeology in general in the twenty first century. The Questions, Answers, and Commentaries Fourteen questions were asked of each member of staff, but answers were not necessarily received for each question. The questions were framed so as to be able to explore perceptions of archaeology in a range of contexts and at a variety of different levels, Departmental, University, nationally, and internationally. Although not all the interview data is reproduced here, where relevant the survey data is extensively quoted in unedited form as it was transcribed during the interviews and each set of answers to every question is followed by a brief analytical commentary.

However, it should also be noted that in analysing the emic perspective the author is not attempting to adopt such a position himself. Instead this has been gained through interviewing in April-May 2008 as many members of staff of the Department as were available using a survey questionnaire, the eleven responses to which, for reasons of protecting privacy, have been anonymised. Hence, an etic (external) perspective might be brought to the comments upon the survey data, but this is acknowledged as such and no phenomenological musings on the blurring of the emic/etic divide are entertained here. It should also be noted that this was an interesting point to have been involved in the Department as it is on the cusp of initiatives in repositioning itself in terms of its disciplinary identity and focus, with a move away from what could be termed ‘pure’ archaeology. Equally, rapid changes both within the wider University and the country as a whole are impacting upon the Department, and new resource streams from home and

1) Is archaeology thought of as a serious subject by other Departments and the central administration in the University of Ghana? •



1

“Archaeology is tolerated – partly as it can provide intellectual space and variety. But not in heart of hearts regarded as a core discipline. The University Council does not understand archaeology. This is the only Department of Archaeology in the whole of Ghana. There are many public universities, but this is still the only department. Private universities don’t want to hear about it at all”. “Until recently, no, but 12 months ago perception changed. Major public relations initiatives assisted in changing this. These

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA included the holding of major conferences and workshops, changes in the academic programme, change of department name, linkage with the British Museum, and the holding of temporary exhibitions on (in conjunction with the British Museum), Ghanaian textiles, (“Fabric of a Nation: Textiles and Identity in Ghana”, March-July 2007) in commemoration of the 50th anniversary of Ghanaian independence, and football in Ghana (“Football in Ghana: People, Places and Social cohesion”, ongoing) to coincide with the African Cup of Nations (January-February 2008). The memo of understanding with the University of Copenhagen and the development of the GDARCH (Ghana-Denmark Archaeological Cooperation) has meant the University is also now taking Archaeology seriously. Students now elect to read Archaeology, whereas before they were offered it by the university administration. The number of postgraduate students and senior undergraduates has also increased. We have also had students requesting via the Dean to change to Archaeology courses. The development of new projects has also been beneficial here, e.g. Krobo Mountain Archaeological Project, Komaland Project, Internal African Diaspora in Ghana Project. We have also involved the media in what we are doing, e.g. in field schools, and this has increased public understanding of what is done here in the Department”. •









“Until very recently it was not. Many thought those doing archaeology were a joke, they were ‘professional grave looters’. This has changed very recently due to department leadership, research drive and direction of research now, and the recruitment of younger people into the department”.

Based upon the responses received it would seem that archaeology is flourishing within the University of Ghana, and is perceived as a strong, healthy, academic subject area which is mirrored in student recruitment trends, both at undergraduate and graduate student levels. 2) How do you rate the Department of Archaeology, University of Ghana, in terms of its international profile in comparison to other departments of archaeology in – i) West Africa? ii) Sub-Saharan Africa? •

“No, historically it has not been. When education started in West Africa the target was to produce lawyers, engineers, accountants, doctors etc., sociology was accommodated in this but archaeology, history, and classical civilisations lumped together as not relevant for development”. “Previously they did not. Now the perception has changed. This is because of the efforts of the Department…..People did not use to know anything about archaeology, only about artefacts. Perception has changed and people now respect archaeology. Students are opting to study and major in archaeology and undergraduate numbers are growing substantially”. “They are now getting that understanding. Students (are) now taking archaeology because before students did not see the economic benefits in taking archaeology, (they) thought it was solely about excavation, but now see it as diverse”. “The general perception was low but the image has improved dramatically especially since 2006 and the conference held in August 2006, and the media was involved in this. Since then people have had a change of heart for the discipline”. 2

“We have a very high profile and based upon information received from colleagues in the subregion the dept is perceived of as the strongest in the region”.



“Ibadan, is ahead of University of Ghana. (The) numbers of staff (are) higher in the archaeology department in Ibadan. Thurstan Shaw & Bassey Andah boosted their profile and the West African Journal of Archaeology was based there. Nigeria also has more archaeology departments – networks. Ghana does not have this, for example, Ife, Nsukka, Ahmadu Bello (Zaria), Legon is the only one in Ghana. Look at archaeology holistically then Nigeria is ahead of Ghana. But quality is another matter, then Legon is better in this”.



“Our specific competitors would be in Nigeria, Ibadan, and they have moved a step further by publishing journals and publications on issues in West African archaeology. We are not active enough here in organising publications in the department”.



“Very high. We attract visiting lecturers of a high reputation. Our own staff have also carved out a reputation for themselves, for example, Professor Anquandah. We have also attracted a lot of funding, for example, Danida for G-DARCH, and the British Museum. Without a high profile we could not have attracted this funding”.



“We have a good reputation in the sub-region”.



“I rate it high”.



“Francophone countries are slightly ahead in cultural management than us, for example, in publications. But this department is emerging as one of the strong departments in the sub-region.

TIMOTHY INSOLL: ARCHAEOLOGY IN THE UNIVERSITY OF GHANA, LEGON We have more students and our numbers are growing”. •

It can be seen that opinions varied as to where the main competitor departments were perceived to be and in some instances there was a self-stated lack of comparative knowledge. However, two main areas emerged in the responses. Firstly, Nigeria, and in particular Ibadan, replicating, in part, answers already provided to question two, and second, South Africa. Although the Anglophone comparative emphasis in the responses is obvious, and must be acknowledged, and indeed was by one respondent, the perceptions of competitors both regionally, and in sub-Saharan perspective would certainly appear accurate.

“Before the 90s we hardly knew of the existence of this department in the country, but internationally it was well known. Comparatively, our department is attracting more attention and recognition than other departments of archaeology in the sub-region”.

Although the responses received predominantly related to the sub region rather than sub-Saharan Africa as a whole, it can be seen that the international profile of the Department of Archaeology is rightly seen as high in comparison to other departments in West Africa. This is supported by the various conferences, exhibitions, and research projects emanating from the Department, with key examples discussed, where relevant, below.

4) Do you believe the net contribution of archaeological research in Ghana is greater in international terms now that it was i) 10 years ago. ii) 20 years ago. iii) 30 years ago? •

“The net contribution of archaeological research in international terms is greater in Ghana than it was 10 or 20 years ago, and is the same now as 30 years ago”.



“It is the same as 10 years ago. The period of Oliver Davies and the Volta Basin Research Project was the highpoint….as we saw efforts to survey wider areas of Ghana. Even today Davies’ work continues to be a reference point and nothing comparable exists today”.



“Yes, more than 10 years ago. Lots of foreigners have played important roles in Ghanaian archaeology and the Kintampo culture complex debate has raised the profile of Ghanaian archaeology”.



“Possibly, there is more going on, whereas in the 80s-90s there were only a few projects. But this is mainly historical archaeology and ethnoarchaeological research and not much prehistory. The 60s-70s were the heyday of prehistory here”.



“This is a difficult question to answer as not much fieldwork has been done, particularly by indigenous scholars. If you include the work by DeCorse and others in Elmina, then internationally, Ghanaian archaeology is better known now than 10 years ago. Compared to 20 years ago, the profile is now higher. Archaeology in Ghana is now well-known because of people like Stahl, DeCorse etc. working here – people know about some aspects of it. And the Koma work has made the archaeology fairly wellknown, the art pieces increased awareness. More people know of archaeology in Ghana than in the past”.



“Yes, more than 10 years ago”.



Yes, the rating now is higher than ten years ago”.

3) Where do you see your main competitor departments in terms of a similar or greater international profile being? • “Our closest competitors in sub-Saharan Africa are in South Africa”. • •

“Continental competitors, I have not thought in these terms before”. “In West Africa there is a difference between Anglophone and Francophone, we don’t know much about what is going on in the Francophone, but in Anglophone Africa, Ibadan. Otherwise, Kenya, Zimbabwe (but trouble there now), and Cameroon because of Nic David’s work there”.



“There are no regional competitors, but I don’t have enough knowledge of the subject”.



“Ibadan, and Witswatersrand in South Africa. No where else, we outclass all the rest”.



“Nigeria – Ibadan, but in sub-Saharan Africa I have no idea”.



“In West Africa, Ibadan, and University of Nigeria, Nsukka. Otherwise South Africa, South African universities are high profile universities”.



“As an English speaking country we compare ourselves with English speaking countries in West Africa and Nigeria would be our competitor, in particular Ibadan and University of Nigeria, Nsukka. Ibadan has a vibrant faculty of all theoretical shades”.



“Ibadan, and Tanzania outside of West Africa, but I have no specific knowledge of this”.

3

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA •

“Twenty years ago in the 1980s the economic situation in the nation was bad. The government had a very bad reputation in international circles, and there was drought. Archaeology was a low priority”.



“I should think so. Ten years ago there were not lots of collaborators and external funders, but it is different now. It is one sign that we are doing better”.

play a role. Yet the expatriate female contribution has been more significant, for example, Dombrowski, Casey, Stahl, (though) now seeing indigenous products”.

The responses to this question varied dependent upon the age and/or experience of the respondent as to whether they felt they could comment upon such a period of time. Interestingly, the answers provided also varied depending on what was deemed significant in terms of research focus; Prehistory, Iron Age, or Historical Archaeology. However, the general perception was that compared to ten years ago, i.e. the chronological point of reference most people felt best able to comment upon, the net contribution of archaeological research in Ghana in international terms was now greater. This would appear to be supported by high-profile publications such as those by Stahl (2001), DeCorse (2001), and Anquandah (2007).



“Yes, when I was a student, we never had one female undergraduate student. Now we have a lot of female undergraduates and for the first time two female MPhil students have finished – a big change”.



“It is now starting. In general terms more females are now interested in archaeology & see the first female MPhil students as role models. Females are also losing their fear of excavation as they are now beginning to understand it. Not just about digging manholes!”



“Oh yes, now we have female students coming in and participating in excavations. Women were fewer before, now women are enthusiastic about the discipline”.



“Yes it is coming. The department produced the first two female MPhil students. Prior to their graduation people always thought archaeology was for men”.



“We are just beginning. Two female MPhil students are joining us as staff and this is a plus for the present generation of archaeologists”.



“Sure. I think there were women doing archaeology up to level 400, but they did not take it seriously in terms of a profession. The first female MPhils have now completed and now more women are expressing an interest in graduate studies. The perception before was that archaeology was all about digging. This is changing, it is not all about this, not only about carrying a pickaxe in the field”.

5) Do you believe women now play a more significant role in Ghanaian archaeology than they did in the past? If so, how? •





“Women do now play a more important role. Two women completed the MPhil for the first time ever in Ghana last year. There are now more female senior undergraduate students than male”. “The greater role of women in Ghanaian archaeology is definitely a development of the last ten years. It is an outcome of the broadening of topics which are treated in archaeology courses encompassing issues of gender, tourism, water management etc. It has a bigger appeal. Also the great increase in the number of female students is a reflection in the overall increase in the number of women students over the past decade, as part of the active policy of the University with regard to female admissions”.

There was almost universal consensus in the belief that women now play a more significant role in Ghanaian archaeology than previously. Specifically, the recent completion of the MPhil by two indigenous female students, Joyce Dartey (2007), and Gertrude (Aba) Eyifa (2007) was signalled, repeatedly, as of great importance. This is a significant development providing the necessary role models to encourage further indigenous female participation in Ghanaian archaeology at all levels.

“(The) current women who have just completed the MPhil are not the first to have been trained, but others left. (The) lack of women in (the) Department in the past is not because of lacking vision but problem is marriage, children etc. Also had expatriate women staff in the past, for example, Joanne Dombrowski and Signe Nygaard”.



“The participation of indigenous female archaeologists is a new development and the female presence is much greater now”.



“I can’t genuinely answer this question, but there is now more of a realisation that women should

6) What for you could be the most significant ways that archaeology in the Dept. could be improved?

4



“Field research should also be designed to make it more public”.



“Developing field schools so that the students can earn credit from these, but not just for 10 days or 2 weeks, it should be for 6 weeks”.

TIMOTHY INSOLL: ARCHAEOLOGY IN THE UNIVERSITY OF GHANA, LEGON •

functions well in Ghana from abroad via partnerships. For example, G-DARCH is a good idea if organised well, and it has $1million in funding”.

“A lot of courses should be longer and more detailed and structured than they are now, and a lot more emphasis placed upon fieldwork. Teaching complex courses in one semester means cutting corners. More publicity surrounding archaeology would also be useful, but Kodzo Gavua has been trying this and having some success”.



“(Include) courses such as economic anthropology, tourism and archaeology, and more courses on Ghana. For example, (the) stone age other than Kintampo, nothing is said about Ghana. We need more on this”.



“The most significant is what has already been started. We are scratching the surface of what could happen”.



“The library needs physically expanding, it is too small, and we need more books from outside”.



“There is little theory in our teaching. We need to focus on this from the undergraduate level upwards”.



“Modern equipment is coming in, e.g. GPS, hence our field techniques and methods should be improved. In the late 70s (to) early 80s we lacked even a dumpy level and had to borrow one from Geology”.



“The student-teacher relationship. Teachers should make themselves available for consultation after teaching. But this is improving”. (Also) the direction of research, there should be a development component in this, e.g. water management in Salaga. This is a good thing”.

In general, the responses to this question indicated that both the teaching of, and research in, archaeology in the Department did not need significant improvement, It was felt that any major changes which might have been required were already in motion, as the responses to later questions (e.g. 7, 9, 10) further indicates. 7) Are resources for teaching, and researching archaeology better now in Legon than they were in the past in terms of - i) 10 years ago. ii) 20 years ago. iii) 30 years ago? •

“The Department does now have much better facilities such as computers, video cameras etc. The situation is better but there is room for improvement. G-DARCH has brought some benefits with colleagues having access to money for research”.



“Yes, in a way, they are better than 10 years ago. People are attempting to make sure archaeology



“Equipment wise it is better, but beyond this I can’t comment”.



“The resources are better than 10 years ago – especially with coming of Danida (G-DARCH). This funds graduate students and lecturers research. In the past we had no money. Lecturers used their own money – it is different now”.



“Yes, resources are better. Some resources are available, but not enough for everyone – could be increased so as to take all of Level 100 to the field. Ten years ago not done at all”.



“Yes we have more money from central university for teaching than in the past. Research, it is not so clear. When (Merrick) Posnansky was here in the 1970s we had money from the Leverhulme Trust for research. G-DARCH is not as simple. I believe in Posnansky’s time it was easier getting money for research than now”.



“Yes in terms of sheer comparison. In, for example, that we now have a PPT projector which is an integral part of modern teaching. Also through G-DARCH computers have been made available to graduate students”.



“Yes, they are better than 10 or 20 years ago. Thirty years ago, archaeology was then dominated by expatriates. The research resources are now better thanks to the British Museum and G-DARCH”.



“The resources for teaching are better than 10 years ago, e.g. access to the internet. There were no computers 10 years ago, except in the general office. The last five years have been significant. The resources are also better for research, e.g. fieldwork. Graduate students are now supported from G-DARCH, irrespective of the problems with the latter. Previously, students could not graduate as they could not get funding for their field research”.

Clearly, based upon the answers received to the question, it is almost universally perceived that resources for teaching, and especially for research in archaeology, are better than they were 10 years ago (again the temporal benchmark most respondents felt able to comment upon). Of especial significance in this respect is the Danida funded Ghana-Danish Archaeological initiative which has provided substantial funding for the development of archaeology in the Department.

5

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA choice and archaeology is still part of Cultural Heritage. (It) also allows people to take more anthropological courses & will allow people to go to the MPhil level without excavating. This is important as because of the financial aspect of excavation if people are not able to cater for this they can’t go into the field, changing the name will allow for ethnoarchaeology etc”.

8) What do you think about the new disciplinary shift suggested by the dropping of the word ‘archaeology’ in favour of ‘cultural heritage’ in the new Department title? •

“The change in Department name to ‘The Department of Cultural and Heritage Studies’ is in answer to the attitude of Ghanaians. It was proposed as archaeology seen as irrelevant by students. Archaeology does not give a job, hence what is the use of archaeology? Most students rather not choose archaeology – students mocked for choosing it, ‘archaeo’ is a mockery. Who can tell what will come out of the new type of Department they are setting up. ‘Heritage & Archaeological Studies’ would have been preferable”.



“The name change broadens the scope of what is done in the Department to cover other disciplines relevant to the work of archaeology – ethnography, gender studies, museum studies, indigenous knowledge systems etc. Archaeology will remain a central component of study in the Department so (it) won’t impact on the international profile. Eventually we want to transform the Department into an institute with separate departments, e.g., visual anthropology, museum studies, archaeology, etc”.



“I have personal reservations about this. Dropping the name archaeology has deflated my ego completely. I am proud to be an archaeologist but the name has gone”.



“The word ‘archaeology’ is implied in ‘Heritage Studies’. The change is right for me personally. The students did not want to study archaeology, by changing the name we will attract more students. Drawing upon social and cultural anthropology will increase the way we interpret archaeology in the Department and improve the theoretical knowledge of the students”.



“If in general it is thought to be beneficial, then it’s a good thing. But I am slightly worried that archaeology will be downplayed in favour of heritage management. For its name to go something serious has happened as this was one of the few departments in West Africa and the only one in Ghana”.



“I think it is for the better, previously our scope was too limited. Now we are involved in areas that are more interesting and have direct bearing on our development as a people”.



“It changes the image and perception of the Department positively. The name ‘archaeology’ sounds archaic for the general public…..It also broadens the courses taught and gives more



I don’t totally agree with the idea, but I agree with a change of name, initially that was meant to be the Department of Archaeology and Heritage Studies, I don’t agree with dropping the word ‘archaeology’…..Core archaeology will have less and less emphasis. We are the only department of archaeology in Ghana, if you take archaeology out the discipline is gone”.



“I think it is good. First and foremost the new name reflects properly the scope of activities carried out in the Department. When you mention archaeology to the general populace, digging comes to mind. This has not helped perceptions amongst the general populace. Heritage is now being exploited for economic and political gains. We need to consolidate ourselves and be accepted, re-branding helps and makes the subject more appealing”.



“To be frank, I do not share the enthusiasm that archaeology be dropped from the title. I would like the Department to be the ‘Department of Archaeology and Cultural Heritage Studies’….Geography changed its name but retained the core name ‘Geography’ in becoming ‘Geography and Resource Development’”.



“It is all about perception. The term ‘archaeology’ had become something not attractive as far as the Ghanaian community was concerned. Other departments have changed their name, e.g., ‘Library and Archival Studies’ became ‘Information Studies’ and now everyone wants to do it. I am for it”.

This was, perhaps, the question which elicited the most detailed and vociferous responses. Although a degree of subjectivity in interpreting the answers has obviously still to be recognised, broadly, the answers given could be categorised, as six in favour of the new Department title, four against, and one ambivalent. Repeatedly, the rationale for and against the change was contextualised within the framework of the wider University and the needs of the country as a whole (see also Posnansky 2006-2007). Additionally, it might also be viewed within the context of department, faculty, and disciplinary readjustment and re-structuring which has occurred in the heritage and tertiary sector internationally, as in the U.K. for instance.

6

TIMOTHY INSOLL: ARCHAEOLOGY IN THE UNIVERSITY OF GHANA, LEGON 9) Do you think this change in Department ‘identity’ will have negative or positive, or no consequences at all, in terms of the international profile of the Department? •

“Only for people who don’t know about archaeology in Ghana. People should know that archaeology will be in Cultural Heritage, but in some instances might look to History for archaeology”.



“Hard to gauge. Possibly it won’t make much of a difference, but depends on the work completed and how much publicity it is given”.



“It will have a positive impact. Starts locally and change will have a gradual impact internationally”.



“Everyone knows that only Legon does archaeology in Ghana, so if (they) do a web search, they might think that archaeology is not now done in Ghana”.









10) Has the character of the Department changed in light of recent appointments – if so, how?

“For someone who has not seen our brochure. Without seeing this you cannot be sure what our course content will be and for that reason I am opposed to the dropping of the name archaeology”. “On the surface some people might have negative perceptions. The dropping of archaeology may be thought of as the dropping of the discipline, but this is not the case. But people with a fair idea of current trends should not have a problem with this”. “In terms of our international profile we will have problems. In what are our students going to graduate – Heritage Studies? How are departments outside going to assess what they do? Hence we should have retained ‘archaeology’”. “Internationally, they think changing the name is not the solution to the problem. There will be some negative elements initially, but the new course content will salvage this”.

The direct cross-over with the previous question is apparent in the responses received. In all instances where an answer was given this correlated with the previous position taken in responding to the change of Department title. Applying an etic perspective to this question, warranted by its international focus, would suggest that initially the adjustment of Department ‘identity’ will take some time to become known internationally, but should hold no adverse consequences once it is acknowledged that archaeology as a discipline is being maintained.



“(We) need to separate the discipline and the staffing (appointments), cannot merge the two”.



“The character of the Department has changed with younger more enthusiastic staff. We now do a combination of social anthropology and archaeology and we need to make our work responsive to the overall development of the nation and in line with the needs of the University of Ghana to revise programmes”.



“Yes, the character has changed. We are employing a different perspective in looking at archaeological issues. People are thinking about the context and not just the material. We are looking more at interpretive perspectives”.



“Don’t know”.



“Yes, there is a young breed here now, the teaching style reflects this and more Audio Visual material is used. Students now socialise more with lecturers as the age gap has reduced”.



“Yes, everyone come with a different objective. The student/lecturer relationship is now much better, e.g. freer access to lecturers is now possible”.



“Yes, considerably. The relationship between the junior lecturers and the Head of Department is healthier than before. There is now interaction and the leadership is more receptive to ideas from outside. The image has changed totally”.



“Yes. We now have younger faces here and a new spirit is arising. There are fresh ideas with this younger generation even in the simple sense of using digital equipment. The increasing numbers of staff also has net benefits for increasing numbers of students”.



“Through diversification in the research agenda”.



“Yes, for sure….The students are more comfortable in dealing with people closer in age. A practical consequence has been that the numbers of students taking archaeology in the last three to four years has increased dramatically”.

Excluding two responses which either expressed no knowledge about the issue or thought the rationale of the question needed to be altered, the majority of the answers (eight) were emphatic in the opinion that the character of the Department had changed, positively, as a consequence of recent appointments.

7

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA 11) Do you think the amount and diversity of foreign researchers involved in Ghanaian archaeology is less, the same, or more than it was - i) 10 years ago. ii) 20 years ago. iii) 30 years ago? •

“In the past five years there has been a lot more activity by foreign archaeologists in Ghana. This is a reflection of the whole atmosphere and conditions which have changed both in the country and the Department.



“It is more or less the same as 10 years ago but with different foci”.



“More than both 10 and 20 years ago”.



“There are now more foreign researchers than 10 years ago”.



“There are more foreigners than 10 years ago. Twenty years ago the research was done by people from the University (even if foreigners) – now there are more”.



“Greater”.



“More than 10 years ago and more than 20 years ago”.



“No, the same sort of people and numbers as 10 years ago”.

As can be seen in the answers, once again the period of time respondents felt able to comment upon varied. This factor accepted, the majority of responses (six) indicate a belief that there is greater diversity and numbers of foreign archaeologists involved in Ghanaian archaeology, certainly in comparison to 10 years ago. Why this might be so was not explored, but from a personal etic perspective, could include factors such as the receptivity of the Department to accommodating foreign researchers, and the overall national environment which is conducive to completing research.



“We cannot do without international participation. Archaeology itself is global and the material we are studying is evidence of global interactions. We also need the input of new ideas and it helps increase the capacity of the department through foreign postgraduate training. The archaeological landscape of Ghana is such that there is no way local archaeologists can make inroads on their own. It brings needed diversity. However, we don’t need school prefects telling us what to do, they need to come here as partners. Unfortunately, the conditions and low pay makes us vulnerable and there is a tendency of some partners to take control of the money and exploit this vulnerability, as some partners think our situation is so hopeless. This has led some people to completely move research direction in order to get money”.



“Yes, it is useful. (It) helps in securing funding, there is no funding locally”.



“Yes, I think it is useful. Brings different educational backgrounds, different approaches, all important”.



“I think it is good. They bring a different perspective to Ghanaian archaeologists, e.g. in teaching methods used by (Timothy) Insoll and (Derek) Watson. We can also learn technical skills from foreigners, e.g. GPS and cartography, interaction helps here”.



“It is somehow useful, if, e.g. G-DARCH, they can come to terms with local circumstances”.



“Yes, I have always stood up for it. We have a big place that Ghanaian archaeologists alone cannot deal with. No-one owns a site in this country. One way archaeology can improve in this country is when we have foreign participation”.



“I think it is a very good thing. It allows for the exchange of information and ideas, and for us to get awareness of current approaches. It is also good on the level of a partnership but problems can occur”.



“It is both useful and not. It is useful in terms of the kind of collaboration that comes in. Mutual interests where both sides are considered as partners, that is very helpful. It is also useful when foreign researchers come with the mindset to also learn, that is good. If otherwise, it can cause problems. Expatriate archaeologists may be versed in theoretical archaeology but we know our culture. Also they bring in resources. One difficult aspect of the Department is the lack of funding. International colleagues help address the inequalities”.

12) Do you think the participation of foreign archaeologists in Ghanaian archaeology is useful or not? If so, how? •

“The participation of foreign researchers is useful in terms of sharing knowledge, keeping abreast of modern trends and makes archaeology in Ghana more international. We will continue to encourage that because of the benefits. But need to see it more in terms of partnership rather than just collaboration, and partners sometimes partake in the planning of the research and its publication, rather than just its execution. The planning of the G-DARCH work was as a partnership and recently students from the American University in Rome have taken part in one of our field schools”. 8

TIMOTHY INSOLL: ARCHAEOLOGY IN THE UNIVERSITY OF GHANA, LEGON •

foodways etc., this will occur. Perhaps also see stone age research again increasing and water management archaeology growing in importance for this is the most important element in order to survive on a site, water”.

“It is useful, but it would be more useful if there was extensive co-operative work, e.g. Tallensi/Tengzug project is a good example. Or G-DARCH giving money to local archaeologists based on their own proposals”.

Of all the questions asked this is perhaps the one most relevant to a foreign archaeologist such as this author. As can be seen, the answers received were generally positive, though with understandable reservations. The latter varied but the most significant and the one recurrently cited was that the participation of foreign archaeologists in Ghana should be at the level of a partnership and not merely of collaboration. 13) What do you see as the significant research themes emerging in Ghanaian archaeology over the next decade? •





“We have moved from Phase 1 of archaeology in Ghana (50s-60s), phase of pioneers such as Davies and his contemporaries doing their reconnaissance surveys, through Phase 2 of indepth testing of selected sites (70s-80s), but still not area excavation, and into Phase 3 (sometime in 70s-80s) when area excavation starts at sites such as Begho, Bono Manso, & Daboya. The latter Phase 3 excavations are pure research excavations to extend the frontiers of knowledge. Now into a fourth phase of developmental or applied archaeological research. Research undertaken, e.g., to benefit tourists, at sites such as Fort St Jago, Cape Coast Castle, and Elmina old town. Sometimes difficult to draw a clear line between pure and applied/developmental research, for although you are advancing the frontiers of knowledge at the same time the data is contributing to tourists’ understanding”. “The Koma archaeological project, the Krobo Mountain archaeological project, the internal African diaspora in Ghana project, and research focussed upon slavery in general. Also a combination of ethnographic and ethnoarchaeological work in communities affected by oil production activities and mining. With regard to the former (oil), the Department is going to actively pursue these research opportunities, (as) a main priority of the Department, as state officials do not clearly understand archaeological work, hence the department has to be pro-active. Heritage and environmental impact assessment will pre-occupy archaeologists in Ghana”. “Historical archaeology. The issue of combining history and archaeology in attaining research objectives has become attractive – focussed on forts and castles, and slave plantations. Also a move toward using archaeology to answer anthropological and sociological questions, taste, 9



“(The) archaeology of ritual and religion with an indigenous knowledge base to draw upon. Historical archaeology and government seems to be supporting research in that direction, e.g. on forts and castles, as it is tourist related through the Ministry of Tourism. This is linked with slavery and (the) African American community and to entice the diaspora to spend foreign exchange in Ghana. In contrast, the north is neglected.



“Personally, I would like to see prehistory. Ethnoarchaeology and historical archaeology will continue but people will eventually realise that X many tens of thousands of years are virgin”.



“Historical archaeology, e.g. plantation archaeology will be continuing over the next 10 years. A revival in prehistory with the appointment of Dr Watson”.



“Getting into grassroots and a concern that the local community should benefit from the research. Research (is) not just for academic researchers but should be for the local community as well. North/south archaeological imbalance is being redressed – before there was a southern emphasis. Students (are) now asking to go to the north”.



“Slavery and gender, and over the next 10 years we will pay more attention to historical archaeology, by this I mean communities researching their own history through archaeology and building their own museums. Many palaces of the traditional rulers want their own museums. They need data from their own sites”.



“Slavery and slave trade, also aspects dealing with indigenous people, i.e. Ghanaian heritage, e.g. the Tengzug project. This helps bring out local elements in the history. The Krobo Mountain project also has a focus on the local aspect, as does the Internal Diaspora project”.



“Themes that are relevant to the present day needs of the nation. We should also not lose sight of sites that are African, e.g. in the north. There was disparity in north-south archaeology, but there are still areas of the south we don’t know about, we might know Akan history but we know little about its archaeology”.

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA •

of the tourism industry. We need both to pursue academic and scientific objectives and meet our social responsibilities”.

“The move toward CRM (Cultural Resource Management) and museum concepts rather than traditional archaeology such as pottery analysis. There should be developmental elements added into research projects – so local community benefits. As in Koma where a heritage research centre is being built with an attached museum for the community – they have provided land for this already. Now have northern appointed staff – helps redress imbalance”.

A wide range of responses were received to this question. With a recurrent theme being that development based archaeological research and assessment will be potentially significant in the coming decade, especially in the local context. The notion of applied archaeology which comes out of the responses is an important one and a very good example of this is provided by the Salaga Research Project which has had as one of its main aims, the investigation of past water management strategies and technologies with a view to what can be (re)learnt from these for the present (e.g. Okoro 2003). A north-south imbalance in archaeological research was also commented upon by three respondents but this was seen as being redressed through staff appointments and increased research in the northern sector of Ghana. But perhaps the most significant research theme repeatedly referred to as being important over the next decade was archaeological research focussed upon slavery, and the slave trade. This can be seen, in large part, to be due to the perceived economic (through tourism) benefits it will potentially bring to the nation. 14) What role, potentially, do you see archaeology as playing in the national development of Ghana? •



“So many roles, deliberate designing of parts of our courses to push something into the minds of students as to how archaeology is relevant for development, publishing in ways that the Ghanaian public can see with their own eyes the value of archaeology. We need to tell people about our results from the President down to the level of the District Assemblies. Applied archaeology – which has two sets of agenda, mainstream (academic archaeology) and development, but they are not mutually exclusive. Applied archaeology is placing archaeology at the centre of contemporary development. As far as the development of Ghana is concerned, aspects of research are relevant and should be pulled out and given prominence. Archaeology can also play a role in developing cultural literacy in Ghana as there are high levels of cultural illiteracy”. “Archaeology has been helping the nation to define its identity. This has been achieved by identifying, conserving, and managing national heritage resources proficiently and efficiently. By so doing we have been contributing to the growth 10



“Work in National Parks, e.g. Mole, need to incorporate an archaeological dimension in their management and to develop museums in Parks. Public archaeology, we need a revision of law, e.g. in advance of development”.



“A million books and articles have been written about the contribution of archaeology but it needs to get out into rural areas otherwise it won’t play a large role in many peoples lives. We need a concerted campaign to tell people about archaeology, but whether the money is available and the effort made is another thing. On the whole, I am slightly pessimistic about this”.



“A very important role in the sense that we are bringing out our past”.



“Archaeology is trying to create the consciousness of culture among our community, creating national consciousness and identity. We have shown people, e.g. via the textiles exhibition held in the Department Museum (see above) that archaeology is not just about excavation. The Department has also helped in forming a heritage council, (and an) important conference (was) held bringing together different interest groups (see below)”.



“I personally believe that an important aspect of nation building is about the consciousness you have of yourself as a people. One way of achieving this is through archaeology. Archaeology helps in nation building. (Also) tourism, cultural tourism in which archaeology plays a role has become the number three foreign exchange earner in Ghana. Museums in traditional areas will help in this process, e.g. the restoration of Danish plantation buildings which Dr Bredwah Mensah is involved in.



“There are many. It helps in creating awareness of ourselves and in the discovery of our identity as a group of people. Krobo, Tallensi, Koma, people get an idea of what went on before, it informs us of our heritage. It instils pride and hope for the future in the populace. All over the world people have relied on the past for their identity and its absence is a disaster. Archaeology helps in social cohesion, archaeology helps breach ethnic imbalances. Also opens up for tourism, as a by-product of our research”.



“Archaeology can play a significant role. By way of public archaeology, by this I mean the conservation and restoration of heritage structures and objects. In educating craftsmen

TIMOTHY INSOLL: ARCHAEOLOGY IN THE UNIVERSITY OF GHANA, LEGON and custodians of culture in the need to take care of buildings, artefacts, and the work that they do. Also archaeologists can engage with government to invest in archaeology and strengthen laws and regulations on heritage management. And in positioning ourselves in relation to engaging international bodies such as UNESCO”. •

consequence which is best remembered by both indigenous and foreign archaeologists alike; otherwise the necessary partnerships which are a prerequisite of the culmination of a successful research process will not succeed. Acknowledgements. I would like to thank all the members of staff of the Department of Archaeology, University of Ghana, Legon, who were asked for being willing to be interviewed. I would also like to thank Dr Rachel MacLean and Dr Kodzo Gavua for comments on the paper, though all misinterpretations, omissions and errors remain my own.

“Lots of things to consider. First, cultural tourism, e.g. in creating sites for tourist presentation, as with the rebuilding of the Frederiksgave Plantation. (And) the revival of indigenous knowledge systems, e.g. various forms of manufacture are no longer used”.

The detailed responses received to this question indicate the positive potential roles that archaeology is seen as playing in the national development of Ghana. These are various, but a repeat theme is the role archaeology should play in nation building and the development of national consciousness. Again the concept of applied archaeology was seen as significant in making archaeology relevant with a key example of this in action being provided by the conference, “Archaeology and Heritage Management in Ghana”, held in the British Council Hall, Accra (August 9-10, 2006). This can be described thus as it brought together a range of key stakeholders interested in heritage, heritage protection, and development including, besides members of staff from the Department of Archaeology, the Ministers of Tourism and Diaspora Relations, and of Chieftaincy and Culture, representatives of UNESCO, the World Bank, the National Commission on Culture, the West Africa Museum Programme, and the Ghana Museums and Monuments Board, as well as from the British Museum, the Institute of Field Archaeology (UK), and English Heritage. It thus served to indicate the role of archaeology in national development but also raised the overall profile of archaeology as well.

References AGBODEKA, F. 1998. A History of University of Ghana. Accra: Woeli Publishing Services. ANQUANDAH, J. 1982. Rediscovering Ghana’s Past. Harlow: Longman. ANQUANDAH, J. (ed.). 2007. The Transatlantic Slave Trade: Landmarks, Legacies, Expectations. Accra: SubSaharan Publishers. DARTEY, J. 2007. An Archaeology of Akyem Abompe: A Gender Perspective. Unpublished MPhil Thesis, Department of Archaeology, University of Ghana, Legon. DECORSE, C. R. 2001. An Archaeology of Elmina: Africans and Europeans on the Gold Coast, 1400-1900. Washington: Smithsonian Institution Press. EYIFA, G.A. 2007. House Construction and the Role of Women in Tongo-Tengzug: An Ethnoarchaeological Study. Unpublished MPhil Thesis, Department of Archaeology, University of Ghana, Legon.

Conclusions

KANKPEYENG. B. W. and DECORSE, C. R. 2004. Ghana’s Vanishing Past: Development, Antiquities, and the Destruction of the Archaeological Record. African Archaeological Review 21: 89-128.

In many ways it is difficult to conclude this review of the profile and future of archaeology in the University of Ghana, Legon without restating what has been already written. Hence it is useful to end by reiterating that this survey was completed as the Department of Archaeology, soon to be the Department of Cultural and Heritage Studies, is at a very significant point in its history as it transits from ‘pure’ archaeology, i.e. one that is dominantly research driven, to a more multi-disciplinary emphasis encompassed within the wider domain of ‘heritage studies’ (see Posnansky 2006-2007: 61-63 for relevant discussion on the concept and application of “Heritage Studies”). The concept of ‘applied’ archaeology described in the surveys eloquently captures what this new direction entails, but in order for archaeology to flourish further in Ghana, it has to be applied to national developmental needs, as well as local community based requirements, otherwise it will become peripheral and superfluous (see Posnansky ibid). Furthermore, it should be noted that this is a point of

KENSE, F. J. 1990. Archaeology in Anglophone West Africa. (In), Robertshaw, P. (ed.), A History of African Archaeology. London: James Currey, pp. 135-154. OKORO, J. A. 2003. Research on Water and Slaves in Salaga. Nyame Akuma 59: 45-53. POSNANSKY, M. 2006-2007. Appreciating the Global Dimension of Ghana’s Past. Transactions of the Historical Society of Ghana 10: 49-63. STAHL, A. B. 1994. Innovation, Diffusion, and Culture Contact: The Holocene Archaeology of Ghana. Journal of World Prehistory 8: 51-112. 11

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA STAHL, A. B. 2001. Making History in Banda. Cambridge: Cambridge University Press.

12

An Investigation of a Kintampo Complex Site at Boyasi Hill, near Kumasi, Ghana James Anquandah Dept. of Archaeology, University of Ghana, Legon, P.O. Box LG3, Legon, Ghana

Editor’s Note. This paper describes archaeological research completed over 25 years ago but which for various reasons never appeared in print. Because of the significance of the research to current debates on the Kintampo complex, because the site has now been destroyed by quarrying activities (J. Anquandah pers. comm.), and because of the preeminent role of Professor Anquandah in developing archaeology in Ghana it is our pleasure to include it in this volume as it was originally written to appear in the West African Journal of Archaeology without any attempt at updating it. Boyasi Hill was originally a forest environment but that heavy settlement accompanied by intensive cultivation in prehistoric times may have led to erosion with the result that the thin soil cover could not sustain the forest vegetation, hence the invasion of the hill by grassland type plants and fauna; (2) Boyasi hill has always had an insular grassland type of vegetation and that the presentday grassland vegetation may in fact be a relic of an original prehistoric vegetation type.

Introduction The Kintampo complex is now fairly well recognized to be an indigenous ‘neolithic’ tradition peculiar to Ghana and the Cote d’Ivoire. The complex has been shown to be characterised by material evidence on: i) Sedentism and building technology. ii) Mixed subsistence economy including hunting, fishing, gathering of wild crops and animal domestication. iii) Cottage industries and crafts such as potting, ground and flaked stone implements, milling equipment and stone beads. iv) Terracotta figurine art.

The existence of sites of the Kintampo complex in the Kumasi area was first established through research conducted by Nunoo along the banks of the Buruburo and Wiwi streams located in the area of the Kumasi University of Science and Technology. In the course of excavations conducted at Buruburo and Wiwi, Nunoo identified an industry characterised by quartz microliths, polished stone axes, stone bracelets, and stone ‘cigars’. Nunoo (1959, 1966) established that the granite outcrops of the area had been exploited for the production of stone artefacts.

Kintampo sites identified in Ghana number some twentyseven while two others have been recently located in the Cote d’Ivoire (Anquandah 1982; Chenorkian 1983). From available radiocarbon age estimates, the Kintampo is known to have flourished sometime during the period 2000-500 BC.

It was against this background that it was decided to undertake test excavations at Boyasi Hill.

The site of Boyasi Hill was discovered from an aircraft in 1969 by Len Newton, of the Department of Biological Sciences, University of Science and Technology, Kumasi. Newton was intrigued by the presence on the site of flora such as Borassus ethiopum palm trees, perennial geophytes such as the ground orchid, Eupholia cuculata and fauna such as the standard-winged Night-jar, Macrodipteryx longipennis Shaw. These were found to be flourishing on an apparently anomalous ‘island’ of micrograssland perched within the Kumasi forest vegetation spreading from the lower flanks of Boyasi Hill. Newton noted that the Night-jar’s normal habitat in Ghana is in the grassland territory north of latitude 9°N (Newton and Woodell 1975; 1976). During a reconnaissance of the site Newton and Woodell identified a number of grinding grooves, pottery sherds, polished stone axes and ‘cigars’ usually associated with Kintampo Complex sites.

Site Location The site of the derelict village of Boyasi (6°44’N, 1°34’W) is located near Kenyasi town, eight kilometres north east of Kumasi (Figure 1). The village stands at the foot of a hill, 330m high and steep-sided, a dome-shaped outcrop of granite which is one of several that form part of the southern Asante dissected peneplain. The southern Asante uplands comprising chiefly Birrimian and Tarkwaian rocks rise gradually from a height of 150m in the south to over 300m in the north east. The hill above Boyasi, named by Newton and Woodell (1975) after the village is a ridge which is oriented in a northeast to southwest direction. The ridge rises to about 60m above its immediate surrounding land. The north east part of the hill is covered by low forest. Other areas of the hill are only forested on their lower parts, the higher areas being covered entirely by grassland type vegetation.

Newton advanced alternative hypotheses to explain the Boyasi hill ecosystem, vis-à-vis the Kumasi forest ecosystem. According to him it is possible that: (1)

13

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

Figure 1. Location of the site of Boyasi Hill, Kumasi.

Aims of the Investigation

Surveying and Mapping

When the site was first discovered it was not immediately certain whether there had been a major Kintampo occupation there. The primary objective therefore was to ascertain whether the site was indeed associated with the Kintampo complex, and if so, to estimate the extent of the settlement. A second consideration was to ascertain to what extent the material culture remains differ from, or resemble, those of other Kintampo complex sites in the Kumasi area and further away in Brong, and Northern and Coastal Ghana. Thirdly, there was the need to discover some concrete evidence for early subsistence practices and sedentism. Finally, the research project hoped to obtain radiocarbon dates so as to obtain some idea of chronology of the Kumasi forest Kintampo tradition vis-à-vis that of Coastal and Northern Ghana.

The first season of fieldwork was carried out in 1975-76 and the second season in 1982-83. There was ground surface survey in Christmas 1975 and in Easter 1976 the site was mapped. A 30m by 30m grid was laid out on the hilltop where grinding grooves and remains of ancient stone huts were exposed and a site map drawn (Figures 2 and 3) prior to excavation. In 1982, a brief reconnaissance survey was undertaken at other Kintampo complex sites, notably Nkukoa Buoho, Wiwi and Buruburo, accompanied by further excavations at Boyasi Hill. As part of the Boyasi survey programme, a final year undergraduate student completed research at the site in 1983 resulting in a Long-Essay dissertation (Bredwa-Mensah 1984). The radial or spoked-wheel model of transect interval sampling was utilised for surveying the site. Transects were laid out from a central datum point along compass bearings 90°, 135°, 180°, 225°, 270°, 315°, and 360°. Sample units (each measuring 40m by 15m) were marked out as alternating ‘boxes’ along the transect ‘rays’.

Research Design The investigation at Boyasi Hill was, from the outset, viewed within the broader context of the search for prehistoric origins of the Asante forest population. Thus the research was seen as a prelude to a long-term project which was expected to entail:

A record was made of physical and cultural features found in each sample unit. Artefacts were collected along the transect lines from the datum point until the point where no more artefacts were visible on the surface. On the basis of the decline in frequency of artefact distribution along the transect lines the rough boundaries of the site were delimited. It was thus possible to estimate the area of the site to be 11.53ha. The survey produced a map illustrating the relative distributions of;

i) The search for Late Stone Age hunter-gatherer sites, Kintampo complex sites and also sites immediately post-dating the Kintampo. ii) Pre-excavation ground survey, mapping of sites and documentation of the prehistoric settlement pattern and distribution of cultural remains and assessment of the extent of occupation. iii) Selection of sites for excavation. iv) Analysis of the data and publication.

i) Present-day grassland and forest vegetation and the incidence of Borassus trees, modern farms, and footpaths. 14

JAMES ANQUANDAH: AN INVESTIGATION OF A KINTAMPO COMPLEX SITE AT BOYASI HILL, NEAR KUMASI, GHANA. ii) Prehistoric pottery, stone implements, beads, bracelets, daub and stone foundations of prehistoric architecture. iii) Exposure of grooved granitic rocks that served the prehistoric community as atelier for artefact manufacture. Excavations The test excavations conducted in 1976 and 1982 were directed by the author assisted by teams from the University of Ghana at Legon and the University of Science and Technology at Kumasi. The excavations were carried out under a licence issued by the Ghana Museums and Monuments Board, Accra. The Chief of Kenyasi granted permission to undertake scientific research on the farm lands of Boyasi Hill. The method employed in the excavations was to lay out pits in well-defined locations within the grid. The pit locations were determined by random sampling. The pits were mostly excavated in the form of trenches, one pit was however excavated by the quadrant method. In total five test pits were opened (Figure 3). PIT A – This is the northeast quadrant of a circular hump having a diameter of 8m. The pit was excavated to bedrock reaching a depth of 1.3m, although after 90cm it produced no cultural materials. PIT B – This unit measuring 2m by 4m was excavated to a depth of 1m and produced pottery, polished stone axes, the so-called ‘rasps’ or ‘cigars’, and a biconically perforated stone bead (Figure 4). A small sample of charcoal was obtained in the 1976 excavation from the basal levels of pit B and later submitted to the Rikagaku Kenkyusho Radiocarbon Laboratory in Tokyo but the sample quantity was inadequate. In fact one of the goals of the 1982 test excavation was to obtain adequate samples for radiocarbon dating but this proved unsuccessful.

Figure 2. Plan of Boyasi Hill.

PIT C – (Location R.102) – This unit measuring 2m by 2m produced pottery in the first level. Level 2 contained large granite boulders interspersed with cultural materials such as polished stone axes and pottery. Underneath the boulders were more potsherds and numerous fragments of daub. Level 2 also produced 37 quartz flakes, the product of a microlithic industry. PIT D – (Location V.93) – A pit measuring 2m by 2m which was excavated in the 1976 season was extended in the 1982 excavations into a larger octagonal pit (Figure 3). This area appears to have been utilised as an

Figure 3. Plan of excavated areas at Boyasi Hill.

15

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

Figure 4. View of the excavations and patterned arrangement of the granite slabs on Boyasi Hill.

atelier for the manufacture of polished stone arrowheads, polished stone axes and stone beads. Strong support for this assumption was found in the presence in the deposits of roughouts of axes, arrowheads and beads in the immediate vicinity of the large granite outcrop located in the middle of Pit D. There is evidence of four elliptical grinding grooves visible on the outcrop, Level 2 of Pit D also produced 12 quartz flakes.

Stratigraphy Figure 5 provides an illustration of the stratigraphy in Pit B. This was largely replicated in the other pits. However the ground at Boyasi Hill is generally sloping and there are also rock outcrops in between soil deposits. Thus the natural stratigraphic levels may not be entirely seen in a straightforward way. The Boyasi stratigraphy may be summarised thus (Table 1).

PIT E and Location F – Figures 3 and 4 illustrate the patterned arrangement of the granite slabs. The roughly rectangular arrangements suggest that they provided the infrastructure for the construction of wattle-and-daub houses. In order to determine whether or not these presumed house locations belong to Kintampo complex times, it was decided to undertake a test excavation at the entrance to Location E. The result was positive as the excavation produced Kintampo complex comb stamped pottery, rasps, polished axes and especially daub fragments. The result does not of course exclude the possibility that the structures could post-date the Kintampo. However no cultural remains apart from those of the Kintampo were found.

Description of the Cultural Materials 1) Stone Artefacts. A total of 12 polished stone axes made of either fine grained greenstone muscovite or sericite were retrieved (Figure 6). They range between 56cm in length, 2-3cm in width and 3-4cm in breadth. Some specimens are notched either unilaterally or bilaterally on their upper flanks probably to facilitate hafting of wooden or bone handles (Figure 7, nos. 9-12). Two roughouts of greenstone polished axes were found in level 2 of pit B (Figure 7, nos. 13 and 14). Also some 56 microlithic quartz flakes, most of them nondescript, were retrieved, nearly all from level 2. Among the most

Level Depth I – 0-20cm

Natural Stratigraphy Topsoil – dark brown humus and clay

II – 20-60cm

Dark brown clay with glassy quartz grains Basal rock; grayish white quartz with feldspar and biotite inclusions

III – 60-100cm

Table 1. Summary of the Boyasi Hill stratigraphy.

16

Cultural Stratigraphy Kintampo pottery, stone arrowheads, stone axes, stone beads, daub and rasps. Quartz flakes, stone axes, arrowheads, pottery, rasps and daub -------------

JAMES ANQUANDAH: AN INVESTIGATION OF A KINTAMPO COMPLEX SITE AT BOYASI HILL, NEAR KUMASI, GHANA.

Figure 5. The stratigraphic sections in Pit B, Boyasi Hill.

Figure 6. Polished stone axes.

17

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

Figure 7. Polished stone axes from Boyasi Hill.

Figure 8. Polished stone arrow heads.

18

JAMES ANQUANDAH: AN INVESTIGATION OF A KINTAMPO COMPLEX SITE AT BOYASI HILL, NEAR KUMASI, GHANA.

Figure 9. Biconically-perforated stone bead.

Figure 10. Calcite rasp.

arrowheads have been reported from the site of Dabakala in Cote d’Ivoire (Chenorkian 1983:134-36).

interesting findings of the excavation was the occurrence of 7 polished greenstone arrowheads and a few fragments of the same (Figure 8). Pressure-flaked stone arrowheads are known from Kintampo complex sites, such as Ntereso (Davies 1966), and one such pressure-flake arrowhead was found in level 1 of Pit D (V93). However, the Boyasi arrowheads are ground and polished bifacially to a fine point, which when hafted renders them potentially deadly weapons. All the specimens were excavated from Pit D (V93) which flanks the multi-grooved granite outcrop. Recently, finds of Kintampo complex polished stone

The distribution of grinding grooves on granite rock exposures at Boyasi has been discussed (Bredwa-Mensah 1984). While these grooves may have partly resulted from grinding of food products, it is evident that the hill community specialised in a variety of stone industries including tools for wood-working and perhaps subsistence practices, hunting equipment and stone beads.

19

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA One large biconically-perforated stone bead was found in Pit B (Figure 9) and a roughout of a quartz stone bead was retrieved from Pit C (R102). Eight fragments of the so-called stone ‘rasps’ or ‘cigars’ were found (5 in level 2 and 3 in level 1) which measure between 0.2 and 0.6cm in thickness. I am grateful to Dr W.B. Jones of the Geology Department, University of Ghana, Legon who carried out mineralogical analysis of these ‘rasps’ as well as specimens of polished stone axes, arrowheads, beads, pottery, and daub. From Jones’ thin-section analysis of the rasps, it appears that they were made from very fine grained calcite or dolomite (Figure 10). These materials, unlike the raw materials used for the other artefacts found at Boyasi Hill, were, according to Jones, probably traded into Boyasi from quite a distance away from the Kumasi area.

3) Pottery. The 1976 and 1982 surveys and excavations yielded some 866 potsherds whose provenance is given in Table 2. Their general characteristics are identical to those of 140 potsherds collected from the sample units in the transect surface survey of Bredwa-Mensah (1984). i) Pottery Fabric. The Boyasi pottery is generally brownish or greyish in colour, gritty and coarse and contains large quartz and mica inclusions. The clay raw material, according to Jones’ mineralogical analysis contains angular fragments of quartz, microdine and quartzite. Most potsherds were found to be heavily weathered. On the whole it appears that the material used for potting is similar to that used for the wattle-and-daub technology. ii) Pottery Forms. 96 potsherds were found to be classifiable on the basis of form. Using form and shape as diagnostic traits, the pottery has been divided into six classes –

2) Architectural Remains. In the course of preexcavation ground survey, concentrations of daub were recorded in several locations below the summit of Boyasi Hill, especially on the gradual slopes around rock exposures. The excavated pits A-E all produced evidence of daub. Jones conducted mineralogical analysis of a specimen of daub from level 2 of Pit D (V93) and reported that it contained angular fragments of quartz, microdine and mixtures of quartz and microdine as well as muscovite flakes. This particular daub specimen, like many others, exhibited broad channels or hollows representing pole impressions, probably the product of wattle-and-daub technology.

Class A (Figure 11) – Large thick-walled pots and dishes with heavy everted rims, typical of Kintampo complex tradition. Average thickness is 1cm-1.7cm. This class constitutes 67% of the total classifiable number of potsherds. Class B (Figure 12) – Large but thin-walled pots and dishes with everted rims. Average thickness is 0.4-0.6cm. This class provides 11% of the total form categories.

Figure 11. Large thick-walled vessels.

Figure 12. Large thin-walled vessels.

20

JAMES ANQUANDAH: AN INVESTIGATION OF A KINTAMPO COMPLEX SITE AT BOYASI HILL, NEAR KUMASI, GHANA. Unit Pit A Pit B Pit C Pit D Pit E Hilltop Surface Total

Level 1 19 28 14 51 33 ---------145

Level 2 67 210 25 319 42 ---------663

Total 86 238 39 370 75 58 866

Table 2. Sherd totals from Boyasi Hill.

Figure 13. Thin-walled bowls, dishes, and globular vessels.

Figure 14. Decoration styles of pottery from Boyasi Hill.

21

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

Class C (Figure 13) – Medium and small-sized thinwalled bowls and dishes with in-turning rims. Constitutes 9.5% of the total. Class D (Figure 13) – A variation of Class C but has a flattened rim. Constitutes 5.5% of the total. Class E (Figure 13) – Bowls with plain rims. 4%. Class F (Figure 13) – Thin-walled pots and bowls with channelled rims. 3%. iii) Pottery Decoration. Figure 14 provides an illustration of the Boyasi ceramic decorative techniques and the resultant patterns. Except for a few minor variations, they follow the general pattern of Kintampo complex decorative style as attested in several Ghanaian sites. Out of 866 sherds, 618 are plain, 5 are characterised by red slip coating (which can be categorized as both a functional surface treatment and as producing a decorative effect) and 243 have ‘plastic decoration’. The pottery is thus divisible into 5 decorative classes –

Figure 15. Terracotta figurine.

Class 1 – Impressed Decoration (n=143, 57.6%). 1a. Square-toothed comb stamp decoration. 1b. Stamped impression (possibly by means of a ‘rasp’). 1c. Carved stamp impression.

Discussion The investigations at Boyasi Hill have so far been largely of an exploratory nature. The data obtained is inadequate to warrant any forthright conclusions. Of course, some positive data has been obtained in relation to:

Class 2 – Incised Linear Decoration (n=44, 17.7%). 2a. Incised vertical and horizontal straight lines. 2b. Incised criss-cross lines. 2c. Incised chevron motif.

i) The question of the cultural identity of the site, namely that it is a Kintampo Complex site. ii) The question of sedentism and the extent of the prehistoric settlement. iii) Technology and industries practiced at the site. iv) Some resources locally available and some of those which may have been imported from outside the Boyasi area.

Class 3 – Impression and Incision/Grooving (n=52, 20.9%). Combinations of class 1 and class 2 illustrated in 3a, 3b, 3c. Class 4 – Broad Channelling with Thumb Impression (n=2, 0.8%).

Although the lithic industry points to a mixed economy of hunting and perhaps cultivation, no faunal or botanical materials were recovered that would provide some concrete biological information on food production or the gathering of wild crops. Also, it has so far not been possible to obtain some chronometric age estimations for the site. However, on grounds of pottery typology as compared to pottery styles from other Kintampo sites, the Boyasi prehistoric settlement may be guess-estimated to around 1500-500 bc.

Class 5 – Red Slip Coating (n=6, 2.4%). 4) Terracotta Figurine. Boyasi Hill furnished one surface find of a terracotta figurine of what appears to be a ‘dog’ which was attached to a pot (Figure 15). The fabric of the piece is identical to that of the domestic pottery found in the excavations. Terracotta figurines of animals such as ovicaprids, tortoise, lizards etc. have been recorded in Kintampo Complex contexts at Ntereso (Davies 1966), Kintampo (Flight 1976), Chekru (Calvocoressi 1975), Mumute and Bonoase (Dombrowski 1976, 1980), and Daboya (Gavua 1985). The Boyasi ‘dog’ supplements not only the accumulating data attesting to the Kintampo as Ghana’s earliest context for production of ‘art’ but also the indirect circumstantial evidence indicative of the practice of animal domestication.

During the 1982 fieldwork at Boyasi it was possible to carry out a reconnaissance survey of the newlydiscovered Kintampo Complex site of Nkukua Buoho (1°38’W, 6°47’W) located 10km northwest of Boyasi as well as the Wiwi-Buruburo complex located 8km southeast of Boyasi. Nkukua Buoho is probably the most impressive of the Kintampo Complex sites located in the Kumasi district. It is a granite hill which rises to about 350m and extends at a guess about 700m in diameter. In the course of the reconnaissance, the site was found to be 22

JAMES ANQUANDAH: AN INVESTIGATION OF A KINTAMPO COMPLEX SITE AT BOYASI HILL, NEAR KUMASI, GHANA.

Figure 16. Pottery and rasp from Nkukoa Buoho, near Kumasi, Asante.

littered with numerous remains of daub, comb stamped pottery (Figure 16), polished axes, rasps, bead polishers, microliths and microcores.

Actes du Ve Congres Panafricain de Prehistoire et de l’Etude Quaternaire. Tenerife: Museo Arqueologico, pp. 27-42.

From the Boyasi and Buoho results, it would appear that the granite inselbergs located in the forestlands of the Kumasi district were heavily occupied in prehistory and the entire area therefore needs to be scoured in detail, followed by large-scale excavations.

DOMBROWSKI, J.C. 1976. Mumute and Bonoase: Two Sites of the Kintampo Industry. Sankofa 2: 64-71. DOMBROWSKI, J.C. 1980. Earliest Settlements in Ghana: The Kintampo Industry. (In), Leakey, R.L. and Ogot, B.A. (eds.), Proceedings of the 8th Pan African Congress of Prehistory and Quaternary Studies. Nairobi: International Louis Leakey Memorial Institute for African Prehistory, pp. 261-262.

References ANQUANDAH, J. 1976a. Boyasi Hill – A Kintampo ‘Neolithic’ Village Site in the Forest of Ghana. Sankofa 2: 92.

FLIGHT, C. 1976. The Kintampo and its Place in the Economic Prehistory of West Africa. (In), Harlan, J.R., De Wet, J., and Stemler, A.H. (eds.), The Origins of African Plant Domestication. The Hague: Mouton, pp. 211-221.

ANQUANDAH, J. 1976b. Excavation at Boyasi Hill, Kumasi – A Preliminary Report. Nyame Akuma 6: 33-35.

GAVUA, K. 1985. Daboya and the Kintampo Culture of Ghana. Unpublished M.A. Thesis, Department of Archaeology, University of Calgary.

ANQUANDAH, J. 1982. Rediscovering Ghana’s Past. Harlow: Longman. BREDWA-MENSAH, Y. 1984. The Kintampo ‘Culture’ of Ghana. A Case Study of the Boyasi Hill Site. Unpublished B.A. Long Essay, Department of Archaeology, University of Ghana, Legon.

NEWTON, L.E. and WOODELL, S.R.J. 1976. A Newly Discovered Site for the Kintampo Culture. Sankofa 2: 1922. NUNOO, R.B. 1959. Archaeological Survey at the College of Technology. Journal of Kumasi College of Technology 1: 17-19.

CALVOCORESSI, D. 1975. A Terracotta Figurine from Chekru, Brong Ahafo Region, Ghana. West African Journal of Archaeology 5: 205-208.

NUNOO, R.B. 1966. Buruburo Factory Excavations. (In), Actes du Premier Colloque International d’Archaeologie Africaine. Fort Lamy, pp. 321-333.

CHENORKIAN, R. 1983. Ivory Coast Prehistory: Recent Developments. African Archaeological Review 1: 127142.

WOODELL, S.R.J. and NEWTON, L.E. 1975. A Standard-Winged Night-Jar Breeding in the Forest Zone of Ghana. Nigerian Field 40: 4.

DAVIES, O. 1966. The Invasion of Ghana from the Sahara in the Early Iron Age. (In), Cuscoy, L.D. (ed.), 23

Molluscs in Archaeological Reconstruction: The Kpone Coastline, Ghana, as a Case Study Fritz Biveridge Department of Archaeology, University of Ghana, Legon, P.O. Box LG3, Legon, Ghana

Furthermore, by studying the composition of molluscs over a wide area, scientists can establish environmental variations of that region over time (Sparks 1969: 323). They therefore provide clues to the severity or magnitude of such changes. Their relevance as agents in reconstructing ancient environmental conditions is thus enormous. Moreover, molluscs feature prominently in the study of human dietary patterns through time and space, providing a similar type of evidence that archaeologists obtain from studying faunal and floral remains at hunter gatherer sites (Meighan 1969: 415). Since prehistoric times, edible molluscs such as clams, oysters and scallops have served as important sources of protein, minerals, and vitamins in many societies in Ghana as attested by the recovery of shells of edible species at the Gao Lagoon site near Kpone and at various Kintampo complex sites.

Introduction Molluscs like potsherds have played a significant role in the reconstruction of human culture and environment in Ghana. Their usefulness can partly be attributed to the hardy nature of their shells which enables them to withstand the vagaries of the weather and other environmental elements, sometimes over several millennia. Another unique feature about molluscs is that their shells retain much of their natural colorations and other external characteristics such as shape and linear markings (Abbot 1998). The species type can thus be identified after several millennia especially when it is preserved in a well aerated and oxidized environment. Molluscs also provide archaeologists with one of the most dependable sources for dating archaeological sites under investigation. This dating method hinges on the principle that shells contain conchylin, an organic protein and calcium carbonate. It has been established by scientists that the latter decreases with time and the quantum or rate of decrease is known. Thus, by dissolving the shell in acetic acid and measuring the ratio of conchylin relative to calcium carbonate, the age of the shell can be established and by association the age of the site. Koptal (1981: 6) has noted that the fossilized shells of molluscs can also serve as chronological time clocks for establishing various chronological ages of the earth’s crust. They are also important for petroleum geologists as they serve as an aid for locating oil and coal fields.

Brief Description of Molluscs Molluscs are soft bodied invertebrate animals, many of which live inside shells (Gupta and Malik 1977). A small number however do not possess shells. Molluscs can be classified broadly into seven classes namely Monplacophora, Aplacophora, Polyplacophora, Gastropoda, Scaphopoda, Bivalvia and Cephalopoda. Twenty six species belonging to four phyla namely Gastropoda, Cephalopoda, Scaphopoda and Bivalvia were recovered from both archaeological and ethnographic contexts in the research area which was the focus of this study. The following species were identified as belonging to Gastropoda. They were Conus papilionaceus, Tympantonus fuscata, Planex lineatus, Conus ambiguous, Cerithium atratum, Cymbium cymbium, Gadina afra, Thais haemastoma, Neritha senegalensis, Cyraea steriomaria, Tympanotonus fuscata var radula, Nassa semistriata, Turritella meta and Semifusus morio. Those identified under the phylum Bivalvia were Arca afra, Arca senilis, Tagelus angulatus, Arca gneisis, Iphigenia rostrata, Donax acutangulus, Pendalion amomiodae and Pecten. Under the phyla Cephalopoda and Scaphopoda, only one species each was identified, namely Sepiella ornate and Dentalium respectively. These mollusc species constituted the focus of the research described here.

Elsewhere, they have served as important clues for reestablishing paleo climatic conditions (Mozley 1982). Unlike humans who have been able to successfully adapt to different ecosystems, the majority of molluscs are adapted to specific environmental conditions and minor changes in environmental conditions can lead to their annihilation (Sparks 1969: 400). According to Edmunds (1978), only a few species types living along the sea shore can withstand alternate exposure to air and submersion in sea water for very long periods. For example, periwinkles cannot withstand long periods of submersion and can only survive on the upper levels of rocky boulders located on the shore. Limpets survive only a few inches below sea level while Drupa are adapted to the sea floor. Mussels bore into rocks and will only be found in areas conducive to such settlement, while other species like Turitella meta survive only in brackish freshwater. Their recovery at places known not to be their natural habitat is suggestive that there may have been a change in environmental conditions.

The distinctive attributes of Gastropoda are that they have spirally coiled shells, heads with tentacles, eyes, and feet adapted for creeping (Figures 1 and 2). They also have a well developed set of radula teeth. They can be found in marine water, brackish freshwater and terrestrial habitats. 25

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

Figure 1. Diagram of gastropod shell showing various parts.

Figure 2. Diagram of live gastropod showing various parts.

Bivalvia are the most numerous after Gastropoda with a population of some 10,000 different species made up mostly of clams, mussels, cockles and oysters. The distinctive feature about them is their shell which is divided into two halves. These are hinged together to enclose the whole body. They also have hatchet shaped feet and have no head or radula teeth.

The Cephalopoda class lives exclusively in the sea. According to Edmunds (1978:13) this group is the most successful of all molluscs on account of their size, defensive mechanisms to ward off predators, speed of movement under water and complexity of their internal organs. Other unique attributes of this class of molluscs are their eight or more arms with many circular suckers and their parrot-like beak. The total number of the phylum Scaphopoda is about 300. Their shells are formed 26

FRITZ BIVERIDGE: MOLLUSCS IN ARCHAEOLOGICAL RECONSTRUCTION through the Kpone area. A letter dated 8th May 1699 from General Jan Van Sevenhuyen, the Dutch DirectorGeneral to his counterpart at Elmina castle attests to this. It reads:

like tusks and are open at either ends. They are generally small and rarely exceed 8cm in length and 4cm in breadth, respectively. They feed by using numerous filament-like tentacles on marine microscopic organisms such as plankton and algae.

“After the urgent request of the Under King of Aquamboe (Akwamu), I have now encouraged the trade at Ponnie (Kpone) for more than one and one half years and I propose that we establish a fort or lodge there. It has provided us with very good slaves and abundant gold and it is of such importance that the English have already insisted with the Upper King that he make us depart from there … As there are abundant and good slaves for sale, so many English and Portuguese ships sometimes get their entire load at that place. We could take over the trade if only we had a fort there … Moreover, much of the gold is also being received there even when the passage towards Accra is closed by the Upper King, because then, the other one authorizes those at Ponnie (Kpone) to stay open” (Dantzig 1978: 70).

The Research Area The Kpone coastline stretches about five kilometers west of the port city of Tema (Figures 3 and 4). Much of the coastline consists of a ridge which is between 18-40m above sea level. Two lagoons, the Gao and Laloi lie immediately west and east of it respectively. The shoreline especially from the Tema New Town end to Kpor Etee, a small fishing community located west of Prampram is characterized by several rocky outcrops with only a few patches of sandy shoreline lying in between. The research area is occupied by the coastal Dangbe, an ethnic group with close linguistic and cultural affiliation to the Ga (Kropp Dakubu 1997). According to early European records (e.g. Barbot 1752; Bosman 1967), the ancient people of Kpone played an important role in political and commercial affairs in this region during the pre colonial era. The area was also an important entrepot where slaves, ivory, and gold from the hinterland were exported to Europe between the 16th to 19th centuries. European export merchandise including items such as glass beads, ceramics, alcoholic beverages, textiles, guns and tobacco pipes were also imported to the Guinea Coast

The foundations of an ancient Danish fort and some canons at Kpone are vestiges attesting to this activity in the past, and according to oral traditions, the main economic activity of the indigenes in the past was fishing. They also undertook farming, salt production and exploited different molluscs types for subsistence and for export to several inland ethnic groups at Shai and Ladoku.

Figure 3. Geographical setting of Kpone.

27

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

Figure 4. Map showing distribution of shell middens.

28

FRITZ BIVERIDGE: MOLLUSCS IN ARCHAEOLOGICAL RECONSTRUCTION Research Approach and Methodology

Research Results and Discussion

Despite the centrality of mollusc research to ecological and cultural heritage reconstruction, very little attention has been paid to it in Ghana. This research was therefore focused upon developing an understanding and appreciation of the place of molluscs in the environmental history and cultural practices of the Dangbe people who occupy the Kpone coastline. The main thrust of the research was archaeological and it was undertaken in phases between 2004 -2006. The field researches were timed to coincide with the dry season when much of the overlying surface vegetation in the area had withered. This allowed for a more detailed observation of surface data. The archaeological aspect of work also involved mapping. A variety of documentary sources mainly from early European records and ethno historical data from selected indigenes at Kpone was also used. Two sites, Sega and Kpone West, both presently unoccupied, were selected for archaeological investigation. These two sites, according to ethno historical traditions of the indigenes, were occupied by the ancestors of the Kpone people.

The study of molluscs recovered from the research area provided valuable information about ancient subsistence patterns and economic activities. For example, unlike mollusc remains which were well represented at all levels of the three units at Sega and the trench at Kpone West, no Pisces remains were recovered from level 3 and levels 5 and 6 (the lowest levels at Sega and Kpone West respectively). Significant numbers however were recovered in the upper levels (1 and 2 at Sega and levels 1, 2, 3, and 4 at Kpone West). This indicates that the exploitation of marine resources such as fish was not a major occupation of the indigenes. It was probably not an integral part of the local economy during the occupation period represented by those levels. The exploitation of offshore and onshore molluscs in the Gao and Laloi lagoons however were well established modes of subsistence during the pre 1650 period. This suggests that the exploitation of molluscs predated marine fishing in the research area and together with farming was probably the lynch pin of the local economy during that period. The research also indicated that the exploitation of molluscs for food was not limited to the coastline. It was undertaken inland as well. Some of the molluscs recovered such as Nassa semistriata are naturally adapted to wet forest and heavy shrub regions which are located north of the research area. The research results also showed that the people who occupied the Kpone coastline in ancient times were extremely dexterous and astute fishermen who had developed advanced navigational skills which enabled them to navigate in deep waters. This was indicated by the fact that some of the molluscs recovered such as Thais haemastoma (Figure 5) and Semifusus morio are deep sea molluscs. The recovery of these species in association with large quantities of onshore molluscs in surface and sub surface contexts suggests that both types were exploited contemporaneously.

Study of the stratigraphy of the three units excavated at Sega namely, Units A, B and C and a trench excavated at Kpone West established two cultural levels for the area. A pre European level dating to the early 16th century and a post European level, late 16th to late 19th century. These dates were established using Ozanne’s (e.g. n.d.) typological dating scheme of tobacco pipes. Each unit at Sega measured 2m x 2m and was chosen randomly for archaeological investigation. The excavations were controlled vertically using 20cm arbitrary levels. The sterile level at Sega was found at a depth of 63cm. The trench at Kpone West measured 4m x 2m and hit the sterile level at 120cm. The archaeological remains recovered from Sega included molluscs, locally manufactured potsherds and tobacco pipes, and stems of imported tobacco pipes. Other finds included 7,922 stone slabs, grinding and milling stones, querns, imported glass beads which have been identified as originating from Holland and Venice, a gaming disc and a wide variety of faunal material consisting mainly of fish, mammalian, and reptilian remains. The finds from Kpone West were the same as those from Sega, albeit in varying quantity except with regard to the stone slabs which were not found at Kpone West. Quantitatively, a total of 1,696 faunal remains were recovered (1,098 from Sega and 598 from Kpone West) and the molluscs comprised 480 in number (28.30%).

The study of molluscs also provided insights into ancient trade patterns in the region. Pottery recovered from Sega and Kpone West has been identified as originating from Shai land which is about thirty kilometers north of the research area. Archaeological excavations at the Shai site of Cherekecherete revealed a wide dispersal of pottery characterized by depictions of elegantly embossed and stylized cattle, goats and human female heads. Anquandah (1982: 116) suggests trade as the medium through which Shai pots were sent to Kpone by the people of Ladoku. Mineralogical analysis of the potsherds supported this notion of trade between the two groups. It revealed a matrix of the following oxides as the principal mineral constituents --- garnet, pyroxene, schist and plagioclase. An ample quantity of garnet in particular was present indicating that the base materials used for their production were from the Basic Dahomeyan geological formation. The presence of epidote, pyroxene and hornblende further supports this assertion. These mineral

The collection of supporting ethnographic data was undertaken in phases using questionnaires. The questions focused on how different molluscs are trapped and exploited in the community. Another phase involved identifying relevant resource persons and respondents in the locality to interview. These included retired and active fishermen, women engaged in the mollusc preservation industry and persons in the building industry who also use molluscs for various purposes. 29

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA been an important activity during prehistoric times. Ethnographic research indicates that beads were used as ornamental jewellery in ancient times. This practice is still prevalent in the area and is a common sight during festive occasions of the people like the Homowo festival held in August every year. Some mollusc beads may have been exported to other ethnic groups living in the hinterland. Bowdich (1819: 38) who visited Kumasi in 1817 and met with the Asantehene Osei Bonsu and his entourage reported, “he wore a fillet of aggrey beads round his temples, a necklace of gold, cockspur shells strung by their largest ends”. Cockspur shells are of marine origin and indigenes who undertook bead manufacturing may have done so during the off season fishing period probably to supplement their income.

constituents have occurred consistently in pottery produced by Shai potters (Anquandah 1982: 115-116). Further archaeological evidence supporting the possibility of trade between the ancient indigenes of Kpone and their inland Dangme neighbours was provided by the excavations at Ladoku hill top by a team from the Department of Archaeology, University of Ghana under the direction of James Anquandah (1978, 1979, 1982b). The artifacts recovered from a circular mud incinerator included several European imports such as – 1. Two pieces of blue and white German stone ware (17th century). 2. Five pieces of pottery manufactured at Stoke on Trent and Bristol in England (19th century). 3. Four pieces of Dutch and English clay pipes (17th18th century). 4. Several glass beads from Holland and Venice (17th century). 5. A large number of cowrie shells of Cypraea moneta and Cypraea annulus types (17th century).

The study of molluscs played another significant role in establishing the dietary patterns of the ancient people of Sega and Kpone West. The combined evidence (archaeological, historical, ethnographic, and ethno historical) indicated that the ancestors of the Sega people exploited a wide variety of natural resources for food. Molluscs constituted an integral part of this variety and their importance in the past can be inferred from the large number of molluscs remains recovered. Arca senilis and Pecten were the most exploited species since they constituted 42% of total molluscs recovered at the two sites. The least exploited varieties were Nassa semistriata, Turritella meta, Semifusus morio and Thais haemastoma. It appears from ethno-historical sources that molluscs were exploited all year round and peaked during the off fishing season. Mollusc meat may have played a significant role as a protein supplement during the prehistoric period.

The large quantity of marine molluscs recovered from Shai land further supports the claim that molluscs may have constituted an important trade item between the two groups who have the same ethnic origin according to oral traditions. Commodities from the research area which may have been exchanged for Shai pottery and terrestrial molluscs like Thais haemastoma may have included salt, fish, and marine molluscs. An early European assertion supporting the notion that molluscs constituted an important trade item between the people of the Kpone coastline and their inland neighbours was documented at Elmina in 1518 by a Spanish lawyer. It was later translated by R. Barlow in 1541. It reads:

The ethnographic study revealed that the feet of molluscs constituted the most prized part and were processed in a variety of ways with cooking and smoking being the most important. During periods of bumper harvests, the excess was strung on small wooden sticks and smoked in large clay hearths to preserve them for future use and for export. The inedible parts like the intestines and opercula were removed and stored in cans and used as bait to trap fish at sea and in the two lagoons located nearby.

“Between this Giaga (Gao) and the castle of the Myne (Elmina) there be goodlie serves where they gather mouche gold and carry it to the castle of the Myne and sell it to the Portugals………in tracke (exchange) of rynges of Latyn (brass) and coper and certain red shelles (shells) of fysshes and clothe of colours” (Barlow 1932: 1056).

The shells of some molluscs were also used in the traditional building industry long before the arrival of Europeans to the Guinea Coast. According to respondents, they were used in a variety of ways. First, mollusc shells were used as a bonding material to hold stone blocks in position during house construction. This involved pounding the empty shell into powder and mixing it with sea sand. Reindorf (1966: 264) for example, attests to this practice. He reports that, “shell picking was a very profitable occupation for women at the time because all forts and tanks built by the Europeans on the coast were made with lime obtained by burning the shells with wood and it was better than that obtained from Europe”. The respondents also stated that the shell and sand mixture made good plastering material.

The name Giaga mentioned by Barlow is the Gao lagoon which lies immediately west of the Kpone township. Dangme oral traditions indicate a long history of salt trading with other ethnic groups. This has been buttressed by Reindorf who wrote that: “…towns along the coast chiefly applied themselves to the salt making industry. Those who did this acquired great riches because the demand from the interior was constant and if peace could be attained the coastal towns could have been the most prosperous on the Gold Coast” (Reindorf 1966: 63). The recovery of locally manufactured mollusc beads at Sega is suggestive that bead manufacturing may have

Ground shell was also mixed with water and used as paint 30

FRITZ BIVERIDGE: MOLLUSCS IN ARCHAEOLOGICAL RECONSTRUCTION to whitewash walls to beautify them. Different colours were derived by cooking the barks of certain trees and mixing the end concoction with the paint. According to the high priest of Kpone (Wolomo), in the past, the body fluids of some octopus species mixed with the ground shell provided an attractive black coloration for painting the outer walls of shrines housing some local deities and gods.

sea shells were in much demand among the coastal people of Ghana “because in those districts, many thunderbolts fell from the sky and those people believed that any man whatsoever who carried one of those shells was protected from the thunderbolts. Accordingly, the Castilians imported them from the Canary Islands exchanging them for 20 to 30 pesos of gold on the Gold Coast”.

A wide variety of bivalves were used as wall protection devices in traditional architecture. They were pressed into the wall while it was still leather hard during construction and became embedded in the wall when the mortar hardened. During the wet season, they served as protection devices preventing the rain from directly eroding and washing away the clay wall and thus prolonging the life span of houses. It was a waterproof mechanism central to clay wall construction in this part of Ghana. They were also used as cooling devices to keep the inside of rooms cool by preventing the sun’s rays from directly heating the walls.

The Woleitse also intimated that in the past, shells were used to portray a man’s social status in the community. For example, on his left wrist, he wore a unique shell bead traditionally called “afli” to show he was the Wolietse of Kpone. Attempts to identify the species type were, however, not possible because the shell was already worn out and all reference marks were lost. Moreover, answers from other respondents in the research area appear to indicate that the use of shells as a method of communicating information was not limited to the Kpone people alone. It was the practice among other ethnic groups located along the coastal belt of Ghana. Reindorf, for example, (1966: 119) has noted that among the Ga people, the number of cowrie shells placed around the middle leg of a stool showed how many slaves a man owned. He noted further that:

Ethnographic research also revealed that the shell of some molluscs served other utilitarian purposes in ancient times. For example, the whole spire and parts of the whorl of an empty Thais haemastoma mollusc was used as a communicating device by fishermen at sea. According to the Wolietse (chief fisherman), this device was very handy in times of distress and was used to draw the attention of other fishermen to oneself especially in the night and during foggy weather conditions when visibility was poor.

“A single red shell (lamie) on a string as a necklace is a symbol of silence after many troubles, either by repeated deaths or any other calamity. A red shell is also broken when hostile parties are reconciled in a king’s court. Each party appoints a representative, the one for the injured takes a stone, the other holds part of the same stone. The king appoints one to represent him who also joins in holding the stone. After that the shell is broken. A bit of the broken shell is given to each party as a memorial of peace and the rest is thrown into the sea. We call this Lamidswa.” (Reindorf 1966: 119).

The mollusc was turned into a communication device by abrading the entire outer lip area which is toothed with 2 to 3 rows of rounded tubercles on a flat stone to smooth the rough edges. A small hole was drilled at the apex and when air was blown into the enclosure, a high shrill and sharply penetrating sound was made. It was also used to make sweet music during festive occasions by continuously blowing air into the apex and controlling the pitch.

According to Madam Amouku, a herbalist and traditional birth attendant at Kpone, the body fluids and secretions of some molluscs such as the ground shell of Sepia have immense curative properties and are thus used to prepare herbal concoctions to cure hydrocephalous. Also, according to the respondent, in the past, the pulverized shells of Arca senelis (Figure 5) and Arca afra were used as facial talc. The process involved carefully washing the shells to remove dirt and sand. They were then roasted in containers at very high temperatures to break them into pieces, and then pulverized and sieved to remove the nonpowdered remains. Furthermore, according to Tetteh Appia, an elder of the Appia We clan at Kpone, the shells of some molluscs served as a medium of exchange and a store of value in ancient times. He, however, could not name the species type that was used for these purposes.

According to the Wolietse, the sharp ventral edges of some molluscs such as Cyrena were also formerly used as a sharpening and cutting tool to produce special spears used for spearing big fish like sharks. They were also used as scrapers to make fiber ropes from vegetal matter which were then used as fish traps. Some molluscs were also used in the performance of ritual in the research area. Because of their esoteric nature these rituals were not revealed to me. However the Wolietse asserted that they had to do with the appeasement of certain marine spirits prior to the commencement of the major fishing season which is between July and September every year.

The ethnographic research also revealed that the body parts of some edible molluscs such as the intestines and opercula were used as bait to trap fish in the Gao and Laloi Lagoons. Some molluscs such as Conus papilionaceus (Figure 5), Cerithum atratum (Figure 5), Turritella meta, Tympanotonus fuscata var radula and

It is a common belief among the fisher folk in the research area that the shells of immature Pecten (Figure 5) provide spiritual protection against calamities and misfortune at sea. They were strung at the base area of the umbo and hung around the neck. Garrard (1980: 25) records that in 1474, the Castilians found that large red 31

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

Figure 5. Some examples of molluscs recovered.

32

FRITZ BIVERIDGE: MOLLUSCS IN ARCHAEOLOGICAL RECONSTRUCTION especially in the fishing industry has adversely affected the traditional mollusc industry along the Kpone coastline. For example, of the twelve different ways in which molluscs were exploited in the past, only four are presently being practiced. These are as food, bait, in goldsmithing and as a toy. Moreover, except as food, where its usage is relatively consistent, the use of mollusks for these other three purposes is gradually declining. Oral information gathered from several resource persons who are all indigenes also revealed that molluscs play a significant role in the non-material culture of the people. For example, mollusks figure extensively in proverbs, with the most commonly used being “ke nshor woo, Ngler woo”. Translated it means, “if the ocean will not sleep, Ngler (pecten) will not sleep”. The study also indicated that different species of molluscs are exploited in various ways by different ethnic groups for different purposes. For example, while cerithium atratum is not exploited for food at Kpone, it is a delicacy among the Efutu and Ewe peoples who live west and east respectively of Kpone. Similarly, among the indigenes in the research area, Turritella meta is widely used as bait while among the Fanti, it is a delicacy. The significance of this for Ghanaian archaeologists is that where accumulations of such shells occur, their use must be considered vis a vis the ethnic group under study and sweeping generalizations that Mollusc “X” is used as food in area “A” and hence used similarly in area “B” can be misleading.

Tympanotonus fuscata were not consumed in the area even though they are delicacies among other coastal ethnic groups. The entire body parts were used as bait. Similarly, the shell of Sepiella ornata is an integral part of the tool kit of goldsmiths in the research area. Its importance lies in the fact that it is not very hard like other shells and allows the goldsmith to carve a hollow of the projected design in the shell into which molten gold is poured. When it solidifies, the ornament takes the shape of the hollowed design and is taken out of the shell to be polished. The shells of some Bivalves such as Arca senilis, Pecten and Arca afra were also used as storage devices for gold dust and vegetable oils like shea butter which in the past were used as body creams. The shells of some molluscs were also modified and utilized as needles to sew torn nets. Molluscs used for this purpose were Dentalium, Murex cornutus and Murex brandarex. They are open at either end with the anterior end being broader than the posterior end. The shell is characterized by several spikes. They are referred to as “abui waah” among fishermen in the area because their shape, colour, and form are reminiscent of spears. The process involved first breaking off the longest of the spike like spines and abrading the broken point to smooth the edges and prevent it from accidentally injuring the user. A small hole was then drilled at the anterior end to allow for stringing. Summary and Conclusions

It is also possible to postulate based on the findings from the study that local environmental conditions in the research area have been relatively stable since the Late Stone Age. This assertion is based on, for instance, the study by Sparks (1969: 400) that indicates that various molluscs are adapted to specific environmental conditions and thus by considering the presence or absence of certain species within a specific region, past environmental conditions can be reconstructed or established. In 1977, Dombrowski (1977) excavated the Late Stone Age site of Gao Lagoon which lies immediately west of Kpone. Among the material remains recovered from the excavation were mollusc shells which were identified as Arca senilis, Arca afra, Tympantonus fuscata and Tympantonus fuscata var radula. These species were also recovered from both surface and archaeological contexts during the course of the research described here. That these molluscs species have survived in the research area since the Late Stone Age to the present is thus suggestive that local environmental conditions have been relatively stable.

The research revealed that the exploitation of molluscs on the Kpone coastline has a profound history and that it constituted an integral part of both the material culture and other aspects of the lives of the people of the region. Hence the indigenes had a profound and intimate knowledge of its importance and exploited it for a wide variety of purposes. Its close proximity to the Gao Lagoon and the Guinea Coast which were the natural habitats of molluscs were important factors which stimulated and prompted its exploitation in the past. The research findings also indicate that mollusks are still an important natural resource which play a central role in the socio-economic and cultural lives of the people. Except for ritual purposes, the exploitation of molluscs is not limited to a particular social stratum but encompassed the entire community and that there exists no taboos or laws prohibiting its use by any section of the society. It has also been noted that the exploitation of molluscs is not limited only to fulfilling the dietary requirements of the indigenes but extends to other aspects of subsistence. The following different uses and contexts were identified in addition to the above, namely as bait, as a trade item, in the traditional building industry (paint and lime), human bodily decoration (bead), as a cosmetic (talc), as a musical instrument, in rituals associated with traditional religion, in traditional medicine, goldsmithing, as status symbols and as gaming objects.

But in concluding it should also be noted that the material expressions of the coastal Dangbe people must be appreciated beyond the domain of functionalism. Indeed, the material expressions which operate in relation to molluscs straddle the spheres of aesthetics and symbolism with reference to one’s position, rank, and prestige in the community. Archaeologists will be missing a lot of great insights into their cultural heritage if they ignore these realities.

The introduction of foreign imports and technologies 33

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

Mollusc Species

Natural Habitat

Vernacular Name kporlor kakladah gbutey

Nerta senegalinsis Pendalion amoniodes Donax acutangulus Sepiella ornate Arca senilis

Marine Marine Marine Marine marine / lagoon

Cymbium cymbium Semifusus morio Cerithium atratum Planex lineatus Arca afra

Marine Marine Lagoon Marine Marine / lagoon

Nsorh waah acloo momochiley aglorh Nshor aordey

Pecten

Marine

Ngler

Conus papilionaceus Cypraea sterioaria Tagelus angulatus Conus ambiguus Gadina afra Tympatonus fuscata Tympantonus fuscata var radula Nassa semistriata Thais haemastoma

Marine Marine Marine Marine Marine Marine Lagoon

akwasioondan Tsina muu afraah kokowah awomiyeh momochiley momochiley

Terrestrial Marine

kokoha

Iphigenia rostrata Turretilla meta Arca gneisis Dentalium

Marine Marine Marine / lagoon Marine

gbutey momochiley shinnie Tsina bla

Food Food / musical instrument / communication device Food Bait Food / traditional architecture Used as needles to sew torn nets

Murex cornutus Murex brandaris

Marine

---



Marine

---



shinnie

Traditional Uses Food Food Food Gold smithing / food Food / traditional architecture / ritual Food Food Bait Food Food / traditional architecture / gold smithing Food / traditional architecture / religion/ ritual / gold smithing Food Food Food Food Food Bait Bait

Table 1.The biological names, vernacular names, uses and natural habitat of molluscs recovered from the research area.

BARBOT, J. 1752. A Description of the Coast of North and South Guinea. (In), Churchill, A. and J., Churchill Collection of Voyages and Travels. Vol. 5. London: Golden Ball, Paternoster Row.

References ABBOT, T. R. 1998. World Publishing Ltd.

Sea Shells. New York: Trans

ANQUANDAH, J. 1982a. Rediscovering Ghana’s Past. Harlow: Longmans.

BARLOW, K. 1932. A Briefe Summe of Geographie. Taylor, E.G.K. (ed.). London: Hakluyt Society.

ANQUANDAH, J. 1982b. Archaeological Reconnaissance in the Shai Hills, Ghana. Nyame Akuma 21: 15-17.

BOSMAN, W. 1967. A New and Accurate Description of the Coast of Guinea. London: Frank Cass. BOWDICH, E. T. 1819. Mission from Cape Coast to Ashante. London: Frank Cass.

ANQUANDAH, J. 1979. Accra Plains Archaeological and Historical Project. Nyame Akuma 15: 14-18.

KROPP DAKUBU M. G. 1997. Korle Meets the Sea. A Socio-linguistic History of Accra. New York: Oxford University Press.

ANQUANDAH, J. 1978. Accra Plains Archaeological and Historical Project. Nyame Akuma 12: 24-27.

34

FRITZ BIVERIDGE: MOLLUSCS IN ARCHAEOLOGICAL RECONSTRUCTION DANTZIG, VAN, I. 1978. The Dutch and the Guinea Coast 1674-1742. Accra: Ghana Academy of Arts and Sciences. DOMBROWSKI, J. C. 1977. Preliminary Note on Excavations at a Small Midden near Tema, Ghana. Nyame Akuma 10: 31-34. EDMUNDS, J. 1978. Sea Shells and other Molluscs Found on the West Africa Coast and Estuaries. Accra. Ghana Universities Press. GARRARD, T. 1980. Akan Weights and the Gold Coast Trade. Harlow: Longmans. GUPTA, R. K. and MALIK, B. S. 1977. A Treatise on Mollusca. Calcutta: Pragati Prakash Press. KOPTAL, R.L. 1981. Mollusca. Calcutta: Rostogi Rastogi Press. MOZLEY, A. 1982. An Introduction to Mollusca Ecology. London: H. K. Lewis and Company. MEIGHAN, C. W. 1969. Molluscs as Food Remains in Archaeological Sites. (In), Brothwell, D. and Higgs, E. (eds.), Science in Archaeology. London: Thames and Hudson, pp. 415-422. OZANNE, P. No Date. Tobacco Pipes of Accra and Shai. Legon: Institute of African Studies (Mimeographed). REINDORF, C. C. 1966. The History of the Gold Coast and Ashantee. Accra: Ghana Universities Press. SPARKS, B. W. 1969. Non Marine Mollusca and Archaeology. (In), Brothwell, D. and Higgs, E. (eds.), Science in Archaeology. London: Thames and Hudson, pp. 395-406.

35

Excavations at Fort Amsterdam, Abandze, Central Region, Ghana J. Boachie-Ansah Department of Archaeology, University of Ghana, Legon, P.O. Box LG 3, Accra, Ghana.

In May 2003, Dr. Roberta Capella, Francesco E. Ventura, Dr. Anna Belardinelli, Edi Grassi and Paola Fioretti, officials of Scuola Edile di Perugia, a school and a nonprofit professional training body based in Italy made a proposal for the restoration and adaptive re-use of Fort Amsterdam (Capella et al 2003). In this proposal, Scuola Edile di Perugia in association with local partners comprising the Abandze Development and Welfare Association and the Vocational and Rehabilitation Centre of Biriwa, intended to initiate and implement a tourismrelated, community-based project for the people of Abandze and neighbouring areas. The project involved the restoration of Fort Amsterdam through on-the-job sharing and transfer of technical know-how to local artisans including the staff of the Ghana Museums and Monuments Board; equipping and training of personnel for the restored Fort Amsterdam to undertake activities linked to tourism, and the establishment of a Vocational Training Centre in Abandze as an income generation centre.

Introduction From 6th to 8th May 2003, a newly appointed Board of Directors of the Ghana Museums and Monuments Board of which the writer was a member undertook a tour of the forts and castles in the Central and Western Regions of Ghana. The purpose of the tour was to enable the board members to familiarize themselves with, and observe at first hand, the state and condition of the forts and castles. Among the forts visited was Fort Amsterdam, located on an isolated hill in the township of Abandze in the Central region of Ghana (see Figure 1). While at Abandze, it was observed that erosion was negatively impacting on both the archaeological record and the standing architecture in the western part of the fort. Several smoking pipes, beads, bottles and pottery (both local and European) had been eroded from their original contexts. Rodents had also made holes in one of the rooms on the ground floor of the fort and exposed pottery and marine shells.

Figure 1. Map showing the location of Abandze

37

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

Carolusburg had been taken over by the Dutch (van Dantzig 1980: iii).

The proposal also made provision for space in the fort for temporary art exhibitions and a permanent exhibition on the history of the fort. The Vocational Training Centre, also to be located in the fort, was to train the youth of Abandze and neighbouring areas in bead-making, basketand cane- weaving, wood- carving, batik- making, textileprinting, catering and house-keeping. Although the memorandum of understanding submitted by the Ghana National Museums and Monuments Board, the body entrusted with the care of Ghana’s historic resources has not yet been responded to by Scuola Edile di Perugia, it is hoped that the restoration works will soon begin - an anticipation borne out of the desire to arrest the decay and destruction of the fort.

To achieve this, the British blockaded that part of the coast. This infuriated the Dutch who made frequent attacks on English ships. In 1664-65 England sent out two fleets against the Dutch colonies in the Americas and Africa. The Dutch colony on the Hudson River known as New Netherlands was defeated by England and the capital of the colony was named New York after James, the Duke of York. On the Gold Coast, the English captured Dutch forts at Takoradi, Shama and Mori and lodges at Anomabu and Egya. The Dutch immediately sent Admiral De Ruyter to the Gold Coast to recapture their lost possessions. With the help of Africans from Asebu, Kormantin and Anomabu, De Ruyter succeeded in capturing the English fort and headquarters at Kormantin and the lodges at Anomabu and Mori which had been abandoned by the English in order to reinforce their position at Fort Carolusburg in Cape Coast (Claridge 1964: 112-118; Ward 1958: 84-85; van Dantzig 1980: 33-34).

However, the proposal for the restoration and adaptive reuse of Fort Amsterdam did not make adequate provision for an archaeological impact assessment and excavations to retrieve data before commencement of the restoration works. It was partly for this reason, and partly for the fact that erosion was doing a great deal of damage to the archaeological record that the author undertook to conduct excavations at Fort Amsterdam. It was hoped that the excavations would provide important data on the history of Fort Amsterdam and thereby provide useful information for the proposed exhibition in the fort.

The Dutch reconstructed their new acquisition at Kormantin and named it Fort Amsterdam to avenge their loss of New Amsterdam to the English. After the fall of Fort Kormantin, the Cape Coast Castle captured by the English from the Dutch in 1664 (Nørregard 1966:25-26), became the new English headquarters on the Gold Coast. The Dutch West India Company, which took over Fort Kormantin (renamed Fort Amsterdam), became bankrupt in 1674 and was replaced by the newly formed New West India Company, which took over the administration of the fort (Lawrence 1963:245; Ward 1958: 145-146). In 1782, the English captured Fort Amsterdam (Lawrence 1963: 245; Ward 1958: 145-146). The Dutch regained possession of the fort by treaty in 1785 (Lawrence 1963:245; van Dantzig 1980: 63). In 1806, Ashanti warriors invaded the fort in search of two renegade Assin chiefs, Kwaku Aputae and Kwadwo Otibu who had shown disrespect to the Asantehene (the King of the Asante) and killed his messengers (Ward 1958: 147-151). The Dutch commander of the fort surrendered the fort without any attempt at resistance. The Fanti of Anomabu who were allies of the English also besieged the fort and captured it in 1811. According to Lawrence (1963: 245), the fort was never occupied after its invasion by the Anomabo. However, Ward (1958:175, footnote) maintains that the Dutch did not abandon the fort until 1816.

History and Description of Fort Amsterdam In 1631, Arent Groot, a disgruntled former employee of the Dutch West India Company on the Gold Coast changed his allegiance to an English Company known as the Company of Adventurers of London Trading to Guynney and Binney and negotiated on behalf of the company for a piece of land at Kormantin to build a fort (van Dantzig 1980: 33; Ward 1958: 77). A small fortified lodge was built on a hill near Kormantin (where the township of Abandze is now located), about 20km east of Cape Coast. The lodge was gradually expanded (van Dantzig 1980: 21) from about 1638 (Lawrence 1963:245), but was destroyed by fire supposedly caused by the Dutch in 1640 (Atta-Yawson n.d: 4; van Dantzig 1980: 21). After the destruction of the fort by fire, a stronger fort was rebuilt on the site (van Dantzig 1980: 22) probably between 1645 and 1647 or certainly before 1661 (Lawrence 1963: 245-247). The fort, named by the English as Fort Kormantin, was transferred in 1660 to a better-organized company, the company of Royal Adventurers of England Trading to Africa. This company, patronized and supported by the English court, engaged in a brisk trade at Kormantin and established new lodges at Anashan (modern Biriwa) about 5km to the west, at Winneba, 60km to the east, and at Accra, 110km to the east (van Dantzig 1980:33). Competition and rivalry ensued between the Dutch and the English as each of the two powers wanted to control trade on the coast (Nørregard 1966:25-26). The English tried to prevent the Dutch from re-settling Cabo Corso (i.e. Cape Coast) where in 1664 a Swedish fort, Fort

In March 1867, the British and the Dutch governments signed a treaty to exchange some of their forts in order to divide the coastline of the Gold Coast into a Dutch and British stretch and to get rid of the inconvenience of having Dutch and British forts mixed together (Claridge 1964:557-575; Ward 1958: 236-240). The exchange which took effect on 1st January 1868, fixed the frontier between the Dutch and English stretch at the Sweet River just to the east of Elmina. By the treaty, the British ceded 38

J. BOACHIE-ANSAH: EXCAVATIONS AT FORT AMSTERDAM, ABANDZE, CENTRAL REGION, GHANA it to “supply 3,000 slaves a year to the West Indies, and to take over the fort at Kormantin and build new forts or lodges at Cape Coast, Anashan, Egya, Komenda, Winneba and Accra” (Ward 1958: 83).

to the Dutch four forts namely Beyin, Dixcove, Sekondi and Komenda, all located to the west of the Sweet River. The British on the other hand received in exchange four forts namely, Mori, Amsterdam (Kormantin), Apam and Accra to the east of the Sweet River (Figure 1). In February 1872, the Dutch ceded all their possessions on the Gold Coast including all the Dutch forts to the British government (Claridge 1964: 626-630; Ward 1958:247248). Fort Amsterdam therefore became the possession of the British. Despite the fort falling back into British hands, it has retained the name Fort Amsterdam.

Enslaved Africans from Kormantin earned a reputation for bravery and stubbornness and organized slave revolts in Jamaica in the middle of the eighteenth century (Boahen1966: 109-110). The name Kormantin is idolized in Jamaica, Suriname, Panama, Antigua and the Virgin Islands. Due to the popularity of the name Kormantin, several descendants of enslaved Africans from the United States, Jamaica, Suriname and Panama have paid visits to Fort Amsterdam (Atta-Yawson n.d. : 16-17).

Fort Kormantin (as it was called before the Dutch named it Fort Amsterdam) was engaged in the gold and slave trade and was probably the first slave prison built on the Gold Coast (van Dantzig 1980: 22). The English relied heavily on the Gold Coast for the supply of labour to the West Indies. After the conquest of Jamaica in 1665, there was a great demand for slaves. Kormantin was the English headquarters and many of the early enslaved Africans sent to the West Indies were shipped from Kormantin. The English West Indian planters, even after the English had lost the fort at Kormantin to the Dutch, continued to refer to the slaves from the Gold Coast as “Cormantins or Coromantese” (van Dantzig 1980: 2223).

Fort Amsterdam (Figure 2) is located on the summit of an isolated hill in the town of Abandze, a settlement which grew out of a small village founded by the local population who provided labour for the building of the fort. The hill on which the fort is located is a little below 61m above sea level. To the south, the hill falls precipitously to the sea, while the slopes elsewhere are too steep for easy walking. By 1682, steps had been cut into the hill to improve the path to the fort (Lawrence 1963: 245). The fort is built of Elminian sandstone (DeCorse 2001: 91), and red and yellow bricks. The fort is entered through an archway in the north curtain (see Lawrence 1963:45).The fortifications consisted of two rectangular bastions in the north-west and south-east corners (see Lawrence 1963:246, Fig. 19; van Dantzig 1980: 22). The walls between the bastions and three of the bastions were filled with earth; but the fourth bastion, the one to the south-west which has now disappeared,

The demand for slaves by the English planters in the West Indies is indicated by the fact that the charter of the company of Royal Adventurers of England Trading to Africa, (a Company formed in 1662) not only gave the company the monopoly of trade from the Straits of Gibraltar to the Cape of Good Hope, but also empowered

Figure 2. Fort Amsterdam showing the north-east bastion in ruins

39

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA College, Achimota, from 1951 to 1956; and by the African Descendants Association Foundation in the 1970s (Kankpeyeng and DeCorse 2004: 109). The restoration did not make use of appropriate architectural reconstruction techniques and a considerable portion of the reconstruction subsequently collapsed. In 1970-72, the Ghana Museums and Monuments Board with donations from the Netherlands restored the entrance to the fort. Criticisms raised against the reconstructions centre on the failure to make detailed plans prior to demolishing portions of the fort in preparation for the reconstruction and the inconsistency of the restored structures with the structural histories of the fort (see Kankpeyeng and DeCorse 2004: 109).

was hollow, and according to van Dantzig (1980: 22), had a grated ventilation hole in its roof. This was where enslaved Africans were kept. The slave prison was reached from a courtyard by a passage vaulted with yellow bricks (Lawrence 1963: 247). A line of two- and three- storeyed buildings was built to the south (Lawrence 1963: 247; van Dantzig 1980: 22). One-storeyed vaulted rooms were built in the western portion of the fort (Lawrence 1963: 245-247). The fort served as a store for trade goods and as living quarters for the permanent commercial and military staff. The two and three storey buildings were used as living quarters. The fortifications on the landside were meant to defend the fort against the local people while those on the side of the sea were defensive mechanisms against European competitors. Two walls of a sub-triangular courtyard joined the north-west bastion to the north-east bastion (see Lawrence 1963: Fig 19; van Dantzig 1980: 22). This courtyard served as the fort’s market (AttaYawson n.d.: 8). A second courtyard was flanked by the rooms of the ground floor and the two storey buildings to the west and south.

The Excavations From 22nd March to 11th April 2006, excavations directed by the writer took place to retrieve data likely to be destroyed and disturbed by erosion, the activities of rodents, and by an impending restoration of Fort Amsterdam by the Scuola Edile de Perugia. In all, four units were excavated. The first pit, labelled Pit 1 (Figure 3) and measuring 2m2 was opened in a room in the ground floor located in the south-west immediately next to the south-west bastion. It was in this room that rodents had exposed ceramics and shells. The second pit labelled Pit 2 (Figure 3) and also measuring 2m2 was opened in a rectangular courtyard immediately adjoining the room in which Pit 1 was sited. The third excavated unit (Figure 3) was a trench measuring 2m x 8m and labelled Pit 3. The trench was opened 6m to the east of the south-east bastion in a sloping area where erosion had exposed ceramics, beads, smoking pipes, broken bottles and shells. The last pit, Pit 4 (Figure 3) was a test pit measuring 1m2 in a room directly opposite and next to the courtyard in which Pit 2 was sited and immediately next to the south-east bastion. It was hoped that the excavations would provide data on the lifeways of the occupants of the fort as well as provide information on the chronology of the fort.

Other features of the fort include two underground cisterns, an ammunition room, a condemned cell, two ponds, presumably for rearing geese, and a tunnel connecting the fort to the beach. In addition to the outer courtyard and its gates, other features which are known from historical sources but which have disappeared are the two southern bastions that defended the fort from the coast, the roofs of the two-storeyed building to the west and the tunnel connecting the fort to the beach. Several renovations and changes were made to the fort as it changed hands from one trading company to the other and from the English to the Dutch. Starting as a small English trading post in 1631, it is said to have converted into a fort in 1638 and after its destruction by fire in 1640, it was rebuilt and provided with bastions by the English between 1645 and 1647 or certainly before 1661 (Lawrence 1963: 245-247; van Dantzig 1980: 21-22). By 1682, during a period of Dutch possession, the steps on the pathway to the fort had been improved (Lawrence 1963: 245). By 1704, the two-storey building to the south had reached its final shape and height. The third storey forming a tower at the west end was probably added sometime before 1786, again probably by the Dutch who regained possession of the fort by treaty in 1785 (see Lawrence 1963: 245-248). A staircase in yellow brick was also added against the courtyard frontage during the time of Governor Swerts, a Dutch Governor from 1684 to 1689. There are indications that parts of the older structures were demolished by the Dutch as can be seen from the drawings of 1682 during a period of Dutch possession (see Lawrence 1963: 248).

Owing to lack of funds and logistical problems, the south-west bastion which was used as the slave prison could not be excavated. This area has not been affected by erosion and was therefore reserved for future excavation. Future research will involve the excavation of the bastion and the old Kormantin Village where the Africans who traded with the Europeans on the coast resided. This, it is hoped, will provide data on the nature of the transformations that occurred in the era of the Atlantic Slave trade. Apart from Pit 3 which was excavated with an arbitrary level of 20cm to a context, all the excavated units were excavated with an arbitrary level of 10cm. Pit 1 had a simple stratigraphy, the colours of which were determined by the Standard Colour Soil Chart. The levels with cultural materials attained a depth of 2.1m. The first level was characterized by a loose black humus soil (Hue 10YR1.7/1). The soil was so loose that it easily gave way at the slightest touch of a trowel. The looseness of the soil

Several attempts have been made to conserve and restore Fort Amsterdam. The ruins of the fort were cleared and conserved by A. W. Lawrence (1963: 245), first Director of the National Museum of the Gold Coast and first Professor of Archaeology at what was then the University 40

J. BOACHIE-ANSAH: EXCAVATIONS AT FORT AMSTERDAM, ABANDZE, CENTRAL REGION, GHANA

Figure 3. Ground floor plan of Fort Amsterdam showing excavated areas

Levels 3 to 5 were also characterized by the same loose dark brown soil of the second level. They produced only a few finds comprising marine shells and sherds from locally manufactured vessels. The remaining levels with cultural materials, Levels 6 to 21, ( i.e. from a depth of 60cm to 2.1m) were characterized by a compact reddish brown soil ( Hue 2.5YR 4/8) which, except for a few shells and potsherds, was devoid of finds. These levels were probably filled with soil when the fort was originally built. Of particular interest is that part of the foundation of the fort was reached at a depth of 1.25m. It was constructed with red bricks and measured 34cm wide. The pit was underlain by compact bright reddish brown sub-soil (Hue 2.5YR 5/8) with glittering specks of mica.

was explained to the excavator by Mr. Ebenezer Eshun (a native of Abandze) in the course of the excavations. According to him, the room in which Pit 1 was dug was filled with soil when Dr. Robert E. Lee, a member of the African Descendants Association Foundation used the fort as his residence in the 1970s. Finds from the first level included 8 broken pieces of yellow bricks, presumably Dutch; the handle of a mason’s head pan, 4 iron nails; unidentifiable iron objects; an iron bolt; a piece of iron with two holes, each measuring about 3cm in diameter; an iron plate measuring 23cm² with a hole measuring 2cm2 ; 2 pieces of blue polythene; 3 small pieces of European glazed pottery and marine shells similar to those incorporated in the walls of the fort.

Pit 2, dug in the rectangular courtyard, also had a simple stratigraphy. The levels with cultural materials attained a depth of 38cm below which appeared a bright reddish brown sub-soil (Hue 2.5 YR 5/8). The levels with cultural materials consisted of a black humus soil (Hue 10YR 1.7/1). Finds from the pit include a corroded iron object, probably a knife blade; a fragment of a yellow brick; a piece of asbestos; a plastic object probably carved from the sole of a bathroom slipper; 2 plain Dutch “twisted square” wound glass beads, all from Level 2; a fragment

The second level consisted partly of the same black humic soil that characterized the first level and partly of a loose dark reddish brown soil (Hue 2.5YR 4/1). Finds from the level consisted of seeds of elaies guineensis (probably brought in by rodents or incorporated in the soil used in filling the excavated area); an iron hoe blade, marine shells; 2 fragments of a coconut husk; a black gun flint (probably English) and marine shells.

41

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA and it is probable that the pottery was produced in the area around Fort Amsterdam .The colour of the inner and outer fabric is grey, black or brown. A few of the sherds contain pieces of quartz, mica, and what appears to be hornblende, a blackish shiny mineral. Several sherds are blackened with soot - an indication that they were used for cooking on open fires.

of a roofing tile and European and locally-manufactured pottery from Level 3, and fragments of European smoking pipes and glass bottles from all the levels. Pit 3, the trench dug outside the fort close to the southeast bastion and for which an arbitrary level of 20cm was adopted was the deepest of all the pits. The layers with cultural materials attained a depth of 2m. The natural stratigraphy consisted of a top layer of the same black humic soil characteristic of the other pits. The humic soil was underlain by dark reddish brown soil (Hue 2.5YR 4/1) which was in turn underlain by a reddish brown soil (Hue 2.5 YR 4/8). The sub-soil was bright reddish brown in colour (Hue 2.5 YR 5/8) and contained weathered sandstone and specks of mica.

Burnishing was unpopular and was applied on 174 (17.7%) sherds. A total of 300 sherds (30.5%) were decorated and the remaining 682 (69.5%) are undecorated. Grooving was the predominant decoration applied on the pottery. A total of 36 sherds constituting 12.0% of the total decorated sherds are decorated with single circumferential grooves while 231 sherds (77% of the decorated sherds) are decorated with multiple circumferential grooves. Grooves therefore constitute 89% of the total decorations on the sherds. The grooves were applied on rims, necks, bodies and carinations (Figure 4: a, e 5: c). Other decorations are single (1.7%) and multiple (3.0%) incisions on rims, bodies, shoulders and necks; applied plastic decoration or embossment (1.0%) on bodies and rims; rim-lip notches (0.3%); dot stamps (0.3%) on the rim; channelling (1.0%) on rims and necks and triangular stamps (1.3%) on body sherds. In some cases, two decorative motifs were combined on the same vessel. The combined decorations are multiple horizontal grooves and triangular stamps (0.7%) on the rim and body; embossment and single incisions (0.3%) on the neck; multiple grooves and dot stamps (0.7%) on the body and rim; multiple grooves and multiple incisions (0.3%) on the body; and multiple incisions and triangular stamps (0.3%) on the body.

As in the case of Pit 1, the soils of the excavated levels were not compact but soft and loose. Specks of charcoal and ash were interspersed with the soil. The looseness of the soil, the numerous broken artefacts and the specks of charcoal and ash suggest that the pit was opened in an area where refuse from the fort was discarded. The specks of charcoal and ash suggest that the refuse dump was burnt to control rapid accumulation. Soil and charcoal were mixed in such a way that it was impossible to find sufficient charcoal to make up samples for radiocarbon testing. The bulk of the finds consisted of fragments of glass liquor bottles, several thousands of which were recovered from the excavations. Bones, shells, and pottery, both imported and locally manufactured, were found in all arbitrary levels. Two fragments of locally manufactured smoking pipes, one consisting of the bowl and the other of the stem, were found in Level 6 of the trench. Whilst Pit 4 which produced fragments of locally manufactured pottery was rather shallow and attained a depth of 17cm. The top 10cm consisted of a black humus soil (Hue 10YR 1.7/1) which was underlain by a dark reddish brown soil (Hue 2.5YR 4/8) of about 10cm depth. The sub-soil consisted of a bright reddish brown clayey soil (Hue 2.5 YR 5/8) with mica and weathered sandstone.

Sherds found on the surface of the site are similar to those retrieved from the excavations in terms of fabric, paste, characteristics and decoration. The surface sherds are decorated with single and multiple horizontal grooves, single incisions and a combined decoration of multiple grooves and dot stamps, and single grooves and triangular stamps. Except for the rims, some of which join the neck at sharp angles, the pottery is predominantly characterized by vessels with flowing profiles. That vessels with angular profiles are few is indicated by the fact that of the total of 982 sherds, only 5 (0.5%) are carinated. Jars and bowls constitute 46% and 53.6% respectively of the total vessel forms.

The Finds In summary, the finds from the excavations consisted of locally-manufactured pottery, European and Chinese glazed pottery, locally-manufactured and European clay smoking pipes, a piece of asbestos, clay roofing tiles, gun flints, iron and cuprous objects, objects made of polythene, what appears to be a hammerstone, fragments of drinking glasses, broken bottles, glass beads, and shells and bones.

The first jar form consists of vessels whose everted rims curve sharply at both interior and exterior (Figure 4: a-b) and with a rim diameter ranging between 12 and 18cm. These constitute 31.4% of the total jar forms and 14.5% of the total vessel forms. 13 out of the 16 sherds belonging to the jar form are blackened with soot from use on open fires, an indication that the vessels were used for cooking. Only 3 of the sherds are burnished. The outer and inner fabric of the sherds contain mica. Decorations consist of circumferential grooves on the neck, inner parts of the rim (Figure 4: a) and shoulders, and oblique incisions on the shoulder (Figure 4: b). Rim

Locally-manufactured Pottery A total of 982 potsherds were recovered from the excavations. Of these, 14 (1.4%) were retrieved from Pit 1; 25 (2.5%) from Pit 2; 933 (95.0%) from Pit 3 and 10 (1.0%) from Pit 4. The fabric of the pottery is coarse and well-fired and contains fragments of mica which were certainly part of the clay used in manufacturing the vessels. Mica abounds in the clays of the Abanzde area 42

J. BOACHIE-ANSAH: EXCAVATIONS AT FORT AMSTERDAM, ABANDZE, CENTRAL REGION, GHANA

Figure 4. Vessel forms of locally-manufactured pottery

lips are rounded. Sherds of this vessel form were recovered from all the 13 arbitrary levels of Pit 3.

from 14 to 28cm, and rim lips are rounded. Decoration consists of grooves on the neck and inner parts of the rim. A few of the sherds are burnished and 16 out of the 26 sherds belonging to the vessel form are covered with soot. Sherds of this vessel form were found in Levels 3 of Pits 1 and 2, and from Level 1, 5 to 8, 10, 11 and 13 of Pit 3.

The second jar form (Figure 4: c) constitutes 60.0% of the total jar forms and 23.6% of the total vessel forms. The everted rim curves sharply in the interior but is smoothly curved at the exterior. The rim diameter ranges 43

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

Figure 5. Vessel forms of locally-manufactured pottery

The third jar form (Figure 4: d) is a vessel with a rounded rim lip and rim diameter of between 14 and 18 cm. The short everted rim curves smoothly at the exterior and interior. Represented by only 3 sherds (all covered

with soot), the vessel form constitutes 5.9% of the total jar forms and 2.7% of the total vessel forms. Sherds are characteristically unburnished and decoration consists of multiple circumferential grooves on the shoulder. Sherds 44

J. BOACHIE-ANSAH: EXCAVATIONS AT FORT AMSTERDAM, ABANDZE, CENTRAL REGION, GHANA

Figure 6. Vessel and base forms of locally-manufactured pottery

belonging to the form were retrieved from Levels 5, 9, and 11 of Pit 3.

deep jar. The jar form constitutes 3.9% and 1.8% of the total jar and total vessel forms respectively.

The fourth jar form (Figure 4: e), represented by 2 unburnished sherds from Level 9 of Pit 3 and decorated with multiple circumferential grooves on the neck and a curvilinear groove on the shoulder, has a rather short, everted squat rim. It is characterized by specks of mica in the inner and outer fabric and has a rim diameter of about 20cm. The curvature of the body wall suggests that it is a

The fifth jar form (Figure 4: a), represented by a single sherd from Level 12 of Pit 3 is also characterized by a short squat rim with a rounded lip, and has a rim diameter of 12cm. It is unburnished and its decoration consists of a single circumferential groove in the inner part of the rim and triangular stamps aligned in a horizontal pattern on the shoulder. It is covered with soot and has specks of mica in the inner fabric. The jar form constitutes 2.0% of 45

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA consists of chanelling immediately below the exterior part of the rim lip. The form constitutes 40.7% and 21.8% of the total bowl and vessel forms respectively. The sherds were retrieved from Levels 4 to 10 of Pit 3. The second bowl form (Figures 5: d-e, 5: a) has an incurved rim with squared lips and a carination (Figure 5: d) or a ridge (Figures 5: e, 6: a) separating the rim from the body. The rim diameter ranges from 16 to 26cm. Represented by 33 sherds, the form constitutes 55.9% of the total bowl forms and 30.0% of the total vessel forms. 19 out of the 33 sherds are burnished. The decoration consists of circumferential grooves on ridges and rim lips; a combination of horizontal and curvilinear grooves just below the exterior of the rim lip; channelling just above the carination, or circumferential grooves immediately below the carination. A few of the sherds are characterized by channelling on the rim lips. Sherds belonging to the bowl form were recovered from Levels 1 to 3, and 5 to 10 of Pit 3. The last bowl form (Figure 6: b) represented by 2 burnished sherds from Level 7 of Pit 3, is a deep hemispherical bowl whose squared rim lip extends a little beyond the body wall. The rim diameter measures 14cm. The vessel form constitutes 3.4% of the total bowl forms and 1.8% of the total vessel forms. In all, 3 base forms were identified. The first (Figure 6: c), retrieved from Level 6 of Pit 3 is elliptical in shape; the second from Level 9 of Pit 3 is round with a concaveshaped bottom (Figure 7: a), and the third (Figure 7: b) from Level 1 of Pit 3, appears to be a fragment of a pedestalled base. European and Chinese Pottery Several European potsherds were found on the surface of the area surrounding the fort. Pictures of these as well as those from the excavations were submitted to Dr. Jerzy Gawronski, Head of the Archaeology Division, City of Amsterdam for identification and comments. The sherds were also submitted to Mr. L. B. Crossland of the Department of Archaeology, University of Ghana, Legon, for his comments. The descriptions of the potsherds are based on their comments for which the author is most grateful.

Figure 7. Base forms of locally-manufactured pottery

the total jar forms and 0.9% of the total vessel forms respectively. The last jar form (Figure 5: b), represented by 3 unburnished sherds and constituting 5.9% of the total jar forms and 2.7% of the total vessel forms, has a short everted rim and a straight neck. Decoration consists of minute triangular stamps and a single circumferential groove on the rim lip. It has a rim diameter of 14cm. Sherds of this jar form were retrieved from Levels 1 and 6 of Pit 3.

The sherds from the surface consist of deep pearlware bowls with underglaze polychrome designs of floral patterns in pinkish red, green, bright blue and orange probably made in the Netherlands and dating to c. 18201840 (see South 1977: 212; DeCorse 2001: 153); pearlware bowls with cut sponge-stamped floral and geometric patterns in blue, red and brown from the Netherlands and of the period c. 1830-1873 (see DeCorse 2001: 153); pearlware plate rim fragments with floral and geometric patterns made in England c. 1790-1840 (see DeCorse 2001: 153) and simply identified as English industrial earthenware by Dr. Gawronski; English pearlware plate sherds with purple transfer-printed decoration in floral design and made c. 1780-1830 (see

The bowl forms are all hemispherical in shape. The first bowl form (Figure 5: c) is represented by 24 sherds, 18 of which are blackened with soot. The rim is incurved or inverted and the rim diameter ranges from 16 to 24cm. Only 5 of the sherds are burnished and the decoration 46

J. BOACHIE-ANSAH: EXCAVATIONS AT FORT AMSTERDAM, ABANDZE, CENTRAL REGION, GHANA

Figure 8. (a) Dutch Delftware; (b) Fragment of a Bellarmine jug; (c) English creamware with feather-edged design; d) Chinese porcelain with underglaze blue decoration and Batavia glaze.

with red paste and covered on one or both sides with a brown lead-glaze; wheel-made white and cream-paste cylindrical storage jars with green lead-glaze and storage jars with handles, buff paste and mustard-yellow glaze all probably made in the Netherlands.

South 1977: 212); a probable English green-edged pearlware plate with single or multiple lines along the rim, and manufactured c. 1795-1840 (see South 1977: 212); an English creamware plate rim with a “royal pattern” decoration of the period c.1765-1785 (see Hume 1976: 116); English creamware of types made c.17601815 in Leeds, Liverpool, and other production centres in England, and characterized by transfer-printed decoration with floral designs in an overglaze black transfer print (DeCorse 2001: 153; Lewis 1999: 112; South 1977: 212); and a fragment of a deep bowl of a refined earthenware with a foot ring and in a deep yellow colour of a type made in Staffordshire and other British centres c. 17621785 (see Hume 1976: 126; Lewis 1999).

Dutch delftware sherds were also recovered from Levels 1, 4, 5, 6, 7, 11, and 12 of Pit 3 and from the surface of the site. These include undecorated Delftwhite tin-glazed sherds of ointment pots and a rim of a plate belonging to the period c. 1630-1800 (see Hume 1970: 62-65; South 1977: 211); fragments of plates and a dish cover (Figure 8: a) with blue monochrome hand-painted decoration of naturalistic, geometric and floral motifs of the period c. 1625 to 1800, and fragments of soup plates with medium buff paste and printed motifs over tin glaze of the same period as the dish cover and plates just described.

Out of a total of 224 identifiable imported sherds from the excavations, 68 (30.4%) are probably from the Netherlands, 58 (25.9%) are probably from England, 53 (23.7%) are German, and 45 (20.1%) are Chinese. Most of the European sherds found in the excavations date to the eighteenth century. They consist of fragments of storage jars, mugs, bowls, ointment pots, plates and soup plates, dish covers, saucers, jugs, bottles with handles and tea bowls. A variety of European coarse earthenware was found from Level 1 through Level 4 and from Level 6 to 10 of Pit 3. All the coarse earthenware sherds date to the eighteenth century and consist of low fired storage jars ( probably from the Netherlands) with unglazed surfaces, brown paste and throwing rings which suggest they were wheel-thrown; red-paste unglazed storage jars with a redslipped exterior probably made in the Netherlands or England; low-fired buff-paste mugs with green leadglazed interior; everted and inverted kitchen utility bowls

English delftware sherds were also recovered from Levels 1, 2, 6, and 11 of Pit 3. These consist of sherds of shallow plates with flat bases and characterized by slate-blue decoration and deep pink body paste which is a peculiarity of Lambeth delft which was made in the late seventeenth, and eighteenth centuries (AngusButterworth 1970:82), and fragments of a deep tin-glazed bowl with “powdered” blue ground pigment of the mideighteenth century (see Draper 2001: 30-31). Other European sherds found in the excavations include refined earthenwares of the period between 1740 and 1870 (see Draper 2001: 41). These include Whieldon colour-glaze ware, creamware and pearlware. Levels 1, 4, 5, 6, 9-11 of Pit 3 produced fragments of small cream47

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA 210). Also found were fragments of a Westerwald blue and grey salt-glazed mug with a square cylindrical neck cordoned and decorated with a single cobalt band, and the body incised with geometric designs and stamped blue floral motifs dating to the period 1650 to 1775 (see DeCorse 2001: 153; South 1997: 210); sherds of an unglazed grey stoneware storage jar and the remnants of a thick-potted grey salt-glazed storage jar of a type manufactured in Frechen-Rhineland in the seventeenth and eighteenth centuries.

coloured saucers in “basket weave” pattern and decorated with tortoise shell coloured glazes of green, yellow, manganese and grey. This type of pottery was made in Staffordshire and other centres in England c. 1740-1780 (see Draper 2001 :45; Lewis 1999: 83; South 1977: 211). Creamware sherds include plate rim fragments with “feather-edged” relief moulded designs (Figure 8: c) of the period c. 1765 to 1820 (see DeCorse 2001: 152; Hume 1976: 116); a cream-coloured plate rim fragment with a “royal”pattern dating to c.1765-1785 (Hume 1976: 116); a fragment of a yellow deep bowl with a foot ring and of a type made in Staffordshire and other British centres c. 1762-1785 (see Lewis 1999; Hume 1976: 126) and a surface find in the form of a creamware plate fragment with floral design in an overglaze black transfer print of a type produced in Leeds, Liverpool and other production centres in England c. 1760-1815 (see DeCorse 2001: 153; Lewis 1999: 112; South 1977: 212).

Stonewares from England were also recovered from the excavations. These consist of white salt-glazed plate fragments with press-mould diaper and floral border decoration of a type produced in Staffordshire, Derbyshire, Yorkshire and Liverpool from about 1720 to 1775 (DeCorse 2001: 153; Draper 2001: 36-39; Miller and Stone 1970: 68-69; South 1997: 211); and a redslipped unglazed, finely grained red-brown sherd from a storage jar probably manufactured in England in the mideighteenth century.

Most of the pearlware sherds were found on the surface of the site. Only a single pearlware sherd was found in the excavations (Pit 3, Level 2). They consist of fragments of Dutch bowls with underglaze polychrome painted floral patterns in pinkish red, green, bright blue and orange made c. 1820-1840 (see DeCorse 2001: 153; South 1977: 212); bowl fragments with cut spongestamped floral and geometric patterns in blue, red and brown, probably made in the Netherlands c. 1830-1873 (see DeCorse 2001: 153); plate rim fragments with floral and geometric patterns and transfer-painted blue “willow pattern” made in England c. 1790-1840 (see DeCorse 2001: 153); purple-coloured transfer-printed plate sherds with floral design from England made c.1780-1840 (see South 1977: 212), and blue green-edged plate fragments with single or multiple lines of either blue or green along the rim, probably made in England c. 1795-1840 (see South 1977: 212).

Pottery from the Orient was also recovered from the excavations. These consist of vessels of Chinese export porcelain in the form of plates, saucers, tea bowls, bowls and mugs. As with the European pottery, most of the sherds date to the eighteenth century. They include white hard paste sherds with underglaze blue decoration with painted motifs of conventional flowers and landscape scenes (Figure 8: d). They date to the period between 1660 and 1800 (see DeCorse 2001: 152; Miller and Stone 1970: 31-90; South 1997: 210). Other sherds found consisted of fragments of saucers and tea bowls which have blue and white interior decoration with conventional flowers and landscape scenes and an overall brown “Batavian” glaze exterior. These date to c. 1690-1790 (see DeCorse 2001: 152; Hume 1976: 260). The last group of the Chinese export porcelain consists of 2 fragments of a small bowl with enamelled polychrome decoration usually described as “Imari style” over the white glaze. The paste is hard and white and the pottery type is usually dated between 1700 and 1780 (see DeCorse 2001: 152; Hume 1976: 258-259; Miller and Stone 1970: 86-87; South 1997: 210).

Fragments of German stoneware were found from Level 1 through 12 of Pit 3. These include Bellarmine or Bartmann bottle fragments with a pronounced neckcordon and a mottled orange brown iron wash under a thin salt-glaze. This kind of vessel was produced in centres on the Rhine-Frechen, Raeren, the Westerwald and possibly Siegburg. It is generally dated to the period c. 1550-1700 (DeCorse 2001: 153; Gaimster 1997: 212222; South 1977: 210). The fragments from Fort Amsterdam are decorated with a stylized beard (Figure 8: b) in the form of a series of strokes, a feature which appeared in the second quarter of the seventeenth century. The vessel type continued to be made in the Rhineland and exported through the first quarter of the eighteenth century (see Hume 1976: 57).

A fragment of thinly-moulded buff coloured plastic plate probably made in the early twentieth century was found on the surface of the site. Similar plastic wares have been found in early twentieth century contexts within the African townships adjacent to Fort Crèvecoeur (Dutch) and James Fort (English) in Accra ( L. B. Crossland, pers. comm., March, 2008). Smoking Pipes

Other German stoneware found in the excavations includes jug rim fragments (together with a handle) decorated with a broad band of manganese (purple) slip around a reeded neck and decorated with incised checquered ornament in cobalt blue. This vessel type was made in the Westerwald District of the Rhineland from the mid-seventeenth century to the eighteenth century (see DeCorse 2001: 153; Hume 1976: 281; South 1997:

Only two locally-manufactured smoking pipes consisting of fragments of the base and the stem, all from Level 6 of Pit 3, were found in the excavations. Given the fragmentary nature of the pipes, they cannot be fitted into Ozanne’s (1962, n.d.) typology which has helped so much in dating several sites in Ghana. 48

J. BOACHIE-ANSAH: EXCAVATIONS AT FORT AMSTERDAM, ABANDZE, CENTRAL REGION, GHANA

Figure 9. (Left) Probable English smoking pipe of the 17th century; (Middle) Dutch smoking pipe of the 18th century; (Right) Smoking pipe of the 19th century

Figure 10. 18th century Dutch Smoking pipe stems

mark IN. GOUDA, the place of manufacture. Other maker’s names found on stems are F. VERZYL, a pipemaker who lived in Gouda (Gerzy Gawronski, pers. comm., March, 2007); DANES; DEIONG, which stands for De Jong, the surname of the pipe-maker (Gerzy Gawronski, pers. comm., March, 2007); I. DANENS together with IN. GOUDA; LUCAS DEIONG together with chevrons in the form of the letter V and inverted V (Figure 10, top); ANB; and ANPUY and A. N. V. BEEK. Also found were partly broken inscriptions which read F. L…IN and …RAMMER.

A total of 1,783 fragments of European smoking pipes were recovered from the excavations. Of these, 1,649 (92.5%) are stems and 134 (7.5%) are bowls. Almost all the pipes are Dutch. Decoration on the stems consists mainly of rouletted lines and a series of open circles. Such decorations are typical of Dutch pipes (Walker 1975: 185-186). A few of the stems have rouletted motifs consisting of denticulated square-like impressions achieved by the use of a cog-like instrument and with the name “GOUDA” immediately below the impressions. Other stems are characterized by curved lines bordered by tiny square impressions.

The decorations and makers marks on the pipe stems clearly betray their origin. The same can be said of the bowls and decoration on the spurs. The pipes have the plane of their rims at a markedly obtuse angle to the line of the stem. The bowls are all forward drooping bowls. These are features characteristic of eighteenth and nineteenth century Dutch pipe bowls (Walker 1975:185). They also have a delicate cog-wheel denticulation round the bowl rim, a feature which was abandoned by the English from the early eighteenth century (Walker 1975:185). The bowls conform to Dutch bowls typical of the period after 1750 (see Atkinson and Oswald 1972 : 178, Fig. 79, Nos. 27 and 29; DeCorse 1998: 2, No. 2). The length of the Dutch bowls from rim to spur, the rim

Some of the stems have the maker’s mark LUCAS DEION. Others marked with the name LUCAS DEION are decorated with chevrons which appear in the shape of V and an inverted V (Figure 10, top). In some cases, the stems with the maker’s mark LUCAS DEION have tiny denticulated square impressions arranged to form a series of open squares. Such stems were recovered from Level 5 of Pit 3. Others from the same level of Pit 3 are decorated with rouletted motifs of sub-rectangular impressions enclosed by names that appear to read IAN. V. BEEK and IN. GOUDA. From the second arbitrary level of Pit 3 was found a stem decorated with rouletted impressions (Figure 10) enclosed by the name I. DANINS and the 49

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

Figure 11. Dutch smoking pipes with; (Top left): crowned 32 on the spur; (Top right): crowned AB on the spur; (Bottom left) D with incised lines; (Bottom right) crowned WS

dating from 1726 to 1925 (Duco 2003: 152); a crowned AB (Figure 11, top right) known to date to 16841873/1800 (Duco 2003: 155) together with the Gouda coat of arms consisting of a shield with two rows of vertical stars known to date from 1739 and probably abandoned sometime before 1814 (Walker 1975: 186); a crowned P, a crowned WD and a Gouda coat of arms on the side of the spur; a crowned L together with a coat of arms consisting of two vertical lines that curve to join at the base to enclose two rows of square impressions; a crowned DS; a crowned WS (Figure 11, bottom right) ; a crowned SH, a Gouda mark of 1686-1849 (Duco 2003: 160); the letter S or a serpentine figure with the Gouda coat of arms; a crowned T, a Gouda mark of 1726-1824 (Duco 2003: 153); a crowned EB, a Gouda mark of 16831720/1725 (Duco 2003: 155); a crowned EP, and what appears to be a crowned hat.

diameter of the bowl, and the diameter of the spur measure on average 4.5cm,1.9cm, and 5mm respectively. A few of the bowls have marks on them. The mark TD appears together with what looks like the letter “S” or a serpentine motif on bowls from Levels 1, 3, and 5 of Pit 3. TD pipes first appeared c. 1755 (probably in England) and were soon extensively copied by several pipe makers in Europe (Calvocoressi 1975: 199; 1977: 138; Walker 1975: 183). By about 1840, the letters were used to denote a bowl shape and TD bowls were made in Glasgow until 1967 (Walker 1975: 183). However, those found in the excavations are of Dutch origin. This is supported by the fact that the bowls conform to the shapes of Dutch pipes and the marks on the spurs are identifiable marks associated with Gouda pipes of Dutch origin. One bowl fragment is decorated with a tulip, a Gouda mark dating from 1643 to 1793 (Duco 2003: 128). Another bowl fragment from the fifth level of Pit 3 is characterized by fluting or ribbing, a feature which appeared in the late eighteenth century (Oswald 1975:9697, 111; Walker 1975: 183).

Other marks on the spur consist of a crowned 2 in use from 1679 to 1925 (Duco 2003:187); a crowned 32 (Figure 11, top left) in vogue from 1729 to 1866 (Duco, 2003: 190); a crowned 33, a Gouda mark in use from 1719 to 1854 (Duco 2003: 190) and a crowned 67, a Gouda mark of 1728-1868 (Duco 2003: 195). Also found on the spur of the Dutch pipes are a windmill, a Gouda mark dating to 1667-1902 (Duco 2003: 124); a coffee kettle, a Gouda mark dating to 1753-1788 (Duco 2003: 138); a clover leaf, in use from 1660 to 1840 (Duco 2003: 127), together with the Gouda coat of arms; a blacksmith hitting an object on an anvil, known to date from 16931753 (Duco 2003: 134); the coat of arms of the city of Gorinchem consisting of a crowned sub-rectangular

Some of the Dutch pipe bowls have unbroken spurs. Marks on the spurs attached to bowls include the letter D crossed by five horizontal lines and a single vertical line (Figure 11, bottom left); DP representing the name Daniel Passet of Gouda and dating from 1749 to 1790 (Duco 2003: 164); a crowned M K, a Gouda mark of between 1682 and 1871/72 (Duco 2003: 161) and the letter S on side of the spur; a crowned L, a Gouda mark 50

J. BOACHIE-ANSAH: EXCAVATIONS AT FORT AMSTERDAM, ABANDZE, CENTRAL REGION, GHANA figure enclosing two rows of zig-zag lines aligned horizontally and known to have been in vogue from 1675-1769 (see Duco 2003: 124); a Gouda coat of arms on the side of the spur; and a crowned pipe, a Gouda mark of 1727-1775 (Duco 2003: 139). Many of the bowls are blackened with soot, an indication that they were used in the fort before they were discarded. Four of the bowls do not conform to the shapes of eighteenth century Dutch bowls. One such bowl is a small forward-drooping bowl with a pedestal foot and a rim diameter of 1.3cm, a swelling on the back of the bowl and rouletted denticulations on the rim. The pipe conforms to the shape of English pipes usually assigned a date of 1600-1640 (see Hume 1969: 302, Fig. 97:4; Oswald 1975: 37, No. 4, Fig 3G: H). The length of the bowl from the rim to the pedestal foot is 4cm. The square platform measures 1.1cm wide. The bowl was recovered from the fifth level of Pit 3. Another pipe from the third level of Pit 3 has an erect bowl and a partly broken stem. It has a mark on the right side of the bowl that appears to read HP. It has no spur and conforms to the shape of nineteenth century bowl forms of Dutch and English pipes (see Atkinson and Oswald 1972: 178, Fig. 79, No. 31; DeCorse 1998, 2001: 164, Fig. 5.8). The remaining two bowls which do not conform to the shapes of eighteenth century Dutch bowls are all erect. One of them, found in Level 1 of Pit 3 has a small partly broken spur. The other picked from the fifth level of Pit 3 and devoid of a spur has denticulated motifs on the rim and is marked by what appears to be a swan at the underside of the bowl, close to where the stem adjoins the bowl. The bowl probably dates to the nineteenth century, a period during which erect bowls without spurs were produced by the Dutch, English, and Americans (see DeCorse 1998: 2).

Figure 12. Case bottle with everted lip, c. 1620 -1650

container for liqueurs, brandy or gin were also found, whilst fragments of a vertically straight-sided green case bottle with blopp (thickened and rolled) rim, high shoulder and a square base, and of a type made in Holland and Belgium for keeping liquor such as brandy, cognac, and gin in the period between 1770 and 1800 (Figure 14) form other glass vessel types represented.

Glass Bottles and Drinking Glasses

Also found among the bottles are full-size (Figure 15) and half size onion-shaped bottles known as “Dutch onions” of the period c.1710-1750. Such bottles were produced in Belgium, Germany and Holland (see van den Bossche 2001: 73, 121, 361) and were used as containers for wine, beer, oil, and in pharmacies as decanters. Other bottle types include a bottle blown in a funnel-shaped dip mould; a “mallet” utility bottle also known as a “Belgian mallet” (Figure 16) and made in Belgium and Britain from about 1703 to 1760, and of a type used as containers for beer, wine and other wet wares (see Davies 1972: 28; Green 1978: 121; van den Bossche 2001: 170-171). Also found were fragments of “bladder-shaped” or flattened “onion” bottles known to have been manufactured in European countries such as England, France, North Germany, Belgium and Holland c. 1730-1760 (see van den Bossche 2001: 334,338) and used as containers of wine and other wet wares; base fragments of narrow cylindrical bottles used for wine storage and probably made in England c. 1780 to 1820 (see Davies 1970: 30; Green 1978: 121), and a utility vial bottle with a flanked lip and in aquamarine colour (Figure 18, top left) probably made in the Netherlands c.1820(see van den.

Several thousand glass fragments were recovered from the excavations. Only 41 fragments could be identified by Mr. L. B. Crossland with certainty. Of these, 35 (85.4%) are probably from the Netherlands, and only 6 (14.6%) could have come from England. As with the European sherds, most of the glass bottles date to the eighteenth century. Among the bottles are a rim and base sherds of a straight-sided light green case bottle with rounded base corners and slanting shoulders (Figure 12) made in Holland or Belgium in the seventeenth century (see Hume 1982: 42; McNulty 1971: 105; Schledorn 2003: 37) and used as a multi-functional vessel for storing oils, liqueurs or medicines, and string rims (Figure 13) belonging to case bottles with nearly square and rectangular bases and of a type made in Holland and Belgium c. 1720-1750 as containers of cognac, brandy and gin. These are probably the forerunner of the case gin bottle. The rim and base fragments from a green straightsided case bottle with a narrow orifice, flared rim and a rectangular base possibly made in Holland, Belgium or Germany and usually dated from 1740 to the 1770s (see van den Bossche 2001: 258-259) and probably used as a 51

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

Figure 13. Case bottles with string rims and square bases, c.1720 -1750

Figure 14. Case bottles with blopp rims and square bases, c. 1770 -1800

52

J. BOACHIE-ANSAH: EXCAVATIONS AT FORT AMSTERDAM, ABANDZE, CENTRAL REGION, GHANA

Figure 15. Full-size ‘Dutch Onions’, c. 1710 – 1750

Bossche 2001: 361) which was possibly used for keeping perfume, balsam, stomach bitters, oil, smelling salts, or medicines.

Glass Beads Three light blue, 1 deep blue, 1 yellow, 1 white, 1 brown, 1 grey and 1 light green nineteenth century Bohemian mould beads were collected from the surface of the site. The only beads from the excavations are two colourless Dutch “twisted squares” dating from 1650 to 1750 (Francis 1994: 65). The two beads recovered from the second level of Pit 3 are turning purplish, an indication that the glass from which they were made contained manganese impurities. When such glass is exposed to the tropical sun for more than a hundred years, they begin to turn to a shade of purple (Fletcher 1975: 93).

Glass drinking vessels represented included a fragment of a colourless balustre wine glass (Figure 18, bottom left) with a finely air-twisted decoration probably made in England c.1735-1765 (see Hume 1969: 32; 1976: 19, Type xxi); a colourless wine glass with a drawn stem and a wide base (Figure 18, top right) dating to c.1725-1760 (see Hume 1976: 191, Type xvii) and probably manufactured in England or in Continental Europe; a fragment of a green balustre wine glass (Figure 18, bottom right) probably made in England c.1700-1760 (see Hume 1976: 191, Type x); and a base fragment of a colourless tumbler. 53

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

Figure 16.’Belgian Mallet’, c. 1703 – 1760

Figure 17. Cuprous Badge with the Coat of Arms of the United Kingdom for the period 1714 -1801

whole. The largest and smallest yellow brick respectively measure 19.2cm length by 9cm width and 4.1cm breadth; and 15.5cm length by 6.5cm width and 4cm breadth respectively.

Gun flints A light brown gun flint was recovered from Pit 3 Level 2, and another gun flint, black in colour, was recovered from the second level of Pit 1. English flints from the East Anglian mines were either grey or black in colour and are easily distinguishable from the French flints which are usually honey-coloured, blond or brown (Hume 1969: 220). It may be that the gun flints from Fort Amsterdam are both English and French.

It has generally been assumed that the red bricks are English and the yellow bricks are Dutch. Doorways and windows at Fort Amsterdam were coigned with thin red bricks, and these thin red bricks were also used for building the parapet of the north-west bastion. According to Lawrence (1963: 245), red bricks were made contemporaneously with yellow bricks in Holland and thus it is unwise to attribute the red bricks at Fort Amsterdam to the English. Lawrence (ibid) also argued that at Mori, which was the nearest Dutch Fort, red bricks were used in conjunction with yellow bricks. However, DeCorse (2001: 25, 91) identifies yellow bricks on the Gold Coast with the Dutch. In view of the fact that both the Dutch and the English imported most (if not all) of

Bricks 52 yellow bricks and 19 red bricks were found in the excavated pits. The yellow bricks were mainly concentrated in Level 2 through 11 of Pit 3, and the red bricks were found mainly from Level 2 to 5 and 7 to 9 of Pit 3. Many bricks were broken into pieces. Ten out of the 52 yellow bricks and 4 out of the 19 red bricks were 54

J. BOACHIE-ANSAH: EXCAVATIONS AT FORT AMSTERDAM, ABANDZE, CENTRAL REGION, GHANA Iron Artefacts All the iron objects recovered from the excavations were corroded. This makes identification rather difficult. They include 33 nails which were recovered from Levels 1 and 3 of Pit 1, the first three levels of Pit 2, and the first five levels of Pit 3. The length of the nails ranges from 5 to 10cm. Nails were among the building materials imported into the Gold Coast by Europeans (Lawrence 1963: 93).

the bricks for the building of forts on the Gold Coast from Europe (see Lawrence 1963: 90-95), it is probable that the bricks from the excavations were imported from both England and Holland.

Eight screws were also recovered from Levels 4, 5, 6 and 9 of Pit 3. The longest measures 7.3cm and the shortest 4.3cm. Other iron artefacts include a door hinge measuring 19cm long and recovered from Level 4 of Pit 3; what appears to be a remnant of a hoe blade with a tang for insertion into a wooden handle from Level 1 of Pit 3; a remnant of a spade with part of the wooden handle anchored to the socket from the second level of Pit 1; the handle of a mason’s head pan from the first level of Pit 1; a flat iron plate measuring 232 cm from the third level of Pit 1; an iron bar or rod measuring 36cm long; and a bolt measuring 7.6cm long from the first level of Pit 1 and a barrel bracket from the third level of Pit 2, the latter probably a remnant of the equipment used for strengthening wooden barrels containing wine or salted meat imported from Europe. Some of the finds from Pit 1 including nails, the spade, the handle of the head pan, the flat iron plate, the iron rod and the bolt may not be contemporaneous with the fort since the soil in the room in which the pit was opened is not the primary context but instead was deposited there and used as a filling for the floor in the 1970s as already noted. The same can be said of a piece of iron with two holes found in the first level of Pit 1 and probably part of an engine of an outboard motor used on dug-out canoes by local fishermen.

Roofing Tiles

Cuprous Artefacts

A total of 79 yellow roofing tiles were recovered from the excavations. All except one from the third level of Pit 2 were recovered from the second through the twelfth levels of Pit 3. They are all European imports as can be inferred from their fine grained paste which is not characteristic of locally-produced ceramics. They are also more highly fired than the local ceramics. A similar observation has been made by DeCorse (2001: 93) on tiles from Elmina. Lawrence (1963: 90) has observed that “none of the many experiments in baking on the spot produced durable bricks (whether for lack of the right clay or of skill), and scarcely ever did anyone [the Europeans on the Gold Coast] try to make roofing- or flooring tiles locally”. The tiles are fragmentary, slightly convex in shape and are probably Dutch.

A remnant of what looks like a cuprous bowl was recovered from Level 7 of Pit 3. It measures 7.7 cm long by 3.7cm wide. Also found in Level 12 of Pit 3 is a cuprous medal (Figure 17) with the Royal Coat of Arms of the United Kingdom for the period 1714-1801, probably brought to the site between 1782 and 1785 when the British held Fort Amsterdam.

Figure 18. (Top left) Utility glass bottle with flanged lip, c. 1820; (Top right) wine glass with drawn stem, c. 1725 -1760; (Bottom left) Colourless balustre wine glass with air-twisted decoration, c.1735 -1765; (Bottom right) Green balustre wine/cider glass

Polythene Artefacts A piece of blue polythene decorated with floral designs and criss-cross lines was recovered from the first level of Pit 1. Like the other artefacts from Pit 1, it post-dates the use of the Fort by the English and the Dutch. Bones

Hammerstone

The bones and shells were analysed by Mr. B.M. Murey, Chief Technician, Department of Archaeology, University of Ghana, Legon. In most cases, the condition of the bones was such that it was not possible to identify them to specific species. A total of 1,278 bones were

A hammerstone was recovered from the first level of Pit 3. It is made of quartzite, and is almost spherical in shape and has several abrasions, suggesting that it was used as a hammerstone.

55

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA recovered from the excavations. The bones were distributed as follows (Table 1). GENUS/SPECIES Rodentia Reptilia Cavia Porcellus (Pig) Aves Equus Caballus (Horse) Bovida Bos Taurus (Cattle) Pisces Canivore Unidentifiable bones TOTAL

BONE COUNT 14 4 2 194 104 452 98 71 2 337 1,278

The shells are predominantly those of molluscs whose habitat is either offshore, low rocky shores, mangroves, lagoons, estuaries and shallow waters and some of which would have been washed up on the beach by waves. The molluscs represented are predominantly those collected for food or known to be edible and must have been a good source of protein to the occupants of the fort. The molluscs collected for food or known to be edible are Donax acutanculus, Cardium costatum, Ostrea denticulata, Thais callifera and Achatina (see DeCorse 2001; Edmunds 1977; Hodasi 1995; Koptal 1981). The shells of these molluscs constitute 88.59% of the total shells from the excavations. Only 40 (1.9%) of the shells are terrestrial, all shells of Achatina.

% OF TOTAL 1.1 0.3 0.2 15.2 8.1 35.4 7.7 5.6 0.2 26.4 100.0 approx.

Table 1. The faunal remains from Fort Amsterdam.

Discussion

Three of the rodentia bones were identified as those of a mouse, and 11 were identified as those of the giant African rat (Cricetomys gambianus). All the reptile bones are tortoise bones. 17 turkey bones, 7 goat bones and 5 bones of a freshwater fish, Eutcopius niloticus were identified.

The locally manufactured pottery from Fort Amsterdam bears close resemblance in paste characteristics, decoration and vessel forms to that excavated by Nunoo (1957:12-14) at Asebu, a pre-seventeenth to nineteenth century site 20km to the north-west of Abandze. As is the case with the Asebu pottery, sherds from Fort Amsterdam contain fragments of quartz and mica (undoubtedly from the underlying rocks of the area) in the inner fabric. Decoration on sherds from the two sites includes grooves, incisions and dotted and triangular impressions. However, the decorations on the Asebu pottery are more variable than that of Fort Amsterdam and motifs such as chevrons, shell, seed and square impressions characteristic of the former site are absent on Fort Amsterdam pottery. Vessels forms which are common to both sites include jars with everted rims and spherical bodies (Figures 3: d , 4: a ; Nunoo 1957: Fig. 9:12,15,16,23); jars with everted rims, straight necks and spherical bodies (Figure 4: b; Nunoo 1957: Fig 9:9-11); hemispherical bowls with incurved rims (Figures 4: c-e, 5: a; Nunoo 1957: Fig.9:3-5,18-21), and splayed and pedestalled bases (Figure 6:a-b; Nunoo 1957: Fig 8:1-3, Plate vi).

Only in a few cases was it possible to determine the minimum number of individual animals represented by the bones. Altogether, 13 birds out of which 2 were specifically identified as turkeys, 1 giant African rat (Cricetomys gambianus), 1 cow, 9 bovids (not identified to species) and 1 horse are represented. 17 bones of the bovid group, 2 cattle bones and 2 horse bones have butchery marks on them. The butchery marks on 2 rib bones from a horse raise the question as to whether it was kept for ceremonial and / or military purposes or for food, and whether it was brought on a ship as butchered meat or as a live animal. Law (1980: 53) has suggested that the Dutch may have taken horses, probably from Allada in the Republic of Benin, occasionally to the Gold Coast since a Dutch writer of the mid-seventeenth century, D’Olfert Dapper, complained that “the people there [Gold Coast], if given a horse, kill it and eat it rather than keeping it for breeding”.

Certain similarities can also be seen in the pottery of Fort Amsterdam and that of Bantama, a site 28km to the west. While the pottery from Fort Amsterdam does not bear any resemblance with the pottery of the earlier occupation levels of Bantama, Phases 1 and 2, it is similar to the pottery of the later nineteenth century occupation level of Bantama Phase 3, during which period the Dutch Fort known as Fort de Veer or Veersche Schans was built in 1810/11 (see Calvocoressi 1977: 122123). At both Fort Amsterdam and Phase 3 of Bantama, pottery with flakes of quartz and mica in the fabric are common. Also common to both sites are curvilinear grooves on the body of sherds; horizontal and circumferential grooves around necks and rims; straightnecked jars with everted rims, and bowls with inturned rims and pedestal bases (see Figures 4: b-d, 6: b; Calvocoressi 1977: 127- 128, Fig. 5:1-5, Fig.5:7).

Shells Two shells of Cardium costatum and one each of Ostrea denticulata and Cypraea annulus were collected near Pit 3. A total of 2,085 mollusc shells were recovered from the excavations. The molluscs represented are Turiella meta (0.04%), Cardium costatum (24.7%), Ostrea denticulata (67.7%), Tympanotomus fusca (0.04%), Tivela tripla (0.7%), Pinna rudis (0.04%), Oliva flammulata (0.04%), Triphora sp. (0.3%), Conus genuanus (0.04%), Thais callifera (1.2%), Gadinia afra (2.5%), Achatina sp. (1.9%), Cardium kobelti (0.3%), Cypraea zonata (0.09%), Cymbium cymbium (0.04), Codokia eburnea (0.04%), Cantharus vierratus (0.09%), Cardium ringens (0.04%), Donax acutangulus (0.09%), Cardium afra (0.09%), and Lithophaga lithophaga (0.04%).

The impressive work by DeCorse (1989: 158-169; 2001: 116-123) at Elmina, 26km to the west, has produced 56

J. BOACHIE-ANSAH: EXCAVATIONS AT FORT AMSTERDAM, ABANDZE, CENTRAL REGION, GHANA pipes there for the African trade because they were cheaper than English pipes, although it was forbidden to import Dutch pipes into England (Donnan 1931: 241242). Besides, Dutch pipes were much used by the French and were also major exports from other European countries such as Denmark, Sweden, and Brandenburg, all of which were active in the West African trade (Walker 1975: 167). However, since the Dutch occupied Fort Amsterdam for a longer period than the English, the bulk of the Dutch pipes described in this paper were probably brought in by the Dutch themselves.

pottery whose vessel forms (except for the ubiquitous spherical vessels with everted rims and a few hemispherical bowls with incurved rims) do not conform to the shapes of the vessels from Fort Amsterdam. On the whole, the locally- manufactured pottery belongs to a tradition that characterizes some of the pottery from seventeenth to nineteenth century Asebu (see Walker [1975:184] for the dating of Asebu) and the nineteenth century levels at Bantama. Many of the sherds from Fort Amsterdam are covered with soot and carbon, an indication that they were used as cooking vessels on open fires. In contrast to sites occupied by local inhabitants, the potsherds from Fort Amsterdam are few, and may have been used by the native residents who provided services in the fort, and by the European residents to supplement vessels from their home countries. Enslaved Africans, particularly those kept in the slave prison are likely to have been served food in local earthenware vessels whose broken edges are not as sharp as European imported ones. Care was taken to ensure that slaves were not supplied with articles that could be used as weapons. For this reason, slaves are likely to have been served food in vessels of local pottery rather than in highly fired European and Chinese ceramics, the broken pieces of which could be used as weapons. Vessels were also likely to be broken in the usually crowded slave prison for which reason it made sense to supply prisoners with cheaper locallymanufactured vessels.

As already indicated, most of the pipes found at Fort Amsterdam are eighteenth century Dutch pipes. DeCorse (2001: 25) has aptly commented that “the predominance of Dutch pipes, wine and gin bottles [on the Gold Coast], and yellow brick on archaeological sites, along with the occasional finds of tobacco boxes and delftware sherds is a testament to the commercial success of the Dutch during the seventeenth and eighteenth centuries”. The pipe stems from Fort Amsterdam are decorated with motifs typical of Dutch pipes and the maker’s marks on the spurs of the pipes are acknowledged marks of pipes made in Gouda. All except three of the identifiable pipe bowls are Dutch, and only one probable English pipe of the seventeenth century was identified among the pipes from the fort. The Dutch presence at Fort Amsterdam can also be seen in the presence of the predominantly yellow bricks and roofing tiles found in the excavations. The predominance of the yellow bricks and roofing tiles in comparison with the red bricks at a site which, for the most part, was occupied by the Dutch would seem to suggest that most of the yellow bricks and tiles were brought in by the Dutch. The fact that they are also characterized by a finer-grained paste and are more highly fired than the local ceramics suggests that they were imported and not locally manufactured (see DeCorse 2001: 93).

If Fort Amsterdam was not occupied by the Dutch again after its invasion by the Anomabu in 1811 and by the English after 1811 as claimed by Lawrence (1963: 245), then the English occupied the fort for only 36 years, from 1631 to 1665 and from 1782 to 1785; while the Dutch occupied it for 143 years, from 1665 to 1782 and from 1785 to 1811. To what extent do the archaeological finds from the excavations reflect this long period of occupation by the Dutch and what evidence is there to show that the fort was not occupied again after 1811?

The predominance of Dutch sherds such as the coarse earthenware storage jars, mugs, and bowls and Dutch delftware plates, ointment pots, dish covers and soup plates all testify to the Dutch presence. The Rhenish saltglazed Bellarmine jug and Westerwald jugs and mugs of the eighteenth century as well as the Chinese export pottery of the same period were also probably brought in by the Dutch who occupied the fort for a greater part of the eighteenth century. It is also likely that some of the items from England such as the creamware with “featheredged” and “royal” pattern designs and the cuprous badge with the Royal Coat of Arms of the United Kingdom of the period 1714 to 1801 were brought in by the British during their short possession of the fort from 1782 to 1785. It is unlikely that the Dutch would have carried a Coat of Arms of the United Kingdom to the fort.

Archaeological evidence suggests that even if the fort was occupied after its abandonment by the Dutch, it was occupied for a very brief period. Sherds of European pottery which continued to be made into the second quarter of the nineteenth century are few and those found were all picked from the surface and could have been brought in before their terminal date of production. They are also not likely to be in their primary contexts and the predominance of eighteenth century finds would seem to suggest that the fort was abandoned in the early nineteenth century. European countries often traded in items which originated from European countries other than their own. As Walker (1975: 184) has rightly noted with regards to pipes traded on the West African coast, “caution should be exercised ... in assuming that Dutch pipes found in West Africa invariably indicate Dutch trading contacts”. This is because in 1759, it was observed that English trading companies were allowed to go to the Netherlands to buy

Eighteenth century European and Chinese pottery as well as smoking pipes were found in the upper and lower levels of Pit 3, an indication that the archaeological finds accumulated within a short period and that the site is a single occupation site. This is supported by the fact that 57

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA They were also consumed by inmates of the forts and castles along the coast. The consumption of alcohol by Europeans can best be explained by considering the mental state of the European in a foreign country. This is aptly described by Selena Axelrod Winsnes (2004: 34):

fragments of the same Bellarmine jug were found in Levels 2, 5 and 6 of Pit 3. During the early seventeenth century, the commercial activities of the Dutch East India Company resulted in the flooding of European markets with good quality Oriental porcelain from China and Japan. Just how Chinese export porcelain began to appear on sites along the Guinea coast is uncertain. It is difficult to tell whether they were direct imports from the Orient through the Cape of Good Hope or re-exports from Europe. DeCorse (2001: 157) has noted that the majority of Chinese pottery in Elmina can be dated to the Dutch period and was probably brought in during the eighteenth century. As at Elmina, the Dutch were probably the importers of the Chinese porcelain found at Fort Amsterdam.

“Along with actual physical symptoms of illness we must consider the mental state of these Europeans. They were plagued by grinding homesickness, in this world so different from Europe. They felt forgotten, neglected because of lack of mail. The infrequency of... ships to Guinea made correspondence exceedingly difficult and although the ships of other nations carried mail to Europe, there would still be delays of forwarding letters from other countries. When the European was not himself suffering from a variety of complaints and illnesses, he was confronted on all sides by his compatriots in distress. Death was so common...The Europeans commiserated with one another, both as countrymen and as friends across national lines, sharing the same minatory fate. This group support invariably resolved itself in drink”.

The beautiful European and Chinese pottery testifies to the lavish life-style of the European inmates of the fort. As already indicated, it is likely that the imported ceramics were supplemented by locally–produced ones. Anquandah (2006: 13) has noted that at Fort Crèvecouer, an important Dutch trade fort in the period 1640-1867, “the overwhelming predominance of local pottery as compared to German, Dutch, Chinese and English imported ceramics in a purely European fort does suggest that the European inmates eventually became acculturated and adapted to use of indigenous ceramics because they were cheaper and more economical” and that “the same acculturation seems evident in subsistence economic lifestyle”.

While the European ceramics, liquor and smoking pipes may have been used by European inmates in the fort, it is likely that similar items were traded to or used by the local inhabitants. The probability that similar items as those found in the excavations were traded into the hinterland is attested by a fragment of a colourless balustre wine glass with air-twisted decoration (similar to the eighteenth century balustre wine glass from the tenth level of Pit 3) found among a collection of goldweights by Crossland (1973 Plate 16; 1989, Plate 16) and illustrated and cited by Garrard (1980: 178) and DeCorse (2001: 161) respectively. Smoking pipes were imported for use by enslaved Africans during their voyage to the New World (Walker 1975: 166; Svaleston 2000: 118, 130, 136). Enslaved Africans were also given alcohol to “lift their spirits”. It is likely that enslaved Africans and servants who worked for the Europeans in the fort were given rations of tobacco and alcohol. It was also a common practice for the governors of the forts and castles to entertain invited guests from nearby settlements especially during Christmas and on New Year’s Day. The numerous broken bottles testify to the consumption of alcohol and perhaps to the debauchery so much characteristic of life in the forts.

The Europeans at Fort Amsterdam may have also become acculturated. Apart from the use of locally-manufactured pottery, molluscs, an important food item and a delicacy to the local inhabitants were exploited for food by the inmates of the fort. At Fort Crèvecouer, the Dutch entered into agreement with the local Ga chiefs to supply fish on a weekly basis to the Fort (Anquandah 2006: 13). It made sense to subsist on domestic animals, meat and fish brought in by local hunters and fishermen. This is because European ships were reaching the coast at long intervals. Thus, needed provisions from Europe could not be relied on and Europeans had to adapt to African food if they were to survive. It is therefore not surprising that molluscs collected for food by the local inhabitants such as Ostrea denticulata, Thais callifera, Cardium costatum, Donax acutangulus, and Achatina are found in the archaeological record. The molluscs may have been sold to the European inmates by the local inhabitants. Cattle may also have been traded from the east coast of the Gold Coast to Fort Amsterdam. In the seventeenth century, Barbot (1732: 215) noted that sheep and cattle did not do well on the west coast but abounded on the Accra coast from where they were exported to the west coast of the Gold Coast.

The most common imported ceramics of the seventeenth and eighteenth centuries in Ghana are Rhenish stonewares. Rhenish stoneware jugs are particularly used in rituals in Ghana and are found in many stool and shrine rooms alongside traditional ritual pots. Their preponderance in burials at Elmina indicates that they were held in particular esteem (DeCorse 1989: 144; 2001: 157). Rhenish stoneware vessels have been found in southern Ghana in such places as Eguafo (DeCorse 2001: 157), Asebu (Nunoo 1957: 16), Ayawaso (Ozanne 1962: 58; Bredwa-Mensah 1990: 130) ,Ladoku (Ozanne 1965: 6) and Adjikpo-Yokunya (Nimako 2005: 108). Further north in the Techiman area, two pieces of blue and white Rhenish pottery were collected by Oliver Davies and

The thousands of broken bottles, some of which were containers of liquor consumed by the inmates of the fort, are to a certain extent a reflection of the life-style of the Europeans on the coast. Apart from being an important trade item, alcoholic beverages were also used as gifts to local chiefs to enhance friendship and to obtain favour. 58

J. BOACHIE-ANSAH: EXCAVATIONS AT FORT AMSTERDAM, ABANDZE, CENTRAL REGION, GHANA inmates adapted to local conditions and probably traded with the local inhabitants in such items as pottery, shellfish and meat.

Anquandah at Tanoboasi in 1964 (Anquandah 1965: 117), and Effah-Gyamfi (1974: 273, 1985: 202) found two pieces of Westerwald salt-glazed pottery of the second half of the eighteenth century at the nearby site of Bono Manso. In a recent excavation, Boachie-Ansah (2005: 73-74, 85-86) also found a fragment of a Rhenish salt-glazed, light gray stoneware jug painted in cobalt blue in an eighteenth century context at Ohene Ameyaw Anim, a site close to the Techiman Secondary School site. While some of the stoneware fragments such as those from Ayawaso and Asebu predate the eighteenth century and others such as those from Adjikpo-Yokunya post-date the eighteenth century, it is likely, indeed probable, that Rhenish pottery probably brought in by the Dutch during the eighteenth century found its way to trading centres in the hinterland.

William Bosman, writing in 1704, described Great Kormantin, a town of the local inhabitants as “large and populous that it very well deserves the name Great...” According to him, “all the inhabitants besides mercantile traders, are fishermen, amounting to the number of 700 or 800 and sometimes 1000” (Bosman 1967: 58). Bosman also maintained that anytime the local inhabitants disagreed with the Europeans, the latter were denied important supplies and provisions. The cessation of trade and interaction between Europeans and natives in times of disagreement perhaps explains the decline in trade described by Bosman (1967: 58) in 1704. However, there are indications that the economic depression described by Bosman was temporary and that the eighteenth century witnessed a great deal of trade and interaction between the Dutch and the hinterland as evidenced by typical vessel forms characteristic of the hinterland in coastal areas such as Elmina, and by European ceramics particularly Rhenish pottery in the hinterland. As already indicated, much of the Rhenish pottery is likely to have been brought in by the Dutch during the Dutch period at Elmina and Fort Amsterdam. It is hoped that the excavations just completed by E.K. Agorsah at Old (Great) Kormantin will throw light on the impact of European trade on the communities close to the fort.

The influence of Rhenish stoneware on Ghanaian societies can also be seen in the shape of a ritual pot known as mogyemogye (Rattray 1959: 304, Fig. 252), and used as a container for the wine poured on the golden stool of Ashanti. DeCorse (2001: 120) has rightly observed that this vessel bears a close resemblance to a Rhenish stoneware jug. Similar pots have been found in Elmina in ritual contexts dated to the eighteenth and nineteenth centuries, at the same period when typical Ashanti vessel forms also appear in archaeological contexts at Elmina (DeCorse 2001: 120-122). From every indication, the pot is a copy of a Rhenish jug and may be regarded as artefactual evidence of influence of European ceramics on the ceramic industry in Ghana. The intensive interaction between the hinterland and the coast during the eighteenth and nineteenth centuries suggests that trade items including European ceramics of types found in Elmina and Fort Amsterdam were traded into the hinterland.

Acknowledgements I am grateful to the Ghana Museums and Monuments Board for funding the excavations; to the late Nana Kwasi Attah IV, Chief of Abandze and Nifahene of the Mankesim Traditional Area and his sub-chiefs for their immense assistance; to Mr. N. Ivor and Dr. P. Y. Dzamefe for their encouragement and logistic support; to Messrs. Iddrisu Abass, Prince Buertey Larweh and Albert Martey, students of the Department of Archaeology who assisted in the excavations, and to Mr. Cosmos Logosu for driving us to the site and for his participation in the excavations. Sincere and special thanks go to Mr. L. B. Crossland, and Dr. Jerzy Gawronski for the detailed comments on the smoking pipes and European and Chinese ceramics; to Mr. B. M. Murey for the faunal analysis; to Mr. Aaron Eshun and my son, Isaac, for drawing the maps and the pottery respectively, and to Joyce Dartey, Getrude A. Mansah, and Aba AttaQuayson for typing the manuscript.

Some of the smoking pipes found at Fort Amsterdam are similar to those excavated at Bantama by David Calvocoressi (1977: 136-139). These include pipes with crowned L and “clover leaf” marks. That Dutch pipes similar to those from Fort Amsterdam were traded into the hinterland is perhaps supported by the fact that some of the pipes excavated by Nunoo (1957) at Asebu have been identified by Walker (1975: 184) as Dutch pipes ranging in date from the seventeenth to the nineteenth centuries. Conclusions The excavations at Fort Amsterdam have uncovered European and Chinese goods, a predominant number of which are dated to the eighteenth century. Pit 3 which produced most of the finds, including thousands of broken bottles, appears to have been a dumping ground for the inmates of the fort during the eighteenth century. The fact that similar artefacts dated to the same period were found in the lower and upper levels suggests that the archaeological debris accumulated within a short period. The use of locally- manufactured pottery as well as the reliance on molluscs from shallow waters, mangroves, estuaries and lagoons clearly indicates that the European

References ANGUS-BUTTERWORTH, L. M. 1970. Pottery and Porcelain. Glasgow: Collins. ANQUANDAH, J. 1965. An Archaeological Survey of the Techiman-Wenchi Area. (In), Goody, J. and Arhin, K. (eds.), Ashanti and the North-West, a special issue of 59

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA DECORSE, C. R. 1989. An Archaeological Study of Elmina, Ghana: Trade and Culture Change on the Gold Coast between the Fifteenth and the Nineteenth Centuries. Unpublished Ph.D Thesis, Department of Anthropology, University of California, Los Angeles.

Research Review, Supplement No. 1. Legon: Institute of African Studies, pp. 111-134. ANQUANDAH, J. 2006. The Accra Plains c. 1400-1800: An Overview of Trade, Politics and Culture from the Perspective of Historical Archaeology. Accra Before Colonial Times: Proceedings of a Colloquium on Early Accra. Research Review, Supplement 17. Legon: Institute of African Studies, pp. 1-20.

DECORSE, C. R. 1998. Artifact Typologies for Coastal Ghana. Unpublished Manuscript. DECORSE, C. R. 2001. An Archaeology of Elmina. Washington: Smithsonian Institute Press.

ATKINSON, D. R. and OSWALD, A. 1972. A Brief Guide for the Identification of Dutch Tobacco Pipes found in England. Post-Medieval Archaeology 6: 175182.

DONNAN, E. 1931. Documents Illustrative of the Slave Trade to America. Cited in Walker I. C. 1975. The Potential Use of European Clay Tobacco Pipes in West African Archaeological Research. West African Journal of Archaeology 5: 165-193.

ATTA-YAWSON, P. n. d. The History of Fort Amsterdam. Amsterdam: The Battlefields Trust. BARBOT, J. 1732. A Description of the Coasts of North and South-Guinea and of Ethiopia Inferior Vulgarly Angola: Being a New and Accurate Account of the Western Maritime Countries of Africa. Paris: Didot.

DRAPER, J. 2001. Post-Medieval Pottery 1650-1800. Buckingham: Shire. DUCO, D. H. 2003. Merken en Merkenrecht van de Pijpenmakers in Gouda. Amsterdam: Pijpenkabinet.

BOACHIE-ANSAH, J. 2005. Excavations at Techiman, Brong-Ahafo Region, Ghana. Ghana Studies 8: 39-101.

EDMUNDS, J. 1978. Sea Shells and other Molluscs Found on West African Shores and Estuaries. Accra: Ghana Universities Press.

BOAHEN, A.A. 1966. Topics in West African History. London: Longmans.

EFFAH-GYAMFI, E. 1974. Oral Tradition and Archaeology: A Case Study of the Bono State. Unpublished M.A. Thesis, Department of Archaeology, University of Ghana.

BOSMAN, W. 1967 (1705). A New and Accurate Description of the Coast of Guinea. London: Frank Cass. BREDWA-MENSAH, Y. 1990. An Archaeological Investigation Conducted at Okai Koi Hill (Ayawaso) and it’s Significance for Iron Age Archaeology in Ghana. Unpublished M.Phil Thesis, Department of Archaeology, University of Ghana.

EFFAH-GYAMFI, K. 1985. (Previously Effah-Gyamfi, E.). An Archaeological Investigation into Early Akan Urbanism. Calgary: University of Calgary Press. FLETCHER, E. 1975. International Bottle Guide. Dorset: Blandford Press.

CALVOCORESSI, D. 1977. Excavations at Bantama, near Elmina, Ghana. West African Journal of Archaeology 7: 117-141.

FRANCIS, P. JNR. 1994. Beads of the World. Atglen: Schiffer Publishing Ltd.

CAPELLA, R., VENTURA, F. E., BELLADINELI, A., GRASSI, E. and FIORETTI, P. 2003. Restoration and Activation of Fort Amsterdam. Perugia: Ghana Museums and Monuments Board and Scuola Edile Di Perugia.

GAIMSTER, D. 1997. German Stoneware 1200-1900. London: British Museum Press. GARRARD, T.F. 1980. Akan Weights and the Gold Trade. New York: Longman.

CLARIDGE, W. W. 1964. A History of the Gold Coast and Ashanti. Vol.1. London: Frank Cass.

GREEN, R. 1978. Bottle Collecting. Staffordshire: John Mackie.

CROSSLAND, L.B. 1973. A Study of Begho Pottery in the Light of Excavations Conducted at the Begho-B2 Site. Unpublished M.A. Thesis, Department of Archaeology, University of Ghana.

HODASI, J. K. M. 1995. Snails in the National Economy. Legon: School of Communications Studies Press. HUME, I. N. 1969. A Guide to Artifacts of Colonial America. New York: Alfred A. Knopf.

CROSSLAND, L.B. 1989. Pottery from the Begho B-2 Site, Ghana. Calgary: University of Calgary Press.

HUME, I. N. 1970. Early English Delftware from London and Virginia. Colonial Williamburg Occasional Papers in Archaeology 3.

DAVIES, D. 1972. English Bottles and Decanters 16501900. London: Charles Letts. 60

J. BOACHIE-ANSAH: EXCAVATIONS AT FORT AMSTERDAM, ABANDZE, CENTRAL REGION, GHANA HUME, I. N. 1982. Martin’s Hundred: The Discovery of a Lost Colonial Virginia Settlement. New York: Dell Publishing.

SOUTH, S. 1977. Method and Theory in Historical Archaeology. New York: Academic Press. SVALESTON, L. 2000. The Slave Ship Fredensborg. Accra: Sub-Saharan Publishers.

KANKPEYENG, B. W. and DECORSE, C. R. 2004. Ghana’s Vanishing Past: Development, Antiquities and the Destruction of the Archaeological Record. African Archaeological Review 2: 89-127.

VAN DANTZIG, A. 1980. Forts and Castles of Ghana. Accra: Sedco Publishing Ltd.

KOPTAL, R. L. 1981. Mollusca: A Text Book for College and University Students. Meerut: Rastogi Publications.

VAN DEN BOSSCHE, W. 2001. Antique Glass Bottles 1500 -1850. Woodbridge: Antique Collectors Club.

LAW, R. 1980. The Horse in West African History. London: Oxford University Press.

WALKER, I. C. 1975. The Potential Use of Clay Tobacco Pipes in West African Archaeological Research. West African Journal of Archaeology 5: 165-193.

LAWRENCE, A. W. 1963. Trade Castles and Forts of West Africa. London: Jonathan Cape.

WARD, W. E. F. 1958. A History of Ghana. London: George Allen and Unwin.

LEWIS, G. 1999. Collectors History of English Pottery. Woodbridge: Antique Collectors Club.

WINSNES, S. A. 2004. A Danish Jew in West Africa. Trondheim: Department of History, Norwegian University of Science and Technology.

MCNULTY, R. H. 1971. Common Bottles: Their Production, Use and Forms in the Seventeenth and Eighteenth Century Netherlands. Journal of Glass Studies: 13: 91-148. MILLER, J. J. and STONE, L. 1970. Ceramics from Fort Michilmackinac. Washington: Smithsonian Institution Press. NIMAKO, E. 2005. Historical Archaeological Investigation at Mount Mary Training College and Adjikpo-Yokunya. Unpublished M.Phil. Thesis, Department of Archaeology, University of Ghana. NORREGARD, G. 1966. Danish Settlements in West Africa. Massachusetts: Boston University Press. NUNOO, R. B. 1957. Excavations at Asebu in the Gold Coast. Journal of the West African Science Association 3: 12-44. OSWALD, A. 1975. Clay Pipes for the Archaeologist. British Archaeological Reports 14. London: British Archaeological Reports. OZANNE, P. 1962. Notes on Early Historic Archaeology of Accra. Transactions of the Historical Society of Ghana 6: 51-70. OZANNE, P. 1965. Ladoku: an Early Town near Prampram. Ghana Notes and Queries 7: 6-7. OZANNE, P. n. d. Tobacco Pipes of Accra and Shai. Legon: Institute of African Studies (Mimeographed). RATTRAY, R. S. 1959. Religion and Art in Ashanti. London: Oxford University Press. SCHLEDORN, L.A., ELIENS, T.M. and VAN AKENFEHMERS, S. (Eds.). 2003. Delftware: History of a National Product Vol. 3. Delft: Gemeente Musea. 61

Researching the Internal African Diaspora in Ghana Kodzo Gavua Department of Archaeology, University of Ghana, Legon, P.O. Box LG3, Legon, Ghana

• Some Agortime people of the Volta Region, who emigrated from the region of the Dangbe (Armah 1994).

Introduction In September 2006, this author initiated a project to research the influence of the Trans-Atlantic Trade in enslaved people on the society and culture of some Ewe communities of the Volta Region of Ghana (Figure 1). The project was supported financially by the French Embassy in Accra. This paper presents a summary of aspects of the work that pertain to a Diaspora of peoples of these communities, which were created in Ghana as a result of activities that were related to the trade. Since the preliminary work in these communities this research has been expanded to cover other African Diaspora in Ghana but discussion will be conveniently limited to the Ewe Diaspora researched so far.

• The Santrokofi and some Twi-speaking peoples of the Volta Region, who emigrated from Bono-Ahafo and Asante (Cobblah 1994). The Internal African Diaspora in Ghana may thus be manifold. However, there is a paucity of information on this Diaspora, and public awareness of its existence is limited. Relevant research has been centred mainly on the African Diaspora in the Americas and Europe, and on forts, castles, and routes that were associated with the trade (Perbi 2004; van Dantzig 1980; Anquandah 1999, 2007).

The term ‘Diaspora’ is derived from the Greek word, diasperein, which means ‘to scatter’, and generally connotes the dispersion or scattering of a people with common origin and culture (Agnes 1999). It also refers to places where dispersed people settle. In the particular case of Africa, the term may refer to peoples of African origin, who relocate to other continents, and to continents and regions to which dispersed Africans have relocated. Diaspora is often used in a strict sense, nonetheless, to distinguish Africans, who have been dispersed and relocated to other continents and regions of the world as a result of the trade in enslaved people. Hence, reference is made, for example, to the African Diaspora in the Americas, Europe, and the Middle East. The term ‘Internal African Diaspora (I.A.D.)’ is used here, therefore, in relation to Africans, whose forebears were relocated from their homelands to other regions of the continent during the period of the Trans-Atlantic trade in enslaved people.

In the sections that follow, the approach to researching the Ewe Diaspora and what the research has revealed, so far to this author, about the character and dynamics of this Diaspora will be discussed. Various strategies by which peoples of the Diaspora have negotiated new identities, while retaining and perpetuating their ancestral memory are highlighted. Though preliminary, the results of the work should add to the discourse on social formation processes and ethnicity in Africa (Ambler 1985; Kumar 2003). Approaches to the Research This research on the Internal African Diaspora in Ghana has been predicated on the view that in spite of distinctions that are made between them, each language (ethnic) group of Ghanaians comprises a mix of peoples of different geographical and ancestral backgrounds. Migrations and relocations of peoples within the geopolitical area of Ghana particularly during the period of the Trans-Atlantic Trade might have resulted in interactions between, and integration of different groups of people. But the character of the society and culture of each group as found today emanates from transformations that have been influenced by innovations and inventions within the group, and from the appropriation and reappropriation of foreign traits to which members of the group are exposed.

The relocation of the forebears of several groups of Ghanaians, during the period of the trade, from their ancestral homelands to regions where they are found today has been documented. Examples of such groups include the following: • Peoples of northern Ghanaian origin, who are found in the Asante Region of Ghana (Wilks 1961) • People of Ahanta origin in the Western Region of Ghana, who relocated to the Volta Region and are now known as the Avatime (Konu 1981)

The society and culture of each language group would be characterized, therefore, by some indigenous traits, a fusion of indigenous and foreign traits, and foreign traits that have been incorporated into prevailing domains of behaviour, or substituted for indigenous traits. Whatever the case may be, the transformation and development of the group’s culture would entail negotiations of behaviour

• People of Denkyira origin, who relocated to Vume in the Volta Region (Danyo 2006), and to Krobo (Wilson 1995). 63

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

Figure 1. Towns of the Ewe Diaspora in Ghana

and identity by the different members of the group (Appadurai 1996) in relation to their quest to survive.

October 2006 to Abesim in the Bono-Ahafo Region and to Nsuta-Mampong, Juaso and Obogu in the Asante Region based on information obtained at Wusuta on the existence of a Diaspora of Wusuta people in those towns. In addition to field notes, use was also made of photography and video in the process of field observation.

Fieldwork Fieldwork was begun on September 2, 2006, in Wusuta, a town of about 6,000 people in the Kpando District of the Volta Region (Figure 1). This work was extended in 64

KODZO GAVUA: RESEARCHING THE INTERNAL AFRICAN DIASPORA IN GHANA Wa 2 respectively. The pit at Wa 1 (W/W-1) was 60cm deep while that at Wa 2 (W/W-2) was only 40cm deep. At Gadze, two trenches were dug at the site of an ancient shrine in the home of the late Togbe Kweku Aba where several potsherds and faunal remains were exposed on the surface. The trenches, measuring 2m x 4m each, were only 20cm deep. A 4m x 6m trench was also dug at Torkpeta on the compound of the Roman Catholic Primary School to 40cm below the datum line.

Wusuta A preliminary survey this author conducted in 1980 of the prehistory of Wusuta (Gavua 1980) indicated that the town was severely depopulated after 1869 when many of its people were taken into bondage in Asante by warriors, who had used the town as a military base to unsuccessfully invade part of the territory that constitutes the Volta Region today. The survey also suggested that the town is located on an ancient route on the eastern corridor of Ghana, which linked northern and southern communities of the country. It once had a market where a variety of goods as well as enslaved people were traded. Guided by this information, an archaeological survey was conducted and ethnographic studies undertaken to gather data in the town and its environs. Oral accounts pertinent to the research were also obtained through semi-formal and informal interviews as well as focus group discussions.

Following the excavations, the elders of some families, who claimed their forebears were forcibly ejected from the town and relocated to Asante and other regions of Ghana, were interviewed, while group discussions were held with others. Family names and relevant folklore were also recorded and studied. A list of towns in the Asante, Eastern and Bono-Ahafo Regions, which informants associated with the Wusuta Diaspora, was compiled. The names of some peoples of the Diaspora were also recorded.

Information was sought specifically on the following: The Wusuta Diaspora

• The material culture of early settlements of the people prior to 1869.

With the information received at Wusuta about places where the Wusuta Diaspora could be found, an exercise was embarked upon to locate the towns and families of the Diaspora. It was only possible to visit Abesim, Nsuta-Mampong, Juaso and Obogu, although the list of towns referred to by informants was extensive. With the assistance of informants, individual persons and families, who identified themselves as part of the Ewe Diaspora in Akan territory, were found. Three major families each at Abesim and at Nsuta respectively claimed Wusuta ancestry. At Juaso, one relatively large family identified with Wusuta, while another identified with Botoku, a town that is found about six kilometers south of Wusuta. A large group of people, who claimed Avetile-Peki ancestry, was found at Obogu. All the families located were royals.

• Slave-trading activities and their effects in the Wusuta area. • The identity of individual persons and families, whose ancestors were directly or indirectly associated with, or affected by the trade. • The location and possible identity of peoples of Wusuta and other Ewe descent, who form part of the Diaspora. • Circumstances that influenced the creation of the Diaspora. • Relationships that exist presently between the people of Wusuta and the Wusuta Diaspora.

Through informal and semi-formal interviews of individual persons, discussions during family gatherings, as well as direct participation in funeral ceremonies and clan rituals, oral accounts were obtained about settlement and family histories, traditions, and the experiences of individuals and groups of the Diaspora. Information was also sought on the processes by which the Ewe Diaspora is integrated into the society and culture of their present homelands, and on any material and other behavioural correlates that may be found between the Diaspora and the peoples of their ‘ancestral homelands’. The information was obtained from not only members of the Diaspora community, but also other persons, including chiefs and school children, who were not part of the Ewe Diaspora.

• Ways in which the relocation to and integration of the Wusuta Diaspora into their ‘new’ communities have influenced the society and culture of the town. In order to gain insights into the early material culture of the town, five sites (Figure 2), located where informants claimed to have been ancient settlements of various clans of the town, were excavated. These sites included Adosime, ‘an ancient slave market’ that was found at the southernmost end of the town’s old settlement area. The rest of the sites, Wa 1, Wa 2, Gadze and Torkpeta, named after clans in whose territory they were found, were within the Wusuta settlement.

After the preliminary work in these towns in October 2006, the author was joined by three elders of Wusuta to revisit the families located and interviewed in order to strengthen links with the families and to seek additional information. The follow-up visit in December 2006 offered an opportunity to partake of various family

Two 6m x 2m trenches were dug at Adosime. Work on the first was stopped 40cm below the datum line where a sandstone-bedrock was encountered, while the second was dug up to 60cm below the datum. Two test pits, measuring 4m x 4m, and 2m x 2m were dug at Wa 1 and 65

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

Figure 2. Sites Excavated at Wusuta

the town to trace their ‘roots’. According to the spokesperson, the group represented a large family of about 1,400 people, the descendants of a woman, who was kidnapped as a young girl from Wusuta in about 1825 and brought to Accra. The group claimed their forebears had re-connected with their relations at Wusuta in the 1930s and provided some names of a few of their possible relatives. They also narrated oral histories that their forebear had passed onto them about their ancestry. With the leads they provided, and in liaison with some elders of Wusuta an intensive investigation was conducted in the community over three months. Eventually, we located a 75-year old woman, who narrated an oral history that corroborated the story of the

reunions and rituals that were organized by the families. The author was, for example, invited to observe and document on video a ritual in the stool room of the chief of Ntomase, the clan of the Wusuta Diaspora at Nsuta. Telephone conversations were also carried out with some members of the families, who were resident in the United Kingdom, while other members, who were resident outside the towns visited but had learnt of the research, visited the author at the University of Ghana to register their names and to tell their story. During the course of the research, a group of five people from the Gă community of Gbese in Accra, were encountered at Wusuta in January 2007. They had visited 66

KODZO GAVUA: RESEARCHING THE INTERNAL AFRICAN DIASPORA IN GHANA disturbance of the archaeological record by natural and cultural agents.

group and claimed to be a relative of the group. The group has since April 2007 reunited with the family of the old lady and has even begun arrangements to build in Wusuta.

Since there is a paucity of information on pottery from archaeological contexts in the Volta Region of Ghana, a broad overview of the pottery that was found in the Wusuta excavations will be briefly presented. A total of 131 everted rim sherds (e.g. Figure 3c), 88 inverted rim sherds (e.g. Figure 3d) and 38 straight rim sherds were recovered. Some of the base sherds found appear to be similar to those recovered from the Krobo Mountain and the Accra Plains in general (Figure 3e). Surface treatment of the pottery ranged from stamped impressions that are akin to the Kintampo cultural tradition (Figure 3f), rouletted impressions, channeled decorations (Figure 3g) and grooved impressions (Figure 3h). Some of the sherds were burnished and others were smudged. As indicated in Table 1, grooving and burnishing appear to have been the most common techniques of treating the surfaces of pottery.

Preliminary Research Results The research is ongoing and analyses of the data obtained to date have not been completed. The findings presented here are thus provisional. i) Archaeological Data The archaeological survey in the Wusuta area yielded a total of 1679 potsherds, 584 beads (Figures 3a and b), 108 cowry shells, 36 fragments of glass, 28 pieces of metal, and 21 faunal remains, including shell. These finds were retrieved in contexts that were not stratified and were probably mixed. No radiocarbon dates have, however, been obtained and there is the likelihood of Decorated Sherds Stamped 26

Rouletted 13

Burnished 148

Total Decorated

Undecorated Smudged 37

Grooved 311

Red-slipped 25 1119

560 Table 1. Pottery decoration

a

b

d

c

e

g

f

h Figure 3. Samples of excavated finds

67

Total

1679

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

REGION

ASANTE

TOWN

MampongNsuta Appiakrom Ejisu Konongo Dubi Juaso Obogu Santase Akumadan

BONOAHAFO Abesim Atebubu

EASTERN

G-ACCRA

CENTRAL

Kwahu (Obo, Nkwatia, Abetifi) Akyem-Oda

Gbese-Accra

BremanAsikumah

Table 2. Regions and towns identified by informants with the Ewe Diaspora

Wusuta men were lured alongside women into Asante following the Asante-Ewe war of 1869. The men followed their women, who were taken into marriage by chiefs and other royals at Mampong-Asante. They revolted after being maltreated in various ways, including some of them being sacrificed to Asante deities. They plundered farms and livestock in the Mampong area to the dismay of the local Asante. As the Mampong royals could not stop the rebels, they only urged the local Asante to peacefully coexist with the rebellious Ewe men. This led to a Twi saying in the Mampong area that “se Wusuta (Nsuta) fuo didi mea, na Mampong fuo hu ato om”, meaning, ‘Mampong people would only be free when the Nsuta people are satisfied’. It appears that the name Nsuta in the Mampong area is a Twi corruption of Wusuta, although there are several other towns know as Nsuta in the Asante region. Nsuta generally means “headwaters” of streams and rivers.

ii) Data on the Ewe Diaspora Data on the Ewe Diaspora comprises a list of towns in which the Diaspora could be found (Table 2), narratives of family histories, oral accounts (histories and traditions) pertaining to the creation and experiences of the Diaspora, and registers of family names. The data strongly points to the existence of a fairly large Diaspora of Ewe people in Akan and other territories of Ghana. According to informants, a number of the Ewe people were recruited as porters and sent to Asante by Akan merchants, while several of them were captured and lured by Asante soldiers, who presented them to the king of the Asante, Kofi Karikari, as prisoners of war. Many of the people interviewed reported that their greatgrandmothers were kidnapped in Wusuta and other Ewe communities such as Avetile-Peki, brought to Asante and coerced into marriage within royal families. Because many of the women who were kidnapped were married into royal families, their off-spring have become royals and wield positions of power in the Diaspora. But mindful of their shaky positions in a foreign society, the Diaspora worked hard to protect property they inherited, and amassed much wealth. Many established relatively large families as a defensive mechanism and as a means of increasing family labour and human resources.

Most of the members of the Diaspora found have never visited the Volta Region and do not know the towns from which their ancestors came. A few have, however, visited Wusuta in the past in search of their relations, while some Wusuta people have also visited Juaso and Obogu to establish links with their relations. Unpublished oral accounts compiled by the late Rev. E.O. Kofi Gavua identify a number of Wusuta families, whose forebears were slave dealers. Enslaved people, whose freedom was bought, were also identified. According to the accounts, local merchants bought and incorporated some enslaved people into their families. A female descendant of a freed slave was, for example, noted to have been instrumental in empowering the women of Wusuta to wear shoes for the first time after gaining much wealth and power. The records also claim that the forebears of many of today’s Wusuta people returned to the town from the Diaspora after escaping their captors. This claim was supported by the field research data. A number of people, who are known to have escaped and relocated to Wusuta from Asante were identified and named by informants. Since the beginning of the research, there have been a number of reunions between some Wusuta families and peoples from Kwahu, who were able to trace their roots to Wusuta. A similar reunion also occurred in 2007 in the town of Sovie, which is about 10km north of Wusuta.

In Abesim for example, the story was told of a man, who in order to maintain his Ewe heritage, worked very hard to amass wealth, married several women and produced about 36 children. He ensured that each of his children became aware and conscious of their Ewe background and identity, in spite of their royal status. In his old age, the man caused a grave to be dug for him in a room of his house and fitted the doors with iron bars. He bought his own coffin and instructed his children to bury him in the grave so he would continue to live with ‘his Ewe people’ even in death. Although the Akan generally are matrilineal, the family of this man has remained patrilineal. Nonetheless, informants were generally silent over the relocation of men from the Ewe territory into Asante expect for those of Mampong-Nsuta. According to a narrative recorded at Mampong-Nsuta, a large number of 68

KODZO GAVUA: RESEARCHING THE INTERNAL AFRICAN DIASPORA IN GHANA Nineteenth Century. International Journal of African Historical Studies 18: 201-222

Conclusions It has not yet been possible to establish any relationship between the material culture of the Ewe community research and the Ewe Diaspora and this might not occur until additional archaeological data is obtained in the towns in which the Diaspora is found. Other information presented, suggests, however, that a combination of factors, including interactions engendered by commerce and warfare between the Akan and the Ewe during the period of the trade in enslaved people were major factors that culminated in the creation of the Ewe Diaspora. Among the processes by which the Diaspora was created were physical coercion and kidnapping. While young men, who were captured, were either sold, executed, or sacrificed to deities, the young women were forced into marriage.

ANQUANDAH, J. K. 2007. The Trans-Atlantic Slave Trade: Landmarks, Legacies, Expectations. Accra: SubSaharan publishers. ANQUANDAH, J.K. 1999. Castles and Forts of Ghana. Paris: Atalante. APPADURAI, A. 2003. Modernity at Large: Cultural Dimensions of Globalization. Minneapolis: University of Minnesota Press. ARMAH, J.K. 1994. The Technology of Traditional Textile Weaving in the Agortime-Kpetoe Area. Unpublished B.A. Long-Essay, Department of Archaeology, University of Ghana.

Marriage into royal families by the women transformed the identity of their offspring. The Diaspora adopted the language and culture of their captors in order to retain and maintain property and status they acquired through inheritance. The new identities have been maintained by the descendants of the Diaspora, who returned to their ancestral homes in the form of names and language. Many still have Akan names and combine Twi with Ewe in speech.

COBBLAH, M. 1994. An Ethnographic Study of Subsistence Patterns of Santrokofi traditional Area – Its Archaeological Implications. Unpublished B.A. LongEssay, Department of Archaeology, university of Ghana. DANYO N. 2006. A Study of Early Settlement History of Vume Gbogame. Unpublished MPhil Thesis, University of Ghana, Legon.

The preliminary findings of the research have to some extent affirmed the conjecture that ethnic group identity is not fixed but fluid (see Franklin and Fesler 1999; Jones 1997). It changes under various circumstances and is influenced by complex variables. In the particular case of the Ewe Diaspora, the variables include physical, social and psychological coercion. Identity may be negotiated as a survival strategy, although people may be mindful of their origins as shown by the royals among the Ewe Diaspora many of whom continue to maintain memories and other cognitive structures that link them to their previous identity, although that have adopted a completely new language, material culture and social relationships. Acquired identity may also be maintained by individuals because of shame and fear of social reprisals reinforced by stereotypes, and specific social relations. This appears to be the case of the Ewe Diaspora, who would not identify with their origins.

FRANKLIN, M. and FESLER, G. (eds.). 1999. Historical Archaeology, Identity Formation, and the Interpretation of Ethnicity. Williamsburg VA: Colonial Williamsburg Research Publications. GAVUA, K. 1980. A Survey of the Prehistory of Wusuta. Unpublished B.A. Long-Essay, University of Ghana. JONES, S. 1997. The Archaeology of Ethnicity: Constructing Identities in the Past and Present. London: Routledge KONU, J. A. 1981. An Archaeological Survey of Amedzofe-Avatime. Unpublished B.A. Long-Essay, Department of Archaeology, University of Ghana. KUMAR, A. 2003. Community and Identities: Contemporary Discourses on Culture and Politics in India. (In), Jodhka, S.S. (ed.), Pacific Affairs 76. London: Sage Publications.

The work has shown, above all, that although people may speak different languages, have different material culture and social systems, they may belong to a common ancestry. It is thus nonsensical for Ghanaians and other people to discriminate on so-called tribal or ethnic lines, as tribalism is a fluke, a creation of people whom it benefits materially and socially.

PERBI, A. A. 2004. A History of Indigenous Slavery in Ghana: From the 15th to the 19th Century. Accra: Sub-Saharan Publishers. VAN DANTZIG, A. 1980. Forts and Castles of Ghana. Accra: Sedco Publishing

References WILKS, I. 1961. The Northern Factor in Ashanti History. Legon: Institute of African Studies.

ABNES, M. (ed.). 1999. Websters New World College Dictionary, 4th edition. New York: Macmillan. AMBLER, C.H. 1985. Population Movement, Social Formation and Exchange: Central Kenya in the

WILSON, L.E. 1995. The Krobo People of Ghana to 1892: A Political and Social History. Monograph in 69

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA International Studies, Africa Series, No. 58. Athens: Ohio University Press.

70

Current Archaeological Research at the Krobo Mountain Site, Ghana William Narteh Gblerkpor Department of Archaeology, University of Ghana, P. O. Box LG 3, Legon, Ghana, email: [email protected]

functions of the stone terraces that characterize the site. It also evaluates the socio-economic, religious, and political developments of the people on the mountain. The main concerns of the current study are the developments that occurred during the last 100 years of the mountaintop settlement which was abandoned in 1892.

Introduction This paper presents preliminary results of an archaeological survey of the south-western section of the Krobo Mountain conducted during 2004 to 2006. The study is the first major attempt at providing archaeological insights into the cultural developments of the Krobo, especially during the terminal point of settlement and cultural development on the Krobo Mountain. Various elements have been examined including indigenous religious practices, burial systems, architectural techniques, economic activities, and political institutions. Historical accounts about the primary function of the site and the circumstances of its abandonment have also been evaluated. Hence the study combines archaeological, ethnographic, and written and oral historical data.

Previous Archaeological Surveys Prior to the current investigation an archaeological team from the Ghana Museums and Monuments Board (GMMB) conducted reconnaissance surveys on the mountain in 1976 and 1982. These expeditions sought to assess the potential of the site for further archaeological research. The team concluded that the site was well preserved, and worth detailed investigation. In a one-page report entitled “Archaeological Survey of the Krobo Mountain Ancient Settlement”, the local pottery and architectural landscape were given prominence as future research objectives (Debrah 1982).

The contribution of archaeology to our knowledge of the historic environment of the Krobo area has been minimal. So far, none of the research publications about the Krobo, examines the archaeological evidence at the Krobo Mountain site. Since 2004, this author has been coordinating archaeological investigation on the Krobo Mountain to provide archaeological information to augment the existing data on the mountain settlement. The preliminary result of the initial fieldwork is found in an M.Phil thesis, An Archaeological Investigation of the Krobo Mountain Dry-stone Terraces (Gblerkpor 2005). The thesis considers the construction techniques and

Site Location and Description. The Krobo Mountain (0º 05'E- 6º 20'N) is located about 70km north-east of Accra, along the Tema-Akosombo highway. It is located about 2km south-east of Okwenya, and about 4km east of Somanya in the Eastern Region of Ghana (Figures 1 and 2). The mountain is separated by a dry-valley into two peaks namely the south-western (Yilo Krobo) and north-eastern (Manya Krobo) sections. It is a

Figure 1. Krobo Mountain. An aerial view from the north-eastern side (Photo. W. N. Gblerkpor)

71

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

Figure 2. A map of the study area showing sites mentioned

Kwa language family (Kropp-Dakubu 1982: 246). They are found in the southeastern part of Ghana. According to one-version of Krobo traditions, their ancestors migrated from a place called L‫כ‬l‫כ‬vor (believed to have been located in the east Accra Plains) to settle on the Krobo Mountain, after the Dangme had separated into several groups (Wilson 1987: 474). Security reasons have been cited for the selection of the rugged but defensive mountain as against the more habitable flat land that surrounds it. The mountaintop settlement was so defensible that since its establishment in about the 14th century, its inhabitants were only defeated in the 1850s after more precise rifles and rockets were developed; and was first occupied by enemy forces in 1892 (Wilson 1995: 7-8; 15-17).

rocky steep-sided terrain with pockets of gentle slopes to flatter portions. The vegetation is grassy, interspersed with mainly big cotton and baobab trees. It also has many caves, rock shelters, sacred groves and water cisterns. The south-western section measures about 1 x 2km and stands at a height of 350m above sea level. Currently, the site faces threats from Fulani herdsmen who graze their animals on top of the mountain, thereby destroying fragile artifacts such as pots and glass bottles whilst the annual burning of the vegetation cover by these herdsmen constitutes an even greater threat to the archaeological evidence for the fire destroys some of the archaeological record and the bare terrain exposes artifacts to treasure hunters. The erosion of the top soil by runoff water from the torrential rains that follow the bush burning dislodges the buried archaeological remains, thereby disturbing their original context (Gblerkpor 2005; 2006). To prevent further destruction of the site, the Yilo Krobo Traditional Council in collaboration with the Department of Archaeology, University of Ghana, Legon, and the Hills Resort Ltd. is addressing the situation through education and by providing security on the site.

In July 1892, the British colonial authorities ejected the Krobo from their mountain settlement for alleged acts of human rights violations, especially ritual murders. Scholars who disagree with this assertion have instead cited economic and political reasons for the ejection (Huber 1993: 36, Omenyo 2001:15, Steegstra: 2005: 31). After the sack of the settlement, the population moved to live permanently on their plantation villages located in the plains, and at the foot of the Akuapem-Akwamu Hills. These locations had served as settlements where the youth, farmers, and traders, periodically stayed to work on their farms and to carry out economic activities such as the exchange of commodities. Somanya, Manyakpogunor and Odumase are among the major successor towns that emerged from these villages after the abandonment of the mountain settlement.

Historical Background Until 1892, the Krobo Mountain was the major political and ritual centre of the Krobo (Steegstra 2005: 29). There were two groups with two separate political units on the mountain; Manya Krobo and Yilo Krobo (Debrah 1982: 1; Wilson 1987: 273). The Krobo are a group of Dangmespeaking people whose language belongs to the Western 72

WILLIAM NARTEH GBLERKPOR: CURRENT ARCHAEOLOGICAL RESEARCH AT THE KROBO MOUNTAIN SITE, GHANA

Figure 3. Local oil-lamps on display at the Kloyosiplemi Festival, 2006 (Photo. W. N. Gblerkpor)

the Yilo Krobo Traditional Council decided in 1992 to launch the Kloyosikplemi festival, a cultural concert, which initially sought to commemorate the centenary of the ejection, and to showcase Krobo traditions and cultural heritage. It is now an annual event and during the celebrations, ancient costumes, food and musical performances are displayed. Indigenous religious rites such as pouring of libations and animal sacrifices are performed. Traditional performance of this nature and their accompanying narrations proved useful for the study. Using interviews, oral traditions were collected from priests, priestesses, royals, teachers, elderly men and women, and others conversant with the oral traditions of the area.

Methodology The research has adopted a comparative study of the history, ethnography, ecology and archaeology of the site. Oral tradition was collected in addition to the study of written historical records. The archaeological research involved surface survey and collection as well as limited excavations. Data from ethnography, and written and oral history provided guidelines for the archaeological survey and excavations. Information from the nonarchaeological sources influenced the selection of specific areas for survey and excavation. The information was also useful in the identification, interpretation, and explanation of the archaeological data. For instance, the ethnographic analogues provided an interpretative framework for the elaboration of the indigenous religious and water management practices on the mountain.

A number of visits to some of the present-day Krobo settlements such as Sra, Somanya and Odumase enabled the collection of information about socio-economic and other cultural practices of the people. Particular attention was paid to local shrines, their structures, associated artifacts and functions. The traditional architecture was also examined. The building materials, structural designs and functions of historic buildings dating to the first decades after the lowland settlements were developed were studied. Additional information on contemporary cultural practices was collected during socio-cultural events such as the Kloyosikplemi festival celebrations of 2003 to 2006. The functions of some contemporary artifacts identical to artifacts recovered from the mountain were observed at first-hand during the traditional performances at the festival. For instance, the making and use of local oil-lamps was demonstrated to the amazement of participants at the 2006 festival (Figure 3).

Historical and Ethnographic Research The Krobo are knowledgeable about their past, especially with reference to the last century of the mountain settlement. This level of awareness of the local traditions may have resulted from the forceful removal of their ancestors from their ancestral home. This consciousness may also have resulted from the retention of some socioreligious institutions and practices such as the Dipo (Krobo traditional puberty rites for young female adults) and Klama (Krobo folk music) by the successor population. These two performances preserve and perpetuate the traditions, culture and history of the Krobo (Coplan 1972). There are great similarities in the material culture (ornamentation, architecture, and pottery) recovered from the mountain and those from the successor settlements.

Historical documentary sources also proved very useful in this study. Publications on Krobo, particularly those based on archival records such as the Basel Missionary

Because they are well informed about their cultural heritage, and the events of the 1892 sack of the Klowem, 73

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA reports were utilised. The works reviewed included Huber (1993); Omenyo (2001); Steegstra (2005); and Wilson (1987, 1995). Each of these works attempts the evaluation of some aspects of Krobo socio-economic, religious and political developments through time and space. For instance, issues about the origins of the chieftaincy institutions; dipo rites; commercial farming; the huza (a traditional Krobo cooperative land acquisition system); the beginnings of the Krobo Mountain settlement; destruction of shrines by the colonial army; architectural designs; and population and external influences on the people are highlighted in these sources.

study, the Manya Krobo authorities under whose jurisdiction the northeastern section falls were not contacted. But arrangements are far advanced to cover both sections in subsequent fieldwork. For this reason, whereas the historical references in this article may sometimes relate to both the Yilo Krobo and Manya Krobo, the archaeological evidence only pertains to the Yilo Krobo (south-western) section. The Excavations The limited excavations undertaken to date provided evidence about ancient burial practices on the mountain. It also supplied data for the development of a relative chronological framework for the settlement. This was achieved through a comparative analysis of finds from the surface survey and excavations. The comparison was undertaken to ensure proper ordering of the surface finds that were mixed up by erosion and human activities on the mountain. Three units comprising a pit (1 x 1.5m) and two trenches (Trench 1 = 2.5 x 4m and Trench 2 = 1 x 9m) were excavated. The excavation was aimed at providing information on the techniques used in building the terraces. For this reason all three units were opened close to terrace walls. The idea behind this was to reveal the hidden parts of the terraces, and thus expose the original terrain for assessment. The excavations also sought insight into the chronology of the settlement through recovery of dateable finds from stratified contexts in the pits.

Site Surface Survey The site survey provided an insight into the nature, distribution, function and chronology of some of the cultural and natural features identified on the mountain. For example, the results from the survey were useful in the determination of the relative age of imported ceramics, glass bottles, and hence associated sociocultural developments. Artifact assemblages, individual artifacts, surface features as well as the landscape were observed, analyzed, and recorded. The distribution pattern of the local pottery, glass bottles, and architectural remains was recorded. The ancient architectural landscape (including stone walls, stone terraces, foundation stones and house mounds), caves, rock shelters, water cisterns, shrines, possible burials and other features have been studied. The surface survey was extensive, covering the accessible portions of the southwestern portion of the mountain. It was done on foot, and involved observation, collection, recording and mapping of selected archaeological features. This survey strategy facilitated the documentation of a substantial amount of archaeological and environmental data for further off-site or laboratory analysis and interpretation.

The excavations were restricted to the Okpε suburb. The suburb was selected for excavation because it had the highest concentration of terrace foundation platforms and artifacts and the main ritual ground of the entire settlement, locally known as Okpεtesa was located there (Huber 1993: 34). It was also believed that the abundance of house foundations in the area as well as the associated dateable artifact assemblage reflected what was below the surface. As expected, large amounts of diagnostic finds were recovered from the excavations.

The artifact assemblage retrieved through the surface collection involved selective sampling of artifacts and ecofacts deemed relevant to the study. Dateable objects such as schnapps bottles, drinking glasses and imported ceramics were preferred. This artifact assemblage bears makers’ marks or seals that indicate source and period of production. The relative chronology for the site was thus obtained through the evaluation of these artifacts. Human and animal bones were also recovered and the discovery of exposed human bones from a number of locations provided data on burial practices in 19th century Krobo.

Features and Finds. A total of over 6,500 archaeological remains was recorded and/or recovered during the survey and excavations on the Krobo Mountain (Tables 1-4). The site yielded cowries, oyster shells, grinding stones, pottery, iron implements, glass bottles, glass beads, and animal and human bones. Seven hundred and seven features including stone terraces and mounds representing collapsed buildings were also found.

The archaeological survey was restricted to the southwestern side of the mountain, because at the time of the Category

Local pottery

Total No.

3000

Imported goods: glass beads, smoking pipes, glass beads, cowries 800

Percentage (%)

51.3

13.7

Human and Animal bones

Glass bottles

40

2005

5845

0.7

34.3

App. 100%

Table 1. Summary of the Archaeological Remains from Krobo Mountain Site

74

Total

WILLIAM NARTEH GBLERKPOR: CURRENT ARCHAEOLOGICAL RESEARCH AT THE KROBO MOUNTAIN SITE, GHANA

Category

Kaserving bowls

P‫כ‬-buε: cooking pots

Kagrinding bowls

Waterstorage pots

Takpacarrying water

GbakuWashing Pot

Kadu‫כ‬bath pot

Others

Total

Total no. Percentage (%)

945

150

390

780

390

150

180

15

31.5

5.0

13

26

13

5.0

6.0

0.5

3000 App. 100%

Table 2. Classification of Local Pottery Schnapp Bottles

Items Total No. Percentage (%)

Medicine, glasses

Cosmetics,

condiments,

wine

Total

1604

401

2005

80

20

App. 100%

Table 3. Glass Bottles/Objects Category

Terraces, building platforms Ruins of houses (including a palace)

Special places: shrines, sacred places, treasury and prison

Total

700

7

707

99.0

1.0

App. 100%

Total No. Percentage (%)

Table 4. Architectural features Recorded

these are connected to form a continuous drainage, and this drainage system prevents the fast-flowing surface run-off from washing away buildings along the slope in the area. Behind, and on-top of most of these stone walls, are mounds formed by collapsed mud structures.

Architecture and Landscaping The architectural landscape of the Krobo Mountain settlement is astonishing. In all, over 700 well preserved architectural features including terrace walls, house mounds, foundations/platforms, walls, and piles of stones blocks were recorded (Table 4; Figures 4 and 5). This figure would have been much higher (probably over 1,000) as the mud houses without foundations that once stood on the flat portions of the land have collapsed and their remains eroded by rain water. The latter estimate being supported by the identification of the scanty ruins of mud walls in one of the flat areas, as well as the discovery of domestic artifacts (storage jars, cooking vessels and grinding stones in concentrated contexts) which supports the former presence of house units in those areas.

Three major types of terrace walls were identified; terrace walls that served as building platforms, which directly supported stone and/or mud houses; terraces that served as retaining walls to check erosion; and terraces that helped to level the steep sloping portions of the landscape to facilitate normal domestic activities and easy climbing (Figure 5). The terraces provided the basic adaptive tool with which the people coped with the rugged mountainous environment. The abundance of naturally occurring stone made stone-terracing the most logical response to the demands of the environment. The structural designs of the terraces, coupled with abundant domestic artifacts suggest that the terrace walls were built for settlement purposes rather than for agricultural use (Gblerkpor 2006). The discovery of two piles of stone blocks similar to those used to build the terraces suggests that stone terracing was in vogue right up to the sack of the settlement.

The mountain is characterized by several stone terraces ranging between 1.5 and 2.0m in height and 6.0 to 30.0m in length. They are located at accessible areas of the mountain, where stone foundations and piles of stone blocks were recorded. But the terraces and foundations are absent at the extremely steep slopes that are often inaccessible, and portions where the gradient is flat and could permit mud structures that did not require the erection of platforms. Wherever the terraces occurred in groups of more than 10 walls, a very interesting landscape develops. This landscape is composed of giant steps or short walls built with stones, capped by a gentle rising clay mound, and separated by a series of gaps. All

In addition to the terraces, two other walling styles were identified. The classification was based on the building materials used. They are clay walls (atakpame) and claystone walls (a clay wall with a number of stone blocks fitted in the corners and other parts of the wall) (Figure 4). The archaeological evidence indicates that the people 75

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

Figure 4. Ruins of a stone-mud house (Photo. W. N. Gblerkpor)

Figure 5. A terrace wall (Photo. W. N. Gblerkpor)

used both clay, and stone masonry, with stone dominating as a construction material in the building ruins recorded. This is evident in the over 650 stone structures as against 30 mud walls identified on the site, and this is understandable, since on the mountaintop stone was and is more common than clay, the latter the usual traditional building material in southern Ghana. It must also be noted that stone structures last longer than mud ones, thus influencing the relatively lower figures recorded for the mud houses.

Based on the available archaeological data, one cannot establish with certainty whether the clay or stone masonry was the first form of house architecture that was developed on the mountain. According to Arlt (1995: 7), during the second quarter of the 19th century, the people built European style houses in stone to entice the Basel missionaries working in the Krobo area to settle on the mountain. This was particularly the case with the area occupied by the Manya Krobo who are said to have embraced the Basel missionaries and western education. 76

WILLIAM NARTEH GBLERKPOR: CURRENT ARCHAEOLOGICAL RESEARCH AT THE KROBO MOUNTAIN SITE, GHANA

Figure 6. Ruins of a stone house (Photo. W. N. Gblerkpor)

presence of locks on the wooden doors of some of the houses. However, most doors and windows would have been secured with mats and woven palm fronds (Wilson 1995: 7).

Piles of stone blocks, possibly for the construction of stone houses were recorded in this area. The two house shapes identified on the ancient settlement site are the rectangular and circular, and these are similar to those found in the present-day Krobo area. Basel missionaries who visited the site in the 19th century confirmed the existence of these architectural designs (Huber 1993: 34). So far, three circular foundation stones representing circular buildings have been recorded, but many more could be uncovered in subsequent surveys. The majority of the house foundations and ruins of buildings recorded measured about 5-8m x 6-10m, which indicates their small size. This archaeological evidence is consistent with Johannes Zimmerman’s (a Basel Missionary) report that the houses on the mountain looked “like eagle’s nest” with one room Wilson 1995: 87). Traditional architects in the study area still employ these designs in the construction of buildings, though the circular shape is restricted to indigenous shrines. According to Steegstra (2005: 29), many of the houses on the mountain were single-storey structures (comprising a ground floor, and a first floor). However, except for the ruins of the palace, the study did not find any archaeological evidence to support this view as only small portions of the walls were found. Nevertheless, there are a number of old single-storey building structures in many of the successor towns such as Somanya and Odumase, and these are believed to have been constructed in line with the architectural designs used on the mountain.

Population Density The archaeological evidence indicates a high population density on the Krobo Mountain during the 19th century. With an estimated figure of about 1000 house units on the mountain, distributed over such a relatively limited habitable area (about 1 x 1km ), the settlement would have been densely populated. Even if 700 out of the estimated 1000 buildings served as dwelling places/rooms, and with each room occupied by an average of three persons, there would have been 2100 people staying on the mountain in the 1800s. This estimate applies to the sedentary mountain population, that is, the people that used the mountain as their permanent settlement. This figure could have substantially increased when the populations of the farming villages ascended the mountain to celebrate the dipo rites or perform funerals. According to Krobo traditions, the mountain population increased considerably during periods of religious and social celebrations, and a small room then accommodated more than five persons at a time. In addition to the houses, there is ample archaeological proof for cave habitation on the mountain. Some of the caves identified are big and comfortable enough to accommodate people. Several domestic utensils such as cooking, storage, serving and drinking ceramic vessels recovered from the caves could have been used by the inhabitants of the caves. Oral history confirms that a lot of the caves were inhabited by the pioneer settlers prior to the construction of houses on the mountain.

The oral traditions also mention wood (obtained from the top of the mountain) and grass (from the surrounding plains) as the main building materials besides stone. Ten pieces of building hardware including door hinges, a keyhole plate and parts of a door-lock were recovered. The recording of the door accessories suggests the 77

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

Figure 7. Gmawe (water-storage vessels) under a rock shelter on the Krobo Mountain (Photo. W. N. Gblerkpor)

It is fed by rainwater that flows from boulders located on higher ground. The surface area of the second pond is about 1 x 6m and about 1.3m in depth. Three depressions believed to have been used as cisterns were also identified. However, if all these water points supplied water at their maximum they may not have satisfied the water needs of the population, especially during the dry season.

After the Krobo had settled on the mountain, they were joined by other people fleeing the Asante invasions and slave raids during the 18th century. These included Denkyera refugees who fled the Dankyera-Asante War and Ewe groups (Huber 1993: 34; Arlt 1995: 9). The influx of these refugees substantially increased the population of the mountain settlement, which led to a shortage of land in the immediate surroundings of the mountain. Subsequently, the Krobo acquired new territories from the Akwepem and other neighbouring groups.

The archaeological evidence suggests the adoption of other water management strategies by the inhabitants. For instance, out of the about 3,000 local potsherds found on the site, as many as 1,070 (39%) belong to vessels for carrying or storing water. Examples include takpa for carrying water; gmawe and did‫ כ‬for storing water (Table 2 and Figure 7). These storage vessels measure over 1m high and 2m in diameter and can hold up to 800 litres of water. It is also possible that the residents harvested rainwater from the roofs of their houses. This traditional water harvesting technique is still practised in the successor settlements, and in many other villages in Ghana. The oral accounts claim that in the dry season, additional water was obtained from nearby streams such as Okue, located on the plains. Although hauling water from the plains up the mountain may have been difficult for the population, it was not impossible.

By the 19th century, the Krobo Mountain settlement was densely populated. The traditions state that the high population received an increase during festive occasions, when many people from the surrounding villages congregated to celebrate festivals or final Dipo rites. When Rev. Heck, a European missionary, visited the Krobo Mountain in the 19th century, he described the settlement as, “a curious town! We stand on the flat roof of a house, we look up, and high above our heads on a majestic rock there stands a second house. We look below, and straight beneath our feet there, a third house. Such are the conditions of the entire town” (Huber 1993: 32-36). Most of the houses described by the missionary are now represented by the stone-foundation platforms capped by clay mounds, ruins of walls, and the door accessories described earlier.

Indigenous Religious Practices There is considerable archaeological evidence on indigenous African religious practices on the Krobo Mountain. The study produced a number of architectural remains, natural features and an artifact assemblage known to be associated with local religious performances. These include three circular house foundations representing “room-kept shrines”, and four “out-door shrines” as well as iron bells, and terracotta figurines.

Water Resources and Management The inhabitants of the Krobo Mountain, located 350m above sea level, and without any major source of water, were faced with the challenge of acquiring water for domestic and other uses. Two small water ponds formed by depressions on rocks were discovered. The first pond measures 1.5m in depth and about 3 x 6m in surface area. 78

WILLIAM NARTEH GBLERKPOR: CURRENT ARCHAEOLOGICAL RESEARCH AT THE KROBO MOUNTAIN SITE, GHANA

Figures 8(a). Priestess wearing a straw hat decorated with cowries; (b) Cowries recovered from excavations on the Krobo mountain (Photos. W. N. Gblerkpor)

shrines. For instance, the Ka-si-tsotsε was used for serving mamunyu (millet or corn flour solution) to the gods. Glass beads of all shapes and colours (particularly white) as well as cowries have been collected from shrine contexts. The ethnographic information suggests that these beads were worn by traditional priests and priestesses in Krobo. White-coloured beads, which signify purity in Krobo tradition, are the most preferred ones by indigenous priests and priestesses (Figure 8 [ab]).

Traditionally, shrines found in Krobo and other Dangme areas are rarely housed or built in permanent building materials. But when they are, the structures are circular in shape, representing a design believed to be a typical Dangme building design. The presence of circular dipo shrines in present-day Krobo and Shai areas provide examples that corroborate the archaeological record and the written and oral historical accounts. The most obvious religious structures identifiable on the mountain are the huge rocks associated with artifact assemblages and plants associated with local worship. These rocks appear in three major categories: 1) ‘large’ flat-surface rocks (40-50m in width and 50-60m in length), 2) rock shelters, and 3) rock boulders. The Okpεtεsa (rock of Okpε) (40 x 60m), located within the Okpε suburb typifies the ‘large’ category. The Okpεtεsa (shrine) was littered with several schnapps bottles, potsherds and cowries. It was also encircled by cactus and jatropha plants. These and other local plants like the buna tree and nyabatso provided herbs for ritual cleansing and curative purposes. These plants are found at present-day shrines in Krobo, including the palace shrine of the Yilo Krobo paramount chief. These plants also serve non-religious functions such as in their use for the construction of fences for gardens. Perhaps the plants ability to withstand drought and fire make them suitable for use in religious practices.

More than 1,600 19th century schnapps glass bottles representing 80% and 27.4% of glass objects and total finds, respectively, were recovered. In addition to the liquor bottles, local pottery vessels, mumui and lik‫כ‬k‫כ‬ (little clay cups) for tapping palm-wine and pouring of libations have also been retrieved (Figure 9[a-b]). This suggests a high use of liquor by the inhabitants. Although social drinking may have contributed to the liquor collections, the constant association of these artifacts with shrines suggests their use in rituals and religious practices. Alcohol libation constitutes an important aspect of Krobo indigenous religion. This is consistent with local traditions and ethnographic practices among the Krobo, whose religious, social and political activities are accompanied by lavish use of schnapps. The ancient Krobo were polytheistic. They had gods that were responsible for various natural and human events. For instance, there were Ohue dumla (rain god), Medoku Aye (god for blacksmithing), Nadu and Kotoklo (war gods) and Likpotsu (god responsible for peace, riches, and prevention of infectious diseases). As an integrating element of the Krobo society, the Dipo rites and the worship of the war gods Nadu and Kotoklo were of great importance.

The use of the Okpεtεsa for the performance of the mandatory dipo ceremonies may have accounted for the prominence of the Okpεtεsa shrine in Krobo traditions. According to the oral traditions, it was specifically here that the dipo ceremonies were performed. As noted by Steegstra (2005: 129), shrines played important roles in ancient Krobo, and as many as 133 ‘fetish haunts’ (shrines) were reportedly destroyed by the Hausa soldiers who enforced the ejection order of the British authorities.

The indigenous priests and priestesses played important roles in 19th century Krobo. They conducted the customs and rituals associated with the worship of the gods. They still play various key roles in contemporary customary rites and festivals in the area. For example, the Yilo Krobo priestess, Mama Adzovi, seems to have a lot of

Artifacts retrieved from the excavations and associated with indigenous religious activities include a terracotta figurine. These items served various purposes in the 79

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

Figure 9(a). Medallions from schnapps bottles; (b) Schnapps bottles and palm-wine tapping/storage vessels (Photos. W. N. Gblerkpor)

influence on the community because her safe ascent and descent of the Kloyo is a pre-requisite for the formal commencement of the Kloyosikplemi. The author witnessed this during the 2006 festival when the durbar of chiefs took place after her descent of the mountain. The archaeology has not produced any significant evidence to show external influence on Krobo indigenous religion. Apart from three rusted iron gong-like bells and some glass beads whose origin may be attributed to the neighbouring Ewe group, very little archaeological evidence was retrieved. Perhaps subsequent study may bring more data to enhance our knowledge on external influences on the local religion. But the historical accounts and ethno-linguistic evidence acknowledge considerable Ewe elements in Krobo religion. For instance, the name Mama Adzovi for the Yilo State priestess is Ewe. The priestess spoke the Ewe language with at least one of her associate priestesses during the 2006 Kloyosikplemi festival (Kodzo Gavua, Personal communication 2006). This can be explained by the fact that the Ewe were among the non-Dangme groups which joined the Krobo on the mountain for the festival.

Figure 10. 19th Century imported glass beads (Photo. W. N. Gblerkpor)

Burial Practices

communities bury their loved ones in their homes. This is however done only with the permission of both the paramount chief and National Health authorities.

A total of nine human bones were recovered from the excavations and from the surface of the site. These comprised two skulls, four forelimbs and a thigh bone retrieved from the excavations, and two more forelimb bones recorded during the surface survey. The bones came from under the foundations of two house mounds, suggesting that they were originally buried in rooms. Grave goods recovered included an earthen pot, Hebron beads, cowries, and 19th century European composite glass beads (Figures 9 and 10). The disclosure of the human remains is noteworthy as intramural sepulture was prohibited by the colonial authorities under the Native Cemeteries Ordinance of 1888 (Steegstra 2005: 123). Although the associated trade goods date the excavated burial to the 19th century, it could not be determined whether the burial took place before or after the passage of the 1888 ordinance. Although intramural burials are not generally practised, some people in the Krobo

Finally, the discovery of some beads around the waist and knee areas of one of the remains shows that at least one of the deceased was female. In Krobo and other Dangme customs, waist beads are the preserve of females of all ages and sometimes infant males. From the archaeological evidence, it is clear that the people buried their dead with grave goods, such as precious beads and domestic vessels, suggesting belief in life-after death. Nowadays less grave goods are added with the dead, because grave looting is rampant in some of the communities in the area. There is strong archaeological evidence for high liquor use on the mountain. For instance, numerous J. H. Henkens schnapps brands were retrieved. They include the Stork, Worm and Star brands (Figure 9[a]). In many 80

WILLIAM NARTEH GBLERKPOR: CURRENT ARCHAEOLOGICAL RESEARCH AT THE KROBO MOUNTAIN SITE, GHANA

Figure 11. Ruins of the 19th century palace for the Kono of Yilo Krobo (Photo. W. N. Gblerkpor)

export value of palm-oil from the Krobo area amounted to 900,000 litres per annum (Anquandah 1985: 21). To sustain their oil-palm plantations, they purchased new territories at the foot of the Akwapem Hills (Arlt 1995: 4; Huber 1993: 32) using the Hunza system. In this traditional land purchasing system, a group of interested land buyers selected a wealthy or rich person in the Krobo society to buy a plot of land on their behalf. The plots were later paid for and shared among the individual contributors.

traditional African societies, funeral ceremonies are usually accompanied by drinking. There were reports by some Basel Missionaries that the Krobo consumed a lot of liquor during funerals to forget their anguish from the loss of their loved ones (Steegstra 2005: 123).   Economic Activities  The archaeological evidence points to a prosperous 19th century Krobo society. The density and diversity of imported artifacts such as schnapps, glass beads, cowries, roofing sheets, and door accessories on the site suggest the active participation of the people in the economic activities of the Gold Coast at that time. Two features identified as palm oil processing cisterns show that there was commercial production of palm-oil on the mountain. This is also a reflection of the economic benefit in the area. The economic benefits were also noticeable in the construction of several well planned rectangular stone buildings, including the one-storey palace (Figure 11) that was constructed in the 1850s during the reign of Kono Ologo Patu (c.1832-69) (Wilson 1995: 278). Although yet to be investigated archaeologically, the presence of a cave claimed by oral tradition to have served as the treasury for the settlement gives further support to the apparent economic prosperity enjoyed by the people.

The overall performance of the Krobo economy can be inferred from the numerous cowries that are found at the site. The use of cowry shells as medium of exchange was widespread in the Gold Coast (Bredwa-Mensah 2002: 160). At the beginning of the 19th century, the English were shipping about 100 tons of cowries to Africa annually. But with the abolition of the slave trade in 1807, their shipment dropped to an average of about six tons a year between 1810 and 1815. The cowry trade however, recovered with the development of the palm oil trade. By the 1830s, between 60 and 100 tons went annually to the Gold Coast alone, and from 1835 at least 50 tons a year to east of the Volta. Later on in the 19th century, the cypraea moneta was replaced by the cypraea annulus in West Africa, and it (cypraea annulus) was accepted by the palm-oil merchants. By 1850 the Gold Coast was importing cowries at the rate of some 150 tons a year (Johnson 1970: 49; York 1972: 96-98). It is not surprising that large numbers of cowries were traded to the Krobo Mountain as the inhabitants participated in the economic activities of the 19th century.

The historical accounts confirm the economic achievements of the inhabitants, as it has been reported that the Krobo were major producers of oil-palm fruits and palm-oil in the Gold Coast. It is estimated that the 81

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA retrieved from the site and other contemporary sites in the East Accra Plains have been credited to the neighbouring Dangme Shai people (a closely related socio-linguistic group) (Anquandah 1985: 22). Even though the results of the mineralogical analysis of the pottery from the Krobo site did not point to the Shai area as the sole source of the ceramics, the ethnographic and historical evidence support a Shai origin for the majority of the pottery recovered (a view contrary to an earlier submission [Gblerkpor 2005: 104]). Whereas a potting tradition persists in the Shai area and the potters still produce ceramic vessels similar to those recovered from the site, the potting tradition in Krobo is weak, and moreover, the act of potting is almost absent. Also, Krobo oral and historical accounts allege that the Shai were commercial potters who supplied the predominantly farming Krobo population with pottery for domestic use and for the export of their palm oil to Erurope (Anquandah 1985: 22). Based on the available data, it is maintained that even if the Krobo produced some pottery while on the mountain, the bulk was probably of Shai origin.

Political Structure The existence of a huge 19th century stone palace on the mountain, the well planned nature of the architectural landscape, as well as the location of two structures described as a prison and a treasury indicate the presence of a strong political leadership or institution. The old palace represents the single largest building on the mountain (Figure 11). This single-storey palace was strategically located at almost the centre of the settlement, among huge boulders with an excellent view of the surrounding plains. The comparatively large size of the palace and its strategic location on the mountain confirms the prominence of the position of the Kono in 19th century Krobo. According to Wilson (1995), the Krobo Mountain settlement was a strong chiefdom by the 19th century, ruled by paramount chiefs (Konos). But prior to the emergence of the chieftaincy system of governance the settlement, like many other Ga-Dangme groups, was ruled by priests and priestesses, a kind of political system Wilson refers to as “priestly oligarchy” (Wilson 1995: 49).

Conclusions There are various local traditions concerning the emergence of the chieftaincy institution in Krobo. One version states that refugees, who fled to the mountain after the defeat of Denkyera by the Asante, introduced the chieftaincy institution. This tradition is supported by a section of the Krobo population and some of the royals of the Yilo paramountcy profess their Denkyera ancestry.

The study confirms earlier historical accounts about cultural developments on the Krobo Mountain and shows adaptive strategies in architecture, landscape designs, indigenous water management and local craft and technology. The well planned nature of the settlement and the identification of features such as the palace, shrines, houses and burial grounds confirm oral and written historical claims that Klowem was the major political and socio-religious centre for the Krobo. The finds also indicate the presence of effective central political and religious institutions on the mountain.

Contacts and Influences The archaeological evidence indicates that the Krobo Mountain settlers had contacts with neighbouring ethnic groups in the Accra plains, as well as with Europeans and other local peoples on the coast. The occurrence of about 30% European imported goods in archaeological contexts suggests direct and indirect European contacts and influences in the area. For instance, apart from the rectangular structural designs of some of the stone houses including the palace, the building hardware such as the roofing sheets, nails, bolts, and door accessories were definitely European. Many more archaeological remains such as the cowries, schnapps bottles, glass beads, wine glasses, ceramic vessels and pipes dating to the first half of the 19th century were certainly European imported goods. The historical evidence confirms this conclusion. By the second quarter of the 19th century, the Basel Missionaries were stationed in the Krobo and as part of their efforts to Christianize the local population, they constructed Western styled houses and introduced formal education. These early European architectural designs influenced the houses recorded on the mountain.

The archaeological data retrieved from the Krobo Mountain site shows the Krobo Mountain was a major settlement. The high numbers of domestic artefacts and the near absence of agricultural implements at the site, suggests that the site was principally used for habitation rather than for agricultural purposes. There is strong linguistic evidence that supports this view. Traditionally, the people call the site Klowem (home of the Krobo), confirming that the site was a settlement rather than an area of farmland. The prevalence of imported trade goods, the presence of what is believed to be the treasury for the settlement and the massive infrastructural development suggests economic, social and political prosperity of the settlement during the 19th century. The study also shows that the stone terraces of the Krobo Mountain were a land reclamation strategy that was aimed at creating more habitable areas for occupation. The terraces provided the basic adaptive tool with which the people coped with the conditions posed by the steep-sided nature of the terrain. The abundance of naturally occurring stone made stoneterracing the most logical response to the demands of the environment. The finds also confirm that the use of stone masonry survived up to the time of the abandonment of the settlement.

Although the archaeological evidence for local contacts and influences are not as diverse and obvious as those for European ones, they are still significant. The stone terrace walls on the mountain have also been identified at neighbouring Shai Hills ancient settlements, the ancestral home of the Shai (also sacked in 1892). Furthermore, the thousands of local pots, and sherds of local pottery 82

WILLIAM NARTEH GBLERKPOR: CURRENT ARCHAEOLOGICAL RESEARCH AT THE KROBO MOUNTAIN SITE, GHANA W. Kankpeyeng, John Ako Okoro, and James BoachieAnsah, all of the Department of Archaeology, University of Ghana, Legon, for finding time to read through this paper and providing valuable suggestions. Kodzo Gavua, Head, Department of Archaeology, University of Ghana, provided the needed institutional support during the field work.

To a very large extent, the present landscape was created by the ancient Krobo. The demand for more houses by the increased population led to the construction of stone terraces and building of single-storey houses to accommodate more people. The landscape demonstrates the industrious spirit of the Krobo, and shows their resolve to make the hostile mountain terrain their dwelling place in the prevailing hostile political environment. The research therefore reveals the complex interaction between the inhabitants and their natural habitat, which demonstrates how nature and culture influence each other in a complex manner.

References ANQUANDAH, J. 1985. Ethnoarchaeological Clues to Ghana’s Great Past and Greater Future? University of Ghana Monographs and Papers in African Archaeology 2. Department of Archaeology, University of Ghana, Legon.

The high representation (39%) of pottery for storing water, as well as the absence of a major source of water on the mountain is indicative of the water crisis the people faced on top of the mountain. This figure also shows that the inhabitants developed rigorous traditional water management strategies. Whilst the hundreds of whole pots found on the mountain support oral and written historical claims that the settlement was abandoned in haste. The terminal occupational layer of the site is dated to the 19th century, but the earliest period of the settlement could not be determined by the study.

ARLT, V. 1995. Diplomacy and Power Politics in MidNineteenth Century Krobo - Krobo Chieftaincy Seen through the Reports by Basel Missionaries. Unpublished M.A. Thesis, Department of History, University of Basel, Basel. BREDWA-MENSAH, Y. 2002. Historical – Archaeological Investigations at the Fredriksgave Plantation, Ghana: A Case Study of Slavery and Plantation Life on a Nineteenth Century Danish Plantation on the Gold Coast. Unpublished PhD Thesis, Department of Archaeology, University of Ghana, Legon.

Acknowledgements I wish to thank the Ghanaian and Danish Co-coordinators of the Ghana – Denmark Archaeological Project (G – DARCH Project), Dr. Yaw Bredwa–Mensah, Department of Archaeology, University of Ghana, and Prof. Klavs Randsborg, Department of Archaeology, University of Copenhagen respectively for providing financial support for the 2004 fieldwork. Mr. Bossman M. Murey, the Chief Technician, Department of Archaeology, University of Ghana, Legon, also deserves a thank you for his support throughout the field and laboratory work. I also benefited from his experience in archaeological field techniques and photography during the initial stages of the study. Furthermore, I thank Mr. Cosmos Logosu, Driver-mechanic, Department of Archaeology, University of Ghana, Legon, for his great support during the fieldwork. Sincere thanks go to the levels 100, 200 and 300 students (2003/2004 academic year) of the Department of Archaeology and other Departments of the University of Ghana, who voluntarily participated in the 2004 field work.

COPLAN, D. B. 1972. Krobo Kalama. Unpublished M.A. Thesis, Institute of African Studies, University of Ghana, Legon. DEBRAH, J. N. 1982. Archaeological Survey of the Krobo Mountain Ancient Settlement. Nyame Akuma 21: 17. GBLERKPOR, W. N. 2005. An Archaeological Investigation of the Krobo Mountain Dry-stone Terrraces. Unpublished M.Phil Thesis Department of Archaeology, University of Ghana, Legon. GBLERKPOR, W. N. 2006. An Archaeological Investigation of the Krobo Mountain Settlement Site. Nyame Akuma 65: 18-23 HUBER, H. 1993. The Krobo: Traditional, Social and Religious Life of a West African People. Studia Instituti Anthropos, Vol. 16. Fribourg: Fribourg University Press.

I am also grateful to the chiefs and people of my study area (Yilo Krobo) for willingly accepting my request to carry out this study at their sacred ancestral site without placing any restriction in my way. They welcomed me to their homes and palaces, provided me with free accommodation, and treated me as their own son. I acknowledge the paramount chief of Yilo Krobo, Nene Narh Dorwutey Ologo VI, Paramount Queen Mother, Naana Korleykuor Adjado III, Padi Ologo, Waku Matse and the Development Chief, Nene Tumey Odonkor for their immense assistance. My local field assistants, Tanko and Emmanuel, who carried out most of the dig, deserve a big Ayekoo (thank you). I also wish to thank Benjamin

JOHNSON, M. 1970. The Cowrie Currencies of West Africa (Parts 1 and 2). Journal of African History 11: 17–49, 133–335. KROPP-DAKUBU, M. E. 1982. The Peopling of Southern Ghana: A Linguistic Viewpoint. (In), Ehret, C. and Posnansky, M (eds.), The Archaeological and 83

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA Linguistic Reconstruction of African History. Berkeley: University of California Press, pp. 245-225. STEEGSTRA, M. 2005. Dipo and the Politics of Culture in Ghana. Accra: Woeli Publishing Services. OMENYO, C. N. 2001. The Ongoing Encounter Between Christians and African Culture: Case Study of Girls Nubility Rites of the Krobos. Accra: Jupiter Printing Press Ltd. WILSON, L. E. 1987. The Rise of Paramount Chiefs among the Krobo (Ghana). International Journal of African Historical Studies 20: 471-495 WILSON, L. E. 1995. The Krobo People of Ghana to 1892- A Political and Social History. Monographs in International Studies, African Series, No. 58. Athens, OH: Ohio University Press. YORK, R.N. 1972. Cowries as Type-Fossils in Ghanaian Archaeology. West African Journal of Archaeology 2: 93101.

84

Placing the Tongo Hills, Northern Ghana, in Archaeological Time and Space: Reflexivity and the Research Process Timothy Insoll Archaeology, School of Arts, Histories, and Cultures, University of Manchester, and Visiting Lecturer, Department of Archaeology, University of Ghana, Legon. Email: [email protected]

excavations had been completed in the Upper East Region (UER). Previous relevant research completed in UER had been either ethnographic or ethnoarchaeological in focus. As attested, for example, by Mather’s (1999, 2003) investigation of Kusasi shrines, Apentiik’s (1997) investigation of Bulsa technology and shrines, or Gabrilopoulos’ (1995) examination of the Tallensi compound. Similarly, Priddy (1974) had studied the extant potting traditions of the then Upper Region, as had Pole (1985: 149) the iron-working technology. Garrard (1986) has examined so-called Frafra brass-casting techniques, the latter study including a historical dimension, whilst Smith (1989) has looked, more recently, at pots and potters and their correlation with earth cults. The architectural heritage of the Tongo Hills has also been examined by Kankpeyeng (2001, 2005) in connection with preparing a proposal to gazette the area as a UNESCO world heritage site. Otherwise, archaeology has been neglected in the Upper East Region in comparison to the southern sector of Ghana, and this comparative archaeological neglect of Northern Ghana similarly extends to large parts of the Northern and Upper West Regions as well (e.g. Anquandah 1982; Kankpeyeng and DeCorse 2004).

Introduction Since 2004 an archaeological research project co-directed by the author along with Dr Benjamin Kankpeyeng and Dr Rachel MacLean has been running in the Tongo Hills of the Upper East Region of Northern Ghana (Figure 1), with Insoll focussed upon the archaeology, Kankpeyeng on architecture and heritage, and MacLean on survey. The aims of this research have been described in detail elsewhere (Insoll, Kankpeyeng, and MacLean 2004, 2005, 2007a, 2007b, in press) and are, by way of summary, twofold. Firstly, to reconstruct the sequence of occupation in the Tongo Hills and secondly, to examine the material culture and archaeological manifestations of African indigenous religions variously labelled as totemism, animism, and earth and ancestral cults among the inhabitants of the hills, the Tallensi ethno-linguistic group. Until the season just completed in March-April 2008, when occupation sites were made the focus of investigation, it is the second research objective which has received more detailed attention (e.g. Insoll 2006, 2007a, 2007b, in press). The purpose of this paper is not to provide further information on the results obtained to date which have already been published in preliminary and other forms elsewhere (see Insoll, Kankpeyeng, and MacLean 2004, 2005, 2007; Insoll 2006, 2007a, 2007b). Rather its purpose is intended to reflect on the project itself. Hence in the spirit of reflexivity which is influencing aspects of archaeological research (e.g. Hodder 1999: 80-104) this paper will consider the hermeneutical, epistemological, and empirical problems and issues which have arisen in attempting to place the Tongo Hills in archaeological time and space. This is not done in a spirit of academic ‘navel-gazing’ or pretension, which this author has criticised elsewhere (Insoll 2007c: 14), but as a genuine attempt to contextualise what is a foreigner’s perspective on, and contribution to, current archaeological research in Ghana. In so doing, the themes already considered in chapter one are returned to but here in an applied manner from an etic perspective.

An exception to this is provided by the pioneering and exhaustive survey completed by Oliver Davies (e.g. Davies 1970). The survey published in the volume on Northern Ghana (ibid) included what is now the Upper East Region and thus some observations on the Tongo Hills. Specifically, Davies (1970: 157) notes in summarising previous publications and reports, records of quartz debris and labrets or lip plugs, and ceramic sherds, and from his own observations (ibid: 157-158) at “Tengzugu”, the presence of a ceramic sherd found on the surface and “several grooved rocks”, and “many tumbles of granite rocks forming shelters” on the way to the village of Yinduri (Davies 1970: 158). Davies (1967: 210) also refers to “many” stone hoes in the “Tong Hills”, which, “were associated with small blades, side-choppers and what are probably mace-heads, there is much smashed quartz on the sites, but few microliths, no pottery or celts”. Yet other than Davies’ observations, which in the coverage of a whole country are allowed to be somewhat cursory, both the Tongo Hills and indeed the Upper East Region as an entity have been left floating in archaeological time and space, devoid of chronology and context. In comparison, the anthropology and history of

Previous Research and Project Rationale Prior to the excavation of the first test-pits in August 2004 in Kpeline and Songpaaveng Caves (Insoll, Kankpeyeng, and MacLean 2004: 25) in the Tongo Hills it is believed that no legitimate archaeological 85

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

Figure 1. Map indicating the position of the Tongo Hills within Ghana.

interest and importance, as well as on Tallensi technology, as, for instance, in regard to iron working by the blacksmiths of the Sakpee section of the Tongo Hills (Fortes 1936: 2, 6-7; Insoll in preparation). The second critical point just made, i.e. the empirical criticism applied to Allman and Parker (2005) has to be qualified in that it only refers to an absence of historical data from before the end of the nineteenth century and not to what is an exemplary study. Hence it has been classified as an ‘external’ limitation beyond the capacity of Allman and Parker (2005) to alter, and moreover a limitation which archaeology in the absence of historical sources should seek to rectify.

the Tongo Hills has been the focus of detailed study, notably by Meyer Fortes (1945, 1949, 1983, 1987) for the former, and Jean Allman and John Parker (2005) for the latter. But the studies resulting from both these disciplines suffer from certain constraints which are either self-imposed in terms of hermeneutics, or subject to external empirical limitations. The former criticism can be applied to Fortes (e.g. 1945, 1949) approach that is in its functionalist emphasis devoid of historical trajectory and sequence (Allman and Parker 2005: 15-16), and certainly from the perspective of the archaeologist or ethnoarchaeologist, is lacking in information on Tallensi material culture. Indeed, Fortes presentation of the Tallensi has been eloquently criticised by Allman and Parker (2005: 16) as having the result “that the Talensi were enshrined in ethnographic discourse as the archetypal stateless society, marooned on the margins of a distant hinterland in a timeless ethnographic present”.

The attempt to reconstruct the occupation sequence and ‘historical’ trajectories in the Tongo Hills in the absence of historical sources, i.e. via archaeology, is precisely what the rationale behind the current project has been. But again a proviso must be added that although Arabic historical sources might be missing, seemingly in entirety, and European ones until the late nineteenth century (Allman and Parker 2005: 18, 25-28), oral history of course exists (ibid: 19). Fortes (1945: x) in fact refers to the historical time span recognised by the Tallensi as five or six generations (Insoll 2006: 225). However, Fortes does not draw much upon oral sources and indeed makes the further point that “the Tallensi, therefore, have no history in the sense of a body of authentic records of past events” (1945: 26). Equally, during the course of the research described here, oral history has not been of great use in providing specific dates, associations, markers of events etc., but it has been useful in attesting to a consciousness of the enduring importance of some of the

Perhaps though, Fortes is an easy target for criticism in this regard, and as Kirby (2006-2007: 76) notes, “on many occasions Fortes railed against what he felt was a too narrow interpretation of his ‘equilibrium models’ and against what had then become practically an epithet among the more youthful generation of anthropologists, namely ‘functionalism’”. Also of interest in dissecting Fortes ahistoricism, and to mis-use a concept derived from elsewhere, what could be called his ‘antimaterialism’, it should be noted that his corpus of published work stands somewhat in contrast to his unpublished archive. The latter is housed in the Manuscripts Room in the University Library, Cambridge, and contains fragments of information of historical 86

TIMOTHY INSOLL: PLACING THE TONGO HILLS, NORTHERN GHANA, IN ARCHAEOLOGICAL TIME AND SPACE pastes, inclusions, decorative attributes etc (see Ashley unpublished). These points of reference in relation to existing work completed would be, excluding the Upper East Region for obvious reasons already discussed, in the Northern Region of Ghana and southern Burkina Faso. For the latter, nothing of relevance in the literature has been found as yet. This in turn reflects, again, a comparative lack of research (but see Madiega and Nao 2003), but also the Francophone-Anglophone West African archaeological divide (see Kense 1990; De Barros 1990), which though not a deliberate construct, nonetheless exists and makes difficult what should be a relatively simple comparison across two frontiers which share contiguous ethno-linguistic groups such as the Kasena around Paga and Navrongo in the Upper East Region of Ghana and north to Po in Burkina Faso.

sites, such as the Nyoo shrine, which have been investigated (Insoll, Kankpeyeng, and MacLean in press). Perhaps this is a general correlate of oral history, i.e. an inherent vagueness as opposed to implicit specificity (e.g. Vansina 1985: 186-190), and the existence of this ‘quality’ of oral history has further emphasised the importance of archaeology, albeit with all its limitations in attempting to reconstruct the more fine-grained aspects of the past (Insoll 2007c). Research Realities Hence having outlined something of the project rationale, the question can be asked as to how successfully these objectives have been achieved to date, and thus are the Tongo Hills beginning to be anchored, archaeologically, in time and space? It would of course be nice to say that all the research objectives are being fulfilled as predicted, but the realities of research are not so straightforward. Various factors impinge, as they do upon all research projects, though not all project directors might admit such so candidly.

Looking south to the Northern Region of Ghana for comparisons is also difficult. Here, the obvious point of first reference is the Gambaga Escarpment southeast of the Tongo Hills. The Gambaga Escarpment, or more specifically the Mamprugu Kingdom centred on the towns of Gambaga and Nalerigu (Drucker Brown 1975) had, historically, links with the Tallensi. According to tradition, the institution of chieftainship was brought by the Namoos from Mamprugu, with the Namoos, along with the Talis, together forming the overall Tallensi ethnic group (Fortes 1945, 1949). When this event occurred is unclear, Fortes (1987: 43) suggests some fourteen to fifteen generations ago, which this author has inferred is presumably from the 1930s-40s, the era of his primary field research which would give a conjectural date of c.1650 (Insoll 2006: 225), but regardless of the chronological imprecision it at least provides a potential geographical marker for comparative purposes. Parts of the Gambaga Escarpment have also been the focus of archaeological research, but whilst publication of the Late Stone Age Kintampo material has been exemplary (see Casey 2001), the same cannot be said, unfortunately, about the Iron Age research. This has been described in outline form (e.g. Kense 1992), but has never resulted in full publication of the work. This means, obviously, an absence of comparative material for the Tongo Hills where the material recovered is of Iron Age date, excluding, perhaps, a small quantity of undated seemingly Late Stone Age material from Hyena’s Cave (see Insoll, Kankpeyeng, and MacLean 2007a).

i) ‘Terra Incognita’ An initial factor of consequence is the fact, already stated, that the Tongo Hills are archaeological ‘terra incognita’. This, today, is an increasingly rare situation, at least in European experience (see Insoll 2007c: 23-28), and this point was reinforced to the author at a seminar he gave to a group of Scandinavian postgraduate students on the project at the Danish Research Institute in Athens, Greece, in April 2007. Following this seminar a student asked, “but how can you work where no-one has worked before, are you not worried about making mistakes”. It would be foolish to answer other than in the affirmative but equally to worry too much would mean the research project would never have been initiated. Being put in the position of a ‘pioneer’ has both advantages and disadvantages. For in beginning research in such an area one is presented with a blank slate, hence the first excavations, the first radiometric dates, the first ceramics sequence become a correlate of the project. The reverse is that there is an absence of comparative sequences and material, and of support networks, informal and ad hoc as they might be, of archaeologists and scholars from other disciplines which are almost taken for granted in better archaeologically investigated areas of the world.

In fact, the closest parallels for aspects of the ceramic assemblage, albeit somewhat later in date if the preliminary chronology reconstructed for the Tongo Hills is correct (see below), are with the ceramics from Yikpabongo in Komaland on the border of the Northern and Upper Western Regions and reported by Anquandah as dating from between c. AD 1200 to 1800 (1998: 82; see also Kankpeyeng, this volume). Similarities, between the two assemblages, which may or may not be meaningful, include decoration with string or plaited grass roulettes, decoration by incision, decoration by comb stamping, and the use of slipping for decoration (see Anquandah 1998: 103-109), and vessel similarities include the presence of what Anquandah (1998: 112)

This has been most evident in relation to studying the pottery from the excavations. In the initial two seasons of essentially reconnaissance survey and trial excavation (Insoll, Kankpeyeng, and MacLean 2004, 2005) the lack of comparative material was not of such consequence. In 2006 when larger area excavations were opened in a shrine, the Nyoo shrine (Insoll, Kankpeyeng, and Maclean 2007a, in press), this became a hindrance. For in utilising basic points of reference in classifying the ceramics it is useful to have comparative assemblages and thus to be able to begin to look to an element of regional standardisation with reference to vessel forms, 87

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA refers to as, “perforated vessel(s) or colander(s)” in both assemblages. Other potential points of comparison, as well as differences, exist, but further analysis of the Tongo Hills ceramic material is required before these can be more fully ascertained.

ii) Survey Factors Another reality of completing research in the region is the factors which intrude upon the intensity of survey coverage that can be achieved which is in part dictated by the nature of the terrain. The Tongo Hills are a small, horseshoe shaped chain of hills c.5km north-south by 3km east-west with a plateau at their centre in which the majority of people, the aforementioned Talis live. This population numbers a couple of thousand and is divided into seven sections, Bonchiog, Gundaat, Kpatari, Nanchieyir, Sakpee, Sameed, and Tamboog (Kankpeyeng 2005: 18). This would thus seem like a relatively circumscribed area which could be completely and intensively surveyed. Certainly, the plateau offers no exceptional natural barriers to completing such an intensive survey and large areas of the plateau including significant components of the Sakpee, Tamboog, Bonchiog Kpatari, and Gundaat sections have been surveyed (Insoll, Kankpeyeng, and MacLean 2004, 2005, 2007a). Yet human factors intrude. For example, some of the rock outcrops are used by adults as places for excretion (the children tend to use the fields). Hence a survey party has to be careful in not surprising someone in the act of defecation or in treading in the waste material.

It is true that it is possible to look further south into Ghana for comparisons, but if the OSL dates so far obtained from the excavations in the Nyoo shrine are correct, the ceramic assemblages required for comparative purposes are uncommon. The three dates obtained thus far as reported from the Oxford Research Laboratory for Archaeology and the History of Art are: • NYOO 06 (A) 5 (B5) (N10º40'30.9" W000º48'39.4"; X2845): OSL age (years) of 951 and an error (years) of 101, with a date range of AD 955-1155 and an age (years AD) of AD 1056. • NYOO 06 (B) 4 (D4) (X2846 - N10º40'32.5" W000º48'40.0"): OSL Age (years) of 1234 and an error (years) of 121, with a date range of AD 650-895 and an age (years AD) of AD 773. • NYOO 06 (B) 6-7 (D4) (X2847 - N10º40'32.7" W000º48'39.9"): OSL Age (years) of 1752 and an error (years) of 176, with a date range of AD 80-430 and an age (years AD) of AD 255.

More importantly, various areas of the plateau are ritually off-limits as they might be shrines or cemeteries. Numerous shrines of various sizes and configurations are found in the Tongo Hills, all of which, rightly, have to be treated with respect. However, this can mean in practical terms that gaps are left in the archaeological map if it is not possible to negotiate access to the shrine for surveying. Gundaat section, for instance, has only been surveyed in a preliminary way in the wet season and thus following paths through the millet fields (Insoll, Kankpeyeng, and MacLean 2005). Dry season survey when archaeological visibility is obviously better because of the absence of crops and other vegetation coverage was not possible in the one season (March-April 2008) when it was feasible because of a lack of teaching commitments to work in this season owing to the elders of this community, as is their right, deciding that they did not want research completed in their section. Hence such factors have to recognised, and built into what has to be a reactive research strategy which literally develops to suit the requirements of the community in the field.

Considering these dates, the material from the Daboya, Kadelso, and Chuluwasi sites on the Volta Rivers in the Northern and Brong Ahafo Regions might be useful for comparative purposes. This is because, besides their relative geographical proximity, Anquandah (1982: 72) describes the sites as “large Iron Age villages” well established by “the middle of the first millennium A.D.”. The use of red slip, decorated with either roulette or comb impression (ibid), all also present in the Tongo Hills, further suggests potential comparative possibilities. In particular, the ceramics from Daboya (Shinnie and Kense 1989; Stahl 1994: 79-81) seem to merit comparative analysis, and this will be initiated shortly. Looking beyond this geographical point, at, for example, Begho (e.g. Posnansky 1987; Stahl 1994: 84-87), Banda (e.g. Stahl 2001: 124-125, 174-175), or Bono Manso (Effah-Gyamfi 1985; Stahl 1994: 89-90) in the Brong Ahafo Region, the issue is raised as to how far one should seek comparative parallels, regardless of possible differences in chronology. Should parallels be sought as widely as possible by dint of both sites being from within the same national borders, i.e. Ghana?, when in fact closer parallels might exist north into Burkina Faso due to the obvious affiliation between the Tallensi, Mossi, and all the other members of what used to be called the Voltaic languages (Naden 1988, 1996). Excluding these broader considerations (and thus comparative possibilities will be explored), what is certain is that for ease of comparison having the published ceramics sequence from the Gambaga Escarpment would have been invaluable, and through its absence the research is placed at something of a disadvantage in being ‘pioneer’ when it need not be quite so ‘virginal’.

Similar but also different factors impinge when attempting to intensively survey the hills which surround the plateau. Here, terrain, masses of tumbled granite boulders, does impinge upon the survey. Literally, a head for heights is required to complete survey and the numerous clefts and rock shelters, many of which contain archaeological material, usually sherds from their probably being utilised as places of refuge during periods of slave raiding, all need to be investigated and recorded. For example, survey in 2008 in Tamboog section recorded a cave complex containing sherd material, Yenzee (Red Cave; N10°40.846’, W000°48.225’), which was also test excavated. This, logistically, was a difficult 88

TIMOTHY INSOLL: PLACING THE TONGO HILLS, NORTHERN GHANA, IN ARCHAEOLOGICAL TIME AND SPACE

Figure 2. View from the Yenzee Cave, Tamboog section, Tongo Hills (photo. T. Insoll).

piece of work, getting to the site (Figure 2), working inside it in a confined space, and also in recognising that it is just one of many such sites, some of which are inaccessible, again owing to terrain, ritual prohibition, or because of the presence of snakes. The latter is not an irrelevant consideration, and one of which direct experience was gained. Specifically, that in climbing up to a shrine, Gaaraug Naab, again in Tamboog section, a very large rock python, distended from having just eaten, was seen returning to its burrow (Figure 3). This was interpreted ritually, however, as a good omen as through its ascending the hill it was thus welcoming us to work in the shrine. Shrines are also located in the hills themselves as well as on the plateau as described, with the same attendant problems for archaeological survey. One of the most important of these hillside shrines is Yaane or Tongnaab. This is formed of a prominent rockshelter shielded by a grove of trees and is of especial economic importance in attracting numerous pilgrims as it is believed to be good in, for example, curing infertility and identifying witches (Allman and Parker 2005; Insoll 2006). The precincts of the shrine and of course the shrine itself are thus off limits for archaeological survey, though following negotiation the interior of the shrine has now been planned. Owing to the relationship between shrines and archaeological sites recurrently seen in the Tongo Hills, as at Yaane, where significant occupation evidence is almost certainly contained in the deposits in the cave, much archaeological material remains unrecorded because of its sacred status. The rockshelter which forms the Gobal shrine in the Kudoro rock outcrop in the

Figure 3. Rock python returning to its burrow, Gaaraug Naab, Tamboog section, Tongo Hills (photo. T. Insoll).

89

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA the Tongo Hills needs to emphasise the preliminary nature of such reconstructions where these are done, for as in much sub-Saharan African archaeology, this is from a baseline of little or no pre-existing archaeological data.

Tambiha area (N10°40’.40.8”, W000°49’.01.3”) was likewise placed out of bounds because of its ritual importance, but a view inside from beyond the immediate precincts of the shrine (and thus not transgressing its boundaries) clearly indicated it was of potential archaeological significance (Insoll, Kankpeyeng, and MacLean 2007a: 45). Whilst survey in the Nyoo shrine, subsequently confirmed by excavation, indicated that significant areas of this large site were formed of enshrined archaeological materials such as standing stones, stone arrangements, and deposits of deliberately interred ceramics (Insoll, Kankpeyeng, and MacLean 2007b; Insoll 2007b). This relationship between shrines and archaeological materials is not, of course, unique to the Tongo Hills. Willett (1967: 24) refers to it in Ife in Nigeria, and it is a relationship that has also recently been discussed in relation to sacred groves by Chouin (2008: 178-181). But it does, if access is not forthcoming (and again it has to be stated that such restrictions are understandable and must be respected) have implications for putative reconstructions of settlement patterns based upon survey data.

iii) Reactive Research Strategies This project started with the formulation of research aims and goals at the outset, as well as their revision and development annually when taking stock of the results achieved and in formulating proposals and reports for funding bodies for the next phase of research. This is standard procedure. However, what has been unusual is the extent to which the research has had to adapt in the field to meet understandable community requirements and also in reaction to wider circumstances. An example of the latter is provided by how one of the initial research aims of the project has had to be significantly curtailed, i.e. an archaeological investigation of the Mamprugu state upon its neighbours (Insoll, Kankpeyeng, and MacLean 2004). Partly this was due to an absence of comparative material, as already described, but also because work in one of the target areas Bawku/Garu, the subject of preliminary survey (ibid), proved impossible following the serious ethnic clashes in the area which were centred upon an ongoing chieftaincy dispute.

This in turn leads to a note of caution being added as to the validity of claims to have covered such and such an area by intensive survey in similar terrain in West Africa (and indeed, elsewhere in the continent). Admittedly, ritual prohibitions might not be of such consequence as they are in the Tongo Hills in most instances elsewhere, but other factors such as those of terrain, vegetation, local co-operation etc., surely intrude. Hence when these are not considered and an area of c.400km east-west and 300km north-south is discussed, with a putative three phase occupation sequence advanced, as by Elouga (1998) for western Cameroon, questions are surely allowed as to the extent the types of factors discussed could have influenced the settlement models proposed . And pertinent in the context of this research would seem to be the issue of where is the Late Stone Age occupation of the Tongo Hills. Quartz has been recorded, sometimes in large quantities as, seemingly, deliberately deposited in a mound at the base of the Hills at Zandoya. But this is in an Iron Age context and the only unequivocally worked quartz objects found so far are two ground quartz labrets or lip-plugs from the same site (the author is grateful to Derek Watson for their identification).

This forms one tier of circumstances which have had to be reacted to in the research process. A second tier is in the Tongo Hills itself and has led to a reactive research strategy being employed in the field on a day to day basis while completing excavation and survey. It is true that all archaeological research projects have to react to the material encountered for it is usually impossible to prejudge what might be encountered. But why this is being given especial consideration here in the context of discussing the realities of research is because of the degree of ritual prohibition and sacred contingency which have had to be negotiated and absorbed. This is far greater than this author has previously encountered where, for example, in investigating Islamic sites in Gao, Mali (Insoll 1996, 2000) or Bilad al-Qadim, Bahrain (Insoll 2005), the sole comparable prohibition (which was obviously respected) was upon disturbing Islamic burials (Insoll 1999). Otherwise, all aspects of the archaeological record, including sacred places such as mosques, could be investigated.

Elsewhere, the possible ‘worked’ quartz can only be tentatively described as such and consists of a small assemblage of quartz flakes excavated in 2005 from Hyena Cave in the Tongo Hills (Insoll, Kankpeyeng, and MacLean 2005; Eyifa 2007). These were recovered in a context without accompanying ceramics and remain undated (Eyifa 2007). Otherwise, any evidence, and the Hyena Cave material cannot be regarded as conclusive, for Late Stone Age occupation is lacking. Or is it, is it there but ephemeral, as yet unrecorded in one of the cave sites awaiting survey, or in an area which cannot be surveyed or excavated because of ritual prohibitions? The answer is not known but it is more certain that any attempt to conclusively date occupation based on survey data even in an area as geographically circumscribed as

In contrast, the enduring sacrality of some of the artefacts encountered in the Tongo Hills has been striking. An instance of this is provided by a group of ‘ritual’ pots which were uncovered in association with a cluster of standing stones excavated in the Nyoo shrine (Insoll, Kankpeyeng, and MacLean 2007a, in press). These pots, though empty, were interpreted as having an ancestral association, and thus following a visit to the site by community elders, they were not removed but were recorded in-situ, excluding two examples which had already been taken out for analysis. Hence the excavation strategy was altered to encompass the wishes of the community and in excavating the unit down to natural levels, a decision was taken to excavate in an area 90

TIMOTHY INSOLL: PLACING THE TONGO HILLS, NORTHERN GHANA, IN ARCHAEOLOGICAL TIME AND SPACE

Figure 4. Crouched adult inhumation, Touwang, Tamboog section, Tongo Hills (photo. T. Insoll).

encompassing neither the ritual pots or standing stones in the north-western part of the unit (D3 and D4). Similar immediate reactions within the research strategy and methods employed have been necessary on various occasions, and one other example from the 2008 season will suffice. Survey in Tamboog section recorded an area of abandoned settlements, Touwang (N10°40.949’, W000° 48.235’). One of these was formed of a slight mound with bone eroding out of the surface. As faunal remains have been lacking from the excavations and surveys thus far, and as it was presumed these were cattle bones by foreign archaeologists and locals alike, a unit was put in on this site. However, the faunal remains ultimately translated into a crouched adult human inhumation (Figure 4). This was of significance because of the association between human remains and the ‘forefathers’ (ancestors). Hence the skeleton was cleaned, photographed, and recorded in-situ, but not removed or otherwise disturbed. Moreover, access restrictions to the site were maintained at the behest of the local community members employed as part of our team both from Tamboog and other sections of Tengzug, so as not to lead to gossip and misunderstanding as to what was occurring. Finally, following the recovering of the skeleton under earth and a cairn of rock, a fowl was provided by the research team and an egg and pito (millet beer) provided by our local team members so that the appropriate rites to placate the spirit of the individual buried could be completed in our absence.

Table 1. Total of sacrifices provided.

and thus their cost. This could be interpreted as the shrines being aware that as the results increase so do the inherent sacrificial costs of completing the research. Furthermore, the results of the divination and sacrifice are far from predictable. They might allow research to take place or they might not, and this has proven a major component of the reactive research strategy. The 2008 season again provides a good example. Prior to the start of this season, as with every year, the process of community and shrine negotiation was started with two specific research objectives in mind; completing excavations in a specific compound, the Goldaana House; and in an area where various abandoned shrines and compounds had been recorded, Gundaat (Insoll, Kankpeyeng, and MacLean 2007a).

This process of placation and consultation with the Gods, the Earth, and the Ancestors, via the agency of the shrines has been a continual theme of the research process. Following divination to ascertain what sacrifices the shrine requires, the appropriate animals have been provided in advance of each stage of the research. To date the following sacrifices have been made (Table 1). What is evident from looking at this table is that as the work has progressed there has been an upward trajectory in the number of animals required, their precise colour, 91

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA Following this process of consultation neither of these objectives was achievable for the funeral rites of the late Goldaan had not been completed and hence he had not yet achieved ancestral status, and secondly, as already described, the Gundaat elders did not want any archaeological work completed in their section. These negotiations and consultations were also taking place in the ritually ‘heightened’ atmosphere of the run-up to the Gologo pre-agricultural season festival. Thus following a brief period of shock by this author that the season’s research objectives needed to be completely altered, a new set were quickly formulated which ultimately proved more useful in addressing the first of the research questions outlined at the start of this paper, i.e. the reconstruction of the occupation sequence in the Tongo Hills. Hence the community and the shrines had steered the research in the direction of this, and through the excavations completed in Zandoya at the base of the hills near Yinduri, and at Touwang in Tamboog, several sequences of dateable material were recovered.

References ALLMAN, J. and PARKER, J. 2005. Tongnaab. Bloomington: Indiana University Press. ANQUANDAH, J. 1982. Rediscovering Ghana’s Past. Harlow: Longman. ANQUANDAH, J. 1998. Koma – Bulsa. Its Art and Archaeology. Rome: Istituto Italiano per l’Africa e l’Oriente. APENTIIK, R. 1997. Bulsa Technologies and Systems of Thought. Unpublished MA Thesis, Department of Archaeology, University of Calgary. ASHLEY, C. 2006. Unpublished Ceramics Field Notes from Tengzug, 2006. Department of Archaeology, University of Manchester. CASEY, J. 2001. The Kintampo Complex. Oxford: British Archaeological Reports.

Conclusions CHOUIN, G. 2008. Archaeological Perspectives on Sacred Groves in Ghana. (In), Sheridan, M. J. and Nyamweru, C. (eds.), African Sacred Groves. Oxford: James Currey, pp. 178-194.

This contribution has been candid about the research process, and it is probable that everyone encounters such challenges during the completion of their archaeological research but most, necessarily, remains unpublished. Even Meyer Fortes provides an insight into his worries in this respect when scribbled in one of his notebooks, he states, “I look at the date, and despair strikes me…….And what of all the elaborate, (unclear word) plans and schemes, and ‘revolutionary ideas’? So far none have reached fruition in research. We don’t even know the language well enough to put them into fruition”– (Fortes 1934: 33). However, within the context of editing and compiling a volume on current archaeological research in Ghana it was decided to present a contribution outlining something of a foreigner’s experience of this. Though by way of comparison with the problems Ghanaian colleagues have experienced in, for example the lack of funds, equipment etc., in attempting to complete their research such factors are, in the sum total of things, minor, and are acknowledged as such.

DAVIES, O. 1966. West Africa Before the Europeans. London: Methuen. DAVIES, O. 1970. Ghana Field Notes. Part 2. Northern Ghana. Unpublished Report, Dept. of Archaeology, University of Ghana, Legon. DE BARROS, P. 1990. Changing Paradigms, Goals and Methods in the Archaeology of Francophone West Africa. (In), Robertshaw, P. (ed.), A History of African Archaeology. London: James Currey, pp. 155-172. DRUCKER BROWN, S. 1975. Ritual Aspects of the Mamprusi Kingship. Cambridge: African Studies Centre. EFFAH-GYAMFI, K. 1985. Bono Manso: An Archaeological Investigation into Early Akan Urbanism. Calgary: University of Calgary Press. ELOUGA, M. 1998. Recherches Archéologiques au Cameroun Méridonal: Résultats des Prospections et Hypothèses sur les Phases de Peuplement. (In), Delneuf, M., Essomba, J-M., and Froment, A. (eds), Paléoanthropologie en Afrique Centrale. Paris: L’Harmattan, pp. 213-224.

Acknowledgements I am grateful to the British Academy (2004-2008) and the Wellcome Trust (2008-) for funding the research and fieldwork. I am also grateful to Rachel MacLean and Benjamin Kankpeyeng for commenting on this paper, though as usual it should be stated that I am responsible for all errors, omissions, and opinions stated herein. Finally, I express my sincere gratitude to the people of the Tongo Hills and the Ghana Museums and Monuments Board for allowing the research to proceed.

EYIFA, G.A.M. 2007. House Construction and the Role of Women in Tongo-Tengzug: An Ethnoarchaeological Study. Unpublished MPhil Thesis, Department of Archaeology, University of Ghana, Legon. FORTES, M. 1934. Unpublished Diary. Vol. 2, 24th May 1934. Manuscripts Room, University Library, University of Cambridge. 92

TIMOTHY INSOLL: PLACING THE TONGO HILLS, NORTHERN GHANA, IN ARCHAEOLOGICAL TIME AND SPACE INSOLL, T. 2007c. Archaeology. The Conceptual Challenge. London: Duckworth.

FORTES, M. 1936. Unpublished Notebook 12 (12/36). Manuscripts Room, University Library, University of Cambridge.

INSOLL, T. In Press. Negotiating the Archaeology of Destiny. An Exploration of Interpretive Possibilities through Tallensi Shrines. Journal of Social Archaeology.

FORTES, M. 1945 (1969). The Dynamics of Clanship among the Tallensi. Oosterhout: Anthropological Publications.

INSOLL, T. In Preparation b. Meyer Fortes, Sex and Material Culture. The Published Image and the Unpublished Resource.

FORTES, M. 1949 (1967). The Web of Kingship among the Tallensi. Oosterhout: Anthropological Publications.

INSOLL, T., KANKPEYENG, B., and MACLEAN, R. 2004. An Archaeological Reconnaissance in the Tong Hills, and Garu Area, Upper East Region, and Nakpanduri, Northern Region, Ghana. Nyame Akuma 62: 25-33.

FORTES, M. 1983. Oedipus and Job in West African Religion. Cambridge: Cambridge University Press. FORTES, M. 1987. Religion, Morality and the Person. Essays on Tallensi Religion. Cambridge: Cambridge University Press.

INSOLL, T., KANKPEYENG, B., and MACLEAN, R. 2005. Excavations and Surveys in the Tongo Hills, Upper East Region, Ghana. July 2005. Fieldwork Report. Nyame Akuma 64: 16-23.

GABRILOPOULOS, N. 1995. Ethnoarchaeology of the Tallensi Compound. Unpublished MA Thesis, University of Calgary. GARRARD, T. 1986. Brass-casting among the Frafra of Northern Ghana. Unpublished PhD Dissertation, Department of History, UCLA.

INSOLL, T., KANKPEYENG, B., and MACLEAN, R. 2007a. Excavations and Surveys in the Tongo Hills, Upper East Region, Ghana. July 2006. A Preliminary Fieldwork Report. Nyame Akuma 67: 44-59.

HODDER, I. 1999. The Archaeological Process. Oxford: Blackwell.

INSOLL, T., KANKPEYENG, B., and MACLEAN, R. 2007b. Shrines, Rituals, and Archaeology. Current World Archaeology 26: 29-36.

INSOLL, T. 1996. Islam, Archaeology and History. Gao Region (Mali) Ca.AD 900-1250. Cambridge Monographs in African Archaeology 39. BAR S647. Oxford: BAR Publishing.

INSOLL, T., KANKPEYENG, B., and MACLEAN, R. In Press. The Archaeology of Shrines among the Tallensi of Northern Ghana: Materiality and Interpretative Relevance. (In), Dawson, A. and Shinnie, P. (eds), Shrines in African Societies. Calgary: University of Calgary Press.

INSOLL, T. 1999. The Archaeology of Islam. Oxford: Blackwell. INSOLL, T (with other contributions). 2000. Urbanism, Archaeology and Trade. Further Observations on the Gao Region (Mali). The 1996 Fieldseason Results. British Archaelogical Reports S829. Oxford: BAR Publishing.

KANKPEYENG, B. 2001. The Cultural Landscape of Tongo-Tengzuk. Traditional Conservation Practices. (In), Anon, (ed.), Traditional Conservation Practices in Africa, ed Anon. Nantes: CRATerre-EAG, pp. 23-31.

INSOLL, T. 2005. The Land of Enki in the Islamic Era. London: Kegan Paul.

KANKPEYENG, B. 2005. The Cultural Landscape of Tongo-Tenzug. (In), Joffroy, T. (ed.), Traditional Conservation Practices in Africa. Rome: ICCROM, pp. 15-21.

INSOLL, T. 2006. Shrine Franchising and the Neolithic in the British Isles: Some Observations based upon the Tallensi, Northern Ghana. Cambridge Archaeological Journal 16: 223-38.

KANKPEYENG. B. W. and DECORSE, C. R. 2004. Ghana’s Vanishing Past: Development, Antiquities, and the Destruction of the Archaeological Record. African Archaeological Review 21: 89-128.

INSOLL, T. 2007a. Natural or Human Spaces? Tallensi Sacred Groves and Shrines and their Potential Implications for Aspects of Northern Phenomenological European Prehistory and Interpretation. Norwegian Archaeological Review 40: 138-58.

KENSE, F. J. 1990. Archaeology in Anglophone West Africa. (In), Robertshaw, P. (ed.), A History of African Archaeology. London: James Currey, pp. 135-154.

INSOLL, T. 2007b. 'Totems', 'Ancestors', and 'Animism'. The Archaeology of Ritual, Shrines, and Sacrifice amongst the Tallensi of Northern Ghana. (In), Barrowclough, D. and Malone, C. (eds.), Cult in Context. Oxford: Oxbow Books, pp. 326-333.

KENSE, F. J. 1992. Settlement and Livelihood in Mampurugu, Northern Ghana: Some Archaeological Reflections. (In), Sterner, J. and David, N. (eds.), An African Commitment. Calgary: University of Calgary Press, pp. 143-155. 93

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA KIRBY, J. 2006-2007. Ethnic Conflicts and Democratization. New Paths toward Equilibrium in Northern Ghana. Transactions of the Historical Society of Ghana 10: 65-107. MADIEGA, G.Y. and NAO, O. (eds.). 2003 Burkina Faso. Cent Ans d’Histoire, 1895-1995. Paris: Karthala. MATHERS, C. 1999. An Ethnoarchaeology of Kusasi Shrines, Upper East Region, Ghana. Unpublished PhD Thesis, University of Calgary. MATHERS, C. 2003. Shrines and the Domestication of Landscape. Journal of Anthropological Research 59: 2345. NADEN, T. 1988. The Gur Languages. (In), KroppDakubu, M.E. (ed.), The Languages of Ghana. London: Kegan Paul, pp. 12-49. NADEN, T. 1996. The Oti/Volta Languages. (In), Naden, T. (ed.), Time and the Calendar. Legon: Institute of African Studies, University of Ghana, pp. 4-17. POLE, L. M. 1985. Furnace Design and the Smelting Operation: A Survey of Written Reports of Iron Smelting in West Africa. (In), Haaland, R. and Shinnie, P. (eds.), African Iron Working – Ancient and Traditional. Oslo: Norwegian University Press, pp. 142-163. POSNANSKY, M. 1987. Prelude to Akan Civilisation. (In), Schildkrout, E. (ed.), The Golden Stool: Studies of the Asante Center and Periphery. New York: American Museum of Natural History, pp. 14-22. PRIDDY, B. 1974. Pottery in Upper Region. National Museum Occasional Papers 6: 41-53. SHINNIE, P. and KENSE, F. J. 1989. Archaeology of Gonja, Ghana. Excavations at Daboya. Calgary: University of Calgary Press. SMITH, F.T. 1989. Earth, Vessels, and Harmony among the Gurensi. African Arts 22 (2): 60-65 + 103. STAHL, A. B. 1994. Innovation, Diffusion, and Culture Contact: The Holocene Archaeology of Ghana. Journal of World Prehistory 8: 51-112. STAHL, A. B. 2001. Making History in Banda. Cambridge: Cambridge University Press. VANSINA, J. 1985. Oral Tradition as History. London: James Currey. WILLETT, F. 1967. Ife in the History of West African Sculpture. London: Thames and Hudson.

94

Rethinking the Stone Circles of Komaland. A Preliminary Report on the 2007/2008 Fieldwork at Yikpabongo, Northern Region, Ghana Benjamin W. Kankpeyeng and Samuel Nilirmi Nkumbaan Dept. of Archaeology, University of Ghana, Legon, P.O. Box LG3, Legon, Ghana

After a series of meetings between Ghanaian archaeologists, social anthropologists, opinion leaders of Komaland and museum professionals, convened by the National Commission on Culture, Ghana, in 2005 and 2006, on the way forward for better understanding of the abandoned cultural traditions, archaeological survey and excavation of the stone mound complexes resumed as a collaborative project between the Department of Archaeology, University of Ghana, and the Ghana Museums and Monuments Board. These exploratory archaeological missions assessed the state of the mounds and also conducted preliminary excavations at Yikpabongu and Tando in the Northern Region. Details of data from the field seasons are being analyzed. However, observations during the excavation process of and around a small stone mound at Yikpabongo (Figure 1) and cursory eye examinations of some artifacts provide additional insights or new perspectives on the mounds, and this will be briefly discussed here.

Introduction In 1985 the discovery of a “forgotten civilization” in Komaland, northern Ghana was announced to the international community (Anquandah and van Ham 1985). Several exhibitions on Koma terracotta figurines were mounted in both private and public galleries accompanied by catalogues in Europe and Canada (Dagan 1989). Figurines on display at these exhibitions were looted pieces from vandalized mounds or sites by local peoples instigated by antique dealers (Brent 1996, ICOM 2000). Looting at some of the sites predated the archaeological survey of the archaeological region in 1985 (Anquandah 1998). Public educational programmes by staff of the Ghana Museums and Monuments Board, public officials of the district councils of the area, and educated opinion leaders from the communities scaled down drastically the looting or disturbance of the mounds (Kankpeyeng and DeCorse 2004: 104-105).

Figure 1. The Location of Yikpabongo in Northern Ghana.

95

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA Previous Research

Identification and Recording of the Mounds

Preliminary research was conducted at the Koma sites in 1985. The 1985 archaeological survey identified and recorded the distribution of the sites over a large area (Anqandah and van Ham 1985). The sites covered areas in three northern regions (Upper East, Upper West, and Northern Regions), distributed in a territory measuring 90 x 100 miles. The majority of the sites were located in the West Mamprusi district in the Northern Region. Test excavations, relying on the quadrant method, at Yikpabongo recovered human and faunal remains and hundreds of terracotta figurines (Anquandah 1987, 1998, 2003; Anquandah and van Ham 1985). The artefact assemblage included local ceramic sherds, grinding stones, metal bracelets and tools, and trade items such as cowries and beads. The terracotta pieces included anthropomorphic and zoomorphic figurines: stylized and naturalized. The human form figurines included full size, single heads, two-faced heads and four-faced heads. Others depicted humans riding camels and horses, female adults breastfeeding children or grinding grain, slaves bound with chains, and individuals with circumcised male genitalia. The zoomorphic figurines included wild and domestic species. The wild species represented included lion, rhinoceros, hippopotamus, buffalo, snake, crocodile, monkey, tortoise, and birds. The domestic dog is depicted. Other figurines depicted objects such as stools, chairs, amulets, and musical instruments. Thermoluminescence (TL) dating of sampled figurines provided dates ranging from 1200-1800 AD.

i) Excavations Test excavations were conducted at Yikpabongu and Tando (Figure 1). Based on surface scatters of ceramics, destroyed figurines, the two sites were sampled for comparative purposes. The surface scatters showed different figurine forms from that of Yikpabongu. The dryness and compact nature of the soil, the large quantities and distribution of stone, local ceramics, and figurines, made the use of trowels very difficult. Using trowels would have abraded the surfaces of ceramics and figurines. Instead the units were watered daily to enable the careful recoverery of the materials in their real contexts. In some situations small hand picks and dental picks had to be used to loosen soil before the recovery of any arrangements and removal of artifacts. A grid system was used to excavate portions of mounds which were judgementally sampled. The mounds were selected because of their small size and location. The choice of small mounds was necessitated by little funds available for the project and also the two weeks available for the fieldwork. Furthermore, the mounds which were sampled were chosen deliberately as they were located within the communities. For example, at Yikpabongo the mound was very close to the house of the Assemblyman, the local political representative at the district level. This choice was intended to allow the people to understand the scientific approach to the study, as against the indiscriminate looting strategies conducted by the antique hunters. The use of the grid system allowed the recovering and recording of the distribution of artefacts and features separately, and thereby facilitated our understanding of the distribution of material culture within the mounds both vertically and horizontally.

The mounds were described as burial mounds because of their association with human remains. The cultural tradition was also identified as a “Forgotten Civilization” (Anquandah and van Ham 1985) because the oral traditions of present ethnolinguistic groups within the archaeological region and other parts of northern Ghana do not identify the producers of the ‘cultural complex’. Spirituality was identified associated with the stylized designed pieces. Anquandah (1987, 1998) suggested that long distance trading created a lot of wealth that stimulated the evolution of socio-political complexity in the area with the mounds and artefact assemblage as the markers.

Two units were excavated at Yikpabongo (Figures 2a2d). The first unit was undertaken to provide material for an understanding of the chronology. As such it was placed between a house mound and a stone circle mound. The second unit covered a stone circle mound located near the house of the Assemblyman of Yikpabongo. This second unit was intended to provide an understanding of the distributions and arrangements of material culture within the stone circle mounds. In 2007, a grid of 4m x 5m was established over the south-eastern portion of the stone mound. A total of 20 units covered by this grid were excavated, reaching sterile level at different depths.

Current Research: 2007 and 2008 Field Seasons Archaeological survey and excavations started in 2006 with a tour of sites to evaluate their potential to provide scientific information. The research followed the series of meetings convened by the National Commission on Culture previously described. Additional fieldwork was conducted at Yikpabongo in 2007 and 2008. The ongoing research at Yikpabongo is intended to provide an understanding of the distribution of material culture within the stone circle mounds and to facilitate the establishment of a chronology for the site.

Ceramic sherd samples and carbon samples were collected for dating purposes in Denmark. Unfortunately, only one sample was sent for TL dating because of limited funding Charcoal samples were also collected for radiocarbon dating. The samples are yet to be sent to laboratories.

96

BENJAMIN W. KANKPEYENG AND SAMUEL NILIRMI NKUMBAAN: RETHINKING THE STONE CIRCLES OF KOMALAND

Figure 2a. Material arrangements exposed during excavations at Yikpabongo.

Figure 2c. Material arrangements exposed during excavations at Yikpabongo.

Figure 2b. Material arrangements exposed during excavations at Yikpabongo.

Figure 2d. Material arrangements exposed during excavations at Yikpabongo.

ii) The Material Recovered

The terracotta figurines collected consist of similar types to those recovered during the 1985 preliminary survey. They included double-faced, double-headed, four-faced, hat-shaped with double-faced human faces at the top, phallic, single heads, others well adorned with ornamental motifs, crocodile-like, bird-like with human faces, as well as clay ritual objects (Figures 3-6). Two complete ritual bowls associated with ancestral veneration were also recovered, whilst others are being pieced together. However, we recovered many bird-like forms with human faces. The different types of figurines were recovered and recorded from different squares, and they may provide insight into different types of rituals.

The material culture assemblage recovered consisted of glass fragments, plenty of local ceramic sherds, terracotta figurines (Figures 3-6), ceramic bowls (Figure 7), fired clay ritual objects related to ancestral veneration, metals, and a human skull. Broken glass fragments dating from the second half of the twentieth century were also recovered at surface level, and in contexts less than 10cm deep. Samples of local ceramic sherds from each level of each unit were taken and are being analyzed in Accra. A large size bag of sherds was taken from each level. The rest were stored at Yikpabongo for future analysis. 97

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

Figure 5, Figurine associated with witchcraft exorcising.

Figure 3: Residues on some figurines

Figure 4a. Ritual conically-shaped objects

Figure 6a. Example of figurine recovered.

Figure 4b. Ritual conically-shaped objects

Figure 6b. Example of figurine recovered.

98

BENJAMIN W. KANKPEYENG AND SAMUEL NILIRMI NKUMBAAN: RETHINKING THE STONE CIRCLES OF KOMALAND powerful substances and by proxy powerful technologies (iron working). There is a likely association of the preeminence of iron ore as a material with the pre-eminence of eyes as part of human beings. Also, a smelter usually, “orchestrates the full panoply of rituals” associated with traditional medical care, sacrifices, and prescriptive behaviour (Herbert 1993: 116). Discussion and Conclusions The general material culture assemblage from the excavations consisted of local ceramic sherds or potsherds, complete pots, quartz querns, muscovitebiotite schist grinders, and fired clay figurines. In some cases, small ceramic gourds and iron objects with knots and points were recovered in association with the other materials listed above. Broadly, two types of material arrangements were observed during the excavations; arrangements within the mound and those immediately around the mound. Arrangements within the mound were recovered between 20cm to 60cm depth whilst those from outside the mound were recovered from between 80cm to 160cm in depth.

Figure 6c. Example of figurine recovered.

Above each arrangement was a 20cm layer of lots of ceramic sherds mixed with stones. Beneath this layer were spreads of red-slipped local ceramics (see Figures 2a to 2d). Upon removal of these red-slipped local ceramics, a layer containing the special arrangements of the various material culture objects was observed. It was almost predictable in that the recovery of one type of object would result in the recovery of the other material associations. For example, anytime a rounded stone quern is encountered, there was always the possibility of finding a large potsherd around it with an associated figurine. An unusual or unpleasant scent was associated with such layers. In the case of arrangements outside the stone mounds, bigger figurines were found with complete small ritual pots to the south of them. There were repeated arrangements of this kind down to a depth of 160cm. Reddish-yellowish soil was located around some of the arrangements outside the mounds. As already noted, these await dating.

Figure 6d. Example of figurine recovered.

The most fascinating finds were located above the area where the human skull was recovered. The bird, the fourfaced headed and hat-like, and well adorned human forms were recovered within and around the square where the skull was encountered. The distribution of the figurines will be clearer when analysis is completed.

The material arrangements were intentionally created by past Komaland populations, and the different types of arrangements may relate to different rituals or purposes. The closest ethnographic analogs in northern Ghana may be those of the Lobi along the Black Volta River in northwestern Ghana. Jack Goody’s discussion of such shrines may be of relevance to future interpretations of the material culture in the stone circle mounds of Komaland (Goody 1972: 14-33).

Iron ore was used to constitute the eye balls of the figurines. Quite accidentally a mattock hit a figurine by the eye and exposed the iron ore. The work men hired added that looted figurines also have iron ore as the eye balls. Though the presence of the iron ore on the figurines may be merely decorative, the importance of iron working and the belief and rituals associated with it (Bisson et al 2000; Haaland and Shinnie 1985; Herbert 1993), makes their use in these contexts very potent and of great significance. It probably serves as more than aesthetics, but is indicative of the contextualization of

Figurines recovered from 100cm to 160cm were observed to have oily and slippery layers on them with a touch of a finger. It was also observed that some of these figurines had reddish to brownish pigmentation on them. Laboratory tests are yet to be conducted to establish the type of substances or liquids that may have been applied on the figurines, but initial visual examination by eye, points to the possible use of blood. 99

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

Figure 7d. Complete ritual local ceramic vessel recovered from excavation at Yikpabongo.

Figure 7a. Complete ritual local ceramic vessel recovered from excavation at Yikpabongo.

The presence of the skull and the manner of its deposition, the distribution of the materials, and the type of figurines recovered during the 2007 field season, as well as in the 1985 season provide a lot of insights and potential scenarios. The burial of a skull alone without any other or traces of remains or parts of the body is an unusual occurrence today in northern Ghana; but may be a ritual activity of a group not living in northern Ghana today or an abandoned tradition for several reasons, including migrations. Also, parallel ritual activities by some groups within West Africa may provide clues as to its meaning. For example, among the Koma people in the south eastern portion of Gongola State region of Nigeria, the skull of dead men are removed from the grave after all funerary rituals have been performed and deposited in family shrines for future consultations (Alfonso 1984: 69). Usually the skull is put into a pot for preservation. The occasion of the removal of the dead man’s skull is an important event which usually takes place two months after the burial in grave. The removal of the skull helps the community to assess whether the deceased was a witch or not based on the belief that a wizard must have some tiny holes in the skull. The skull once preserved is used for future consultations marked by different activities which are “among the major customs and traditions of the Koma” (Alfonso 1984: 70).

Figure 7b. Complete ritual local ceramic vessel recovered from excavation at Yikpabongo.

Similarly, amongst the Yungur of Nigeria the skulls are removed after a year, washed and kept at special shrines reserved for the purpose (Berns 1990: 57). The skulls of deceased individuals of advanced age, as well as chiefs are removed from graves (in the form of pot burials) and preserved, often in caves. The skulls are not returned to the graves but kept at the shrines for future ancestral veneration rituals. The skulls are kept in special pots made by women before they are deposited in the special shrines. There is a special relationship between pots and ancestral spirits among the Yungur and also among the Mafa and Bulahay, and Chadic speakers in northern Cameroun (David, Sterner, and Gavua 1988: 365). These sculptural ancestral pots are related to the structuring of

Figure 7c. Complete ritual local ceramic vessel recovered from excavation at Yikpabongo.

100

BENJAMIN W. KANKPEYENG AND SAMUEL NILIRMI NKUMBAAN: RETHINKING THE STONE CIRCLES OF KOMALAND material assemblages. This strategy is what is planned for the future.

the communities and their social relations. Such examples may help explain the presence of the abundance of ceramics in the stone mounds of Northern Ghana, because of their highly charged symbolic meanings (Berns 1990; David, Sterner and Gavua 1988). Some figurines recovered at Yikpabongo are modelled on top of ceramics that look like lids or hats further suggesting their symbolic and ritual importance (see Figure 6a).

Acknowledgements Funding for the fieldwork and dating of the samples was provided under the Ghana-Denmark Archaeolgical (GDARCH) project provided by DANIDA. A visiting fellowship from The British Academy tenable at the School of Histories, Newcastle Unversity facilitated the writing-up of the draft of this paper by Dr. Benjamin Kankpeyeng. He is very grateful to Dr. Jane Webster for facilitating the visiting fellowship. Dr. Kankpeyeng also thanks Samuel Nkumbaan who co-directed the project. He is also very grateful to Mr. Ben Baluri Shaibu of Yikpabongu, Raymond Agbo and R. H. Boateng of the GMMB, and Bossman Murey of the Department of Archaeology, University of Ghana for their support in diverse ways during the research. Thanks also to Professor Timothy Insoll and Christopher R. DeCorse.

Also, the mounds and their contents may have served many purposes in the past perhaps as curative, witchcraft exorcising, and fertility centres. These roles are potentially represented by some of the different figurines. Some other West African examples perhaps serve as examples. In the Inland Niger Delta region of Mali the appliqués on the crocodile-like or other amphibians are indicative of ritual healing of diseases. The appliqué motifs signal the disease transfer from the human being to the animal. Also among groups in northern Ghana witches are considered to transform themselves into birds in the night. Exorcising witchcraft involves the removal of clothes and their deposition on shrines. Clothes do not survive in the archaeological record in the tropics because of the high humidity and acidic soil levels. Therefore, they are archaeologically irretrievable, but some materials or artefacts may indicate witchcraft. The bird-figurine with human faces may have something to do with witchcraft exorcising (Figure 5). Also in the case of the Koma of Gondo State, Nigeria, the reason for removal of the skull after two months of burial is to check whether the deceased was a witch or not. If the deceased was a witch the skull would usually have holes on it. After verification, different rituals are performed. In the same way, the phallic objects and those figurines that show genitalia may relate to fertility healing rituals. In many areas of West Africa fertility healing is today represented by fertility wooden dolls.

References ANQUANDAH, J. 1987. Stone Circle Sites of Komaland, Northern Ghana, in West African Archaeology. The African Archaeological Review 5: 171-180. ANQUANDAH, J. 1998. Koma - Bulsa. Its Art and Archaeology. Rome: Istituto Italiano Per L’Africa e L’Oriente. ANQUANDAH, J. 2001. The Arts of Koma-Bulsa. Ghana Yesterday and Today. Paris: Musee Dapper, pp 135-149.

Moreover, the presence of ritual objects, as in Figure 4, is perhaps related to personal protective deities or fertility rituals. The holes in them facilitated the pouring of liquid offerings. Such deities are deposited in shrines after performance of rituals by the heir to the deceased. Generic ones were recovered in shrines at Tengzug (Insoll 2007: 331; Insoll, Kankpeyeng, and Maclean In Press.) and the Krobo Hills (Gblerkpor pers. comm.., Gblerkpor this volume) in the Upper East and Eastern Regions of Ghana, respectively. Also generic clay objects associated with fertility rituals are common among the Batammaliba of Togo (Preston Blier 1987: 119).

ANQUANDAH, J. and VAN HAM, L.1985. Discovering the Forgotten Civilization of Komaland, Northern Ghana. Rotterdam: Ghames Foundation. BERNS, M.C. 1990. Pots as People. Yungur Ancestral Portraits. African Arts 23 (3): 50-60. BISSON, M.S., CHILDS, S. T., DEBARROS, P.D., and HOLL, A. 2000. Ancient African Metallurgy. The SocioCultural Context. Walnut Creek: AltaMira. BRENT, M. 1996. A View Inside the Illicit Trade in African Antiquities. (In), Schmidt, P. R. and McIntosh, R. J. (eds.), Plundering Africa’s Past. Oxford: James Currey, pp. 63-78.

Based on the distribution of material culture (figurines, ritual objects, and ceramic bowls) and the orientation of the skull with the ceramic discs, and the very limited research conducted to date we can only hypothesize at this stage that the mounds functioned as shrines in the past. The proximity of the mounds to house mounds may also suggest the shrines were associated with ancestral worship and beliefs, probably at family levels. Future work will entail the extension of the grids to cover the entire mound and thus subsequent excavation to throw more light on the type of mounds and their associated

DAGAN, E. A. 1989. Spirits without boundaries. Twenty-six Single Heads from Komaland, Ghana. Montreal: Galerie Amrad African Arts. DAVID, N., STERNER, J. and GAVUA, K 1988. Why Pots are Decorated. Current Anthropology 29: 365-89. 101

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA GOODY, J. 1972, The Myth of the Bagre, Oxford: Oxford University Press. HAALAND, R and SHINNIE, P. (eds.), 1985. African Iron Working. Oslo: Norwegian Universities Press. HERBERT, E, 1993. Iron, Gender, and Power. Bloomington: Indiana University Press. ICOM 2000. Terracotta from Northern Ghana (Komaland and the Cote d’Ivoire). ICOM Red List. http://www.icom.org/redlist/english/page07.htm. INSOLL, T., 2007. ‘Totems’, ‘Ancestors’, and ‘Animism’. The Archaeology of Ritual, Shrines, and Sacrifice Amongst the Tallensi of Northern Ghana. (In), Malone, C. and Barrowclough, D. (eds.), Cult in Context. Oxford: Oxbow Books, pp. 326-333. KANKPEYENG, B. W. and DECORSE, C. R. 2004. Ghana’s Vanishing Past: Development, Antiquities, and the Destruction of the Archaeological Record. African Archaeological Review 21: 89-128. PRESTON BLIER, S. 1987. The Anatomy of Architecture. Ontology and Metaphor in Batammaliba Architectural Expression. Cambridge: Cambridge University Press.

102

The Archaeology of Slavery: A Study of Kasana, Upper West Region, Ghana Samuel Nilirmi Nkumbaan Department of Archaeology, University of Ghana, Legon, P.O. Box LG3, Legon, Ghana

Hence the activities of the slave raiders and traders in northern Ghana have thus far received little attention.

Introduction Until recently, archaeology in Northern Ghana had been neglected. With the exception of, for example, research in the area by Anquandah and Van Ham (1985) in Komaland, Gavua (1985) at Daboya, and Okoro (2003a, b, c) at Salaga and (2001, 2002) at Sakpuli, and a few expatriate scholars like Davies (1963, 1970, 1973, 1980), Goody (1954), Kense (1981, 1987), Shinnie and Ozanne (1962) and Insoll, Kankpeyeng, and Maclean (2004, 2005, 2007), little is known of the archaeology of Northern Ghana in comparison to other areas of Ghana.

Although some histories have been written of northern Ghana that relate to the slave trading activities and wars, e.g. Howell (1998), Wilks (1961), Der (1975, 1983, and 2000), the primary focus of some of these writings has not been on the issue of slavery. Howell for instance, tried to relate the issues of slavery to the Bible so as to get the people converted to Christianity. Similarly, Wilks’ (1961) concern was with the development of Ashanti history resulting from the northern influence and not the slave trade per se in northern Ghana. However, the works of Der (1975 and 1998) on Samori’s relations with North West Ghana and the slave trade in northern Ghana respectively, are important reference sources for this research.

The research described here forms part of the International Slave Route Project initiated by UNESCO. The UNESCO project assesses the place of Ghana in the slave trade in relation to other parts of Africa and the world at large. This has been done by identifying evidence for Ghana’s involvement in the slave trade, it’s aim being to trace the sources of slaves and the routes, as well as the temporary camps through which and from where the slaves got to the coast for final transportation across the Atlantic. Another project aim is to examine the effects of the trade on the people involved. The necessity for this kind of research has been because of the increasing awareness that the slave trade had profound effects on the people involved and there is still a feeling of hurt in many areas.

More generally, Perbi (1996, 19`97), Miers and Kopytoff (1977), Kelly (1997a and b) and others make references to the slave trade both in Africa and Ghana. These are useful references in as far as they reflect the variety of activities involved. However, as noted, there has been almost no archaeological study of the slave trade in Northern Ghana. Hence there was a need for archaeological research to augment the historical accounts on slave trading activities. The research thus had the following aims. (i) Collecting oral accounts to ascertain what the inhabitants of Kasana knew of their involvement in the slave trade activities of the 19th century. (ii) Conducting archaeological investigations at selected sites in this area to collect material evidence that might reflect the slave trading activities. (iii) Assessing the effects of the slave trading and raiding activities on Kasana and its inhabitants.

Furthermore, there has been increasing publicity in the Ghanaian news media (especially with reference to the yearly Pan-African Festival [PANAFEST] in Ghana) as to the impact of the slave trade on the country. Southern chiefs have poured libations to their ancestors to atone for the part their ancestors played in the slave trade and most significantly, increasing numbers of African-Americans are visiting Ghana to try and learn something about their identity. This helps explain why UNESCO has initiated a research project to reopen slave routes and redevelop slave markets for the purposes of tourism (Howell 1998: v).

To achieve these aims, a multi-disciplinary approach was utilized. The Study Area

Kasana in the Upper West Region of Ghana (Figure 1), is hardly mentioned in the history of Ghana (Der 2000), but it was a place that had a significant share of the slave raiding and trading activities. Kasana played a major role as a slave market of great importance in the 1880s because of the Zambarima establishment there. Unfortunately, the slave histories of Ghana have seemingly concentrated on the trade forts and castles along the coast of southern Ghana, without much regard for the interior which served as the source of supply.

Kasana (1º 52'W, 10º 58' N) is located 16km northeast of Tumu, the District capital of the Tumu District in the Upper West Region (Figure 1). The settlement currently has about eight households or compounds and a population of less than 500 people. They speak Grusi, one of the sub-families of the languages of Ghana known as Gur (Naden 1988: 15-16). The mainstay of the people is farming and the keeping of livestock for domestic use. Kasana lies in the Tropical Continental or Guinea 103

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

Figure 1. Map indicating the location of the study area in the Upper West Region and of Kasana.

the knowledge of the current inhabitants on these slave raiding and trading activities during the nineteenth century. In doing this, ten people were interviewed in Kasana. The people selected for interview were generally the aged, and younger people who held some position of authority in the community. Through the questionnaire it was discovered that the archaeological potential examining slavery of Kasana was enormous.

Savannah woodland. The area has quite extensive agricultural lands, watered principally by the Pichira, a tributary of the Sisili River, and associated streams such as the Sakombuli, Mogu, and Chira. Oral Information and the Landscape Besides completing archaeological field survey, oral tradition was collected. The available historical documents indicate that the Zambarima slave raiders engaged in raids in northern Ghana for slaves in the nineteenth century. They are said to have established their headquarters in Kasana where they kept and marketed slaves captured in the raids (Der 1975 and 1998). A questionnaire was prepared with the aim of ascertaining

In the first place, the story was told that Kasana had a key role in the slave raids and trade. The Zambarima who came first as traders in agricultural and mineral resources as well as horses, soon turned their attention to slaves when they became aware of the market value of slaves through being involved in the Dagomba wars to capture 104

SAMUEL NILIRMI NKUMBAAN: THE ARCHAEOLOGY OF SLAVERY: A STUDY OF KASANA, UPPER WEST REGION, GHANA

Figure 2. House remains at the former slave market site, Kasana.

Figure 3. House remains at the former slave market site, Kasana.

slaves for Ashanti as tribute. The Zambarima are said to have raided the Gurunsi area for slaves. Later they sacked the people of Kasana and established their headquarters there. Here, they established a slave market, and also raided raiding surrounding villages to capture slaves for the market at Kasana. Some informants indicated that the Zambarima were in two groups, the first led by Gazare and the other by Samori. Samori and his group were based at Sati, a village to the north of Kasana (now in Burkina Faso). This group was supposed

to be the slave-raiding group. The group led by Gazare was said to be the slave holding and marketing group. They were resident at Kasana where slaves were brought, kept, and sold. There was thus a conscious, coordinated effort whereby Samori and the raiding forces brought the slaves they captured to Gazare and his followers to keep and sell them to the mutual benefit of the two groups. Owing to this trade, houses were built around the market to 105

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA The Zambarima slave raiders are said to have operated in this area for at least a decade. They attacked surrounding villages such as Sati, Gwollu, Kunkoku, Jawia, and Bullu. From Kasana the Zambarima under the leadership of Babatu also invaded Dolbizon and Jaana to the southeast. Hence measures were taken in the 1800s by the people of Gwollu to defend themselves. This took the form of building a fortification wall round the town. Surrounding villages such as Jawia, Lillie, Kunkoku, and Nyiemeti, which thought of seeking refuge within the wall, gave support to its building. The first wall was built around the town without thought as to including the farmlands and water bodies within it. But it happened that raiders kidnapped women and men who went out of the wall for water and farm activities, hence an outer wall was built subsequently.

Figure 4. Desert plant with Zambarina association.

accommodate the merchants who came to either sell their slaves or buy slaves and other goods. It is interesting to note that the remains of houses can be seen at the market site (Figures 2 and 3). Other features recorded included a depression of about 2m depth and 4m width, dug some 50-100m north of Gazare’s house where slaves were believed to have been bathed before they were taken to the market. Also important is a spot believed to be the grave of Gazare. This is regarded as a sacred site and one has to perform a sacrifice or give money to be shown the location. The landscape of Kasana is characterized by house and slag mounds. Each house mound is associated with a slag mound and a depression, probably the places from where earth was dug for the construction of houses or furnaces and where water collected in the wet season for household use. There are also eleven water cisterns within the settlement which may have served as sources of water supply for the community and later the slave dealers.

From within these defences, the people of Gwollu and the surrounding villages formed a strong confederacy under the able leadership of Tanjia in the 18th century. Having secured their territory and people, they marched on the Zambarima and dislodged them from their invasion base at Sati in the second half of the 19th century. This weakened the Zambarima and the base at Kasana fell back into the hands of the Kasena. Gazare was killed in one of the uprisings and Babatu took control. Following the defeat of Sati and Kasana, the Zambarima moved south into the region of Wa and further south to the area around Yendi in the Northern Region of Ghana. The Kasena people of Kasana saw their victory over the Zambarima as an eschatological intervention. In the first place the sword marks on the stone already described were seen as an indication that their prayers to the gods and ancestors were answered. Also they claim the intervention of their river god, Pichira in the defeat of the Zambarima raiders. The inferior weapons of bows and arrows which they used in the defeat of the Zambarima who had more sophisticated weapons and horses gives credence to a religious based explanation for their victory over the Zambarima forces.

Another feature on the site is a midden that was used by the people of Kasana before the Zambarima invaded them, and which the Zambarima used when they established their settlement and slave market. It measures some 100m in length and 40m in width. A scatter of bones, probably of animals, was observed all over the surface of the site where the slave market was located. The bones, according to informants, came from the animals captured by the slave raiders and used to feed the slaves and the Zambarima merchants and warriors themselves.

Excavation at Kasana Two loci; the market square and an area claimed to be the slave master’s quarters, were identified as relevant to the slave raiding and trading activities. Located between the market square and the slave master’s quarters, which are about 800m apart, was a huge midden. The market square and the merchant’s quarters were designated as locus one, and the slave masters quarters and the rubbish mound as locus two. Given the nature and extent of the site, it was difficult to establish a grid system over the entire site to aid the selection of units for excavation. Therefore, based on direct observation of surface configurations and personal intuition, four units of 2m x 2m were excavated i.e. two units each at both locus one and two. The excavation strategy involved the use of arbitrary levels of 10cm to a context or level. The excavations were carried out in July 2002.

A plant believed to be a desert plant is also claimed to be one of the true manifestations of the presence of the Zambarima (Figure 4). Other manifestations are in the form of marks on stones on the site. One of them located at the market square has marks, which informants claim were specially designed in the form of a checkerboard, so that the merchants could play the game of draughts. Another stone with three striations or grooves is said to be from the sword marks of one of the great warriors who helped expel the Zambarima. It is believed that this warrior struck the stone with his sword while calling to the ancestors and the gods of the land to come to their aid to defeat the Zambarima. The people are of the opinion that the marks signify an answer to his prayers.

106

SAMUEL NILIRMI NKUMBAAN: THE ARCHAEOLOGY OF SLAVERY: A STUDY OF KASANA, UPPER WEST REGION, GHANA Artifacts Pottery Shell Bones Smoking Pipes Metal Beads Spindle whorls Stone Ivory Glass Rhenish ware

Units Locus 1 Pit 1 1,244 14 25 -

Locus 1 Pit 2 3 403 -

Locus 2 Pit 1 3,722 17 173 45

Locus 2 Pit 2 1, 397 785 91 17

Total 6,366 816 692 62

Percentage 79.78% 10.23% 8.67% 0.78%

16 5 1

-

9 1 1

2 2 -

27 8 2

0.34% 0.10% 0.03%

1

-

3 -

1 1

3 1 1 1

0.04% 0.01% 0.01% 0.01%

Table 1. Artifact totals from Kasana.

Stratigraphic Profile

Artifact Summary

A. Locus One, Pit One. This pit, located to the south of the merchants’ quarters, had five levels. Levels 1 and 2 (0-20cm) had the same soil texture and colour, a mixture of dark and ashy loose humus soil with specks of charcoal and rootlets. From the 20cm depth at the north corner, and some 40cm into the south wall of the excavation unit, purely ashy loose loamy soil with rootlets was noticed. The third level was of brown loose soil with specks of charcoal. Level 4 had the same kind of soil that is characteristic of the north corner of levels 2 and 3. The fifth and the sterile level were of compacted red brown sand (see Nkumbaan 2003 for stratigraphic profile drawing).

A total of 7,979 artifacts were recovered from the excavations. These comprised 6,366 pot sherds (79.78% of total artifacts); 816 shells (10.23%); 692 bones (8.67%); 62 smoking pipe fragments (0.78%); 27 metal objects (0.34%); three fragments of stone bracelets (0.04%); eight beads (0.10%); two spindle whorls (0.03%); and one each of an ivory bracelet, a fragment of a cobalt blue rhenish ware, and a glass bottle fragment, each of which constitute 0.01% of the artifact totals. Pottery The jar and bowl forms were established on the basis of rim, body, and base sherds, and various decorative patterns were outlined or described. These included rouletting, grooving, comb impression, wavy line impression, dot stamping, perforation, lip notching, incision, and squaral stabs or stamps. Surface finishes included burnishing, slipping, smudging, and handsmoothening. The predominant jar forms had everted rim curvatures and rim diameters ranging between 12cm to 28cm. The bowl forms are characterized by either inverted or straight rim curvatures. They generally are hemispherical in shape with rim diameters ranging between 12cm to 24cm. The base forms were of four kinds. One type is represented by a cone-like pedestal with a flat or straight bottom. The second type is characterized by a round bottom, supported by or sitting on solid projections. The third form has a round base which is quite difficult to distinguish from ordinary sherds. Finally, there is a type that has a kind of concave depression at the bottom of a ring-like projected base (Nkumbaan 2003).

B. Locus One, Pit Two. This pit, located at the market square, had only three levels. The first and second levels of the unit had dark-brown loamy soil that contained only bones. The third and sterile level was composed of reddish brown and coarse sand containing a few bones. C. Locus Two, Pit One. The location of this unit was at the lower part of the rubbish mound between the market square and the slave master’s quarters. Fifteen levels were excavated (see Nkumbaan 2003 for stratigraphic profile drawing) D. Locus Two, Pit Two. This unit was located between two compounds in the area claimed to be the slave master’s residence. Ten levels were excavated. The first layer was characterized by a brown gravelly soil. Underneath this level was a dark brown humus soil with rootlets. A greyish, gravelly soil with rootlets characterized the next layer. The fourth layer had greyish sand that was lighter than that described for the preceding layer. The fifth level did not differ from the fourth one except that it had specks of charcoal in it. Below this layer was a thin strip of dark ashy soil with specks of charcoal. Orange brown earth characterized the sterile level.

Spindle Whorls A total of two spindle whorls were recovered from the excavations, one from level 2 of locus 1, pit 1, and the other from locus 2, pit 1. They have been labelled and described as types A and B. Type A is generally spherical in shape. It has a single hole measuring about 5mm in diameter running through it. The hole at the base is larger 107

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA B, and C (Figure 6). Five beads are identified as type A. These are disc beads manufactured by the technique of winding and are thus called wound beads. There are also unique features associated with some of them that cause variation, prompting further classification. For this reason, the beads in type A are further designated as A1, A2, and A3.

at the top and it seems to have been pierced before firing. This is shown as ‘A’ in (Figure 5). The middle portion of the spindle whorl is characterized by a kind of carination and the diameter of the spindle whorl at this point is greater than at other parts of the body. The surface finish is a black burnish without decoration. This is similar in description to type ‘A’ established at Begho by Crossland (1973: 90-91). Type B is also spherical in shape. It is more rounded, and bigger in size, almost as big as a table tennis ball (Figure 5). It is greyish in colour, undecorated and with a single even hole of about 5mm going through it. The description is similar to type ‘D’ established at Begho by Crossland (1973: 90–91).

Those belonging to A1 are blue in colour and have irregular large holes. Type A2 is only one bead and its unique feature is that the probable manufacturing technique is by a mould press. It has a regular round hole through it and is green in colour. The A3 type is also the only one of its kind. It is a drawn bead, slashed or cut into disc. The use of the winding manufacturing process is also likely. It is made of a white translucent glass with a greyish tinge. The surface or edges have been ground to produce a four-sided faceted bead (Figure 6). Glass bead type B is only one in number, and was recovered from level 2 of locus 1, pit 1. It is a wound bead, a truncated bi-cone in shape, with a flat end. It is a mix of blue and grey colours. The perforation tapers in from the flat end. This is shown as B (Figure 6). The last type, C, is best described as a drawn, multi-layered bead. It is the only one of its kind, recovered from the surface of locus 2, pit 2. At least three clear outlines or layers are visible; the outer, middle and inner layers. The outer and inner layers have a deep or dark blue colour, whereas the middle layer has a light blue colour. The bead is cylindrical in shape and has a seven-sided facet.

Figure 5. Spindle whorls from Kasana.

Beads

Metal Objects

A total of eight beads were recovered in Kasana. They were recovered from locus 1 pit 1 levels 1, 2, and 4; locus 2, pit 2, levels 1, and the surface of locus 2, pit 2. They are grouped into two main types as ceramic and glass beads. Out of the eight beads, only one is ceramic and the remaining seven are glass. The ceramic bead, labelled ‘D’ in Figure 6 was found in level 1 of locus 2, pit 2. It has quite rough edges and a perforation which is about 7mm in diameter. It is brownish in colour with a decoration, more like roulette, but it is badly eroded from the exterior profile such that the decoration is unclear. This could have been made by simply abrading a potsherd on a harder substance to attain a desired shape and thereafter, perforated.

A total of 27 iron objects were recovered. These include iron weapons and equipment used for hunting, and as body adornments. The composition of the metal assemblage is; five arrowheads, six rings, four knife fragments, one fragment of archers loose, two nails, one fragment of a scoop, one stretched bangle and five undiagnostic pieces. Although some of the artifacts are badly corroded, a few of the arrowheads show the complete features or nature of the arrowheads. They are made with tangs, and are meant to have been hafted or fitted onto stalks or wood to form arrows. They also possess scales located on the sharp edge that pierces the objects shot at. The scales have the functional attribute of hooking the arrow onto the object shot. The tangs measure some 10cm in length and the piercing edge measures around 6cm in length. The scales are at least a centimetre long (Figure 7).

The glass beads have been analysed based on types and techniques of manufacture. They are labelled as types A,

Figure 6. Beads from Kasana.

108

SAMUEL NILIRMI NKUMBAAN: THE ARCHAEOLOGY OF SLAVERY: A STUDY OF KASANA, UPPER WEST REGION, GHANA of them are red slipped. They have round bases with the stem and the bowl sitting on top of the base. This type is termed the “double-angled” pipe (Shinnie and Kense 1989: 146). Two of the bases have elaborate decorations on the necks. One of them has fine tiny multiple grooves around the back of the bowl ending at the point where the stem departs from the bowl at the front. The other has a single groove around the edge of the base and vertical multiple incisions on the stem. These are very similar to pipes labelled as d and Db 78 36 by Shinnie and Kense (1989: 148, figure 71a), (Figure 8).

Figure 7. Arrowhead from Kasana.

Six rings were found during the Kasana excavations, one of which is very thin and made of either brass or copper wire. Three of the rings are rectangular in shape and a gap exists at the point where the two ends of the rings meet. These ends are smaller in size than the middle part of the rings. They look more like pendants attached to shingles for dancing rather than body adornments. A twisted iron finger ring was also found. In addition, a half piece of a bracelet made of brass was found. It has a rounded bulb at the point where it begins and reduces in size towards its middle portion.

Figure 8. Tobacco pipe bases from Kasana.

An archer’s loose was also found. This is so named because archers wore it on the thumb to ensure a firm grip of the arrow when shooting. There is also a belief system associated with the use of the object. It is said that people who possess appropriate spiritual powers regarding the shooting of arrows use it to produce some form of music by hitting it against the bow when an arrow is shot. This will ensure that they do not miss their targets. A spoon-like fragment which is termed a scoop was also found in the excavations. It is said that these objects were used in fetching tobacco and related items for sale at the market. Figure 9. Tobacco pipe base from Kasana.

Smoking Pipes The type designated B is composed of only one fragment. It was recovered from the level 4, of locus 2, pit 1 and has a plain burnished and red slipped finish. It is the same as that described by Shinnie and Kense (1989: 156) as having a multi-lobed segment at the base and labelled as D2 Db 454 (Figure 9). Type C is a pedestal base in the sense that it is a solid high one on which sits the bowl and the stem. The bases are semi-round and bend or slant down towards the area where the stem departs from the bowl. There were three fragments of this base type that were recovered from the surface and level 2 of locus 2, pit 2, and level 6 of locus 2, pit 1. They are all burnished

Sixty two smoking pipe fragments were retrieved from the excavations. A number of characteristics have been identified and used as a basis upon which the pipes were analysed. These are stylistic attributes associated with the bases, the bowls, and the stems. In all, 15 fragments, making up 24.19% of the collection are identified as base fragments. Twenty-nine of them, constituting 46.77%, are the bowl forms and the remaining 18, constituting 29.03%, are stems. i) Base Types. Six different base types exist. Type A has three fragments, which are beautifully burnished, and two 109

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

Figure 10. Tobacco pipe stems from Kasana

and Kense and shown in figures 71b, 72b and 73a as k, f B7 829, and b Db 465 c1 respectively (Shinnie and Kense 1989: 149, 152, 154). They are burnished but one is not red slipped. The other two, which are red slipped, are also characterised by very elaborate decorations on the bowls. One of them is marked by grooves and ridges round the bowl on which are very fine squares of, probably, cord roulette. The external profile of the bowl seems to suggest a coil technique of manufacture. The second one has a complex inter-weaving of cross hatches, marked by short vertical and horizontal incisions on raised backgrounds achieved by a complex web of grooves. ii) Stems. A total of 18 fragments of stems were found in the excavations. Seven of these were diagnostic and formed the basis of establishing stem forms. They are all plain stems, burnished and red slipped, except one, which has a leaf shaped design. However, the collars of the stems are different, resulting in the types. In type A, the collars of the stems are shaped like a circumcised penis (Figure 10). They have a pointing end with a uniquely marked head. The B type is made of an open end with flaring collars (Figure 11). Out of the seven diagnostic pieces, four were identified with the type A and the remaining three with type B.

Figure 11. Tobacco pipe stem from Kasana.

and red slipped. They have about 5 incisions each running vertically across the bases ending at the point of the neck’s attachment. Type D is also a pedestal base. An example was recovered from level 5 of locus 2, pits 1. It has five-lobes or facets and is similar to that shown as D1 Db 79 216 in figure 73b (Shinnie and Kense 1989: 156). It also angles or slants down towards the area of the stem attachment. Type E is composed of a single fragment recovered from the surface of locus 2, pits 1. It is a round, pedestal base with multiple vertical incisions running up the base, ending at the neck joint. The type is unique in the sense that the base looks circular when viewed from the top, but from the bottom a gap is noted in the base marking out the stem’s departure from the base.

iii) Bowls. A total of 29 bowl fragments were found. Three of them are represented in Figure 12. The bowls exhibit quite a range of variations in terms of shape and decoration. Some have an externally thickened lip, but most do not. Some may also be described as slightly tapered, conical or convex in shape. The decorative patterns associated with the bowls range from vertical or horizontal incisions or grooving, to patterns of impressions made by stamping or roulette. Bones

Base type F is composed of three fragments. They represent the types described as single-angled bases. This is because the underside of their bases and the stem form a continuous surface. They are similar to those of Shinnie

A total of 692 bones were recovered from the excavations at Kasana. The bones were found on the surface, and across the various levels that were dug. Out of the 692 110

SAMUEL NILIRMI NKUMBAAN: THE ARCHAEOLOGY OF SLAVERY: A STUDY OF KASANA, UPPER WEST REGION, GHANA

Figure 12. Tobacco pipe bowls from Kasana.

occurred within the levels of locus 2, pit 2, and numbered 785. Stone, Glass, and Ivory Three fragments of stone bracelets were found at Kasana locus 2, pit 1; one on the surface, and the other two from level 6. Two are of green stone and one of quartz. Besides the glass beads, one fragment of glass was found at Kasana. It is a plain clear glass neck of a small bottle. It was recovered from level 3 of locus 2, pit 2. One of the special finds from Kasana was an ivory bracelet recovered from level 3 of locus 1, pit 1, the merchants’ quarters. Though in three fragments, these can be joined together to form a whole (Figure 13). Rhenish Stone Ware

Figure 13. The ivory bracelet from Kasana.

A fragment of a cobalt blue rhenish stone ware, probably of German origin was also found in the excavations from level 3 of locus 2, pit 2.

bones, 381, representing 55.06%, were identified and classified as Bos (cattle). The Bovid (sheep and goat) bones numbered 183, constituting some 26.44%. Aves (birds) are represented by 21 bones, and rodents by six bones, making 3.03% and 0.87% respectively. Amphibian and insectivores had only a bone each (0.14% each). In addition, there were bones of reptiles and carnivores, numbering three and 13 respectively. The percentage distribution of these is 0.43% and 1.88%. A total of 83 bones could not be classified and are designated as unknown.

Discussion In the documentary sources, Kasana is described as a slave market and the headquarters of the Zambarima (Der 1998: 20-23). It was a place where the transaction of buying and selling of slaves was done. Merchants were believed to have come from all over the country and beyond the borders of Ghana to partake in this business. The residences of the slave masters and the merchants quarters mentioned earlier reflect these. Aside from the description as a market, Kasana can also be described as a transit point. This is because there is every indication that some slaves were only rested there and were transported further south to markets like Salaga, and on the coast. Such an interpretation is supported by the water bodies recorded, Okoro (2003d, e) for instance, recorded eleven water cisterns. Rivers and lakes nearby could also serve a similar function. Moreover, ditches near houses were

Shells A total of 816 shells were recovered. They were identified and classified into only two groups by specialist analysis by Mr. Bossman Murey. These are designated as Donex acutangulus and Cypraea moneta. The latter is made up of only two shells constituting 0.24% of the total collection. These were found in levels 1 and 3 of locus 1, pit 1. The Donex acutangulus mostly

111

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA loved women, their association with them caused them to temporarily forget their security. Taking advantage of this and using the local women as bait, they caught the Zambarima off guard and killed them. This song is still sung during the annual festivities to commemorate the slave trading activities.

recorded and are said to have functioned as water holders to wash or bathe the slaves before they were taken to the market. It is generally expected that a slave traffic site be littered with shackles, chains and the like for people to identify or accept it as such. Though these are important artifactual manifestations worth considering, archaeology is not a mutually exclusive source in their identification. Other sources such as oral traditions and documentary sources provide very useful data to supplement the archaeological record.

The Zambarima are very bad. The Zambarima are very wicked. They killed our father. They killed our mother. They killed our people. They killed our livestock. They stole our foodstuffs, They destroyed our households and property.

It can also be asserted from the archaeological finds that the contents of the sites, that is, the relics excavated, predate the era of the slave traffic in Kasana. This explanation is supported by the fact that the smoking pipes found in the excavations belong to types believed to have been introduced to the coast from the north, and dated to the 1660s on the coast at sites such as Asebu. Similarly, cowries found in the excavations and identified as Cypraea moneta were not shells common in the nineteenth century.

These oral accounts give useful information about the extent of Kasana’s involvement in the slave trade and raiding activities of the nineteenth century. The songs are regarded as sacred and a sacrifice is made before they are performed. They can only be said when there is war or during the rituals to commemorate the slave raids. Ethnographic materials in the possession of the inhabitants of the present settlement also reflect the slave traffic. These include shackles that chained the slaves by the legs and horse riding accoutrements believed to have been used by the Zambarima horsemen, particularly the raiders, as well as weapons, a gun, spear, and a long iron rod described as used by the local warriors in defending themselves against the raiders. Additionally, the people of Kasana have in their possession certain tattered clothes believed to represent the war attire used by their forbearers during the slave raids. These are used during annual rituals or festivals to re-enact what happened during the time of the slave raids (Figure 14).

Oral accounts from Kasana show that the people have knowledge about their involvement in the slave trading and raiding activities. Stories are told of the capture of people who were sent away from the communities. Tales are also told of how their livestock were captured and used to feed the slaves, the raiders, and merchants at the slave markets. These are told on the free-text basis which means that fabrication, exaggeration, forgetfulness and other distortions resulting from personal interests can come into play in the narratives. However, these are backed by the fixed text types existing in the form of songs passed down the generations with the words remaining unchanged. Some of the songs reflect names of the key players in the slave raiding and trading activities in northern Ghana such as Babatu, Samori, and the Zambarima. A few of these songs were collected at Kasana and are translated as follows:

As described, the majority of the artefacts retrieved from the excavations consisted of pottery. These were well made, well fired and some were burnished and slipped. These characteristics are evenly distributed across the site, the units, and levels excavated. This implies cultural consistency in the pottery making, so that the slave raiders or traders who are claimed to have come into the area did not influence the industry in anyway. Similarly, the local smoking pipes have consistency in manufacture, decoration etc., across the sites. Slave trading activities can also be inferred from some of the other archaeological finds recovered. For instance, the quantity of bones of domesticated animals corroborates accounts in oral traditions and documentary sources that the raiders captured animals and killed them to feed slaves, whilst, the glass beads are signs of external or foreign influence since there is no record of local manufacture of glass beads in the area.

Ah! Are there no men? Are there no warriors? The Zambarima are coming. If there are no men, they should hand in the weapons to the women to fight for them. Ah! Are there no men? This song is said to have been sung during the days of the raids with the aim of expelling nervous energies or fears in local warriors, and to boost their confidence and morale to encounter the enemy. The Zambarima loved women. And that was what killed them. That was what killed Babatu, That was what killed Samori, That was what killed the Zambarima.

The metal objects in the form of arrowheads, and other weapons could have been used in raiding activities or to guard the slaves from escaping. They thus reflect an element of the slaving activities. It is also indicated in some documents that the people of Sankana defeated the Zambarima slave raiders using bows and arrows (Der

This song indicates that women were used as bait at a point in the war against the Zambarima. Because they 112

SAMUEL NILIRMI NKUMBAAN: THE ARCHAEOLOGY OF SLAVERY: A STUDY OF KASANA, UPPER WEST REGION, GHANA

Figure 14. Clothing said to have been worn during slave raids.

2000). It is clear from the landscape and written documents that Kasana was an established iron smelting and smithing site before the Zambarima established their post there. The existence of such weapons may have nothing to do with the slave traffic but rather represent the iron working industry and local exploitation of the environment.

hostilities from the Zambarima who are on record as having established friendly relations with the Islamised areas of Wa and Tumu and of course, Islam forbade the use of fellow Muslims as slaves (Der 1975). With the adoption of Islam, a number of ceremonies including naming, and marriage ceremonies, and the dress code now conform to Islamic ideals.

This question of cultural continuity or consistency should not be surprising because, according to documentary sources, the Zambarima influence in Kasana lasted for just a decade or so, that is, between the 1880s and 1900. This is too short a period to cause any meaningful cultural change in existing ways. More so, given the concentration of the Zambarima on their trade interests, slaves, they probably had little time to think of other activities, which could have affected the material culture of the indigenes.

The effects of the slave traffic in the form of raids and trade on Kasana is two fold; positive and negative. On the basis of existing evidence, the negative elements far out weigh the positive effects. The positive effects may be explained in terms of the introduction of certain new trade items such as glass beads. Negatively, the people of Kasana lost a great deal of their young men and women who were captured in the slave raids. These members of the active population who were to contribute to the development of Kasana were either killed or taken away. Informants intimated that, the Gwala (Kasena word for the slave raiders) made the people as slaves do any kind of job. For example, one person said, “if there was something frightening or fearful under a tree, one would not send his or her son but a slave because he is just any other person (inferior person)” (Howell 1998: 18). Kasana as a community was devastated.

Yet this point noted, the influence of the Zambarima on other aspects of Kasana culture cannot be denied. Socially, though the people still practise their indigenous religion, they imbibed the practice of Islam from the Zambarima who were predominantly Muslim. They may have accepted the religion in order to avoid further 113

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA shed tears as some of the songs reflecting the event are sung.

A further far-reaching effect of the slave traffic on Kasana was the loss of its iron working industry. Kasana before the coming of the Zambarima slave raiders was a very important iron smelting centre. This activity was a household centred business; indicated by the slag mounds and furnace fragments found around the houses. Kasana supplied agricultural implements, weapons, and other household utensils to neighbouring markets. The trade in iron objects at Kasana may well explain the interest of the Zambarima in making the settlement their headquarters.

Acknowledgements I am grateful to Professor Anquandah, Department of Archaeology, University of Ghana for his support and supervision which has culminated in the work. I would also like to thank my dear wife, Phyllis Mapram for her moral support and encouragement. A debt of gratitude is owed the Cultural Attaché of the French Embassy, which through UNESCO, provided part funding for my research. I am grateful to Mr. Bossman Murey for his assistance photographing the artifacts shown in this paper. Thanks to Aaron Eshun to whom the drawing of the map of the study area is credited. I am also very grateful to Professor Timothy Insoll for spending some time to edit this paper for me. All who have contributed in diverse ways and of whom I cannot fully mention are duly acknowledged

It has been said that, following the intensification of the slave wars, the people of Kasana concentrated more upon the production of weapons such as arrows and spears than agricultural implements (Howell 1998: 52). When Kasana was finally invaded, the people moved further south and settled closer to Tumu, thus losing contact with the iron resources and the iron working technology. After the Zambarima were defeated, the iron working industry never resumed. Currently, Kasana has only four people with knowledge of blacksmithing. It should also be noted that, on the return of the people to Kasana after the Zambarima had left, their water cisterns were dried up. The people now travel about 2km to the Pichira River to draw water.

References ANQUANDAH, J.R. and VAN HAM, L. 1985. Discovering the Forgotten Civilization of Komaland, Northern Ghana. Rotterdam: Ralph Schuurman Productions.

Conclusions Based on the use of sources such as oral traditions, ethnographic materials, and written documents, as well as the landscape evidence, slave trading can be inferred from the archaeological evidence. In fact, the pointers to the existence of the slave trade or the involvement of Kasana far out weigh arguments against. The landscape and associated oral and documented tales emphasize this. This brings a new perspective to the conception of the slave traffic. A number of questions arise from this. First, to what extent can archaeologists rely on oral data? Secondly, what are the markers one expects to find on a slave traffic site?

CROSSLAND, L.B. 1973. A Study of Begho Pottery in the light of Excavations Conducted at the Begho – B2. Unpublished M.A. Thesis, Department of Archaeology, University of Ghana, Legon. DAVIES, O. 1966. The Invasion of Ghana from the Sahara in the Early Iron Age. (In), Cuscoy L. D. (ed.), Actas del V Congreso Panafricano de Prehistoria Estudio del Cauternario. Santa Cruz de Tenerife: Museo Arqueologico Santa Cruz de Tenerife, pp. 27-42.

However, the inter-face of archaeology and the evidence from other sources clearly highlights the circumspection with which we have to use these sources in drawing conclusions. It is clear that, not all expectations based on oral accounts may be corroborated by archaeology. Oral accounts are subject to inconsistency through exaggeration, prejudice, forgetfulness and distortions resulting from personal interests, but it should also be noted that not all cultural materials are preserved in the archaeological record. This suggests that a single source cannot provide sufficient basis for drawing conclusions. The use of a combination of these sources is essential in arriving at reliable conclusions.

DAVIES, O. 1970. Ghana Field Notes: Part 2; Northern Ghana. Unpublished Report, Department of Archaeology, University of Ghana, Legon. DAVIES, O. 1973. Excavations at Ntereso, Gonja, Northern Ghana. Final Report. Unpublished Report, Natal Museum, Pietermaritzburg, South Africa. DAVIES, O. 1980. The Ntesero Culture in Ghana. (In), Swartz, B. K., and Dumett, R. A. (eds.), West African Culture Dynamics. The Hague: Mouton, pp. 205-225. DER, B.G. 1975. Samori’s Relations with North-West Ghana, 1892-1898. Unpublished Manuscript.

The loss of relatives, settlements, animals, craft industry and the resulting psychological effects cannot be overstated. The people of Kasana attribute the low living standards in the area to the slave raiding activities, which ravaged their settlement, and vandalized their property. The thought of this event brings bitter memories to even the present generation and it is not surprising to see them

DER, B.G. 1983. Missionary Enterprise in Northern Ghana, 1906 – 1975: A Study in Impact. Unpublished PhD Thesis, Department of History, University of Ghana, Legon. DER, B.G. 1998. The Slave Trade in Northern Ghana. 114

SAMUEL NILIRMI NKUMBAAN: THE ARCHAEOLOGY OF SLAVERY: A STUDY OF KASANA, UPPER WEST REGION, GHANA Accra: Woeli Publishing Services, NKUMBAAN, S. N. 2003. The Search for Evidence on the Slave Traffic: Historical Archaeological Study of Kasana and Sankana, Upper West Region, Ghana. Unpublished M.Phil Thesis, Department of Archaeology, University of Ghana, Legon

DER, B.G. 2000. The Sankana Kalibi Festival: A Background. Wa: Wa Catholic Press. GAVUA, B.K. 1985. Daboya and the Kintampo Culture of Ghana. Unpublished M.A. Thesis, Department of Archaeology, University of Calgary.

OKORO, J.A. 2001. A Report on a Research on Ancient Cisterns at Saakpuli, Northern Ghana. Unpublished Report submitted to UNESCO (Accra).

GOODY, J.R. 1954. The Ethnography of the Northern Territories of the Gold Coast, West of the White Volta. London: Colonial Office.

OKORO, J.A. 2002. Report on a Research on the Slave Market of Saakpuli. Unpublished Report submitted to UNESCO (Accra)

HOWELL, A. (ed.). 1998. The Slave Trade and Reconciliation: A Northern Ghanaian Perspective. Accra: Bible Church of Africa and SIM Ghana.

OKORO, J.A. 2003a. An Archaeological Research at Salaga (1). The Ghanaian Times, Tuesday April 8, 2003, p. 5.

INSOLL, T., KANKPEYENG, B., and MACLEAN, R. 2004. An Archaeological Reconnaissance in the Tong Hills, and Garu Area, Upper East Region, and Nakpanduri, Northern Region, Ghana. Nyame Akuma 62: 25-33.

OKORO, J.A. 2003b. An Archaeological Research at Salaga (2). The Ghanaian Times, Wednesday April 9, 2003, p. 5.

INSOLL, T., KANKPEYENG, B., and MACLEAN, R. 2005. Excavations and Surveys in the Tongo Hills, Upper East Region, Ghana. July 2005. Fieldwork Report. Nyame Akuma 64: 16-23.

OKORO, J.A. 2003c. An Archaeological Research at Salaga (3). The Ghanaian Times, Thursday April 10, 2003, p. 5. OKORO, J.A. 2003d. Archaeology, Water and Tourism in Kasana (1). The Ghanaian Times, Tuesday May 13, 2003.

INSOLL, T., KANKPEYENG, B., and MACLEAN, R. 2007. Excavations and Surveys in the Tongo Hills, Upper East Region, Ghana. July 2006. A Preliminary Fieldwork Report. Nyame Akuma 67: 44-59.

OKORO, J.A. 2003e. Archaeology, Water and Tourism in Kasana (2). The Ghanaian Times, Wednesday May 14, 2003.

KELLY, K.G. 1997a. Western African and Western Saharan Trade. (In), Vogel, J.O. (ed.), Encyclopedia of Precolonial Africa: Archaeology, History, Languages, Cultures and Environments. Walnut Creek: Altimira Press, p. 529

PERBI, A.A. 1996. The Legacy of Indigenous Slavery in Contemporary Ghana. FASS Bulletin 1: 83-91 PERBI, A.A. 1997. A History of Indigenous Slavery in Ghana from the 15th to the 19th Centuries. Unpublished Ph.D Thesis, Department of History, University of Ghana, Legon.

KELLY, K.G. 1997b. Slave Trade in Africa. (In), Vogel J.O. (ed.), Encyclopedia of precolonial Africa: Archaeology,, History, Languages, Cultures and Environments. Walnut Creek: Altimira Press, pp. 529532.

SHINNIE P. L., and KENSE, F.J. 1989. Archaeology of Gonja, Ghana. Excavations at Daboya. Calgary: Calgary University Press.

KENSE, F.J. 1981. Daboya: A. Gonja Frontier. Unpublished Ph.D. Dissertation, Department of Archaeology, University of Calgary.

SHINNIE, P. L and OZANNE, P.C. 1962. Excavations at Yendi Dabari. Transactions of the Historical Society of Ghana 6: 87-116

KENSE, F.J.1987. The Impact of Asante on the trade Patterns of Northern Ghana and the Ivory Coast. (In), Schildkrout, E. (ed.), The Golden Stool: Studies of the Asante Center and Periphery. New York: American Museum of Natural History, pp. 29-35.

WILKS, I. 1961. The Northern Factor in Ashanti History: Begho and the Mande. Journal of African History 2: 2534.

MIERS, S. and KOPYTOFF, I. 1977. Slavery in Africa. Historical and Anthropological Perspectives. Madison: University of Wisconsin Press. NADEN, T. 1988. The Gur Languages. (In), Krobb Dakubu, M.E. (ed.), The Languages of Ghana. London: Kegan Paul International, pp. 12-49. 115

Clay Toys of the Grandchildren of a Potter in Salaga: Insights for Archaeology in Ghana J. Ako Okoro Department of Archaeology, University of Ghana, Legon, Accra, Ghana. Email: [email protected]

Management and Slavery in Salaga for a doctoral thesis in 1992, after preliminary research in August 1990 had identified the research potential of the town and its surrounding area (Okoro forthcoming). Among the sites investigated were the Slave Market, the Slave Bath, and many ancient and contemporary indigenous wells and cisterns as well as mounds that mark the landscape of the settled and adjoining areas.

Introduction Pottery is one of the most ancient artefacts in Ghana and pots have been used in cooking and serving of food, water management, ritual performances, ancestral and shrine worship. In the moist or plastic state, potters use clay to mould many simple and complex items and produce objects ranging from mundane ones for everyday use through to exquisitely beautiful works of art made to delight the observer. Aside from stone tools, the most preserved archaeological remains are fired-clay artefacts and these are normally found in considerable quantities from surface collections and excavations. In the description, analysis, and interpretation of the archaeological pottery database archaeologists in Ghana use information from ethnographic research conducted both by themselves and by non-archaeologists. However, the issue of children making and using clay toys both in the past and at present has not been considered.

Archaeological, anthropological, historical, and ecological research in the Salaga area of East Gonja has become a long-term project for generating innovative concepts and instructive information for issues on rural water management, guinea worm eradication, slave heritage studies, and tourism in Ghana. Other areas of research have been the determination of the significance of the oral traditions of minority ethno-linguistic groups in the Salaga area and Gonjaland (see Okoro in press a). Whilst pots have been examined not only from the point of their archaeological implications but also to identify the relevance of the traditional potting industry in the preservation of indigenous knowledge (see Okoro in press b).

This paper is focused on two teenage girls who are the grand and great grand daughters of a potter aged about seventy years and resident in Salaga in the Northern region of Ghana (Figure 1). The grand daughter called Faram is fourteen years old and the great grand daughter called Nabeo (the names of the children are pseudonyms used to conceal their identity) is twelve. The old woman is called Adjoa Abortorwro, hereafter referred to as the ‘Potter’, and, because she is not an indigene of Salaga, the circumstances regarding her residence in the town are presented. How the teenage girls became members of the household of the potter is also described. Within a short time, the girls acquired the basic skills of using clay to mould toys that exist within the households and community as play things of children.

According to the 2000 Housing and Population census of Ghana Salaga had a population of 16,196 (Ghana Statistical Service 2005) which is lower than the 19th century estimate of between 20,000 and 50,000 persons (Levtzion 1986:342; Okoro forthcoming). A higher population was recorded during the trading season from December to May when traders and dry-season visitors congregated in Salaga. Salaga is presently one of the major market towns in Northern Ghana and has served as the commercial capital of East Gonja since the beginning of the 18th century. The traditional capital of East Gonja is Kpembe which is located 3km to the east of Salaga, and serves as the residence of the paramount chief called Kpembewura.

Some of the toys are based on the concepts and ideas of the girls and reflect things in the society. Others show the girls mimicking in miniature sizes the items the Potter produces for sale. Even if these toys remain in use for a long time, eventually they have to enter the archaeological record. It is suggested that there could be a longer antiquity for the making and use of similar toys in Ghana.

In the 19th century, Salaga was the nerve centre of important commercial networks that connected the Upper Volta (present Burkina Faso), Hausaland, and the Niger Bend area in the north and northeast to the forest and coastal regions of Ghana (Gold Coast) (Johnson 1986:360; Levtzion 1966:209). By the middle of the 19th century, Salaga had developed into an emporium of slaves and kola nuts in the Volta Basin. Salaga was the major nodal centre in the Gold Coast and Asante kola trade (Abaka 2005:5) and the principal place for Hausa traders looking for kola nuts in huge quantities.

Brief Background Information on Research in Salaga Salaga is the capital of the East Gonja District and it is located about 116 kilometres south of Tamale, the Northern Regional capital. The insights for this paper derive from research completed on Indigenous Water 117

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

Figure 1. Map of Ghana showing Salaga and the places mentioned in the text.

potters whose pots are patronized by the people of Salaga and surrounding villages.

Because of its historic trade background, the population of Salaga is very diverse. There are Hausa, Dagomaba, Mossi, Wangara, Baribari, Chokossi, Kotokoli, Gruma, Twi, and Yoruba people. The town is traditionally a Gonja settlement but the semantic pressure of the Hausa language is so strong that within Salaga Hausa has become the first language of general communication. The bulk of the people of Salaga practice crop and animal production, but trading and provision of services are the predominant economic life-ways. Aside from trading, food preparation and selling, hairdressing, dressmaking, and the brewing of local beer (pito) are the main activities for women. Artisans perform tasks such as house construction, repair of bicycles, and the digging and cleaning of indigenous wells. Blacksmithing and potting are two indigenous industrial activities practiced in Salaga by men and women respectively. Since 1994, there have been only two old migrant and non-indigenous

Gonja women in Salaga and surrounding villages do not make pots and before the arrival of these non-Gonja potters, pots produced by Dagomba potters in villages around Tamale were imported into Salaga. Other pots in found Salaga are made by the Brong and Konkomba potters in Yeji and surrounding villages in the Brong Ahafo region. Similarly, pots made by Konkomba potters living at the Grubi Quarters in the southeastern part of the East Gonja District are sold and used in Salaga. In spite of the stiff competition from cheaper plastic and metal substitutes, there is high usage of pots for storing water, providing water for fowls and pigeons, for cooking, bathing children, shea butter processing, and preparing herbal and ritual medicine in Salaga and the 118

J. AKO OKORO: CLAY TOYS OF THE GRANDCHILDREN OF A POTTER IN SALAGA: INSIGHTS FOR ARCHAEOLOGY IN GHANA

Figure 2. The old potter in Salaga in January 1994 making fresh pots. On her right is the clay and an arrangement of the tools she uses in making the pots.

There was an interval of nine years (January 2003) before the author met with the potter again and found two primary school-going teenage girls permanently living with her. These are Faram (fourteen years), the granddaughter, and Nabeo (twelve years) the great grand daughter. The potter was by then seventy years old. The girls had lived in the house of the potter for a couple of months but they had made clay toys. While the toys set the stage for the enquiries of this author, the circumstances of the arrival of the girls in Salaga changed his understanding and perceptions on potting traditions in Ghana.

many surrounding villages. Pots of diverse sizes and designs therefore feature not only as goods transported by traders into the town but as commodities sold at the Salaga Market that is held every seventh day. The Encounters with the Potter This author met with the potter for the first time on January 21, 1994 whilst surveying and mapping ancient wells inside Salaga (Okoro 2003a, 2003b). The potter is from the Nchumuru ethno-linguistic group which has other variant names such as Nchumurung, Nchumburu and Kyongborong. Nchumuru is one of the dialects of the Guan language of Ghana and the speakers of this language are found in the northern part of the Volta region and south-east part of the Northern region. Until early 1994, she was living at Sabongida, in the Jambuai Area Council of the East Gonja District Assembly as a recognized potter. The potter was aged sixty years when an ethnic conflict forced her to migrate to Salaga where her “direct” daughter aged about thirty-nine years (as at 1994), has been living and working as a dressmaker. As a parent, the potter had three sons and a daughter (the dressmaker in Salaga).

What is a Toy? A toy is an object made or purchased with the sole intention that it would be played with by children. Modern toys are made from plastic, rubber, wood, metal or paper, and they are on sale in markets, malls, shops, and by the roadside. Globally, the production of modern toys has become a multi-million-dollar enterprise and the items range from simple and cheap ones to exquisitely beautiful, complex, and expensive products. Toys show plants, animals, vehicles, buildings, human beings, and aeroplanes. Each year, huge volumes of new toys are produced, sold, used, reused, and discarded.

When I first met with the potter, she had been potting in Salaga for only one week (Figure 2). The mother of the potter did not make pots because she did not learn the skill from her mother (that is the grandmother of the potter). As a result, the potter learnt the potting skills from her grandmother. Against the backdrop of the lack of interest by the only daughter to pursue potting as a career, when the potter was asked about the continuity of her profession, she remarked that, “after my death, that will be the end.” It was in this state of hopelessness and uncertainty about the future of her profession that this author left the potter in 1994.

The ultimate purpose of giving toys to children or grand children is to provide them with items to play with and learn from. While taking away boredom, the use of toys enables children to develop their talents. In urban areas, the amount of money spent on toys by some households can be quite substantial. In Ghana, children who are privileged to have parents or relations who buy them toys normally show off these items to their friends and playmates. Their friends admire the toys by holding or trying their hands on them. When children from rich homes play together, a show-off of toys occurs. This shows that among children, toys are more than play 119

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA grand and great granddaughters. It is the analysis of these cultural relationships that make the information gained on the toys significant.

things. They are items for displaying the wealth and class of their parents or relations. It is common for school children to save money to buy popular toys. Normally, children in rural areas of Ghana do not have easy access to imported and fanciful industrial toys. They use any piece of item they can lay their hands on as improvised toys. For example, they can use sticks as guns in sharp contrast to their urban counterparts who use toy-guns that can spray water or produce sound to mimic the real thing. Thanks to the versatility of clay, children who lack access to ‘modern toys’ can also make their own toys from clay and use them for play.

A Brief Review of Clay Artefacts Archaeological Record of Ghana

There are several reasons why pottery artefacts have been recovered and used for the reconstruction of past culture histories and technological development. One is that, aside from stone tools, clay-based artefacts preserve better under the tropical conditions and acidic soils of Ghana. Organic materials do not preserve for them to be recovered in any significant quantities or suitable forms. In most Iron Age sites pottery is the artefact found in the most considerable quantities as shown below in Table 1.

The Children and Toys in this Study

The enormous size of the archaeological pottery database has contributed to the situation whereby pottery study is a fundamental research activity in Ghana. So dominant has been the concern for describing, analyzing, and interpreting pottery artefacts that, ‘pottery analysis’ forms a popular area of Masters’ and PhD research in the Department of Archaeology, University of Ghana, Legon (see Anquandah 1967; Agorsah 1976; Effah-Gyamfi 1980; Boachie-Ansah1985, 1986; Crossland 1986; Bredwa-Mensah 1990; Shinnie and Kense 1989). It equally attracts the attention of undergraduate students doing thesis research for Bachelors degrees in Archaeology in the department (Tawiah 1988; Aryee 1998). Non-thesis researches on the Iron Age have also devoted adequate attention to the description and analysis of pottery artefacts for the reconstruction of past cultural developments (e.g. Shaw 1961; Davies 1964; Mathewson 1968; York 1973; Boachie-Ansah 1998, 2000, 2007).

The toys under discussion here are made from clay by children and are manufactured to be played-with by children. It is therefore the grand children of the potter who are the best persons to tell adults what the toys represent in real life situations and the source of their ideas and motivations. Non-potter children who use the toys are also the ones who can describe their impressions regarding the toys they receive and play with. Like the potter, the children are exploiting the plastic properties of moist clay to make a range of items based on the skills acquired through observation and a few direct instructions. The young and emerging potters cannot be described as ‘children’ of the potter because as will be described, the only real daughter of the potter did not learn potting. The transfer of the potting knowledge is not from mother-todaughter. It is from grand and great grandmother to Region in Ghana Eastern Ashanti Northern Northern Brong Ahafo

the

Renfrew and Bahn (1996:52) observe that, “fired clay, such as pottery and baked mud brick or adobe, is virtually indestructible if well fired. . . . It is therefore, not surprising that, for the period after the introduction of pottery-making, ceramics have traditionally been the archaeologists main source of evidence.” In Ghana, the list of archaeologically recovered pottery artefacts is long. It includes full pots and their fragments, pot stands, hearth pots, pot lids or covers, lugs, oil lamps, pedestals and pottery pipe drains. Others are tuyerés (full pieces and fragments), furnaces (and furnace wall fragments), and mud bricks (normally sun dried). The rest are intriguing bits of artisanship in the form of smoking pipes, terracotta figurines, clay beads, clay discs, clay bracelets, and spindle whorls.

Adults normally accord toys with little or no value because they see the toys as miniature versions of the real thing lacking the importance of the latter. But for children, it is this element of miniature or tiny versions that creates the conceptual framework for their understanding and use of the toys accordingly. The way and manner in which toys are admired, handled, and used by the children can fascinate even the very adults who bought them. When adults imitate children in the use of toys, it creates enormous fun and laughter among children who may tell their friends about that experience. Children learn about the real world through the ownership and use of toys. Toys are therefore an element of the material culture of children in ancient and contemporary times.

Site Dawu Twifo Hemang New Buipe Daboya Begho-B2

in

Total potsherd 500,000 100,00 75,000 70,000 24,000

Table 1. Total numbers of potsherds from excavations in Ghana

120

Researcher(s)/source Shaw (1961) Bellis (1972) York (1973) Shinnie and Kense (1989) Crossland (1989)

J. AKO OKORO: CLAY TOYS OF THE GRANDCHILDREN OF A POTTER IN SALAGA: INSIGHTS FOR ARCHAEOLOGY IN GHANA

Site

Pot/sherd

Ahwene Koko Begho-B2 Daboya New Buipe

6151 24,077 70,000 75,000

Smoking Pipes 16 31 216 150

Spindle whorls

Researcher(s)/source

6 5 32 30

Boachie-Ansah (1986) Crossland (1989) Shinnie & Kense (1989) York (1973)

Table 2. Components of the archaeological pottery database Site Bonoso New Buipe Daboya Begho-B2

Full pots 115 100 58 14

Total pottery 2137 75,000 70,000 24,000

Researcher(s) Boachie-Ansah (1986) York (1973) Shinnie & Kense (1989) Crossland (1989)

Table 3. The number of full pots from excavations Site

Pottery

New Buipe Daboya Begho-B2 Ahwene Koko

75,000 70,000 24,077 6,151

Copper objects 8 21 14 4

Iron objects 185 163 64 1

Glass Beads 123 63 21 1

Researcher(s) York (1973) Shinnie and Kense (1989) Crossland (1989) Boachie-Ansah (1986)

Table 4. Comparison of the figures of potsherds and other artefacts recovered from excavations in Ghana

Potsherds have been analysed to show the material and non-material features of ancient communities as well as the ideas, artistry and concepts of the potters who made them. Secondly, because pottery has served as utensils and containers in day-to-day activities of humanity for millennia (Rice 1987: 113), insights from archaeologically analyzed pottery have been used to determine trade relations, food preparation (Anquandah 1982, 1986), and physical migrations and diffusion of knowledge (Davies 1964, 1980; Mathewson 1968) in Ghana. Pottery has also been used to explain the sudden abandonment of sites (Shinnie and Kense 1989), and to estimate the periods and duration of their occupation (York 1973; Shinnie and Kense 1989).

Non-pot artefacts make an insignificant contribution to the total number of pottery remains recovered as shown in Table 2. Because of the big size of the database, researchers sometimes segregate potsherds according to size and surface decoration in thefield to determine which criteria to be admitted to the preliminary and final analysis. This is understandable in a situation where large bodies of pottery evidence have to be managed prior to being transported to base to be examined manually without the aid of computer based inventorying and analytical programmes. Equally, the literature shows that in comparison with potsherds, the number of full pots recovered from excavations that can be used for direct description and classification is normally limited. In Table 3 information on complete and near complete vessels recovered from sites is shown. The bulk of the knowledge to be derived from pottery has been based on the analysis of fragmented parts as a result of preservation conditions.

Terracotta figurines from sites of Late Stone Age sedentary and food production communities at Boyasi Hill (see Anquandah, this volume), Bonoase and Ntereso have been described as art works and representations of life forms of human beings as well as providing evidence of the domestication of animals (Anquandah 1982:60-62). When Boachie-Ansah (2000:65) studied the funerary terracotta and pottery of the Akan of Ghana, he examined their distribution, function, chronology, origins, and their significance to the student of Akan culture history. Like many such investigations, he uses the artefacts of clay to obtain information on the socio-religious life of the societies of the past. The issue of understanding the lifeways of the potters who made them is not considered. Normally it is assumed that the potters who made pottery in Ghana are adults and also female.

When total pottery (pot and non-pots) is compared to non-clay artefacts, the same skewed representation is evident as shown in Table 4. Because the figures that are registered for non-clay artefacts are low, their analysis often takes little of the researchers’ effort. From the table, only one piece each of glass bead, iron objects and four copper objects were found at Ahwene Koko compared to over six thousand pottery artefacts from the same site. It is clear that pottery has dominated the archaeological record in most Iron Age sites (dated between 500B.C to 1500 A.D.) and also historic sites (dating from 1500 A. D.) in Ghana (Anquandah 1982:67). The attention pottery has received over time is therefore not by accident.

This author’s review of the available archaeological reports has failed to find any mention of toys as an item being identified from pottery assemblages in Ghana. 121

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA Wenchi, Boachie-Ansah (1986:258) studied contemporary potting in addition to completing his excavations so as to obtain information to aid his archaeological interpretation. Accordingly, he gathered information on the traditional methods of potting, the vessels forms, surface treatment and decorations to determine similarity, continuity and change between the past and the present.

Children have not been conceptualized as being present and hence represented in the archaeological data as makers and users of clay artefacts. The observation based on this is that pottery analysis in Ghana has been treated as a ‘serious business’ with no room for ‘kids’ stuff.’ From its inception pottery studies in Ghana has aimed at the description and classification of potsherds to obtain information on vessel shapes and uses, surface finish and decoration. This aim seems to have been carried out too rigidly and religiously with little room for other concepts.

His work is one of a number of similar research projects completed by senior archaeologists and graduate students from the 1970s onwards in the Department of Archaeology, University of Ghana, Legon (Effah-Gyamfi 1980; Crossland 1986; Bredwa-Mensah 1990; Anquandah 1986, 1992). In 1977, Anquandah (1992) initiated the Accra Plains Archaeological and Historical Project to study Iron Age settlements. In addition to using survey and excavation to generate data, he conducted research on existing potting activities in three Shai settlements: Doryumu, Kodiabe and Agomanya. He identified multiple-zoned decoration as a common feature of Dangme-Shai pottery and he used this to establish a connection between the three settlements (Anquandah 1992:5). He did the ethnoarchaeological study because he wanted to use the knowledge derived from the answers to his questions and personal observations on potting to determine his research design on archaeological pottery, and the field methodology for investigating archaeological sites associated with Dangme-Shai (Anquandah, pers. comm).

Why Archaeologists do their own Ethnographic Pottery Studies in Ghana Besides archaeology, information on pottery can be gained from ethnography and research on the traditional pottery industry of Ghana is not the preserve of archaeologists. Ethnographic pottery studies derive their origin from the work of non-archaeologists who are historians and cultural anthropologists (e. g. Quarcoo and Johnson 1968; Priddy 1974). Whenever applicable, archaeologists have used information from these sources for direct historical or general analogical inferences. But, in many cases the assumptions about the archaeological value of the sources have been affected by their failure to provide suitable answers to many archaeological questions. The non-archaeologists study pottery industries first, perhaps, to satisfy their innate curiosities, and secondly to provide information useful for historical discourses, public education, and heritage preservation. Because the perspectives are not designed for producing middlerange-theories for archaeological purposes, gaps in such works are not only to be expected but also they have to be filled through ethnoarchaeological research by archaeologists (David and Kramer 2001:1-2).

In the 1990s, undergraduate students of archaeology were encouraged to investigate potting industries in villages and towns in Ghana (Mensah 1990; Ainooson 1993; Amevor 1993; Mensah 1994; Boateng 1996; Akuoku 1998). The literature shows that the study of traditional potting by archaeologists has extended into the twenty first century (Agbeko 2004; Atta-Quayson 2007). Since the 1970s, it has therefore become fashionable and acceptable for archaeologists to conduct ethnoarchaeological investigations on local potting tradition. One theory is that the data gathered from the ethnographic present serves as an important baseline against which the description, analysis, and interpretation of the archaeological pottery can be made (see David and Kramer 2001: 303-328 for relevant case studies and discussion).

The desire by archaeologists to investigate on-going potting traditions in many villages and also a few towns in Ghana can be described as a response to the admonition by Kleindienst and Watson (1957:77) that, “intensive analysis of the archaeological material is not sufficient for generating interpretation of the past and is possible only by analogy with living cultures.” In their opinion, archaeologists should investigate living communities to compile ethnographic data relevant to the specific needs of the discipline (see also David and Kramer 2001). With reference to pottery analysis, it is suggested that, rather than spending several hours, days, months, or even years in the laboratory deriving insights from the analysis of potsherds, ethnoarchaeological studies could provide knowledge that can inform the analysis and improve interpretation.

Another theory to explain the growth in ethnographic potting studies is that, “the number of communities in which pottery manufacture continues to flourish within the technological system is dwindling due to the availability of durable metal and plastic containers” (Rice 1987:113). Hence this has provided the impetus for the documentation of potting traditions encountered in the field. This author, personally, was influenced by these theories to do a complete ethnographic study when the two potters were found working in Salaga in 1994. The study was also completed to satisfy the archaeologists curiosity on what could be learnt about the presence and activity of the potters. Perhaps it can be said that curiosity about potters and their activities is found in

Tringham’s (1978:170) definition of ethnoarchaeology, “as the structure for a series of observations on behavioural patterns of living societies . . . designed to answer archaeologically-oriented questions,” describes the conceptual framework for research on traditional pottery by archaeologists in Ghana. For example, at 122

J. AKO OKORO: CLAY TOYS OF THE GRANDCHILDREN OF A POTTER IN SALAGA: INSIGHTS FOR ARCHAEOLOGY IN GHANA bits of clay given to the toddlers could be described as ‘toys’ or playthings in the hands and eyes of the toddlers.

most archaeologists and hence it could be interpreted as a major force driving research on the traditional potting industry of Ghana.

Under a programme initiated by the Ghana Museums and Monuments Board in the 1970s, i.e. the completion of a survey of traditional pottery making in Northern Ghana, Barbara Priddy (1974) investigated nine potting villages; Puffien, Ampur, and Charia in the Upper West region, and Fumbisi, Bongo, Kamiega, Binduri, Nakun and Widana in the Upper East regions of Ghana. The objective was to identify potters who “were to work in a newly constructed craftwork area on the grounds of the museum and sell their products in the attached shops” (Priddy 1974:52). Although the research was done to derive findings for practical applications in public education and heritage preservation, the range of issues reported upon makes her work a valuable one for archaeological purposes.

Looking for Children in the Ethnographic Literature on Pottery The human behaviour and cultural contexts regarding the transfer of potting traditions within families of potting groups has been of interest to both archaeologists and non-archaeologists. Normally, the desire is to understand how the learning and transfer of potting skills results in continuity and change in pottery cultural traditions. The ethnographic and ethnoarchaeological literature shows a transfer of potting from female potters to their female children in Ghana. During his research in four potting centres in the Begho area in the Brong Ahafo Region, Crossland (1989:53) noted that: “Pottery making is an hereditary craft passed down from mother to daughter”. He did not observe the role of children in making any items although he identified the support and cooperation between fathers, mothers and children in the performance of various potting tasks. Fathers gather clay and mothers do the fabrication, firing and selling (Crossland 1989:53). The children could have assisted in any of these stages and activities associated with household potting activity.

Priddy (1974:44) noted about the villages she visited that, “hand-made pottery is used for every conceivable purpose: there are enormous jars for brewing and storing corn-beer and storing water; . . . covered in knobsglobular pots with lids and ventilation holes for storing salt and meat; tiny toy pots for children – the list is endless, . . .” Priddy was giving a list of all the clay items made in the villages and fortunately for this study, the “tiny toy pots” for children were included. She does not say whether the toy pots were produced by the adults or by teenage and small girls. If produced by the adult potters, it introduces an interesting twist to the concept that underpins this study. The report shows that the production of clay toys formed part of the items that the adult potters produced for sale. It also suggests that toy clay pots were displayed and sold in the market, or in the homes of the potters, from where adults could buy them for their children. Alternatively, the children could have bought the toy pots themselves from the potters.

This review failed to find archaeological research on traditional pottery that highlighted the role or even the presence of children. Rather, references to children were found in the ethnographic research conducted by nonarchaeologists, specifically Quarcoo and Johnson (1968) at Shai in the Eastern Region, and Priddy (1974) in the Upper East and Upper West Regions. Quarcoo and Johnson (1968:66-67) recorded the following features of Shai potting: Pot making was one of the ‘domestic duties’ of Shai women; from childhood a girl is surrounded by pots and potting activity; this exposure of girls to the art and craft kindles an interest in potting in them; and, that girls learn pottery from observation and a limited amount of ‘actual teaching.’ Further points of relevance are that; first as toddlers, girls are given bits of clay to play with as they sit by their mothers, and other adult relatives doing potting; by the age of six to ten, the girls are already employed in the beating and kneading of clay; and finally, they learn to sprinkle water on heaps of clay brought from the pits.

If the adult potters did not produce the toy pots, the children of the potters could qualify for the position. The children of the potters are perceived as persons who used their rudimentary potting skills to make an input to the products of the household. It could mean that children produced toy pots for children in a fashion similar to the results gathered from Salaga. In terms of the gender and age of the children, although no information is provided by Priddy, the guess of this author is that they were girls, and could be aged fifteen years at the maximum. If they are children who had lived in potting homes from childhood, they could produce such items from age five and above. This author hesitates to include boys because there is no data to make that analogy. The ethnographic literature is mute on the issue of boys as potters. Whether made by adults or children, the key point is that tiny pots were produced purposely as toys. Furthermore, because Priddy (1974:44) indicates that, the list of hand-made pots is endless, it is it acceptable for one to guess that other clay toys (aside from pot forms) could have been present even if not on display for a researcher to see.

If Quarcoo and Johnson (1968) had identified the items that the toddlers (girls) made from the bits of clay they were given to play with, and what happened to the items, they would have provided some useful comparative data but they did not do this. Incidentally, when they made their observations on girls above six years, they shift their attention to the labour support that the girls provided and not the clay items they made. This author is inclined to assume that in Shai, girls (including toddlers) made clay toys not only because were they surrounded by pots and potting activity from their childhood, but also because they actually ‘play’ with bits of potting clay. Moreover, given the description by Quarcoo and Johnson (1968), the 123

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA interview that I did using the Twi language without an interpreter. Aside from presenting the interview in a raw form, I have endeavoured to transcribe and translate the answers of Faram and Nabeo literally to maintain their originality and relevance. Information on places mentioned and definitions of local terms have been provided in parentheses.

Is the Grand Daughter the Child of the Grandmother? One of the key questions that confronted this author when examining the various ethnographic reports was whether the mention of ‘children’ was always a reference to direct biological off spring. With regard to a potter and the children associated with her, this author was interested in using the word ‘child’ in a specific sense to refer to a son or daughter of a male and female human parent. Because the basic concept describes potting skills as being learnt from ‘mothers,’ there is the danger of the word ‘mother’ being applied widely to include grandmothers and great grand mothers. Culturally, these are all mothers of a daughter but biologically they are not. For this study, the distinction is critical because biologically and culturally, a mother is not the same as a grandmother. The term mother is used here as a title for addressing the female human parent through whom the child was borne.

Q1. Can you tell me the things you have produced? Ans. a mobile phone (#1), a bucket (#6), a mortar (#2), a top grinder stone (#7), a pestle (#3), a base grinding stone (#8), a microphone (#4), a table (#9), a scooter and its rider (#5), a chair (#10) (Figure 3). Q2. Why did you use clay to make a cell phone? Ans. I want to use it to call my sister who is far away in Accra. (Accra is the capital of Ghana and the distance from Accra, on the coast, to Salaga is about 600km).

Secondly, between a daughter and a mother there is roughly a minimum age interval of two decades and a maximum of four decades assuming a woman first gives birth at age 20 years and 40 years respectively. This interval has implications for the differences in experience, exposure, worldviews and aspirations between a daughter and a mother. The age interval widens when the same daughter is related to the grandmother. Furthermore, the age gap between the daughter and the great grand mother becomes even wider because there are two generations separating the world of a great grand daughter from that of the great grand mother.

Q3. What about the mortar and pestle? Ans. I will use it to pound fufu. (Fufu is a popular Ghanaian dish made from yam, cassava, cocoyam or plantain and is normally served with soup. In Northern Ghana, yam is used). Q4. What about the microphone? Ans. I will use it to sing. When I see people using the microphone in singing I admire it. Q5. Is it a man or woman who is riding the scooter you have made? Ans. A man.

Intellectually, the age differences introduce a generation gap in the perceptions and knowledge of the aged parents, and the children who live with them as grand daughters. This is one area where this study makes a contribution to pottery studies in Ghana. It questions the assumption that, because pottery making is a hereditary craft, it is passed down from mothers (only) to daughters (only). It also challenges concepts of potting traditions that describe elderly women as working at home with their children (mainly girls). The women (potters) may not be the ‘mothers’ of the girls.

Q6. Why did you not put a woman on the scooter instead? Ans. It is a man that came into my mind. Q7. What is this object? Ans. It is a bucket Q9. What kind of bucket is it? Ans. It is a baby-bath-bucket. babies in this bucket.

Mothers bath their

Interaction with the Grandchild of the Potter Q9. What will you use the grinding stone for? Ans. I will use it to grind pepper.

At the age of fourteen years, Faram is still in primary school class six instead of being in Junior High School class two or three in line with the educational system of Ghana. Her father is one of the three sons of the potter. In the middle of 2002, the mother of Faram died and Faram was transferred to Salaga to live with the grandmother because the father had passed away earlier on. When this author first saw Faram in January 2003, she had lived with the potter for only six months and had made her own collection of ten clay items. To this author’s surprise, not only had the potter fired five of the ten items but also she had kept all the toys in one pot in a room meant for pot storage. After observing the items for a while, I asked Faram a few questions during an

Q10. Why did you make a table and chair? Ans. I will sit on it and do my homework. Q11. What will you want to be when you complete school? Ans. I like hairdressing because it makes women look nice. Q12. Who taught you how to use clay to make objects? Ans. It is my grandmother. 124

J. AKO OKORO: CLAY TOYS OF THE GRANDCHILDREN OF A POTTER IN SALAGA: INSIGHTS FOR ARCHAEOLOGY IN GHANA

Figure 3. The 10 clay toys manufactured by Faram arranged from top left as follows: a mobile phone (#1), a mortar (#2), a pestle for the mortar (#3), a microphone (#4), a scooter and the rider (#5), a baby bath bucket (#6), a base grinding stone (#7), a top grinding stone (#8), a table (#9) and a chair (#10).

An Overview of the Information Obtained from Faram

Q13. Who determined the things that you have made? Ans. I did them myself.

The total number of single items listed for Faram is ten but conceptually, seven composite items are evident on the basis of the information derived from the interview (Table 5 also see Figure 4). The presence of civil servants, business-people, officials of a bank, a few Non Governmental Organizations (NGO’s), and Church organizations in Salaga results in a high use of mobile phones. The mortar and pestle are common household equipment and Faram sees fufu pounding activity on a daily basis. The microphone is used in churches, mosques, at marriage ceremonies and festivals in Salaga. The grinding equipment (a base stone and a top grinder) is normally placed in a stationary position inside the kitchen or within the compound. It may be placed on the ground or on top of a 50-80cm high platform. When placed on the ground, women sit on a short seat to grind but they kneel or stoop when it is set on a platform. Items that are ground include vegetables such as tomato and pepper, nuts and herbal medicine.

Q14. Does your grandmother make the kind of things you have made? Ans. No. Q15. Would you like to do the work of your grandmother when you grow? Ans. Yes. Q16. But, earlier on you mentioned hair dressing which according to you makes women look nice? Ans. I will do my grandmother’s work, but I will do the hair dressing when I am free. Q17. Why did you make these things? Ans. I made some for myself. The rest I will sell to my friends in school. They asked me to make toys for them to play with. Because the items were needed for closer examinations and analysis, they were purchased at the end of the interview. This was after she had indicated the prices for which she would sell them to some of her playmates at school and at home.

What Faram makes in clay, in (concept #6), is neither a representation of a plastic or metal bucket of a type that is commonly used in the house of the potter. Instead, Faram is using clay to represent a traditional item that is actually fashioned from clay. It is the only item in the list that carries a direct connection between item and its 125

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

Figure 4. A grouping of Faram’s 10 single items into 7 functional and composite items arranged from top left: cell phone (#1), Mortar and pestle (#2), microphone (#3), stone grinding set (#4), baby-bath-bucket (#5), scooter and rider (#6), table and chair (#7).

Concept # 1 2 3 4 5 6 7

Concept item/description

Purpose/meaning

Cell phone Mortar and pestle Microphone A set of stone grinding equipment A scooter and its rider A baby-bath-bucket A table and a chair

Making and receiving call Pounding fufu Singing Grinding pepper

Concept framework or aspect of society Tele-communication/status Food preparation/female activity Music/mass communication Food preparation (soup)

Representing Landlord Bathing babies Do homework

Status/transport Child care Furniture/school education

Table 5. The concepts, purposes or uses of the items produced by Faram

learning in the classroom and the doing of assignments at home. It shows that Faram had captured not only the furniture of school, but also, the uses to which they are put. Mobile phones and motorbikes are associated with status in Salaga. The ownership of a bicycle by a smallscale farmer may come through hard work, and a period of savings. People see bicycle ownership as a sign of improved economic standing of the owner.

actual regular material component (also see Table 9). In homes in Salaga and surrounding villages, the claybucket is used for bathing babies. This traditional practice is considered spiritually useful for babies. The clay item showing a scooter and its rider is very interesting. Before Faram’s work was seen the motorbikes that had been observed in Salaga were the Honda, Suzuki and Yamaha types in which the rider sits on the bike with his/her legs open and placed on opposite sides. Thus, this author was at a loss to understanding the type of motorbike that Faram had made in clay and hence she was asked questions about the motorbike including the gender of the rider (Figure 5). Also, this author was unable to determine the source of her inspirations until her answer indicated that it was a representation of the owner of the house (the Landlord) where she and her great grandmother live. The table and chair relate to

Practically, however, the bicycle is more of a necessity than a luxury in Salaga. It increases ones mobility and reduces the time in commuting to the farm, mosque, or a nearby village. It is used in carrying human beings, farm produce, firewood, and gallon jugs full of water from the borehole, well, or dam site. People who can afford the motorbikes have higher recognition. Not many local people have their own cars. Local residents normally graduate from ownership of bicycles to motorbikes with 126

J. AKO OKORO: CLAY TOYS OF THE GRANDCHILDREN OF A POTTER IN SALAGA: INSIGHTS FOR ARCHAEOLOGY IN GHANA the time the potter was visited on January 22, 2003. Further inquiries showed that Nabeo had not made any clay toys yet. A year later in February 2004, this author travelled to Salaga again and went to the house of the potter unannounced. Nabeo had since made ten toys that been fired by the potter. Below is a transcript of a short interview conducted with her. Q1. Can you show me the things that you have made? Ans. a mobile phone (#1) a base grinding stone (#6) a seat (#2) a top grinding stone (#7) a cup (#3) a foam mattress (#8) a kalangu drum (#4) a pillow (#9) a kalangu drum-stick (#5) a pillow (#10) (see Figure 6). Q2. Why did you use clay to make a mobile phone? Ans. I see people using it on their ear and talking. Figure 5. An example of a scooter and its rider.

Q3. What about the seat? Ans. People will sit on it to sell things in the market

only a few moving up to own cars. Ownership of a motorbike is therefore an important terminal status for most residents of Salaga and the surrounding villages.

Q4. What about the cup? Ans. I will use it for drinking water and beverage.

Interactions with Nabeo, the Great Grandchild of the Potter

Q5. What will you use the grinding stone for? Ans. I will use it to grind pepper to prepare soup.

Twelve-year-old Nabeo is the granddaughter of the dressmaking daughter of the Potter. Nabeo is therefore the great grand daughter of the potter. According to the Potter, Nabeo had lived in the house for only six weeks at

Q6. What about the kalangu drum and stick? Ans. I see them being used by some men to play music for people to dance at ceremonies.

Figure 6. The 10 clay toys items manufactured by Nabeo arranged from top left: a mobile phone (#1), a seat (#2), a cup (#3), a kalangu drum (#4), the drum stick for the kalangu drum (#5), a base grinding stone (#6), a top grinding stone (#7), a foam mattress (#8), a pillow (#9) and a pillow (#10).

127

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA Ans. Yes.

Q8. What about the foam mattress? Ans. A baby will sleep on it.

Q14. But, earlier on you stated that you would like to be a nurse? Ans. I will do both.

Q9. Who taught you how to use clay to make objects? Ans. My great-grandmother.

An Overview of the Information Obtained from Nabeo

Q10. Who determined the things that you have made? Ans. Myself.

After living with her great grandmother for only a year, Nabeo had moved from a girl who could not make clay items in 2003 to one who made toys to reflect her gradual acquisition of potting skills. Unlike Faram, Nabeo did not make the items to meet a demand from her class and play mates. This shows the diversity of factors motivating Faram and Nabeo to produce clay toys.

Q11. Did any one make a request for you to make any of the items for him or her? Ans. No Q12. What would you want to be when you complete school? Ans. A nurse.

From the ten items of Nabeo, six single or composite items are discernible as shown in Table 6 and Figure 7. Nabeo made the mobile phone because she sees it being

Q13. Will you like to do the work of your great grandmother when you grow?

Concept # 1

Item/Description

Purposes/Uses

Aspect of household or society

Cell phone

Communication

2 3 4

Seat Cup Drum and stick

Making and receiving phone calls Seating Drinking water and other liquids Musical sound production

5

A set of grinding stone equipment A foam mattress

Grinding pepper and other food substances Sleeping

6

Furniture Food and Nutrition Music, Cultural ceremonies and Entertainment Food preparation Furniture

Table 6. The concepts and aspects of household and society depicted by Nabeo

Figure 7. A grouping of Nabeo’s’s 10 single items into 6 functional and composite items and arranged from top left: cell phone (#1), seat (#2), cup (#3), drum and its stick (#4), stone grinding set (#5), foam mattress with pillows (#6).

128

J. AKO OKORO: CLAY TOYS OF THE GRANDCHILDREN OF A POTTER IN SALAGA: INSIGHTS FOR ARCHAEOLOGY IN GHANA On its own, or in accompaniment with other instruments, this drum is played at marriage, child naming, and religious ceremonies. There are persons in Salaga who earn their living from playing the ‘armpit drum’ and other traditional musical instruments. The material components of the ‘armpit drum’ include wood, leather/skin, cloth and string. These are all organic materials that do not preserve on archaeological sites in Ghana. The presence of the grinding equipment shows food preparation and domestic activities performed by women. The appearance of items associated with food preparation in the list of items of both girls is significant. They see this activity on a daily basis and often perform grinding duties as part of the training given to them by the potter. Clay is providing a material for moulding toys that convey single or multiple meanings. The toys also reflect the thought process of the children who made them. An example of this is provided by Nabeo’s description of one of the toys as a foam mattress with two pillows. Normally, mattresses and pillows are soft items for human comfort. But, these are presented in clay that has been hardened into ‘stones’ by fire. When she told the author that the rectangular-shaped clay piece measuring 6cm long, 4cm wide, and 1.2cm high was a foam mattress (or vono mattress as it is commonly called in Ghana), this was surprising because that possibility had never crossed this author’s mind.

Figure 8. A demonstration of how the hourglass instrument (kalangu) is played.

used by adults. Unlike Faram, she did not relate the reason for making the mobile phone to the desire to call her siblings elsewhere. This is another difference evident in the ideas behind the making of a mobile phone by Faram and Nabeo. The potter did not own a mobile phone at the time Faram made the clay toy phone in January 2003. Nabeo had also made a type of seat that is one of the two types found in homes in Salaga. These are the low-seat of about 30cm high on which women sit when they are cooking, bathing a child, washing, or conversing. The other one is a high seat which stands to about 80cm. The latter is normally suitable for sitting behind tables that are about 1m high. Market women and dressmakers normally use this seat. There are no clay cups in contemporary use in either Salaga or the surrounding villages.

Moreover, it is the toy-clay-foam-mattress and the accompanying two pillows that attracted the most surprise from the people who saw the items on display in the author’s office. A colleague in the Department of Archaeology, Legon told me that: “One could easily take these things that Nabeo calls pillows for ‘game pieces’. I am surprised at the ethnographic information it carries” (J. Boachie-Ansah, personal communication, Dec. 3, 2005). Another colleague, L. B. Crossland, was equally surprised at what he termed as the ‘soft concepts’ behind the hard rectangular or square-shaped clay piece of Nabeo (personal communication, February 14, 2006). For greater insight on these pieces, their dimensions are presented in Table 7 (also see Figures 9 and 10). One (A) measures 2cm long, 1.5cm wide and 0.5cm high. The other (B) measures 1.5cm on the two long sides and is 0.5cm thick. According to Crossland the rectangular or square shape of the items is significant because, “if they had been oval in shape, several interpretations could be given to them” (L. B. Crossland, personal communication, February 14, 2006). At New Buipe, grounded, smoothed or chipped clay potsherds were described either as potting tools, or items for cleaning bodies in line with ethnographic evidence of such uses for potsherds in the settlements found in the area (York 1973). At the Begho-B2 site, Crossland (1989) initially described such clay pieces as counters, but after weighing them, he agreed with Garrard (1980) in suggesting that they were used as gold weights.

The cup Nabeo made could be representing a plastic or metal type. It could also represent a ceramic mug made in Ghana or imported. The handle she has attached to the toy cup makes it suitable for use not only in drinking water and liquid foods such as porridge, but also indigenous herbal medicine. The toy clay hourglass shaped drum made by Nabeo is a typical traditional musical instrument found in the predominantly Muslim population of Salaga called kalangu in Hausa, and longni in Gonja. The drum is tucked in the armpit of the drummer who strikes one of the two leather surfaces with the drumstick (see Figure 8). The way the drummer uses his arm to increase or relax the pressure on the strings of the drum determines the sound produced.

129

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA Item Mattress

Length 6cm

Pillow A Pillow B

2cm 1.5cm

Breadth 4.5cm and 4cm 1.5cm 1.5cm

Width 1.2cm

Remarks Rectangular

0.5cm 0.5cm

Rectangular Square

resting or sleeping. However, this is where the typical adult-oriented interpretation of pottery ends because according to Nabeo, it is a bed for a child. This means, although one can describe the toy mattress and pillows as tiny objects or miniature representation of the larger real ones, according to the ‘manufacturer’ they carry the additional message of usage by a child who can be described in this circumstance as a ‘tiny human being.’ Hence a further acknowledgement of childhood is deliberately being made.

Table 7. The dimensions of the Clay Mattress and two Pillows made by Nabeo

Discussion Several other important points can be made. These are concerned with, first, the proportional relationship of composite toys and second, the findings from a comparative analysis of the toys made by the two girls. Third, the effectiveness of enculturation can be questioned between adults and children when the generation gap is excessively wide as in the case of the potter and her grand daughters and where enculturation is a basic requirement for the transfer of knowledge in traditional societies. Fourth, the interest of the girls in taking up potting as a career as adults can be examined. Fifth, the aspects of household and society represented by the toys can be further considered and finally, a summary of the variety of raw materials that are represented by clay is examined.

Figure 9. The toy foam mattress and two pillows made by Nabeo shown individually

i) Proportional Relationships: There are instances of a lack of proportional relationships of artefact they produce between the composite toys and their regular or original ones. An example is provided by the size of the toy ‘armpit drum’ and the drumstick produced by Nabeo. Normally, the drumstick is a small stick bent at one end to form an angle for hitting one of the surfaces of the leather covering the two ends of the drum. The stick is of a lightweight so as not to burden the arms of the drummer. In the case of Nabeo’s toys, the drumstick with a length of 11cm is far bigger than the 6cm long drum (see Figure 11). The lack of proportional relationship between composite toys as compared to the original objects is one of the ways that clay works of children may be identified in the archaeological record. It shows lack of an appreciation of the value of proportion and this may apply even in instances where they are moulding things they are seeing. Although Nabeo was aged 12 years and had lived with the potter for a year, she fails to ensure proportional relationship in some of her toys. In this respect, the potting experience of Nabeo can be compared to a five year old girl who has lived in a potting home all her life.

Figure 10. The toy foam mattress and pillows of Nabeo as a composite item

Without the benefit of the concept behind a particular item made by the children, description and interpretation of the toy clay foam mattresses and pillows is difficult. Though made in small sizes between 1cm and 6cm, Nabeo is representing larger household furniture. The normal mattress is 1.5m or more wide, 2m long and up to 25cm thick. In line with the normal size relationships, Nabeo made the mattress larger than the pillows. The ability of an archaeological pottery expert to decipher these details and linkages is impossible to guess. Yet these are functional insights that can be derived from this study of children’s use of clay. The mattress and the two pillows represent household furniture and specifically,

ii) Comparing the Common Themes in the Toys of Faram and Nabeo: (a) The Mobile Phones: The mobile phone made by Faram measures 8cm long, 3cm wide, and 0.7cm thick (Figure 7). It has three knobs arranged in a line in the middle, and a hole on the top right hand corner of the face into which Faram put a straw to represent an antenna. There is another feature that looks like an antennae or a 130

J. AKO OKORO: CLAY TOYS OF THE GRANDCHILDREN OF A POTTER IN SALAGA: INSIGHTS FOR ARCHAEOLOGY IN GHANA

Figure 11. The regular kalangu and its accompanying stick (top) compared with Nabeo’s toys (bottom) to show the lack of proportional representation in the toys.

control knob on the left side top end that does not appear evident on regular phones. The dimensions of the two mobile phones are shown in Table 8 (also see Figure 12). The mobile phone “made” by Nabeo is about the same size and thickness as that of Faram. It measures 8cm long, 5cm wide and 8cm thick. Discernible similarities are that Nabeo’s phone has two features covering the face. One half has five knobs while the other has punctuations all over the surface and the position of the antennae is on the right hand side. The ‘antennae’ was broken before it was collected from Nabeo. However, the point of attachment is visible. It shows that a straw was pushed through the base of the ‘handset’ and into the “antenna” to hold the latter firmly attached to the ‘handset.’ Faram’s toy mobile phone also has a straw holding the antennae made from clay placed in a vertical angle in relation to the lie of the phone. Producer Faram Nabeo

Length 8cm 8cm

Breadth 3cm 5cm

Figure 12. The clay toy mobile phones of Faram (left or #1) and that of Nabeo (right or #2) compared.

iii) Enculturation and the Generation Gap between the Children and the Potter: During the examination of the toys made by the girls, this author was interested in two things. One is the overt and covert input by the potter to the creation of concepts and the moulding of the toys. The other is to identify independent thought and creativity by the girls and discern processes of enculturation in potting traditions. Enculturation in potting is a process whereby knowledge, skill or tradition is transferred from older members of a family, household, or community to younger ones. Enculturation can be done through invitation, inducements and rewards, compulsion, sanctions and punishment. Whichever method is used, the aim is to get a younger generation to think, behave, and do things in a way that ensures continuity of traditions.

Width/thickness 0.7cm 0.8cm

The potter does several things such as giving the girls unhindered access to the clay, actually teaching them how to mould things in general, and helping them to make things that particularly mimic her work. She also gets them to sit down and look at her perform various potting activities such as preparing the clay for work, moulding pots and decorating them. Furthermore, she fires some of the items the children produce and she keeps the toys safely in the pot-room. She notes that she does these things to give the children a sense of recognition and acceptance.

Table 8. Comparison of the dimensions of the mobile phones made by Faram and Nabeo

b) The Grinding Equipment: The second common item made by both girls is the grinding stone equipment which shows grinding activity as a concept in their minds. Several reasons can account for this. Grinding activity goes on daily in the house and it normally involves the grinding of pepper, vegetables, nuts, herbs, and cereals into paste or powder forms. At their ages, the teenage girls could be called upon to perform grinding tasks. Grinding may also be a common form of childs play among girls who are practicing how to cook. The use of stones as grinders is common because it enables the girls to grind grasses, leaves, roots, seeds and even flowers. When they are able to find vegetables and spices on the ground or from the kitchen it makes the ‘cooking’ more realistic.

When both girls were asked the question, “who taught you how to use clay to make objects” (see Question 11 of Faram, and 9 of Nabeo), they both mentioned the potter. However, when this author wanted to know the one who determined the things they made, they pointed at themselves (see Question 12 of Faram, and 10 of Nabeo). 131

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA interview, Faram seems to have retracted her lack of a career interest in potting by saying that she would do potting alongside her hairdressing job. What remains clear is that none of the girls selected potting as a first choice career.

There is therefore an inter-play between elements of enculturation and the generation gap between the potter and her grand children. Although there is actual teaching by the potter, some of the perspectives that influenced the works of the girls are not known to her. An obvious example is the mobile phones produced by both girls.

The ethnographic literature on potting industries in Ghana confirms this observation at Salaga. During her research in Northern Ghana, Priddy (1974) noted that there was a steady demand for pots in the communities that could provide a basis for sustainable local production, but she added that the women making the pots were old and there was no assurance that replacements were possible because the young girls did not like pot making. The girls complained that although, “potting is hard work,” it neither brings any attractive income nor enables them to interact with a wider spectrum of members of the community (especially the males) as is found for instance in the brewing and selling of locally brewed alcoholic beverages (Priddy 1974:50).

The use of mobile phones in Salaga coincided with the generation of the potter’s daughter and with subsequent ones, not the generation of the potter herself. Thus, the girls are aware of the use of phones. While the old potter continues to make items typical of her trade for a living, the girls are on the other hand using clay to make a ‘modern’ gadget that has been in use in Salaga since 1999. By 2003, mobile phones had become a popular medium of communication in the town. Margaret Mead is among the first people to call attention to this issue of gaps between an older generation and next or subsequent ones. She notes that, “today, nowhere in the world are there elders who know what the children know; no matter how remote and simple the societies in which the children live. In the past there were always some elders who knew more than any children in terms of their experience of having grown up within the culture system. Today there are none” (Mead 1970:77-78). The potter does not know many things that her grandchildren know of. The world into which the granddaughter (Faram, 14 years), and great granddaughter (Nabeo, 12 years) of the potter were born would remain a distant one to the potter. She sees toys of ‘recent things’ made in clay by her grandchildren who obtain ideas and concepts from the larger community.

Can a liberal traditional society be able to make girls take up potting as a profession? To what extent have western education, Christianity, Islam, and modernization made traditional potting less attractive to girls? These are some of the questions that engaged the mind of this author when he examined the responses of Faram and Nabeo though no attempt is made to address these questions in this paper. However, it is clear that in Ghana there are no strict traditional rules for girls to do potting. The only exception occurs amongst the Dangme-Shai of the Eastern Region of Ghana (Quarcoo and Johnson 1968). From ancient times, the Dangme-Shai used traditional conventions, ceremonies, and sanctions to secure a future for the potting industry of the area. In the past it was only women who had performed the puberty rites (dipo) and were thus permitted to dig clay meant for potting (Quarcoo and Johnson 1968:55). The dipo was in the first place a compulsory rite of passage for all girls. But to ensure their total compliance in the learning of potting, all participants were under an obligation to use the period of confinement associated with the rites to perfect their skills in potting. Quarcoo and Johnson (1968:66-67) established the point that a Shai woman was regarded as a true native of the soil only when she was able to make pots. In other words, “it is practically obligatory for every girl to learn the art of pottery-making in order to be accepted as a full member of the community” (Priddy 1974:41).

iv) Will Faram and Nabeo continue with the Potting Tradition of the Potter? One reason for enculturation is to ensure continuity and this was a critical concern for the potter when she was interviewed in January 1994. She had given up any hope of finding a replacement because her only daughter did not learn potting. A feature of the potting heritage of the potter is that direct daughters normally fail to learn. The mother of the potter did not learn potting, which is why the Potter learnt it from her grandmother. Although the potter had wanted to pass on the tradition to her own daughter, she failed because the daughter preferred dressmaking. The entry of Faram and Nabeo into the life and household of the Potter gave her the hope of continuity but from the analysis of this author’s interaction with the girls this might be doubtful. The circumstances of the grandchildren becoming the “inheritors” of the potting tradition of the potter was by accident. Faram lost both parents and, as an orphan, she was given to the grandmother. In the same way, Nabeo lost her mother and came to live with her great grandmother. Even at their ages, and with all the benefit of being associated with the potter the girls did not mention potting as their number one future career option. Faram has identified Hairdressing and Nabeo has set her eyes on Nursing. Their preferences for these careers may change in future but what may not change is their interest in doing something other than potting. From the

Formerly, this traditional practice worked to perfection not only in getting the girls to know about potting but also in encouraging them to use it as a career. This was at a time when most girls lacked access to western education or opportunities to travel to urban areas. But at the time of the research by Quarcoo and Johnson (1968), there were enough signs that western education was impacting negatively on the future of potting in DangmeShai and elsewhere. The educated young girls of Dangme-Shai of the 1960’s were not active in potting. That is why Quarcoo and Johnson (1968:71) recorded 132

J. AKO OKORO: CLAY TOYS OF THE GRANDCHILDREN OF A POTTER IN SALAGA: INSIGHTS FOR ARCHAEOLOGY IN GHANA occasional celebrations. Hence the archaeological and ethnographic study of clay toys could, therefore, contribute to the achievement of the central core function of archaeology – that is, the reconstructing of ancient lifeways and cultural processes. vi) Variety of Materials Represented by the Clay Toys: An examination of the raw materials that normally form the components of the items that the toys represent shows an interesting variety. However, clay has made it possible for these to be captured in toys as shown in Table 10. The list shows that it is only the baby bucket that is actually produced with clay. It is an example of what can be described as a specialized clay container recognized for use in bathing babies in a traditional way. The use of the clay bath bucket connotes only childcare and also relates to traditional beliefs and practices that ensure the welfare and proper growth of the baby. This is an ongoing cultural practice in many homes in Salaga and surrounding villages. Those who use the clay bucket to bath babies do so not because they do not have metal or plastic substitutes but because tradition dictates that the practice continues. Otherwise the materials represented in clay are not actually clay related.

that, “most, if not all, of the middle-aged women who derive their livelihood from potting are illiterates.” After examining the circumstances of educated girls in the area, Quarcoo and Johnson (1968:72) concluded that, “the question arises whether the present generation of girls will in future find pottery a worth-while career.” Another question could be added as to whether it would be possible for the girls to find potting as a worthwhile hobby or a part time career? The answers for these questions are not encouraging. For example, at Kpandu, Amevor (1993:40) found that, “the younger girls who should have learned potting from their mothers and grandmothers, do not show any interest at all in the industry with the excuse that it will make them dirty.” Western education and modernization seem not to accord with the typical rural nature of traditional potting. The career choices made by Faram and Nabeo suggest that potting was not a first choice. Once educated, many rural girls look for jobs that give them better recognition or enable them to move to urban centres. There is little or no chance that Faram and Nabeo would do potting as adults. But in order to answer this, maybe it would be necessary to do a check on this after 15-20 years when the girls would become adults with defined careers. Aspects Communication Music and production

sound

Furniture Water, nutrition

Food

and

Cultural Ceremonies Food preparation/gender roles Schooling Child care Transport and status Trade by women in the market

Examples Mobile phone, Microphone Hour glass drum and Stick, Microphone Table, Chair, Foam Mattress, Pillow Drinking Cup, Baby bath bucket, Stone grinders Hour glass Stone grinders Table and Chair Baby bath bucket Motorbike showing Landlord, Mobile phone Seat

Material Wood

Made by Faram, Nabeo Nabeo

Leather/skin Metal Plastic/Rubber Foam Stone Clay

Faram, Nabeo Faram, Nabeo

Examples Drum, drum stick, table, chair, bed, bench, mortar and pestle Drum Cell phones, cup, motorbike Cell phone, motor bike Mattress, pillow Grinding equipment Clay baby bath bucket

Table 10: The variety of raw material components represented in the clay toys

Nabeo Faram, Nabeo Faram Faram Faram

Conclusions Three insights are discernible from the potting tradition of the septuagenarian potter. The first is that not all daughters of potters become interested in learning potting from their mother. The second insight is that, when they learn potting, it is not automatic that they take it up as an occupation. Thirdly, while the daughter(s) of potters can choose to stay away from the potting activities of their mothers, the same cannot be said of the grand and great granddaughters who are obliged to do so from the perspective of their parents and also their relationship with the potter.

Nabeo

Table 9. A general list of aspects of society shown in the clay toys made by the grandchildren

v) The Aspects of Society Represented by the Toys: A list of the aspects of the society represented in the clay toys of Faram and Nabeo is provided in Table 9. It includes mobile communication, furniture, childcare and female household activities. A longer list can be made when the multiple representations that some items carry is taken into consideration. The wealth of information about society that can be derived from the study of clay toys made by children is potentially enormous. For example the microphone is used for singing and preaching in churches and mosques in the town. The microphone is also used at social gatherings, annual, and

Several reasons account for men and women sending their children to live with their mothers and fathers (that is the grandfather and grandmother of the children). One of the things a potter would want to bequeath to her grand children (who get to live with her) is the potting tradition. Similarly, a mother who fails to learn potting may harbour the aspiration that her daughter may become interested in potting through living with the grandmotherpotter.

133

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA The study of the clay toys has shown an important attribute of clay which is its versatility. The children have therefore used clay as a universal material to represent nineteen items that are normally made from non-clay material and one item made from clay. Children and grandchildren of potters get access to moist clay being used by mothers, sisters and other relations. It has been shown that potters who are busily making pots can give clay bits to toddlers to play with to allow the potters to concentrate on their work. As they grow up and become knowledgeable in conceptualizing ideas and moulding them into reality, the potters’ clay becomes a material for the physical expression of ideas and the display of aspirations. Such children could make a great variety of toys at different times and ages. For the potters, the ability of ‘children’ to make toys is a signal that they are picking up the potting skill.

Acknowledgements

Pottery analysis may remain the dominant perspective for archaeological research in Ghana but the findings from the toys call for the need to expand the list of known pottery artefacts. They show lifeways and material culture associated with children in particular and society in general. The miniature size of clay toys is the most distinguishing characteristic of such items. Another one is the lack of proportional relationships between composite toys and their attributes as seen with the example of the drum and the drumstick which has been cited. A third feature relates to the rough exterior and interior features of the toys because burnishing and meticulous smoothing may not be important to the children. Potters do these for technical and aesthetic reasons.

AGBEKO, A. S. 2004. The Pottery Industry of Torgome. Unpublished B. A. Thesis, Department of Archaeology, University of Ghana, Legon.

I owe much gratitude to the old Potter whose worry over the lack of a successor for her potting activity in January 1994 created the framework for me to take a serious note of the presence of the grandchildren in her house in 2003. I also thank my local assistants, Issah Seidu and Moses Seidu of Salaga, for being available at all times to offer the local support that I needed. The surprise shown and comment made by Professor James Boachie-Ansah and Mr. L. B. Crossland of the Department of Archaeology, University of Ghana, Legon, and Mrs. Inga Merkyte of the University of Copenhagen, Denmark on the clay toys that I displayed on a table in my office, provided additional impetus for me to write this paper. References

AGORSAH, E. K. 1976. Some Consideration on the Archaeology and Traditional History of the Begho area. Unpublished M. A. Thesis. Department of Archaeology, University of Ghana, Legon. AKUOKU, M. K. 1998. Pottery Industry of Mpraeso Kwahu. Unpublished B. A. Thesis, Department of Archaeology, University of Ghana, Legon. AMEVOR, P. L. K. 1993. Traditional Pottery Industry of Kpandu. Unpublished B.A. Thesis, Department of Archaeology, University of Ghana, Legon.

Some of the clay toys that were collected had not been fired. But whether fired and unfired, it is incumbent on researchers to look out for the handiwork of children in the pottery database. These insights were communicated to Mr William Gblerkpor of the Department of Archaeology when he was analysing pottery from ancient settlements on the Krobo Mountain in the Eastern Region of Ghana for his MPhil thesis (Gblerkpor 2005; Gblerkpor this volume). He saw pottery artefacts that he was inclined to associate with the works of children rather than adults (William Gblerkpor, pers. comm. February 12, 2004).

ANQUANDAH, J. R. 1992. Accra Plains Dangmeland: A Case Study in the Eclectic Approach to Archaeological and Historical Studies. Archaeology in Ghana 3: 1-7. ANQUANDAH, J. R. 1986. Ethnoarchaeological Clues to Ghana’s Great Past. Universitas 8: 15-17. ANQUANDAH, J. R. 1982. Rediscovering Ghana’s Past. Accra: Sedco Publishing Ltd. ANQUANDAH, J. R. 1967. Cultural Developments in Western African in the light of Pottery Studies. Unpublished M. Litt Thesis, University of Oxford, Oxford.

In conclusion, it is stated that in both ancient and contemporary times clay has been a useful material for moulding objects. Clay toys do not only have a long antiquity in Ghana but in general terms they represent one of the fore runners of modern toys. They show that in ancient times potters and their children and grand children were producers of toys for their communities. Contemporary clay toys form part of the ethnographic pottery database that would eventually become part of the archaeological pottery record. Because clay toys reflect the material culture and lifeways of the children, the household and the community, their study has the potential to establish a new and interesting archaeological research agenda in Ghana and elsewhere.

AINOOSON, J. 1993: Traditional Pottery Industry of Okwenya. Unpublished B. A. Thesis, Department of Archaeology, University of Ghana, Legon. ARYEE, N. N. A. 1998. Decorative Designs on Ladoku, Ayawaso and Dawu Awukugua Pottery. Unpublished B. A. Thesis, Department of Archaeology University of Ghana, Legon. ATTA-QUAYSON, A. T. 2007. The Wood Carving and Pottery Industries of the African Market at the Arts 134

J. AKO OKORO: CLAY TOYS OF THE GRANDCHILDREN OF A POTTER IN SALAGA: INSIGHTS FOR ARCHAEOLOGY IN GHANA Centre in Accra. Unpublished B. A. Thesis, Department of Archaeology, University of Ghana, Legon.

Unpublished MPhil Thesis, Department of Archaeology, University of Ghana, Legon.

BELLIS, J. O. 1972. Archaeology and the Culture history of the Akan of Ghana. Unpublished PhD. Thesis, Indiana State University, Indiana, USA.

GHANA STATISTICAL SERVICE. 2005. 2000 Population and Housing Census of Ghana. Accra: Ghana Publishing Corporation.

BOACHIE-ANSAH, J. 2007. Traditions and Changes in Pottery from Mount Mary Training College and AdjikpoYokunya, Eastern Region, Ghana. EthnogragraphischArchaologische. Zietschrift 14: 83-107.

KLEINDIENST, M. R. and WATSON, P. J. 1956. “Action Archaeology”: The Archaeological Inventory of a Living Community. Anthropology Tomorrow 5: 75-78. MATHEWSON, R. D. 1968. The Painted Pottery Sequence in the Volta Basin. West African Archaeological Newsletter 8: 24-31.

BOACHIE-ANSAH, J. 2000. Funerary Terracotta Figurines and Pottery of the Akan of Ghana. In: Ethnogragraphisch-Archaologische. Zietschrift 14: 6582.

MEAD, M. 1970. “Letters from the Field 1925-1975.” New York: Harper and Row.

BOACHIE-ANSAH, J. 1998. Preliminary Report on Exavations at Wodoku, East Legon, Accra, Ghana. West African Journal of Archaeology 28: 1-18.

MENSAH, A. K. 1994. Traditional Pottery Industry of Vume. Unpublished B. A. Thesis, Department of Archaeology, University of Ghana, Legon.

BOACHIE-ANSAH, J. 1986. An Archaeological Contribution to the History of Wenchi. African Occasional Papers No. 3. Calgary: Calgary University Press.

MENSAH, A. M. 1990. An Ethnographic Study of potting at Dodowa, Shai and its Relevance for the Archaeology of the Shai Area. Unpublished B. A. Thesis, Department of Archaeology, University of Ghana, Legon.

BOACHIE-ANSAH, J. 1985. Pottery from Bonoso and Ahwene Koko, Ghana. West African Journal of Archaeology 17: 41-72.

OKORO, J. A. In Press (a). Some Perspectives influencing Research on the Traditional Pottery Industry of Ghana. Ethnogragraphisch-Archaologische Zietschrift.

BOATENG, T. 1996. The Pottery Industry of Esiease. Unpublished B. A. Thesis, Department of Archaeology, University of Ghana, Legon.

OKORO, J. A. In Press (b). Reflections on the Oral Traditions of the Nterapo of the Salaga Area. History in Africa 35.

BREDWAH-MENSAH, Y. 1990. An Archaeological investigation conducted at Okai Koi Hill (Ayawaso) and its significance for Iron Age Archaeology in Ghana. Unpublished M.Phil Thesis, Department of Archaeology, University of Ghana, Legon.

OKORO, J. A. Forthcoming. Archaeology of Indigenous Water Management in Salaga. PhD. Thesis, Department of Anthropology, University of Toronto, Toronto.

CROSSLAND, L. B. 1989. Pottery from the Begho-B2 site, Ghana. African Occasional Papers No. 4. Calgary: The University of Calgary Press.

OKORO, J. A. 2003a. Archival Research on Salaga at the Public Records Office, London. WARA Newsletter p. 13. African Studies Center, University of Florida, USA.

DAVID, N. and KRAMER, C. 2001. Ethnoarchaeology in Action. Cambridge: Cambridge University Press.

OKORO, J. A. 2003b. Research on Water and Slaves in Salaga. Nyame Akuma 59: 45-53.

DAVIES, O. 1980. The Ntereso Culture of Ghana. (In), Swartz, B. K. and Dumett, R. A. (eds.), West African Culture Dynamics. The Hague: Mouton, pp. 205-225.

PRIDDY, B. 1974. Pottery in Upper Region. National Museum Occasional Papers 6-8: 41-53. QUARCOO, A. K. and JOHNSON, M. 1968. Shai Pots: The Pottery industry of the Shai People of Southern Ghana. Baessler-Archiv, Neue Folge, Band 16: 47-87.

DAVIES, O. 1964. Gonja Painted Pottery. Transactions of the Historical Society of Ghana 7: 4-11. EFFAH-GYAMFI, K. 1980. Traditional Pottery Technology at Krobo, Techiman, Ghana: An Ethnoarchaeological Study. West African Journal of Archaeology 10: 103-116.

RENFREW, C. and BAHN, P. 1996. Archaeology: Theories, Methods and Practice. New York: Thames and Hudson. RICE, P. 1987. Pottery Analysis: A Sourcebook. Chicago: University of Chicago Press.

GBLERKPOR, W. N. 2005. An Archaeological investigation of the Krobo Mountain Dry Stone-Terraces. 135

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA SHAW, T. 1961. Excavations at Dawu. Edinburgh: Edinburgh University Press. SHINNIE P. and KENSE, F. J. 1989. Archaeology of Gonja: Excavations at Daboya. Calgary: University of Calgary Press. TAWIAH, A. M. 1988. An Archaeological Survey at the Wodoku Site. Unpublished B.A. Thesis, Department of Archaeology, University of Ghana, Legon. TRINGHAM, R. 1978. Experimentation, Ethnoarchaeology, and the Leapfrogs in Archaeological Methodology. (In), Gould, R. A. (ed.), Explorations in Ethnoarchaeology. Albuquerque: Albuquerque Press, pp. 169-199. YORK, R. N. 1973. Excavations at New Buipe. West African Journal of Archaeology 3:1-189.

136

The Late Stone Age in Ghana: The Re-excavation of Bosumpra Cave in Context Derek Watson Dept. of Archaeology, University of Ghana, Legon, P.O. Box LG3, Legon, Ghana

are inconclusive, but preliminary varve counts from the upper sediments suggest a date of ca. 3,000bp which is more consistent with the later date suggested by Russell et al., (2003; cf. Shanahan et al., 2006: 296). This arid phase was followed by a return to humid conditions with Late Holocene lake level highstands recorded at ca. 2,200bp and 1,700bp (Shanahan et al., 2006). A study of sea-surface temperatures (SST), based on ratios of δ18O and Mg/Ca in planktonic foraminifera, from a core extracted off-shore near Cameroon, indicates a similar pattern with warm periods between 12,000-8,700bp and 7,500-6,700bp, with temperatures ca. ≤27.5 oC, followed by gradual cooling until 2,500bp with the lowest SST (ca. 25 0C) occurring between 3,000-2,500bp (Weldeab et al., 2005: 982-983). A subsequent rapid rise in SST occurred at 2 ka, with a less pronounced decrease from 1,7001,000bp (ibid.).

Introduction Since the inception of the Dept. of Archaeology at the University of Ghana in 1951, the Late Stone Age (LSA) was one of the main focal points of archaeological investigations. In recent years, the focus of research has shifted towards Historical and ‘Iron Age’ archaeology, and Ethnoarchaeology. Despite the considerable efforts of a number of stalwart archaeologists over the last few decades, especially Oliver Davies (e.g. 1964, 1967, nd), the LSA in Ghana remains poorly documented. The publication of this volume provides a timely opportunity to present a critical review of LSA research in Ghana within the wider context of West Africa. This article also takes the step of announcing the commencement of Phase 1 of the “Late Stone Age and Early Iron Age Forest Occupations Project” in southern Ghana, which has as its focus the re-excavation of Bosumpra Cave in the Eastern Region.

After ca. 9000bp, moist semi-deciduous forests were established around Bosumtwi during a period of decreased seasonality, characterised by high precipitation and longer wet seasons (Maley 1991; Talbot et al., 1984). Indeed, it appears that ‘forests’ persisted throughout most of the Holocene even during arid periods as annual precipitation may have stabilised at ca. 1200 mm/year which might have been evenly distributed throughout the year (ibid.). However, unpublished data from K.R.M. Beuning and M.R. Talbot quoted by Russell et al., (2003: 7) indicates a marked increase in the remains of grasses above the 3200bp transition at Bosumtwi suggesting the “spread of grassland at the expense of closed-canopy forest”. This is part of a widely documented and nonlinear environmental decline across the Sahara and adjacent areas from ca. 6,000bp, involving the southwards retreat of the monsoon system and the development of an abrupt arid phase from ca. 5,000bp that led to the formation of the present day Sahara (cf. Brooks et al., 2005; Hassan 1997). To the south, according to Salzmann and Hoelzmann (2005), palynological data from sediment cores from Lac Sélé in southern Benin (Figure 1) indicate an abrupt climatic change to drier conditions and the subsequent spread of Sudano-Guinean savannas and deterioration of rainforest vegetation within the general area from 4,500 - 3,400 cal. BP. This led to the development of the Dahomey Gap and was followed by a return to wetter conditions from c. 3,300-1,100 cal. BP (ibid.). The timing and impact of this ‘aridification event’ was variable across West Africa to the Congo Basin (see reviews in Russell et al., 2003:7; Salzmann and Hoelzmann 2005: 197; Shanahan et al., 2006: 297-299).

Holocene Palaeoecological Conditions Palaeoecological data available for Ghana, and the savanna/forest(s) in general, is limited. This summary will be confined to the few studies relevant to the reconstruction of palaeoecological conditions obtaining during the Holocene in the area of modern Ghana. Studies of pollen and leaf fossil, mineralogy and sedimentology of sediment cores extracted from Lake Bosumtwi (Figure 1) by Talbot et al., (1984: 188-190) suggested high lake levels and low mean annual temperatures between 12,500-9,000bp, followed by a period of increased humidity and higher lake levels prior to ca. 5,000bp. Subsequently, a precipitous lake regression of ≥120m occurred at ca. 3,750bp, and this arid phase may have been part of a general climatic shift that eventually established a longer dry season and a more seasonally differentiated climatic regime similar to present-day conditions within the region (Maley 1991: 91–92; Talbot et al., 1984: 189-190). However, Russell et al., (2003) reassessed this data and argue that the nitrogen isotope composition and core sedimentology indicates that the onset of pronounced change to more arid and seasonally climatic conditions in this region should be redated to c. 3,200bp. Recent geomorphological analysis and radiocarbon dating of in situ charcoal deposits at Lake Bosumtwi by Shanahan et al., (2006) suggests deep overflowing lake conditions from ca. 11,000bp with a significant lake regression occurring ca. 8,800bp which lasted until ca. 7,200bp when deeper lake levels were again re-established. Unfortunately, existing data and dates for the termination of the Holocene humid period 137

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA

Figure 1: Location of sites mentioned in the text and contemporary ecological zones (after Church 1980 and White 1983).

pre-occupied with this tradition (e.g. Casey 2000; Davies 1980; Dombrowski 1976; Flight 1967; Stahl 1985; Watson 2005a). To date, about 50 sites (rock shelters and open air sites) have been located but many comprise undated surface scatters and less than 20 have been excavated (Watson in prep). The known site distribution (cf. Flight 1976) evinces a general pattern of lowland (open air) riverine settings and upland rock shelters. Open air sites range in size from large material scatters of ca. 60ha (Daboya) to smaller sites of ca. 11.53ha - 0.11ha (Boyasi, Birimi), often with variable rectilinear or circlular concentrations of burnt daub (some with ‘pole’ impressions) and/or foundation stones suggesting wattle and daub structures (e.g. Bonoase, Mumute, Ntereso; Anquandah 1976, 1982, 1993, This volume; Casey 1998, 2000; Davies 1967, 1973; Dombrowski 1976, 1980; Gavua 1985). Some of the ‘village’ sites appear to have been deliberately destroyed (e.g. Casey 2000: 119-121) perhaps on a seasonal basis. Rock shelters, confined to the southern half of Ghana, may have also been (re)occupied seasonally; some degree of sympatry with local foragers is indicated by the co-occurrence of mixed Punpun/Kintampo material culture at many of these sites in central Ghana (see below; cf. Watson 2005a). However, Kintampo material culture, site distribution and internal spatial arrangements, and its subsistence economy remain poorly understood and inadequately described as most available reports provide only relatively vague and cursory descriptions or interpretations (e.g. Dombrowski 1976), and few projects have undertaken systematic sampling for organic remains (cf. D’Andrea et al., 2001, 2005, 2007; Stahl 1985; Watson 2005a). Radiocarbon dates for the Kintampo cluster within the early to mid 2nd millennium BC

D’Andrea and Casey’s (2002: 152-154) review of the environmental requirements of domesticated pearl millet (Pennisetum glaucum) recovered from the Kintampo site of Birimi (Figure 1) and palaeoecological research in contiguous areas (e.g. Burkina Faso, K6 in central Ghana) suggests that the ecological conditions obtaining during the Kintampo period (see below) were broadly similar to those found within modern Ghana, i.e. drier wooded savanna in the north with more humid forested area(s) to the south. As the present day distribution of the forest in Ghana (Figure 1) is partly a result of the formation of the Dahomey gap it is possible that the forest extended significantly farther north prior to 3,400-3,200bp, especially during humid periods. The Kintampo Tradition The earliest evidence for food production and semipermanent ‘villages’ in the savanna/forest of southern West Africa are associated with the Kintampo Tradition of Ghana (3,600-3,200bp). Originally described by Oliver Davies (1962) principally on the basis of surface collections, its distinctive material culture has been located throughout Ghana from the Gambaga Escarpment (i.e. Birimi) to the Coast (Figure 1) within a variety of contemporary ecological zones. Remarkably, its distribution appears predominantly confined to the area of modern Ghana, as few diagnostic artefacts have been found in Togo (de Barros 1983) and Côte d’Ivoire (Chenorkian 1983), whilst no trace has yet been discovered in Burkina Faso (cf. Watson 2005a). The Kintampo is the most well defined ceramic LSA Tradition in Ghana, and perhaps in southern West Africa. Indeed, most LSA research undertaken in Ghana has been 138

DEREK WATSON: THE LATE STONE AGE IN GHANA: THE RE-EXCAVATION OF BOSUMPRA CAVE IN CONTEXT

Site

Birimi

Daboya *

Ntereso *

Layer/Spit/Horizon

cal BP

95.4%

95.4%

B-099308

3310 ± 110

1890 - 1387

3839 - 3336

Kintampo Tradition

B-099306

3240 ± 90

1742 - 1316

3691 - 3265

45-50cm (pit)

Kintampo Tradition

TO-8173

2960 ± 370

2199 - 235

4148 - 2184

50-60cm (pit)

Kintampo Tradition

TO-8172

3460 ± 200

2433 - 1316

4382 - 3265

62cm (pit)

Kintampo Tradition

B-099307

3550 ± 40

2016 - 1756

3965 - 3705

120-130cm (daub layer)

Kintampo Tradition

B-104756

3490 ± 50

1937 - 1690

3886 - 3639

RM 20(14) Layer 5

Kintampo Tradition

S-2370

3405 ± 155

2141 - 1388

4090 - 3337

RM 20(14) Layer 6

Kintampo Tradition

6-CX

3870 ± 60

2487 - 2145

4436 - 4094

RM 20(14) Layer 7

Kintampo Tradition

S-2375

3195 ± 325

2457 - 775

4406 - 2724

RN 21(19) Layer 6

Kintampo Tradition

S-2376

4235 ± 150

3338 - 2471

5287 - 4420

Layer 3

Kintampo Tradition

SR-52

3580 ± 130

2336 - 1609

4285 - 3558

Layer 2

Kintampo Tradition

SR-81

3270 ± 100

1873 - 1316

3822 - 3265

Layer 1b (pit)

Kintampo Tradition

SR-61

3190 ± 120

1747 - 1129

3696 - 3078

I-2697

3220 ± 110

1771 - 1213

3720 - 3162

I-2698

3560 ± 100

2198 - 1641

4147 - 3590

Birm-30

3339 ± 35

1733 - 1526

3682 - 3475

I-2699

3530 ± 100

2141 - 1616

4090 - 3565

UCR 1690

3495 ± 100

2128 - 1535

4077 - 3484

Layer 8b Layer 9 Square A2 Unit 2, Layer 2b Unit 1, Layer 2b-d Unit 1 & 2, Layer 3a-b K6 Units 1, 2 & 3, Layer 4

Punpun and Kintampo Traditions Punpun and Kintampo Traditions Punpun and Kintampo Traditions Punpun and Kintampo Traditions Punpun and Kintampo Traditions Punpun and Kintampo Traditions Punpun and Kintampo Traditions Punpun and Kintampo Traditions

UCR 1691

3700 ± 90

2456 - 1831

4405 - 3780

UCR 1692

3550 ± 127

2278 - 1536

4227 - 3485

UCR 1693

3605 ± 100

2279 - 1691

4228 - 3640

Punpun Tradition ?

UCR 1694

6100 ± 250

5519 - 4463

7468 - 6412

Layer 3 (= Stahl L2bd)

Punpun and Kintampo Tradition

Birm-29

3750 ± 84

2458 - 1951

4407 - 3900

Layer 2

Punpun Tradition

Birm-31

3401 ± 74

1890 - 1521

3839-3470

GX 29105

3380 ± 40

1770 - 1532

3719 - 3481

GX 28710

3230 ± 40

1609 - 1428

3558 - 3377

TO 11883

3410 ± 60

1884 - 1536

3833 - 3485

TO 11238

3460 ± 50

1907 - 1693

3856 - 3588

GX 29106

3470 ± 40

1891 - 1689

3840 - 3638

N-1961

705 ± 75

AD1180 - 1409

771 - 541

Unit 2, Horizon 2 (EP)

Unit 2, Horizon 4 (EP) B6B Unit 2, Horizon 4 (EP) Unit 1, Horizon 2 (EP) Layer 2

Punpun and Kintampo Traditions Punpun and Kintampo Traditions Punpun and Kintampo Traditions Punpun and Kintampo Traditions Punpun and Kintampo Traditions ?

Level 5

?

N-1962

165 ± 75

ca. AD1529 onwards

from 421

Level 7

?

N-1963

125 ± 60

ca. AD1667 onwards

from 284

Level 9

?

N-1964

190 ± 75

ca. AD1522 onwards

from 429

20-30cm

?

N-1804

775 ± 75

AD1044 - 1389

907 - 561

90-100cm

?

N-1805

5370 ± 100

4437 - 3975

6386 - 5924

Deposit

?

GlF-4241

4180 ± 140

3321 - 2347

5270 - 4296

Deposit

?

N-2982

5850 ± 80

4932 - 4519

6881 - 6468

Bosumpra Cave

Gao Lagoon Kpone

Reference

D'Andrea et al., 2001; Quickert et al., 2003

Gavua 1985; Shinnie and Kense 1989

Davis 1973, 1980

Rahtz and Flight 1974

Unit 3, Layer 5

Unit 2, Horizon 5 (LP)

Apreku Rock shelter

cal BC

Kintampo Tradition

Layer 8 Square A2

Akyekyema Buor Rock shelter

Radiocarbon Date

42cm (pit)

K1

B5C

Lab No.

32-42cm (pit)

Layer 8 Square A2

K8

Affiliation

Stahl 1985

Stahl 1994; Flight 1968 Stahl 1994; Flight 1968 AORB??

D'Andrea et al., 2006, 2007; Watson 2005a

Musonda 1976

Smith 1975; Stahl 1994 Nygaard and Talbot 1977, 1984

Table 1: Radiocarbon dates for LSA sites in Ghana (calibrated using OxCal 4.0; INTCAL 4 curve). Dates from Kintampo sites marked with '*' are either inverted or stratigraphically suspect (cf. Watson 2005a: Table 2).

(Table 1; Watson 2005a, in prep). Indeed, at the Boase sites (B-sites), a series of loci excavated around four large sandstone outcrops in central Ghana (Figure 1), two apparently distinct phases of occupation were noted with an Early phase dating to approximately 1,900-1,600BC and a Late phase at ca. 1,600-1,400BC (Table 1; Watson 2005a: 10).

Kintampo material culture comprises a suite of distinctive artefacts that are variably represented at sites across Ghana: including polished edge-ground celts (nyame akuma or ‘god’s axes’, ≤10cm in length); ‘objets de parure’ such as polished stone rings, pendants, possibly lip-plugs (labrets), and biconically perforated shell or quartz discs (beads); figurative art consisting of plastic 139

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA 2007) as the relative importance of wild versus domestic resources is entirely unknown. However, this pattern of food producers relying on wild resources continues in Africa until today (e.g. Posnansky 2004: 39-40). The Kintampo is defined here as ‘agro-pastoralist’ (cf. Watson 2005a: 6-7) as until more definitive data is available, this term is meant as shorthand to emphasise the distinctive elements of this tradition’s subsistence economy in an area that was previously populated by forager groups, i.e. the exploitation of domesticated crops and livestock, and possible arboriculture.

attachments on pottery and/or figurines depicting anthropomorphic and/or animal species; ground and/or polished bifacial projectile points; flake-based geometric and non-geometric microliths; grooved stones (handy sized blocks of suitable rock with incised channels, possibly used for the manufacture of beads etc); and the enigmatic rasps or “terracotta cigars” (cf. Anquandah 1993; Watson 2005a). The latter are unique to the Kintampo, and typically are manufactured on relatively friable and fine-grained sedimentary rock rather than ‘terracotta’ as suggested by Davies (1962), but examples of the latter have been reported (e.g. Agorsah 1986). These implements are frequently scored with intersecting horizontal, vertical and/or oblique lines that are usually a few millimetres in depth; but their primary function remains unknown. Kintampo pottery was often elaborately decorated with registers of channelling, rocked/stamped (rigid) comb and rocked/impressed peigne fileté. Previously, motifs created using rigid-comb decor were thought to have been diagnostic of Kintampo pottery decoration, but peigne fileté, identified so far on only pottery from central-west Ghana (Watson 2005a: 1115, in prep) is the most frequently used decorative tool/technique. Kintampo vessels were fashioned using a beating technique(s) (e.g. paddle and anvil) to form globular and relatively large pots (diameters ca. ≤40cm, with open to wide-open apertures) with, predominantly, thickened and everted-thickened rim types (ibid.). Manufacturing and finishing techniques include the use of red-slipping, burnishing and, at the B-sites, the possible use of a reduced firing atmosphere for the production of black-ware (Watson 2005a: 11). A comparative analysis of Kintampo material culture and synthesis of available data is currently under preparation by the author (Watson in prep).

Two opposing hypotheses have been postulated concerning the origin of the Kintampo Tradition. Davies (e.g. 1964, 1966, 1980) hypothesised that the Kintampo represented an ‘intrusive’ group, originating in the Sudanic belt as its material culture (e.g. pottery decoration and bifacial projectile points) showed affinities with material from the Tilemsi Valley and the Niger bend in Mali. Subsequently, an indigenous syncretic Kintampo, coalescing via stimulus diffusion and the adoption of northern elements, was suggested by Posnansky (1984: 150) and Stahl (1985: 145-147). Alternatively, extensive correlates in Kintampo material culture and aspects of their economy with northern groups and the lack of any evidence for the interaction and valency required for syncretism formed the basis of a two stage migration and adaptation model (cf. Watson 2005a: 25-32). According to this model, the event(s) described by Davies was in principal correct, as the Kintampo formed from the migration of a northern, possibly Sahelian, population that moved rapidly southwards and re-adapted aspects of their ‘arid-based’ economy to the prevailing environmental conditions of the savanna/forests of West Africa.

Animal and plant remains have only been recovered in small quantities from Birimi, K1, K6, Ntereso and the Bsites (cf. Carter and Flight 1972; Carter in Rahtz and Flight 1974; D’Andrea et al., 2001, 2006, 2007; D’Andrea and Casey 2002; Stahl 1985; Watson 2005a) indicating a subsistence regime based on the exploitation of wild and domestic resources; with most of the latter comprising non-indigenous species introduced from the Sahara/Sahel (cf. Watson 2005a: 25). Zooarchaeological remains include various wild species and domestic ovicaprines and cattle (Carter and Flight 1972: 280; Gautier and Van Neer 2005: 196-197; Stahl 1985: 138142, 1993:265; Watson 2005a: 6-7;). Palaeobotanical data comprises incense tree (Canarium schweinfurthii), hackberry (Celtis sp.), small legumes (Fabaceae), domesticated cowpea (Vigna unguiculata), nonindigenous domesticated pearl millet (P. glaucum) (D’Andrea and Casey 2002, D’Andrea et al., 2001, 2006, 2007) and oil palm (Elaeis guineensis) (D’Andrea et al., 2006; Stahl 1985). Large quantities of oil palm remains found at the B-sites suggest the practise of arboriculture (D’Andrea et al., 2006); but it remains unknown if vegeculture was also practised by the Kintampo or other proximal groups (cf Watson 2005a: 31 - 32). Characterisation of the socio-economic basis of the Kintampo is problematic (cf. D’Andrea et al., 2006,

The Punpun Tradition The only relatively well-documented LSA forager group in Ghana is the so-called Punpun Tradition. First described by Flight (1967) as the “Buobini Culture”, it was subsequently re-named the “Punpun phase” (Rahtz and Flight 1974). No justification was ever provided for this term, but the long-term continuity evident in the pottery associated with this group at the B-sites suggests that it is more correctly defined as a tradition (cf. Watson 2005a). To date, its material culture has only been located in the hills surrounding the modern village of Kintampo in central Ghana (cf. Watson 2005a: 21). However, the material culture of this tradition has usually only been described from sites where it is mixed with Kintampo Tradition material (e.g. B-sites; K1 and K6, Figure 1; Flight 1970, 1976; Rahtz and Flight 1974; Stahl 1985; Watson 2005a). Yet Punpun material culture (lithics and ceramics) has been found without any mixed or overlying Kintampo Tradition material at K4, K7, and K8; whilst scant remains of Kintampo Tradition material at the latter site were found only on the surface (Flight 1967, 1968, 1970, 1976). Unfortunately, contextual details and material culture from these sites were only mentioned in brief by Flight (ibid.). 140

DEREK WATSON: THE LATE STONE AGE IN GHANA: THE RE-EXCAVATION OF BOSUMPRA CAVE IN CONTEXT Middle and Upper Yengemen previously cited; Coon 1968), Yagala (ca. 1070 ± 100; Atherton 1972) and Kamabai (ca. 2560 ± 115; Atherton 1972) in Sierra Leone which have non-geometric microliths and polished/unpolished celts/axes; Kokasu and Sopie in Liberia (1500 ± 200bp and 3400 ± 200bp respectively; Gabel 1976) where groundstone tools were absent; and Agarade in Togo which dates from ca. 4,200bp (de Barros 1992).

Available radiocarbon dates overlap with the Kintampo Tradition except for the date of ca. 6100bp from the ceramic bearing Layer 5 at K6 (Table 1). If not anomalous, then this suggests that the occupation of central Ghana by this group long predates the Kintampo Tradition and that these are the earliest dated ceramics in all of Ghana, and perhaps southern West Africa. Across West Africa (Figure 1), the LSA is divided into aceramic and ceramic phases (Shaw 1978/79, 1985). Pre-ceramic industries have been found in the Côte d’Ivoire at Bingerville (13,050 ± 230 bp; Chenorkian 1983); in Burkina Faso at Rim I (ca. 12,000-5000bp; Andah 1978, 1979, 1980) and at Maagada (ca. 7590 ± 90bp - 7000 ± 100bp; Breunig and Wotzka 1993); and Iwo Eleru in Nigeria (from ca. 11,200 ± 200bp; Shaw and Daniels 1984). Aceramic groundstone and microlithic industries (both geometric and non-geometric) overlain by later ceramic LSA occupations, are found at widely distributed sites; such as Korounkorokalé in Mali (the preceramic layers underlie the date 5185 ± 95bp; MacDonald 1997); Rim II in Burkina Faso (ca. 3630 ± 120bp Andah 1978); Iwo Eleru (3620 ± 60bp; Shaw and Daniels 1984), Dutsen Kongba (4905 ± 170bp; York 1974, 1978) and Afikpo II (3030 ± 75bp; Andah & Anozie 1980) in Nigeria; Yengema in Sierra Leone (TL dates: 4150 ± 470 bp - 3550 ± 350 bp; Coon 1968) and apparently at Bosumpra (ca. 5300bp, Shaw 1944; Smith 1975; Table 1) which is the only known site in Ghana that appears to span the aceramic/ceramic transition. In Ghana, the onlyother ostensibly early ceramic site is Kpone (Gao Lagoon) ca. 5800bp (Table 1; Nygaard and Talbot 1984). Ceramic technology appears across a wide area, possibly as early as the 8th millennium bp in Ghana, but by the 7th millennium on it is present at dispersed sites situated in Nigeria, Mali and Ghana. Later sites/occupations with ‘microlithic industries’ and ceramics are distributed all over West Africa, including: Yengema (see TL dates for

In regards to Punpun pottery it is worthwhile quoting Flight’s (1967: 68) original description of the material found near the modern village of Kintampo (cf. Rahtz and Flight 1974: 15): “The pottery is highly distinctive, and not to be confused even in a small sample with that of the Kintampo Culture. Almost every sherd bears some decoration, usually incised or stabbed [i.e. punctate]. In a few instances the stab-marks are filled with a red paste…Cord-impression is rare, but a useful trait for purposes of definition”. Indeed, the decoration has been emphasised as the most “diagnostic” means for differentiating it from that of the Kintampo Tradition. Stahl’s (1985) analysis of the Punpun ceramic assemblage obtained during the re-excavation of K6 emphasised the principal use of mat impression, with lower frequencies of triangular punctation, including some with impressed red pigment, and various forms of cord impression. Stahl (1985: 128-134) used differences in decoration to assign the pottery (assemblage total = ca.1014, with 553 analysed) to either the Kintampo or Punpun and excluded nearly all undecorated (body)sherds including, potentially, some of the earliest ceramics in southern West Africa i.e. only 3 of 13 sherds from layer 5 were recorded and all 5 sherds from the undated layer 6 were excluded (cf. Watson 2005a: 22-23, Table 21). As Punpun/Kintampo potsherds from the B-sites were predominantly undecorated (Watson 2005a) this suggests

Figure 2: Punpun Tradition pottery decoration from the B-sites, Locus B5C, Horizons 1-4 (cf. Watson 2005a).

141

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA wrapped rocker”, cf. Flight 1970: 71; Rahtz and Flight 1974: 15; Stahl 1985b: 134) if not mis-assigned (Watson 2005b: 65-66), evince a more varied decorative tradition than previously suspected. Punpun pottery is easily differentiated from Kintampo ceramics as the fabric is friable, sandy, and relatively coarse and the vessels were fashioned using a coiling technique (Figure 3) which is easily discernible (Watson 2005a: 11; cf. Rahtz and Flight 1974: 22; Stahl 1985: 134) Morphologically, these vessels are characterised by simple rim types and predominantly open vessels (Figure 4) with reconstructed diameters ranging from 13-42cm.

that the reconstructed sequence at K6 is unreliable (cf. Watson 2005a: 22-24, 2005b; Stahl 2005, for relevant discussion). The Punpun lithic industry is similarly ill-defined, partly due to the fact that assemblages of Punpun material are generally from mixed deposits. The most detailed characterisation provided by Flight (1970: 71; cf. Rahtz and Flight 1974: 22) for the lithic assemblage from ‘Punpun’ layers at K6 is of a “prolific quartz industry including lunates and backed pieces”. Stahl (1985: 136137) described the lithic assemblage from the later K6 reexcavation as “informal at best” and only a few flakes that “may have been utilized or retouched” were identified.

Data from the B-sites demonstrates that the pottery of the Punpun and Kintampo Traditions were entirely different in terms of their decorative motifs, morphology and in all aspects of their respective manufacturing technologies (Watson 2005a: 11-15). This suggests the existence of two technologically distinct and coeval pottery traditions, based on different chaînes opératoires (Watson 2005a: 21-22). Following Gosselain (2000), this formed the basis for inferring the existence of two socio-economic and cultural groups (foragers and agro-pastoralists) separated by some form of ‘social boundary’ (Watson 2005: 21-22).

At the B-sites (principally locus B5C), Punpun sherds were decorated only with punctate motifs (Figure 2; cf. Watson 2005a: 15) and no indications were evident for any use of red pigment. Puntate decoration was said to be “diagnostic” of the material from K1 (Rahtz and Flight 1974: 15) but it is possible that some of the motifs/techniques ascribed to the Punpun Tradition at K1 and K6 (e.g. “red pigment”, “cord impressions” and “cord

Figure 3: Punpun Tradition coil manufactured pottery. Photographs show the edge of pot sherds with concave/convex ('U-shaped') grooves/ridges indicative of the coil manufacturing technique whereby 'coils' of clay are moulded together. Nearly all Punpun sherds bear these distinctive marks on one or both opposing edges. These examples are from the B-sites, Locus B5C, Horizons 14 (Early to Late phase; cf. Watson 2005a). The examples from Horizon 1 conjoin.

Figure 4: Punpun Tradition rim types and reconstructed vessels. Examples shown are from the B-sites, Locus B5C (Late phase at top and Early phase at bottom; cf. Watson 2005a).

142

DEREK WATSON: THE LATE STONE AGE IN GHANA: THE RE-EXCAVATION OF BOSUMPRA CAVE IN CONTEXT found in Layers 6-5 of Shaw’s (1944: 3) excavation were probably intrusive and, as he found no ceramics below 100cm in his test-pit, this was inferred as evidence for an aceramic microlithic horizon. Unfortunately, Smith (1975) does not explicitly state the basal depth of the testpit, but the lowest depth stated in the article is 110cm which correlates with spit details (surface-110cm depth; compared to ca. 2m maximum depth of Shaw’s excavation) recorded on labels accompanying material culture from this excavation still held at the Dept. of Archaeology, University of Ghana. Consequently, as this site is by no means securely dated and whilst an aceramic/ceramic transition is widely documented across West Africa, Smith’s (1975) ‘aceramic phase’ should be viewed with caution as it is based on conjecture and a ca. 10cm spit.

Indeed, if we consider the radiocarbon dates for the cooccurrence of the material culture (i.e. pottery) of both groups at K1 and K6 (Table 1) in relation to the data from the B-sites, this suggests that this ‘boundary’ was maintained over a long period of time. As the largest assemblages of lithic material from the Bsites were recovered from mixed Punpun/Kintampo contexts (i.e. Loci B5c and B6B) it has not been possible to differentiate or define the industry utilised by the foragers (Watson 2005a: 15-18). The lithic industry from the B-sites comprised flake-based microliths, with a small geometric component, manufactured using core reduction (ibid.). Both traditions appear to have been using a relatively similar technology, but it is one that is common across sub-Saharan West Africa (cf. MacDonald 1997). Nevertheless, the only artefactual link discerned between the Punpun and Kintampo Traditions by Stahl (1985: 142, 144) is in the “lithic technology (e.g. struck quartz)” which she suggests “remains constant throughout the sequence”. However, the widespread use of quartz in West Africa (e.g. MacDonald 1997) means that it is difficult to accept this as evidence for continuity between these traditions (cf. Watson 2005a: 23). Yet it should be noted at this point that it is difficult to determine if the lithic industry of the Punpun or any pre- or contemporary non-Kintampo LSA group in Ghana have any technological or typological similarities or otherwise. Consequently, we must consider this in relation to available data concerning the pottery traditions and the remarkable similarities between the Kintampo (e.g. material culture and economy) and Sahelian/Saharan groups and, despite the paucity of data from the area, there are no known material similarities between the agro-pastoralist and any other LSA groups in Ghana (Watson 2005a).

Interestingly, the original description of the pottery and its stratigraphic distribution by Shaw (1944: 13-26) is reminiscent of that found at mixed Punpun and Kintampo Traditions at more northerly rock shelter sites (e.g. Bsites, K6; Stahl 1985; Watson 2005a). Shaw (1944: 1314) describes the earliest pottery (“ware A”, layers 6-1, but concentrated in layers 3-2) as comprising “straight [i.e. simple] rims”, decorated with “some system of impressed square, rectangular or oval dots, of small size” or “small punctate ornament”. Mixed with and overlying “ware A” is “ware B” (layers 3-1, but concentrated in layer 2) which has “outward curving rims [i.e. everted], sometimes with overhang, sometimes a wide flattish flange” and which are usually decorated with “grooves, comb or thumbnail”. In later articles (Shaw 1978/79: 70; Shaw and Daniels 1984: 63) the rims and decoration were described as comprising “squarish, pointed or slightly thickened rims” which are “commonly a little everted; the only decoration is by comb-stamping”. Shaw’s (1944) wares C-E appear to be restricted in distribution to layers 2-1 and may be later LSA/‘Iron Age’ assemblages. The description of wares A-B shows affinities with the pottery of the Kintampo and Punpun Traditions, although no mention is made of coil-manufacture this may not have been recognised. The lithic and pottery assemblage(s) recovered by Smith (1975) have never been analysed, but a preliminary examination of the ceramics by the author suggests that the stratigraphic sequence may be more complex than suspected as most of the material appears similar to ‘ware B’ whilst no certain ‘ware A’ sherds were present in this collection.

Late Stone Age Sites in Ghana Comparatively few non-Kintampo LSA sites have either been excavated or well documented. Indeed, no regional sequences have been developed and the chronological/spatial distribution of many artefact types or details of subsistence regimes remain sketchy or entirely unknown. To date, the most comprehensively investigated site is Bosumpra Cave situated on the Kwahu Uplands, in the Eastern Region of Ghana (Figure 1). Originally excavated by Shaw (1944) this site yielded pottery, polished celts and geometric and non-geometric quartz microliths. Smith’s (1975) smaller re-excavation (ca. 2.2m3 compared to ≥ 8m3 excavated by Shaw) provided an upper radiocarbon date of ca. 775 ± 75bp (spit - 20-30cm) associated with pottery and chipped quartz and a date of ca. 5370 ± 100bp (Table 1) from the lowest spit (90-100cm) yielding pottery and microliths; these spits correspond in depth to Shaw’s (1944) layers 1 and 3 respectively. Botanical remains comprised incense tree (C. schweinfurthii) and oil palm (E. guineensis) but these were not quantified. These details notwithstanding, the socio-cultural and technological sequence and broader relevance of this site is difficult to assess. Smith (1975: 180) suggested, quite reasonably, that the two potsherds

Two other artefact types (celts and biconically perforated stone beads) recovered from Bosumpra are often associated with the Kintampo Tradition. The celts (n=5 layers 4-2) resemble the‘ Type 1’ biconvex celts found at many Kintampo sites (Watson 2005a, in prep), but celt ‘rough-outs’ (n=6; layers 5-2) and apparent fragments (n=24, layers 6-2) suggest a deeper time-depth for the manufacture of these implements. Similarly, the biconically perforated quartz beads (n=2) appear relatively early in the sequence (i.e. layers 5-4). No rasps were found by either Shaw (1944) or Smith (1975), but these were also absent from some Kintampo Tradition sites such as K1 (cf. Rahtz and Flight 1974). 143

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA dispersed, mobile bands of hunter-gatherers that utilised few non-portable artefacts. Broad similarities appear to exist in lithic technologies and the exploitation of oily fruits (Canarium and oil palm) across West-Central Africa (cf. Cornelissen 2003; Lavachery 2001; MacDonald and Allsworth-Jones 1994, Mercader 2003a). Nevertheless, as most forager sites in Ghana have not been well documented, it is difficult to reach any definite conclusions regarding the distribution of hunter-gatherer groups/sites or potential regional/ecological-based differences in their material culture and economies. This is partly a result of the low visibility of the remains left by these groups in the archaeological record and also that few systematic attempts have been made in the last few decades to locate LSA hunter-gatherer sites in Ghana. Still, it seems probable that Ghana was inhabited by foragers from a relatively early date. Indeed, they may have been localised variant(s) of the geographically widespread “West African Microlithic Technocomplex” posited by MacDonald (1997: 192), and constituted by broad-based foragers situated in different ecological zones that were utilising geometric and non-geometric quartz microliths.

Consequently, it is difficult to be certain of the cultural affiliation or significance of the stratigraphic distribution of the material culture, but apparent similarities to the Punpun and Kintampo Traditions are remarkable. LSA sites situated on the Accra Plains/coast also suggest an early appearance of pottery. The shell midden at Kpone (Gao Lagoon; Figure 1, Table 1) yielded a “small scale flake industry” and pottery dated between 5,850 ± 80bp and 4180 ± 140bp (Nygaard and Talbot 1984: 3434) but the material and its contextual associations have never been adequately described. Similar, but undated, material was also reported at the nearby Nungua Formation, Palaeosol 2 at Asokrochona, and Onukpa Wahe stream section (Figure 1; ibid.). The undated site at Legon Botanical Gardens (Figure 1) yielded a variety of material culture including macrolithic and microlithic assemblages and may document the aceramic/ceramic transition, but the stratigraphy was extensively disturbed and the upper assemblages contained mixed ceramic LSA (a ‘Kintampo’ component is suggested by the presence of beads, celts and grooved stones) and ‘Iron Age’ materials (Davies 1968).

Oliver Davies (1964, 1967, nd) records the occurrence of a large number of, predominantly, undated sites including potential forager occupation/activity sites (mostly as surface scatters of lithics with or without associated pottery) distributed throughout Ghana (also see Shaw 1944). From extant illustrations and descriptions of assemblages it appears that some had geometric microliths (e.g. “Guinea Neolithic”), whilst others also contained bifacially flaked ‘hoes’ that form the so-called forest-based “Neolithic hoe cultures”. Davies descriptions and classifications have been extensively criticised (cf. MacDonald and Allsworth-Jones 1994; Swartz 1972) and the West African LSA was reclassified by Shaw (1978/79, 1985) under the rubric of aceramic/ceramic associated with microlithic and/or non-microlithic industries. Indeed, Davies’ (e.g. 1964, 1967, nd) designations of Middle Stone Age assemblages (e.g. “Guinea Aterian”, “Sangoan”, “Ultimate MSA” with macrolithic components and/or use of levallois technique) are also problematic and details of this period are less well known than the LSA (cf. Casey 2003). MacDonald and Allsworth-Jones (1994: 92-100) argue that many of the sites with macrolithic industries/components, including many previously designated as Sangoan, are manifestations of a widespread LSA phenomenon dating to the Terminal Pleistocene-Holocene and that sites with “core tool” or “axe preforms” (e.g. “hoe cultures”) may date between ca. 6,000-1500bp. In Ghana, this may include at least a component of the lithic assemblage(s) from an undated site near Hohoe, which yielded ‘core tools’ and apparent microliths, some of which may have been geometric (cf. Swartz 1974), and Legon Botanical Gardens (cf. Davies 1968). Moreover, contrary to Shaw’s (1978/79, 1985) suggestion that these industries were entirely limited to forested environments, similar industries have been found as far north as the Sahel (MacDonald and Allsworth-Jones 1994: 99-100).

Our knowledge of the post-Kintampo LSA and/or ‘Early Iron Age’ (EIA) socio-cultural and technological sequence in Ghana is rudimentary. Available dates for socalled EIA sites generally cluster within the early 1st millennium AD, whilst mid-to-late 1st millennium BC dates for ‘metallurgical’ sites (e.g. Daboya, Shinnie and Kense 1989) are difficult to address as such sites are generally poorly recorded and described (for a fuller discussion of EIA dating problems and sites see GodfreySmith and Casey 2003; Stahl 1994: 79-82). Scant details exist concerning the EIA and such sites are often defined simply on the basis of ceramic assemblages distinct to those found at Kintampo or ‘Late Iron Age’ sites (Stahl 1994: 79). Thus, an apparently lengthy hiatus exists between the terminal radiocarbon dates for the Kintampo (Table 1) and the beginning of the EIA. Nevertheless, there is a possibility of long-term socio-economic diversity and persistence of foragers within southern Ghana, suggested by radiocarbon dates from the rock shelters Akyekyema Bour and Apreku which range from the early to mid 2nd millennium AD (Table 1; Musonda 1976). Associated material included pottery, microliths, and polished/ground and ‘flaked’ celts/axes, similar to those found at early ceramic LSA sites, and the remains of ‘wild’ flora and fauna (the site of Tetewabon is undated but contained similar material culture). However, Musonda (1976) rejected the dates and such a recent LSA occupation as implausible. Discussion and Conclusions: The Late Stone Age in Ghana and the Re-excavation of Bosumpra The scant remains of forager groups found in Ghana are broadly similar in extent to those recorded at geographically widespread LSA hunter-gatherer sites across the savanna-forest/forested zone of West Africa (cf. Casey 2003, 2005; MacDonald 1997; Watson 2005a). It is widely assumed that these comprised small, 144

DEREK WATSON: THE LATE STONE AGE IN GHANA: THE RE-EXCAVATION OF BOSUMPRA CAVE IN CONTEXT Mercader and Brooks 2001; Van Noten 1977). In West Africa, the earliest evidence available for the appearance of an ‘LSA industry’ is the small flake industry found at Bingerville at ca. 13,000bp, with geometrics attested by ca. 11,200-9100bp at Iwo Eleru (cf. Chenorkian 1983; Shaw and Daniels 1984). As previously discussed, we also do not know when metallurgy and the so-called ‘Early Iron Age’ actually began, or any specific details concerning the ramifications of the development of this technology in Ghana. Thus, the chronology at least, of the ‘LSA’ in Ghana, and much of West Africa generally, is ambiguous and requires clarification.

Available data for Ghana are insufficient for estimating the chronology or distribution of early human occupation of the forest/savanna. Yet, any consideration of the distribution of the widespread occurrence of macrolithic/microlithic industries located in Ghana (cf. Casey 2003 Figs 1.2-1.3; Davies 1964: Figure 69, 1967; Swartz 1974) suggest that this area was inhabited by diverse forager groups utilising a variety of lithic technologies within, potentially different ecological zones. It is possible that this reflects similarly diverse economies, but floral and faunal remains are generally limited to mixed Punpun/Kintampo sites, though these indicate the exploitation of a wide range of wild animal species and oily fruits (e.g. oil palm) (cf. Stahl 1985 138142). Models of pre-Kintampo incipient cultivation, such as vege- and arbori-culture (i.e. oil palm and yams; cf. Andah 1987; Sowunmi 2002, Shaw 1976; Stahl 1984), are difficult to address, as no substantive data exists to attempt any evaluation.

In sum, many important questions raised by the archaeology of Ghana remain unresolved. Investigations of the Kintampo, perhaps the most well studied LSA group in all sub-Sahelian West Africa, are hampered by limited data concerning site organisation, subsistence basis, material culture, and uncertainty regarding the origin(s) and development of food production (Casey 1998, 2000, 2005; D’Andrea and Casey 2002; D’Andrea et al., 2001, 2006, 2007; Stahl 1985, 2005; Watson 2005a, b). However, research by D’Andea et al., (2001, 2006, 2007) demonstrates that the economic basis of the Kintampo, especially the range of domesticates exploited, is more complex and diverse than hitherto realised. Yet, the prehistory of human occupation within sub-Sahelian West African is poorly understood and even with the advent of farming and, eventually, metallurgy, we know little concerning early settlements within the savannaforests and forests. Whilst foragers and the Kintampo certainly co-occupied the forest/savanna ecotone (Watson 2005a), available archaeological data are insufficient for determining the potential diversity of forager groups, the existence of possible forest-based adaptations and associated technologies, or modelling long-term foragerfood producer interactions, the spread and development of food production, or the formation of new social/economic 'groups' within Ghana.

Recent global research on the archaeology of tropical rain/forests has emphasised the long-term occupation and diversity of socio-economic adaptations within these zones (Mercader 2003a, b). In Central Africa, settlement of savannas and (rain)forest generally began by c. 40,000bp, but there are indications that human penetration of the latter environment occurred even earlier (cf. Cornelissen 2002, 2003; Lavachery 2001; Mercader 2003b; Mercader and Martí 2003; Van Noten 1977). In sub-Sahelian West Africa, only a few widely dispersed LSA sites are known from the late Pleistocene (e.g. Bingerville in Côte d’Ivoire ca.13,000bp and Iwo Eleru ca. 11,200bp in Nigeria; Chenorkian 1983; Shaw and Daniels 1984). To date, scant data exists concerning early occupations and technological adaptations (e.g. Middle Stone Age) within this vast area and the antiquity of human occupation of the forested regions is unknown (cf. Casey 2003). Indeed, even dating the advent of the LSA in West Africa is problematic and based on relative assessments (cf. Quickert et al., 2003: 1292). Shaw (1981) suggested that the MSA in West Africa dates between 35,000-15,000bp. Indeed, Optically Stimulated Luminescence dates of MSA materials found at Birimi suggests northern Ghana was inhabited from as early as ca. 35,000bp (cf. Casey 2003; Quickert et al., 2003). Moreover, suggested dates for MSA (i.e. ‘Sangoan’) materials from Asokrochona and Tema II (Figure 1), based on the presumed deposition period of the matrix, are 25,000-20,000bp and 20,000-13,000bp respectively (Nygaard and Talbot 1984). Generally, the appearance of microlithic industries are thought to delimit the ‘beginning’ of the LSA, but this is confused by whether geometric or small-flake based microliths are viewed as diagnostic (cf. Casey 2003: 4849; Clark and Cole 1957). Indeed, both forms have variable temporal representations as small flake industries appear to form components of ‘MSA’ assemblages, and may continue until the EIA, whereas geometrics are thought to be temporally restricted with the earliest appearing at ca. 40,000bp in Central Africa and ca. 30,000bp at Shum Laka (Casey 20003; Cornelissen 2003; Lavachery et al., 1996; McBrearty and Brooks 2000;

In an attempt to address some of the issues outlined here, the author in collaboration with Dr Yaw Bredwa-Mensah, Dept. of Archaeology, University of Ghana, will conduct a series of excavations and surveys within the forested zone of southern Ghana under the rubric of the “Late Stone Age and Early Iron Age Forest Occupations Project”. The first phase of this project commences in June 2008 with the re-excavation of Bosumpra (Table 1) and we hope to combine this with a survey of 100km2 of the surrounding area to locate further appropriate sites for excavation. If the existing interpretation of the occupation sequence and extant radiocarbon dates for Bosumpra are correct then this is the only known site within Ghana which spans the aceramic/ceramic LSA transition to the EIA. This project seeks to examine long term issues relating to LSA and EIA forager and food-producer populations, and an essential aspect of this research will be characterisation of material culture(s), systematic sampling and identification of macro- and microscopic organic remains (e.g. seeds, starch grains and phytoliths), and radiocarbon dating of the occupation sequence(s) investigated. A major objective of this initial phase of the 145

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA project is to begin to develop a regional archaeological sequence for the Kwahu Uplands which constitute the main physical relief of southern Ghana.

ANQUANDAH, J. 1982. Rediscovering Ghana's Past. Harlow: Longman Group Ltd. ANQUANDAH, J. 1993. The Kintampo Complex: A Case Study of Early Sedentism and Food Production in sub-Sahelian West Africa. (In), Shaw, T., Sinclair, P., Andah, B. and Okpoko, A. (eds.), The Archaeology of Africa: Food, Metals and Towns. London: Routledge, pp. 253-260.

Anthropological and archaeological research on foragers and ethnicity in other parts of Africa, and globally, has re-examined the history of forager-farmer relationships and revised the ways we understand forager communities and the processes by which different socio-economic and ethnic groups interact and/or develop (cf. Amselle 1990; Gosselain 2000; Grinker 1994; Kent 2003; Panter-Brick et al., 2001; Stark 1998). These issues are of paramount importance, as unlike elsewhere in sub-Saharan Africa, West African foragers disappeared, within the 1st millennium AD (cf. MacDonald 1998) or later, either by displacement or acculturation by farmers. These issues have formed major research topics elsewhere on this continent (e.g. Kent 2003), but as our knowledge of these peoples of sub-Sahelian West Africa consists of only a basic outline of their prehistory this hampers discussion of these topics. It is hoped that the Forest Occupations Project will enable insight into the socio-economic adaptations of forager/food producers, the long-term impact of the penetration of agriculturalists and also allow us to begin reconstruction of the ethnic and economic mosaics existing within the savanna/forests of the West African Late Stone Age and Early Iron Age.

ATHERTON, J. H. 1972. Excavations at Kamabai and Yagala Rock Shelters, Sierra Leone. West African Journal of Archaeology 2: 39-74. BREUNIG, P. and WOTZKA, H.-P. 1993. Archäologische untersuchungen im südosten Burkina Fasos 1989/90: vorbericht über die erste grabungskampagne des Frankfurter Sonderforschungsbereiches 268 "Westafrikanische Savanne". Beiträge zur Allgemeinen und Vergleichendedn Archäologie 11: 145-187. BROOKS, N., CHIAPELLO, I., DI LERNIA, S., DRAKE, N., LEGRAND, M., MOULIN, C., and PROSPERO, J. 2005. The Climate-Environment-Society Nexus in the Sahara from Prehistoric Times to the Present Day. The Journal of North African Studies 10: 253–292. CARTER, P. L. and FLIGHT, C. 1972. A Report on the Fauna from the Sites of Ntereso and Kintampo Rock Shelter Six in Ghana with Evidence for the Practice of Animal Husbandry During the Second Millennium BC. Man 7: 277-282.

References AGORSAH, E. K. 1986. Material Characterisation of “Kintampo cigars”. Nyame Akuma 27: 6-9.

CASEY, J. 1998. Just a Formality: The Presence of Fancy Projectile Points in a Basic Tool Assemblage. (In), Kent, S. (ed.), Gender in African Prehistory. Walnut Creek, Ca: Altimira, pp. 83-103.

AMSELLE, J-L. 1990. Mestizo Logics: Anthropology of Identity in Africa and Elsewhere. Stanford: Stanford University Press. ANDAH, B. W. and ANOZIE, F. 1980. Preliminary Report on the Prehistoric Site of Afikpo (Nigeria). West African Journal of Archaeology 10: 83-102.

CASEY, J. 2000. The Kintampo Complex: The Late Holocene on the Gambaga Escarpment. Northern Ghana. (BAR International Series 906, Cambridge Monographs in African Archaeology). Oxford: BAR Publishing.

ANDAH, B. W. 1978. Excavations at Rim, Upper Volta. West African Journal of Archaeology 8: 75-138.

CASEY, J. 2003. The Archaeology of West Africa from the Pleistocene to the Mid-Holocene. (In), Mercader, J, (ed.), Under the Canopy. The Archaeology of Tropical Rainforests. New Brunswick: Rutgers University Press, pp 35–63.

ANDAH, B. W. 1979. The Later Stone Age and Neolithic of Upper Volta Reviewed in a West African Context. West African Journal of Archaeology 9: 87-110. ANDAH, B. W. 1980. Excavations at Rim, north-central Upper Volta: A Palaeoecological Study. (In), Swartz, B. K., and Dumett, R. A. (eds.), West African Culture Dynamics. The Hague: Mouton, pp. 41-65.

CASEY, J. 2005. Holocene Occupations of the Forest and Savanna. (In), Stahl, A. B. (ed.) African Archaeology: A Critical Introduction. Oxford: Blackwell, pp. 225-248.

ANDAH, B. W. 1987. Agricultural Beginnings and Early Farming Communities in West and Central Africa. West African Journal of Archaeology 17: 171-204.

CHENORKIAN, R. 1983. Ivory Coast Prehistory: Recent Developments. African Archaeological Review 1: 127142.

ANQUANDAH, J. 1976. Boyasi Hill - a Kintampo ‘Neolithic’ Village Site in the Forest of Ghana. Sankofa 2: 92.

CHURCH, R. J. H. 1980. West Africa: A Study of the Environment and Man's Use of It. (8th Edition). London: Longman. 146

DEREK WATSON: THE LATE STONE AGE IN GHANA: THE RE-EXCAVATION OF BOSUMPRA CAVE IN CONTEXT DE BARROS, P. 1983. A Kintampo Rasp Fragment from Bassar (Togo). Nyame Akuma 23: 35-36.

CLARK, J. D. and COLE, S. 1957. The Third Panafrican Congress on Prehistory, Livingstone 1955. London: Chatto and Windus.

DE BARROS, P. 1992. Preliminary Report on Excavations at Agarade Rockshelter, Togo, West Africa. Unpublished paper presented at the 11th Biennial Society of Africanist Archaeologists Conference, March 26-29, 1992, Los Angeles, California.

CORNELISSEN, E. 2002. Human Responses to Changing Environments in Central Africa between 40,000 and 12,000 B.P. Journal of World Prehistory 16: 197 - 235. CORNELISSEN, E. 2003. On Microlithic Quartz Industries at the End of the Pleistocene in Central Africa: The Evidence from Shum Laka (NW Cameroon). African Archaeological Review 20: 1-24.

DOMBROWSKI, J. C. (1976). Mumute and Bonoase Two Sites of the Kintampo Industry. Sankofa 2: 64-71. DOMBROWSKI, J. C. 1980. Earliest Settlements in Ghana: The Kintampo Industry. (In), Leakey, R. E., and Ogot P. (eds.), Proceedings of the 8th Panafrican Congress of Prehistory and Quaternary Studies, Nairobi, September 1977. Nairobi: International Louis Leakey Memorial Institute for African Prehistory, pp. 261-262.

COON, C. S. 1968. Yengema Cave Report. Museum Monographs. Philadelphia: The University Museum. D’ANDREA, A.C., and CASEY, J. 2002. Pearl Millet and Kintampo Subsistence. African Archaeological Review 19: 147–173.

FLIGHT, C. 1967. Prehistoric Sequence in the Kintampo Area of Ghana. In, Hugot, H. J. (ed.), Actes de 6e Session, Congrès Panafrican de Prehistoire. Chambery, Les Imprimes Reunies de Chambery, pp. 68-69.

D'ANDREA, A. C., KLEE, M., and CASEY. J. 2001. Archaeobotanical Evidence for Pearl Millet (Pennisetum glaucum) in sub-Saharan West Africa. Antiquity 75: 341348.

FLIGHT, C. 1968. Kintampo 1967. West African Archaeological Newsletter 8: 15-20.

D’ANDREA, A.C., LOGAN, A.L., and WATSON, D. J. 2006. Oil Palm and Prehistoric Subsistence in Tropical West Africa. Journal of African Archaeology 4: 195-222.

FLIGHT, C. 1970. Excavations at Kintampo. West African Archaeological Newsletter 12: 71-72.

DAVIES, O. nd. Ghana Field Notes. Unpublished Manuscript, Dept. Of Archaeology, University of Ghana.

FLIGHT, C. 1976. The Kintampo Culture and its Place in the Economic Prehistory of West Africa. (In), Harlan, J. R., deWet, J. M. J., and Stemler, A. B. L. (eds.), Origins of African plant Domestication. The Hague: Mouton, pp. 211–221.

DAVIES, O. 1962. Neolithic Cultures of Ghana. (In), Mortelmans, G., and Nenquin, J. (eds.), Actes du lVe Congres Panafricain de Prehistoire et l'Etude du Quaternaire (1959). Tervuren: Museé Royal de l'Afrique Centrale, pp. 291-302.

GABEL, C. 1976. Microlithic Occurrences in the Republic of Liberia. West African Journal of Archaeology 6: 21-35.

DAVIES, O. 1964. The Quaternary in the Coastlands of Guinea. Glasgow: Jackson, Son and Co.

GAUTIER, A. and VAN NEER, W. 2005. The Continuous Exploitation of Wild Animal Resources in the Archaeozoological Record of Ghana. Journal of African Archaeology 3: 195-212.

DAVIES, O. 1966. The Invasion of Ghana from the Sahara in the Early Iron Age. (In), Cuscoy L. D. (ed.), Actas del V Congreso Panafricano de Prehistoria Estudio del Cauternario. Santa Cruz de Tenerife: Museo Arqueologico Santa Cruz de Tenerife, pp. 27-42.

GAVUA, K. 1985. Daboya and the Kintampo Culture of Ghana. Unpublished M.A. Thesis, Department of Archaeology, University of Calgary.

DAVIES, O. 1967. West Africa before the Europeans. London: Methuen. DAVIES, O. 1968. Mesoneolithic Excavations at Legon and New Todzi (Ghana). Bull. Inst. Fond. Afr. Noire (B) 30: 1147-1194.

GODFREY-SMITH, D. I. and CASEY, J. 2003. Direct Thermoluminescence Chronology for Early Iron Age Smelting Technology on the Gambaga Escarpment, Ghana. Journal of Archaeological Science 30: 10371050.

DAVIES, O. 1973. Excavations at Ntereso, Gonja, Northern Ghana. Final Report. Unpublished Report, Natal Museum, Pietermaritzburg, South Africa.

GOSSELAIN, O. P. 2000. Materializing Identities: An African Perspective. Journal of Archaeological Method and Theory 7: 187-215.

DAVIES, O. 1980. The Ntesero Culture in Ghana. (In), Swartz, B. K., and Dumett, R. A. (eds.), West African Culture Dynamics. The Hague: Mouton, pp. 205-225.

GRINKER, R. R. 1994. Houses in the Rainforest: Ethnicity and Inequality among Farmers and Foragers in Central Africa. Berkeley: University of California Press. 147

CURRENT ARCHAEOLOGICAL RESEARCH IN GHANA MUSONDA, F. B. 1976. The Later Stone Age in Ghana in the Light of Excavations along the Voltaian Scarp. Unpublished M.A. Thesis, Department of Archaeology, University of Ghana.

HASSAN, F. A. 1997. Holocene Palaeoclimates of Africa. African Archaeological Review 14: 213–230. LAVACHERY, P. 2001. The Holocene Archaeological Sequence of Shum Laka Rock Shelter (Grasslands, Western Cameroon). African Archaeological Review 18: 213-247.

NYGAARD, S. E., and TALBOT, M. R. 1984. Stone Age Archaeology and Environment on the Southern Accra Plains, Ghana. Norwegian Archaeological Review 17: 19-38.

KENT, S. (ed.) 2003. Ethnicity, Hunter-Gatherers, and the “Other”. Association or Assimilation in Africa. Washington (D. C.): Smithsonian Institution Press.

PANTER–BRICK, C., LAYTON, R., and ROWLEY– CONWY, P. (eds.). 2001. Hunter-Gatherers: An Interdisciplinary Perspective. Cambridge: Cambridge University Press.

MACDONALD, K. C. 1997. Korounkorokalé Revisited: The Pays Mande and the West African Microlithic Technocomplex. African Archaeological Review 14: 161200.

POSNANSKY, M. 1984. Early Agricultural Societies in Ghana. (In), Clark, J. D., and Brandt, S. A. (eds.), From Hunters to Farmers: The Causes and Consequences of Food Production in Africa. Berkeley: University of California Press, pp. 147-151.

MACDONALD, K. C. 1998. Archaeology, Language and the Peopling of West Africa: A Consideration of the Evidence. (In), Blench, R. and Spriggs, M. (eds.) Archaeology and Language II. London: Routledge, pp. 33-66.

POSNANSKY, M. 2004. Processes of Change - A Longitudinal Ethno-archaeological Study of a Ghanaian village: Hani 1970–98. African Archaeological Review 21: 31-37.

MACDONALD, K. C. and ALLSWORTH-JONES, P. 1994. A Reconsideration of the West African Macrolithic Conundrum: New Factory Sites and an Associated Settlement in the Vallée du Serpent, Mali. African Archaeological Review 12: 73-104.

QUICKERT, N. A., GODFREY-SMITH, D. I., and CASEY, J. 2003. Optical and Thermoluminescence Dating of Middle Stone Age and Kintampo Bearing Sediments at Birimi, A Multi-component Archaeological Site in Ghana. Quaternary Science Reviews 22: 12911297.

MCBREARTY, S. and BROOKS, A. S. 2000. The Revolution that wasn’t: A New Interpretation of the Origin of Modern Human Behavior. Journal of Human Evolution 39: 453–563.

RAHTZ, P. A. and FLIGHT, C. 1974. A Quern Factory near Kintampo, Ghana. West African Journal of Archaeology 4: 1-31.

MALEY, J. 1991. The African Rain Forest Vegetation and Palaeoenvironments during the Late Quaternary. Climatic Change 19: 79–98.

RUSSELL, J., TALBOT, M.R., and HASKELL, B.J. 2003. Mid-Holocene Climate Change in Lake Bosumtwi, Ghana. Quaternary Research 60: 133-141.

MERCADER, J. 2003a. Introduction: Paleolithic Settlement of Rain Forests. (In), Mercader, J, (ed.), Under the Canopy. The Archaeology of Tropical Rainforests. New Brunswick: Rutgers University Press, pp. 1-31.

SALZMANN, U., and HOELZMANN, P., 2005. The Dahomey Gap: An Abrupt Climatically Induced Rain Forest Fragmentation in West Africa During the Late Holocene. The Holocene 15: 190–199.

MERCADER, J. 2003b. Foragers of the Congo: The Early Settlement of the Ituri Forest. (In), Mercader, J, (ed.), Under the Canopy. The Archaeology of Tropical Rainforests. New Brunswick: Rutgers University Press, pp. 93-116.

SHANAHAN, T. M., OVERPECK, J. T., WHEELER, C. W., BECK, J., JEFFREY, W., PIGATI, J. S., TALBOT, M. R., SCHOLZ, C. A., PECK, J., and KING, J. W. 2006. Paleoclimatic Variations in West Africa from a Record of Late Pleistocene and Holocene Lake Level Stands of Lake Bosumtwi, Ghana. Palaeogeography, Palaeoclimatology, Palaeoecology 242: 287-302

MERCADER, J. and BROOKS, A. S. 2001. Across Forests and Savannas: Later Stone Age Assemblages from Ituri and Semliki, Democratic Republic of Congo. Journal of Anthropological Research 57: 197-217.

SHAW, T. 1944. Report on Excavations Carried out in the Cave known as ‘Bosumpra’ at Abetifi, Kwahu, Gold Coast Colony. Proceedings of the Prehistoric Society (NS) 10: 1-68.

MERCADER, J. and MARTI, R. 2003. The Middle Stone Age Occupation of Atlantic Central Africa: New Evidence from Equatorial Guinea. (In), Mercader, J, (ed.), Under the Canopy. The Archaeology of Tropical Rainforests. New Brunswick: Rutgers University Press, pp. 64-92.

SHAW, T. 1978/79. Holocene Adaptations in West Africa: The Late Stone Age. Early Man News: Newsletter for Human Palecology 3/4: 51-82. 148

DEREK WATSON: THE LATE STONE AGE IN GHANA: THE RE-EXCAVATION OF BOSUMPRA CAVE IN CONTEXT SHAW, T. 1981. The Late Stone Age in West Africa and the Beginnings of Food Production. (In), Roubet, C., Hugot, H.-J., and Souville, G. (eds.), Prehistoire Africaine: Melanges Offerts au Doyen Lionel Balout. Paris: Editions ADPF, pp. 213-35.

WATSON, D. J. 2005a. Under the Rocks: Reconsidering the Origin of the Kintampo Tradition and the Development of Food Production in the Savannaforest/Forest of West Africa. Journal of African Archaeology 3: 3-55.

SHAW, T. 1985. The Prehistory of West Africa. (In), Ajayi, J. F. A., and Crowder, M. (eds.), History of West Africa, Volume 1. New York: Longman, pp. 48-86.

WATSON, D. J. 2005b. Straws within a Glass House: A Reply to Stahl. Journal of African Archaeology 3: 65-68. WATSON, D. J. In Preparation. Within the Savanna and Forests: The Late Stone Age Kintampo Tradition, Ghana.

SHAW, T. and DANIELS. S. G. H. 1984. Excavations at Iwo Eleru. West African Journal of Archaeology 14: 1269.

WELDEAB, S., SCHNEIDER, R.R., KOLLING, M., WEFER, G., 2005. Holocene African Droughts relate to Eastern Equatorial Atlantic Cooling. Geology 33: 981984.

SHINNIE, P. L. and KENSE, F. J. 1989. The Archaeology of Gonja, Ghana. Calgary: University of Calgary Press.

WHITE, F. 1983. The Vegetation of Africa. Paris: UNESCO.

SMITH, A. B. 1975. Radiocarbon Dates from Bosumpra Cave, Abetifi, Ghana. Proceedings of the Prehistoric Society 41: 179-182.

YORK, R. N. 1974. Excavations at Dutsen Kongba near Jos, Nigeria (Preliminary Notice). Nyame Akuma 4: 17209.

SOWUNMI, M. A. 2002. Environmental and Human Responses to Climatic Events in West and West Central Africa during the Late Holocene. (In), Hassan, F. A. (ed.), Droughts, food and culture: ecological change and food security in Africa’s later prehistory. New York: Kluwer/Plenum Publishers, pp. 95-104.

YORK, R. N. 1978. Excavations at Dutsen Kongba, Plateau State, Nigeria. West African Journal of Archaeology 8: 139-163.

STAHL, A. B. 1985. Reinvestigation of Kintampo 6 Rock Shelter, Ghana: Implications for the Nature of Culture Change. African Archaeological Review 3: 117150. STAHL, A. B. 1994. Innovation, Diffusion and Culture Contact: The Holocene Archaeology of Ghana. Journal of World Prehistory 8: 51-112. STAHL, A. B. 2005. Glass Houses under the Rocks: A Reply to Watson. Journal of African Archaeology 3: 5764. STARK, M. T. (ed.). 1998. The Archaeology of Social Boundaries. Washington (D.C.): Smithsonian Institution Press. SWARTZ, B. K. 1972. An Analysis and Evaluation of the Yapei Pebble Tool Industry, Ghana. The International Journal of African Historical Studies 5: 265-70. SWARTZ, B. K. 1974. A Stratified Succession of Stone Age Assemblages at Hohoe, Volta Region, Ghana. West African Journal of Archaeology 4: 57-81. TALBOT, M., LIVINGSTONE, D. A., PARKER, P. G., MALEY, J., MELACK, J. M., DELIBRIAS, G., and GULLIKSEN, S. 1984. Preliminary Results from Sediment Cores from Lake Bosumtwi, Ghana. Palaeoecology of Africa 16: 173–192. VAN NOTEN, F. 1977. Excavations at Matupi Cave. Antiquity 51: 35-40. 149

CAMBRIDGE MONOGRAPHS IN AFRICAN ARCHAEOLOGY No 1 No 2 No 3 No 4 No 5 No 6 No 7 No 8 No 9 No 10 No 11 No 12 No 13 No 14 No 15 No 16 No 17 No 18 No 19 No 20 No 21 No 22 No 23 No 24 No 25 No 26 No 27 No 28 No 29 No 30 No 31 No 32 No 33 No 34 No 35 No 36

BAR S75, 1980 The Niger Delta Aspects of its Prehistoric Economy and Culture by Nwanna Nzewunwa. ISBN 0 86054 083 9 BAR S89, 1980 Prehistoric Investigations in the Region of Jenne, Mali A Study in the Development of Urbanism in the Sahel by Susan Keech McIntosh and Roderick J. McIntosh ISBN 0 86054 103 7 BAR S97, 1981 Off-Site Archaeology and Human Adaptation in Eastern Africa An Analysis of Regional Artefact Density in the Amboseli, Southern Kenya by Robert Foley. ISBN 0 86054 114 2 BAR S114, 1981 Later Pleistocene Cultural Adaptations in Sudanese Nubia by Yousif Mukhtar el Amin. ISBN 0 86054 134 7 BAR S119, 1981 Settlement Patterns in the Iron Age of Zululand An Ecological Interpretation by Martin Hall. ISBN 0 86054 143 6 BAR S139, 1982 The Neolithic Period in the Sudan, c. 6000-2500 B.C. by Abbas S. Mohammed-Ali. ISBN 0 86054 170 3 BAR S195, 1984 History and Ethnoarchaeology in Eastern Nigeria A Study of Igbo-Igala relations with special reference to the Anambra Valley by Philip Adigwe Oguagha and Alex Ikechukwu Okpoko. ISBN 0 86054 249 1 BAR S197, 1984 Meroitic Settlement in the Central Sudan An Analysis of Sites in the Nile Valley and the Western Butana by Khidir Abdelkarim Ahmed. ISBN 0 86054 252 1 BAR S201, 1984 Economy and Technology in the Late Stone Age of Southern Natal by Charles Cable. ISBN 0 86054 258 0 BAR S207, 1984 Frontiers Southern African Archaeology Today edited by M. Hall, G. Avery, D.M. Avery, M.L. Wilson and A.J.B. Humphreys. ISBN 0 86054 268 8. £23.00. BAR S215, 1984 Archaeology and History in Southern Nigeria The ancient linear earthworks of Benin and Ishan by P.J. Darling. ISBN 0 86054 275 0 BAR S213, 1984 The Later Stone Age of Southernmost Africa by Janette Deacon. ISBN 0 86054 276 9 BAR S254, 1985 Fisher-Hunters and Neolithic Pastoralists in East Turkana, Kenya by John Webster Barthelme. ISBN 0 86054 325 0 BAR S285, 1986 The Archaeology of Central Darfur (Sudan) in the 1st Millennium A.D. by Ibrahim Musa Mohammed. ISBN 0 86054 367 6. BAR S293, 1986 Stable Carbon Isotopes and Prehistoric Diets in the South-Western Cape Province, South Africa by Judith Sealy. ISBN 0 86054 376 5. BAR S318, 1986 L'art rupestre préhistorique des massifs centraux sahariens by Alfred Muzzolini.. ISBN 0 86054 406 0 BAR S321, 1987 Spheriods and Battered Stones in the African Early and Middle Stone Age by Pamela R. Willoughby. ISBN 0 86054 410 9 BAR S338, 1987 The Royal Crowns of Kush A study in Middle Nile Valley regalia and iconography in the 1st millennia B.C. and A.D. by Lázló Török.. ISBN 0 86054 432 X BAR S339, 1987 The Later Stone Age of the Drakensberg Range and its Foothills by H. Opperman. ISBN 0 86054 437 0 BAR S350, 1987 Socio-Economic Differentiation in the Neolithic Sudan by Randi Haaland. ISBN 0 86054 453 2 BAR S351, 1987 Later Stone Age Settlement Patterns in the Sandveld of the South-Western Cape Province, South Africa by Anthony Manhire. ISBN 0 86054 454 0 BAR S365, 1987 L'art rupestre du Fezzan septentrional (Libye) Widyan Zreda et Tarut (Wadi esh-Shati) by JeanLoïc Le Quellec. ISBN 0 86054 473 7 BAR S368, 1987 Archaeology and Environment in the Libyan Sahara The excavations in the Tadrart Acacus, 19781983 edited by Barbara E. Barich. ISBN 0 86054 474 5 BAR S378, 1987 The Early Farmers of Transkei, Southern Africa Before A.D. 1870 by J.M. Feely. ISBN 0 86054 486 9 BAR S380, 1987 Later Stone Age Hunters and Gatherers of the Southern Transvaal Social and ecological interpretation by Lyn Wadley. ISBN 0 86054 492 3 BAR S405, 1988 Prehistoric Cultures and Environments in the Late Quaternary of Africa edited by John Bower and David Lubell. ISBN 0 86054 520 2 BAR S418, 1988 Zooarchaeology in the Middle Nile Valley A Study of four Neolithic Sites near Khartoum by Ali Tigani El Mahi. ISBN 0 86054 539 3 BAR S422, 1988 L'Ancienne Métallurgie du Fer à Madagascar by Chantal Radimilahy. ISBN 0 86054 544 X BAR S424, 1988 El Geili The History of a Middle Nile Environment, 7000 B.C.-A.D. 1500 edited by I. Caneva. ISBN 0 86054 548 2 BAR S445, 1988 The Ethnoarchaeology of the Zaghawa of Darfur (Sudan) Settlement and Transcience by Natalie Tobert. ISBN 0 86054 574 1 BAR S455, 1988 Shellfish in Prehistoric Diet Elands Bay, S.W. Cape Coast, South Africa by W.F. Buchanan. ISBN 0 86054 584 9 BAR S456, 1988 Houlouf I Archéologie des sociétés protohistoriques du Nord-Cameroun by Augustin Holl. ISBN 0 86054 586 5 BAR S469, 1989 The Predynastic Lithic Industries of Upper Egypt by Liane L. Holmes. ISBN 0 86054 601 2 (two volumes) BAR S521, 1989 Fishing Sites of North and East Africa in the Late Pleistocene and Holocene Environmental Change and Human Adaptation by Kathlyn Moore Stewart. ISBN 0 86054 662 4 BAR S523, 1989 Plant Domestication in the Middle Nile Basin An Archaeoethnobotanical Case Study by Anwar AbdelMagid. ISBN 0 86054 664 0 BAR S537, 1989 Archaeology and Settlement in Upper Nubia in the 1st Millennium A.D. by David N. Edwards. ISBN 0 86054 682 9

No 37 No 38 No 39 No 40 No 41 No 42 No 43 No 44 No 45 No 46 No 47 No 48 No 49 No 50 No 51 No 52 No 53 No 54 No 55 No 56 No 57 No 58 No 59 No 60 No 61 No 62 No 63 No 64 No 65 No 66 No 67 No 68 No 69 No 70 No 71 No 72 No 73

BAR S541, 1989 Prehistoric Settlement and Subsistence in the Kaduna Valley, Nigeria by Kolawole David Aiyedun and Thurstan Shaw. ISBN 0 86054 684 5 BAR S640, 1996 The Archaeology of the Meroitic State New perspectives on its social and political organisation by David N. Edwards. ISBN 0 86054 825 2 BAR S647, 1996 Islam, Archaeology and History Gao Region (Mali) ca. AD 900 - 1250 by Timothy Insoll. ISBN 0 86054 832 5 BAR S651, 1996 State Formation in Egypt: Chronology and society by Toby A.H. Wilkinson. ISBN 0 86054 838 4 BAR S680, 1997 Recherches archéologiques sur la capitale de l’empire de Ghana Etude d’un secteur d’habitat à Koumbi Saleh, Mauritanie. Campagnes II-III-IV-V (1975-1976)-(1980-1981) by S. Berthier. ISBN 0 86054 868 6 BAR S689, 1998 The Lower Palaeolithic of the Maghreb Excavations and analyses at Ain Hanech, Algeria by Mohamed Sahnouni. ISBN0 86954 875 9 BAR S715, 1998 The Waterberg Plateau in the Northern Province, Republic of South Africa, in the Later Stone Age by Maria M. Van der Ryst. ISBN 0 86054 893 7 BAR S734, 1998 Cultural Succession and Continuity in S.E. Nigeria Excavations in Afikpo by V. Emenike Chikwendu. ISBN 0 86054 921 6 BAR S763, 1999 The Emergence of Food Production in Ethiopia by Tertia Barnett. ISBN 0 86054 971 2 BAR S768, 1999 Sociétés préhistoriques et Mégalithes dans le Nord-Ouest de la République Centrafricaine by Étienne Zangato. ISBN 0 86054 980 1 BAR S775, 1999 Ethnohistoric Archaeology of the Mukogodo in North-Central Kenya Hunter-gatherer subsistence and the transition to pastoralism in secondary settings by Kennedy K. Mutundu. ISBN 0 86054 990 9 BAR S782, 1999 Échanges et contacts le long du Nil et de la Mer Rouge dans l'époque protohistorique (IIIe et IIe millénaires avant J.-C.) Une synthèse préliminaire by Andrea Manzo. ISBN 1 84171 002 4 BAR S838, 2000 Ethno-Archaeology in Jenné, Mali Craft and status among smiths, potters and masons by Adria LaViolette. ISBN 1 84171 043 1 BAR S860, 2000 Hunter-Gatherers and Farmers An enduring Frontier in the Caledon Valley, South Africa by Carolyn R. Thorp. ISBN 1 84171 061 X BAR S906, 2000 The Kintampo Complex The Late Holocene on the Gambaga Escarpment, Northern Ghana by Joanna Casey. ISBN 1 84171 202 7 BAR S964, 2000 The Middle and Later Stone Ages in the Mukogodo Hills of Central Kenya A Comparative Analysis of Lithic Artefacts from Shurmai (GnJm1) and Kakwa Lelash (GnJm2) Rockshelters by G-Young Gang. ISBN 1 84171 251 5 BAR S1006, 2001 Darfur (Sudan) In the Age of Stone Architecture c. 1000 - 1750 AD Problems in historical reconstruction by Andrew James McGregor. ISBN 1 84171 285 X BAR S1037, 2002 Holocene Foragers, Fishers and Herders of Western Kenya by Karega-Mũnene. ISBN 1 84171 1037 BAR S1090, 2002 Archaeology and History in Ìlàrè District (Central Yorubaland, Nigeria) 1200-1900 A.D. by Akinwumi O. Ogundiran. ISBN 1 84171 468 2 BAR S1133, 2003 Ethnoarchaeology in the Zinder Region, Republic of Niger: the site of Kufan Kanawa by Anne Haour. ISBN 1 84171 506 9 BAR S1187, 2003 Le Capsien typique et le Capsien supérieur Évolution ou contemporanéité. Les données technologiques by Noura Rahmani. ISBN 1 84171 553 0 BAR S1216, 2004 Fortifications et urbanisation en Afrique orientale by Stéphane Pradines. ISBN 1 84171 576 X BAR S1247, 2004 Archaeology and Geoarchaeology of the Mukogodo Hills and Ewaso Ng’iro Plains, Central Kenya by Frederic Pearl. ISBN 1 84171 607 3 BAR S1289, 2004 Islamic Archaeology in the Sudan by Intisar Soghayroun Elzein. ISBN 1 84171 639 1. BAR S1308, 2004 An Ethnoarchaeological Study of Iron-Smelting Practices among the Pangwa and Fipa in Tanzania by Randi Barndon. ISBN 1 84171 657 X. BAR S1398, 2005 Archaeology and History in North-Western Benin by Lucas Pieter Petit. ISBN 1 84171 837 8. BAR S1407, 2005 Traditions céramiques, Identités et Peuplement en Sénégambie Ethnographie comparée et essai de reconstitution historique by Moustapha Sall. ISBN 1 84171 850 5 BAR S1446, 2005 Changing Settlement Patterns in the Aksum-Yeha Region of Ethiopia: 700 BC – AD 850 by Joseph W. Michels. ISBN 1 84171 882 3. BAR S1454, 2006 Safeguarding Africa’s Archaeological Past Selected papers from a workshop held at the School of Oriental and African Studies, University of London, 2001 edited by Niall Finneran. ISBN 1841718920 BAR -S1537, 2006 Excavations at Kasteelberg, and the Origins of the Khoekhoen in the Western Cape, South Africa by Andrew B. Smith. ISBN 1 84171 969 2. BAR –S1549, 2006 Archéologie du Diamaré au Cameroun Septentrional Milieux et peuplements entre Mandara, Logone, Bénoué et Tchad pendant les deux derniers millénaires by Alain Marliac ISBN 1 84171 978 1. BAR –S1602, 2007 Chasse et élevage dans la Corne de l’Afrique entre le Néolithique et les temps historiques by Joséphine Lesur. ISBN 978 1 4073 0019 1. BAR –S1617, 2007 The Emergence of Social and Political Complexity in the Shashi-Limpopo Valley of Southern Africa, AD 900 to 1300 Ethnicity, class, and polity by John Anthony Calabrese ISBN 978 1 4073 0029 0. BAR –S1658, 2007 Archaeofaunal remains from the past 4000 years in Sahelian West Africa Domestic livestock, subsistence strategies and environmental changes by Veerle Linseele ISBN 978 1 4073 0094 8. BAR –S1667, 2007 Il Sahara centro-orientale Dalla Preistoria ai tempi dei nomadi Tubu / The Central-Oriental Sahara. From Prehistory to the times of the nomadic Tubus by Vanni Beltrami con le fotografie e i riassunti in inglese di Harry Proto / with English summaries and photographs by Harry Proto. ISBN 978 1 4073 0102 0. BAR –S1679, 2007 Memory and the Mountain: Environmental Relations of the Wachagga of Kilimanjaro and Implications for Landscape Archaeology by Timothy A. R. Clack. ISBN 978 1 4073 0117 4. BAR –S1736, 2008 Archaeological Investigations of Iron Age Sites in the Mema Region, Mali (West Africa) by Tereba Togola. ISBN 978 1 4073 0178 5.