Protecting architectural heritage in Djenné : a civil society point of view

The "Old Towns of Djenné" were listed as a UNESCO World Heritage site in 1988. However, in 2016, this urban landscape was included in the List of World Heritage in Danger. The paper aims to examine the neglect that led to this listing by identifying the actors involved in the protection of the town's unique architecture, describing their constraints and interventions, and demonstrating their persistent inability to define a common strategy. We show how most international and national interventions in the past two decades have ignored the restrictions and commitments imposed by the World Heritage status of the town. A striking feature is the lack of awareness and the progressive loss of knowledge, among decision-makers at local and national level, as well as among external aid funding managers, of the value of the cultural heritage that Djenné architecture represents; and among the professionals (Malian architects and local masons), of the technical know-how needed to maintain this architecture. The paper addresses the need for concerted efforts to educate and sensitize various actors in the conservation of Djenne’s architecture to enable them to contribute effectively to political and technical decisions regarding the protection of


Introduction
In 1988 UNESCO listed the old towns of Djenné -the present-day town and the archaeological sites (Djenné-Djeno, Hambarkétolo, Kaniana and Tonomba) -as a world heritage site (McIntosh 1995).In 1992 the site became protected under Malian law (République du Mali 1992), although the zone of protection was so poorly defined (Brunet-Jailly 2003) that the Hambarketolo site was only included in 2016 (République du Mali 2016).
A diverse range of inherited construction techniques and traditions characterize the architecture of Djenné (Maas, Mommersteeg 1992;Brunet-Jailly, Scherrer 2011).Since its inscription, the site has attracted sustained international interest and involvement.But this did not prevent the World Heritage Committee, in July 2016, from including the Old Towns of Djenné in the List of World Heritage in Danger (UNESCO 2016a).This paper 1 aims to address this question strictly from the perspective of the architectural rather than the archaeological heritage.It will: (i) examine the actors involved in the conservation of Djenné's heritage; (ii) look at the context in which these actors operate; (iii) review information on the degradation of this heritage over the past thirty years; (iv) identify the main shortcomings which have compromised conservation interventions; and (v) offer some 1 SciencesPo, Paris, Joseph.Brunet-Jailly@wanadoo.fr 2 Independent Researcher suggestions for how to create a more effective conservation strategy for the built heritage of Djenné.
This article builds upon more than twenty years of work by the non-governmental organization DJENNE PATRIMOINE, created in 1994 by a group of Djenne dignitaries and European funders.The main objective of DJENNE PATRIMOINE is to promote the participation of the local population in the protection and promotion of the tangible and intangible heritage of the town (De Jorio 2016, 95-115).Its membership consists of about twenty people in Djenné -dignitaries, farmers, craftsmen, cattle farmers, civil servants and young people -and some thirty members in Europe 2 .

Actors in the Maintenance and Conservation of the Djenné Heritage Site
Numerous actors, including DJENNE PATRIMOINE, play a role in the maintenance and conservation of Djenné's architecture.

Local actors
The annual re-plastering of the façades of the Djenné mosque mobilises the entire Djenne population (See Figure 1).Men participate as masons or unskilled workers, young girls fetch water for the work site, and women prepare the meal to be eaten at the end of the event.
Similarly, the maintenance of private residences relies on kinship groups, but as it is not a perceived religious obligation, it depends on the goodwill and the financial means of the house owners themselves.Their understanding of the advantages and constraints of the conservation effort is therefore crucial.
Figure 1: Once a year in Djenne, everybody participates in replastering the mosque (Photo: Joseph Brunet-Jailly) The masons of Djenné have a unique and irreplaceable role in the conservation of the architecture of Djenné: they are the repositories of heritage memory and the essential professional skills.The tradition by which each family had its designated mason, responsible for house construction and repairs, is gradually disappearing.It has become a monetised relationship, responding to the changing political economy and market forces.The bareyton, the masonry guild responsible for apprenticeships, represents the profession vis-à-vis the local authorities.Professional training (Marchand 2009) is organised through a form of apprenticeship to a master mason, and creates strong bonds between the master mason and his protégés, but also between young masons and their elders.This professional cohesion is also ensured through continual social interactions between its members, for instance in the distribution of work, or in assisting masons' families in ceremonies during marriages, births and deaths.However, with the growing involvement of government in commissioning building projects, and also since the short-lived creation, under the socialist regime of Modibo Keita (1960-68), of a competing cooperative which gave rise to longlasting internal divisions, and not least because of the loss of authority of its leaders, the barey-ton no longer plays the role it did in the past. 3  The city's public dignitaries also retain considerable powers.The traditional office of village chief still exists, and a council of prominent citizens from the city's districts supports the chief.These dignitaries receive information about all government proposals and projects, and their approval is requested for these interventions, in return for a wide variety of compensations and incentives.They may or may not have moral authority or influence over other villagers.Their roles are not challenged, except by those citizens who have acquired a standard of living or inherited positions that mark them out from the majority of the population.
Alongside this traditional authority structure, an elected local government council, which chooses one of its members as the mayor, manages the commune of Djenné.It is the preferred but not exclusive interlocutor with the national government.The introduction of an electoral process has reactivated old political conflicts and rivalries between leading Djenné families (DPI17 2004), as election is primarily perceived to be a means of acquiring social status and access to personal advantage.Local conflicts between incompatible interests have thus become much more clearly visible than any collective ambition.Finally, the commune is highly dependent on resources from the central government (DPI16 2004), and so it has very little autonomy in practice.

External actors
First in this category is the Mission Culturelle de Djenné, which is the local agency of the Ministry of Culture (République du Mali, 1993).Its mandate includes: i) to catalogue the fixed and movable cultural assets of the site; ii) to develop and implement a conservation, restoration and promotion plan; iii) to ensure the participation of community organisations and cultural associations in the management of the site; iv) to exploit and disseminate oral and written sources of local history.The second and third mandates are of particular interest here, and the most difficult to fulfil.
Another important external actor is the Préfet.As the local representative of central government, and in this guise privy to all sensitive government matters, the Préfet in fact has a controlling hand in devolved services -such as urban planning, allocation of building land, granting of building permits and in public finance matters.In practice, with decentralisation having been brought to a halt since 2002, supervision by service ministries of the central state has replaced the control of local government by elected officials (Coulibaly 2010).
Other government ministries have a significant role to play in the protection -or the degradation -of the architectural heritage of Djenné.There is no coordination between the Ministry of Culture and other Ministries (education, urban development, roads) to provide for the special case of Djenné (Figure 2).In addition, there is often undue interference from national ministries and their local offices.For example, the Mayor's office is responsible for approving building permits, and there is an official representative of the Ministry of Urban Development in Djenné.However, if landlords are refused permits at local level, they can easily apply to the regional urban planning office, or even solicit any government minister in Bamako to overturn, via a simple telephone call, a refusal by officials based in Djenné.
Malian architects are also to be considered as external actors.Most of them have been trained in foreign countries 4 and some do not attach value to the architectural heritage of Mali, or have no knowledge of mud architecture and building.Moreover, they are often distrustful both of local building materials and of the magical aura with which the masons of Djenné consciously imbue their profession (Marchand 2009).Significantly, in 2006, of the eight bids presented in response to the call for proposals to build a museum in the heart of Djenné old town, seven expressly favoured the option of building in concrete.

Foreign actors
The most important of these are the Government of the Netherlands, which has funded a large restoration programme of a hundred houses typifying the architecture of the city.The Aga Khan Trust for Culture carried out a major restoration of the Djenné mosque.In both cases, the projects were negotiated and agreed in Bamako, at ministerial or presidential level.Local actors were allowed no part in the definition or the focus of the project.Moreover, in both cases, the expertise was essentially international.
A number of different funding agencies have allocated part of their total aid budgets to projects dealing with urban infrastructure in Djenné: the European Union, the Islamic Development Bank, the World Bank and the Kingdom of Denmark.Other foreign actors have documented the architectural heritage of Djenné and the problems of its conservation including Maas and Mommersteg (1992), who continued the work of Prussin (1993), LaViolette (2000), Bedaux 5 and Ducoloner (2009).The Technological University of Delft, GTZ 6 , and later KfW 7 have attempted to address the urgent problems of public sanitation.Donations of equipment like wheelbarrows, shovels, rakes, gloves for the weekly cleaning of the central square next to the mosque after market day, have also been made about every three years by various funding agencies.
From the foregoing, it is clear that the protection of Djenne's heritage depends on a multitude of stakeholders.However, there is no shared vision of Djenne's conservation which is partly due to the transition from a traditional economy to a cash-based and aiddependent society.Indeed, the economic context of monetarisation and market forces has been demolishing the traditional structures of production, in particular those relating to housing and the urban environment (in this respect this recalls the situation in France during the 19th century (see Silberman, 2012).

Contextual Constraints
Local actors offer many different explanations when they are asked about the difficulties they face in their actions, or lack of thereof, on the architectural heritage.The most frequent answer is "we are poor", or "they are poor", but this argument requires further examination.At present, insecurity -a new situation, but one that is likely to persist -has become another obstacle.However, a deeper diagnosis leads to a different explanation, involving the fragmentation of an existing dense network of relations of social interdependence, and its gradual and ongoing replacement by commercial relationships.Marchand (2009) provides a fascinating description of the masons with whom he worked on site for many months.He paints some admirable portraits, which are deeply explored and charged with empathy.He has perhaps not highlighted enough the extreme vulnerability of the living conditions of these craftsmen, some of whom he views as real artists and even nicknames one of them Michelangelo (Marchand 2009: 167-188).What is striking today, is the extremely low standard of living of these well-recognised professionals (Smithsonian 2017).This degree of poverty is associated with a very low level of education.All of Djenne's boys go to Koranic school, learning religious texts by rote (Diakite 2009;Mommersteeg 2009).But most of the masons are considered illiterate and innumerate, as most of their professional training involves no written documents, texts, plans or estimates.As a result, only a handful of Djenné masons are capable of responding to advertised notices or invitations to tender 8 .Masons thus have no choice but to accept all employment opportunites irrespective of the remuneration offered.Similarly, the majority of house owners opt systematically for the least costly solution at a time of diminishing supplies (and rising prices) of materials such as rice husks and rônier wood.Both parties are unable to exploit or transmit precise written information.Construction work begins when a little money, not the full cost amount, is available; the decision is not based on a global estimate or budget of any kind.Moreover, it is extremely rare for a house owner to keep accounts of building costs.When, using direct site observations, DJENNE PATRIMOINE published a short document comparing the full costs of rendering a facade using the best traditional techniques with those of cladding it with fired bricks using cement plaster (Scherrer 2009, Scherrer 2015), no masons or house owners were able to follow the calculations.Professional knowledge itself is increasingly disappearing, for example knowledge of the correct proportions for mixing the various ingredients in mud (banco) plaster 9 .

Precarity
In these conditions of extreme poverty, notions of public interest or of collective heritage, and their associated values, have disappered in thought and behaviour as each individual struggles to get the best from the present without long-term plans or any collective concern.In reality, the population of Djenné is probably not as impoverished as it seems.About a sixth of Djenne's old houses now have fired brick facades, installed at fourteen times the cost of the best quality of traditional mud render.This difference in cost is in no way justified by the durability of this kind of cladding, and this is clear even to those who are innumerate.Nevertheless, the numbers of this type of facade continue to grow (Scherrer 2015), as a fashionable trend has prevailed over economic argument.Some experts stubbornly overlook the cost-effectiveness argument (Joy 2007:154;Joy 2012:126); but the fact that one has some money at one's disposal at a given time is not a justification for an expenditure which will waste a large part of it.We must conclude that a significant proportion of the population possesses the financial means to invest purely for ostentatious reasons (for another example of ostentatious expenditures, see DP12a 2002).In addition, some families have been able to build big concrete houses in the immediate vicinity of the old city, at a cost incomparably higher than that of a modern mud house.More generally, intervention by foreign actors who, in the eyes of the local population, command unlimited resources has created a universal feeling that it would be stupid not to seize these opportunities.If merely requesting, and complaining about one's vulnerable situation, is enough to loosen the foreign purse strings, why not take advantage of this?This is the context in which the opinions on the Netherlands project reported by MacFarquhar (2011), Joy (2012) or Rowlands (2007: 136-7) need to be understood.

Insecurity
There are today two quite different aspects to the insecurity of Djenne's architectural heritage.The first is rarely referred to, but is essential because it is independent of changing circumstances, is the fact that the majority of the houses belong to multiple hereditary owners.This results simultaneously in unwillingness to sell a family house, out of respect for those from whom it was inherited, and at the same time in inabiliy of the owners to undertake maintenance work because of the difficulty of reaching agreement to share the costs.
The second aspect is more obvious.Although Djenné was not occupied during the 2012 armed conflicts in northern Mali, the deterioration of the security situation has now spread towards the Mopti and Ségou Regions.Since 2010, all western embassies in Mali have severely restricted the movement of their citizens in the country.This negative travel advice (UNESCO 2016b) has killed off tourism in the area, greatly impoverishing the commune (tourism tax in Djenne previously represented 20 to 30% of its revenue). 10Moreover, tourism supported a small number of hotels and restaurants and provided a livelihood for people working to serve the seasonal tourist market.All this has now disappeared.

A society under stress
Djenné society maintains very close and asymmetrical relationships between its members.In the current difficult circumstances, diversity of ethnic origins reinforces the revival of claims to separate identities, and perpetuates a hierarchical and fragmented society.This diversity of status is reactivated in a range of private and public circumstances, giving the leading families or religious authorities the opportunity to assemble large numbers of dependants (Holder& Olivier, 2004, Olivier, 2004).Dense family ties, especially those created through marriage, are the subject of carefully maintained knowledge dating back over 150 years in this insular city.The traditional village chief and his council of dignitaries, as well as the religious authorities, all have their own clienteles.Everyone in Djenné is situated within this inextricable network of ties and obligations.
In the past these networks also had an economic meaning.The leading families maintained their power through their capacities to resolve the difficulties their dependants encountered (illnesses, poor harvests etc.).Similarly, pupils of Koranic schools provided free or low-cost labour for the building of houses in djenne ferey 11 .Today, individual interests and market forces definitely take precedence.External interventions, whether by the Government of Mali or by foreign partners, have all played a part in this change, through their requirements for contract tenders, or the individual recruitment of masons, etc. Traditional social structures are collapsing, and the economic and social balance they helped to achieve has been disturbed.
The complexity of local alliance and dependency networks explains the great difficulty that civil servants from other regions of Mali have in understanding who to consult to get advice, and who the real decision makers are.These officials are locally accepted to the extent that they provide those they administer with individual advantages; they are feared only so long as their hierarchical superiors do not overrule them when they impose sanctions; and they are immediately imitated if they prove to be corrupt.So it is very difficult for them to promote cooperative actions with the population of Djenné to protect the city's heritage.The traditional authorities, the chief and his council, and the imam, appear to have no vision of the future for Djenné, but simply seek to maintain their own power, limited as it now is by the official capacities of the commune and regional councils as well as of the national government.
Instability, very low levels of education, lack of access to information, and the rumours and jealousies that develop in a closed milieu, undermines ideas such as collective interest in things like heritage.Where everyone thinks solely of the personal gain to be derived from a situation, market pressure transforms mentalities.So it is not surprising that the heritage of Djenné has seriously deteriorated.

Conspicuous Degradation of the Heritage
The architecture of the old city of Djenné has suffered multiple forms of degradation (Maas 2014).Before the listing of Djenné as a World Heritage site, the only cement buildings were the post office (where the concrete structure was masked by a wall of mud brick) and the warehouse of the agricultural service.In the mid-1990s, the hospital was rebuilt in concrete, given its position on the edge of the old city.The most conspicuous violation was the 1988 construction of the school compound, large concrete buildings rendered with fired bricks, on the former site of the ancient mosque of Sekou Amadou, in the very centre of the historic city.
Figure 3: Private commercial building in concrete faced with fired bricks, and tin shacks on the street (Photo: Joseph Brunet-Jailly) The government continued in the same vein by rebuilding the market in concrete, also in the centre of the city ( 2004), allowing a bank (Banque Internationale pour le Mali) to build its local branch (Figure 3 above) in concrete faced with fired brick (2006), agreeing to the construction of the Djenné Museum, a huge concrete building with mud brick facades (2008), (Figure 4 below) and also erecting a meat market built in concrete also in the town centre in 2015 (see Sarin 2015).These decisions were made despite national commitments to the protection of the listed site.They reflect an addiction to concrete building on the part of public decision-makers and their foreign partners, and also the fact that investing in a World Heritage listed city like Djenné has become so rewarding for the donor countries and such an important source of funding for the recipient country that government officials and local dignitaries have become savvy players in the recipient chains of these funds.
Another attack on the architectural heritage is the fashion, which was begun in the early 1980s by American volunteers, 12 of covering banco facades with fired bricks.This completely alters the urban landscape, and ultimately causes structural damage to the houses. 13 However, the practice is rapidly spreading with no opposition from the regulatory authorities (Scherrer 2015).A further assault is represented by coarse pastiches of Djenné's architectural style, in which the balanced proportions of the old buildings are no longer respected, and the finishing details (parapet walls, pilasters, porches) are poorly executed.Examples are the Crafts Centre, build in 2014, "whose template and materials are not in keeping with the local architecture", or the meat market, a concrete building of "discordant architectural style" according to a recent UNESCO report (UNESCO 2016b).Foreign actors can thus be seen to have as little concern for protecting architectural heritage as the national decision makers who allow these initiatives to go ahead.DJENNE PATRIMOINE has defended the idea that the architecture of Djenné is a work of art (Maas, Mommersteeg 1992) which should be preserved and conserved as such.Leading citizens of Djenné -at least those who are able to have a care for heritage because they are in a position to see beyond daily concerns -support this point of view and lament the dilapidation of the town or the construction of administrative buildings in concrete in the town centre.They are aware that it is possible to modernise the town by providing electrical and potable water installations, TV antennae or modern bathrooms in a manner that perfectly respects the technical specifications of earthen architecture and preserves the urban landscape.Above all, it would appear impossible for anyone to claim attachment to the architectural heritage of Djenné while allowing the town to become a slum, with permanent tin shacks on the mosque square, corrugated iron doors, rusty metal shutters, and electric cables hanging haphazardly.
Local people's criticisms of the restrictions imposed by UNESCO (Rowlands 2007: 135-137) are mostly a means to express their lack of understanding of the restoration project's objectives, and their lack of trust in the officials who manage the restoration work.The common opinion is that these officials gain more than the house owners do from these external funds.The most criticised restriction was that of respect for the traditional internal floorplan of private houses.But in fact this condition was applied flexibly by the Netherlands programme, which in any case affected only one house in twenty.On all privately funded restoration or reconstruction sites there is no enforcement of this condition.But the fact that foreigners are funding the restoration of private assets creates a general miscomprehension.
Figure 4: Oversized museum building; its concrete structure is covered with mud brick façades (Photo: Joseph Brunet-Jailly) For its part, DJENNE PATRIMOINE argued for a restricted definition of the zone of identical conservation (DPI12c 2002), and also, when the issue of building a museum arose, for an international competition for modern earth architecture proposals which would match the quality of the ancient setting of the town.The Ministry of Culture of the time appeared to be in agreement but the donor agency (the European Union) was unwilling to adopt this procedure.World Heritage listing implies commitments by Governments to the World Heritage Convention, including for example a commitment not to increase housing density in the town. 14But these commitments have not been carried out.The question which arises is therefore: what has been missing in the numerous external interventions, to enable them to make a real contribution to protecting the architectural heritage of Djenné?

The Shortcomings of International Interventions
The architecture of Djenné owes its survival to the determination of a few actors who over the past decades have deployed convincing advocacy and considerable means to carry out urgently needed interventions on Djenne's architectural heritage.The rapid decay, between 1984 and 1995, of houses deemed emblematic of the secular architecture of Djenné was a cause of concern to all observers (Bedaux et al 2003).The unexpected collapse of one of the mosque towers during restoration work corroborated the urgency of an intervention to save this monument.Despite these efforts, the architecture of Djenne remains endangered, especially given the economic decline of the city and the breakdown of its social cohesion.
But what is needed for the sustainable survival of Djenne architecture is much more demanding than the restoration of a few buildings.The decisive factor has to be a behavioural change by local, national and international actors, leading to increased awareness and wide dissemination of knowledge and the promotion of traditional knowledge among all the stakeholders.This is the only basis for sustainable protection (Rowlands 2007).Have international interventions lent themselves to this effort?The answer has to be negative.Neither the Netherlands nor AKTC was willing to become involved in this fundamental problem.These foreign actors dedicated themselves to the goal of successful restoration or reconstruction of buildings, concentrating on "simply finding the easiest way through to get the job done" (Rowlands, Joy 2007), while leaving to national authorities the responsibility for managing all the relationships with the local population.As early as 1997, the Netherlands project had given rise to anxiety among the inhabitants and questions like "foreigners are coming here to restore our houses, so have they become the owners?" were asked (DPI4 1998;Joy 2012: 68-69).Information and education activities had been much too cursory.Indeed, donors seek to fund projects which will be favourably viewed in specialist circles outside the country, rather than devoting any time to discussing with local people, either before or during the implementation (as we have personally witnessed).They do not organise themselves so as to ensure that the necessary language skills are available, at the same time their national partners also make no effort to explain, but are there to carry out decisions taken in Bamako.This is the major failure of external interventions: they want to completely ignore the local social context, and this can have dramatic consequences.Thus, during the early 2000s government officials and politicians repeatedly asserted and wrote that the masons of Djenné are the builders of the architectural heritage and have maintained it with care for centuries.But on 20 September 2006, the day on which the AKTC's mosque restoration project was due to start, there were only foreigners (including masons from Burkina Faso) on the roof of the mosque.The violent riot which immediately ensued was proof of the deep mistrust felt by the population, and particularly by the young, 15 towards the municipal administration, the religious authorities, and the Mission Culturelle.These incidents resulted in an intervention by the army, one death and about forty arrests.In the preceding months, Malian civil servants had held innumerable meetings with Djenné's dignitaries (DPI21 2006), but the masons had not been consulted, still less associated with the project.The government had taken all the decisions on its own, as had been the case ten years before (DPI3 1997); and the foreign partners had been naive enough to believe that everything was clear to the local population and the masons.
A second failing of external interventions has to do with their management.Firstly, the masons employed on this project were paid at a higher rate than the norm for Djenné (Marchand 2003), with no consideration for the resulting unfavorable consequences (such as higher labor costs for the maintenance of private houses).Secondly, the management lacks transparency and creates mistrust towards those in charge of it.Although the Dutch project began in 1997, it was not until 2002 that each houseowner was informed of the amount to be made available through the project for his house, and this sum was handed over to the mason in the presence of the houseowner (DPI12b 2002).
In the case of the Djenné mosque restoration, management was the responsibility of the AKTC employees on the spot, without the local authorities or the masons having any knowledge of it.The still-living tradition of local management (by neighbourhood), based on trust (by leading citizens known for their honesty), which is reactivated each year for the replastering of the mosque, was not followed here (DPI11 2001).This disdain shown by external partners for local social structures goes along with a similar lack of consideration on the part of the Government of Mali: local authorities and community members are systematically left out of project design and management.Projects benefitting from external funding are thus in no way real cooperation projects.
A third problem has been to assume that national representatives speak for the local community, or that the governing authorities are capable of setting up committees and commissions to inform and consult the population.These suggestions are based on the illusion that there is local democracy.In fact the level of mistrust is such that officials are unable to speak for the community, and that the authorities prove unable to run such bodies.In 2003, the Dutch project admitted that it suffered from a lack of capacity on the part of its steering committee to listen to local communities (Sanogo, Sidibe 2003); more recently, in 2015, the mayor's office was in conflict with the current Djenné heritage management committee, which demanded the demolition of the meat market.Since 2013, local actors have been unable to cooperate and to revise the site management plan for the period 2014-2018.
Foreign and national actors speak of communities, using the politically correct language of the international aid milieu, but in practice communities are kept away from discussions and decisions.The consequences of this lack of consensus building are a failure to gain the support of the population, and an inability to apply sanctions.The first head of the Mission Culturelle set the tone as early as 1999: "there is the law, and then there is the application of the law […] We have never gone to see someone who is enlarging his house […] I have the right and the power as head of mission to […] stop the work.But I won't do it […] we have never been demanding with the house owners" (DPI8 2000).In the end, the question of heritage is left to external stakeholders: neither local elected officials nor the public feel responsible for it.
A fourth problem results from the competition between external actors, and the pressures they exert on national decisions, as can be seen in the example of public sanitation.In 2000-2002 the Technological University of Delft proposed a domestic waste water infiltration system, whose installation in 1200 houses the GTZ and KfW funded in 2003-2004.Concomitantly, the Environment Ministry commissioned a consultancy group to prepare a Strategic Sanitation Plan.This plan involved stone paving 14 km of streets as well as construction of embankments and outlets, and the drainage of waste water to 22 outfalls fitted with pre-filters and filters on the embankments, and from there to a treatment plant (DPI14 2003).The cost of this pharaonic project, 6.5 billion CFA (about $10 million), made it all the more desirable in the eyes of the government.Locally, it was argued that the existing domestic infiltration systems required regular maintenance, but at the same time, the huge public olfactory pollution caused by the new project's settling ponds, to be built in the immediate vicinity of the old city, was overlooked.In such a climate of aid-dependency and competition between donors, the main aim of the government is to acquire the maximum amount of financial assistance, not to promote an appropriate solution in the context, nor a form of endogenous development.This same mindset accounts for the fact that, within a few years, Djenné has acquired a Museum (funded by the European Union), a Heritage Centre or Maison du Patrimoine (funded by the French Government and the DJENNE PATRIMOINE and Acroterre associations) and a Crafts Centre (funded by the Danish Government), three establishments whose functions are very similar.
The record shows that the failure of the international programmes to protect the architectural heritage of Djenné can be explained by the obvious deficiencies of the interventions.Foreign actors have primarily promoted their expertise in delivering beautifully restored buildings, paying high prices for short-lived restorations of the material heritage, therefore distorting the market for locals; they have neglected the grounding of the conservation effort on a solid basis: targeted training for the local community, heritage professionals and decision-makers at all levels.

Long Term Interventions Aiming at Future Sustainability
The preceding analysis reveals that the core of restoration project activities should not be restricted to work on the buildings themselves, but should be aimed at creating the conditions for sustainable maintenance to take place.As the restoration of the past social structure is highly unlikely, efforts should be focused on the abilities and knowledge of the national and local actors: they must redefine their roles and responsibilities, instead of simply complying with foreign advice.This implies a massive informative and educative effort at national and regional level, where lack of awareness and knowledge -or comprehension-of the conservation task is evident; but also a strong focus on the capacity of local actors to grasp the objectives of a protection strategy, to subscribe fully to it, and to contribute actively to its implementation.
DJENNE PATRIMOINE has conducted many information and education activities.Those aimed at the population of Djenné at large have included conferences on the history and architecture of Djenné, exhibitions of old photographs of the town, and displays of photographs of earth architecture across the world.For the craftsmen there have been initiatives to showcase their expertise (Brunet-Jailly 1997;Brunet-Jailly 2009), events to open up new markets (exhibition/sales, assistance with supply of raw materials, publications of photographs of their work on the web).For young people there have been school visits to the exhibitions of historic photographs, classroom activities etc.Not to mention specific activities for women (particularly potters, a typically female profession) and tourist guides 16 .
In terms of architectural protection, the primary need is to counteract the addiction of Malian decision makers 17 at national and regional level to cement construction by enabling them to become familiar with and to appreciate the Malian architectural heritage, and the qualities of earth-and mud-based materials.Malian architects, a large majority trained in Europe, have a decisive role to play by rediscovering their architectural heritage and applying their technical knowledge to the conservation of this heritage.
At the local level, an essential role must be given to a new organisation of the masons: they cannot survive without creating small enterprises combining technical and managerial skills, in order to respond to the new demands of the market, and to engage in technical and aesthetic innovation in earthen architecture as they have done in the past (Marchand 2010).This is why we now introduce two practical suggestions for structured training of different stakeholders.We will also propose a third approach, which has yet to be tested.

Functional literacy and supplementary professional training for masons
The first intervention focused on improving the masons' capacity to participate in discussions in more formal, official settings.At present they only can take part in very simple discussions, conducted preferably in Sonrai or Bambara, because it is difficult for them to speak French and their capacity to read and to calculate is very limited.Putting the masons back at the centre of the process of conservation and maintenance of Djenné's architectural heritage depends on their acquiring comprehension, communication and information skills.This intervention was tested as part of the construction of the Heritage Centre.It aimed at developing reading, writing and calculation abilities, first in Bambara (the language of daily exchange) and then in French (the official language) for all masons, with the support of the local office of the Ministry of Education 18 .A group of masons were able to attend three hours of instruction five days a week for five months, each year from 2008 to 2014, after completing their jobs on-site.
Although acquiring reading and writing skills is an extremely difficult and daunting task for adults, the masons of Djenné showed an extraordinary willingness to learn.In six years, these functional literacy courses enabled about thirty of the two hundred or so masons in Djenne to master writing, reading and calculation in Bambara and to begin to become familiar with French (DPI28 2010, DPI30b 2011).This approach will need to be carried on and replicated in order to equip more masons with the skills needed to participate in local decision-making and to complete and submit tender documents.Complementary professional training has so far been limited to a comparative evaluation of the recipes used in preparing the plaster for rendering walls.Other subjects could include installing electrical supply and bathrooms in earthern construction houses.

Training of other professionals in the construction sector and heritage conservation
The second intervention was a series of AKTC initiated training workshops at the Centre d'Architecture de Terre (Centre for Earthen Architecture) in Mopti, for architects and construction professionals and Malian officials responsible for built heritage conservation.The first of these workshops brought together government staff, masons and architects, for two weeks in 2013.The objectives of the workshop were fairly general in nature: to bring together actors in architectural heritage and create a platform for exchange of technical knowledge.In practice, large scale models were constructed, precise information on traditional plaster mixes were collected, and information was disseminated on mud architecture and building across the world.
In February 2015, a first training session for architects specialising in earthen architecture and in conservation of architectural heritage was held at the Centre.Participants included 6 architects (among them a representative of the architects association of Mali) as well as 6 final year architecture students from a technical school in Bamako.The one-week training course included theory sessions, visits to emblematic buildings, case studies and practical sessions on the topics of heritage, sustainable architecture, local materials and traditional and improved mud building techniques.
At the technician level, a total of 25 persons were also trained in November 2014 and January 2015: a group of 12 young building technicians was trained in various traditional and improved construction techniques, and a second group of 13 teachers and trainers from the building trades schools of Mopti, Sevare, Bandiagara and Koulikoro.This coordinated and progressively structured initiative 19 (though limited) would have to be extended to meet the aim of supplying the need for sufficient numbers of appropriately qualified specialists.

Raising the awareness of house owners in each neighbourhood.
During the first phase of the Dutch project, M. Sebastien Diallo, a Malian architect working for that project in Djenné, developed, as a personal initiative, a broadcast programme on local radio on the subject of heritage, which proved to be the most effective tool in educating citizens on the topic.Alongside the construction of the Heritage Centre in Djenné a detailed proposal was drawn up by DJENNE PATRIMOINE for a series of meetings for house-owners, to be held in each neighbourhood, to hear and discuss first-hand accounts from those who have had negative experiences of fired brick cladding and from masons who are convinced of the damaging nature of this practice.Potential interlocutors were identified and had agreed to take part.It only remained to find the small amount of funding needed; this was requested from UNESCO, but no response has been received.

Conclusion
The failure of international interventions to protect the architecture of Djenné has resulted from the fact that foreign actors have been mostly concerned with producing impressive restoration works, while completely neglecting the social context in which they have been operating.At the same time, local and national decision makers have been incapable of securing the interest and commitment of the population.
The foremost concern of external actors in Djenné has been to achieve an aesthetic appreciation of the results of their projects, and thereby international recognition for their own skills.Around these objectives they have developed a discourse claiming that local skills have been decisively important and are able to ensure the upkeep of the restored heritage "for future generations" (Schmidt 2007 ;Schmidt, Fane 2013).There is no doubt that the restoration projects have contributed to maintaining and developing local skills, but the claims made for their capacities are clearly over-optimistic, as evidenced by the delays and faulty work experienced on the new museum site (DPI30a 2011).Moreover, upkeep of the buildings has been left to the goodwill of the home-owners -with predictable results -as the Netherlands proposal for a foundation to ensure adequate funding for this ongoing work never developed (DPI2 1997).Some authors have indulged in demagogic discourse, claiming for example that "in some ways Djenne's World Heritage status serves to under-develop the town and stands in the way of people's livelihood strategies" (Rowlands, Joy 2007: 5-6).No argument at all supports the accusation of under-development: a description of the poverty of Djenné's inhabitants (Joy 2007: 149-151) does not suffice to condemn the Dutch restoration project.As for family strategies to improve their living conditions, the claim that these are blocked relies on a sensitive point, the prohibition on modifying the internal space of houses during the restoration.But everyone in Djenné knows that this is not what actually happened, starting from the first house to be restored, that of the brother of the village chief (Rowlands 2007: 135-137), and then for other houses.Another criticism is that the income from tourism does not benefit all the inhabitants, but primarily those in power.But even supposing that this were true -indeed the amounts and the distribution of tourism income have never been accurately mesasured -is it the role of a heritage protection project to correct income distribution?A pertinent critique of development aid is to argue that some of its interventions have had dramatic consequences, in particular in Mali, through undermining social structures (Bergamaschi 2014), but this was not the case in Djenne.
More generally, it is argued that raising the living standards of the local population will enable them to maintain the built heritage (Joy, 2012: 124-125).This is a plausible hypothesis, but it has not been shown to be evidence-based; at best it may be a necessary but not sufficient condition.Improvement in living standards offers no guarantee of a heightened level of cultural attachment to goods inherited from preceding generations and to the social values they represent.One may prefer to buy a car instead of replastering one's house, especially if foreigners are ready to fund this work in one's place.And finally, heritage protection funds are tiny when compared to the amount of aid required for inducing an economic development.
Indeed, a naive belief in the capacity of the population to find the best solutions for conservation of living heritage appears to overlook the fact that some popular demands are disastrous for heritage, as the example of the French Revolution reminds us (Silberman 2012).The idea that everyone should get some economic benefit from cultural heritage is a dramatic narrowing of the ambition for popular participation in heritage protection.International organisations are thus placing risky constraints on local strategies when they suggest -indeed impose -that heritage should depend for its preservation on its market value to local people: they introduce the characteristic rationale of capitalism into a social domain which has largely avoided this up to now.In other words, insisting on the fate of the poor is used here to justify the introduction of individual personal interest (locally) and of the demand for cultural artefacts (on international markets) into cultural heritage management.
The fact that the United Nations, in a resolution passed in December 2010 on "culture and development", adopted this strategy should come as no surprise in this era of unfettered neo-liberalism.Its recommendation to States is "to actively support the emergence of local markets for cultural goods and services and to facilitate the effective and licit access of such goods and services to international markets" (United Nations, 2011: 2).However, the UN are ten years behind the thinking of the best informed economists, who have been rediscovering the role of the state and of public goods since the beginning of the 21st century, after many decades of triumphant neo-liberalism (Lin, 2012).They also forget, when they claim that heritage can be a driver of development through tourism, that the conditions for this to come about have been clearly identified and are very difficult to fulfil (UNCTAD 2014).
Moreover, the proposal to use the market value of heritage as a criterion in heritage management is justified by a strange anthropological argument: that the notion of heritage would not have the same links with identity in Djenné as it does elsewhere in the world.
"It seems that the link between identity and cultural heritage that is so implicit in much Western discourse on cultural heritage -the discourse of cultural rights, dignity and heritage value -is not so implicit to people in Djenne and not particularly well understood in a local setting.In fact it seems as though sometimes assumptions have been made that these universal heritage values will have local meanings or that sensitisation programmes would be enough to instil them."(Rowlands 2007: 2) No comparative research supports the argument for an absence, in Djenne, of an implicit link between identity and cultural heritage: not only has this author not carried out a specific anthropological investigation of this question in Djenné but his assertions  in De Jong, Rowlands 2007) overlook, for example, the parallel which might be drawn between the interest in heritage that appeared in France following the industrial revolution (Silverman 2012) and that which emerged when Mali was faced with globalisation during the 1990s (Arnoldi 2003, Traore 1997, Traore 2012).Furthermore, as evidenced by the European experience, this linkage develops very slowly and progressively (Choay 1999).
An opposing argument has to be considered: that the protection of architectural heritage is to be pursued from the perspective of values other than those of individual interest and the market.In practice, all nations protect their heritage for reasons other than tourism and market value or capitalist free trade: historians observe that "heritage is the natural and historical birthplace and home of individual and collective identities" (Le Goff 1998: 12).And if anthropologists are right to claim that "heritage is produced in a context of discourse on roots, ownership, nationalism and global politics of recognition" (de Jong, Rowlands: 25), it is hard to see what role the market might play in this discourse, or that it might have a favourable influence.
From this point of view, the training of professionals and the education of the population and of decision makers are of more crucial importance in protecting heritage than the successful outcomes of building restoration projects.In Djenne, the divergent international interventions have ended in failure because they did not address the root of the problem, which is that the very idea of heritage protection has to be appropriated and implemented by local people.A large part of the failure of international interventions lies in the lack of awareness, knowledge and sensitivity of many of the actors involved -foreign as well as national experts included -and in the weakness of governmental authorities.

Acknowlegments
This article could not have been written without the contributions made by our friends in Djenné over the past 20 years, in particular the late Papa Moussa Cissé, the first President, and Amadou Tahirou Bah, the current President of DJENNE PATRIMOINE, and by other leading members of this association.We are also indebted to the Djenné masons with whom we have worked.Our thanks are due to Pierre Maas for his remarkable work on the architecture of Djenné, and to Duncan Fulton who worked with patience and commitment to translate successive drafts of this article.Notes 1 A first version of this paper was presented as a panel contribution to a session on "Islamic Architecture and (Contested) Cultural Heritage Management in West Africa" at the 59 th Annual Meeting of the African Studies Association in Washington DC, 1-3 December 2016 .Thanks to Barbara G. Hofmann for her invitation. 2The leading author of this paper (JBJ) was a founding member of this association and has been active as one of its facilitators alongside its presidents, Papa Cissé and later Amadou Tahirou Bah; an economist by profession, he lived in Mali from 1986 to 1995 and from 2000 to 2004, and has made frequent and repeated visits to Djenné, in particular for the writing, with leading personalities and craftsmen in Djenné, of a book describing the various dimensions of the town's heritage (Brunet-Jailly 1997, Brunet-Jailly 2007), and also later to supervise the building of a private residence designed to respect tradition in terms of plan and decoration while incorporating modern levels of convenience.This house project has been jointly carried out by JBJ, Benedicte Wåhlin, architect (Wåhlin Arkitekter AB, Stockholm), Papa Cisse and Amadou Tahirou Bah, leading citizens of Djenné, and by Boubacar Kouroumanse, known as Bayere, and Ladji Kouroumanse, two renowned master masons (Marchand 2009: 149-150, 167-188).The second author (OS), heads a cooperative of master masons using traditional building materials in the South of France, and was centrally involved in the projectinitiated by the same association -to build the Maison du Patrimoine de Djenné; he lived for one month each year in Djenné between 2007 and 2014, working daily with local masons, as well as participating in training workshops for masons, building technicians and architects organised by the Aga Khan Trust for Culture (AKTC) in Mopti.The two authors have maintained close and continuous relations over many years with local professional associations (of livestock herders, masons, embroiderers, shoemakers, potters, weavers…), and with the Mission culturelle de Djenné, the commune council, teachers, health service workers, house owners, Ministry of Culture, religious leaders etc. 3 The history and anthropology of the barey ton is still to be undertaken 4 A non-governmental School of Engineering, Architecture and Urbanism (ESIAU) opened in Bamako only in 2006, http://esiau-mali.com/ 5 During the past decades, R M A Bedaux was the curator of the Africa section of the Leyden Ethnological Museum and organiser of archaelogical researches by the Dutch teams in Djenné, the Central Niger Delta and the Dogon country.Deutsche Gesellschaft für Technische Zusammenarbeit (GTZ) is the German agency for international cooperation; from 2011 on, its new name is Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ) 7 The Kreditanstalt für Wiederaufbau (KfW) is a German public bank now devoted to the funding of German exports and international cooperation for development. 8Projects carried out by the government are well publicised and advertised in a form which the Dogon entrepreneurs, who are better educated, know how to respond to, whereas the majority of the Djenné masons do not.For major projects which are foreign funded, external agencies fulfil the role of entrepreneur and only employ Djenné masons as labourers. 9Banco : mud construction material (a mixture of clay, sand, various vegetable materials, with rice husks, shea butter, powder from baobab and nere fruits) 10 Except when the mayor put his brother in charge of the collection and claimed that the inflow of tourists was suddenly in decline (2004-2009)  11 Traditional handmade cylindrical mud brick, characteristic of the inherited construction technique in Djenné 12 According to Papa Cissé, interview 28 December 1988 13 Water can penetrate behind the cladding of fired bricks and get into the mass of the wall, causing damage which becomes apparent only much later, when the covering falls from the wall.14 Indeed, due to the economic context, the population growth is very limited. 15Perhaps 200, certainly not several thousand (Rowlands 2007 p. 142) 16 The reader can refer to the website of DJENNE PATRIMOINE, http://www.djenne-patrimoine.org/ 17The example of Morocco shows that producers and importers of cement are the most determined lobby against earth architecture (Naji 2012(Naji , 2017)). 18Centre d'animation pédagogique, in charge of education activities outside the regular schools. 18Olivier Scherrer and co-authors wrote a report on each training session ; these reports are considered confidential by AKTC.

Figure 2 :
Figure 2: House in Djenne, with façade in the style known as Moroccan ; note the electrical and phone cables installed haphazardly in the streets (Photo: Joseph Brunet-Jailly)

Figure 5 :
Figure 5: Djenne dignitaries gathering for a meeting with foreign partners involved in tourism development (Photo: Joseph Brunet-Jailly)