mcintosh by lanyuehua



                                   Roderick J. McIntosh
                                     Yale University


and Locales] Frankly, we prehistorians are not terribly good at exorcising our

(intellectual) ghosts. By intellectual ghosts, I do not mean the quasi-looters, those

Schliemanns or those Austin Henry Layards, or even the communal-fantasies, those

Indiana Joneses or Laura Crofts (who make the public swoon and give archaeologist the

willies). Rather, we seem unable to put to rest forever the excesses of our intellectual

forbears, the Grafton Elliot Smiths / Lord Raglans and the Ellsworth Huntingtons.

Earlier, more extreme versions of migration madness (of the former two) and of the

latter’s environmental determinism so haunt our field that any explanations of culture

change that even mention these bugbears can risk being dismissed without fair hearing.

Will we ever be able to leap off the vast pendulum of our discipline's intellectual history

that demands an emotional counter-reaction to anything even hinting, however kindly and

gently, of the themes so abused by past pseudo-scientific explanations, namely hyper-

diffusionism linked to migrating "superior races" in the former case and the exogenic

agency (and material primacy) for cultural change of the latter? The reactions are

exaggerated, if that is even possible, in the case of Africa because of the reaction against

a Colonial-period tradition of explanation that stripped Africans of their own active

Yet surely there are some cases of prehistoric change and process that have to take into

account demonstrable climate change and documented large-scale movements of peoples.

This is why I found it so disturbing, and discouraging, recently to see The Way the Wind

Blows (eds., R.McIntosh, J.Tainter and S.K.McIntosh) described as “neo-environmental

determinism” in a recent chapter by William Balée and Clark Erickson.

Yikes! How could the editor (Balée) of the series in which Wind appeared (who had

heaped such praise on the book at its coming-out) have so misinterpreted the book’s

thrust? This was a book that was intended to unite a bakers-dozen of the archaeologists

around the world who were, at a time of increasingly better resolved palaeoclimate

knowledge of their areas, experimenting with an explicit rejection of an old-time

materialist determinism, while at the same time assessing the role of climate change in

the creation of ancient peoples’ perception of their environment? I remain doubly

confused because they write:

       Historical ecology is an interdisciplinary approach. It focuses on the
       historical landscape, a multi-dimensional physical entity that has both
       spatial and temporal characteristics and has been modified by human
       activity such that human intentions and actions can be inferred, if not
       read as material culture, from it. The landscape is like a text, but not one
       that is readily accessible to historians’ and epigraphers’ methods because
       it is not written in a decipherable script, but rather is inscribed in a subtle,
       physical sense by learned, patterned behavior and action – what
       anthropologists traditionally refer to a culture. Culture is physically
       embedded and inscribed in the landscape as nonrandom patterning, often
       a palimpsest of continuous and discontinuous inhabitation by past and
       present peoples. …Our historical ecology also stands in sharp contrast to
       the neo-environmental determinism popular in archaeology today.

In Wind we were attempting to provide examples from global prehistory of how to

reconstruct “social memory”, a concept meant to describe the ways by which

communities understand, curate, and transmit knowledge of past environmental states

(that is, the culturally filtered and constructed knowledge of climate change and attendant

states of the landscape) and possible responses to them. If we accept that all

archaeologists (and indeed many historians) have ever done is to interpret the traces of

past peoples’ actions on the world as those peoples perceived the world, then the concept

of social memory seemed to us a powerful – and explicitly non-environmentally

deterministic – tool to pries away the veils of time and forgetfulness separating the

prehistorian from the culturally-mediated understandings and motivations of past peoples

as they acted upon their perception of climate change and its causations. This is a

concept, at the very core or so I will argue here, of my attempts to understand the

transformation from the Late Stone Age (LSA) landscape of hunters/gatherers/

herders/fisherfolk (where a very nascent, of any, community concentration of exploitation

was the rule) to the emergent specialized and lightly clustered communities at the

transition of urbanism, and finally to the massive, intensively-specialized and ferociously

heterarchical Clustered Cities of the late first millennium BC and first millennium and a

half of the Common Era.

Allow me to summarize argument for a deep-time Mande social memory that I believe

will one day be fundamental to our reconstruction of the trajectory of emergent Middle
Niger urbanism. Fundamental to this argument is the apparently deep historical link to an

ethos of resistance to accumulation of power. These cities were lodged deep in the heart

of the northern precinct now and traditionally occupied by the language and cultural unity

called Mande [FIGURE 2: Saharan and Sahelian Regions of the Ancestral Mande].

Mande scholars and the Mande themselves are unusually united in their conception of the

peoples’ relationship with their history. And this theory of their relationship to their

history is key to thinking about the curation and transmission of packages of ideas about

the past (changes to the bio-physical sphere, past actions and events, past relations among

peoples). In the case of climate change, social memory provides the link between a

peoples’ perception of climate change and attendant changes to the bio-physical

landscape, the “real world”. People act not upon some abstract “real world”, but upon

their perceptions of same, as filtered and mediated by social memory. Mande social

memory has the extra feature of being not just the link, in complex but decipherable

ways, between actions of people on the world and that real, objective climate change, but

it also provides the link to a theory, better, perhaps a calculus, of the authority of those

sanctioned to deal with and make decisions about environmental stress and surprises.

The trick for the student of Mande history and prehistory is to use the concept of fluid

transformations with continuity explicit in the Mande view of their own past in order to

develop hypotheses about linked alterations to the nature in the way peoples used their

environment and developed institutions of authority for dealing with the highly

unpredictable climate modes of their Sahelian world (again, as modulated by social

memory). This is what, in Wind, I called the Mande weather machine. I suggest that

there is a traceable trajectory, granted with multiple shifts, at least back into the first
millennia BC, that is, back into the long-pre-urban LSA – and possibly even earlier.
Before that is it simply foolish to talk about trying to demonstrate strict continuity to a

cultural construction of causality in the world – but here for the first time I adduce some

field data that one day may tempt the less- squeamish! It is one thing to recognize the

enduring-while-changing history of certain elements in Mande settlement pattern and

modes of authority through time. It is quite another to understand what lies behind the

particulars of symbols, legends, and landscape markers that are the stuff of appropriation,

manipulation in legitimation sequences, and negotiations in the dynamic relations

between various interest groups in society.

These are the dynamics behind changes to the Mande weather machine, dynamics that are

never purely random in their evolution through time. In other words, certain symbols or

perceptions of the physical or social environment can effectively crystallize social action

only if there is a deeper organizing principle that gives a historical trajectory to the

society's social construction of reality. This must never be allowed to devolve into an

historicist exercise of looking for the unchanging essences of essentially static peoples.

The historicist project has always been to minimize the influence of change and to ignore

processes of social history, including the production of cultural perceptions of landscape

or of climate change. Rather, my intention here is to explore what gives trajectory, what

fundamentally structures the deep-time dynamism of environmental social memory.

Social memory here is used to mark the arena in which metaphors, symbols, legends and

"attitudes" crystallize social action.

Mande social memory is highly complex, but here let us boil it down to three meta-

themes most pertinent to how I have come to think about critical junctures in the

emergence of Middle Niger clustered cities:
    1) The Mande see their history (and present) as a calculus of interlocked core values

        and concepts (the DNA of historical trajectory, if you will) that are continuously
   mined, re-invented, and reinterpreted to adjust to present-day realities. [the core

   value critical to social memory as applied here is kuma koro (literally, “ancient

   speech”) or the idea that a line of ancestors continue to influence the world by

   acting through the behavior of descendants in the here-and-now]. This is, by the

   way, entirely consistent with the modern definition of landscape that allows the

   present physical construct (of geomorphology, environment and settlement traces)

   to be seen as a palimpsest of multiple, overlapping, cumulatively-modified former

   spatial arenas for human interaction with others and with the many gyres of bio-
   physical imputs.

2) The cause of climatic (environmental) surprise and stress is (not discoverable in

   physical laws of Nature, but rather…) to be found in disruptions/perturbations in

   the flow and distribution of a vitalizing force of the world, nyama. The

   disturbances in nyama can be understood by, predicted, and manipulated (for

   social good) by “nyamakala” (“handles of nyama”) – individuals or collectives

   (esp. secret societies) and sometimes by occupation corporations (most famously

   the mande blacksmiths and hunters) trained through a lifetime apprenticeship in

   the authority to deal with the most dangerous occult forces of the world. [My

   long-time collaborator, Téréba Togola, was named “téréba”, i.e., “(shocking

   environmental) surprise” after a grandfather born in the midst of the 1914-1918

   drought, conditions repeated meteorologically in the 1968-(1985) Sahel Drought,

   for which Dr. Togola was a living archive in his village. He was thus the

   walking-talking curation for three generations of recondite knowledge pertaining

   to what material responses worked and did not work – and curation of the store of

   occult events thought to be at the root cause of the devastation].

3) Major sources of perturbation to the flow of nyama are the uncontrollably violent
   members of society (e.g., the soul-sucking Xon-faaf of the Serer, who leave a

   string of damaging storms in their wake) and, esp. those who try to accumulate
        the authority that is more properly dispersed amongst the occupation corporations,

        the secret societies, and individual nyamakala. [A persistent political narrative of

        later, “imperial” Mande history is of the nyamakala gone bad – e.g., Sumaworo,

        the twisted blacksmith “king” of the Sunjata chronicles at the rise of the Mali


Shift now to the archaeological dilemma of the emergence of a distinctive form of

urbanism in the dune breaching palaeo-channels of the (present-day) Sahara and the
Niger River’s floodplain of Mali’s Sahel. We struggle with the prehistorian’s (and oral

historian’s) curse: that we can never know the full intentions and motivations, the full

play of the social memory of past peoples. Yet, to deny that archaeologists can decipher

some clues, perhaps from seemingly anomalous actions upon the world, to past

perceptions is to say that we are forever cut off from knowing at least the bare bones of

how culturally powerful mechanisms like social memory may have worked in the past.

The more optimistic among us think that all is not lost and that we have some help in

models of self-organizing landscapes.


several distinctive signatures of Middle Niger cities, their clustering (multiple specialists’

mounds in orbit about a “focus” tell ), the resolutely non-agglomerated personality of the

landscape, their lack of evidence for political hierarchies (or of elites, for that matter) –

these are classic “cities without citadels” -- and especially their stability and

sustainability for many hundreds of year (without demonstrable evidence of warfare) are

superb expressions of resilience. Resilience (defined as the ability of a system to maintain

its structure in the face of disturbance and to absorb and, indeed, to utilize change) is key
to my longstanding argument of Middle Niger sustainability. Here sustainability is

defined as these communities’ capacity to continue a desired condition of process, social
or ecological) in the face of the reality of life in what several palaeoclimatologist have

called the planet’s most unpredictable, most stressful climatic regime in terms of

frequency of anomalous events, of high inter-annual to inter-century variability – not to

mention (today) of marginal pastoralism and agriculture at the fringe of the world’s

largest desert.

A word is in order about three of the shorter-scale climate complications that are the

realities of the Mande world. Let us begin with the shortest scales, the inter-annual and
inter-decadal. Jean Koechlin has called the arid West African climate among the world’s

most variable and unpredictable. Shukla provides the classic description of the

instrumental record: “In the recorded meteorological data for the past 100 years, there is

no other region of the globe of this size for which spatial and seasonal averaged climatic

anomalies have shown such persistence”. Taking a longer chronological and a

continental-scale perspective, Sharon Nicholson despairs: “The largely semi-arid African

continent has undergone extreme climatic changes which are probably unmatched in their

magnitude and spatial extent”. Given this relentless saga of unpredictable weather and

climate, the Mande peoples have truly traced a history of progressive strategies of

flexibility “at the edge of chaos”.

More difficult to resolve to any detail is climate variability at the century/millennial scale.

We lack the lake cores, the speleothems, and the long history of stratified archaeological

“ecofactual” sequences that bless other regions and make tenuous our reconstructions

from lakes at great remove (L. Bosumtwi and L. Chad, in this case), from deep-sea cores

(ODP program) off the Mauritanian coast, and from spotty anecdotes and oral traditions.

FIGURE THREE (Sahelian Palaeoclimate) presents a crude per iodization of several
thousand years of stable and unstable modes, each stage lasting over five hundred years.

This figure shows the enormity of the southern Saharan/Sahelian shifts in climate, as
illustrated by precipitation shifts over 4,000 years of from +200% of present rainfall to

more recent persistent episodes of up to -25% less, for the broad Sahelian belt. (Because

West African rainfall is monsoonal, with little altitudinal complication, the shifts

northward and southward of rainfall isohyets run roughly parallel at the various time-

scales discussed here.) These are the dramatic realities of climate change that had to be

responded to by the LSA peoples of this story – and we need a visual model to help us

think about the multiple, interlocking dimensions of that response.

FIGURE FOUR [Idealized Lorenz phase transform diagram] forces one to think

about Sustainability (Low to High on the vertical axis) in terms of two graduated,

correlated scales (on the horizontal). The first is a measure of Broad Demography, or

population diversity from high to low. Please note that this scale is not segmented off

into “urban”, “village”, “hamlet” population sizes; that would be defeating the subtlety

implied. Broad Demography here works best as an organic balancing of gross population

numbers against numbers or other attributes of segmentation (in the case of the Middle

Niger, ethnicity is important, but corporate belonging even more so). The second

horizontal axis is “Authority” – spanning from “low” (chaos) through the full gradations

of persuasive, to hierarchical power, to a stifling “high” despotism that would spawn

mass revolt and out-migration.

The plateaux-like polygons encased within this Figure Four are, of course, merely

illustrative of a possible infinity of potential pathways to urbanism. They illustrate three

core beliefs I have come to harbor, and hope someday will be testable, about the incipient

stages of Middle Niger urbanism as a case of self-organizing landscape:

   1) Ascent (and descent) from one level of Sustainability to the next is not gradual. A
       significant transformation of landscape beliefs (or realization of a new precipitate

       of social memory) has to happen for a new mode of the (Lorenz) transform to
       work and I believe that these events were fairly sudden, separated by long periods

       of status quo – the classic metastable equilibrium events. When I first introduced

       this model in my 2005 Ancient Middle Niger book, this was to illustrate my belief

       that emerging urbanism in the Middle Niger could usefully be described as a self-

       organizing phenomenon. I remain convinced that this is a useful heuristic.

   2) As one “ascends” the Sustainability scale through these stacked polygons, there is

       no right or best positioning vis-à-vis the Population (Diversity) against the

       “Authority” axes. Clearly, some cartesian positions just won’t work. But a
       pathway described for pre or early Hellenic Greece or Mesopotamia would

       certainly look quite different than the one to be introduced here (as would the

       polygons describing the Aegean) – and that one is just a guess at that.

   3) As one gets higher on the Sustainability ranking, the polygons are smaller (i.e., it

       would be easier to tip Population (Diversity) or Authority (or both) too far – and

       the complexity of the system “falls” off to a lower level of sustainability. Recall

       in this the classic definitions of “collapse”. Keeping one’s balance is a tricky

       thing and even if one keeps one horizontal axis variable constant, maneuvering

       along the other axis risks a tumble (as here, with the uppermost level, if Authority

       is kept roughly in the middle and one tries to tightrope-walk along that narrow

       defile towards higher Population [Diversity]).

Why bother at all with such a model? This is where the previous discussion of social

memory comes in. It is my belief that the successful discovery by ancient peoples of a

complexity trajectory that will allow a highly sustainable form of urbanism (as I believe

we can argue Middle Niger cities were) is a highly fraught exercise. Success is never

determined or pre-ordained from a previous state. Blown away as we are by the palaces,
the trappings of power, and the monumentality of, say, Mesopotamia or Mesoamerica, we

forget that these cities were prone to periodic collapse, or were relatively short-lived
when compared to the two thousand (plus) year history of urbanism in the Middle Niger:

sustained, ever growing in population and complexity. Every new field reconnaissance in

each previously under-appreciated basin of the Niger’s interior floodplain reveals

hundreds more tells of urban dimensions – depressingly (!), they seem to become ever

larger and older the more just-basic discovery we do! This is all the more impressive

given the evidence that the chronic climate (hence, physical landscape) instability we

know from the instrumental record was rarely less, often even more. At the millennial

time-scale climate change was complemented by episodes of climate-forced long-distant
migration from now barren reaches of the Sahara up to 24 degrees north latitude down to

the distal reaches of the Niger Bend palaeochannels (c. 17 degrees) and ultimately to the

last two “live” (inundated) basins, the Upper Delta and the Macina (at 14 to 15 degrees).

I believe there were transforms to the social memory, and to the general landscape

conception, of these ancient peoples that allowed the successive settlement and landscape

transforms I discuss below. Key would have been transformations to the conception of a

landscape that not only would have peacefully accommodated the presence of other

ethnicities and subsistence and artisan corporations, but also would have required easy

access to the goods and services rendered by these new neighbors. Within a generalized

and deep-time trajectory of greater complexity to the rules of reciprocity would have

been (self-organizing) moments of invention of non-pre-ordained expansions to the

worldview. This was the key to resilience – and eventually to sustainable urbanism.

Purely as a illustrative device, I use here the Four Transforms of Middle Niger

Clustering [FIGURE FIVE] from Ancient Middle Niger to project an idea of the subtle

mix of evolving specialization, sedentarism, introduction of new subsistence practices
and artisan crafts that went behind the dimension of Population (Diversity).
How, then, to assemble the empirical evidence for abstractions such as “social memory”

or “deep-time trajectory of self-organizing landscape”? What has changed since this

model was first introduced in 2005 has been the publication of reconnaissance and

geomorphological research in the Méma and (especially) several seasons of basic

exploration, with survey and excavation, along the major palaeochannel in the Timbuktu

region of Mali’s Niger Bend and along several of the major lakes, to the west, in the

Lakes Region. Given the conditions under which it was undertaken, this research

represents an heroic expansion of our knowledge of the origins of Middle Niger back
some 1,500 to 2,000 years earlier than the better-understood foundation of Jenne-jeno in

the southern Middle Niger. It now allows us to say more about the linkage of the third

and second millennium BC LSA pastoral settlements of the Mauritanian Tichitt region –

posited by Kevin MacDonald [FIGURE SIX: MacDonald thesis of Migration into the

Méma] as a potential source of migrants, via the palaeochannels of the Dhar Nema, into

the Lakes Region and Mema, as the Sahara became less and less hospitable [FIGURE

SEVEN: Hodh to Mema/Lakes Region].

STAGE ONE: How frustrating, decades after the first systematic survey and dating of

hundreds of distinctive stone ruins in the Dhar Tichitt region of Mauritania, still only to

be able to speak of a generalized “sedentary agro-pastoral” adaptation [FIGURE

EIGHT: LSA Stone Cliff-top Ruins of Dhar Tichitt]. Hundreds of multi-component

LSA ruins dot the bottom (pediment) slope and tops of the Dhar’s escarpment, with other

non-stone, small tells and scatters apparently associated on the floodplain of the once-

vast playa. Sadly Munson’s systematic survey and test excavations, thorough -- however

much criticized by some as resulting in an overly-precise radiocarbon and cultural
sequence -- have not been repeated since. So, bedeviled as they are with issues of

contemporaneity of sites of the same class, but spread out over hundreds of linear
kilometers along the scarp, discussion has devolved into a dichotomous argument of

“seasonality” versus “sequential driven by 1500 years of desiccation”. That is, some

scholars have argued, sequentially, for a hunter-gatherer and then herding adaptation that

moved (1) from the banks of the high lake (sometime before 4,000 bp) to (2) a desperate

“down-bathymetry” movement following the retracting lake, to (3) establishment of large

stone architectural sites on the pediment (where early grain cultivation soon complements

cattle herding) and finally to (4) even larger sedentary herding-oriented villages on the

scarp tops (eventually collapsing into marginal defensive positions and eventual
abandonment by c. 2400bp [5]). Others conflate the entire sequence into one unchanging

“Neolithic seasonality” unity, with the playa settlements representing dry

season/declining lake occupations, and both the pediment and the scarp-top architectural

sites represented a wet season time of herding and millet-cultivation.

Ignored in these reconstructions (and what must be tested with far better chronological

and survey data) is that the Dhar Tichitt sequence occupies the latter half of more global,

long, drawn-out flight of Saharan peoples, beginning c. 6400 bp with high amplitude arid

episodes disrupting the Sahara’s second Holocene pluvial. The instability (century-scale)

of the climate intensifies after c. 4,200 bp, while the drying trend (millennial-scale)

intensifies. These peoples, and those occupying the >20 degree north latitude reaches of

the once inundated Azawad basin of the ancient Middle Niger, will eventually find refuge

first in the now-senescent “dead” basins of the Middle Niger (the Niger Bend/Lakes

Region and in the Méma discussed soon) and lastly in the still inundated “live” basins to

the south. MacDonald hypothesizes that many peoples, with many emergent

specializations, were thrown together in these more southerly refuges – how… and why,

under these conditions of extreme dislocation and stress did they interact apparently so
To this we might add the former inhabitants of the hundreds of LSA “agro-pastoral”

settlement of the Tagant to the northwest of the Tichitt region – whose adaptations and

eventual destination are poorly understood, but who most probably fled south as well.

But what we lack for the Middle Niger generally (and have evidence for in spades in the

Middle Senegal River Valley, another fluvial refugia, far to the west) is evidence for

conflict among the new (and original?) arrivals. How curious is this lack of conflict for

these areas still with water, granted, but with the same or intensified contexts of climatic

stress and surprise.

It would be entirely conceivable (once the political and banditry issues are resolved) to

construct an empirical test of another model of interaction. FIGURE NINE [Stage One

pathway] is an attempt, within the framework of the “Sustainability” transform model

presented above, to predict an initial situation of short contact and exchange – leading

conceivably to a seasonal, predictable co-occupation of a locale. In this model, this is a

critical moment of discovery of the “rules of engagement” for sustained interaction of a

few communities of proto-specialists. All communities might do a little of everything,

but – if successful – the promise of exchange and reciptrocity amongst the groups in

times of stress would conceivably lead to a greater thrust towards true specialization (in

Figure Nine we see fisherfolk, herders, and gatherers of the local grains, wild millets and

sorghums, perhaps even fonio and red rice that will someday become domesticates).

STAGE TWO [FIGURE TEN: Stage Two pathway]: This is the stage for which we

currently possess the least empirical evidence. It would represent the critical evolution of

landscape forms for which we are beginning to find evidence in the field. Starting as

early as perhaps 4000bp, we can imagine the movement of some of the Tichitt herders
(and the proto agro-pastoralists of the Tagant?) (and very likely the generalized, acquatic-

oriented hunters /gatherers /fisherfolk of the great playalands of the Azawad?) to the
south, attracted by the still-vibrant lakes and waterways of the Middle Niger’s Lakes

Region/Niger Bend. In some cases the movement would have been facilitated by the

connecting series of (generally) north-south oriented palaeochannels that crosscut the 54

degree-trending late Pleistocene dunes of the (now) southern Sahara (then, of course, the

northern Sahel). On present evidence, these peoples would have moved down

palaeochannels such as the Wadi El-Ahmar (visible just east of Timbuktu) and,

especially, through the vast anastomosing channel network “draining” the southeastern

Hodh into the Méma.

From a suite of complementary satellite imagery we can easily see stone architectural

sites analogous to those better known around Tichitt, on the scarps of the Dhar Nema

region of southeast Hodh [FIGURE ELEVEN: Dhar Nema architectural sites] and on

the inselberg exposures of the Lakes Region. In the Lakes Region [FIGURE TWELVE:

Lakes Region], where preliminarily surveys and surface recordings have been

completed, and in the Dhar Nema (where unsafe conditions still impede work on the

ground) these architectural sites lie near “tell” sites of varying size (some, like Fait 6 in

the Lakes Region Gorbi Valley reach an impressive 5 ha.) [FIGURE THIRTEEN: Fati

6 seen from nearly stone ruin]. Lakes Region sites in apparent close proximity include

(1) mountain-top architectural sites, (2) elevated tells with stone architecture, and (3)

small to large tells with lesser accumulation – probably because construction was of

wattle-and-daub or dry mud walling (tauf).

The amplification of the seasonal co-occupation and co-evolution implied by Figure Ten,

where more and more specialists spend more and more months of each year in proximity,

would only work, of course, if the sites are contemporaneous. This we cannot yet verify
for the Dhar Nema; we only begin to have suggestions of this for the Lakes Region. But

we do believe this is the case (as well as the limits of radiocarbon dating allow (!)) for the
Méma. Here acquatically-oriented hunters/fisherfold (present at least as early as 4,000

bp.), by 3300 bp “welcome” semi-specialized herders to share lake-beach and levee

localities. And, by 2850 bp, they are joined by proto-cultivators. Figure Ten implies that

not only are the “rules of engagement” for peaceful interaction amongst these groups

maturing, but also that the sense of security these networks of reciprocity provide

encourage certain communities to become even more specialized (with, perhaps the

appearance of specialist artisans, such as potters). These later Stage Two sites have all

the look of essential permanency about them, meaning that the majority of the
heterogeneous super-community may have stayed put all year.

STAGE THREE [FIGURE FOURTEEN: Stage Three pathway]: Our best evidence

for the next step of the progression comes from extensive survey and excavation of

several large and small tells of the Méma [FIGURE FIFTEEN: Kobadi, Méma &

FIGURE SIXTEEN: Akumbu settlements and geomorphology & FIGURE
SEVENTEEN: Akumbu Cluster, Méma] and, especially of the Timbuktu region,

dating to the LSA-Iron Age transition (or, about 2500 bp). Clustered LSA sites in

preferred locales of the Méma transmute into clustered Iron Age mounds (including,

eventually, several new satellites devoted solely to the loud and dirty [and polluting?]

craft of the iron smelters) [FIGURE EIGHTEEN: Méma Geokistics].

In the Niger Bend, LSA peoples still with the acquatic-skills of the Azawad descend

rapidly into the region, occupying small but growing tells preferentially sited on edges of

Pleistocene dunes in close proximity to the (still-flowing) palaeochannels [FIGURE

NINETEEN: Maximal Niger Bend/Lakes Region]. This flushing of the now-Sahara is

encouraged, surely, by the climatic Big Dry. This is the trying period from c. 300 BC to
AD 300 of probably -20% precipitation, during which many if not all of the large lakes of

the Lakes Region dry out completely.
This climate crisis, surely, would have encouraged permanency of occupation at those

restricted numbers of locales blessed with year-round water – making the continuation of

the reciprocal, peaceful relations amongst proximate communities implied by Figure

Fourteen all the more critical. If the implied social contract is successful, then, the stage

is set for the appearance of true cities. As precipitation again improves (increasing

steadily to perhaps +20% of modern baseline rainfall by AD 1000-1100), the explosion

of site size in the Timbuktu region from c. AD 450 becomes entirely understandable. We
understand this best with the completion of three seasons of extensive survey and

excavation of several tells flanking the major palaeochannel, the Wadi El-Ahmar just east

of Timbuktu [FIGURE TWENTY: Tells of the Wadi El-Ahmar]. Tells are of urban

proportions (50-100 ha) during the later parts of the climatic Stable Optimum. (c. 770-

1100). [FIGURE TWENTY-ONE: Stage Four pathway].

Almost as an afterthought to this story of urban emergence, from the beginning of the Big

Dry and accelerating during the subsequent amelioration of climate in the mid-first

millennium of the Common Era, population streams into the previously unexploited

“live” basins of the southern Middle Niger [FIGURE TWENTY-TWO: “Live” basins

of the Middle Niger]. Jenne-jeno explodes on the scene, already large (>20ha.) at its

foundation. But here the focus tell [FIGURE TWENTY-THREE: Jenne-jeno] remains

at a “mere” 33 ha throughout its maximal prosperity during the mid- to later first

millennium AD – the urban population being “proximally dispersed” amongst scores of

specialist-oriented satellites (more on which below). Pity poor Jenne-jeno, apparently

overshadowed in mid-first millennium AD by the contemporary 100ha-plus tells of the

Méma and Lakes Region/Niger Bend.
But all good things… As robust as our “sustainability” model has been for over 2,000

years, things are severely tried by the end of the climatic Stable Optimum at c. AD 1000

– 1100. Now comes the commencement of many centuries of unstable, climate

oscillations (high amplitude and high temporal). This (and, surely, the incursions of

martial peoples from southern Mande, most spectacularly during the expansion of the

Mali Empire in the early fourteenth century) shakes the bonds linking specialist

communities. Some Middle Niger basins suffer massive and rapid depopulation. The

Timbuktu region, at about AD 1000, records a remarkable population drop-off and that
date is our best guess for a similar depopulation in the Mema as well. It appears that

people simply transplant to the still-“live” basins, especially to the Upper Delta and the

Macina. This adds to the already large and urban Jenne-jeno which , however, becomes

just one of 70 contemporaneously-occupied mounds in a four kilometer radius –

comprising what we have taken to call the Jenne-jeno Urban Complex [FIGURE

TWENTY-FOUR: Jj Urban Cluster - excavations]. It is here that we have our best

evidence for the functional differentiation of the cluster’s components [FIGURE

TWENTY-FIVE: Jj Urban Cluster – functional]. [FIGURE TWENTY-SIX: Aerial

Jj] Dia in the Macina, may have been occupied even earlier and is similarly surrounded

by a complex of satellite mounds – the Clustered City.

FINAL THOUGHTS ABOUT SOCIAL MEMORY: Three and a half decades of

research in the Jenne region and we can say several things, with considerable confidence,

about the final form of ancient Middle Niger urbanism. Firstly, clustering appears to

have been a time-established solution to the problem of how different specialist

corporations might maintain their distinct identities, while at the same time having

immediate access to other providers of needed goods and services and access to those to
whom they themselves provide. Occupation by some members, and “corporate

ownership” of individual mounds by all members of an occupationally-defined corporate
group would have been a strong symbolic reinforcement of identity. The archaeological

expression of this phenomenon is the near-exclusivity of occupational debris at the

satellites. (Whereas all subsistence and artisan activities took place at the large focus

mound of Jenne-jeno.)

Secondly this was a highly sustainable form of urbanism – from its foundation in the

third (or earlier) century BC, the city of Jenne-jeno and, indeed, the Jenne-jeno Urban

Cluster, was large and remained so right up to a century or two (at most) of its
abandonment at c. AD 1400. As opposed to other similar Sahelian tell-occupation

situations, such as the Middle Senegal Valley, there is no evidence during that long urban

flourish for conflict or warfare.

And thirdly, despite some twenty-two units sunk into the focus tell (and 35 more in

satellites) during that three-plus decade period since Jene-jeno’s “discovery” in 1977, we

have no evidence of a state-like, top-down, elite-driven political engine powering this

urbanism along through time. No kings, no citadels, indeed, no obvious elites – the

alternative term heterarchy appears the best description of the political and economic

organization of late first millennium BC and later urbanism in the Middle Niger. If our

four-stage model of initiation of urbanism proves to be empirically confirmable in the

field, then we will be confident to talk about some degree of proto-heterarchy or

initiation of a trajectory of self-organized landscape going back perhaps two millennia

earlier. That presumes a lot of hard-slog in the desert.

To what extent are lessons from the Middle Niger applicable to other parts of the

urbanizing ancient world with at least superfially-similar clustering (Lungshan northern
China and pre-Uruk northern Mesopotamia come immediately to mind)? To what degree

can we say, elsewhere also, that a pre-urban landscape of (putative specialist?) satellites
revolving about a nominal focus-community (or, for that matter, an arena of tightly-

packed equivalent-size, but occupationally differentiated communities) implies a long

history of multi-corporate accommodation? Driven by what? Here we descend into the

murky depths of the motivations and socially-constructed perceptions of the long dead.

Are the motivations of the ancients forever a closed book to archaeologists? Or is there

any room for optimism that past intentions can be at least partially revealed through an

investigation of how symbols and objects function as devices or insignia communicating

peoples’ view of themselves and of their bio-physical landscapes? The concept of social
memory, as used here, is predicated on the hope that we can go beyond a simple-minded

use of climate change (including stress and surprise) and migration as causal conditions

for the development of the Middle Niger’s distinctively heterarchical, deeply sustainable

(and resilient) clustered urbanism.

I believe that the four stages of the developing Middle Niger “pre-urban” landscape

describe the creation of a self-organizing landscape that both nurtured specialization (that

is, multiple specialists articulated in a generalizing regional economy) and displayed

great resilience in the face of a demonstrably and brutally surprise-laden bio-physical

environment. Key to the emergence of hundreds of clustered cities over the vast Middle

Niger floodplain(s) was an evolving social memory -- the ways by which communities,

understand, curate and transmit knowledge of past environmental states and possible

responses to them.

Now begins a long empirical campaign to test whether the social codes and social

memories etched into the settlement patterns leading to full urbanism can be fully

revealed by dirt archaeology. To again quote Balée and Erickson: “The landscape is like
a text, but not one that is readily accessible to historians’ and epigraphers’ methods

because it is not written in a decipherable script, but rather is inscribed in a subtle,
physical sense by learned, patterned behavior and action – what anthropologists

traditionally refer to a culture. Culture is physically embedded and inscribed in the

landscape as nonrandom patterning, often a palimpsest of continuous and discontinuous

inhabitation by past and present peoples.”

To top