GaWC Research Bulletin 284

GaWC logo
  Gateways into GaWC

This Research Bulletin has been published in R.E. Lee (ed) (2012) The ‘Longue Durée' and World-Systems Analysis, Albany, NY, State University of New York Press, pp. 35-64.

Please refer to the published version when quoting the paper.


History and Geography: Braudel's 'extreme longue dureé' as Generics?

P.J. Taylor*

Histories (Time)

Over the last decade I have been studying contemporary globalization through urban lenses. In this literature globalization is often interpreted in geographical scale terms as a pincer movement on the state with power moving both upwards and downwards. This is entirely consistent with my long-term concern for thinking beyond the state, which originally brought me to world-systems analysis some three decades ago. As a result I have found studying globalization to be intellectually exhilarating because it has sharpened my thinking outside the knowledge box that is modernity. Globalization is a severe challenge to the embedded statism of the orthodox social sciences: put simply, to understand contemporary globalization it is sometimes more useful to take stock of social change processes before the 200-year modernity geocultural interlude. This is the case when the specifics of social change in modernity blind us to the wider generics of social change in human history.

This train of thought has led me to Braudel's (1972a) reference to the ‘extreme longue durée ' (p. 41, p. 45) in his classic 1958 paper on the three time spans of history (events, cyclical movements and longue durée ). In his discussion of the extreme longue durée as a possible fourth time span, Braudel explores ‘'timeless realms of social mathematics' (p. 47) and ‘myths … developing slowly' as examples of what he also calls ‘very longue durée ' (p. 46) and ‘excessive longue durée ' (p. 47). Wallerstein (1991) has interpreted these as ‘normative social science' (p. 144) and, following Braudel, ‘the time period of the sages' (p. 137). Both scholars are sceptical about the relevance, and even existence, of this ‘eternal time'. However, it seems to me that it does not have to be quite so eternal. The extreme longue durée can be returned to the historical if it is interpreted as the generic form of institutions, collective processes that can be traced over very long time periods. This is, of course, Marxism's transhistorical concepts; in world-systems analysis we might call them trans-system processes. The idea is that there is an institution/concept/process that has a specific reality within a historical era, which has been created from an existing generic reality. The exception is first creations of the process, the historical beginnings that have their own specificities but also lay the foundation for all subsequent historical developments. The classic Marxist example is class conflict; Braudel's discussion of Lévi-Strauss's work on communication/language fits this position well (pp. 43-4); in this paper I concentrate on city/state relations in the extreme longue durée.

This choice derives from my engagement with the globalization literature wherein world cities are often described as breaking free from their state containers. In particular, in my own research the role of cities has been found to be at variance with the established paradigm of macro-urban studies. Here cities were deemed to form ‘national urban systems' with strict city hierarchies ultimately derived from central place theory and related simple notions such as the rank size rule. Cities in globalization simply do not fit into such conceptualizations. In other words the specifics of the modernity interlude seems to be giving way to a new global specifics requiring new models and theories. But, as is the nature of history, the (modern) present had been widely projected backwards to other periods – historians appear to love central place theory. And, of course, central place theory and the rank size rule have featured prominently in urban historical research in world-systems analysis. Hence part of the purpose of this paper is to update urban models in world-systems analysis while at the same time interpreting city/state relations in terms of the specifics and generics of cities and states.

One final point on histories: Braudel is adamant that his three time spans ‘fit into each other neatly' (p. 48) and therefore the ‘cardinal error' is ‘to choose one of these histories to the exclusion of all others' (p. 34). If we rehabilitate the fourth span by returning the extreme longue durée to history then the above positions need to be extended to four histories, none of which can be ignored.

Geographies (Space)

According to Braudel (1972a, 47) ‘time sticks to [the historian's] thinking like soil to a gardener's spade'; the same can be said for the geographer and space. Braudel ends his essay with a plea to bring geography into his proposed history-social sciences dialogue in terms of ‘spatial models' and ‘a concentration on place' (p. 52). Of course, Braudel's books are awash with geographical concepts integrated into his history: for instance, he begins what is arguably his most influential book, The Perspective of the World (Braudel (1984), with a discussion entitled ‘Economies in space' wherein geography is treated as a set of ‘ground rules': ‘tendencies which will clarify and even define [world economies'] relations with geographical space' (p. 25). Thus are world economies treated generically and classed with other concepts such as ‘civilizations, states and even empires' (p. 24). But geographical thinking has moved on from treatment of geography as a backdrop to history and social sciences.

Social spaces are constructions that, like social times (Braudel's historical times), are integral to social change: societies do not pass through time in space but are constituted as time and space. This is explicit in Wallerstein's (1991) conception of historical systems with their cycles and core-periphery structures and is spelt out clearly in his extension of Braudel's historical times into TimeSpaces. These are defined by adding space-scopes (geographical scales) to Braudel's time-spans (p. 139). But his identification of this ‘space parallel' to Braudel's times does not take into account space's extra dimensions compared to time's uni-directional nature. This difference between space and time means that there are non-parallels between construction of social spaces and social times; there is more to geographical spaces than geographical scales.

I have found Castells' (1996) theory of social space to be particularly useful for identifying spatial concepts that encompass the dimensions of space. He interprets globalization as a new network society that is replacing industrial (modern) society. With the advent of instantaneous electronic communication, he argues that the necessity for contiguity and simultaneity in social relations (place) is being superseded by simultaneity only (flows) to define a new social era. In terms of the construction of social spaces, this is interpreted as old ‘spaces of places' giving way to new ‘spaces of flows'. Without buying fully into this social theory, these two concepts of spatial form are extremely useful tools for understanding how social spaces are constituted in social change. Thus although Castells' interest is in the changing strategic balance between these two space constructions in contemporary globalization, the two concepts can be deployed more broadly.

In order to use Castells' (1996) space concepts generically it has to be understood that they are abstractions from a single social space construction. All spaces of places depend of spaces of flows for their construction and vice versa. This is proven by the limiting cases that turn out to be impossible social spaces: a space of places without flows is an unliveable mosaic space; a space of flows without places is an unliveable fluid space. Hence in this theory of social space both construction processes are always in operation and the interest lies is in the relation between them in specific instances (like Castells' network society). In this paper I am interested in the spaces of flows that are city networks, the spaces of places that are state territories, and their interaction in different historical systems. But in order to use Castells' spatial concepts in this way it is necessary to imbue them with relevant social substance, in other words a theory of cities and a theory of states.

Making a Living (Work)

In giving the two spatial forms social substance I take a materialist approach that privileges work. Social spaces are what are formed through the everyday work processes in making a living. It is individuals in households who make a living but they can only do so in a collective enabling context that has to be ongoing: social spaces are a critical element of this necessary reproduction. According to Jacobs (1992) there are, in fact, only two basic ways of making a living: as part of the commercial world (making and trading) or as part of the guardian world (taking and governing). These are very different and require contrasting work ethics enabling their reproduction.

Trust is the basis of any ongoing social enterprise; once trust between participants breaks down so to does the social institution. Jacobs (1992) delineates two moral syndromes that she argues are necessary for the reproduction of commercial and guardian work. Jacobs identifies commercial and guardian moral syndromes through which trust is established and reproduced to make work possible. Each syndrome consists of a set of behavioural ‘precepts', one set that enables the reproduction of commercial work, and a very different one for guardian work. These syndromes are best introduced through what Jacobs terms their ‘key precepts', be honest and be loyal respectively. Successful commerce requires participants to be honest in their transactions; for instance, markets require honesty to build the trust to operate. Guardian work, on the other hand, has its trust built upon loyalty between participants; for instance, wars require loyalty to justify the sacrifice. To be sure there will be dishonesty in commerce and disloyalty in guardian work but the point is that when these reach a certain level a tipping point occurs and reproduction of their respective work practices becomes impossible: markets collapse, armies disintegrate. To emphasis that these are two contrasting moralities note that honesty and loyalty are not inter-changeable: loyalty in a market is bad business; honesty in war preparations is a recipe for defeat.

The two complete sets of precepts are listed in Table 1 where their divergence is all too apparent. I have grouped precepts into clusters to ease interpretation and comparison. Where do these syndromes come from? Jacobs has gleaned these precepts from a very large-scale textual analysis of reports on different appropriate and inappropriate work behaviours and numerous texts about what is and what is not suitable behaviour in myriad work contexts. Jacobs claims that the syndromes derive from ‘millennia of experience with trading and producing, on the one hand, and with organizing and managing territories, on the other hand' (p. xii). In other words they are generic to human work practices. Whether trading and producing in Uruk Mesopotamia, Inca Peru, Ming China or contemporary USA this work requires honesty to be supported by not forcing agreements, non-discrimination, keeping to contracts, enterprise, inventiveness, dissent, and productive investment. Whether raiding and governing in Sargon's Akkad, Republican Rome, Shogunate Japan or the USSR this work requires loyalty to be supported by exerting power, building tradition, respecting hierarchy, deceiving, discriminating, being ostentatious and dispensing largesse.

One final point: although the commercial syndrome appears to be innately more preferable to modern sensibilities, Jacobs emphasizes that in any society both commercial and guardian moral syndromes are necessary. And each syndrome is implicated in a different social space construction.

State Territories (A Constellation of Guardian Practices)

Jacobs (1992) is explicit that the guardian syndrome is implicated in territories in terms of ‘territorial responsibilities … the work of protecting, acquiring, exploiting, administrating, or controlling territories' (p. 29). But she does not develop an associated theory of states. For this I turn to Scott's (1998) treatise on how states operate to ‘radically simplified designs for social organization' (p. 7). This is a generic theory of states.

For Scott (1998, 2), ‘legibility is a central problem of statecraft'; the more legible a society, the better a state can carry out its ‘classic' functions of ‘taxation, conscription and prevention of rebellion'. This is directly reflected in the construction of state spaces: quite simply, for state activities, spaces of places are more legible than spaces of flows. Thus are policies predicated upon spaces of places. Furthermore, under the right conditions, spaces of places can be redesigned in the image of the state itself as ordered, hierarchical and regimented. Spatially this simplification takes geometric forms such as the Roman Empire's network of straight roads and grid-iron cities, Post-Revolutionary France's equal-size ‘hexagonal' departments, the straight-line boundaries common to European and European-settler imperialisms in the Americas, Africa and Australia, and the planning of ‘high modernist' cities like Brasilia (Taylor 1999; Scott 1998). These are all guardian spaces reflecting political power.

According to Scott (1998, 79) the techniques for projecting state power have been generic: ‘appropriation, control, and manipulation … remain the most prominent' (p. 79). These have made centralization and territoriality ubiquitous to the political process. Accumulating power and defending it is most easily accomplished through territorial behaviour: demarcating a space, controlling access to that space and influencing the composition of the space. The modern inter-state system is a straightforward example of this whereby mutual acceptance of territorial sovereignties is the foundation of international law. But this worldwide political mosaic is very specific to modernity; more generally political territories are separated by frontiers rather than lines. Whereas modern states aspire to territorial homogeneity behind boundary lines, previous political territories more typically consisted of core zones of vital people and resources surrounded by a land buffer: the Roman Empire's division of its British province into a productive civilian zone in the south and east behind a military defensive zone in the north and west is a classic example of such territorial behaviour. Smith (2007) describes exactly the same structure for ancient states. And the United Provinces operated in this way in its 80-year war of independence. Dutch territory was treated as two zones, a core zone of maritime cities and buffer zone to act as a bulwark against invasion (Taylor 1994); Braudel (1984, 202) provides a map of the ‘fortified island'. Of course, it was the success of this ‘un-modern' territorial policy that made a major contribution to the rise of the modern inter-state system formally adopted at Westphalia in 1648.

Scott (1998) argues that states' practices of social simplification cannot adequately deal with ‘essential features of any, real functioning social order' (p. 6). Whatever states have in mind, societies remain very complex matrices of flows and function on the basis of myriad ‘informal practices and improvisations'. In other words guardian state practices creating spaces of places cannot properly deal with the complexities of spaces of flows, as exhibited by cities in particular. All states, because of their inherent guardian territorial imperative, invariably treat ‘people who move around' badly (p. 1). Scott lists the following who ‘have always been a thorn in the side of states': ‘slash and burn farmers, nomads, pastoralists, hunter-gatherers, Gypsies, vagrants, homeless people, itinerants, and runaway slaves and serfs' (p. 1). In other words, states have problems dealing with people whose lives are centred on spaces of flows. Note, however, that Scott does not include long-distance merchants in his list, presumably because their wealth usually precludes them from becoming victims of states. Nevertheless, they too have a history of undermining state policies. But guardian practices cannot ignore such flows. For instance, Braudel (1972b, 308-9), after listing the fall of several cities to territorial expansion of states in early modern Europe, illustrates the limitations of state territoriality:

‘The victorious states could not take control of and responsibility for everything. They were cumbersome machines inadequate to handle their new superhuman tasks. The so-called territorial economy of textbook classification could not stifle the so-called urban economy. The cities remained the driving forces. States that included these cities had to come to terms with them and tolerate them. The relationship was accepted the more naturally since even the most independent cities needed the use of the space belonging to territorial states.'

This restates the points that spaces of places and spaces of flows are always entwined and that guardian and commercial practices are inevitable inter-locked. Let us now see how a theory of cities can show how spaces of flows are created through commercial practices.

City Networks (A Constellation of Commercial Practices)

The theory of cities I use is that of Jacobs (1969, 1984) supplemented by GaWC research on contemporary cities in globalization. The basic starting point from both Jacobs (1969) and Castells (1996) is to treat cities as a process: networking practices that create spaces of flows. Thus cities are never singular, they come in groups that form city networks.

According to Jacobs (1969) cities are the source of economic expansion; it is through the unique co-mingling of myriad people from different contexts (cosmopolitanism) that economic changes are invented, adopted and adapted. Put simply, this expansion in commercial life is the result of two economic externalities – extra-market advantages – that occur through cities. The first are externalities within cities resulting from knowledge-rich economic milieu wherein commercial practices generate and maintain the latest ways of conducting business in both production and services. These are economic clusters through which Jacobs ideas have made inroads into economic theory (Glaeser et al. 1992). Second, there are externalities between cities resulting from knowledge–rich connections through which commercial practices are informed of and adopt or adapt new business. These are economic networks in which cities are the nodes of this commercial work. In effect I am generalising Hicks' (1969) internal and external growth through cities in the ‘Mercantile Economy' to commercial activity across all societies.

In and of themselves cluster externalities and network externalities do not generate the full potential of city economic expansion. It is when they interact that something special happens through import replacement. For Jacobs (1969) import replacement in a city frees up new resources for cumulative growth in massive spurts of new work. Thus it is that economic expansion occurs in cities through their economic clusters within their economic networks. As well as expansion resulting in city growth, the process also produces geographical extension through growth of the city network. It is the latter we have been studying in GaWC where we have developed an interlocking network model of inter-city relations through analyses of the worldwide office networks of advanced producer service firms (Taylor 2001, 2004). Unlike other network models this one has not two but three levels: the net level in the world economy, the nodal level (which are the cities), and, in addition, a sub-nodal level of firms with offices in numerous cities. It is these financial, professional and creative service firms that ‘interlock' cities through their everyday work generating a largely electronic space of flows between offices. We now treat this model as a generic theory of cities – central flow theory – since the interlocking process can be found in all vibrant city networks: ‘agents' in some form or other are required across cities to facilitate and augment commercial work (Taylor et al. 2008). Good examples where this process is clearly evident are in the commercial practices of medieval Europe (Pirenne 1969; Spufford 2002) and ancient Mesopotamia (Stein 1999; Algaze 2005). We call this network space-making the process of city-ness.

City-ness does not exhaust the category of urban work and activities; we identify also the process of town-ness. This is a space-making process that encompasses the work that relates an urban settlement to its hinterland. Whereas city-ness is non-local and complex because of its extra-market basis, town-ness is local and simple based upon just markets. This is the process that central place theory describes with its market areas organised into a hierarchy. It is not relevant to economic expansion because hinterlands alone cannot generate new work. But the simplicity of central place theory makes it popular with state guardians in their planning practices and administrative organization. However, their neglect of central flow theory and therefore cities is not a problem because economic expansion is not a matter of guardian work through its spaces of places; it is what commercial work creates through its spaces of flows. This is why state economic development planning is generically doomed to failure as ‘second' and ‘third' world guardian practices in the last quarter of the twentieth century illustrated all too well (Taylor 2006a).

The division between local and non-local is a common-place distinction. For instance, Hicks (1969) refers to ‘petit commerce' and ‘grand commerce', and Braudel (1982, 376) shows this to be reflected in different terms used for trade in different languages (Table 2). Interestingly he notes that the ‘long distance' merchants have typically come to form a ruling group (p. 68), that is to say their original commercial prowess is converted into guardian status. The theoretical arguments above allow us to develop this further. Jacobs (1969, 135) notes that in medieval Europe, guilds were divided into three types: ‘merchant guilds, strictly local guilds, and craft guilds'. The first two correspond to Braudel's list of ‘traders' and the third one is about producers. The merchant guilds ‘settled down with warehouses, counting houses and agents in this city and that' (p. 135) indicating that they produced networks but not necessarily economic expansion as they transmuted into ‘merchant-princes'. The local guilds – bakers, butchers, coopers, and other shopkeepers – were concerned with urban hinterland relations, buying off farmers and merchants and facilitating exchange. The crafts guilds – weavers, saddlers, smiths of various hues – are further divided by Jacobs into ‘shopkeeper-craftsmen' and ‘merchant-craftsmen'. The former are local (hinterland) but some expand their trade into non-local markets to become their own merchants (p. 136). Unwin (1963, xxx1, 61) identifies these crafts, especially the weavers, as being as the main economic challengers to the merchant princes. And it is their combination of producing and trading that Jacobs singles out as particularly important for economic expansion: they are able to seek out and expand specific new markets for their goods in ways that general merchants did not and really could not do. Thus are merchant-craftsmen central to expanding city economies in medieval Europe. But where trade and crafts are kept apart (e.g. in caste systems) cities are stultified and economic expansion diminished (Jacobs 1969, 134-5). The key point is that this argument about division of work in terms of space-making is generic:

‘Taken together, the three types of guilds give us a picture of the economy of the city that is not only valid for medieval times but for our own times as well: local goods and services that remain local; exports, starting with the initial exports; and local goods and services that become exports' (Jacobs 1969, 137)

City-ness is about economic clusters and economic networks combining in commercial work to expand space-economies.

Town-ness is also generic. Because we conceive of city-ness and town-ness as processes, they can be treated as happening together simultaneously in urban settlements. Thus we argue that all urban settlements encompass both city-ness and town-ness processes but to very different degrees (Taylor et al. 2008). But our focus here will be on city-ness and its making of spaces of flows.

Using History (Beyond Modern Geographies)

If we take globalization seriously as a collection of processes taking us beyond modernity (i.e. the demise of the modern world-system) then history becomes especially important to social scientists. If new structures are emerging then past modern experiences (and theories based thereon) are becoming less salient and may even become obstacles to strategic understanding. In contrast knowledge of times before social knowledges were nationalized in modernity may become relevant. These histories are not tied to events but are likely to be longue durée times or phases thereof (cyclical time).

In this section I employ two brief case studies to illustrate the previous theoretical arguments. I use some detailed historical scholarship to first answer a question about commercial practice and then second to shed light on guardian practices. My chosen examples are economic expansion in Leeds during the early industrial revolution and the governance of Genoa in the middle ages. In both cases I use other cities to show alternative paths of change.

Commercial Practice: Leeds Woollen Iindustry in the 18th Century

The industrial revolution in England is associated with rapid urbanization in the eighteenth century in the north culminating in the great Victorian industrial cities of the nineteenth century. But the process was very uneven: Leeds became a vibrant centre of a great woollen textiles industrial region in the eighteenth century while Newcastle was becoming a stagnant dependent city in a coal exporting region at the same time. The latter city and its port were controlled by an oligarchy of coal landowners and coal merchants prospering from sending coals to London. Through this guardian process other work with non-local potential was prevented from developing (Middlebrook 1950, 64-66; Jacobs 1969, 155). The result was the “Black Indies”, a dependent region like the other two more well known ‘Indies' (East and West) with a one commodity economy and urbanization only sufficient to service the simple trade pattern (Ellis 2001). (Newcastle's great industrial spurt came, unusually, a century later).

Far away from such guardian constraints, Leeds' commercial experience was in complete contrast: it was one of the great economic success stories of the eighteenth century. Woollen textiles were, of course, not a new production of the industrial revolution; they had been the main industrial sector in England since the late middle ages with production centred on the West Country and East Anglia. This geography was still extant in 1700. But by the end of the century the Yorkshire woollen industry centred on Leeds had become dominant. Why? The first point to make is that was not simply because of the rise of the factory system; as Gregory (1982, 47) argues ‘its drive to national supremacy was largely completed in the eighty tears or so before the advent of the factory system'.

It is particularly interesting to find that the historical explanation echoes Jacobs' ideas on production and distribution being combined. In the case of the West Country and East Anglia the process was dominated by merchants who transacting their business through London. In contrast in Yorkshire ‘local mercantile houses were much more important' (Gregory 1982, 57) with the whole process dominated by ‘independent master clothiers'. The process was entirely as Jacobs describes:

‘(Merchants) in Leeds and Wakefield concentrated entirely on the export of cloth; they knew their business as woollen exporters backwards, and they had a deep knowledge of the actual production of cloth unlike their counterparts in London, who relied on factors for their cloth purchases' (Wilson 1971, 52)

The outcome is exactly as Jacobs predicts:

‘The difference between the ways in which the West Riding trade was handled by the active merchants of Leeds, Wakefield (and eventually Halifax) and the exports of every other production area from Norwich down, which were all monopolised by non-specialist London traders often working within the restrictions of the trading companies themselves accounts in good measure for Yorkshire's growing supremacy in the eighteenth century' (Wilson 1971, 52)

Thus it is not just that the traditional value chains in the south of England transferred value to London (another dependency relation) whereas chains in the north retain much value in local cities; in addition chains in the north are expanded to produce new networks of business in ways impossible for London general merchants. Hence the massive economic expansion of Leeds, its region and its business networks that were a critical element of the English industrial revolution.

Guardian Practice: Governing Genoa in the Middle Ages

For the guardian case study Genoa is contrasted with the other three leading Italian cities: Florence, Milan and Venice. As Arrighi (1994) has demonstrated so clearly whereas the latter cities finally went down the territorialist path of city-state formation, Genoa embarked on an altogether different trajectory for its guardian requirements that resulted in what Braudel calls its ‘silent rule' through finance under Habsburg guardians in the sixteenth century. This early-modern separation of economic networks centred on Genoa from territorial governance based upon Castile is truly un-modern (in the sense of no spatial congruence of economy and polity as in the nation-state, see Taylor 2007) and therefore surprises contemporary political scholars. But the roots of this ‘strange' governance arrangement lay deep in Genoa's history.

The rise of ‘outside' governance of Genoa takes place in two phases that broadly reflect the medieval European economic cycle. Arrighi (1994, 88) has shown that in the growth period, city cooperation was commonplace whereas with the onset of stagnation competition and rivalry became the norm and this is reflected in the changing form of Genoa's governance. In the first phase Genoa is not untypical of other northern Italian cities, but diverges from other cities in the dangerous second phase.

The norm of governance for many Italian cities in the thirteenth century was what Waley (1969) calls ‘the institution of podesteria'. This was an instrument to solve the problem of conflict inherent in the concentration of people into cities. This problem had become endemic and critical in northern Italy reflecting imperial versus papal rivalries. A podestà was a ‘single executive official… from another (not neighbouring) commune trained in law and holding his office for six months or a year' (p. 68) who came with a team of lawyers and guards. The key point is in the time (short-term, temporary) and space (from distant, non-competing city) natures of the guardian position that were deemed to provide government ‘above the fray'. Various rules were instituted to reinforce the neutrality through space-of-place making including prohibition on trading while in office and not leaving the city without permission (pp. 68-9). Genoa decided to suspend its government and change to the podestà system in 1190 ‘because of so much civil discord, conspiracies and divisions wracked the city' (Epstein 1996, 88). The first podestà in 1191 was from Brescia (p. 88), a second in 1194-5 was from Pavia (p. 89), and the system continued through the next century until the last podestà in 1296 (p. 325). Epstein (1996, 88) refers to the podestà as ‘a temporary foreign ruler' but in the next century Genoa turned to foreign rulers that were not so temporary.

After the neutrality of the podestà system, foreign guardianship was still sought but without the neutrality: important political figures were offered the government of the city to reflect the current winners in the local Guelf versus Ghibelline conflicts. Thus was Emperor Henry IV made ruler of the city in 1310:

An agreement was made: Genoa agreed to cede itself to Henry for twenty years, and he agreed to end divisions in the city and restore peace. … the Genoese renounced their own autonomy, not in the face of external force, for Henry's power and resources were limited, but because self-rule no longer worked. (Epstein 1996, 184).

He was followed by Robert of Naples (with Pope John XXII) in 1318 (p. 196); the Archbishop of Milan in 1354 (p. 220); Charles VI of France in 1396 (p. 245); Francesco Viscounti, Lord of Milan in 1421 (p. 265); and Charles VII of France in 1458 (p. 286). This serial love of ‘foreign rule' is even more curious when we consider how Charles VI was chosen:

After public deliberations and backstairs dealings in 1396, the Genoese decided to hand themselves over to the king of France, even though they knew that from time to time he suffered from mental illness. (p. 245).

He may not appear to be a good candidate but he was not without competition; in the debate the proposal was that:

the doge and a group of citizens negotiate with the king of France, or England, or the emperor, or another prince, or the Tuscan league, to find a leader. (p. 247)

But this is only unusual to our modern sensibilities. What we have here is a traditional separation of commercial and guardian work in the specific circumstances of church and state rivalry in medieval Italy. First, the podestà system was invented to counter internal conflict and second, powerful foreign rulers were brought in to provide protection also from outside enemies. However neither the podestà as a ‘hired hand' nor powerful rulers were ever allowed to control the Bank of San Giorgio:

(Machiavelli writing in the 1480s) astutely observed that even as the Genoese handed over their state several times to various foreign rulers, San Giorgio remained firmly in local hands and carefully secured its privileged status from whatever ruler controlled the state. (Epstein 1996, 220)

And so the Genoese making their money under Hapsburg guardianship in the sixteenth century is not so surprising after all. They will have felt quite at home in the commercial world of Seville while guardian matters were attended to in Madrid (Pike 1966).

But this governance arrangement is not completely unheard of in our contemporary modern world. There is one very important ‘modern political economy anomaly': Hong Kong, which has been arguably the most economically successful city of the last half-century. And through this period it has never had sovereignty: it was ruled until 1997 as a British colony and since then as part of Communist China but with a guaranteed separate economic system. The separation of commercial and guardian processes seems not to be such a bad thing; as the modern interlude comes to a close the new ‘global governance' might become awash with such ‘modern anomalies'.

Using Geography (Beyond History's Reach)

Using geography in history has two different meanings. Most commonly it is a reference to physical environmental conditions. For instance, when the main channel of the Euphrates changed course about 2350BC it had a massive effect of relations between ancient Mesopotamian cities (Nissen 1988, 131). However geography references are sometimes to human geography in the sense of socio-spatial patterns. In this case the overwhelming example is the use of central place theory to explain settlement patterns. This theory became a major source of research work in geography from the 1950s to 1970s and then largely disappeared as a research topic in the 1980s. However it continued to be used by historians without reference to new ideas about inter-urban relations in globalization research. As portrayed above, central place theory defines a process that is essentially local; for non-local inter-urban relations, city network formation (central flow theory) is the process to study. And yet we find, post-Pirenne, in the very rich urban historical tradition of Belgium, central place theory being deployed to understand Flemish cities such as Ghent and Bruges whose importance was constituted by their location in Europe-wide urban networks (Stabel 1997; see Taylor et al. (2008) for historian's debates on this). In this section ‘using geography' means deploying central flow theory to understand inter-city relations.

By ‘beyond history' I mean debates about periods when there is little or no documentary evidence. The mixture of archaeological and historical evidence that is available is massively incomplete, fragmentary and, most definitely non-random. Although more evidence is potentially available and is being obtained, with interpretation modified accordingly, there will never be enough data coverage to overcome the systemic dearth of information on the societies being studied. Thus in these cases there is a real sense that we ‘can never know': all interpretation must perforce be modest interventions into debates. Thus all studies are only as good as the evidence available and the theory being deployed to stretch it into a plausible social story.

The two case studies I discuss here are definitely from the evidenced-challenged end of historical scholarship. First, I use the debates about urban development in the period between the fall of the Roman Empire in the west and medieval commercial revolution of the eleventh century. This is the debate sparked by Pirenne (1962). Second, I engage with the question of urban origins. This is about whether Mesopotamia has the honour of having the first cities, as conventionally asserted, or whether there were earlier city networks. The two things these debates have in common is the use of central place concepts to the almost total neglect of city network formation. Hence this section is about what happens to the interpretations in these debates when central flow theory is introduced.

The Debate over Cities after the Fall of Rome in the West (the Pirenne Thesis)

Pirenne (1962, 3) views the Roman Empire as ‘an essentially Mediterranean commonwealth'; it was constituted as a federation of ‘city-states' combining existing (‘allied') cities in the East with new planted cities (founding colonies and municipia ) in the West (Grimal 1983, 109). Thus in terms of its urban constituents the empire existed as two parts with the greatest cities in the East (Pirenne 1962, 4). However, Pirenne's thesis is that the deposing of the last Emperor in the West was not decisive because the Mediterranean commonwealth continued: ‘a clear-cut, direct continuation of the economy of the Roman Empire' (p. 12). The break occurred not with the Germanic invasion from the north but rather three centuries later with the Islamic invasion from the south. The result is a shift to northern Europe where Charlemagne's empire is categorised as ‘a closed state; a state without foreign markets, living in condition of almost complete isolation' (p. 29). It is only with the revival of trade with the east via Venice and the Flemish coast in the eleventh century that cities revive as commercial centres: ‘Just as the trade of the west disappeared with the shutting off of its foreign markets, just so it was renewed when these markets were reopened' (p. 82). In this way Europe ‘regained that essential characteristic of being a region of cities' (p. 103).

This is the barest bones of the Pirenne thesis that has been highly controversial since its definitive statement in 1925 (Hivighurst 1976). It is a thesis broadly adopted by both Jacobs (1969) and Braudel (1994, 92-3); according to McCormick (2001, 2) ‘it predominates today' but without the assumption that it was the Arab conquests that are a primary cause. The core of the majority view is that the Carolingian economy was the nadir of post-Roman economic changes in Europe. But, asks McCormack (2001), what does this downturn really represent? Was there such an isolated Carolingian state?

McCormack's (2001) impressive answer to this question is both honest about the evidence and path-breaking in its dealing with it. He argues that ‘the volume of scholarly discussion' on the Pirenne thesis ‘far outweighs the evidence to which it is devoted' (p. 3). But he counters this with a prosopographical research exercise that puts together what evidence is available in one large data collection exercise in order to create a geography of communications for this period. Although there is little information on commerce there is a lot of evidence for the movements of travellers, information and various artefacts. He creates data sets based upon nearly 700 journeys between 700 and 900, over 400 letters, and over 800 movements culled from documentary sources from 600 to 970, plus other evidence such coin hoards. The result is an immense amount of information on this data-challenged research area: the whole book describes an historical space of flows not equalled for any other period to my knowledge. And it is shown conclusively that Pirenne's rhetoric on the isolation of the Carolingian world needs reconsideration and, at least, further careful specification. On the basis of this new evidence McCormack is mindful to agree with Pirenne's critics in this reopening of the debate.

From a geographical perspective there is a problem with this interpretation of the new data. I take issue with McCormack's work in two respects. First, the focus is on communications. I am interested in the key economic process described above that is about city networks linking city clusters. Both are needed to define a world economy. In other words showing reams of communication in the form of people, knowledge and commodities is simply not enough. These flows have to be tied together in productive nodes of economic expansion – cities. Cities as a theoretical concept is absent from McCormack's arguments; they just appear as beginnings, stops or ends in the flows with no discussion of how the flows might have created an economy (p. 575). And this is, of course, because there were no Carolingian cities to carry an economy. Furthermore, this is precisely what Pirenne (1972a) understood though his explicit focus on cities. It is the absence of cities that is the key not the absence of communication even where the latter are commerce. Carolingian society had ecclesiastical and fortress urban settlements (for guardian functions) but no commercial cities (p. 56), just towns with weekly markets (pp.65-6). Pirenne contrasts Kiev with its strong commercial links to Constantinople and Baghdad with Aachen, which was merely Charlemagne's favourite residence – ‘it was to become a real city only four centuries later' (p. 62). Later scholarship suggests the size of Aachen never got beyond a couple of thousand (Barraclough 1979, 109), whereas Constantinople is estimated to have had some 250,000 people in 800 growing to 300,000 in 1000 when Kiev reached a population of 45,000 (Chandler 1987, 14-5). These are different economic worlds and should be understood as such.

Second, McCormack mixes commercial and guardian work flows together in a data set where the former are in a very small minority, for instance only 2% of ‘eastern travellers' were merchants (p. 212). From the perspective adopted here his prosopographic approach is theoretically naïve for understanding his chosen subject: ‘the origins of the European economy'. We need to acknowledge that there are two processes that lead from the demise of the Western Roman Empire to the creation of Europe: a guardian story that embodies continuities from Rome and a commercial story that, as Pirenne argues, does not: Venice is the beginning (Taylor 2006b). They are entwined but separate and should be understood as such.

The Debate over Urban Origins (the Jacobs Thesis)

One of the most established assumptions in ancient history is that the first cities were created in Mesopotamia five thousand years ago; large agricultural surpluses from the exceptional fertility of the alluvial lands created the ‘urban revolution' (Childe 1950). But given the evidential basis in very early history, we can never be certain. This is a world of scholarship consisting of only of ‘reasonable conjecture' and ‘limited finality' (Jacobsen 1970). And yet the urban revolution remains an unmovable watershed in conventional thinking on human evolution. It looks very much like ‘the ideological wall … that has been placed at 5,000 years ago' that Rudgley (1998, 209) claims.

Jacobs (1969) has nimbly scaled this wall in her controversial ‘cities first, rural development later' thesis. Since cities are the basis of economic change, it follows that agriculture is an urban invention. This means that it is hunter/gatherer societies who will have created the first cities as a prelude to the agricultural revolution. In other words cities should appear at the beginning of the Neolithic period when agriculture was established. And this is what Jacobs finds at Catal Huyuk in Anatolia. With a concentration of population of perhaps 10,000, some 9,000 years ago, she interprets this settlement as an early city trading with other unknown cities. Being near an important source of obsidian, initial trade would have included exchanging this mineral for food to feed the growing population. But at some point the innovative process of import replacement generated new work in local food production (the Neolithic agricultural revolution) thus freeing exports to buy additional new goods (which is Jacobs' definition of economic expansion). This is the primary example of cities moulding their environs to their own needs, a process that has culminated in contemporary globalization.

Although garnering support in geography – Soja (2000) treats Catal Huyuk and its ilk as the first urban revolution, with Mesopotamia relegated to second revolution, and industrial modernity representing the third – Jacobs' thesis is dismissed out of hand by archaeology scholars (Balter 1998).

One of the interesting things about these archaeologists is that they themselves draw heavily on geographical tools in their interpretations. For instance, Renfrew (1975, 3) develops his influential theory of the emergence of civilization based on the use of ‘quantitative methods inspired by geography'. But when he argues that ‘the rise or origin of civilization can profitably be considered in terms of the genesis of … spatial organization' (p. 12) he is being lured into a dominant spaces-of-places geography. Thus we are told that ‘a permanently functioning central place system is a feature of every civilization' (p. 12) in which ‘the hierarchy of central places … carries with it a hierarchy of central persons' (p. 24). This is an endogenous theory wherein exogenous processes ‘outside the civilization territory' are not necessary (pp. 26-31). Since it is not part of a central place system, Renfrew is able to dismiss Catal Huyuk as merely an ‘early farming village' albeit of ‘almost urban size' (p. 7). Renfrew's ideas are broadly applied by Nissen (1988) in his history of ‘ancient Near East' from 9000 to 2000BC wherein both Catal Huyuk and early Mesopotamia feature in his story. I will critique his space-of-places history from a more recent spaces-of-flows perspective.

Nissen's story is as follows: from 9000 to 6000 BC the Near East consisted of ‘separated settlements' with few connections (p. 32-3) wherein Catal Huyuk, and Jericho (another inconvenient early large settlement) are treated as exceptions because they are not part of a settlement system (pp. 36-8). Having no ‘real centrality' (p. 38), Catal Huruk and Jericho are deemed not to be urban settlements but rather they are interpreted as ‘extremes of a … spectrum' of settlement sizes. And because ‘further development seldom proceeds from the extremes', such large settlements have only minimal relevance. For instance, Jericho is considered to be just a ‘river oasis' (p. 37). Such dismissal does not address why such large settlements were created and what effect such concentrations of people might generate. Why should an oasis on river be so large? (Catal Huyuk was also on a river (Balter 1998).) Could non-local trade be a factor? Such a thought does not occur here because there is no sense of a space of flows. Urban development must await the making of a central place hierarchy in Susiana several thousand years later. And this phase of the story is also instructive in showing the narrow spatial thinking. We are told that a ‘c ompletely developed settlement systems can, for the first time, be identified in the period we call the Late Susiana period' (p. 52) and the locale is described thus:

The plain lies at the junction of two natural routes through an otherwise impassable mountain area … The early settlements certainly took advantage of it, because the relationships we can prove to have existed at that time between the civilizations west and east of the Zagros must have made use of this route. It is then easy to suppose that Tepe Sohz, by far the largest site, had a part to play on the connecting route, about which, for the time being, we cannot be more specific. However, the location of this site near the spot where the Marun River flows into the plain shows that its primary importance had a different basis. Because of its location, the site was in a dominant position for an irrigation system that drew its water from the Marum, a little to the north of Tepe Sohz. (pp. 53-4)

He concludes that ‘there is no doubt that we are dealing here with a simple settlement system' (p. 54) on the grounds that settlements are in a line (a possible canal course) and consist of three size-groups (a possible hierarchy).

Note that Nissen has a choice of what to emphasize: the non-local space of flows (an important trade route) or local space of places (a three tier settlement system). Of course, the fact of irrigation and hinterland in no way disqualify the process of city network first, local development second. Nissen proceeds to a four-tier settlement pattern in the Uruk period so that Mesopotamian urbanization is itself largely interpreted in central place terms (see also Gregory (1975).

The key point in all this is that the Jacobs thesis should not be so easily dismissed. Nissen's (1995, 36) ‘certainty' of ‘even development' (meaning no urbanization) before 6000BC does not square with his explicitly recognition of critical problems relating to evidence – he devotes all of his first chapter to this topic and refers to it throughout the book. This makes his archaeological orthodoxy look like an ideological response to evidence. Certainly it is clear that Catal Hayuk and Jericho are not alone as large Neolithic settlements; important examples have been found in Japan (Sannai Maruyama (Hudson 1996)), in North America (Cahokia (Chappell 2002)) as well as elsewhere in the Near East (Jawa (Helms 1981)) (other examples are to be found in Soja (2000) and Hansen (2000, 2002)). From a space-of-flows position these are all evidence of cities first (central flow theory) and towns second (central place theory) with the latter urban settlements as minor service centres to agricultural villages further away from the city but still in its hinterland (Taylor 2008). Of course, none of this can be known in any definitive sense but it is a ‘reasonable conjecture' based on geographical theory and growing evidence.

Conclusion (in the Spirit of Braudel?)

The substantive parts of this paper have been quite conventional with human geography (representing social science) using historical scholarship (detailed evidence (specifics) is used to support generic arguments) and history (or archaeology) using human geography for its theoretical insights (generics). Braudel is, of course, a major exception to this commonplace interaction between historians and social scientists. He uses his historical concepts to straddle times and spaces: he thinks generically. It is appropriate, therefore, to end with a further linkage between Braudel's history and current geographical work. In particular, I consider Braudel's (1979) most contrarian ideas on capitalism (Wallerstein 1991) to Jacobs' moral syndrome analysis of work.

In my use of Jacobs' moral syndromes I have applied them to cities and states in a way that equates the guardian syndrome with politics and the commercial syndrome with economics. However Jacobs' ideas are somewhat more subtle than this: like Braudel she has no truck with the modern disciplinary division between political and economic processes. Although superficially the guardian syndrome appears to be about politics and the commercial syndrome about economics this is not the case. Certainly commercial processes are about part of economics (markets) but the guardian syndrome also involves economic players in the world of monopolies, as well as the state. Big city financiers in globalization as ‘masters of the universe' were, until very recently, running a truly opaque ‘global financial market' in largely guardian ways. Of course, this is very much like Braudel's (1979) definition of capitalism based upon both political and economic monopolies in the top tier in his tripartite structure of society (Wallerstein 1991). Braudel separates his capitalism from the market of his second tier, which, for Jacobs, reproduces itself via her commercial syndrome of ethics.

It seems to me that here is an opening for further thoughts on history and geography via the ideas of Braudel and Jacobs …..


Algaze, G (2005) The Uruk World System: The Dynamics of Expansion of Early Mesopotamian Civilization. Chicago: University of Chicago Press

Arrighi, G (1994) The Long Twentieth Century: Money, Power, and the Origin of our Times. London: Verso

Balter, M. (1998) ‘ The first cities: Why Settle Down? The Mystery of Communities' Science 282, 1442

Barraclough, G. (1979) The Times Atlas of World History. London: Times Books

Braudel, F. (1972a) ‘History and the social sciences: the longue durée ' in P Burke (ed) Economy and Society in Early Modern Europe. London: Routledge and Kegan Paul

Braudel, F (1972b) The Mediterranean and the Mediterranean World in the Age of Phillip II. Volume 1 (London: Collins)

Braudel, F. (1973) Capitalism and Material Life 1400-1800. London: Weidenfeld and Nicolson

Braudel, F (1982) The Wheels of Commerce (London: Collins)

Braudel, F (1984) The Perspective of the World (London: Collins)

Castells, M (1996) The Rise of Network Society. Oxford: Blackwell

Chandler, T. (1987) Four Thousand Years of Urban Growth: an Historical Census. Lewiston: Edwin Mellen

Chappell, S A (2002) Cahokia: Mirror of the Cosmos. Chicago: Chicago University Press

Childe, V D (1950) “The urban revolution” Town Planning Review 21, 3-17

Ellis, J (2001) ‘The “Black Indies”: economic development of Newcastle, c.1700-1840' in R Colls and B Lancaster (eds) Newcastle upon Tyne: a Modern History. Chichester: Phillimore

Epstein, S A (1996) Genoa and the Genoese. Chapel Hill, NC: University of North Carolina Press

Glaeser, E L, Kallal, H D, Scheinkman, J A and Shleifer, A (1992) ‘Growth in cities', Journal of Political Economy 100, 1126-52

Gregory, D (1982) Regional Transformation and Industrial Revolution. A Geography of the Yorkshire Woollen Industry. London: Macmillan

Gregory, G A (1975) ‘Locational Analysis and the investigation of the Uruk local exchange systems', in Sabloff, J A and Lamber-Karlovsky, C C (eds) Ancient Civilization and Trade Albuquerque: University of New Mexico Press

Grimal, P (1983) Roman Cities. Madison: University of Wisconsin Press

Hansen, M H (ed.) (2000) A Comparative Study of Thirty City-State Cultures. Copenhagen: Det Kongelige Danske Videnskabernes Selskab

Hansen, M H (ed.) (2002) A Comparative Study of Six City-State Cultures. Copenhagen: Det Kongelige Danske Videnskabernes Selskab

Helms, S W (1981) Jawa: Lost City of the Black Desert. London: Methuen

Hicks, J (1969) A Theory of Economic History Oxford: Clarendon

Hudson, M. (1996) ‘Sannai Maruyama: A New View of Prehistoric Japan' Asia-Pacific Magazine 2, 47-48.

Jacobs, J (1969) The Economy of Cities. New York: Vintage

Jacobs, J (1984) Cities and the Wealth of Nations. New York: Vintage

Jacobs, J. (1992) Systems of Survival. New York: Vintage

Jacobsen, T (1970) Toward the Image of Tammuz Cambridge, MA: Harvard UP

McCormack, M (2001) Origins of the European Economy: Communications and Commerce AA 300-900. Cambridge: Cambridge University Press

Middlebrook, S (1950) Newcastle upon Tyne: Its Growth and Achievement. Newcastle upon Tyne: Newcastle Journal

Nissen, H J (1988) The Early History of the Ancient Near East: 9000-2000BC Chicago: University of Chicago Press

Pike, R (1966) Enterprise and Adventure: the Genoese in Seville and the Opening of the New World Ithaca, NY: Cornell University Press

Pirenne, H (1925) Medieval Cities: Their Origins and the Revival of Trade Princeton: Princeton University Press

Renfrew, C (1975) ‘Trade as action at a distance: questions of integration and communication', in Sabloff, J A and Lamber-Karlovsky, C C (eds) Ancient Civilization and Trade Albuquerque: University of New Mexico Press

Rudgley, R (1998) Lost Civilizations of the Stone Age London: Century

Scott, J C (1998) Seeing Like a State. New Haven, CN: Yale University Press

Smith, M L (2007) ‘Territories, corridors and networks: a biological model for the premodern state' Complexity 12,(4), 28-35

Soja, E. (2000) Postmetropolis: Critical Studies of Cities and Regions. Oxford: Blackwell

Spufford, P (2002) Power and Profit: the Merchant in Medieval Europe. London: Thames and Hudson

Stabel, P (1997) Dwarfs among Giants: The Flemish Urban Network in the Late Middle Ages. Leuven: Garant.

Stein, G J (1999) Rethinking World-Systems: Diasporas, Colonies, and Interaction in Uruk Mesopotamia. Tucson: University of Arizona Press

Taylor, P J (1994) ‘Ten years that shook the world? The United Provinces as first hegemonic state' Sociological Perspectives 37, 25-46

Taylor P J (1999) ‘Places, spaces and Macy's: place-space tensions in the political geography of modernities' Progress in Human Geography 23, 7-26

Taylor P J (2001) ‘Specification of the world city network' Geographical Analysis 33, 181-94

Taylor, P J (2004) World City Network: a Global Urban Analysis. London: Routledge

Taylor, P J (2006a) ‘Development as a “monstrous hybrid”: an essay on the primacy of cities in the expansion of economic life' Environment and Planning A 43,

Taylor. P J (2006b) ‘Prologue: a lineage for contemporary inter-city studies' in P J Taylor, B Derudder, P Saey and F Witlox (eds) Cities in Globalization. London: Routledge

Taylor, P J (2007) ‘Problematizing city/state relations: Towards a geohistorical understanding of contemporary globalization” Transactions, Institute of British Geographers NS32, 133-50

Taylor, P J (2008) ‘Cities in Transitions and Transformations: Exploring a Jacobsean approach to Macro-Social Change' GaWC Research Bulletin No. 237

Taylor, P J, Hoyler, M, and Verbruggen, R (2008) ‘External Urban Relational Process: Introducing Central Flow Theory to Complement Central Place Theory' GaWC Research Bulletin No. 239

Unwin, G (1963) The Gilds and Companies of London. London: Frank Cass (4th ed. ; 1st in 1908)

Wallerstein, I. (1991) Unthinking Social Science Cambridge: polity Press

Waley, D (1969) The Italian City-Republics. London: Weidenfeld and Nicolson

Wilson, R W (1971) Gentleman Merchants: The Merchant Community in Leeds, 1700-1830 Manchester: Manchester University Press



Table 1: Clusters of precepts within Jacobs' moral syndromes


Table 2: Braudel's local and non-local commercial work in different languages

*These are terms used in Cotrugli's book on commerce published in 1456/8

Source: derived from Braudel (1982), p. 377



* Peter J. Taylor, Department of Geography, Loughborough University, UK. Email:


Edited and posted on the web on 3rd October 2008

Note: This Research Bulletin has been published in R.E. Lee (ed) (2012) The ‘Longue Durée' and World-Systems Analysis, Albany, NY, State University of New York Press, pp. 35-64.