HOW CLIMATE CHANGE IS DRIVING TECHNOLOGICAL PROGRESS AND SOCIAL COMPLEXITY�

HOW CLIMATE CHANGE IS DRIVING TECHNOLOGICAL PROGRESS AND SOCIAL COMPLEXITY

[Note: thanks to CCNet]

Quaternary International, Article in Press, Corrected Proof
HYPERLINK "http://tinyurl.com/h268j" \t "linkWin" http://tinyurl.com/h268j

Cultural responses to aridity in the Middle Holocene and increased social complexity

Nick Brooks,

Tyndall Centre for Climate Change Research, Saharan Studies Programme and School of Environmental Sciences, University of East Anglia, Norwich NR4 7TJ, UK

Abstract
The first complex, highly organised, state-level societies emerged in the Afro-Asiatic monsoon belt and northern South America during the 6th and early 5th millennia BP. This was a period of profound climatic and environmental change in these regions and globally, characterised by a weakening of the global monsoon system and widespread aridi which complex societies emerged during the Middle Holocene: the central Sahara (focusing on the Libyan Fezzan), Egypt, Mesopotamia, South Asia (Indus-Sarasvati region), northern China and coastal Peru. Links between environmental and socio-cultural change are explored in the context of archaeological and palaeoenvironmental data and a theoretical framework of increasing social complexity as a response to enhanced aridity, driven largely by population agglomeration in environmental refugia characterised by the presence of surface water. There is direct evidence of adaptation to increased aridity in the archaeological literature relating to the Sahara and Egypt. In the other regions examined, the data are consistent with the notion that increased social complexity was largely driven by environmental deterioration, although further local-scale archaeological and palaeoenvironmental data are required to clarify the processes involved.

1. Introduction

Environmental catastrophes, particularly severe, rapid or abrupt changes in climate, are often associated in the academic literature with the collapse of civilisations. The fall of the Akkadian Empire and the end of the Egyptian Old Kingdom 4.2 thousand years before present (kyr BP) have both been attributed to climatic change resulting in a period of pronounced regional desiccation (Hassan, 1997; Weiss, 1997; Cullen et al., 2000). The collapse of the Tiwanaku state in the Titicata Basin around 1 kyr BP has been interpreted as a result of drought (Ortloff and Kolata, 1993), as has that of the Mayan civilisation (Haug et al., 2003).

All dates are given in calendar years before present (defined as 1950) to the nearest 100 years; the majority of dates are estimates based on radiocarbon dating rather than precise dates referring to a particular year before or after the beginning of the Common or Christian Era (i.e. BC/BCE or AD). Where dates are given in the original source material in terms of uncalibrated radiocarbon years before present, these have been converted to calendar years before present using the conversion tables from Hassan (2002, pp. 8-9). These calibrated dates are then assumed to be equivalent to calendar years BP. Where such a conversion has been applied, the date as quoted in the original source is given in parentheses. Conversely, where uncalibrated dates, or years BC, are quoted directly from the original sources, they equivalent date in calendar years before present is given in parentheses, e.g. "...3300 yr BC [5.3 kyr BP]."

A widely held view is that the process of increasing "social complexity" associated with the development of agriculture, large settled communities and the earliest states, was made possible by the relatively benign climate of the Holocene (e.g. Fagan, 2004; Burroughs, 2005). In this model of human social evolution, severe or abrupt changes in climate are associated with interruptions to or reversals in the progressive development of human societies. However, there is widespread and increasingly abundant evidence that pronounced increases in social complexity in the Middle Holocene coincided with climatic and environmental deterioration, and in particular with increased aridity. This paper is concerned with the climatic and environmental changes, and the contemporaneous social and cultural developments, associated with the emergence of the earliest "civilisations", defined here as complex, highly organised societies exhibiting a high degree of urbanisation, social stratification, specialisation of production, and centralisation of power. From here on the term "complex societies" is used as a shorthand for such entities. It is not the purpose of this paper to address definitional issues relating to states and chiefdoms and the differences between them; here we are concerned with the emergence of the first societies characterised by large urban centres (i.e. the first "cities") and exerting influence or control over large geographical areas, most probably through the exercise of some centralised power and associated formal institutions of governance. For detailed treatments of issues of complexity and "social evolution" the reader is referred to Chapman (2003) and Yoffee (2005).

The first such complex societies emerged in the region currently occupied by the desert belt stretching from North Africa in the west to South Asia and China in the east, and in northern South America, during the late 6th and 5th millennia BP, a period characterised by a shift towards aridity in these regions. This paper examines the emergence of such societies in Egypt, Mesopotamia, South Asia (the Indus-Sarasvati region), northern China, and coastal Peru, and argues that their development was largely a response to increased aridity driven by global-scale changes in climate commencing around 6 kyr BP. However, the discussion of linked environmental and social change begins with the Sahara, where the evidence for large-magnitude changes in the physical environment and the associated impacts on human societies is relatively clear in comparison to other regions.

At this point it is worth stating some of the assumptions underpinning the following analysis. First, it should be stressed that the term "complex societies" is used simply for convenience and brevity; the rigid classification of societies into successive evolutionary types representing "stages" of social development is rejected here (see also Chapman, 2003; Yoffee, 2005). Secondly, the aim is not to argue in favour of environmental determinism or reductionism. Rather than seeing social and cultural change simply as being determined by changes or fluctuations in the physical environment, the view taken here is essentially a materialist, coevolutionist, one in which the physical environment is seen as setting the context within which social change occurs, providing both opportunities for, and constraints on social, cultural, economic and technological innovation. Social and cultural contexts inform responses to environmental change, and are themselves influenced by the physical environment in which a society exists (Chapman, 2003).

One motivation for an examination of the role of the environment in the emergence of complex societies is the rather unsatisfactory nature of earlier models of increasing social complexity. These have invoked a wide range of causal factors in the development of what are popularly thought of as the first civilisations, including cultural diffusion from an original centre of innovation-usually Mesopotamia (e.g. Wheeler, 1968), increased agricultural productivity resulting in food surpluses capable of supporting non-producing members of society (e.g. Possehl, 2002), the need for central organisation of irrigation systems (Wittfogel, 1957), fortuitous technological innovation (e.g. Ratnagar, 2001), and the growth of trade networks (e.g. Possehl, 2002). Warfare has also been proposed as a driving factor in the emergence of state-level societies in other contexts (Spencer, 2003).

None of these factors provide a very satisfactory explanation for the emergence of the first complex societies when viewed in isolation, and it has been argued that many of the phenomena proposed as drivers of social complexity are more reasonably viewed as its products, albeit products that feed back into the process of increasing complexity (e.g. Fagan, 1999). More complex, subtle and multidimensional models, involving combinations of the above, have also been proposed (see Fagan, 1999; Chapman, 2003; Yoffee, 2005), but these do not satisfactorily explain the contemporaneous (within the space of about a millennium) emergence of complex societies in widely separated parts of the world. The problem of explaining such closely spaced developments becomes more acute in the light of recent work that places the emergence of large urban centres in northern South America at the beginning of the 5th millennium BP (Solis et al., 2001; Mann, 2005).

Further impetus for a re-examination of the role of the physical environment in the emergence of the first complex societies comes from developments in our understanding of Holocene climatic and environmental change. It is increasingly apparent that the Middle Holocene was a time of profound change in environmental as well as cultural terms (Steig, 1999). In the early Holocene, the currently arid zone in the northern hemisphere extra-tropics was considerably wetter than today, and regions that are today covered by desert were well vegetated, supporting significant faunal and human populations. A wealth of palaeo-environmental evidence, reviewed in more detail below, indicates that these regions became progressively more arid after the end of the 7th millennium BP, reaching a state similar to that existing today during the 5th millennium BP. Increasing social complexity, culminating in the emergence of highly organised state-level societies, parallels this process of environmental desiccation. Given the overwhelming importance of the physical environment in determining the suite of available livelihood options for prehistoric and proto-historic societies, it would be remarkable if such changes did not influence societal development.

This paper argues that global climate change in the Middle Holocene stimulated societies to become more complex and organised as they responded to its local and regional manifestations. While this process is described in terms of broadly common responses to environmental change, the paper also addresses differentiated responses and diversity in outcomes resulting from the development of coupled social and environmental systems. While associations between social and environmental change in the Middle Holocene have been proposed in the past, to the author's knowledge there has been no attempt to synthesise regional archaeological and palaeoenvironmental data with indicators of past climate change at the global scale. The increase in the availability of archaeological and palaeoclimatic data in recent years also puts such a synthesis on a firmer footing, although large gaps in our knowledge still remain.

The paper concludes with a brief discussion of the implications of the model of coupled socio-environmental change for our understanding of socio-cultural trajectories, and for current concerns about climate change and associated human adaptation.

[...]

9. Discussion

It is now well established that the earliest complex, highly organised, state-level societies emerged at a time of increasing aridity throughout the global monsoon belt. This trend towards desiccation commenced around 8 kyr BP as a result of declining solar insolation associated with changes in the Earth's orbital parameters, but accelerated around 6 kyr BP after a widespread centennial-scale arid episode that may have been the result of transient cooling in the North Atlantic. The following millennium was a time of profound cultural change that saw the development of the world's first states in Mesopotamia and Egypt, and laid the foundations for similar developments in South Asia, northern China and northern South America. The available data suggest further abrupt changes in climate in the late 6th millennium BP, when regional records indicate environmental and cultural discontinuities. The period around 5.2 kyr BP seems particularly significant. At this time a unified Egyptian state emerged, and the Uruk culture of Mesopotamia collapsed and gave way to the transitional Jemdet Nasr period, characterised by fragmentation and regionalism. In South Asia the beginnings of the Early Harappan phase have been placed at 5.2 kyr BP (Possehl, 2002) and the beginnings of urbanisation are evident at the site of Harappa (McIntosh, 2002). In northern China there is evidence for an abrupt drop in temperature and accelerated aridity coinciding with the Yangshao-Longshan transition in the final centuries of the late 6th millennium BP (Liu, 1996). In South America this period was characterised by profound changes in the ENSO cycle and an increase in coastal upwelling (Reitz and Sandweiss, 2001; Andrus et al., 2004) prior to the emergence of large urban centres exhibiting monumental architecture. Changes in the behaviour of monsoon systems appear to be one aspect of a wider reorganisation of the global climate.

The association between environmental desiccation and increasing social complexity is particularly striking in the central Sahara and Egypt. Trajectories of environmental change are quite clear in the Sahara, and are based on a variety of proxy data from a large number of locations. Geoarchaeological studies in the Libyan Fezzan tell a coherent story of demographic and cultural change driven predominantly by changes in water availability, although different adaptive responses are evident. In Egypt there is abundant evidence of the abandonment of the deserts flanking the Nile Valley, increased population densities in the Nile Valley itself, and attempts to transform lifestyles rooted in mobile cattle pastoralism to ones suited to a more sedentary existence in a confined geographical area. The model of competing "proto-state entities" eventually coalescing in the late 6th millennium BP is compatible with the expected impacts of the process of desiccation that began around 6 kyr BP and culminated around or soon after 5 kyr BP. This desiccation, and the resulting scarcity of water and pasture, eventually would have led to the collapse of livelihoods based on mobile pastoralism in the Western and Eastern deserts of Egypt. Those pastoral populations that did not perish would have migrated to refugia such as the Nile Valley, either with their animals in search of water and pasture, or as what we might term "environmental refugees" as a last resort. Increases in population density, competition over resources and trade routes, and the emergence of social strata associated with the presence of advantaged and marginalised groups (for example, established groups and migrants from the surrounding deserts) are likely to have provided the conditions for the development of political power and competition between political entities.

The links between environmental and social change are less well defined in the other regions discussed in this paper, and are to a large extent inferred by comparing distinct archaeological and regional palaeoenvironmental records. In the case of Mesopotamia, palaeoenvironmental data in the southern alluvial lowlands are scarce, and interpretation of past climatic and regional environmental change is further complicated by the role of changes in river channels in local environmental change. Nonetheless, there is abundant evidence of regional environmental desiccation and, while the trajectories of social change in Mesopotamia and Egypt are very different, in both cases increasing social complexity is associated with increases in local population densities, the congregation of populations along rivers, and evidence of competition or conflict at the end of the 6th millennium BP. In Egypt this occurs within a context of political unification, whereas in Mesopotamia the context is one of fragmentation. In both cases environmental deterioration coincides with trends towards greater urbanisation and the development of ideologies of political power.

The transition to highly urbanised societies exhibiting the characteristics of states occurs later in the Indus-Sarasvati region and northern China. In both regions increasing social complexity is associated with a concentration of settlements along rivers at a time of increasing aridity, followed by the emergence of large urban centres. The transition to urbanisation was particularly rapid and discontinuous in the Indus-Sarasvati area, and strongly suggests a time of social upheaval during which many existing settlements were deliberately destroyed. Whatever the role of the environment in this transition, there were clearly other processes at work, although their precise nature (e.g. violent conflict and/or radical ideological change) is unknown.

Evidence for increased social complexity during periods of regional climatic and environmental change in the Middle Holocene is not restricted to the Afro-Asiatic desert belt. It appears that similar developments were occurring in coastal Peru in the early 5th millennium BP. Reconstructing environmental trajectories in this region is complicated by the high level of heterogeneity in the palaeoenvironmental record, associated with geographical and topographic factors. Nonetheless, responses to increased coastal aridity provide a convincing explanation for the cultural developments evident in the archaeological record, and are consistent with what is known about the nature of Middle Holocene environmental change in this region.

Explanations of increasing social complexity involving environmental change will remain controversial until (and probably even if) they can be supported further by archaeological data. In particular, they must ultimately be assessed through field studies that examine linked socio-cultural and environmental trajectories at the local scale. This will require numerous studies of individual settlement sites in order to illuminate the nature of changes in livelihood strategies, social organisation and ideological frameworks, supported by palaeoenvironmental proxy data from the vicinity of the settlements in question. While regional palaeoenvironmental data can provide us with a broad context within which to interpret cultural change, they tell us little or nothing about the nature of the interactions between people and the physical environment at the scales which would have been important to the communities from which complex societies arose.

While much more work needs to be done on the links between environmental deterioration and the emergence of complex, organised, state-level societies, the case for climate change induced aridification as the principal driving force behind the development of the first "civilisations" is strong. The evidence for dramatic and sometimes abrupt climate change in the Middle Holocene is overwhelming, and we would expect such changes to have profound impacts on societies that were directly dependent on their immediate physical environment for their livelihoods and food security. Changes in this physical environment would have resulted in changes in the available options for and constraints on livelihood strategies. We may view social change within such a context through the prism of adaptation to climatic and environmental change, mediated by other factors such as pre-existing socio-economic relations, livelihood strategies, and ideologies.

A recurring theme in this discussion has been population agglomeration in environmental refugia, necessitating the development of new social institutions and relations, and technological and institutional adaptations related to water extraction, food production and distribution. As suggested by Fagan (1999), these developments are not necessarily dependent on the prior emergence of an organising elite as proposed by Wittfogel (1957); it might be argued that the emergence of elite groups is a by-product of, rather than a prerequisite for, such adaptations, and results from the exploitation by certain groups of both geographic advantage and emerging social relations and institutions that provide opportunities for exerting control over other, relatively disadvantaged, social groups. People arriving in refugia would have been more likely to constitute such disadvantaged groups, providing a pool of "human capital" available for exploitation through organised labour or military activity by emerging elites. Ideological systems developed to legitimise the power of elites and support the emerging social hierarchy (Yoffee, 2005), while innovations in administration, production, and distribution would have be required in order to maintain the emerging social system. In the model presented here, increases in social complexity leading to the emergence of urbanisation and state-level societies are not driven by surpluses in food production. Instead, the capacity of agricultural systems to generate surpluses (from which high-density, specialised, urban societies can be supported) is harnessed out of necessity. While surpluses may have played a role in the first steps towards complexity, for example in the Ubaid and earlier periods in Mesopotamia and in the development of village agriculture in pre-Harappan times in South Asia, the rapid increases in complexity in the late 6th and early 5th millennia BP are interpreted as precipitated by hardship rather than abundance. This interpretation reflects the main conclusion of Marshall and Hildebrand (2002) regarding the domestication of plants and animals in Africa, which they argue was driven by a desire for greater predictability in the food supply rather than a deliberate attempt to increase yields and thus volumes of available foodstuffs. They point out that "The end results of agriculture-visible today as larger yields, higher carrying capacity, denser stands of crops, larger seeds and seed heads, or greater animal productivity-were not necessarily realized during the earliest phases of the domestication process" (Marshall and Hildebrand, 2002, p. 101). As with agriculture, so with urbanisation and the development of the institutions of state-the consequences of such developments would not have been anticipated by those populations that unwittingly precipitated them.

It should be stressed that this paper addresses the development of a specific set of societies at a particular period in time; its purpose is to illuminate the processes associated with the emergence of these particular societies and to draw some very broad general lessons rather than to identify any universal laws of human-environment interaction. The emergence of complex, urban, state-level societies is not an inevitable outcome of environmental deterioration; in the Middle Holocene, increased mobility proved in many respects to be a more sustainable response to aridity than technological innovation and urbanisation (di Lernia and Palombini, 2002). Neither should we assume that urbanisation and state formation only occur as a response to environmental change. There are numerous other instances of state formation, and it is not proposed that the model elaborated here applies to all of them. Conversely, there are instances of urbanisation during the Early Holocene in western Asia that did not culminate in the development of complex, state-level societies (Akkermans and Schwartz, 2003).

Furthermore, the diversity of the societies discussed above, and the different trajectories leading to them, should caution us against reductionism. While environmental desiccation provided the context for, and was arguably the principal driving force behind, their emergence, a multitude of other factors influenced their development. Each complex society discussed here emerged from a different cultural context. For example, the Garamantes ultimately emerged from the cattle herding cultures of the Sahara, and mobile cattle-based societies appear to have formed a component of proto-Dynastic Egyptian society; there was no significant tradition of agriculture in the prehistoric Sahara. This may be contrasted with the long-established and geographically widespread tradition of village agriculture in Mesopotamia, and livelihoods based on marine resources on the pre-urban Peruvian coast. The different physical environmental contexts would have provided each precursor society with different constrains and opportunities as they became more complex. Ideological systems would have developed with reference to the physical environment, but would also have been influenced by personalities and by contacts with other cultures. Trade would also have played a role in shaping livelihoods, in the development of economic and social structures, and in shaping world views. Each of the mature complex societies discussed here was distinct from the others, exhibiting its own unique "personality"; the unified Egyptian state in which the bulk of the population remained rural may be contrasted with the much more urbanised society of Mesopotamia, characterised during much of its existence by competing and cooperating city states. The high levels of social stratification in Egyptian and Mesopotamian societies, dominated by powerful ruling elites and individual leaders, suggest very different social structures to the apparently far less hierarchical Indus culture (McIntosh, 2002).

In terms of responses to "abrupt" climate change, it should be noted that the climatic changes apparently associated with the emergence of complex societies are qualitatively if not quantitatively similar to those blamed for societal collapse. In the case of Egypt, increased aridity is associated with the emergence of the unified Dynastic state in the late 6th millennium BP and with the widespread collapse of social systems during the First Intermediate Period some thousand years later at 4.2 kyr BP. In Mesopotamia, aridity appears to be associated with the emergence of the Uruk culture, its subsequent collapse and the transition to the Dynastic period, and also with the later collapse of the Akkadian Empire. Transient environmental shocks thus appear to have a variety of outcomes depending on the nature of the societies on which they impact, and the relationship between environmental and socio-cultural change is not straightforward, with a given environmental phenomenon having a single type of societal outcome. For example, it has been argued that the development of agriculture, which formed the basis for the increases in social complexity in the Middle Holocene discussed here, was associated with a succession of responses to both climatic amelioration and deterioration (e.g. Hole, 1991).

The model of linked environmental change presented here has a number of implications. Firstly, it provides another challenge to linear models of social evolution in which societies evolve through a number of predefined developmental states. There is a significant element of convergent evolution in the societies examined here, and they emerged as a result of increasing complexity which might be interpreted as taking them through "stages" of development. However, the trajectories that led to their emergence exhibit considerable diversity and discontinuity, associated to a large extent with their regional environmental contexts. In this model, increases in complexity are not a result of the inevitability of progress, but are driven by responses to changes in the physical environment on which people depend for their livelihoods. Social complexity may therefore be viewed largely as emerging from adaptation during the Middle Holocene, although it would clearly be nonsense to claim that complexity is only ever driven by adaptation to environmental change. Furthermore, not all adaptation in the Middle Holocene leads to increased social complexity as defined here; responses involving increased mobility can be at least as successful as those based on urbanisation. The social systems emerging from adaptation to aridity based on mobile pastoralism are in some respects more resilient to further changes than their complex and in many respects fragile urban counterparts, as apparent from the persistence of pastoralism long after urban civilisations have disappeared (di Lernia and Palombini, 2002).

The resilience of responses based on mobility holds important lessons for the emerging fields of adaptation research and policy, which are developing in response to concerns about current and future anthropogenic climate change. In today's globalising world, traditional livelihoods are under pressure from economic liberalisation, monetisation of local economies and development programmes based largely on western models. Factors such as drought and conflict can also threaten traditional systems. Given the uncertain nature of future changes in climate in semi-arid regions such as the Sahel (Brooks, 2004), the building of flexible and resilient livelihoods is a priority. The creation of "enabling environments" for adaptation and the building of "adaptive capacity" have been stressed as priorities in order to facilitate adaptation (Brooks and Adger, 2005); support for traditional livelihoods that have evolved to cope with climatic variability is one way of achieving this. Any measures to enhance resilience and promote adaptation must take account of local contexts and recognise that imported developmental models may be inappropriate; indigenous livelihood strategies have often emerged from centuries or even millennia of linked environmental and social change.

The archaeological record can offer more general lessons about adaptation to climatic and environmental change. Today, adaptation is seen as a means of minimising or neutralising the adverse impacts of anthropogenic climate change, of preserving existing social systems, and of supporting predefined developmental goals. In this sense, adaptation is seen as something that is manageable and, by implication, predictable. This is evident in the language of adaptation research and policy, which talks of "adaptation goals" and the design of "adaptation strategies, policies and measures" (e.g. UNDP, 2005). While such approaches are necessary to confront the consequences of climatic changes that are the inevitable results of past and unavoidable near-future greenhouse gas emissions, they may be unrealistically optimistic. The archaeological record emphasises that adaptation has in the past been associated with great social upheavals that could not have been foreseen by those who were undertaking the adaptation. The consequences of adaptation were unplanned and unpredictable, arising from the ad hoc responses of a variety of actors to environmental change. Planned adaptation strategies designed to help societies cope with anticipated future changes in climate represent a radical shift in the way human beings respond to and interact with the physical environment. Furthermore, the view of adaptation as a means of neutralising the impacts of environmental change is a naïve one, particularly when such change is abrupt in nature. Rather than being the result of deliberative processes, past adaptations have as a rule emerged from upheavals triggered by environmental change. Past adaptation has occurred out of necessity, after damages have already been incurred, and itself is not without cost. We may be justified in viewing civilisation as a form of adaptation to climate change, but the negative aspects of the transitions with which its development is associated, for example increased inequality and violent conflict, suggest that it should perhaps be viewed as a "sub-optimal" adaptation by the standards of present-day aspirations.

10. Conclusions

The archaeological and palaeoenvironmental evidence is consistent with the notion that the development of complex societies in the Middle Holocene was largely the consequence of the responses of the precursor societies to deteriorating environmental conditions. This deterioration was associated principally with the orbitally driven weakening and southward retreat of the northern hemisphere monsoon belt. In all the regions examined in the preceding sections, the emergence of complex societies coincided with or followed a period of increased aridity. A general trend towards desiccation after about 8 kyr BP was punctuated by shorter (decadal to centennial scale) episodes of increased aridity. While desiccation trajectories were mediated by local and regional factors and feedback processes such as the collapse of vegetation systems, there appear to have been episodes of accelerated aridification that were coherent throughout the monsoon belt. One such event occurred around 6 kyr BP, and may have been associated with cooling in the North Atlantic. Enhanced regional aridity following this event coincided with socio-cultural change, particularly in the Eastern Sahara, Egypt and Mesopotamia. Regional data suggest another episode of accelerated change at the end of the 6th millennium BP, when discontinuities are apparent in archaeological records from across the Afro-Asiatic desert belt, and abrupt environmental changes are suggested by records from northern Africa, China and northern South America.

The evidence for linked environmental and social change is very strong in the central Sahara and Egypt, where responses to aridity are evident in local archaeological records. While the data are consistent with the hypothesis that social complexity was stimulated by increased aridity in Mesopotamia, the Indus-Sarasvati region, northern China and coastal Peru, further field-based research is required in order to link social change explicitly with environmental change at local scales, for example as represented by individual settlements. In particular, more high resolution palaeoenvironmental data are required from these regions in order to establish local trajectories of environmental change that may be related to local archaeological records.

FULL PAPER at HYPERLINK "http://tinyurl.com/h268j" \t "linkWin" http://tinyurl.com/h268j

doi:10.1016/j.quaint.2006.01.013
Copyright © 2006 Elsevier Ltd and INQUA All rights reserved.