The central proposition of Toby Pillatt is that in developing an understanding of past human affairs weather is as important as, or more so than, climate. Climate may be simply defined as average weather, whilst weather is the day-to-day occurrence of atmospheric phenomena which impact in perceptible ways on people's lives. The general proposition is sound enough; the challenges come in implementing these ideas in ways which advance our understanding of past people–environment relationships.
The paper is based on historical sources: two 18th-century diaries recording weather in the Lake District compared with the Central England Temperatures series compiled from instrumental records since 1772 (Parker, Legg and Folland 1992). It is shown that perception on the ground may differ from instrumental records or secular trends (i.e. major period of distinctive climate). Furthermore, agricultural changes concerned with land use may relate more to economic and social factors than to weather or secular climatic episodes. Pillat makes a fair point in arguing that archaeologists have given great emphasis to climate change, without much consideration of weather. Also in need of consideration is the timescale of climate change, whether it would have been perceptible to human communities or, if it was not perceived, by what precise mechanisms it engendered less conscious social or economic change. However, the argument as presented is founded largely on historical sources and it does not resolve how we can investigate weather phenomena in those periods for which no historical sources are available and where we are often reliant on proxy evidence (i.e. indirect palaeoclimatic records).
In order to factor weather into our thinking we need to consider the basic distinction between climate and weather and the evidence which is available to us as archaeologists. Of particular significance is the timescale of the weather and climate phenomena in question and thus the extent to which changes would have been perceptible. This includes consideration of extreme events which may, or may not, be part of secular climatic trends. Such questions are a familiar part of current debate between those who identify global warming as a major environmental and political issue and the small minority of scientists who are global-warming sceptics. It is debated whether particular episodes of more extreme weather – e.g. warmer summers, colder winters, or more stormy conditions – are parts of longer-term secular trends or further evidence of the natural variability of climate. The sequence of international meetings, beginning with Rio and Kyoto, and most recently in Durban, highlight the significance of developing an increasingly refined understanding of the timescales of climate phenomena and extreme events and also how they affect human communities in various economic stages and geographical situations. As Mitchell (2008) and Van de Noort (2011) have emphasized, archaeology can, and should, be playing a far more active role in the ongoing debates about climate change and the coping strategies of past and future societies.
Of particular interest are periods of especially marked climate change, or extreme weather conditions, which are increasingly recognized, closely dated and likely to have impacted on human communities. Outstanding among these are the very rapid climate changes at the end of the last glaciation within less than a human generation (Alley 2000). Within the Holocene, millennial-scale cyclical cooling episodes are marked by ice-rafted debris in deep ocean cores, the so-called Bond events (Bond et al. 1997). The most marked of these Holocene events at 8200 B.P. was apparently caused by catastrophic discharge of water from the Laurentide ice sheet and its effect on North Atlantic oceanographic circulation.
The increasing availability of well-dated, sometimes annual palaeo-environmental records from ice cores, tree rings and laminated sediments creates opportunities for correlating cultural changes with both secular climatic trends and weather phenomena represented by extreme events. Ice cores with high, at times annual, resolution for much of the Holocene preserve isotopic records as well as trace gasses, atmospheric dust records, geochemical sequences and carbon-particle records of palaeofire occurrence, all of which contribute as proxy sources to palaeoclimate studies (Bell and Walker 2005). Tree ring sequences also provide palaeoclimatic records, including evidence for extreme weather events in particular areas (Baillie 1995). Annually laminated sediments can provide long datable annual-scale palaeoclimate records in areas such as Scandinavia where lakes are frozen for part of the year. Shorter, and not precisely datable, coastal laminated sequences have also been shown to exhibit annual banding and to have potential for the investigation of climate/weather variability (Dark and Allen 2005). It must be acknowledged that, however precisely dated a particular environmental change is, it often still poses significant challenges of correlation with cultural changes. Baillie (1991) highlighted the problem which he labelled ‘suck in and smear’, whereby precisely dated palaeo-environmental events are correlated with cultural changes which are sometimes not at all precisely dated. An example is a well-attested climatic deterioration c.800 B.C. for which there is good evidence in peat bog and other palaeo-environmental sequences over a wide geographical area (Van Geel, Buurman and Waterbolk 1996). To this deterioration has been attributed the abandonment of burnt mounds and field systems in many upland areas of the British Isles. However, the abandonments are much less precisely dated and no correlation or climatic causation can necessarily be assumed without specifically testing the hypothesis. Indeed, as Pillatt identifies, many factors apart from climate can lead to abandonment, or establishment, of upland fields.
During the 19th century and most of the 20th, environmental and biological science was founded on gradualism (processes operating at very slow rates by tiny increments), and a literal (or substantive) adherence to uniformitarianism (the present is the key to the past). In early ecology this led to the factoring out of human agency in the quest for ‘natural ecology’, which became increasingly problematic with the recognition that so many habitats, once considered pristine, have been affected by long histories of human activity. Today there is greater recognition of the significance of contingency (chance factors) and punctuated equilibrium (periods of rapid change and stasis), and thus awareness of the transformative potential of high-magnitude, low-frequency events (Gould 1999), which include weather phenomena. These have been significant developments for archaeology because human agency can now take its place as part of a spectrum of environmental disturbance factors, including not only weather, but also faunal agents, disease and many others (Bell and Walker 2005, chapter 6). For palaeo-environmental scientists working in Britain a particularly significant event was the great storm, or hurricane, which hit southern England on 16 October 1987, felling 15 million trees (T. Brown 1997; Lamb and Frydendahl 1991, 189). Prior to this there had been a tendency among palaeo-environmentalists working in Western Europe to attribute environmental disturbance uncritically to human agency, perhaps because there such extreme events are – by comparison with the Caribbean or southern United States, for instance – so rarely observed.
The effects of extreme weather conditions are particularly apparent to archaeologists working in a coastal and maritime context. Most shipwrecks are the result of extreme storm events (Fenwick and Gale 1998). In some storms very large numbers of ships were lost, the series of storms which destroyed the Spanish Armada in A.D. 1588 being a particularly significant historical example for which the meteorological conditions have been reconstructed using historical sources (Lamb and Frydendahl 1991, 40). Coastal sediment sequences also reveal evidence of extreme storm events in the form of coarser sediment increments, and many of the major coastal changes will result not from average conditions, but from the high-energy conditions associated with high-magnitude, low-frequency weather-related events. Examples would be the inundation of coastal forests to create submerged forests, the breaching or formation of coastal barriers and lagoons, and so on. Similarly, in coastal dune contexts, burial of archaeological sites and landscapes will have been concentrated during storm events. Such episodes seem to have been particularly frequent during the Little Ice Age secular climatic episode, from 1550 to 1850 A.D. (Grove 2002), when many coastal settlements along the Atlantic seaboard from Brittany to Scotland were inundated by sand dunes. It is of interest that there is also evidence for episodic dune deposition punctuated by stable conditions more locally in western Britain during the Bronze Age.
Although not always so obvious, weather events are similarly significant in many fully terrestrial contexts. Many sediment increments which archaeologists encounter may not derive from average conditions over extended timescales but from particularly intensive periods of high geomorphic activity related to particular weather events. This will apply to riverine processes, including extreme overbank flooding, changes of river course and the deposition of major sediment increments. These can sometimes be recognized as the result of major flood events, as in the case of the dendrochronologically dated medieval Hemington bridges in the River Trent (A.G. Brown 1997, 2009). Likewise, in the case of colluvial slope processes, comparison of present and past erosion evidence on arable land highlights the role of infrequent high-magnitude rainfall events (Bell and Boardman 1992).
Archaeologists have had a tendency to make deterministic assumptions about particular climate or weather phenomena. As Pillatt shows, historical records can be valuable in revealing how particular events were perceived and the reaction this produced. The concentration of witch burnings in years of particularly adverse weather represents an extreme example (Behringer 1999). A major flood event which impacted in the Bristol Channel and Severn Estuary in A.D. 1606 was attributed to God's warning (anon. 1607). It is debated whether this event resulted from a major Atlantic storm or a tsunami (Haslett 2007). The Great Till Flood in a Wiltshire valley resulted from storm runoff from agricultural land in 1841, leaving three dead and 200 homeless; it prompted a local vicar to write a poem, sold to aid the distressed, which suggested the event was retribution for the ‘wasted hours and squandered days’ of those affected (Cross 1967). Such perceptions as late as the mid-19th century indicate the probability of past responses to events very different from those which might be anticipated from a post-Enlightenment scientific perspective.
This contribution supports Pillatt's proposition that archaeologists and palaeo-environmental scientists need to give greater consideration to weather-related phenomena. This is becoming increasingly realistic as high-resolution and well-dated paleoclimatic records from tree rings and annually laminated sediments facilitate the identification of periods of rapid climate change and unusual years, even seasons. The sediments within which archaeological contexts lie are probably, more often than we recognize, the products of high-magnitude, low-frequency events. It is acknowledged that archaeologically based palaeo-environmental investigations are seldom directed towards establishing whether a particular deposit reflects a short-term event or gradual deposition. This highlights the value of more event-specific modern analogue studies such as comparisons with biota or sediments from recent floods. There is also a tendency to make simple, generally deterministic, assumptions about the effects of particular climate and weather-related phenomena on people. The perceptual points made augment Pillatt's argument concerning the complexity and close relationship between environment and society. The current significance of debates about climate change and sustainability, and the contribution which archaeology has to make to them, require development of a more sophisticated conceptual toolkit than has generally been employed in archaeological discussions of these issues, as Van de Noort (2011) has argued. This should develop beyond the identification of simple deterministic relationships between weather, climate and social change and give greater consideration to the precise timescale of the changes we record, to the question whether they are likely to have been perceptible to human communities at the time, to the role of human agency (Mitchell 2008) and to the evaluation of the diverse spectrum of possible coping strategies (Bell and Walker 2005, 140) which past societies may have employed.
