2 RESEARCH FRAMEWORK
2.1 Research tradition
2.1.1 Tree-ring data in historical research
As early as 1914 Andrew Ellicott Douglass proposed that climate-sensitive tree-ring records provide novel material to study the environmental conditions of certain his-torical events.21 Yet it took several decades before tree-ring data were included in historical research, and where this was the case historians tended to consider den-dro-sciences mainly as a dating method.22 In his early works, Emmanuel Le Roy Ladurie, one of the pioneers of climate history, was sceptical of the suitability of tree-ring material in historical research, and especially its ability to reflect past grain yield fluctuations. This was mostly because of the poor temporal and spatial distri-bution of tree-ring measurement series available from Europe at the time.23 Over the last four decades, the growing number, sophistication, and easier accessibility of tree-ring based reconstructions have opened up new archives for historians. How-ever, integrating human history with the information captured in the tree rings is a task that has largely been left to natural scientists.24 Historians, although referring increasingly to tree-ring studies in their work, commonly treat tree ring material more as supplementary information than as source material.25 Yet, the full potential of including tree-ring material in historical research was demonstrated to the histori-cal community by Bruce M.S. Campbell, first in his study on the connections between environment and society in pre-modern England (2010) and later in his monograph
‘The Great Transition’ (2016).
In addition to reconstructing past climate variability, tree-ring series have been successfully used to reconstruct rain-fed agricultural production and food reserves in America.26 In Finland, too, Ilmari Hustich found already in 1947 the existence of a strong correspondence between the annual variations of Scots pine (Pinus sylvestris) growth and crop yields.27 A few years later Peitsa Mikola (1950) noted that years of crop failure, especially those that were caused by frost damages, were also poor in terms of tree growth.28 However, although generally agreeing that adverse weather and climate were the primary cause of the various hunger years of the country, Finn-ish historians have questioned the reliability of the tree-ring data as source material, particularly in research on pre-industrial crop failures.29 Eino Jutikkala (1994), for
ex-21 Douglass 1914, 322.
22 Kolchin 1967, 30–31; Bannister & Robinson 1975, 210–211.
23 Le Roy Ladurie 1972a, 2–3; 1972b, 142–148.
24 See, e.g., Baillie 1999; Holopainen & Helama 2009; Büntgen et al. 2011, 2016; Schmid et al. 2015.
25 See, e.g., Behringer 2010; White 2011.
26 Burns 1983; Andersson et al. 1995; Therrell et al. 2006; Maxwell & Knapp 2012; Maxwell et al. 2013;
Bocinsky & Kohler 2014.
27 Hustich 1947, 9–10.
28 Mikola 1950, 205.
29 Tornberg 1989, 59; Keränen 1989, 226–227.
ample, doubted whether tree growth and agriculture are sensitive to the same climat-ic factors. He was especially suspclimat-icious of tree-ring material corresponding to grain yield fluctuations resulting from short-term weather events, like night-frosts during harvest time.30 Mirkka Lappalainen elaborated Jutikkala’s argument by rationalizing that whereas tree rings are indicative of the average growing conditions over the whole summer season, a single fatal frost night over the harvest time – which does not affect tree-ring growth – can destroy the whole harvest.31
2.1.2 Grain yields and hunger in the northern margin
Fluctuations in climate can be expected to have a considerable impact on peasant so-cieties in the marginal areas of agriculture, where either temperature or precipitation is clearly the principal yield-limiting factor.32 In areas near the thermal limit of crop cultivation, like in north-east Europe, the rate of plant growth is strongly regulated by temperature, and lower temperatures reduce growth speed.33 In these areas even small temperature departures can potentially cause severe crop failures. Severe crop failures, in turn, can be expected to have considerable impacts on the medieval and early modern peasant population, as their food availability was heavily based on local, self-sufficient agricultural production.34
However, Le Roy Ladurie noted in his essay ‘History and Climate’ that we should not naïvely assume the effects of any given meteorological conditions, for instance the alter-nation of wet and dry cycles, on crops.35 Therefore, instead of relying on general assump-tions, crop yield responses to climate variability should be carefully evaluated and defined in the region and period of study, as demonstrated by Christian Pfister.36 Lit-erature on yield responses to climate variability in the northern margin of agriculture is widely available.37 Yet, statistical data for the research on the relationships between climate and crop yields have been gathered commonly from the most recent decades alone. It may not be feasible to derive the understanding on the relationships solely from the era of industrial agriculture when aiming to understand medieval and early modern yield responses to climate variability because of technological and method-ological developments which have taken place in agricultural practices during the recent past. Moreover, the genetic plant material changed after the World Wars.38 For example, breeding has brought new crop varieties that are more resistant to weather stress and (weather-related) plant pests and diseases. Thus, industrialisation of agri-culture has not only increased the yield but may have also altered the crop responses to climate variability.
30 Jutikkala 1994, 11.
31 Lappalainen 2012, 32.
32 Parry & Carter 1985, 97; Toohey et al. 2016, 408.
33 Rantanen & Solantie 1987, 19.
34 Orrman 2016, 161.
35 Le Roy Ladurie 1972b, 146.
36 Pfister 1978, 224. See, also, Le Roy Ladurie 1972a, 289–292.
37 See, e.g., Mukula & Rantanen 1987, 1989; Carter & Saarikko 1996; Kaukoranta & Hakala 2008; Pelto-nen-Sainio et al. 2009a, 2009b, 2011, 2016a, 2016b, 2016c; Himanen et al. 2013.
38 Holopainen & Helama 2009, 217.
Moreover, it is known that all food shortages – or crop failures – do not inevita-bly lead to hunger. In addition, not only declines in food availability but also enti-tlement failures, which are caused by human and institutional actions, are known to cause hunger.39 This entitlement approach was pioneered by Amartya Sen in his work ‘Poverty and Famines’ (1981). Since then, the economic nature of hunger has gained profound intrest among scholars of famine.40 However, a strong economic focus may not be the most fruitful approach to better understand hunger among self-sufficient peasant societies.41 For example, in western Finland, only one out of six years of elevated mortality over the period 1750–1850 coincided with a year of high grain prices.42 Furthermore, I have proposed earlier that if grain prices were as-sociated with food shortages in early modern Finland, rising prices were most likely the consequence of – rather than the cause for – food shortage. This is because the price peaks commonly lagged crop failures by a couple of years.43 An entitlement ap-proach might be only applicable if an institution or other actor can control the entitle-ment to food, for example through exchange, markets, or trade.44 In the northeastern periphery over the study period this was not always the case.45
An alternative approach, following Thomas Malthus’s theory, is to see hunger (and related mortality) as inevitable consequence of population growth when human population grew faster than food production.46 This approach long dominated the in-ternational famine studies focusing on pre-industrial societies.47 And last, hunger can also be considered as a hazard, when exposure to an event, like extreme climate, has harmful consequences.48 In Russia and Finland, medieval and early modern hunger has been largely understood through the latter approach, primarily as a consequence of production failure due to adverse climate and weather.49 However, this generally agreed conception of hunger as a ‘natural hazard‘ originates from rather old litera-ture, and has been hardly revisited.
2.1.3 The ‘Kingdom of Frost’
Miikka Voutilainen noted sharply in his recent study on the Finnish hunger years of the 1860s that the international literature published after the 1970s has had little impact on Finnish famine studies.50 Indeed, the main discussion in Finland has focused on wheth-er malnutrition or diseases wwheth-ere more fatal for the pre-industrial population during
39 Millman & Kates 1990, 9.
40 Voutilainen 2016, 67.
41 See, Soininen 1974, 351–365, for the discussion on the little impact market economy had on the livelihood and subsistence of pre-industrial peasant in Finland.
42 Jutikkala 2003, 504.
43 Huhtamaa 2018, 56.
44 Sen 1981, 155.
45 See, Korpela 2014b, 185–188.
46 Malthus 1798, 44.
47 Slavin 2016, 434.
48 Millman & Kates 1990, 7.
49 See, Melander & Melander 1924 (Finland) and Kahan 1968 (Russia).
50 Voutilainen 2016, 22.
the hunger years.51 Yet, the initial cause of hunger has not received careful analysis.
Instead, food shortage, more precisely crop failure due to harvest time night-frost, is seen almost unanimously as the primary trigger for hunger in medieval and early modern Finland. Dramatic titles like ‘The Constant Danger of Frost’, ‘Land of the Frost-Devil’, and ‘The Kingdom of Frost’ are common in the literature.52 The direct causality between frost, crop failure and hunger is such a well-established hypothe-sis in the Finnish literature that records of frost year, year of scarcity, and hunger year (hallavuosi, pulavuosi, and nälkävuosi in Finnish, respectively) are commonly regarded as indications of crop failure years.53
The research on pre-1700s hunger in Finland has focused almost solely on the famine of the 1690s.54 In these studies, the interpretations on the role of the institu-tions, authorities, population pressure, economic factors, and disease, inter alia, in contributing to the development and severity of the famine differ. Yet there is a con-sensus that severe crop failure due to night-frost was the decisive trigger of the crisis.
However, these studies are not simply works of environmental determinism, as all of these consider the significance of pre-crisis socio-economic conditions and possible relief measurements. Nevertheless, the relationship between crop failure and hunger is rather unidirectional.55
Research on crop failures and hunger extending prior to the 1690s is scanty. Viljo Voipio (1914) explored the severity and human consequences of the 1601 crop failure in south-west Finland. Väinö Johanson (1924) assessed the frequency and causes of crop failures in Finland from the 17th century onwards. Matleena Tornberg (1989) examined yield ratio variability in the 16th and 18th centuries south-west Finland.
To date, Kurt Reinhold Melander’s and Gustaf Melander´s (1924) work is the only overview on the topic that extends back to the medieval period (to the 14th century) from Finland. Because of the lack of early sources from the region, the catalogue in-cluded some known Swedish and Russian famines as indicators of crop failure years in Finland over the earlier centuries. However, no references to any primary sources are given in the chronology. In their chronology of Finnish crop failures, three crop failure years in the 14th century, ten in the 15th, 16 in the 16th, and 33 crop failure years in the 17th century were identified. The catalogue was later reproduced large-ly in Timo Myllarge-lyntaus’s study (2009). These works saw harvest time night-frost as the most common reason for severe crop failures in Finland. Yet other factors like drought or excessive precipitation were also suggested to have caused considerable yield damages and hunger.56
The increasing number of recorded crop failure events in the 17th century has been further connected to the cooler climatic conditions of the Little Ice Age (LIA),
51 See, e.g., Turpeinen 1986; Jutikkala 1987, 2003; Soikkanen 1991; Häkkinen 1994; Kallioinen 2005; Lappa-lainen 2014a. For further review on the ‘malnutrition-disease’ dabate, see, VoutiLappa-lainen 2016.
52 Jutikkala 2003, 292; Solantie 2006, 185; Lappalainen 2012, 47, respectively.
53 See, e.g., Melander & Melander 1924; Myllyntaus 2009.
54 Jutikkala 1955, 2003; Neumann & Lindgren 1979; Mäntylä 1988; Muroma 1991; Katajala 1994; Lappalainen 2012, 2014a.
55 Voutilainen 2016, 24.
56 Melander & Melander 1924, 352–353; Myllyntaus 2009, 81.
especially in more recent reseach.57 However, the occurrence of night-frost, the most commonly named cause of crop failure in Finland and Russia,58 is more dependent on weather than climate. In the studied area, harvest time night-frosts are usually radi-ation frosts. These result because of temperature inversion, which occur under clear sky and calm winds. In this situation, radiant heat rapidly rises after sunset, and cold air, being heavier, flows down and stays trapped below the warmer inversion layer.
The topography of the fields affects the frost prevalence, valleys and depressions being the highest risk areas, whereas a close proximity to waterbodies decreases the likelihood of frost occurrence. These radiation frosts are local and short-term weath-er phenomena, lasting only hours ovweath-er the night.59 Therefore, instead of long-term mean temperature, the occurrence of night frost in the region is mainly dictated by daily weather: calm, clear and dry atmospheric conditions. Thus, perhaps previous historical research has failed to make clear distinction between short-term whether phenomena (like radiation frost) and long-term climatic regimes (like the LIA).
In Finland, the relationship between crop failures and hunger is commonly seen as generic.60 Solely the fact that records of hunger and famine from Finland, or even from Sweden or Russia, have been used as proxy data for crop failure years sug-gests that these two variables are seen inherently connected. Consequently, based on the crop failure chronology by Melander and Melander, Myllyntaus proposed that during the early modern period at least part of the population faced a famine every third year during those two [17th and 18th] centuries.61 The frequent famines are associated with the cool climate of the LIA. However, the frequency of early modern crop fail-ure and hunger years was soon challenged by Jari Holopainen and Samuli Helama (2009), who noted that the information on past crop failures has been gathered from spatially fragmentary sources. By analysing the spatial synchrony of early modern yield ratios, they suggested that the early modern crop failures listed in these cata-logues might have been much more local than previously assumed. Consequently, they have suggested that the Finnish famine history may actually not be as dark as as-sumed based solely on the listed years [by Melander and Melander].62