Making a feast for the deceased

(1)

Master thesis, 30 ETCS

Environmental Archaeology Masters programme, 120 ECTS Spring Semester 2019

Making a feast for the deceased

Archaeobotanical investigation of plant remains in ovens used for drying cereals and food

consumption through ovens, graves, and bog bodies

Daniel Smeds

(2)

(3)

Foreword

First, I would like to express my deepest gratitude towards those who helped me in this endeavour. As I was able to conduct an internship at Moesgaard Museum at the department of conservation and natural science I owe a strong gratitude to all the personnel and staff,

especially Peter Mose Jensen, who worked as my supervisor and board game friend as well as Peter Hambro Mikkelsen who allowed me to perform my internship there, as well as offerings his excellent noodle dinner. Tobias Danborg Torfing at Sydvestjysk Museer for allowing me to use his material for my thesis also deserves thanks.

I also owe a big thanks to all the teachers I had throughout my university life, including Dr.

Holmblad and Dr. Rundkvist. Dr. Buckland who has worked as my supervisor now for two theses. He has been a great support to have in my many times of need as well as filled with fun anecdotal information about archaeology and beetles. I also owe gratitude to Dr. Ivanka Hristova, who worked as co-supervisor and guided me through many hours of environmental archaeology sessions in the lab. Dr. Linderholm, who has been motivating all of us master’s students and constantly pushing us using his fermented herring and carrot method, deserves a sincere thanks.

A special thanks to those two students who have been through all five years of university with me, I would like to thank, Ida Lundberg, who has also helped with a lot of knowledge about plants and wildlife and Love Eriksson. The other students from the master’s programme have also been of tremendous help such as Ola Lindgren with his GIS knowledge together with Eva Kourelas computer help. Stefano Vilardi and Balint Toth have also been of great help and fun during the many classes we had together. Stressing about work with all of You is more fun together than stressing on your own.

Mats Eriksson, working at the environmental archaeology laboratory (MAL) in Umeå who has been supporting with motivation, statistical help, and good times.

My parents should also be thanked as they have supplied me with endless encouragement throughout my life. Last but not least, I am also in debt to Ming-Hsin Lü, who has been a pillar during my theses work and the best partner one can have.

(4)

1

Abstract

Det finns många frågetecken kvar kring konsumtionsbeteende inom förhistoriska kulturer.

Målet med denna studie var därför att studera och förbättra förståelsen av det arkeobotaniska materialet i ugnar gjorda för att torka sädeskorn och hur ugnarnas innehåll korrelerar med den växtbaserade matkonsumtionen, samt att försöka se potentiella samband mellan ugnarna, deponerade växtrester i gravar och de två mossliken Grauballemannen och Tollundmannens maginnehåll.

De 14 ugnarna, belägna i Sydskandinavien och daterade till järnåldern, analyserades och jämfördes för att se hur deras innehåll förhåller sig både i tid och rum men även jämfört med det övriga åkerbruket från bronsåldern till medeltiden. Det arkeobotaniska fynden visade att i yngre bronsåldern odlades det Hordeum vulgare var. vulgare, vilket följdes av en period då Hordeum vulgare var. nudum odlades. Detta pågick till slutet av romersk järnålder då

Hordeum vulgare var. vulgare igen blev den vanligaste grödan att odla. I slutet av järnåldern odlades både Avena sativa och Hordeum vulgare var. vulgare. Övriga sädeskorn som

Triticum sp. och Secale cereale har även odlats, dock endast till en mindre grad enligt fynden från ugnarna.

De arkeobotaniska fynden jämfördes sedan med den rådande bilden av matkulturen under järnåldern, vilket visar indikationer på att Chenopodium album, Fallopia convolvulus, Persicaria maculosa/lapathifolia, Plantago lanceolata, Poacea, Polygonum aviculare och Spergula arvensis troligen har konsumerats tillsammans med de funna sädeskornen. Flera av dessa påträffades i de samtliga analyserade fornlämningarna.

Jämförelseanalyser i dendrogram mellan gravarnas och ugnars makrofossila rester samt de båda mosslikens maginnehåll visade inga definitiva samband. Dock fanns det gravar som korrelerade relativt starkt med ugnarnas innehåll, möjligen på grund av dess stora

fyndmaterial av makrofossil. Dessa kunde visa indikationer på mat och måltid har deponerats i gravar.

Keywords

Archaeobotany, Cereal drying ovens, Food consumption, Grave deposition, Bog body stomach content

(5)

2

Figure index

Figure 1. Potential operation sequence of processing cereals in southern Scandinavia during the Iron Age. Source for the figure: redrawn after Viklund 1998:35, by Sofia Lindholm ... 8 Figure 2. Chronological description of time periods in south Scandinavian prehistory (Based upon Grabowski 2014)... 12 Figure 3. Distribution map of southern Scandinavia, showing the analysed features (Map by Ola Lindgren)... 19 Figure 4. Agricultural species of the archaeobotanical remains in the ovens according to periods .... 24 Figure 5. Agricultural species of the archaeobotanical remains in the ovens according to their dated period ... 25 Figure 6. Percentage of cereals found in each oven according to their dated period ... 26 Figure 7. Non-agricultural species of the archaeobotanical remains in the ovens according to their dated period ... 27 Figure 8. Non-agricultural species of the archaeobotanical remains in the ovens according to periods ... 28 Figure 9. Archaeobotanical remains of the cereals found in graves, grouped according to the graves dated period ... 29 Figure 10. Cluster analysis (A) of all ovens, graves and bog bodies using the Jaccard similarity

coefficient. The analysis uses presence or absence of the taxa as comparison. Cluster analysis (B) of ovens, graves and bog bodies using the Bray-Curtis similarity coefficient. The analysis uses the abundance of the taxon in the features. Highlighted features are discussed in more depth (analysis done with PAST3 (Hammer & Harper 2001)) ... 30 Figure 11. Ecocodes of the archaeobotanical material found in the ovens, graves and the two bog bodies. ... 32 Figure 12. Bar chart of the botanical remains of the features ... 33

(7)

4

1. Introduction

Archaeologists often think about what made a culture able to survive, and the procurement of food is without a doubt one major aspect of this (Hambro Mikkelsen & Nørbach 2003).

Archaeobotany and research concerning agriculture have greatly increased the knowledge of cultural development. It enables a reconstruction of the food habits and can be a testimony to social hierarchy through e.g. importation of exotic food (Viklund 1998:31). Although

archaeobotany only contains the plant-based side of food, the study of diet and subsistence is a vital aspect of culture as it also reflects the social context of consuming food (Palmer & van der Veen 2002). There is, furthermore, also a skewed research bias in consumption habits.

There is a great deal more work done on consumption patterns and feasting habits based on osteological remains than of botanical macrofossil remains (Hansson 1997:11), although this has started to change. Thus, it is important to continue this research into the food habits of the prehistoric people.

During the authors’ internship at Moesgaard Museum at the department for Conservation and Natural Science, a set of samples from two Jutlandic sites with Iron Age ovens used for drying cereals were analysed archaeobotanically. Ovens are a very interesting set of feature type as they are a deliberate selection of gathering from the agricultural fields. Activities ranging all the way from sowing and harvesting to selection of fuelwood for the ovens and subsequent drying of the cereal grains are represented. Therefore, the samples from the ovens will serve as a base for this thesis about the Iron Age ovens and consumption habits. The four ovens will be compared to other similar ovens from prehistory and the Iron Age agriculture.

The food aspect will be investigated primarily in earlier studies to get a grasp on what the consumption habits have been. Depositions in graves and investigation of the last meals of the bog bodies, yielding potentially indications of a ritual meal before death occurred. Both these feature types are good examples of both rituals and consumption habits. Therefore, an attempt in order to investigate a potential correlation between ovens, graves and bog bodies last meal will be carried out.

(8)

5

1.2 Aims and research questions

This thesis aims to investigate Iron Age ovens used for drying cereal grains in south Scandinavia through their archaeobotanical remains, both in changes through time and the consumption potential of the remains. This will, in turn, be compared against archaeobotanical remains of graves and two bog bodies, in order to investigate if a potential relationship of food can be detected between these structures and remains.

1. How do the archaeobotanical remains in ovens used for drying grains change through the Iron Age and how do they compare to the

agricultural landscape of south Scandinavia?

2. How do the archaeobotanical remains found in the ovens fit according to the current knowledge about Iron Age plant-based food?

3. What kind of similarities and differences are there between the ovens, graves, and bog bodies in terms of archaeobotanical remains and can any ritual meal be discerned?

(9)

6

1.3 Theoretical perspective

Food is consumed every day if possible. Therefore, it is of no surprise that some food has also developed into a ritual context, such as funerary rites and food offerings to deities (Palmer &

van der Veen 2002). Archaeobotany often focuses on the gathering of wild plants and

reconstructing the agricultural practices of which tends to be based on the habitat and physical properties of the plant. These uniformitarian theories of past agricultural lack to some degree, the necessary social and cultural interpretation that is needed when dealing with food.

Together with ethnographical, anthropological and historical studies, a more holistic approach can be achieved (Heiss & Hansson 2014).

In anthropology, consumption and food processing is a non-verbal language that all human societies participate in. The food system with whom, when, where, and what is being eaten, is a cultural tradition and rules of a people, mirroring the values and belief systems (Hitch 2015:116). Furthermore, there is also a difference worth noting between a diet and a meal. A meal is a single occurring event, while a diet is a consisting pattern of meals (Hansson 1997:37). This means that both graves and stomach contents of bog bodies only represent a single meal, which may not have been the norm for food intake. As theories of a special last meal for the bog bodies have been proposed, deliberate selection plants have also been proposed (Palmer & van der Veen 2002). Graves also represents a deliberate deposition, although they may not merely contain the last meal, cereals and seeds found may also have been deposited for symbolical purposes, not intended as a meal (Hansson 1997:37).

Therefore, and for the lack of accessible archaeobotanical data of graves, both cremation and inhumation graves from multiple periods will be investigated, although there are clear biases in the taphonomy in different kinds of graves, as will be discussed below.

Consuming food is a very nuanced way of expressing oneself. It can represent the cultural identity such as ethnicity and social and economic situation. For instance, Palmer and van der Veen (2002) proposed that in Roman societies and other societies that are more structured and have a strong hierarchy, the emphasis lies on exotic food and beverages, while archaic

societies tend to focus on the quantity of food. There is also a huge variation in what kind of food that is served and consumed, with elaborate or ritualistic food more likely to be reserved for celebrations or an honoured guest and have a highly selective process of which ingredients to use. On the other hand, more mundane food is eaten on a more daily basis. This, of course, is also dependent on how a meal is prepared and how it is consumed. Food is an important way of expressing one’s identity, this also includes food deposits in religious and burial

(10)

7 practices. Festivities and religious ritual are also known to include drinking and feasting (Heiss & Hansson 2014).

Ethnographical studies have taught archaeobotanists a great deal about consuming habits.

These kinds of research can very well be of use for interpreting the consuming habits during the Iron Age in Scandinavia (Viklund 1998:17). Hillman (1984) assessed that prehistoric agriculture had around 30 different steps of process before an edible meal can be achieved.

During these steps, species are divided, or part of the plant are separated, and preservation of plant material can occur at any time. Although the same procedures might not have been done in the same way through time and space, as the method and steps of agriculture may have changed or become unnecessary. Cereal cultivation and food habits are, however, both considered to have a stronger connection to the humans and thus less susceptible to change (Viklund 1998).

There is still a lot to uncover in terms of producing cereal, as questions around the chain of processes remain unclear. It is safe to assume that most of the grains dried in ovens came for the local surrounding (Viklund 1998:16), but this is just one step of a large chain of operations (see figure 1). Houses can sometimes be considered as a direct insight into the material of the house just before it burnt down. Ovens, the other hand, are more often thought to be of repeating scenarios, meaning that it has been exposed to multiple usages before it was abandoned. Therefore, the ovens are able to present the agricultural practices in a more nuanced way. In that sense, ovens can also show a more diverse picture of the past food culture (Grabowski 2014). The ovens are in contrast compared to graves depositions and the bog bodies last meal which would hold a single time deposition or consumption of a meal.

(11)

8

Figure 1. Potential operation sequence of processing cereals in southern Scandinavia during the Iron Age. Source for the figure: redrawn after Viklund 1998:35, by Sofia Lindholm

Ritual is often used when dealing with scenarios that are considered extraordinary, when in fact a ritual can also be done as a mundane or everyday action. In this thesis, I am to use the word Ritual as an action with a greater degree of a formality than an ordinary day-to-day activity, although these two actions can overlap with each other (Bradley 2005:104, 106).

Therefore, there are no clear distinctions on what rituals and the ordinary are. It very much depends on the situation the scenario takes. A deposition, or a meal, can have different meanings depending on what the outcome is desired (Bradley 2005:119-120). Although it is important to remember that not all preformed rituals are associated with religious beliefs or the supernatural (Bradley 2005:28).

Care is also due when an archaeologist is trying to examine prehistorical ritual acts using contemporary models of explaining actions on societies when there seems to be a lack of a clear goal or outcome of the ritual action (Brück 1999). This could be because the Christian ritual often takes place in a special place, performed by special people for a clear and well- established outcome (Bradley 2005:35). Sometimes a ritual act does not appear to do anything, and therefore, they must have some another purpose or needs to be reinterpreted (Brück 1999:318). Thus, it can be quite difficult to find out whether or not a consumed meal or drink had a ritual tone to it.

Depositions in lakes and bogs have been an ongoing process from the Mesolithic in Sweden (Bradley 2005:143). These acts, involving agricultural tools, people and artifacts in wetland areas, are very closely linked with rituals and the prehistoric people’s beliefs systems. This

(12)

9 too applies to fertility and death during the Danish Bronze Age and Iron Age. Although one should be careful not to overinterpret artifacts and finds as there has been a number of researchers suggesting that plant remains in e.g. postholes could be considered as a ritual deposition (Regnell 1997).Although, this has been criticized and is more likely it had to do with the layout of the interior structure and layout of the house (Viklund 1998).

Taphonomic processes are always something to consider when interpreting archaeological remains. This, too, applies for extraction of archaeobotanical remains. This includes 1) the use of an artifact before deposition, 2) the preservation capability of a site, 3) processes of

deposition, 4) potential post-depositional actions and 5) the excavation of the material. This will all factor in towards the result, and vary depending on the archaeological methods, even though sites may originally have had similar remains (Grabowski 2014:29). One problem with the handling of data also exists when researchers are biased towards their results. This may lead to researcher choosing a part of the data which is either confirming or disproving a hypothesis (Grabowski 2014:36).

Preservation of botanical remains is dictated by a number of criteria. Carbonised plant

remains are, perhaps, the most common form of preservation. Seeds, or other plant parts, need to be exposed to sufficient heat in order to become carbonised. The process will also factor in the most in this thesis as both ovens and graves include great heat. Other preservation

processes include waterlogging which will factor in towards the bog bodies, but also mineralisation, desiccation and imprints of plant remain on e.g. ceramic pots (Grabowski 2014:6). In order for seeds to be preserved as non-charred remains, a special kind of

environment must be present. These environments must be oxygen-poor such as waterlogged, frozen environments, or sometimes even deeply buried in soils, depending on the acidic level.

Seeds are, however, better preserved as carbonised than untreated, as non-charred seeds usually are easier subjected to fungus, rot, and decay (Viklund 1998:31).

The carbonisation process must be of a certain temperature for the seeds to be able to be preserved. If the temperature is too high, then the seeds contain a high amount of fat, e.g.

Linum usitatissimum and Camelia sativa, can be completely or partially destroyed. In heat with a high amount of oxygen will usually result in only ash remains (Grabowski 2014:7).

Between 250 – 500°C in a low oxygen environment complete combustion of the cereals is achieved (Viklund 1998:97). However, the charring of seeds affects both the morphological characteristics, and their shape or surface pattern may be distorted (Grabowski 2014:7). Some seeds disintegrate, while others may only have slight distortion. The charring process can

(13)

10 happen in many various ways, e.g. accidental burning of a house often results in botanical remains in postholes and deliberate drying of cereals in ovens can produce charred grains (Viklund 1998:90).

2. Background

2.1 Early farming in south Scandinavia

The first traces of agricultural practises in Denmark are from the latter half of the fourth millennium BC. Although agriculture is presumed to be a small supplement to hunter-

gatherers, this new way of acquiring food soon came to revolutionize the environment and the population. The transition to agriculture, not only changed the food production but had also an enormous impact on the economic and social structure. It allowed for an increase in

population but demanded an increase in effort of food production (Jensen 1982:70). The areas which were first farmed were previously covered in forest and humans, thus, impacted the local vegetation and fauna. Although modest and sparse at first, around 3200 BC that the agrarian culture started to rapidly expand (Jensen 1982:73, 77).

With the introduction of ard-ploughing technology in the third millennium BC, a gradual transition from slash-and-burn technique started (Jensen 1982:80). The Neolithic agrarian cultivation consisted of spelt (Triticum spelta), emmer (Triticum dicoccum), einkorn (Triticum monococcum), and naked barley (Hordeum vulgare var. nudum), although by the late Neolithic einkorn was only sparsely cultivated and consisted in a similar manner until the Late Bronze Age (see figure 2 for year classifications). In Jutland, these types of cereals remained longer in use than in Scania. There, a new type of grains was also being cultivated more and more, hulled barley (Hordeum vulgare var. vulgare), and eventually becoming the dominant type of cereal grain (Grabowski 2014:13). Around 1000 BC a new type of agriculture started to appear. The use of manuring became more and more prevalent along with the introduction of the three- aisled longhouse (Viklund 1998:139).

Around 500 BC – 1 AD, the cultivation of emmer and spelt decreased drastically while both hulled and naked barley seems to have been cultivated to a comparable degree. In the subsequent years, however, naked barley starts to decrease as well in importance and by the Late Roman Iron Age, hulled barley became the prevailing cultivated cereal (Grabowski 2013). The iron sickle was introduced sometime later in the Pre-Roman Iron Age, which

(14)

11 enable harvesting the straws closer to the ground, and the subsequent gathering of straws as fodder for the animals. Before the introduction of the iron sickle, hand-plucking of the cereal grains may have taken place, or by crude flint or bronze sickles (Hambro Mikkelsen &

Nørbach 2003:137).

Celtic field systems were in use from around 800 BC – AD 200 and were in use in Denmark, Gotland, and southern parts of Sweden, while stone clearance systems were in use in the south and central Sweden. Due to the large size of the stone clearance setting and Celtic fields, questions whether all of the fields were cultivated at the same time has arisen. Potentially they were cultivated for a few years to decades and then migrated to another field, while the

previous ones were left to serve as a fallow field (Hambro Mikkelsen & Nørbach 2003:125).

Around 400 – 500 AD, rye (Secale cereale) started to become more and more common, although hulled barley still remained the dominant crop until the Viking Age (Viklund 1998:174). Rye cultivation culminated around the Viking Period – Medieval Period when it seems to have had the same high status as hulled barley. Oat (Avena sativa) also saw an increase in importance during the Late Iron Age and indicating towards more intentional cultivation (Grabowski 2014:15).

The latter half of the Iron Age, 600-1100 AD, saw a transition to autumn-sown crops on a three-field or two-field system (Hambro Mikkelsen & Nørbach 2003:127, Viklund 1998:37), with rye being the autumn-sown crop and hulled barley being sowed in the spring (Grabowski 2014:22). This change also brought upon new weed species (Robinson 1994). A permanent field most likely relayed on repeated manuring as the nutrients in the ground was not to be depleted. These types of fields would have been sustainable for a few years up to a few decades. After the nutrients disappeared or certain weed species would have taken over the field, a migration of the farm would have taken place. Then the old agriculture land would have turned in to a fallow while a new field was being cultivated (Grabowski 2013). Regional differences in south Scandinavia, of course, occurs throughout both regions and time.

(15)

12

Figure 2. Chronological description of time periods in south Scandinavian prehistory (Based upon Grabowski 2014)

The development of infields and outfields during the Iron Age was not as developed as in the Medieval and Post-Medieval Period was. The infield is, in its roughest form, the area of cultivation and possibly the meadow used for fodder. The outfield, in contrast, was the area where the animals grazed or where people gathered other resources such as wood and wild plants from wood- and wetlands (Hambro Mikkelsen & Nørbach 2003:129, Grabowski 2014:24).

When the fields where harvested and the grains collected, threshing was used to remove unwanted parts of the cereal from plant e.g. the awns, either by clubbing the straws with a mallet or a club on a hard surface (Hambro Mikkelsen & Nørbach 2003:138). The use of animals trampling the grains from the awns has also been documented. Finds of both threshed and unthreshed have been found suggesting that threshing could take place around the year (Viklund 1998).

The threshing was followed by sieving which aims to remove the cereal grains from the weed seeds and other plant parts. This could be achieved by three different methods, pouring, winnowing or flinging. Winnowing method is used by throwing the cereals up in the air in a place with wind or draught. The lighter particles would be flown away with the wind while the larger would fall into the basket or troughs. The larger unwanted particles could then be plucked away by hand. The pouring method is done in a similar matter as it also utilizes the wind and pouring the cereals containing chaffs, awns, weed seed and cereal grains from one container to another from a height. Flinging, or throwing, the material along a floor could also have been utilized as a method. Throwing the grains separates the lighter particles from the heavier using their own weight. The lighter particles such as awns, chaffs and smaller weed seeds land closer to the person, while the heavier cereal grains are thrown further (Viklund

(16)

13 1998:60-62). The use of sieves for cleaning the crops is also most definitely used (Hambro Mikkelsen & Nørbach 2003:138).

2.2 Iron Age ovens

The processing of vegetal matter has a variety of desired outcomes. Amongst the desired outcomes are to change the physical structure, making the food tastier, more stable and less toxic. The ovens used for drying the cereals can be considered to contain a few of these

outcomes i.e. a change physical structure and more stable when storing the grains (Metheny &

Beaudry 2015:411). Botanical remains in the ovens would most likely have been subjected to a range of different crop processing methods before (Hansson 1997:46).

Due to the moist climate of northern Europe, ways of drying the cereal grains were necessary to implement. If the air is too moist, the seeds are subjected to develop mold, or they can even start to germinate. Therefore, several ways of drying and roasting were common in historical times such as designated houses, bastu. The cereals could also be dried by suspending them in a pot or kettle above a hearth or rolling hot stones on the grains. Drying of hulled species in ovens was probably done as the grinding, and the subsequent, food making was both easier and tastier, and of course, for storing (Viklund 1998:90-92). Drying of germinated cereals has also taken place in order to make beer (Ross et al. 2016). There are not many sites found in Sweden that are interpreted as oven used for drying cereal grains or other similar structurers used for drying the crops dating to the Iron Age (Viklund 1998:93), therefore, the majority of ovens are composed of Danish material.

Ovens can have many different shapes as rectangular U-shaped, oval dome-shaped or keyhole-shaped are known from the northern and western Europe (Vanhanen & Mikkanen 2013. Ross et al. 2016). They can be located both inside and next to a settlement and sometimes situated close to the fields for easier and faster drying when harvesting the crops (Vanhanen & Mikkanen 2013). In the large chamber, often made out of clay and stones, the harvest is placed upon a suspended floor and then subsequently roasted (Ross et al. 2016).

There are ovens that contain several ceramic sherds and thus been interpreted as ceramic ovens (e.g. Høyem Andreasen 2016). These types of ovens will also be included as they can have been used for both pottery making and drying cereals. Other oven types that existed are iron extraction furnaces. These ovens are used, as the name implies, for extracting iron from bog ore. These will, however, not be investigated to a large extent but they often contain an

(17)

14 abundance of plants from cereals, such as hulled barley or rye, and weed species which were placed at the bottom of the extraction oven (Mikkelsen & Nørbach 2003:173).

2.3 Iron Age plant food

The prehistoric food culture encompassed not only the basic sustenance but also a social role.

It served as a social insurance and marker, symbol of power, but also formed a religious element (Hansson 1997). With potentially more than half the food coming from agricultural lands, not factoring in the gathered plants, agriculture played a huge role in daily life

(Pedersen & Widgren 2004). Although for other areas in Scandinavia, agriculture would not have been feasible in a similar manner as colder regions would have relied more heavily on meat from animals (Hansson 1994:5).

Amongst the most common type of food used from cereals is porridge, bread, and grain-paste (Zachrisson 2014:183). Grain-paste can sometimes be a dish by itself but can also be refined further into porridge. The difference between grain-paste and porridge is the latter is cooked.

The historical account of Pliny (23 – 79 AD) states that the Greek people used to soak their barley grains in water over one night, drying it over a fire and then grinding it mixed with flax seeds into a grain-paste (Hansson 1994:6). Grain-paste and porridge is hard to find in

archaeological remains, while the bread is more commonly found, as it is baked. For porridge and grain-paste to be preserved, they are almost required to have been found in a pot or similar vessel, while bread and beer can sometimes be found in graves (Hansson 1994:10).

The coffin of the Egtved girl from the Bronze Age contained a container which had had a mixture of beer and fruit wine. It was made from wheat along with cranberries or

lingonberries and honey (Glob 1970).

The earliest known bread in the Nordic countries is from around 200 AD, potentially

introduced by returning men from Roman auxiliary service. Without a doubt, a large number of new species was introduced to central and southern Europe by the Romans and

subsequently gradually introduced to the north (Rohde Sloth et al. 2015:33). The first types of bread were most likely eaten on special occasions, as the bread seems to have been associated with higher powers (Hansson 1997). Furthermore, the first find rotary querns are found together with carbonised bread in central places like Helgö and Uppåkra. It is only in the

(18)

15 Viking Age that rotary querns become a common practice, while the pre-existing grinding stone was common until the medieval period (Zachrisson 2014:181).

Most commons edible plant food was undoubtedly from cereals, though, finds of storages containing weed seeds have also been encountered. A large find of Chenopodium album found in a heap in a granary building along with separate heaps of oats and barley are known from Danish prehistory. The weed seeds would be most likely gathered in the fields and fallows to supplement the diet (Helbæk 1959b). In total, it has been proposed that Spergula arvensis, Bromus secalinus, Polygonum lapathifolium (nowadays known as Persicaria lapathifolia), Chenopodium album, Fallopia convolvulus, Rumex acetosella, and Glyceria fluitans have most likely played a substantial part of the human food intake of the plant foods (Behre 2008:65).

Garden plants, or horticulture, has been documented to have existed in the Viking Age, but may well have taken place further back in prehistory as well. These gardens placed just outside the house would include herbs, vegetables, spices and also medicinal plants. The use of medicinal plants is, however, convoluted as some plants may have acted as both food and medicine depending on the desired outcome (Karg & Robinson 2000:133). Plants like peas and beans are poorly preserved, as they are harvested before they set seeds, but has most likely been cultivated. Sowing of garden plants for the next year’s harvest would likely not have required as many seeds to be preserved as required for cereals (Rohde Sloth et al. 2015).

Without a doubt was other parts of the plants used as well for food besides the seeds. Leaves, roots, and bulbs, however, rarely gets preserved without exceptional preservation (Behre 2008:65).

Gathering of plants in the outfield was also an important task, e.g. sweet gale (Myrica gale) grows in bogs and on heathland and had to be gathered for use in brewing beer along with hops (Hansson 2005). Fruits, nuts, and berries such as wild strawberry (Fragaria vesca), elderflower (Sambucus nigra), various Rubus species along with hazel and acorns of oak consisted to a large degree of the traditionally gathered species that are not classed as a weed (Karg & Robinson 2002. Robinson 1994:545). As shown by both the Kayhausen bog body which contained apple seeds, millet, flax, Persicaria lapathifolia, Spergula arvensis and Chenopodium album in large quantities (Behre 2008:67) and the Neolithic bog body called

‘Raspberry girl’, which was found with large amounts of raspberry seeds in her stomach (Sjögren et al. 2017), the gathering of wild plants and weeds was common.

(19)

16 Bog bodies have often been under the discussion when considering sacrifice and offerings during the Iron Age. This too applies to the archaeobotanical aspects of the deceased (Palmer

& van der Veen 2002). Tacitus wrote in his book Germania that bog bodies are a form of penalty for desertion or punishment of homosexual acts (Tacitus 2016:25). As there are several countries and different periods from when the bog bodies belong to, from the Mesolithic to the Iron Age, it is difficult to group them all together. There have most likely been a difference in the social, political, and religious aspect throughout these periods (Turner 1999:227). Bog bodies from earlier periods than the Iron Age, however interesting, lay

beyond the scope of this thesis but can serve as a comparison, as it would be unjust for a comparison as there is a time difference of several centuries and millennia (Bennike 1990:29) The famous bog body, Tollund Man, was found in 1950 in Bjældskovdal and dated to 300 BC (Glob 1969). Tollund Man was found with a leather noose around his neck, which indicated that he had been strangled to death. He had worn a pointed leather cap on his head and a belt of oxhide around his hips, but in other regards, naked (Fischer 1999:94). Parts of Tollund man’s body had suffered from taphonomic processes but due to his well-preserved intestines were examined and indicated that his last meal consisted of vegetarian food that was ingested about 12-14 hours before death occurred. Tollund Mans last meal was porridge consisting of ground wheat and barley grains completed by flax seeds and the seeds of up to 30 different wild plants such as spurrey, hemp nettle, willow-herb, gold-of-pleasure and plantago (Fischer 1999:94). The meal is unable to provide any certain time of the year that Tollund man was sacrificed, but it has been suggested that he died in winter or early spring (Coles & Coles 1989:181). The porridge could have been boiled in bog water as indicated by moss fragments.

There was also a quite high abundance of sand found in his stomach.

Grauballe man was found in 1952, in a very well-preserved state, although some distortion to his face and parts of his inner organs were decayed a bit. His death is, however, different from Tollund man as Grauballe man had had his throat cut from ear to ear, along with a broken leg and a blow to the face. He is relatively contemporarily deposited into a bog, similar to

Tollund man. Tollund man and Grauballe man had eaten a similar meal, which contained cereals and over 50 wild species. The meal also indicated the time of death to winter or early summer, due to there being no summer or autumn fruits or berries (Fischer 1999:96. Coles &

Coles 1989: 181).

The Borre man has many of the characteristics of the previous bog bodies presented. He was found with a noose around his neck, indicating that he was either strangled or hanged, which

(20)

17 according to C¹⁴ date, took place around 700 BC. His last meal consisted of gruel with

knotweed and Spergula arvensis amongst other weed seeds (Fischer 1999). The English Lindow Man (Lindow II) was found in a bog in 1984 and believed to have met his end in the first century AD. He was found naked except for a fur band around his upper left arm. He also had a trimmed beard and hair and well-polished nails. Due to no finds of insects, it would appear that he was buried immediate after receiving two blows to the head and strangled until his neck broke (Turner 1999:229). His last meal consisted of a bread made of a mixture of emmer wheat, spelt, and barley. His stomach also contained mistletoe pollen and Sphagnum, indicating an unclean beverage. In the same bog, Lindow man III was found in a similar state.

His last meal consisted of crushed cereal grains and hazelnuts (Fischer 1999:96).

Graves offers a good context for ritual and symbolically interpreted food remains (Heiss &

Hansson 2014). These graves offerings are usually intended for use in the afterlife, either as whole meals or just as ingredients, representing a meal, drink or medicine (Karg et al. 2014).

Grave offerings are considered deliberate depositions and can indicate socioeconomics and belief system of past cultures. Food offerings related to death does not, however, necessary involved the deceased person but can also include feasting for the participants. Ibn Fadhlan, an Arabic traveller describes a tenth-century funeral feast of a band of Rus which included animal sacrifice and heavy drinking (Metheny & Beaudry 2015:345). Food and drinks appear to have been common in graves dated to the Viking Age in Birka, although perhaps a more special meal than the everyday meal (Hansson 1996:62). Graves does not necessarily always contain a meal aimed for the afterlife, but rather a symbolic deposition. The remarkable burial of Peder Windstrup from the 1600s contained a large number of hops, along with cereal grains of rye, barley and oat, and other pleasant-smelling plant parts in both his pillow and bedding material (Lagerås 2016).

Several cremation graves from the latter half of the Iron Age in Sweden have contained bread (Zachrisson 2014:185). Excavation at sites in central Sweden and Helgö, Ljunga and

Boberget have also resulted in numerous finds of breads (Hansson 2005). These breads have, however, only been preserved due to the charred state that they were in. Not only are cereals used for making bread throughout the prehistoric times, but also vetches, flax, gold of pleasure, field weeds and even peas, while in the Medieval Period, pure cereal bread is the most common types of breads (Hansson 2005). A study from the 1950s done by Åke

Campbell suggests that the contemporary bread dates back all the way to the Medieval Period.

Before that bread was consisted mostly as a thin and flatbread, baked over a fire (Viklund

(21)

18 1998:142). Hulled barley was the most common type used for making bread followed by oats, but not these were the only type of cereals. The bread contained often two or more types of cereals (Hansson 1996:71). It is only during the 16^th century that rye became the most commonly harvested grain in central Sweden. However, rye bread had been a common sight in southern Scandinavia, especially Denmark, since the Late Iron Age (Viklund 1998:147).

Carbonized bread found in graves can be hard to distinguish from porridge or grain-paste, but both are a good source of information into the mindset of prehistoric people, as both the content and technique of making the food are carefully chosen (Hansson 1997). Charred bread is often found in graves they are assumed to be an essential part of a ritualistic deposition in burials. Most of the charred botanical remains found in graves can be assumed to have been buried along with the body. The most common burial practise of Late Bronze Age to Late Iron Age was cremation burial (Viklund 1998:145,155), although the religious beliefs and practices may greatly have changed during these periods (Glob 1969:105). This may also include the practice and purpose of the bog bodies.

Excavation and soil analyses of graves from Birka have also resulted, not only bread but also finds of Crataegus calycina, Prunus spinosa, and Prunus instititia and hazelnuts (Arwidsson 1984:274). Arrhenatherum elatius var. bulbosum have also been recorded in graves in Sweden. Furthermore, there are finds of several types of different cereals in some graves, along with straw fragments. Sometimes even appearing to be in leather pouches (Viklund 1998:155). Peas and flax are rarely found in graves but are credited to have had magical powers. Grave depositions of flax and pluses are known from in burial rituals from 19^th century Sweden. Nuts, millet seeds, and other cereals being more common as depositions (Viklund 1998:159).

(22)

19

3. Material and method

The geographical setting is the south part of Scandinavia, i.e. Denmark, and the southern parts of Sweden. However, due to the low amount of accessible archaeobotanical material of

graves, a wider area was forced to be accounted for, such as the grave from Para in

Ångermanland, northern Sweden and cairns from a gravemound in east-central Sweden (see figure 3 for map). The time frame is mostly based on the Iron Age, with a few features dating both to earlier and later periods in order to show a more distinct diversification. For a full description of the features, the readers are referred to table 1 and 2 and the references there.

While the first research question is mostly a comparison between ovens throughout the prehistory and the third research question focusing on a potential similarity between ovens, graves and the two bog bodies. The second research question is foremost a literary research topic which aims to give a more holistic interpretation into the food culture of the Iron Age when combined with the other research questions.

Figure 3. Distribution map of southern Scandinavia, showing the analysed features (Map by Ola Lindgren)

(23)

20 The material is collected from Moesgaards museums department for conservation and natural science laboratory reports, the environmental laboratory at Umeå University (MAL), as well as SEAD database (Buckland & Eriksson 2018). These served as bases for the research.

Additional literature and the stomach content of the bog bodies were found in books, articles and other papers related to the subject. All samples and features were analysed by other researchers and staff personal at the various institutions, except two sites. Cobra Kablet and Tjæreborg were analysed by the author with standard procedure of stereomicroscope and reference literature as well as the supervision of Peter Mose Jensen at Moesgaard Museum.

As for the bog bodies, Tollund man and Grauballe man only have an estimation of the amount contained in their stomach as the seeds were too heavily fragmented for a clear distinction. In the appendix, 1) means that there are less than 5 seed, 2) represents 5-15 seeds and 3) over 15 seeds and 4) means that it is the majority of the sample. See table 1 and 2 for the

archaeobotanical remains of the analysed features that have more than 20 seeds (15 for the bog bodies) of a single species and the appendix for a complete table of all species from the features.

The dendrogram (see figure 10) was created in PAST3 using Paired group (UPGMA) and both the Bray-Curtis and Jaccard’s similarity coefficients. The dendrogram is a way of

presenting how datapoints may correlate or cluster. Bray-Curtis similarity coefficient is based upon the abundance of each taxon in the features. When a taxon is absent, the tool ignores that taxa and few or rare taxa adds very little to the coefficient (Krebs 2014) (see Hammer 2013 for calculation details). Jaccard’s similarity coefficient is somewhat similar, although it only factors in if a sample has a presence or an absence of taxa in the sample of feature (Krebs 2014).

The plants were grouped into their habitats (which the author will refer to as Ecocodes), according to botanical literature such as Mossberg & Stenberg (2010) and Nylén (2000), as well as based on Buckland’s (2007) thesis of developing a system of categorizing fossil insects, although adapted for an archaeobotanical usage. Cultivated species are composed of cereals species. Oil, fibre, and horticultural plants are made up of species that are cultivated but not cereals. This group contains plant such as Linum usitatissimum, Camelia sativa and Daucus carota. Gathered plants consist of species that has collected, but are not cereals, oil, fibre, or horticultural species, e.g., Corylus avellana, Rubus idaeus and Juniperus communis.

Meadow and grass species are more likely to grow in areas affected by grazing or similar conditions. This group includes Poaceae and Rumex acetosella, among others. The weeds and

(24)

21 ruderal species such as Chenopodium album and Persicaria maculosa, thrives in fields and anthropogenic soils (see appendix for the Ecocodes of the plants). There is, however, a potential overlapping area of the Ecocodes. Certain species can also exist in two or more groups; thus, these plants were duplicated and group into both Ecocodes.

(25)

22

(26)

23

(27)

24

4. Results

Figure 4. Agricultural species of the archaeobotanical remains in the ovens according to periods

Figure 4 shows the difference in cereals species throughout the periods investigated. Ovens (in both figure 4 and 5) that had dates spanning from one period to another was grouped into both periods as it is possible that they were used in a prolonged period and overlapping periods. The numbers inside and next to the bars indicate the frequency of species in the ovens. The number of ovens used for each staple in figure 4, 5, 7 and 8 is displayed underneath the bars.

The Late Bronze Age is dominated by barley (Hordeum sp.) with hulled barley (Hordeum vulgare var. vulgare) being the main species identified. The following period, Pre-Roman Iron Age, on the other hand, is dominated by naked barley (Hordeum vulgare var. nudum) although it contains several other kinds of cereals as well in smaller quantities. There is a noticeable change from naked barley (Hordeum vulgare var. nudum) to hulled barley (Hordeum vulgare var. vulgare) in the Germanic Iron Age. Oat (Avena sp.) is found in the Roman Iron Age and the two periods preceding and succeeding periods, as well as a large amount of oat in the Viking Age and Medieval Period. Rye (Secale cereale) is only found in samples from the Roman Iron Age and Germanic Iron Age, and then it has a clear presence.

(28)

25 Wheat (Triticum) is only represented a few times in the Pre-Roman Iron Age and then in the latter half of the Iron Age.

Figure 5. Agricultural species of the archaeobotanical remains in the ovens according to their dated period

Figure 5 was achieved by grouping the ovens according to periods in a more comprehensive scope than figure 4. Figure 4 and 5 shows similar results, though figure 5 show a more gradual change in agricultural and cereal crops. Late Bronze Age has a clear abundance of cereals only determined to genus, Hordeum, followed by Hordeum vulgare var. vulgare. The early part of the Iron Age contains the large amount of Hordeum vulgare var. nudum, and a change towards Hordeum vulgare var. vulgare can be seen from late Roman Iron Age. The late Roman Iron Age – early Germanic Iron Age contains more Hordeum vulgare var. vulgare than Hordeum vulgare var. nudum, the majority is, however, Secale cereale. The latter half of the Iron Age is dominated by hulled barley although Avena sp. and Avena sativa is

overshadowing the other species in the youngest oven.

(29)

26

Figure 6. Percentage of cereals found in each oven according to their dated period

Figure 6 shows the difference within the ovens in different periods. In periods with multiple ovens, there are some variations between the ovens as well. The ovens in Pre-Roman Iron Age contain large amounts of oats (Avena sp.) while CO.A4343 contains over 25 000 naked barley (Hordeum vulgare var. nudum) grains, thus it will factor into a huge degree in other graphs. However, figure 5 shows a similar trend with a change of naked barley to hulled barley, while also highlighting the various smaller quantities of other cereals.

(30)

27

Figure 7. Non-agricultural species of the archaeobotanical remains in the ovens according to their dated period

Figures 7 and 8 show the non-agricultural plant composition that occurred with a higher frequency than 50 in the ovens over time. Species that prefer the same kind of soils and belong to the same family (e.g Persicaria maculosa and Persicaria lapathifolia) were grouped together as to decrease the number of species that would otherwise be in the graph and to make it more comprehensible. Species that were undetermined but likely to be of a certain species (e.g. Chenopodium cf. album) were also grouped according to the species or family. This will make the graphs more easily understandable, although may be a bit skewed.

One needs to bear in mind the uneven distribution of the number of species in the ovens, as some contain a higher abundance of certain species. The Early Roman Iron Age staple e.g.

contains up towards 8000 seeds of Spergula arvensis while the following periods does not equal that amount altogether. The Late Bronze Age oven contained no data of non-cultivated species.

(31)

28 The sudden decrease of Spergula arvensis from the Early Roman Iron Age is very noticeable.

Persicaria sp. is consistent through most periods, although they do not appear in all periods.

Chenopodium sp. and Poaceae existed throughout all the periods as well. Other species appear more seldom throughout the periods but can nonetheless be found in quite large quantities.

Figure 8. Non-agricultural species of the archaeobotanical remains in the ovens according to periods

Figure 8 is similar to figure 7, although the ovens are grouped according to the chronological order, they occur in. Ovens overlapping between periods has been grouped, similar to figure 4, 5, 7 and 8 was done. The first two staples that contain non-cultivated species have a much larger quantity of seed than the rest of the staples combined. Similar to figure 7, Spergula arvensis is in a drastic decrease in frequency after the Roman Iron Age. Other species have roughly the same frequencies throughout the periods, although the latter periods contain much

(32)

29 fewer seeds in total. This is due to the fact that there are more ovens in the Pre-Roman Iron Age analysed.

The figures presented above shows that cereals in the ovens have all been cultivated and eaten to a large degree during certain periods. Weed seeds have been consumed in quantities as well and non-agricultural remains with more than 20 seeds in ovens and graves, 15 for bog bodies, are Chenopodium album, Fallopia convolvulus, Persicaria maculosa/lapathifolia, Plantago lanceolata, Poacea, Polygonum aviculare and Spergula arvensis. Other non-agricultural seeds that may have been consumed have lower frequencies but are also present in all three feature types.

Figure 9. Archaeobotanical remains of the cereals found in graves, grouped according to the graves dated period

Figure 9 should only be taken as a very broad indication of the cereals deposited in graves according to their dates. As there are so few graves and very little remains in all but one grave, this should not be stretched further. There are, however, a similar indication as to then ovens. i.e. a trend of naked barley in the beginning, which changes towards hulled barley.

Worth noting is that all the graves are very close in time and should therefore not be seen as definitive evidence. Bread wheat Tritium aestivum is furthermore considered to be a luxury compared to hulled barley and could be a sign of a high-status burial gift (Hansson 2005).

0

116 2

15 18

1

196 2

1 12

134

9

1 1

1

0 % 20 % 40 % 60 % 80 % 100 %

CG.1063 CG.43 CG.231 CG.A2:2 CG.15248 CG.A1:1 CG.A1:2

LRIA - eGIA eGIA eGIA GIA eGIA - VIK lGIA VIK

ARCHAEOBOTANICAL REMAINS OF CEREALS IN GRAVES

Avena sp. Hordeum vulgare var. nudum Hordeum vulgare var. vulgare

Hordeum vulgare sp. Secale cereale Triticum aestivum

Triticum sp.

(33)

30

Figure 10. Cluster analysis (A) of all ovens, graves and bog bodies using the Jaccard similarity coefficient. The analysis uses presence or absence of the taxa as comparison. Cluster analysis (B) of ovens, graves and bog bodies using the Bray-Curtis similarity coefficient. The analysis uses the abundance of the taxon in the features. Highlighted features are discussed in

more depth (analysis undertaken with PAST3 (Hammer & Harper 2001))

Figure 10 presents two correlation dendrograms created in the PAST3 program (Hammer &

Harper 2001). (A) is Jaccard’s similarity coefficient which focuses on the presence and

(34)

31 absence of the species in all of the features, while Bray-Curtis (B) analysis relies on the raw data of the features. In both dendrograms, the closer to 1 on the left scale, the closer two features correlate with each other. This means that they have similar contents in e.g. an oven and a grave. The use of Jaccard and presence/absence of the features is a solution to overcome the fact that ovens can have been used many times, while graves and the stomach content of the two bog bodies are most likely just a single occurring event.

In the Jaccard dendrogram, none of the features correlate with each other higher up than 0.6.

However, the two features that correlate the most with each other are CO.12357 and

CG.15248, followed by CO.1030, CO.125, and CO.2743. The ovens CO.A4343, CO.17,190, and CG. 231 are also features that are more closely correlated with each other. In (A) there are graves that are more closely correlated with other graves than with ovens and bog bodies, although there are graves (CG.231 and CG.15248) that correlate with ovens. The two bog bodies also correlate the most with each other. CG.1063 is the feature that is the least similar to any other features.

In Bray-Curtis similarity correlation dendrogram, three features, CO.2743, CO.A1460 and CG.231, followed by CO.A1968 are very similar to each other, contrasting to (A)

dendrogram. CO.POKS and CO.KS are fairly similar to each other. Tollund man and Grauballe man are also closely correlated with each other as in (A). The graves and bog bodies are not correlating with each other, other than previous mentioned CG.231.

(35)

32

Figure 11. Ecocodes of the archaeobotanical material found in the ovens, graves and the two bog bodies.

Figure 11 presents the difference in ovens, graves and the two bog bodies of Tollund man and Grauballe man, with both presence/absence and their total data. The majority in all the bars are non-cultivated plants, with bog bodies having the least in both the presence/absence and in their total frequencies. The weed and ruderal plants have the largest presence of all features, but meadow and grass plants retain a higher presence in the presence/absence staples.

Wetland plants have relatively low frequencies in all features, although more apparent in presence/absence.

Roughly 40% of the total composition in the ovens are of cultivated species. A large amount of oil, fibre & horticulture in the ovens are from large finds of Camelia sativa, and thus, smaller in the presence/absence. The wetland plants are also more representative in the presence/absence than in the oven than in the raw data staple.

The graves have almost 20% of cultivated species in both presence/absence and in the raw data. Gathered plants (5%) and meadow and grass plants (20%) are more representative in the presence/absence while in the raw data the weeds and ruderal plants are in clear majority. The wetland plants are also more represented in the presence/absence bar.

The two bog bodies have the smallest amount of cultivated species in their composition. They have no gathered plant species, but large amounts of weeds and ruderal plants and roughly 5%

0 % 10 % 20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 % 100 %

Oven frequencies Grave frequencies Bog bodies frequencies Oven (Presence/absence) Grave (Presence/absence) Bog bodies (Presence/absence)

Ecocodes of the plants found in the features

Cultivated plants Gathered plants Meadow & Grass plant Oil, Fiber & Horticulture Weeds & Ruderal plants Wetland plants

(36)

33 of oil, fibre and horticulture plants. The wetland plants are less than 10%, while meadow and grass plants are 15-20%.

Figure 12. Bar chart of the botanical remains of the features

Figure 12 shows the abundance of the different botanical remains found in all the features.

There is a clear representation of Hordeum vulgare var. nudum. The oven CO.2349 from the Late Bronze Age contains the more than 25 000 grains and, thus, is in clear majority. The Spergula arvensis consists of up towards 15 000 seeds, followed by Camelina sativa.

(37)

34

5. Discussion

5. 1 How do the archaeobotanical remains in ovens used for drying grains change through the Iron Age and how do they compare to the agricultural landscape of south Scandinavia?

It is difficult to answer how and why changes in agriculture took place during the prehistory.

However, when and where it took place is easier to differentiate as we have a solid

background of archaeobotanical data (Grabowski 2014:17). Different views of as to why have been debated and a single cause of event have not been able to justify the change of

agricultural practices. There have been many discussions on whether a climatic deterioration was the sole cause or other natural and anthropogenic factors was in play (Hjelmqvist 1992, de Jong & Lagerås 2011).

The archaeobotanical remains from ovens have most likely gone through several stages of threshing, cleaning, and sorting before being dried in an oven, all of which would have an impact on the recovered material (Viklund 1998:60).

During the Neolithic to Late Bronze Age, hulled barley was cultivated in smaller scale in parts of Denmark, although in general naked barley was the primary crop (Robinson 2003:145). The earlier cultivated grains such as emmer, spelt, and naked barley slowly succumbed to the cultivation of hulled barley (Grabowski 2013). Finds of hulled barley in Scania and in Zealand, however, shows that hulled barley was also an important crop in certain regions (Grabowski 2013:168). The oven CO.2349 from the Late Bronze Age was located in Zealand and contained a substantial part of hulled barley with an even greater amount of barley, Hordeum sp., thus, further suggesting to regional cultivation of hulled barley in the Late Bronze Age. This feature containing hulled barley could also indicate a specialised purposed. Some have argued that a switch to hulled barley can be seen already in the Late Bronze Age (Henriksen 2003), which the earliest oven could potentially indicate as seen in figure 4 and 5. The Late Bronze Age oven CO.2349 is most likely a drying of cereal event that accidentally became carbonized through fire as it has an incredibly high amount of barley compared to any other feature and plant (see figure 6 and 12). Furthermore, the ovens contained no records of weed seeds, though this may be a question of where the samples were taken from and of how it the oven was analysed.

(38)

35 From Pre-Roman Iron age until the Roman Iron, naked barley is the most common crop. The naked barley in the oven samples, in figure 5 and 6, occurs all the way until the Early Roman Iron Age, although only as very few seeds. Late Roman Iron Age – Early Germanic Iron Age ovens also have two seeds of naked barley but contains more seeds of hulled barley. Grains of Hordeum sp. have a large presence though it is dangerous to draw species conclusions of these seeds as they are too fragmented to distinguish to either species. Arguably the large amount of Hordeum sp. in the Late Pre-Roman Iron Age – Early Roman Iron age could be of naked barley though they were unable to be more interpreted that to the species. It is worth noting that some of the periods have only one oven or they contain very few finds of cereals.

The transition from naked barley to hulled barley was most likely not a question of

availability of grains, as the change of cereal was also not being implement at the same time all over south Scandinavia and it was cultivated in earlier times in certain regions. During the early Roman warm period in the Roman Iron Age, parts of Denmark still cultivated naked barley, while other parts had switched to hulled barley (Grabowski 2014:17). One reason for switching from naked barley might be that it is more susceptible to a parasite and insect infestation compared to hulled barley, so in more humid climate this may have had a play a part of the change (Buxo I Capdevila et al. 1997).

Arguments concerning different processing and harvesting techniques could have brought upon a switch from naked to hulled barley has also been suggested. The fact that hulled barley grains tend to be more steadily attached to the spikelets, and thus may have given a more flexible time of harvest is also a possibility of change (Mikkelsen & Nørbach 2003). The start of manuring may also be a reason for the gradual change from naked to hulled barley as has been proposed by Engelmark (1998). The use of manuring is argued to be more beneficial for hulled barley than naked barley. Together with the iron sickles introduced, the straws of the cereal enabled the gathering of animal fodder and further benefitted the manuring capability (Viklund 1998). The increase in certain weed species, that prefers soils rich in nitrogen, is in accordance with such a hypothesis. Although huge quantities of certain weed species may indicate gathering, or even cultivation, of weed species, may have taken place instead of manuring (Grabowski 2014:18). Other researchers have argued that manure was not a

decisive change of cultivation of hulled barley instead of naked barley. Nor that manuring was a new agriculture technique, but rather had existed earlier, albeit in perhaps a smaller scale (Lagerås & Regnell 1999).

(39)

36 Explanations that a different type of food is more easily accessible and used have also been used as a hypothesis for switching cereals in the late Bronze and early Iron Age. Skoglund (1999) argues that perhaps a crushed type of grain paste was eaten rather than a bread-based cuisine. He further argues that crushing, instead of grinding, the grains would symbolize and reflect the treatment of a deceased person. Although the introduction of the rotary querns would potentially over time switch to bread-based cuisine around 200 AD (Zachrisson 2014).

Hulled barley has also been discussed in relation with an increase in beer consumption (Grabowski 2014:20).

Secale cereale, rye, is a plant that starts to become more and more common during the Iron Age in southern Scandinavia. It is, however, in the latter half of the Iron Age that rye becomes a major crop (Grabowski 2014:22). As rye is also able to grow as weed species amongst other cereals, it is hard to prove an exact time of cultivation (Behre 1992). Early finds of rye, Secale cereale, are sometimes ascribed to be of an intruding species in the fields. The few finds in the Late Pre-Roman Iron Age – Early Roman Iron Age can potentially be seen as a weed species as it is only two grains found. It is during the Late Roman Iron Age - Early Germanic Iron Age that rye becomes a more dominant species both in the analysed oven CO.348 and according to other studies (Grabowski 2014:49). Archaeobotanical analysis of iron extraction furnaces have shown that rye was potentially cultivated already in the Roman Iron Age (Hambro Mikkelsen & Nørbach 2003). The majority of the cultivated species found in iron- extraction furnaces are rye and hulled barley. Oven CO.348 contains 118 grains of rye and 131 rachis fragments of rye. This would suggest that an intended cultivation of rye took place.

Therefore, the rye in the earlier phases that have low quantities can most likely be considered a weed species, albeit a not harmful but tolerated weed, as noted by Behre (1992:149). There are interestingly no finds of rye in the latter half of the Iron Age, as by then it was the major crop along with hulled barley (Robinson 1994). Although rye is argued to have been one of two major crops in the latter half of the Iron Age, rye is only found in one oven in large

quantities. This could be because there are too few ovens found from these periods, or perhaps rye was being dried in a different manner.

The fact that rye is possible to cultivate on poorer soils may be one of the reasons for the start of rye cultivation (Grabowski 2014:23). As rye is thought to have been sown during the autumn, while barley would have been sown during spring are signs of a rotating agriculture taking place instead of a settlement slowly migrating (Grabowski 2014). There were no finds

Making a feast for the deceased