Human Ecology

, Volume 39, Issue 3, pp 351–360

Never Mind the Bottle. Archaeobotanical Evidence of Beer-brewing in Mediterranean France and the Consumption of Alcoholic Beverages During the 5th Century BC

Authors

    • CNRS-Centre de Bio-Archéologie et d’Ecologie (CBAE)-UMR 5059-Institut de Botanique
  • Philippe Boissinot
    • EHESS-Travaux et Recherches Archéologiques sur les Cultures, les Espaces et les Sociétés (TRACES)-UMR 5608
  • Philippe Marinval
    • CNRS-Archéologie des Sociétés Méditerranéennes (ASM)-UMR 5140
Article

DOI: 10.1007/s10745-011-9395-x

Cite this article as:
Bouby, L., Boissinot, P. & Marinval, P. Hum Ecol (2011) 39: 351. doi:10.1007/s10745-011-9395-x

Abstract

This article reports on an example of early archaeobotanical evidence for beer-making in Iron Age South-Eastern France. An archaeological sample from a fifth century BC house at the site of Roquepertuse produced a concentration of carbonized barley (Hordeum vulgare) grains. The sample was taken from the floor of the dwelling, close to a hearth and an oven. The barley grains are predominantly sprouted and we argue that the assemblage represents the remains of deliberate malting. Malt was most likely related to beer-brewing. The neighboring oven could have been used to stop the germination process at the desired level by drying or roasting the grain. Beer-making evidence in Roquepertuse is discussed in the context of the consumption of alcoholic beverages in the Iron Age Western Mediterranean using archaeological and historical data.

Keywords

BeerArchaeobotanyIron AgeMediterraneanSocial consumption of alcoholic beverages

Introduction

The last millennium BC was a very dynamic period in the Mediterranean. Eastern civilizations, especially Greeks and Phoenicians, founded many coastal colonies across the western Mediterranean, maintaining strong cultural and economic relations with indigenous societies (see Dietler 1997, 2007; Hodos 2006). In the French Mediterranean area of the Etang de Berre, Iron Age communities had intense economic relations with the neighboring Phocean Greek colony of Massalia (modern Marseille), a city founded around 600 BC. Numerous wine amphorae and Greek ceramics retrieved from indigenous settlements provided strong evidence of imports from Massalia. The indigenous counterparts to this exchange are poorly known, but it is often hypothesized that agricultural resources, and above all cereals, could have played a large part in this trade (Garcia 1987, 1999; Py 1993; Dietler 1997). New Mediterranean products, especially wine, could have had a significant cultural and economic influence on indigenous societies (Dietler 1990, 2007), but it is actually very difficult to assess exactly what this might have meant for local agriculture and alimentation.

Carbonized fruits and seeds preserved in archaeological sediments are essential sources of information concerning economic plants, diet and agriculture of past human populations (see Renfrew 1973; Zohary and Hopf 2000). A major difficulty lies in the identification of the ways in which specific plants were used. We know, for example, that six-row barley was the main cultivated plant of the Iron Age. The reasons for this have been the subject of a number of hypotheses (e.g., Jones 1981; Marinval 1988), but we lack direct evidence about its uses. In this article, we discuss the first archaeobotanical results obtained from the well-known site of Roquepertuse, which suggest that six-row hulled barley was employed to make beer.

Iron Age Economic Plants

In the Provence region of southeastern France archaeobotanical studies have been carried out, especially on hillfort settlements excavated during the 1970s and 1980s located in the vicinity of the Etang de Berre (Marinval 1988; Bouby and Marinval 2000). These results indicate that local agriculture in the Late Iron Age (ca. 450–25 BC) was based on hulled six-row barley (Hordeum vulgare) and naked wheat (Triticum aestivum/turgidum). Cereals also included broomcorn millet (Panicum miliaceum), Italian millet (Setaria italica) and emmer (Triticum dicoccum). Pulses seem to have played an important role along with cereals (Marinval 1988). Bitter vetch (Vicia ervilia), lentil (Lens culinaris) and grass pea (Lathyrus cicera/sativus) are predominant in storage places. Two oil plants, gold-of-pleasure (Camelina sativa) and flax (Linum usitatissimum), were probably also cultivated.

This range of cultivated plants is largely similar to those identified at the same time throughout the French Mediterranean area, especially in the Languedoc region, on the opposite shore of the Rhône River (Marinval 1988; Buxo I Capdevila 1996; Bouby and Marinval 2000; Py and Buxo I Capdevila 2001; Alonso et al.2007). This agricultural pattern largely developed from a more ancient Early Iron Age (ca. 750–450 BC) and Late Bronze Age (ca. 1,400–750 BC) tradition that focused on hulled six-row barley. The most striking innovation concerns the diffusion of viticulture, at least from the beginning of the fifth century BC onwards. This can be inferred from the increasing frequency of grape pips and from the presence of the domesticated morphotype at most sites, particularly in the Etang de Berre area (Marinval 1988; Bouby and Marinval 2001; Bouby et al.2006). Relations with Phoceans from Massalia and with other Mediterranean agents probably had an influence on the rapidity of the spread of cultivated grapes, but little is known concerning the other species and how they were culturally employed.

Site and Sampling

Roquepertuse (Velaux) has been a well-known archaeological site since the nineteenth century and the discovery of stone statues of two cross-legged warriors. It is located on the southwestern slope of a rocky hill dominating the valley of the Arc River, some 15 km west of the modern city of Aix-en-Provence (Fig. 1), and 7 km east of the Etang de Berre. New excavations carried out during the 1990s (Boissinot et al.2000) largely rejected the earlier “sanctuary hypothesis” and interpreted the prestigious art and symbolic elements within the context of the dynamics of a small hill-fort village over a period lasting from the fifth to the early second century BC (Fig. 2).
https://static-content.springer.com/image/art%3A10.1007%2Fs10745-011-9395-x/MediaObjects/10745_2011_9395_Fig1_HTML.gif
Fig. 1

Location map of Roquepertuse (Velaux, Bouches-du-Rhône, France) and of the main sites mentioned in the text

https://static-content.springer.com/image/art%3A10.1007%2Fs10745-011-9395-x/MediaObjects/10745_2011_9395_Fig2_HTML.gif
Fig. 2

Chronological stages of the small agglomeration of Roquepertuse (after Boissinot et al. 2000, simplified)

The Iron Age occupation began during the second half of the fifth century BC with the erection of two partially rounded structures made of wood and clay. The presence of hearths, of an oven and of various large and small storage vessels show that both buildings were used as human habitations. They were destroyed by fire before the end of the fifth century.

The site underwent major changes around 300 BC, the most obvious being the construction of a fortification wall made of stone, encircling a 1,300 m² surface around the plateau. During the third century BC, the fortification wall was made thicker and more monumental. Human habitations spread outside the wall along the access path in the small valley. It is probably at this time that the carved cross-legged warriors and the stone portico retrieved from ancient excavations were erected in the village (Boissinot 2004). Real or sculpted human heads were exhibited in niches carved in the pillars of the portico, in the Celtic tradition. Portico and sculpted warriors could have constituted a kind of gallery of heroes. This village was the object of a military assault during the second half of the third century BC, when it was destroyed by fire.

By the end of that century, occupation was much more limited and the fortification would have lost most of its prestige and defensive value, becoming an isolated rural dwelling-place with substantial storage activities and probably wine production. Beyond the presence of grape pips, various archaeological features point to local wine production: a plantation pit was found close to the entry of the site, along with a cellar with various storage jars (dolia), clay vats that could have been used to tread grapes, and a pruning knife (Boissinot et al. in prep). This settlement was in turn violently destroyed about the end of the third or the early second century BC, as were other settlements in the area.

Excavations during the 1990s sampled sediment from three archaeological contexts where macrobotanical remains were visible to the naked eye. These were water-sieved through 2 mm and 0.5 mm meshes. One sample comes from the floor of one of the fifth century BC houses, close to a hearth and an oven. The other two samples were taken from the contents of a ceramic vessel and from a pit. Both were located in the vicinity of two different storage areas.

Botanical Results

The density of carbonized plant remains is high in the three samples (Table 1). A total of 7780 plant remains have been identified, providing evidence for 19 taxa. In general, hulled six-row barley is the most common species and is well represented in all samples. Both dense-eared and lax-eared six-row barley are in evidence. Naked wheat and millets—broomcorn and Italian millet—are well represented, the former in sample B and the latter in sample C, with both samples dated to the last period of occupation of the site. A few seeds of pulses (grass pea and bitter vetch) and, probably, of flax, an oil and textile plant, were recovered. Fruits are represented by hazelnut and grape remains consisting of various pips and one single pedicel. Grape is mentioned in all samples. Morphometric analysis of the pips (see method in Bouby et al.2006) allows identification of at least seven specimens as cultivated morphotype and three as wild morphotype. The cultivated subspecies is already present during the fifth century BC (sample A). Wild herbaceous plants are mostly segetal and ruderal weeds and are only reported in samples B and C (ca. 225–175 BC).
Table 1

Archaeobotanical results from Roquepertuse

Sample

A

B

C

Total

Origin

House floor

Vessel

Pit

Age (years BC)

450–400

225–175

225–175

Volume of sediment (l.)

ca. 5

3

10

ca. 18

Cereals - grain

Cerealia

fg.

570

600

111

1281

Hordeum vulgare (hulled)

ungerminated caryopsis

10

319

199

528

germinated caryopsis

93

-

-

93

caryopsis (germination undetermined)

134

-

-

134

fg.

2140

838

620

3598

Panicum miliaceum

caryopsis

-

-

464

464

 

fg.

-

-

40

40

Setaria italica

caryopsis

-

-

802

802

 

fg.

-

-

62

62

Triticum estivum/turgidum

caryopsis

3

238

5

246

underdeveloped caryopsis

-

3

-

3

fg.

4

216

1

221

Triticum sp.

caryopsis

2

-

1

3

fg.

16

-

-

16

Cereals - chaff

Cerealia

fg. glume

9

-

-

9

Hordeum vulgare (hulled)

lemma dense-eared

11

-

-

11

lemma lax-eared

-

15

2

17

Oil plants, pulses

Linum cf. usitatissimum

seed

-

-

1

1

Lathyrus cicera/sativus

seed

-

-

3

3

fg.

-

-

4

4

Vicia ervilia

seed

3

4

7

14

cotyl.

-

-

3

3

fg.

-

2

1

3

Fruits trees/Hedges, woodland margins

Corylus avellana

fg.

1

-

-

1

Vitis vinifera

seed

12

3

-

15

fg.

5

10

8

23

pedicel

-

-

1

1

Segetal weeds

Asperula arvensis

seed

-

-

1

1

Bupleurum lancifolium/rotundifolium

seed

-

-

3

3

fg.

-

-

1

1

Lolium temulentum

caryopsis

-

7

5

12

fg.

-

1

1

2

Vaccaria hispanica

seed

-

-

9

9

fg.

-

-

1

1

Ruderal weeds

Anagallis arvensis

seed

-

1

-

1

Echinochloa crus-galli

seed

-

-

32

32

Wet places

Scirpus lacustris

seed

-

1

-

1

Other

Avena sp.

fg.

-

4

-

4

fg. awn

-

1

2

3

Fabaceae

fg.

2

1

2

5

Malva sp.

seed

-

-

1

1

Panicum/Setaria

caryopsis

-

-

18

18

fg.

-

-

87

87

Poaceae

fg.

-

1

-

1

Silene sp.

seed

-

1

-

1

Undetermined

seed

-

1

-

1

Total (number of remains)

3015

2267

2498

7780

Nb of remains/litre

ca. 600

755.67

249.80

-

Fragmentation rate (%)

91.48

74.50

37.87

69.32

The fifth century BC sample is strongly dominated by barley (Fig. 3). In spite of favorable conditions for the preservation of seeds in the sediment, which was rich in ash, charcoal and generally carbonized plant material, the grains are corroded and highly fragmented. This fairly poor preservation seems rather specific to the barley grains from this sample. Conditions are therefore not suitable for a satisfactory observation of all the morphological features of most of the grains. However, the best-preserved specimens allow the assessment that slightly more than 90% of the grains were sprouted when carbonized (Table 1, Fig. 3) and this is easily visible from the groove-like channel imprinted on the dorsal side of the grains (Fig. 4). As far as we can see, sprout length is not perfectly uniform, but it seems generally to reach half or two thirds of grain length.
https://static-content.springer.com/image/art%3A10.1007%2Fs10745-011-9395-x/MediaObjects/10745_2011_9395_Fig3_HTML.gif
Fig. 3

Composition of sample A from the floor of a 5th c. BC house. a, Overall relative abundance of taxa represented by seeds and expressed as Minimal Number of Individuals (MNI = Nb of seeds + 1/2 Nb of fg.). b, Proportion of sprouted six-row barley grains

https://static-content.springer.com/image/art%3A10.1007%2Fs10745-011-9395-x/MediaObjects/10745_2011_9395_Fig4_HTML.gif
Fig. 4

Germinated carbonized six-row barley grains (Hordeum vulgare) from sample A

The poor preservation of barley grains is probably largely caused by germination, which tended to render them brittle. Fragmentation occurred mostly after carbonization and is therefore not due to Iron Age human practices. The bad preservation of germinated carbonized grains has been confirmed experimentally (Stika 1996). Even if they are weakly represented, the presence of hulled caryopses and glume fragments must be regarded as evidence that grains were not dehusked before carbonization. We might note that chaff is actually destroyed faster than grains during carbonization (Boardman and Jones 1990). Charring is certainly also responsible for the disappearance of coleoptiles, coleorhizae and roots from our sample.

Discussion

The Remains of Deliberate Malting

Germination of cereal grains can potentially be either accidental or deliberate in origin (van Zeist 1991; Stika 1996). Accidental germination can happen in the field or during storage, in both cases because of overly damp conditions. On the other hand, germination can be actively induced by steeping grains in water in order to promote the development and activity of an enzyme that transforms the starch of the grain into fermentable and more easily digestible sugars. Malt (the sprouted grain) has various alimentary uses (Maurizio 1927): it can be eaten as is or used to make bread, but its most common use is certainly related to the production of fermented beverages. Beer results from alcoholic fermentation. Other kinds of fermentations can also occur, especially lactic fermentation, which is generally regarded as responsible for the spoilage of the initial product but can sometimes be promoted to produce acidic beverages or soups like the Russian kvas (Sigaut 1997) or the African kaffir beer made of malted millet or sorghum (Hornsey 2003). These acidic beverages generally also contain at least a small amount of alcohol (Maurizio 1927). Evidence from the site at Hochdorf suggests that, due to the slowness of the drying process, the activity of lactic acid bacteria would be high, both helping preservation due to acidity but also giving a sour taste to the beer (Stika 2011).

Various reasons lead us to argue that sample A from Roquepertuse represents the remains of malt and therefore a probable evidence of beer-brewing. Due to its high density (more than 500 plant remains/l.) and to its purity (barley grains make up more than 97% of the sample), it can be characterized as a close find assemblage or concentration (see Rösch et al.1992), which implies a single product, carbonized during storage or during its transformation process. Caryopses were still hulled but totally free of weed seeds which points to a careful processing of the grain. This does not fit with the hypothesis of a spoiled crop. Even more conclusive is the very high proportion of germinated grains. It cannot reasonably be argued that a germination rate above 90% could be achieved accidentally. Van der Veen (1989) assumes that above 15% of germinated grains, one can probably rule out accidental sprouting. In fact, mature and well-developed caryopses germinated at a rate above 75% are regarded as a reliable evidence of malt. According to the traditional literature dedicated to beer-brewing (Laubenheimer et al.2003), as a prerequisite to this, barley grains are carefully cleaned of every sort of contaminant. Then, ideally, 95% of the grains should germinate to provide good starch degradation. This description closely fits the situation displayed by the Roquepertuse sample. In the classical process, however, germination should be uniform and stopped when the sprout reaches two thirds of caryopsis length. However, it seems likely that it would have been difficult to obtain such an even germination before the establishment of rigorous methods and processes of industrial brewing (van der Veen 1989). Moreover, it has been demonstrated experimentally that when hulled barley is malted, the length of the dorsal groove is constrained for a large part by the tightening glumes. The malting level of the grains is reflected only to a minor extent in their morphology (Stika 1996). We can therefore estimate that barley grains were carbonized at the end of the malting process and before grinding of the dried malt.

Beer Brewing: A Domestic Activity?

There may thus be sound evidence that the barley deposit from sample A represents remains of malt and therefore probably of beer brewing. The building where this evidence was found displays all the features of regional Iron Age human habitations. We consequently consider that, in this case, beer brewing could have occurred as an ordinary domestic activity. Beer-making does not require specialized structures or implements (Arthur 2003; Laubenheimer et al.2003). All that is needed is an amount of grain, some water, containers (commonly pottery vessels) in which to soak the grain, a flat paved area—possibly the floor—to spread out and turn the grain during germination, an oven to dry it in order to stop germination, domestic grindstones to grind the malted grain, hearths and again containers for fermentation and storage. Evidence for all of these sorts equipment is reported from the Roquepertuse dwelling.

Of special interest is the oven discovered less than 2 m from the barley concentration. In brewing activity, an oven is often employed to stop germination at the desired level by drying or roasting malted grain (van der Veen 1989; Laubenheimer et al.2003). The Roquepertuse oven (Fig. 5) would have been especially suitable for this operation. It is of a quite elaborate kind, typical of the French Mediterranean area during the Iron Age and particularly well-known in settlements from the Etang de Berre area (Py 1993; Chausserie-Laprée 2005). Such ovens are composed of three parts (Fig. 6): embers are contained in the lowest part while food can be cooked in the upper part by hot air circulating through the middle chamber. This would allow gentle air-drying which is perfectly suited to stop the germination process, as malt should not be overheated to avoid damaging the chemical properties of enzymes (Laubenheimer et al.2003). Archaeologists still regard the specific functions of these typical ovens as unknown and they could have been multi-functional. A particular function of the drying or roasting of cereals has often been assumed (Py 1993; Chausserie-Laprée 2005). According to the Roquepertuse evidence, the drying of malt may well have been one of their specific functions.
https://static-content.springer.com/image/art%3A10.1007%2Fs10745-011-9395-x/MediaObjects/10745_2011_9395_Fig5_HTML.gif
Fig. 5

Fragments of the oven retrieved in Roquepertuse, close to the concentration of carbonized barley grains

https://static-content.springer.com/image/art%3A10.1007%2Fs10745-011-9395-x/MediaObjects/10745_2011_9395_Fig6_HTML.gif
Fig. 6

Schematic representation of the type of elaborated oven identified in Roquepertuse (approximate height 1 m) (after Domallain/Chausserie-Laprée 2005)

In the case of Roquepertuse, it is not possible to state whether carbonization results from the house being destroyed by fire or whether carbonization was accidentally caused during beer production, which involves heating at various steps of the sequence, especially to stop germination, as we have seen.

Beer and Wine in Provençal Celtic Society

The consumption of beer among Celtic communities of Western Europe is well known from classical Greco-Roman written sources (André 1981; Garnsey 1999; Laubenheimer et al.2003). For example, according to Pliny (N.H., XIV, 29) in the first century AD, a wide variety of beers was brewed in Gaul and Spain, and some of the Spanish ones could be kept for several years.

However, most of these texts are quite late. We do not know of any mention earlier than the first century BC. Archaeobotanical evidence is therefore of great interest. Nevertheless, convincing evidence from germinated grain assemblages is not common and most dates to Roman or later times (van der Veen 1989; van Zeist 1991; Stika 1996). Except for Roquepertuse, the most ancient evidence seems to come from the early Celtic site of Hochdorf in modern-day Germany (ca. 600–400 BC; Stika 1996, 2011). In France, assemblages of barley and millet grains contained in pots retrieved from the Late Bronze Age cave (ca. 1,200–900 BC) of Planches-Près-Arbois (Jura; Pétrequin et al.1985) and from the Early Iron Age (ca. 750–700 BC) tumulus of Saint-Romain-de-Jalionas (Isère; Verger and Guillaumet 1988) have been regarded as possible evidence of beer-brewing. But in those cases, since the grain is not sprouted, various other equally plausible explanations could be put forward. Brewing of beer has been suggested from earlier Spanish sites relying on other kinds of evidence: chemical analyses of Bell Beaker ceramics from various funeral settlements (Fabregas Valcarce 2001; Bueno Ramírez et al.2005; Rojo Guerra 2006) and starch grains analysis in the Late Bronze Age site of Genó (ca. 1,000–800 BC; Juan Tresseras 1998). Archaeobotanical records of plants traditionally used to flavor beer, especially sweet gale (Myrica gale) and hop (Humulus lupulus), strongly increase during medieval times (ca. 500–1,500 AD) in northwestern Europe, but some evidence of sweet gale in Iron Age settlements from the Rhine estuary area could point to early use of this additive and indirectly to beer-brewing (Behre 1992).

Diodorus of Sicily (Book V, 26) clearly mentions the making of a barley beer in Gaul, which he calls zythos. However, beer would not have been made solely from barley but also from at least wheat (e.g., André 1981). Each kind of beer would have had its own use and value, in addition to the consumption of wine. This is reflected in the writings of Posidonius of Apameia (135–51 BC) (Athenaeus, Deipnosophists, 4, 152 d in Laubenheimer et al.2003). Rich people would have drunk wine from Italy and from Massalia. Moreover, they would have drunk it pure (which, in the eyes of the Greeks, was the sign of absolute barbarism) or lightly mixed with water. Less wealthy people would have drunk beer made from wheat and prepared with honey. The lower classes would have had to content themselves with a simple beer, without honey, called corma, which other ancient authors, such as Marcellus and Pedanius Dioscorides, describe as barley beer. About the end of the first century BC, Strabo (IV, 6, 2) mentions barley beer as the drink of Liguria, a Mediterranean region which included the hinterland of Massalia at the time.

According to archaeobotanical data, there already seems to have been much diversity and complexity in the consumption of alcoholic beverages at Roquepertuse about three centuries before the birth of Posidonius. Of course, even if naked wheat and, later, millets are archaeobotanically identified in the site, we still know nothing about their possible local use in brewing. On the other hand, wine would have been quite commonly consumed. Local people from fifth century BC sites in the vicinity of Etang de Berre were accustomed to buying considerable amounts of wine from Mediterranean merchants and above all from the Massaliots (e.g., Dietler 1990, 1997; Py 1993; Sourisseau 2000). Although only small amounts of Massaliote amphorae shards have been recovered from the fifth century BC habitations of Roquepertuse (Boissinot et al.2000). Moreover, we have to take into account the consumption of an indigenous, locally produced wine. At Roquepertuse, domesticated pips are evidenced. Of course we do not know whether grapes were used to make wine on the site or eaten fresh, dried or used in some kind of preparation. Nevertheless, neighboring Late Iron Age sites provided unusual archaeobotanical evidence of winemaking. In Coudouneu (5th c. BC), Île de Martigues (4th c. BC) and le Castellan (2nd c BC) (Fig. 1), assemblages of pips, pedicels, and pressed grape skins have been recovered from storage and domestic contexts (Marinval 1997; Bouby and Marinval 2001; Marty and Del Corso 2002). Such assemblages are clearly residues of wine-pressing activities (Margaritis and Jones 2006). They consistently demonstrate the existence of local wine-making.

Beer drinking seems to have had a strong cultural connotation. Beer is regarded as a long-standing traditional beverage of Western Europe (Dietler 1990), even if, as we have seen, the evidence remains scarce. On the contrary, Greco-Latin authors considered beer with a fair amount of disdain (André 1981). Mediterranean people are regarded as wine-drinkers and in Southern France, wine would have been an alien alcoholic beverage borrowed from Mediterranean colonial agents and rapidly indigenized (e.g., Dietler 1990; 1997). It is true that no firm evidence of wine production or of grape cultivation is available in Southern France and Eastern Spain before the establishment of colonial trade and interactions (Bouby and Marinval 2001; Brun 2004; Buxo 2008). However, considering how fast grape cultivation and wine production had spread in Mediterranean Southern France, we should not totally rule out the possibility that wine may have been known and used on a small scale by local societies before the emergence of the Mediterranean wine trade.

What could have been the respective status and functions of these various alcoholic beverages? A kind of social classification similar to the one portrayed later by Posidonius of Apameia could be assumed: Massaliot wine for rich people, local wine for common people and barley beer for the poor. The social, cultural and economical factors at work may however have been more intricate, subtle and interdependent. In non-industrial societies, alcohol is not only a foodstuff but commonly has many sociocultural, political and symbolic values (e.g., Dietler 1990, 2006; Jennings and Bowser 2008; McGovern 2009). The fact that beer was probably produced in a domestic context in Roquepertuse does not necessarily mean that it was just an everyday food. For example, in today’s Gamo society in Ethiopia, beer is at the same time a common subsistence item, a high status food and an important feasting and ritual drink. It is a household product, even if more especially made in the wealthiest households (Arthur 2003). In Zimbabwe, we also encounter home-brewed beer which is importantly utilized during ceremonies, but is also drunk in everyday life, even if less commonly (Vermeulen et al.1996). We should not regard beer as only or primarily a poor man’s drink. Processing grain into beer instead of bread requires investing a significantly amount of labor and time, which adds to its value. Beer could even have represented a commercial good in Celtic societies. We mentioned above the passage in Pliny (N.H., XIV, 29) referring to some Spanish beers that could be kept for several years. Hence, if people were able to preserve it effectively, beer could have been exchanged. It has been argued that wooden barrels may have been invented and produced by Celtic people to store and transport beer before being used for wine (Tchernia 1997).

Conclusion

In non-industrial societies, brewing beer does not necessitate any special implement or structure. It can easily be carried out in a household context. The production and use of beer therefore leave hardly any discernable trace in classical archaeological records. In this instance, archaeobotany can be of great use to us, provided that favorable preservation conditions are present.

This is what indeed seems to have been the case in the fifth century BC human habitation at Roquepertuse. An assemblage distinctly dominated by germinated barley grains could be interpreted as accidentally carbonized malt. A parallel can be drawn with much more recent Greco-Roman texts which cite the use of barley beer among Celtic populations in Gaul and Spain. The Roquepertuse example suggests that beer was already produced within the context of domestic activities. Compared to other archaeobotanical and archaeological evidence, findings from Roquepertuse contribute to portraying a society that combined an intricate use of various alcoholic beverages including beer, which was probably of long-standing local tradition, and wine, which was, at least in part, promoted by colonial contacts with Mediterranean agents.

Acknowledgements

We would like to thank Cozette Griffin-Kremer for her help in improving the English text. Many thanks are also due to Michael Dietler and Hans Peter Stika, for their useful comments and help in improving the manuscript.

Copyright information

© Springer Science+Business Media, LLC 2011