Alonso, N. et al. 2013 The effect of dehusking on cereals: experimentation for archaeobotanical...

ACTES DES XXXIIIe RENCONTRES INTERNATIONALES D’ARCHÉOLOGIE ET D’HISTOIRE D’ANTIBES

ÉDITIONS APDCACULTURES ET ENVIRONNEMENTS.

PRÉHISTOIRE, ANTIQUITÉ, MOYEN ÂGEVILLE D’ANTIBES

R E G A R D S C R O I S É S S U R L E S O U T I L S L I É S A U T R A V A I L D E S V É G É T A U X

A N I N T E R D I S C I P L I N A R Y F O C U S O N P L A N T - W O R K I N G T O O L S

Sous la direction dePatricia C. Anderson, Carole Cheval et Aline Durand

REGARDS CROISÉS SUR LES OUTILS LIÉS AU TRAVAIL DES VÉGÉTAUX

AN INTERDISCIPLINARY FOCUS ON PLANT-WORKING TOOLS

ACTES DES RENCONTRES23-25 octobre 2012

Sous la direction dePatricia C. Anderson, Carole Cheval et Aline Durand

Avec le concours du CEPAM : Cultures et Environnements. Préhistoire, Antiquité, Moyen Âge

et de la ville d'Antibes

XXXIIIe RENCONTRES INTERNATIONALES D’ARCHÉOLOGIE ET D’HISTOIRE D’ANTIBES

Éditions APDCA – Antibes – 2013

9

REGARDS CROISÉS SUR LES OUTILS LIÉS AU TRAVAIL DES VÉGÉTAUX. AN INTERDISCIPLINARY FOCUS ON PLANT-WORKING TOOLS.XXXIIIe rencontres internationales d’archéologie et d’histoire d’AntibesSous la direction de P. C. Anderson, C. Cheval et A. DurandÉditions APDCA, Antibes, 2013

Sommaire15 Patricia C. ANDERON, Carole CHEVAL

Introduction

25 Marion FELIX BERNARD

Étude des traces d’utilisation d’une faucille en bronze expérimentale corrodée

37 Nicolas BERNIGAUD

Systèmes agro-pastoraux et utilisation de la faux en Dauphiné depuis le second âge du Fer

49 François POPLIN

La faucille falx veruculata denticulata de Columelle : une énigme bien verrouillée

61 Isabelle BROUILLET

Un outil de moisson du mil au Mali aujourd’hui : le « kebele »

75 Mondher M’HAMDI et Patricia C. ANDERSON

Approche ethnoarchéologique d’outils et techniques de moisson de l’alfa (Stipa tenacissima) dans la région des Hautes Steppes en Tunisie :Premières observations et analyses

89 Patricia C. ANDERSON

Neolithic Tools Used For Stripping Ears From Hulled Cereals : An Update

103 Perrine MANE

Fléau, tribulum ou foulage…, différentes techniques d’égrenage au Moyen Âge

117 Natalia SKAKUN, Natalia VINOGRADOVA et Vera TEREKHINA

New data on agricultural tools of the late Bronze Age from Southern Tajikistan (on materials of the settlement Kangurttut)

129 Florent JODRY et Emmanuelle THOMANN

« Autour du moulin… » Réalisation d’un moulin rotatif manuel expérimental de la fin du deuxième âge du Fer et premiers résultats de mouture

139 Pascal VERDIN et Gilles DESRAYAUD

Les structures de séchage et de grillage de céréales : un outil dans le processus de traitement des céréales

REGARDS CROISÉS SUR LES OUTILS LIÉS AU TRAVAIL DES VÉGÉTAUX

10

155 Natàlia ALONSO, Ferran ANTOLÍN, Daniel LÓPEZ, Francisco José CANTERO et Georgina PRATS

The effect of dehusking on cereals : experimentation for archaeobotanical comparison

169 Soultana MARIA VALAMOTI, Danai CHONDROU et Lambrini PAPADOPOULOU

Plant food processing and ground stone equipment in prehistoric Greece : An experimental investigation using seeds of einkorn and grass-pea

189 Alexandre CHEVALIER et Dominique BOSQUET

Culture matérielle, exploitation du territoire et identités socio-culturelles rubanées en Belgique : analyses de microfossiles sur des instruments de mouture

205 Marta PORTILLO, Maria BOFILL, Miquel MOLIST et Rosa M. ALBERT

Phytolith and use-wear functional evidence for grinding stones from the Near East

219 Maria BOFILL, Hara PROCOPIOU, Roberto VARGIOLU et Hassan ZAHOUANI

Use-wear analysis of Near Eastern prehistoric Grinding stones

237 Xavier RODA GILABERT, Jorge MARTÍNEZ-MORENO et Rafael MORA TORCAL

La gestion des végétaux dans les Pyrénées : la consommation des noisettes sur le site mésolithique de Font del Ros

251 Narjys EL ALAOUI

De la pierre non débitée à l’outil : la vie humaine des pierres dans l’extraction domestique des huiles végétales au Maroc

267 Caroline HAMON et Valérie LE GALL

Le végétal outil, le végétal transformé : fabrication et usages des mortiers en bois en pays Minyanka (Mali)

279 Olivier LANGLOIS, Christian SEIGNOBOS et Patricia ANDERSON

Vers une histoire du « sel de potasse » dans le nord du Cameroun : observations préliminaires

297 Justine MAYCA et Maxence BAILLY

L’écorce, le liber et les pointes plates. Caractérisation tracéologique de la pérennité fonctionnelle d’un outil emblématique des stations littorales de la baie d’Auvernier (NE, Suisse)

311 Thibault LEGRAND, Renaud LEROY et avec la collaboration de Camille LEHNEBACH

Des aménagements de berges à Douai (Nord), « La Motte Julien »

323 Carole CHEVAL et Giovanna RADI

Les lames de tissage, critères de détermination et perspectives de recherche

341 Emmanuelle MARTIAL, Fabienne MÉDARD, Nicolas CAYOL, Caroline HAMON, Yolaine MAIGROT et Cécile MONCHABLON

Chaîne opératoire textile au Néolithique final dans le nord de la France :méthodologie et premiers résultats de l’approche pluridisciplinaire

355 Cozette GRIFFIN KREMER

A bruising experience : tools to soften furze (Ulex spp.) for fodder

367 Émilie CLAUD, Céline THIÉBAUT, Aude COUDENNEAU, Marianne DESCHAMPS, Vincent MOURRE et David COLONGE

Le travail du bois au Paléolithique moyen : nouvelles données issues de l’étude tracéologique de plusieurs industries lithiques d’Europe Occidentale

383 Antoni PALOMO, Raquel PIQUÉ, Xavier TERRADAS, Oriol LOPEZ, Ignacio CLEMENTE et Juan F. GIBAJA

Woodworking technology in the Early Neolithic site of La Draga (Banyoles, Spain)

397 Sylvain BURRI, Aline DURAND, Vanessa PY et Christophe VASCHALDE

Les outils pour acquérir et transformer la matière ligneuse dans les chaînes opératoires techniques des artisanats forestiers en Provence et Haut-Dauphiné au Moyen Âge

415 Toomai BOUCHERAT

Démonstration : approche technologique pour un usage de valves de moules crantées et de bâtons percés au Paléolithique supérieur et au Mésolithique dans la réalisation de fils et de cordes d'origine végétale

11

SOMMAIRE

155

REGARDS CROISÉS SUR LES OUTILS LIÉS AU TRAVAIL DES VÉGÉTAUX. AN INTERDISCIPLINARY FOCUS ON PLANT-WORKING TOOLS.XXXIIIe rencontres internationales d’archéologie et d’histoire d’AntibesSous la direction de P. C. Anderson, C. Cheval et A. DurandÉditions APDCA, Antibes, 2013

comparisonNatàlia ALONSOa, Ferran ANTOLÍNb, Daniel LÓPEZc, Francisco José CANTEROd and Georgina PRATSa

AbstractArchaeobotanical investigations on several archaeological sites have been used to set a pattern of the fragmentation of cereal grains and chaff remains (nodes and rachis seg-ments, spikelets and glume bases) prior to charring. Such a pattern could be the result of threshing activities as well as the dehusking practice of hulled cereals. During this first experimentation, the effect of the latter is evaluated, considering the possible pre-treat-ment of the grains (“fresh”, soaked, toasted), the type of action that is conducted (milling or pounding) and the tool used (wooden mortar, stone mortar, pit, saddle quern, rotary quern). The analyzed cereals are 2-rowed and 6-rowed hulled barley, emmer and einkorn.Keywords : Archaeobotany, Experimentation, Dehusking cereals, Glume wheats, Hulled barley.

À partir d’échantillons carpologiques issus de plusieurs sites archéologiques a été établi un schéma de fragmentation, avant carbonisation, des caryopses de céréales et restes de vannes (nœuds et segments de rachis, bases de glumes et glumelles) qui pourraient correspondre aux processus de battage et aussi de décorticage des céréales vêtues. Une première expérimentation a permis d’analyser l’effet du décorticage en fonction d’un possible prétraitement des grains (« frais », détrempés ou grillés), de l’action réalisée (broyage ou mouture) et de l’outil utilisé (mortier en bois, mortier en pierre, fosse, meule à va-et-vient, moulin rotatif). Les céréales analysées sont l’orge vêtue à 2 ou 6 rangs, l’amidonnier et l’engrain.Mots clés : archéobotanique, expérimentation, décorticage, blés vêtus, orge vêtue.

a. Grup d’Investigació Prehistòrica, Departament d’Història, Facultat de Lletres, Universitat de Lleida, Pl. Víctor Siurana, 1, 25430 Lleida, Catalonia, Spain.b. Laboratori d’Arqueobotànica, Universitat Autònoma de Barcelona / Integrative Prehistory and Ar-chaeological Science (IPNA/IPAS), Universität Basel. Spalenring, 145, CH-4055 Basel, Switzerland.c. Món Iber ROCS SL C. Santa Anna, 25, 08800 Vilanova i la Geltrú, Catalonia, Spain.d. Universitat de Barcelona, Facultat de Geografia i Història C. Montalegre, 6, 08001 Barcelona, Catalonia, Spain.

Natàlia ALONSO et al.

156

Cereal remains are one of the most frequently encountered archaeobotanical finds since the beginnings of agriculture. They are recovered, primarily, in a char-red state and their state of preservation can be very diverse, according to several variables such as the degree and intensity of the heating treatment, the type of sediment in which they were located or the postdepositional processes that they suffered. Both the caryopses (grains) and the chaff fragments (that is to say, other parts of the ear, e. g. rachis fragments, rachis nodes and internodes, spikelet forks, glume bases…) can be recovered.

Within the agricultural process, once harvested, cereals must be threshed in order to separate the grain for human consumption. Free-threshing cereals are those in which, during this process, the ear easily frees the grain. These would include, for instance, bread wheat, Triticum aestivum (the ear immediately breaks into grain and light chaff) or hulled barley, Hordeum vulgare ssp. vulgare (the ear breaks into grain, awn fragments and rachis segments). The hulled wheats, conver-sely, do not free their grain so easily and they break into spikelets. Some examples are emmer, Triticum dicoccum, or einkorn, Triticum monococcum (the ear breaks up into separate spikelets and awn fragments) (see for example Hillman, 1984).

On the other hand, and according to the degree of adhesion between the caryopses and the glumes (bracts which form a husk around the grain), cereals are classified into naked varieties (i. e. bread wheat), where the grain and the glumes easily separate from each other, and hulled varieties (i. e. einkorn, emmer or hul-led barley), where the grain and the glumes do not separate despite the threshing process. Therefore, hulled cereals need a supplementary process of dehusking, even if they are of the free-threshing type, in order to free the grain from the hull (in the case of barley) or from whole spikelets (in the case of glume wheats). In the latter case, the spikelet ends up breaking into free grains, light chaff (glumes and awns), rachis segments, spikelet forks and glume bases.

As far as archaeobotanical remains are concerned, one can recover cereal grains with glumes, without glumes, spikelets, glume bases, rachis segments, etc. Caryopses can be found complete or fragmented. Their fragmentation can happen after the charring, due to post-depositional processes (trampling, sedi-mentary pressure, soil acidity, etc., or the process of excavation and recovery of the remains) or prior to the charring (threshing and dehusking processes, but also milling). The identification criteria for distinguishing both types of fragments were already established by other authors (see, for instance, Valamoti, 2002), for which it is possible to carry out an archaeobotanical evaluation of the fragmenta-tion of the grain produced prior to charring.

Fragments of grain produced both prior to and after charring can affect the grain in transversal or longitudinal direction. In the first case, apical, embryo-nal or medial fragments can be obtained. In the second case, ventral, dorsal or ventral-dorsal fragments can be produced (for further details, see Antolín and Buxó, 2011).

THE EFFECT OF DEHUSKING ON CEREALS : EXPERIMENTATION FOR ARCHAEOBOTANICAL COMPARISON

157

In this work, we will focus on the effects of dehusking on archaeobotanical remains. Several dehusking methods are known through ethnographic research. Some of them are applied on a larger scale (for instance, large wooden mortars), where more than one person can work at a time, while others are used on a day to day basis, being performed by single individuals for their own consumption. The implications of each system are of high economic and social significance, for which the identification of these techniques might improve our knowledge of past societies.

Our working hypothesis is that the different types of fragmentation of hulled cereals observed in carbonized, archaeological remains could be caused by the different dehusking methods applied.

In order to evaluate this issue, an experiment was carried out with the fol-lowing objectives :1. Observing the diverse types of fragmentation patterns produced as a result of :

— The pre-treatment of the grains or the spikelets.

— The type of action that is conducted on the grains or spikelets.

— The type of tool used.

2. Examining the dehusking efficiency of the different methods.

3. Analyzing the breakage of chaff remains.

It was not our intention to carry out an experiment of the process of dehus-king in itself. Therefore, the working time, the number of actions or the use of archaeological replicas as working tools, for instance, are not aspects that were considered within this work.

The experiment was conducted within the framework of the Camp d’Experimen-tació de la Protohistòria (CEP) of Verdú (Catalonia), where experimental research in archaeobotany was recently started in order to target several questions raised after the analysis of archaeological seed and fruit remains.

The experiment was conducted on four cereal taxa : two glume wheats, einkorn (Triticum monococcum) and emmer (Triticum dicoccum), and two hulled barleys, 2-rowed barley (Hordeum vulgare distichum) and 6-rowed barley (Hordeum vulgare hexastichum). These are the most frequently encountered hulled cereals in the Western Mediterranean. In order to proceed with their dehusking, which can be conducted in a variety of ways, several variables were considered (fig. 1) following ethnographic references (see discussion chapter).

On the one hand, three different pre-treatments were applied : raw or “fresh” (without pre-treatment), roasting and soaking. Five sets of 0.5 litres of each taxon were treated with each method. Roasting was done without temperature control, using a metallic pan. Soaking lasted about 30 minutes.


158

Two different types of action and five different utensils were used for dehusking :1. Pounding :

– Pounding with a wooden mortar with a concave base and a wooden pestle.

– Pounding with a stone mortar with a concave base and a wooden pestle.

– Pounding in an earthen pit with a wooden mallet.

2. Grinding :

– Milling with two saddle querns of granite.

– Milling with two rotary querns of conglomerate, without specific separation between the stones.

Sixty sets were processed in total. The experiment was performed by us, toge-ther with several students from the Dept. of History of the Universitat de Lleida,

Fig. 1. Table of samples and images of the experimentation process : (1) roasting ; (2) soaking ; (3) wooden mortar and pestle, and dehusking with it ; (4) stone mortar and pestle, and dehusking with it ; (5) saddle quern, and dehusking with it ; (6) rotary quern, and dehusking with it ; (7) earthen pit, and dehusking with a mallet.


159

with no previous experience in this task. The working time was around 30 minutes per set, although it was shorter in some cases.

— After dehusking, the product obtained was sieved with a 3 mm mesh. The resul-ting fraction recovered inside the sieve was considered the desired product, including the vast majority of the grain, together with other parts of the spike-let (glume fragments, spikelet forks, etc.). The analysis was largely performed on the results of this fraction. Around 400 items were aimed per set in order to have a representative population for the analysis (following Van der Veen and Fieller, 1982). For that purpose, a riffle box (sample splitter) was used.

— The residue of less than 3 mm was sieved again with a sieve of 1 mm mesh size in order to obtain two fractions. Only the fraction above 1 mm was analyzed. Both analyzed fractions were winnowed using a mechanical system of venti-lation, which allowed a rapid separation of the heavy (mainly grain, but also spikelets and some spikelet forks) and the light (mainly glume fragments, glume bases and awns) residues.

— This procedure allowed the quantification of the variables that were of interest for this study, which were recorded on specific forms for each taxon. These variables are : the number of spikelets, the number of grains with glumes, the number of husked grains, the number of complete grains, the abundance of small fragments of grain, of longitudinal fragments, of transversal fragments, and the abundance of glumes, spikelet forks, glume bases and rachis segments.

Evaluation of the fragmentation of the grains

The evaluation of the degree of fragmentation and the type of fragmentation (transversal or longitudinal) of the caryopses was carried out in a semi-quantified

Fig. 2. Evaluation of grain fragmentation : (a) according to cereal taxon ; (b) according to the type of action and the implement. Each point represents a sample (present 25 %, abundant > 25 % < 50 %, dominant 50 %...).

0%

0%

0%100%

100%

100%

who

le g

rain

s

transversal fr.

longitudinal fr.einkorn emmer 6r-barley2r-barley

a

0%

0%

0%100%

100%

100%

who

le g

rain

s

transversal fr.

longitudinal fr.wooden mortarstone mortar

saddle quernrotary quern earthen pit

b


160

way (1, present ; 2, abundant ; 3, dominant). When representing the results in a triangular diagram (considering the percentage of whole grains, transversal and longitudinal fragments) (fig. 2a), one can observe a specific general dis-tribution according to cereal type :

— Einkorn fragments intensively and it presents a higher tendency towards longi-tudinal fragmentation (fig. 3e),

— Emmer presents higher indices of frag-mentation and transversal fragmentation predominates (fig. 3f), even though longitudinal fragmentation is also well represented,

— 2 rowed-barley is the taxon which was less fragmented and, when fragments were pres-ent, these were of transversal type primarily,

— 6-rowed barley presents a less clear pat-tern, rather fragmented, both transversally and longitudinally.

When considering both the type of action and the implement used in this graph (fig. 2b), it can be observed how stone tools tend to break the grains more intensively than wooden tools or earthen surfaces. Grinding also frag-ments more than pounding, but this might be correlated with the fact that the implements used for this were made of stone. The type of fragmentation seems to not be correlated with the type of tool.

Finally, it seems that the pre-treatment does not have an influence on the type of fragmen-tation (graph not shown).

The efficiency of the dehusking process was calculated considering the ratio between the grain actually dehusked (at least half of the grain) and non-dehusked grain. We have

Fig. 3. (a) dehusked grains of einkorn ; (b) dehusked grains of emmer ; (c) dehusked grains of 2-r barley ; (d) dehusked grains of 6-r barley ; (e) longitudinal fragments of einkorn ; (f) transversal fragments of emmer.


161

considered as individuals any com-plete caryopsis or fragment which conserved the embryo area (including longitudinal fragments, which were counted and divided by two). The ones which remained inside spikelets were equally counted.

The percentage of dehusked grain is low, in a general view (fig. 4). This might be due to the lack of experience of the performers of the experiment. It is probable that increasing the working time would have produced signifi-cantly better results. Nevertheless, we consider that the results obtained are comparable, due to the similar level of inexperience of all the volunteers who participated in the experiment.

Interesting trends can be observed when looking at the results per taxon :

— The glume wheats (emmer and einkorn) gave better results either without any pre-treatment or roasted (not soaked). Both einkorn and emmer gave higher indexes of dehusking efficiency with non pre-treated grain and dehusking by pounding with a wooden mor-tar (84 % for einkorn and 77 % for einkorn) (fig. 4a). Both species also gave good results when toasted then pounded with a stone mortar (76 % of dehusked grain for einkorn and 68 % for emmer) (fig. 4b). Emmer seems to be dehusked efficiently also without any pre-treatment and using the rotary quern (index : 70 %). It is very interesting to note how both taxa react similarly in

Fig. 4. Comparison between the indexes of dehusking efficiency according to the utensil and the pre-treatment of the cereals. The percentages shown correspond to the percentage of dehusked grains in each case (es = einkorn ; pi = emmer ; od = 2-r barley ; oh = 6-r barley).

pi: 77%

pi: 48%

rawindex: 41%

roastedindex: 33%

soakedindex: 41%

es: 84%

es: 52%

oh: 78%

oh: 50%

a. Wooden mortar (n: 5697 index 37%)

rawindex: 9%

roastedindex: 17%

soakedindex: 21%

es: 22%

oh:44%

oh: 24%

oh: 26%

od: 21%

e. Earthen pit(n: 4834 index 15%)

rawindex: 42%

roastedindex: 59%

soakedindex: 22%

pi:68%

es: 76%

oh: 96%

oh: 58%

b. Stone mortar(n: 3088 index 39%)

od: 85%

rawindex: 21%

roastedindex: 34%

soakedindex: 28%

es: 35%

es: 58%

pi:52%

pi:34%

es:31%

pi:41%

oh: 36%

c. Saddle quern(n: 4585 index 28%)

rawindex: 27%

roastedindex: 16%

soakedindex: 33%

es: 41%

pi:70%

oh: 45%

es: 51%

pi:50%

d. Rotary quern(n: 4188 index 27%)


162

each of the different experiments (fig. 5). Despite some divergences in percentages, the reaction of both taxa to the pre-treament and tools used are quite similar.

— 2-rowed barley gave the lowest indices of dehusking efficiency, except in one case, when toasted and pounded with a stone mortar (index : 85 %) (fig. 4b).

— 6-rowed barley seems to give the best results when the grains are soaked. Nevertheless, the best index was obtained with the stone mortar and without pre-treatment (index : 96 %) (fig. 3b). Soaked grain gave good results when dehusked by pounding, especially with the woo-den mortar (index : 78 %) as well as with the stone mortar (58 %).

Regarding the type of tool used, the overall indices are very low, but mortars seem to have a slightly higher efficiency, while querns seem to be somewhat less efficient. The earthen pit was, accor-ding to our results, the least efficient method (fig. 4e). Whatever the pre-treatment system considered, it does not seem to affect the results obtained for each tool. Only in the case of the stone mortar, does it seem to be less efficient on soaked grains, and the earthen pit is less efficient for grains in a raw state.

Some final observations can be put forward for each implement according to the variables which showed a higher efficiency (fig. 4) :

— The wooden mortar gave the best results with glume wheats in raw state (84 % and 77 %) and toasted hulled barley (78 %).

— The stone mortar gave the highest indexes of efficiency, especially with 6-rowed barley in raw state (96 %) and roasted cereals (einkorn, 76 % ; emmer, 68 % ; 2-rowed barley, 85 %). As already mentioned, this is the only technique that produced acceptable results for 2-rowed barley.

— The saddle quern produced, in general, low indexes of efficiency. Only toasted einkorn and emmer produced indexes of over 50 % of dehusked grain.

— Finally, for the rotary quern, emmer in raw state gave the best results (70 %), followed by toasted einkorn (51 %) and soaked emmer (50 %).

0

20

40

60

80

100

raw roasted soaked

einkornemmer

n1 n2 n3 n4 n5 t1 t2 t3 t4 t5 m1 m2 m3 m4 m5

Fig. 5. Comparison between the indexes of dehusking efficiency of the glume wheats.

Fig. 6. (a) spikelet forks of emmer ; (b) glume bases of einkorn ; (c) rachis segments of emmer ; (d) light chaff of einkorn.


163

Chaff

Concerning the evaluation of chaff remains, the most interesting results refer to the type of breakage of the spikelet forks of glume wheats (fig. 6). The abun-dance index of glume bases of einkorn is higher, which indicates that the spikelet forks of this taxon tend to get fragmented more easily than those of emmer. On the other hand, in roasted wheats, the remaining spikelets and the glume bases are more similar to the archaeological ones. Roasting probably made the glumes more brittle and fragile, for which complete glumes were hardly ever recove-red. The opposite results were observed with raw and soaked spikelets, where the glumes are fully preserved. For a proper evaluation of the archaeobotanical significance of these observations, an experimental charring under controlled conditions would become necessary.

In the last decades, several experimental works on dehusking cereals (wheat and barley) have been carried out, as well as complete ethnographic observations that cover the whole process that we have experimented with in this work.

The pre-treatment of the grain is widely documented, with parching, roasting and soaking the most commonly used. Parching and roasting make chaff more brittle. It is applied both to glume wheats and barley. This practice is documented, for instance, in Ancient Greece with barley grains (Amouretti, 1986 : 135). In dry climates, parching is usually not practiced, and spikelets are just dried in the sun (Hillman, 1984 : 129 ; Peña-Chocarro et alii, 2009 : 106). On the other hand, roasting barley also allows a better preservation of the product, which can be kept in jars after that (Ferchiou, 1985). Parching and roasting are usually carried out on the fire with ceramic or metallic pans (Amouretti, 1986 : 135 ; Alonso et alii, in press) or in ovens (Hillman, 1984 : 129). In our observations roasting produced better results when using a stone mortar (especially with 2-rowed barley) and a saddle quern (particularly with glume wheats).

On the other hand, soaking grain facilitates the glumes and light chaff being stripped from the rachis internodes and freeing the grain which mostly stays intact. Besides, water softens the chaff and makes it flexible, so that whole grains often pop out of the spikelet (Nesbitt and Samuel, 1996 : 52-53). This fact has been observed during our experimental work, since, as already mentioned, spikelet forks remained largely intact. According to our results, soaking has positive effects on the dehusking of 6-rowed barley, in particular.

If grain was soaked, then it required subsequent drying, whether naturally (in the sun) or artificially (using fire), which would facilitate, together with parching, the carbonization of assemblages of chaff and grain, making these remains more likely to survive in the archaeological deposits. The experimental charring of our products obtained after soaking is necessary in order to test whether this practice can be identified on archaeobotanical material.


164

Regarding utensils and their use, mortars and pestles/mallets are the more common equipment for dehusking emmer all over the Mediterranean, Europe and Iran or India (Gast, 1968 ; Hillman, 1984 : 130 ; Fenton 1992). They were made on stone or wood as well in ancient Sumer (Grégoire, 1992 : 328), Egypt (Nesbitt and Samuel, 1996) or in the Classical World (Moritz, 1958 : 22-28 and 146-147 ; Amouretti, 1986 : 135-137). In Turkey, those most commonly used are those made from wood, albeit their use life is shorter (Hillman, 1984 : 130 ; 1985 : 20).

Several experiments with mortars have been carried out by researchers. G. Hillman (1984 : 130) used the replica of a wooden mortar of modern age from Wales, which, according to the author, was extremely efficient to dehusk wild ein-korn. Unfortunately, no further details were provided.

H. Procopiou conducted another experiment using a wooden mortar where 2-rowed barley and emmer were processed in raw state, roasted and soaked (2003 : 121-122). Both for barley and emmer, the resulting indexes of dehusking effi-ciency were very high, between 80 and 90 %. Roasted grains, according to the author, tend to fragment more (even though no quantified data were supplied), while soaked cereal does not break and the operation was carried out faster.

The data that we have obtained for these cereals with the wooden mortar also give high indexes of dehusking efficiency for emmer in raw state (77 %), but not for roasted and soaked emmer (48 % and 34 %, respectively). Besides, in our case, the resulting grain was much more fragmented notwithstanding the applied pre-treatment. In contrast, soaked 2-rowed barley is one of the least frag-mented, followed by raw 2-rowed barley, but the percentage of dehusked grains is extremely low, with a maximum of 23 % for the soaked barley. We lack a convin-cing explanation for this phenomenon at the moment and further experiments are needed.

Stone mortars are equally used, even though they may increase the fragmenta-tion of the grain. Hillman (1984 : 130) introduces the case of the rice producers of northwest Anatolia, who sell their product after dehusking it with a stone mortar and without breaking the grain. This leads this author to conclude that stone mortars could have been used in the past as well for dehusking emmer.

The size of the mortars is determined by the quantity of grain that is to be processed. In Turkey, two sizes are documented (Hillman, 1984 : 130 ; Ertug-Yaras, 2002 : 212-214) : the large ones, carved stone mortars, and the small ones, of around 10 cm in diameter, always combined with the use of the pestle, not the mallet. M.-Cl. Amouretti distinguishes two types of mortar also for Ancient Greece (1986 : 135-137) : the small circular one, with a short pestle, and the large and deep one, with a long pestle. The former could be made of stone or ceramic, and it was not particularly linked to cereals. Yet it could be used to finish the grinding process or to mix this product with other ingredients. The latter is more specifically for cereals and they could be made of wood or stone.

Nesbitt and Samuel (1996 : 52-53) used a shallow ancient stone mortar to dehusk emmer and they splashed it with water so that its contents did not spread out. This practice is also observed in Turkey (Hillman, 1985 : 20 ; Ertug-Yaras,


165

1992 : 214). In our experiment, the stone mortar gave the best results for all taxa, for 6-rowed barley in raw state (96 %), and for the rest after toasting (69-85 %).

Stone mortars are well known from archaeological contexts, even before the beginnings of agriculture (Kraybill, 1977) and they were used until classical Antiquity. Their relation with the process of dehusking is reinforced through the finding of a mortar with chaff remains of emmer in Egypt (Nesbitt and Samuel 1996 : 52).

The use of a saddle quern for dehusking was also experimentally tested. H. Küster (1984 : 310) performed an experiment with parched and soaked einkorn, using a saddle quern with a light upper stone, taking special care not to practice any excessive pressure that could break the grain. In the first case, the results were not considered as successful by the author, since the grain and chaff remains fragmented a lot and they became more difficult to separate. In the second case, the author observed that the grain and the chaff stuck together and that they were impossible to separate. In our experiment we obtained better results with roasted einkorn (58 %) than with the soaked sample (38 %) but frag-mentation was intense, resulting in longitudinal fragments, primarily.

J. Meurers-Balke and J. Lüning also carried out an experimental dehusking of raw and parched emmer using a saddle quern (Lüning and Meurers-Balke, 1980 : 338-344 ; Meurers-Balke and Lüning, 1992 : 346-348). For the former, sets of 30 gr of emmer were used and 8 back-and-forth movements were done. After 4 movements, 50 % of the grain was dehusked, yet 50 % of it was cracked or crus-hed. The authors considered these fragments as lost product, since they could not be isolated from the chaff. Parching emmer (at 50, 100, 150 and 200 ºC) seemed to accelerate the process when increasing the temperature of the pre-treatment, but fragmentation also increased during dehusking. In the results we have presented here, the fragmentation pattern is similar, although the dehusking efficiency index is slightly lower (41 % for the raw grain and 52 % for the toasted one). We also obtained overall better results with roasted cereals (over 50 % for the two glume wheats).

A more exhaustive experiment was carried out by H. Procopiou (2003 : 119-121). In the experiment, the author used a saddle quern with handstones and wooden handles, dehusking 2-rowed barley, emmer and wild einkorn. The grain was manipulated in three different states : raw, roasted and soaked. In the first case, raw grains tended to break and reduce to groats. We did not observe this behaviour, since in our case it seems that pre-treatment did not influence frag-mentation. Only a small percentage (10-15 %) of the wheats were dehusked in Procopiou’s work, while barley yielded negative results. The type of handstone did not seem to affect the results. The low index of dehusking efficiency of emmer does not coincide with the data obtained by Meurers-Balke and Lüning, nor with our own results (around 50 %). In contrast, the results obtained for 2-rowed bar-ley clearly match ours. The more positive results were obtained with soaked grain (after 30 min of soaking). The percentage of dehusked grain was around 60-70 %. Concerning toasted grain, the results for emmer were similar to those in our


166

experiment, just slightly superior to those obtained with the raw grain (41 % with raw emmer and 52 % with toasted). The influence of soaking is, nevertheless, lower in our case.

Data concerning dehusking with rotary querns are much scarcer. It has been documented, for instance, in Turkey, in Kar province (Hillman, 1984 : 138), where the separation between the two stones is about 2-3 mm and the operation pro-duces a mixture of fragments of grain (grits), floury particles of endosperm, fine chaff and weed. In the Moroccan Rif, a sheet of cork was introduced between the stones, which avoided the breakage of the grain (Peña-Chocarro et alii, 2009 : 106-107). On some occasions this sheet of cork could be used as lower stone, by fixing it to the ground. In our experiments, where such a sheet was not used, good results were obtained for raw and soaked emmer (70 and 50 %, respectively) and toasted einkorn (51 %).

The processing of barley seems not to include dehusking before grinding the grain. In Crete, in the island of Melos, Turkey, Sardinia or Tunis (Hillman, 1985 : 20 ; Procopiou, 2003 : 117 and 124 ; Alonso et alii, in press), barley is ground with a rotary quern when still hulled and glumes are discarded later by sieving. In our observations, barley tended to produce flour more easily when using the rotary quern (except when soaked in the case of 6-rowed barley). It might be pos-sible then, that rotary querns were not used to dehusk barley in the past.

Finally, the earthen pit is the most commonly used system for dehusking ein-korn in the Moroccan Rif (Peña-Chocarro et alii, 2009 : 106-107). The pit is around 20-30 cm in diameter and has a depth of 20-40 cm. It is plastered with clay, inclu-ding a wide outer rim. The mallet is made with a light wood, so that the grain is not broken and that the bottom of the pit is not cracked. Hillman (1984 : 130) comments on the use of this sort of pit in Sumer, and in Turkey a similar system is documented, the so-called bedrock mortars (Ertag-Yaras, 2002 : 212-214). In our experiment it is the least efficient method, probably due to the breakage of the pit walls during the process. This could be have occurred because the mallet was not made of such a light type of wood. The repetition of this experiment is, therefore, indispensable.

The experiment that has been carried out permitted a first approximation to the response of grain and chaff remains to dehusking operations. The initial hypothesis that the type of fragmentation can be determined by the dehusking system was not fully proved, even though some differences in fragmentation, but especially according to species, were observed. Differences in the intensity of the fragmentation, on the other hand, were observed. Querns gave the highest frag-mentation while wooden mortars produced the lowest. The same happens with chaff remains, where differences were observed mostly between taxa. Spikelet forks of einkorn broke more easily than those of emmer. The pre-treatment used did not influence the results.


167

The dehusking efficiency seemed quite low in general, maybe due to the inexperience of the performers or the working time. Nevertheless, it was highly satisfactory in several cases, reaching figures over 80 %.

In order to offer a suitable reference for archaeological work, the carboniza-tion of the obtained products becomes indispensable and it will be the object of a future work.

Acknowledgements

This study was funded by the projects RecerCaixa 2011ACUP00230 and HAR2012-36877. We acknowledge the collaboration of the students of Prehistory from the University of Lleida and also Aitor Moreno, who, despite having been criticized for their inexperience, have played a major role in this work.

References

ALONSO N., CANTERO F. J., JORNET R., LÓPEZ D., MONTES E., PRATS G., SOLTANI N., VALENZUELA S., SANMARTÍ J., BELARTE C., KALLALA N., in press. Milling Wheat and Barley with rotary querns : the berber Ouarte women (Dahmani, El Kef, Tunisia). Seen through a millstone : Geology and Archaeology of Quarries and Mills, Bryggens Museum, Norwege.

ANTOLÍN F., BUXÓ R., 2011. Proposal for the systematic description and taphonomic study of carbonized cereal grain assemblages : a case study of an early Neolithic funerary con-text in the cave of Can Sadurní (Begues, Barcelona province, Spain). Vegetation History and Archaeobotany, 20 : 53-66.

AMOURETTI M.-Cl., 1986. Le pain et l’huile dans la Grèce antique. De l’araire au moulin, PARIS, CRA, 67, 357p.

ERTUG-YARAS F., 2002. Pounders and Grinders in a Modern Central Anatolian Village, In : PROCOPIOU H., TREUIL R. (dir.). Moudre et broyer. L’interpretation fonctionnelle de l’outillage de mouture et de broyage dans la Préhistoire et l’Antiquité : 211-225, Paris, CTHS.

FENTON A., 1992. Small scale technique in processing cereals for food. In : MESNIL M. (ed.), Du grain au pain. Symboles, Savoirs, Pratiques : 83-95. Bruxelles, Coll. Ethnologies d’Europe.

FERCHIOU S., 1985. Conserves cérèalières et rôle de la femme dans l’économie familiale en Tunisie, In : GAST M., SIGAUT F., Les techniques de conservation des grains à long terme : 190-197. Paris, Ed. CTHS.

GAST M., 1968. Alimentation des population de l’Ahaggar. Étude ethnographique. Memoires du Centre de Recherches Anthropologiques préhistoriques et ethnographiques, 8, Paris.

GRÉGOIRE J.-P., 1992. Les grandes unités de transformation des céréales. L’exemple des minoteries de la Mésopotamie du sud à la fin du IIIe millénaire avant notre ère. In : ANDERSON P. (dir.), Prehistoire de l’Agriculture. Nouvelles Approches expérimentales et ethno-graphiques : 321-339 Monographie du CRA, 6.

HILLMAN G., 1984. Traditional husbandry and processing of archaic cereals in recent times : the operations, products and equipment that might feature in Sumerian texts. Part I : the glume wheats, Bulletin on Sumerian Agriculture, 1 : 114-152


168

HILLMAN G., 1985. Traditional husbandry and processing of archaic cereals in recent times : the operations, products and equipment that might feature in Sumerian texts. Part II : the free-threshing cereals, Bulletin on Sumerian Agriculture, 2 : 1-31

KRAYBILL N., 1977. Pre-Agricultural Tools for the Preparation of Foods in the Old World. In : REED C. (ed.), Origins of Agriculture : 485-521, The Hague, Mouton.

KÜSTER H., 1984. Neolithic plant remains from Eberdingen-Hochdorf southern Germany, 6th

LÜNING J., MEURERS-BALKE J., 1980. Experimenteller Getreideanbau im Hambacher Forst, Gemeinde Elsdorf, Kr. Berheim/Rheinland. Bonner Jahrbücher, 180 : 305-344.

MEURERS-BALKE J., LÜNING J., 1992. Some aspects and experiments concerning the process-ing of glume wheats. In : ANDERSON P. (dir.), Prehistoire de l’Agriculture. Nouvelles Approches expérimentales et ethnographiques : 341-362, Monographie du CRA, 6.

MORITZ L. A., 1958. Grain-mills and flour in Classical Antiquity, Oxford, Clarendon Press, 231 p.

NESBITT M., SAMUEL D., 1996. From staple crop to extinction ? The archaeology and his-tory of the hulled wheats. In : PADULOSI S., HAMMER K., HELLER J. (ed.), Hulled Wheat : 40-99. Proceedings of the First International Workshop on Hulled Wheats, July 1995, Castelvecchio Pascoli, Tuscany, Italy.

PEÑA-CHOCARRO L., ZAPATA L., GONZÁLEZ J. E., IBÁÑEZ J. J., 2009. Einkorn (Triticum mono-coccum L.) cultivation in mountain communities of western Rif (Morocco) : An ethno-archaeological projecte. In : FAIRBAIRN A., WEISS E. (eds.), Ethnobotanist of distant pasts : Archaeological and ethnobotanical studies in honour of Gordon Hillman : 103-111. Oxford, Oxbow.

PROCOPIOU H., 2003. Les techniques de décorticage dans le monde égéen. Étude ethnoar-chéologique dans les Cyclades. In : ANDERSON P. et alii (dir.), Le traitement des récoltes : un regard sur la diversité, du Néolithique au présent, XXIIIes Rencontres internationales d’archéologie et d’histoire d’Antibes, 2003, Antibes, Éditions APDCA : 115-115.

VALAMOTI S. M., 2002. Food remains from Bronze Age Archondiko and Mesimeriani Toumba in northern Greece ?. Vegetation History and Archaeobotany, 11 : 17-22.

VAN DER VEEN M., FIELLER N., 1982. Sampling seeds. Journal of Archaeological Science, 9 : 287-298.

ACTES DES XXXIIIe RENCONTRES INTERNATIONALES D’ARCHÉOLOGIE ET D’HISTOIRE D’ANTIBES

REGARDS CROISÉS SUR LES OUTILS LIÉS AU TRAVAIL DES VÉGÉTAUXAn interdisciplinary focus on plant-working toolsSous la direction de Patricia C. Anderson, Carole Cheval et Aline Durand

L es activités liées au traitement des végétaux restent mal connues en raison de leur nature périssable. Au-delà de l’archéobotanique et de l’étude de l’outillage lithique, plutôt

utilisées pour rechercher ces activités, ce livre porte un regard interdisciplinaire sur des artefacts dont la détermination n’est pas toujours aisée. Les auteurs examinent comment ces outils, tirés de différents matériaux (bois, matières dures d’origine animale, céramique, métal, structures en terre, mobilier lourd) ont été employés pour la récolte, le battage, le meulage, le broyage, la cuisson, la préparation et l’utilisation des fibres et du bois dans des domaines divers : alimentation, cosmétiques, textiles, vanneries, et architecture. Les méthodes engagées combinent l’ethnologie, l’histoire, l’archéologie, l’expérimentation et aussi l’analyse tracéologique, tribologique, métrologique, ou de chimie élémentaire, ainsi que l’étude des restes végétaux tels que phytolithes et amidons. Ces 28 documents d’auteurs originaires de six pays constituent un ensemble de données inhabituelles sur l’histoire des techniques et de leur contexte socioéconomique dans treize pays en Europe, Afrique et Asie du Sud-Ouest, du Paléolithique à nos jours. De nombreux résultats présentés ici sont le fruit de deux projets de recherche, le GDR 2517 du CNRS et le programme EARTH de la Fondation Européenne de la Science (ESF).

Much of the evidence for plant processing escapes us, due to the perishable nature of most plant remains as well as of many of the tools used. Beyond archaeobotany and

analyses of better understood flaked stone tools, this book provides an interdisciplinary look at how tools that are difficult to identify as such were made and used to work plant materials. Authors explore how unusual tools in wood, bone, ground stone, earthen structures, ceramic, and metal were used for harvesting, threshing, grilling, grinding, pounding, fiber preparation and use, extractions from plants, and woodworking, related to animal fodder, human food and condiments, cosmetics, textiles, basketry and building, for example. The interdisciplinary methodologies used combine archaeology, experimental archaeology, history, social and cultural anthropology, microwear analysis, tribology and metrology, elemental chemical analysis, and study of plant macro remains, phytoliths and starch. 28 papers by authors from 6 countries, comprise an unusual data set of new studies on the history of techniques and their social and economic context in 13 countries in Europe, Africa and Southwest Asia, from the Paleolithic to the present. Presented at the Colloque d’ Antibes, many papers result from two collaborative research projects, the GDR 2517 of the CNRS and the EARTH Program of the European Science Foundation (ESF).

Prix : 40 €ISBN 2-904110-53-4

nice S O P H I A A N T I P O L I S

université APDCA

APDCA

FID

E

SERVAND EE X EMPLUM

I 1815

Alonso, N. et al. 2013 The effect of dehusking on cereals: experimentation for archaeobotanical...

Documents

Transcript of Alonso, N. et al. 2013 The effect of dehusking on cereals: experimentation for archaeobotanical...