56

Introduction (La Draga)

The site of La Draga is located in thenorth-east of Iberia, on the eastern shoreof the “Estany de Banyoles” (BanyolesLake), a small lake 50 km from theMediterranean coast and 40 km south ofthe Pyrenees (Fig. 1). I t was discered in1 990 and has been excavated from 1 990to 2005, and again from 201 0 to 201 2,under the direction of the MuseuArqueològic Comarcal de Banyoles(MACB) and the Centre d'ArqueologiaSubaquàtica de Catalunya (CASC).Recently, the research team has beenenlarged with the participation of theConsejo Superior de Investigaciones

Científicas (CSIC), the UniversitatAutònoma de Barcelona (UAB) and theMuseu Arqueològic de Catalunya (MAC).The lake is a karst landform and is fed byunderground waters. Original ly it wasdrained on its eastern side by a smallriver. This river, as it exited the lake, musthave created an area of marshes on thenorthern shore, which is the location of LaDraga archaeological site. During theNeolithic occupation this shore wouldhave been in the form of a peninsulawhich stretched out into the lake, with agentle continuous slope, from east towest and from north to south. Accordingto the surveys which have been carriedout, it is bel ieved that the settlement

Experimentel le Archäologie in Europa - Bilanz 201 2, S. 56-65

Kategorie: Experiment und Versuch

Woodworking technology and functional experimentationin the Neolithic site of La Draga (Banyoles, Spain)

Oriol López, Raquel Piqué, Antoni Palomo

Zusammenfassung – „La Draga“ ist ein vom Wasser eingeschlossenes Gebiet in Ba-nyoles (im Nordosten der iberischen Halbinsel). Die anaeroben Gegebenheiten bieten ei-ne unglaubliche Art der Konservierung von organischen sowie hölzernen Überresten. Dieältesten Siedlungen in diesem Gebiet sind von 5.300-5.130 v. Chr. (frühes Neolithikum).Wir haben ein experimentelles Programm entwickelt, dass durch den Nachbau und dieKonstruktion der hölzernen Artefakte sowie die Benutzung dieser Gegenstände Auf-schluss über ihren genauen Gebrauch und Nutzen liefern soll. Wir benutzen dieses Pro-gramm ebenso um die unterschiedliche morphologische Beschaffenheit sowie die Ge-brauchsspuren der Gegenstände besser zu erkennen und die Artefakte besser zuordnenzu können.Mit diesem experimentellen Programm war es uns somit möglich, die verschiedenenSchritte des Arbeitsablaufes der Holzwerkzeuge zu verstehen und die verschiedenenGebrauchsspuren im Abgleich mit den nachgebildeten Werkzeugen an den Gegenstän-den zu charakterisieren.

57

occupied an area of about 8000 m2.The archaeological excavations carriedout to date have documented a habitationlevel situated immediately above thelacustrine chalk, which corresponds to anoccupation by a Neolithic populationl inked to the Cardial Culture (BOSCH,CHINCHILLA, TORRUS 2006; BOSCH,CHINCHILLA, TORRUS 2011 ) (Fig. 2).Numerous radiocarbon dates have beenobtained from samples of wood, bone andcharcoal, whose results place the oldestoccupation between 5.300 and 5.000 calBC. Equally, based on thedendrochronological analysis of woodenposts, we may consider that theoccupation perhaps went through differentphases, during a period of approximately80 or 1 00 years.Above this habitation layer are the

remains of a more recent, and possiblyless important, occupation, dating to thebeginning of fifth mil lennium cal BC. Atthe moment this second occupation, incontrast with the first, seems to havebeen located further from the shore andhigher above the lake. In areas closer tothe lake this level appears to be verydiffuse. In these areas, some of thestructures dug in the ground affected theearl ier archaeological level.The location of the site does not seem tohave been chosen at random andcorresponds to a pattern that is repeatedin early Neolithic occupations in thewestern Mediterranean. They are wetlandlocations, on the shores of lakes, lagoonsor marshes, yet close to agricultural land.This pattern has been clearly documentedin I taly (La Marmota) (FUGAZZOLA,

Fig. 1: Situation of La Draga and Banyoles in the Iberian Peninsula.

58

D’EUGENIO, PESSINA 1 993), southernFrance (Leucate) (GUILAINE, FREISES,MONTJARDIN 1 984), by lakes in the Alpsand the Jura, and inland in the IberianPeninsula (ROJO ET AL. 2008).The fact that the site is now partial lycovered by the waters of the lake hasfavoured the extraordinary state ofconservation of organic remains madefrom plant matter, which range from thewooden posts in the buildings to thesmallest objects made or gathered by theoccupants. We can highl ight dozens ofwooden utensils, of types that arerepeated in different modules and sizes(BOSCH, CHINCHILLA, TORRUS 2006). Of al lthese, the most important are agriculturaltools (pointed sticks and sickles), bui ldingtools (adzes, wedges), hunting

instruments (bows, arrows, spears) anddomestic utensils (wooden bowls, basketsmade from aquatic plants, mixers, combs,spindles, spoons and spatulas).In the area nearest the shore of the lake,in the Cardial level, the archaeologicalexcavation has recovered hundreds ofposts that supported large huts, sunk inthe ground to depths of over 2 metres.Next to these, small posts or stakescorrespond to complementary structures.The archaeological layer hasaccumulated all the wood from thecollapse of the huts, including severalboards, fragments of ropes made fromplant matter and rol ls of l ianas used tohold together the different buildingelements (posts, beams and boards).Additional ly, fragments of oak bark havebeen found and these may have beenpart of the roofs or the floors of the huts.At the base of the stratigraphy, we believethat the large huts that surrounded thelakeshore stood a little above the originalground level (the lacustrine chalk). Thisbuilding method would have avoidedflooding if the level of the lake rose. I t isnot surprising, therefore, that no hearthsare found in this area (any there mighthave been inside the huts would havedisintegrated at the time of the collapse ofthe huts). However, we should note thataccumulations of charcoal andcobblestones affected by heat have beenfound, and these are the result of thehearths inside the huts being cleaned outwhile the huts were in use.In contrast, in the area furthest from thelake, as it is at a higher altitude, thepresent-day phreatic level l ies below thearchaeological level. Therefore, the onlyplant remains that have been preservedare the posts located 40 cm below thisarchaeological level. Since it was a much

Fig. 2: Capture of the last excavationseason.

59

drier area it was possible to build dozensof hearths directly on the ground, inhol lows that are ful l of charcoal andquartz and sandstone cobbles. Somestructures built directly on the ground, l ikethe two hut pavements made withtravertine flagstones, were aimed atinsulating the hut floor from the damp.Due to the good preservation of organicmaterial it has been possible to analysethousands of samples of plant remainsthat, together with fauna remains, provideexceptional information about thesubsistence of the first farmingcommunities. The population of La Dragaprimari ly practised agriculture and animal-husbandry, whilst hunting and thegathering of wild plants was a secondaryactivity. Cereals l ike wheat and barleywould have made up the main agriculturalbase, always accompanied by pulsessuch as peas and broad beans (BUXO,ROVIRA, SAÜCH 2000; ANTOLIN, BUXO 2011 ).The excellent state of conservation of theorganic matter al lows us to study aspectsof reaping techniques that cannot beinferred solely from the analysis of l ithicartefacts, which are all that are normallypreserved at most archaeological sites.Nevertheless wood material should bepreserved in order to stop the degradationprocess once out of the archaeologicallevel. In the case of La Draga the woodenimplements were lyophil ized (freeze-dried) and stored in a control led ambient(AGUER 2006).All this makes La Draga a privi leged placeto carry out subsistence, technological oreven experimental studies in order tounderstand these first Neolithicpopulations in the western Mediterraneanin greater depth. The excellent state ofconservation of the organic matter al lowsthe study of different aspects of

woodworking techniques that cannot beinferred solely from the analysis of l ithicartefacts.The aim of the experimental process is tounderstand the way those wooden toolswere made, how and why were theyused, and also difference between thedifferent types of traces (tool-marks anduse wear).In order to reach those objectives anexperimental program was developedbased on replicating the technologicalprocess of production of those woodenartefacts, and contrasting its functionalhypothesis. This paper focuses on theagricultural tools (pointed sticks andsickles), as well as the adzes, which arethe most abundant type of tools at thesite.

Study of the technological process of theagricultural tools of La Draga

Raw materials

The determination of the raw materials isthe first step in the study of the woodenartefacts. The aim is to determine boththe taxon and the anatomical part of thetree or bush used, in order to establishregularities and recurrences in whichcertain properties of the material weretaken advantage of (flexibi l ity, strength,resistance, . . . ), as well as how much effortwas invested in order to obtain thesupport. In the experimental program isvery important to work with the same rawmaterials as the ones that are found atthe archaeological site. In this way it ispossible to more ful ly understand theelaboration process and to contrast theefficiency of those materials for certainuses. The observation of those features inthe archaeological material have made it

60

possible to consider hypothesis about theprocess of obtaining these raw materialsthat have been used as a base for thedevelopment of the experimentalprogram.The raw material used in La Draga isdiverse; at least 1 7 taxa were used tomanufacture the wooden objects (BOSCH,CHINCHILLA, TORRUS 2006).

The artefacts’ fabrication

Once the raw materials have beenstudied, the next step in the analysis ofthe artefacts is to establish the fabricationhypothesis of those artefacts. Thehypothesis can be formed from observingthe different tool marks that can be seenon the artefacts’ surface and that havebeen produced during the different stepsof the elaboration process. Even thoughthe tool marks can be hidden because ofother tool marks or use-wear, enoughinformation can be obtained about thedifferent actions that have produced it.During visual inspection of thearchaeological tools from La Draga it waspossible to identify different kinds of toolmarks. Depending the size and shape ofthose marks, it was possible to distinguishbetween the ones produced during thesplitting process, the use of the adze, oreven the surface regularization processby polishing.The aim of this experimentation was toverify the hypothesis about the differentprocesses that produced those toolmarks. In order to do these two differentartefacts from La Draga site werereconstructed: a double pointed stick, andan adze. For this experimentation the kindof tools that can be found on thearchaeological site were used, such aswooden wedges of boxwood (Buxus

sempervirens), a wooden hammer and anadze with an oak handle and a hornfelsblade. These tools were specifical ly madefor the experimental work. Below are thedetai ls of the protocol fol lowed during theconstruction of the adze handle and thepointed stick.

Adze handle

First of al l the pine branch must be splitlongitudinal ly (Fig. 3). This was doneusing silex chips, wooden wedges and awooden hammer. Once the piece of woodwas split to an adequate size, the excesswood was eliminated with the hornfelsadze blade. The tool marks from this partof the process could not be observed inthe archaeological adze handles due tothe perfect polished finished that can befound on the surface of those tools. So,afterwards, this adze replica was polishedwith a piece of sandstone (PALOMO ET AL.2011 ).

Double pointed stick

The case of the pointed sticks is sl ightlydifferent, because of the large variety ofthis tool that are found at the site:

Fig. 3: Fabrication of an adze: snappingthe trunk.

61

examples with one pointed end and theother with a bevelled edge, only onepointed end, two bevelled edges, and soon. Even a wide range of sizes and theraw materials can also be found. I t wasdecided to replicate the elaborationprocess of one of the pointed sticks whichhad one pointed end and the other onebevelled. Moreover in these kinds of tools,a notch with unknown function can befound in the central part of the “body”.For this reproduction boxwood (Buxussempervirens) was used, because it is theonly species that was used for the double-pointed sticks. The same tools that wereused for the adze handle were also here:fl int chips, boxwood wedges and hammerand an adze.The branch was again spl it longitudinal ly.But in this case this process was repeatedat least three more times in order toobtain the adequate size. Afterwards thesurplus of wood was eliminated with anadze unti l the pieced reached the desiredsize and shape (Fig. 4). Fol lowing thearchaeological example, the ends werethen polished (PALOMO ET AL. 2011 ).The experimental work allowed theverification of the elaboration processhypothesis that was made from the

observation of the archaeological tools,and at the same time provided new dataabout the initial steps of the splitting of thewood (which are not so obvious in the toolmark). Likewise the amount of timeinvested in the creation of the tool couldalso be studied. Additional ly the wasteproduced by this work could also becharacterized, which could then be usedto distinguish between tools and workingwaste.

The artefacts’ function

Another of the factors that was to bestudied was the function of thearchaeological wooden tools. As stated,before their function was inferred fromethnographic and archaeologicalparal lels. However, the wood suffers acertain kind of wear, due to the use, theso-cal led use-wear. That’s whyexperimentation is fundamental in orderto understand the different actions thathave produced certain use-wear, andtherefore the function of those woodenartefacts. The aim of using anexperimental approach was to compareand contrast the most l ikely hypotheses.The pointed sticks of La Draga are one ofthe artefact types that needed to bechecked. At first glance, based onethnographic paral lels, it was thought thatthey were digging sticks, related with theagricultural task of removing the soilbefore sowing. Based on this hypothesisthe reconstructed pointed sticks wereused for this task: removing soil . Thisprocess not only had the aim to ascertainthe efficiency of those tools in this specificactivity, but also to obtain a good recordof the use-wear that this “removing thesoil” left on the active parts of the tool.

Fig. 4: Fabrication of a pointed stick:rough-hew.

62

Experimental work of removing the soil

This work was carried out on a 5x5 meterplot. The factors that were recordedinclude the time of use, and kind of workwhich was done. The main activity wasdriving the end of the stick into the soiland levering it up so it can be oxygenated(Fig. 5). One of the bi-pointed sticks washafted as a pick as well , using theaforementioned central nick. Both the useof the pointed sticks as a digging stick aswell as a pick (Fig. 6) turned out to bepretty effective. Although with the pick thework was faster, at the same time theimpact at the end was also abrupt,meaning that this tool suffered much moreabrasion that the other. This experimentalprocess of using the pointed sticks hasbeen also useful in order to obtain therequired reference collection of use-wear,which can later be used in comparativestudies of the archaeological tools.

The study of the tool marks and use-wearon the experimental tools

The technological, as well as thefunctional study of the archaeologicalwooden tools can't be performed without

the use wear analysis. The objects of thisanalysis are the marks and traces thatcan be found on the tool’s surface, whichwere produced during fabrication or use.This method has been considered as oneof the most objective approaches tocertain work processes so far. In thissense the use-wear and tool-marksanalysis contribute very valid knowledgeto the study of the production andreproduction social relations (CLEMENTE1 997).On the surface of the tools two kinds ofmarks and scratches can be found: thetechnological ones made during theelaboration process (tool-marks), and thefunctional caused by the erosion use(use-wear). I t is by the observation of thefirst type that we can infer what kind oftool was used to work the wood and inwhat way. And studying the second typeof mark (use-wear) establishes whatmaterial these archaeological tools wereused in and in which way.In addition to these types of marks, onesproduced by pre- and post depositionalprocesses must also be taken intoaccount (CLEMENTE 1 997). I f those are notwell recognized, they wil l hide or evendestroy the use-wear and tool marks.Furthermore, in order that this study canbe developed, it is necessary topreviously obtain an experimentalreference collection of marks and

Fig. 5: Use of a pointed stick to turn overthe soil.

Fig. 6: Pointed stick as a pick.

63

scratches. Experimentation in use wearstudies is the only way in which theinvestigator can establish a comparativedatabase so the hypothesis can be tested(GIBAJA 1 993).In agreement with those approaches themarks and scratches produced during theexperimental processes were recorded.These different features werecharacterized so that they can be used inthe future to identify similar marks onarchaeological remains. These include:- Tool marks such as splitting, adzemarks, scratches and sanding marks (Fig.7).- Use-wear such as fractures, flattenedareas, use polish, abrasion and alsoscratches (Fig. 8-9).

Problems in wood traceology analysis

Wood has different qualities than otherraw materials in which the use wear

approach has traditional ly been applied.That's why this methodology had to beadjusted for the peculiarities of wood. Thewooden elements from a waterlogged sitel ike La Draga are as much valuable asthey are weak, meaning that they mustundergo a certain amount of restoration(AGUER 2006). However, during thisrestoration process of the archaeologicalmaterial , some of the marks andscratches are distorted and becameunrecognisable. So the traceology has tobe executed before restoration.Manipulation of the fragile archaeologicalmaterial before its restoration has to becarried out very careful ly. In order to solvethis issue, work with 3D scanner anddigital reconstruction (Fig. 10) has begun,but before that si l icone and resin clayswere used.

Conclusions

1 ) The experimental process has beenused to test hypotheses about theproduction process of La Draga tools. I thas been possible to establish the

Fig. 7: Tool-mark: adze mark.

Fig. 9: Use-wear: fracture.

Fig. 8: Use-wear: use starches.

64

different steps of the wood workingprocess, as well as register the wastecreated during this process.2) I t was also possible to characterize thetool marks and use-wear from theexperimental tools, even when the marksleft from the first activities were erased ormasked by the superposition of laterones.3) The experimental approach proves thatthe technology and use study of thearchaeological wooden tools can be doneby the examination of the morphologicalfeatures, as well as the marks andscratches from the tools surface.4) The problems that the archaeologicalwood presents for its study can be solvedwith 3D scanning, or even with clayimpressions. Both methods provide acopy of the artefact with enough quality souse wear analysis can be applied.5) Additional ly it is worth emphasizing thatal l of these experimental processes canbe supplemented with other new proposalto the pedagogical project of the NeolithicPark from La Draga.

Literature

ANTOLIN, F., BUXO, R. 2011 : L’explotacióde les plantes: Contribució a la història del’agricultura i de l ’al imentació vegetal delneolític a Catalunya. In: A. Bosch, J.Chinchil la, J. Tarrus (Hrsg.), El poblatlacustre del neolític antic de La Draga.Excavacions 2000-2005. Girona 2011 ,1 47-1 74.AGUER, C. 2006: La restauración delsobjectes de fusta de La Draga. In: A.Bosch, J. Chinchil la, J. Tarrus (Hrsg.), Elsobjectes de fusta del poblat neolñitic deLa draga. Excavacions de 1 995-2005.Girona 2006, 1 63-1 66.BOSCH, A., CHINCHILLA, J., TARRUS, J.(Hrsg.) 2000: El poblat lacustre neolíticde la Draga. Excavacions de 1 990-1 998.Girona 2000.BOSCH, A., CHINCHILLA, J., TARRUS, J.(Hrsg.) 2006: Els objectes de fusta delpoblat neolític de la Draga. Excavacionsde 1 995-2005. Girona 2006.BOSCH, A., CHINCHILLA, J., TARRUS, J.(Hrsg.) 2011 : El poblat lacustre delneolític antic de La Draga. Excavacions

Fig. 10: 3D model of a pointed stick.

65

2000-2005. Girona 2011 .BUXO, R., ROVIRA, N., SAÜCH, C. 2000:Les restes vegetals de l lavors i fruits. In:A. Bosch, J. Chinchil la, J. Tarrus (Hrsg.),El poblat lacustre neolític de la Draga.Excavacions de 1 990-1 998. Girona 2000,1 29-1 39.CLEMENTE, I. 1 997: Los instrumentoslíticos de Túnel VI I : una aproximaciónetnoarqueològica. Barcelona 1 997.FUGAZZOLA, M., D’EUGENIO, G., PESSINA,A. 1 993: La Marmotta. Scavi 1 989. Unabitato peri lacustre di età neolitica.Bul letino di Paletnologia I tal iana 84, 1 993,1 81 -304.GIBAJA, F. 1 993: El como y el porque dela experimentación en análisis funcional.Revista de Arqueologia 1 48, 1 993, 1 0-1 5.GUILAINE, J., FREISES, A., MONTJARDIN, R.1 984: Leucate-Corrège, un habitat noyédu Néolithique cardial. Toulouse 1 984.PALOMO, A. ET AL. 2011 : Tecnologia ifuncionalitat dels objectes de fustaneolñitics de La Draga. Una aproximaciónexperimental. In: A. Bosch, J. Chinchil la,J. Tarrus (Hrsg.): El poblat lacustre delneolític antic de La Draga. Excavacions2000-2005. Girona 2011 , 89-96.ROJO, A. ET AL. 2008: Paisaje de lamemoria: Asentamientos del neolíticoantiguo en el Valle de Ambrona (Soria,España). Arte y Arqueologia 23. Valladol id2008.

Picture creditsFig. 1 , 6-1 0: Oriol LopezFig. 2: Antoni PalomoFig. 3-4: Raquel Piqué

AuthorsOriol López, Raquel Piqué,Antoni PalomoPrehistory DepartmentEdifici B, Facultat de Filosofia i LletresE-081 93 Bellaterra (Barcelona)