Beckmann S. — A Minoan Experimental House

Повний текст. OCR-шар PDF розріджений (літери розділені пробілами: t he, m ount ain) — при цитуванні звіряти. Нумерація ## PDF N.

Бібліографія: Beckmann S. «A Minoan Experimental House Paying Tribute to Middle Bronze Age Cretan Vernacular Architecture» // EXARC Journal 2016/4. — journal.exarc.net (members only).

Третя робота Бекман у vault; дві інші — Beckmann_Domesticating_Mountains_Appendices і Beckmann_Terebinth_Bronze_Age_Cult. Розбір → Eksperymentalnyi_dim_znannia_cherez_roblennia.

PDF 0

A Minoan exper im ent al house paying tribut e t o Middle Bronze Age Cret an vernacular archit ect ure

Prot ot ypes, sit uat ion, general descript ion

I n t he m ount ains sout h of Agios Nikolaos, nort h - east Cret e, t he Minoans of t he Middle Bronze Age ( 2000 - 1650 B.C.) left behind several kinds of ruins, which were st udied in m y PhD t hesis ( Beckm ann 2012a) . The 337 ancient sit es discovered during t his invest igat ion were arranged in a loose set t lem ent pat t ern, wit h dwellings ca. 150 m from each ot her, com posed of m anifold field- enclosures and anim al pen walls surrounding variably sized houses and connect ed by an ext ensive net work of pat hs. Especially im pressive are t he oft en m assive ruins of dwelling foundat ions referred t o ( for det ails see Beckm ann 2012a) .These foundat ions neut ralize t he oft en st eep slope incline t hey are built upon, so t hat t he original houses would have st ood in a horizont al posit ion. The set t lem ent pat t ern and surface findings indicat e t hat t hese inst allat ions were vernacular in charact er and probably housed m ixed agricult ural farm s ( ibid.) . Traces of upper st ruct ures ( especially t im ber) could not be det ect ed apart from t wo sit es where fragm ent s of what seem ed t o be t ypical Minoan ( albeit accident ally fired) m udbricks cou ld be seen ( Figure

. Consequent ly, I suggest ed t hat upper st ruct ures m ust have consist ed of unfired m udbrick. This st udy was based exclusively on surface findings, t herefore no analyses of any raw m at erials det ect ed exist .

To m y knowledge no reconst ruct ions of Minoan vernacular ( or ot her) houses have been undert aken yet . As a landscape archaeologist , I had no m ore experience in experim ent al archit ect ure t han having re- built part of a t radit ional Germ an farm house wit h a light kind of cob ( at t he t im e st il l a pioneering feat ) in t he 1980 .

Archit ect ure

Apart from t he preserved foundat ions of undressed m assive blocks ( Figure 2) , ca m ount ain clim at e ( snow and rain in wint er, heat in sum m er and wind year round) had left no upper st ruct ures. None of t he ruins show t races of window openings in t he som et im es high rem aining foundat ions, while few door openings ( ibid. 123) exist , t he lat t er always on t he upslope sides. The few preserved m udbrick fragm ent s seen in t he m ount ains were sim ilar t o t he unfired m udbricks known from excavat ed Minoan sit es such as Quart ier Mu in Malia, t he largest current ly known and excavat ed Middle Minoan urban sit e wit h preserved m udbrick st ruct ures ( cf. Poursat et al.

PDF 1

Figure 1: Accidentally fired fragments of mudbrick at a M iddle M inoan site.

1978, Poursat 1996 1 ) . These urban m udbricks wit h sizes of ca. 40- 50 x 20- 30 x 10- 15 cm - ( Figuress. 5- 7) were usually t em pered wit h organic m at t er ( st raw or sea- grass, depending on availabilit y) . The very sm all sam ples from m y st udy area did not show obvious inclusions of plant m at erial, alt hough t iny round holes/ voids m ight indicat e t he use of pine needles as t em per, suggest ing ut ilisation of t he pine forest close t o t he set t lem ent region. Ot her sm all, accident ally fired pieces of m ud seen on t he surface indicated larger plant inclusions ( Beckm ann 2012a: 135) and m ay have originally belonged t o bricks, plaster, or even roofing. The scarcit y of m udbrick findings also su ggest s t hat t here were no m aj or conflagrat ions t o end Minoan set t lem ent in t he m ount ains.

Figure 2: M iddle M inoan oncolithic site, goats standing on the horizontal level of the ruin, indicating steepness of slope.

1 For mudbrick in t he Aegean and Cret e in general cf. Guest -Papamanoli 1978, in M iddle M inoan M onast iraki cf. Como & M arazzi 2006, in Late M inoan M ochlos cf. Soles& Davaras 1994, in Palaikast rocf.Devolder 2006. Det ailed analyses in Nodaru et al. 2008. Please not e t hat t hese excavat ed sit es were all of urban nat ure.

PDF 2

Experim ent al Possibilit ies

Wit h a relat ively large num ber of available ruins for st udy, I could det erm ine t hat t he dwelling part of t he houses m ust have sat horizont ally on t heir levelling oncolit hic foundat ions, alt hough t he basic digit al equipm ent used in t his research did not allow t he slope t o be present ed ent irely accurat ely in a digit al reconst ruct ion. I f one t ries t o visualize t hat

plane, one can im agine how st eep t he slope is in t his exam ple ( Sit e 33, in Beckm ann 2012a: 299,542) .

While experim ent ing wit h t he available com put er reconst ruct ion possibilit ies, I realized t hat even t he m ost sophist icat ed soft ware would leave t oo m any quest ions unanswered. As a result , I decided t o at t em pt t o reconst ruct a sm all house in a corner of m y own ( part ly covered in wilderness) Cret an garden, som e7 km from ( and ca. 600m and m ore below) t he m ount ain originals. Alt hough we know of t he m udbricks exist ence, t he possible appearance of t he houses has been inform ed by

( Figure 3) .

Figure 3: Late M inoan house model from Archanes. (https:/ / commons.wikimedia.org/ wiki/ File:Tonmodell_Haus_Archanes_01.jpg )

I t is likely t hat t hese exam ples only show urban houses. Vernacular houses probably did not have a second floor wit h colum ns, and som e basic archit ect ural det ails can be ident ified from t hem , such as t he roofing wit h round wood beam s, t he exist ence of open court s, window const ruct ion, and t he possibilit y of balconies.

PDF 3

Preparat ions

The ext ent of experim ent al work for such a house nat urally depends on t he expect ed out com e. I n t his case, t he desired result was t o reconst ruct a sim ilar foundat ion t o t he Middle Minoan m ount ain ruins, along wit h a m udbrick st ruct ure, t o discover if/ how t he t wo could be connect ed. We were also int erest ed in det erm ining what im plicat ions t he challenging landscape had on const ruct ion - would specialized t ools and knowledge have t o be used on a slope?

Mudbrick A visual inspect ion of Quart ier Mu ( Figures 5- 7) provided im port ant inform at ion on t em pering cont ent , as considerable am ount s of sea grass ( Posidoniaoceanica) was used in t he bricks of t he building, a t ypical choice for a sit e on t he seashore ( cf. Guest - Papam anoli 1978: 6) . Much was also learned about building and posit ioning t echniques. For t he m udbricks we obt ained several cubic m et ers of local soil ( t erra rossa) which, aft er a first screening using a shake t est ( Weism ann& Bryce 2006: 45) proved t o be ideal for building. We t hen built a m ould m easuring 40x20x12 cm and produced sev eral t est bricks ( Fig ure 4) , em ploying various t em pers t o reduce shrinkage. I t was observed t hat added elasticit y m ay be gained t hrough t he use of organic t em per, an im port ant fact or in eart hquake- prone regions like Cret e.

Figure 4: Test bricks (from top left): straw, sand, plain, pine needles (ca. 4 weeks old)

For our experim ent several variet ies of bricks were t est ed ( Figure 4) :

Plain m ud: These broke int o m any sm all pieces aft er drying aft er t wo weeks.
Mud wit h ca. 30% sand: These also suffered several deep cracks wit hin t wo weeks.
Mud wit h locally sourced pine needles: These produced a st able brick 4

PDF 4

which rem ained dam p and soft aft er t wo weeks; t he needles had m inim al filling abilit y, so t he brick was com parat ively heavy. 4. Mud wit h ca. 50 volum e % wheat st raw : The st raw soaked up m uch of t he m ixing wat er wit hin hours, so t hese bricks were useable, alt hough slight ly pliable aft er t wo days in t he sun 2 .

Judging from t he rem aining exam ples it seem s probable t hat t he m ount ain Minoans m ay have used sand or a 3 sandier soil t han our com parat ively pure clay from a colluvial lowland polj e, rat her t han st raw. This seem s a logical choice when t aking int o account t hat st raw m ight have been a precious com m odit y in m ixed agricult ure. St ill, being sit uat ed in t he lowlands, we decided t o use st raw as did m ost of t he analyzed Bronze Age cases in Nodarou et al. ( 2008) .Theoret ically pine needles would also be a good choice for t he m ount ain surroundings, according t o t he usually observed principle of local availabilit y ( ibid.) .We t ried t o apply t hat principle as oft en as possible, t o avoid various labour- int ensive feat ures as t his was a privat ely financed experim ent ( see m ainly t he paragraph on t im ber below) . We have since discussed if anim al hair such as sheep or goat would be an opt ion as addit ional t em per. I n a shepherding societ y it would also be econom ically j ust ifiable as an am ount of discarded wool is always left over from shearing.

Subsequent ly, we chose a m edium of t he known brick sizes from all over Minoan Cret e of 40x20x12cm . This allowed t he possibilit y t o change direct ion in t he laying of t he bricks ( header- st ret cher) , alt hough t he Quart ier Mu rem ains were inconclusive in answering how and when header- st ret cher direct ions were changed ( Figures. 6 and 7) . My observat ions suggest t hey m ainly used pure header const ruct ions, while bonding was obviously oft en done wit hout paying at t ent ion t o where t he perpend ( t he vert ical j oint bet ween t wo bricks) was posit ioned ( Figure 6) , even t hough it should have been obvious already in t he Bron ze Age t hat aligned perpends produce com parat ively unst able walls. I t should be int erest ing in t his cont ext t o com pare t hese Middle Minoan const ruct ions wit h any available Lat e Minoan const ruct ions t o see if eart hquake experiences im pact ed on t his observed indifference. While our brick t hickness of 12 cm was great er t han t he exam ple known from t he m ount ains ( ca. 50% m ore t han t he ca. 8 cm t here) , it was a m ore econom ical solut ion in t erm s of m anpower. Wit h t he sam e num ber of m ovem ent s in laying bricks, 50 % m ore wall height is reached - and for us, m anpower was by far t he m ost expensive elem ent in t he experim ent .

2 Depending on t he amount of plant t emper used, t he weight and temperature isolat ing abilit ies change nat urally. I adjust ed t he mixt ure t o a macroscopically roughly similar one wit h t hat visible in M alia Quart ier M u cf. Fig. 5-7. 3 For t he directly used, not mixed charact er of soil in M inoan mudbricks see Nodarou et. al 2008.

PDF 5

A sim ilar effect m ay have been achieved by m aking t he bricks as large as t hey were in Malia, alt hough bricks of t his size would have t o have been m uch drier before building, so as not t o fold when lift ed.

Figure 5: M udbricks in Quartier M u, M alia (the beams here as floor base for the second floor - are reconstructions filling the voids remaining in the mud parts of the ruins). Headers and stretchers seem to be randomly mixed.

Figure 6: M alia Quartier M u mudbrick wall with several perpends aligned.

Figure 7: M alia Quartier M u mudbrick wall with offset perpends.

PDF 6

The use of st ones t o fill gaps was a feat ure we adopt ed from t he Middle Bronze Age exam ple in Malia, as it t urned out t o be m uch easier t han t rying t o use m ore m ort ar, which was oft en im practical, especially when t he com parat ively fluid m ort ar would j ust run out of larger gaps. I n ret rospect , had we had t he luxury of drying t im e and adequat e space, t he pine needle t em per bricks would have been m y preferred building m at erial, due t o t he durabilit y of t his version. Aft er t wo years of cont inuous exposure t o t he Cret result ed in t he following:

Sand: Surface rounded by erosion but m ain m ass part ly int act , alt hough wit h now widened cracks.
Pine needles: Surface slight ly eroded, whole shape st ill int act , no cracks.
Plain: Nearly t ot ally dissolved and hardly recognizable due t o t he m any cracks t hat all enlarged t he erodable surface.
St raw: About 0.5cm of surface eroded so t hat t he st raw t em per is m ore visible, general shape int act ( no cracks) .

From t he com parat ively flat shape of t he Minoan m ud bricks we concluded t hat Minoan builders m ust have had a longer drying period t han we used ( t wo days was enough in our case) , and t hus harder bricks. This m ust have had ot her consequences, such as t he need for great er am ount s of m ort ar t o be used, as t he bricks are less adapt able t o already built , uneven layers underneat h, a fact t hat can be seen in Figures 5 and 6, where t hick layers of m ort ar are obvious, t oget her wit h t he regular use of st ones as chinking in larger gaps.

As a base for m ort ar we used different ly coloured clay t hat was recom m ended by a specialist in clay m ort ar ( m ainly baking ovens) due t o it s except ional qualit y. Geologically it is a calcarious m arl ( for t he geological det ails of Cret an soils see for exam ple John A. Gifford in Shaw & Shaw 2014: 59) , and called kouskouras or asprochom a Cret e. I t consist s of a whit ish, clayey, very dense and nearly t ot ally wat erproof when wet kind of soft st one.I n sit u our version seem ed t o be h practically it is densely packed clay t hat when passed by a t ract or wit h a plough t urns int o fine dust in sum m er or deep m ud in wint er. I t was sourced som e 12 km from our sit e, and offered t he advant age of clearly seeing t he difference bet ween t he basi c, red- brown bricks and t he beige m ort ar and plast er ( Figure 8) . An addit ional advant age is t hat t he various elem ent s and procedures of building can be dem onst rat ed m uch m ore easily.

Tim ber and ot her raw m at erials While we know from Lat e Minoan urban sit es like Knossos t hat cypress t im ber was used ( Evans 1921: 341) , t his seem s im probable for t he building of vernacular houses, especially as cypress has always been a 7

PDF 7

valuable export product ( Chaniot is 1999: 209, wit h bibliography) . Et hnographic parallels show t he logical choice of t im ber from what ever wood is locally available. We did not experim ent ally prepare any wooden elem ent s ourselves as t he locally available t im ber is m ost ly kerm es oak ( Quercus coccifera) or olive ( Olea europaea) ; bot h ext rem ely hard, heavy kinds of t im ber, which, if worked in a Bronze- Age- like m anner, would have surpassed our work- hour schedule and hence our budget considerably.

The raw m at erial we want ed t o use for t he lat er roof covering, giant cane ( Arundodonax) , cam e from t he coast , som e 5 km from t he sit e, but could have grown closer at ot her t im es. I t is a bam boo- like plant t hat has been used in archit ect ure in t he Bronze Age Aegean ( Thornt on 2007: 77, Vlachopoulos&Zorzos 2014: 194) and seem ed like a good choice, even t hough we do not know m uch about Minoan roof const ruct ions. There are som e sm all fragm ent s of accident ally fired m ud wit h plant im pressions ( m ainly sm all irregular t wigs) in t he surroundings of t he m ount ain sit es, as m ent ioned above, but t hey are t oo sm all t o est ablish what part of t he const ruct ion t hey m ay have belonged t o. I n any case, Bronze Age roofs m ust have been part ly m ud, alt hough hardly anyt hing can be learned from Middle Minoan sit es.S as t hey are usually called in t he bibliography, only Vlachopoulos&Zorzos call t hem Arundodonax based on phyt olit hs - on t op of t he beam s in t he reconst ruct ion inFigure 5, based on plant im pressions in t he m ud.( For t his and ot her possibilit ies cf. also Shaw 1977: 229 and Passim .) These bam boo- like grasses were put aside t o dry unt il t hey were t o be used. We did realize t hat t ransport ing all t h ese m at erials on donkeys and hum an backs would also m assively effect t he m an/ donkey- power needed. Bronze Age builders would have m ost likely, wherever possible, chosen m at erials from as close by as possible

Anot her im port ant com ponent needed for m aking m ud bricks is wat er. While m odern Cret an condit ions bring wat er via net works t o m ost set t lem ent s, t he m ore or less free use of wat er during t he Bronze Age would have be Warm and sunny weat her is needed for t he bricks t o dry, and t he roof cannot be const ruct ed as long as t he bricks have not been laid.Therefore, t he quest ion of t he t im e of year when building would h ave t aken placerem ains unclear.

Beginning work: t he foundat ion

The first st ep of building is nat urally t he choice and preparat ion of t he space for t he fut ure house. As could be expect ed for m ost cases of Minoan Bronze Age plot s, cut t ing t rees and bushes occurred aft er choosing t he space, which sim ult aneously produces a cert ain am ount of t im ber. I n our case, as m ent ioned earlier, we did not use t his t im ber but bought ready cut im port ed t im ber for econom ic reasons, such as t he 8

PDF 8

insert ion of t he horizont al lint el int o t he upper layers of t he brick wall over Figures 8, 22 and 24.

Figure 8 amount of dissoved by rain bricks visible in the darker stains on the beige surface (produced during closing perpends). The foundation stones are not filled in.

We also t ried t o preserve as m any of t he surrounding t rees on t he plot as possible, t o prot ect t he house f rom driving rain. This t urned out t o be a

a prot ect ive t ree, st ill was not prot ect ed enough, as lat er developm ent s proved. We also chose a place wit h as m any large rocks as possible, so as t o be able t o build our foundat ion wit h m inim al effort s. The space cleared of big st ones on t he upslope side of t he house, dem onst rat ed clearly why and how m any of t he m ount ain houses act ually seem ed t o have had a court on t he upslope side. This side is in m any cases also t he lee- side in t erm s of wint er weat her condit ions, which gives t he space an ext ra qualit y for use as an out door act ivit y area. I n our experim ent al case t he downslope side is t he weat her side and so event ually required prot ect ive m easures t o be im plem ent ed for t he exposed wall.

We used t wo kinds of t ools for our foundat ion, a big sledge ham m er t o clear t he bedrock part s which had t o be broken off, and t hen a lever for m oving t he blocks t o t heir final posit ion. For a few days we em ployed only one worker, but discovered t hat t wo workers were t he m inim um requirem ent , as t oget her t hey could m ove m any of t he big blocks ( Figure 9) .One working individual m ost cert ainly required a lever, a t ool which would have been difficult t o produce, ev en if using fire- hardened hardwood, before t he use of iron. The weight of t he st ones we m oved would have broken m ost wooden levers, while wit h t wo ( or m ore) workers

PDF 9

t he need for a lever was m uch sm aller. The quest ion of what m ight have been used in t he place of a sledge- ham m er in Bronze Age Cret e rem ains open for debat e, as all ham m er- like t ools t hat I know of would be t oo sm all for t his t ask. Nat urally uncont rolled cracking of bedrock by fire ( usually only used t o get rid of bedrock) or cont rolled cracking by wooden wedges swelled wit h wat er, as known from et hnography, would have been a possibilit y, t oo. Bedrock, frequent ly int egrat ed int o Minoan m ount ain foundat ions, gives t he buildings great er st abilit y, especially in t he case of eart hquakes.

Figure 9: M oving oncolithic blocks from the upper level of the later foundation. Upslope working does not seem like an option.

Having erect ed t he fram e of t he foundat ion as m uch as possible in t he Middle Minoan way ( Beckm ann 2012a, chapt er 2, Beckm ann 2012b: 36- 37c) , we filled t his wit h sm aller st ones. Som e m ount ain houses m ay have used t he space inside t he foundat ion as a cellar vel sim ., while ot hers were filled t o creat e a level base for t he upper st ruct ures ( ibid., and Figure 10) .

We realised t hat rainwat er running over t he surface above t he house would inadvert ent ly lead t o an event ually unst able upslope wall during t he rainy season. Therefore we cleaned about a m et er’s breadt h of soil from t he upper foundat ion and filled t hat space wit h rubble. Then, on t op of t hat , we added several layers of wat erproofing clay wit h an out ward 10

PDF 10

incline ( Figure 10, under t he yellow bucket ) . I nt o t he final layer, const ruct ed for a clean surface only aft er t he whole house was finished, we insert ed wit h som e 20cm dist ance from t he final, plast ered wall a shallow drainage channel ( m ade from ridge- t iles) which has funct ioned perfect ly since t hen ( Figure 27) . Drainage channels like t his are known from various ancient sit es ( see t he list in Shaw 2004: 173 - 4, n4) .

Figure 10: Filling the oncolithic foundation with rubble and the upslope side with waterproofing clay over rubble.

The realizat ion of t his problem prom pt ed m e t o ret urn t o one of t he unexcavat ed prot ot ype sit es in t he m ount ains, where an unaut horized bulldozer had scraped flush along t he ruin wall and exposed t he upslope part of t he filling behind. Short ly aft er t he bulldozing I had discovered t hat t here were several st one slabs em bedded in t he soil close t o t he m odern surface and hence probably not exactly in t heir original posit ion. At t he t im e, I had considered t hese t o have been t races of an ancient flagged t errace or t he like. When revisit ing t his sit e, unfort unat ely m ost of t he originally clear- cut edge was gone and only one slab rem ained, so I revert ed t o m y original, older phot ographs 4 . I could see now t hat t he slabs m ust have been part of a const ruct ion divert ing wat er from t he house wall ( Figure 11) . A good exam ple t o show of how well suit ed experim ent al archaeological w ork is t o inform ing us about underst anding ancient feat ures. 4 This was an occasion t o enjoy t he exist ence of digital phot ography which made it possible for me t o have now 14.000 phot ographs, which have at least virt ually preserved condit ions since t hen gone forever on several occasions already.

PDF 11

Figure 11: Upslope part of a M iddle M inoan ruin accidentally cleaned by bulldozer (2008). The slabs in the soil right of the big foundation blocks near to the modern surface must have been part of an ancient drainage construction (channel?), may be even in two phases. Note also the filling of small rubble (for drainage as well?) underneath and the deep-reaching foundation blocks under the building’s corner (at walking st ick, 90cm high for scale).

Mudbrick const ruct ion

Aft er finishing t he st one- building sect ion up t o a height w here splashing wat er from wint er rains would not harm t he m udbrick const ruct ion easily ( one big block layer above t he surface, as in t he m ount ains) , t he whole raised plat form was levelled wit h layers of fine rubble- sand- m ud t o creat e a good base for t he lat er floor. This kind of com posit ion can also be seen on t he broken edges of floors in Bronze Age Malia Quart ier Mu.

We st art ed producing t he m udbricks ( which had been decided upon aft er reviewing t he t est bricks) , by filling a wooden m ould wit h a m ixt ure of 50% red clayey soil and 50% wheat st raw, wit h as m uch wat er as was needed for a sm oot h, even consist ency where all t he st raw was well covered in m ud. I n t he beginning, we produced t his m ixt ure in t he t radit ional way by t reading wit h our feet . Aft er t he first bat ch t he workers suggest ed using a cem ent m ixer for t he clay- wat er m ix, t hen pouring t he m ix over t he st raw in a wheelbarrow, m ixing it by hand and t hen t ransport ing t he wheelbarrow t o t he space where t he m oulding and drying was carried out ( Figure 12) . Aft er t wo days, t he m udbricks were dry enough ( while st ill agreeably pliable) t o be used in building ( Figure 13) . Aft er a few days, we had reached a rhyt hm of building, where one bat ch of bricks ( ca. 90- 100 pieces produced by t he t wo workers in one work session) would dry out for t wo days and t hen be ut ilised in const ruct ion. 12

PDF 12

Following t his a day of pause for producing m ore bricks was also adv isable t o give t he built bricks m ore t im e t o dry and t hus not be deform ed t oo m uch by t he weight of t he following layers. For t he m ort ar we used a m ixt ure of kouskouras in t he sam e m ixing rat io as t he bricks, but freshly m ade every day and t hus m ore fluid for inst ant building. The st raw for t he m ort ar was chosen from short er pieces t han t he brick m ixt ure t o im prove it s workabilit y. As an easy Bronze Age parallel t he use of m ud m ort ar in Malia Quart ier Mu is well visible e.g. in Figures 6 and 7.

Figure 12: Finished mudbricks with the mould (centre right) and its lid that allowed for easier pressing of the brick from the mould (on bottom right brick).

Figure 13: The first layer of two day old mudbricks, using kouskouras-based mortar mixed with finer straw.

The first layer of bricks was laid wit hout furt her bonding ont o t he upper 13

PDF 13

layer of rocks, equalized by plent y of m ort ar and a few filling st ones where necessary. I t seem s t hat t he rough surface of t he un - worked blocks, t oget her wit h t he weight of t he wall, produced am ple cross- bracing t o m ake t he connect ion st able, as t he first eart hquake showed about six m ont hs aft er t he house was finished. At 6.1 on t he Richt er scale, it was st rong enough t o produce a slight crack on t he bui weakest part , over t he door, but did not cause cracks elsewhere 5 .

I n t he following t wo layers of brick, som e of t he inequalit ies rem aining from t he first layer were adj ust ed where necessary, wit h m ore m ort ar and a few ext ra st ones ( as also visible in t he Bronze Age wall in Malia, Figure 6) , which were also used t o fill ot her gaps ( Figure 14) .

Figure 14: M udbrick wall with kouskouras mortar using additional stones to fill gaps.

The layers were changed bet ween header and st ret cher rows and perpends were m ost ly offset t o increase bonding. Due t o an oversight , one sm all sect ion of t he wall was built wit hout offset , ( sim ilar t o t he part Figure 16) , producing a m aj or argum ent bet ween t he t wo workers who want ed t o rebuild t he whole part of t he wall. I insist ed on not rebuilding, as t here are clear exam ples t hat t he Minoan builders e.g. in Malia also had not always t hought t o avoid vert ically cont inuous perpends ( seeFigure 6) . Aft er plast ering, none of t his rem ains visible and t he eart hquake did not produce any crack in t hat area

5 A similar (but more severe) damage appeared in t he cob experimental house of the St anley Park Eart hen Archit ect ure Project in Vancouver BC at 7.2 Richter, also at first over t he door (cf. ht t ps:/ / www.yout ube.com/ wat ch?v=ChbccUQhpJc at minute 5.00, last seen on 29.8.2016). Cf. also Dimiziania and Ort olani 2014: 6. 14

PDF 14

of t he wall. This was part icularly int erest ing as it seem s t he const ruct ion wit h st ill hum id and pliable m udbrick m ight im prove t he m ort ar - block int erface and t hus t he shear st rengt h, as t he walls becom e m ore solidified.

Figure 15: M udbricks with cut edges to provide a better bonding with the door casing.

We had erect ed t he door casing from t he beginning, so as t o be able t o bond it wit h t he bricks as t he layers were rising. Header rows were adj oined t o t he fram e whereas in st ret cher rows t he bricks were cut on t heir ends in order t o fit int o t he space bet ween t he inner and out er part of t he fram e. Som e wooden pieces, like t wigs or st rips of cut t im ber, were also insert ed t o im prove t he connect ion ( Figure 15) We know t hat t he Minoans oft en const ruct ed t heir doors wit h a sim ilar wooden casing ( for Malia cf. Poursat and Schm id 1996: 74) , as even t hough t he t im ber has vanished, t he hollows oft en rem ain. However, I am not aware if t hey did or did not use t his ext ra reinforcem ent .

The Minoans seem t o have regularly included a rat her peculiarly shaped window in t heir houses, as observed in t wo of t he walls ( wit h four and six openings) of t he Lat e Minoan house m odel from Archanes ( Figure 3) .

The workers ceased building layer by layer aft er t he first few rows, and inst ead raised som e part s higher t han t he rest , creat ing a st epped appearance ( Figure 16, 17) . One evening when I revisit ed t he const ruct ion sit e aft er everyone was gone, I not iced t hat t hey had left one loose brick lying on t op of t wo ot hers which had som e space in bet ween, next t o one such st epped wall- part . The sim ilarit y in pat t ern wit h t he 15

PDF 15

Archanes m odel’s windows was st riking, and I realized how logically t his kind of opening could be int egrat ed int o such a brick wall wit hout large dist urbances t o t he building process, and wit hout giving up t he st abilizing or weight - bearing qualit ies of t he brick const ruct ion. Originally I had not planned t o include any windows in t he const ruct ion of such a sm all house

following et hnographic exam ples only a sm all opening for cross vent ilat ion was scheduled.

Figure 16: M inoan style “window” created by spacing a single row of bricks.

Figure 17: Raising the mudbrick walls -header and stretcher alternating- but not by layer. This first rain came early in 2014.

PDF 16

On t he next day we const ruct ed, in t he widt h of a window and on t he last even header- layer underneat h, a single out er row of st ret cher bricks, leaving an opening of half a brick in t he m iddle, covered by anot her st ret cher wit h quart er- brick openings on bot h it s sides. This was again covered by t wo st ret chers spaced wit h a half brick dist ance in t he m iddle ( Figure 16) . On t he out side t his produced a level wall wit h openings, and on t he inside it appeared as a recess in t he wall of a whole window’s size of 1.5 bricks widt h and 3 bricks height ( Figures 16, 17, 24) . This was sat isfact ory, even t hough som e of t he perpends t urned out again aligned ( Figure 16) , but aft er t he earlier argum ent one of t he workers cam e early t he next day knowing t hat I would probably not want t o change t hings - and reconst ruct ed on his own t he whole t hing wit h correct ly offset perpends. Subsequent ly, const ruct ion cont inued sm oot hly for a day and t hen t he unexpect edly early first rain cam e ( Figure 17) , while t he window- const ruct ion was only spanned by t he t im ber lint el, t o bet t er spread t he weight , t hat hadn’t been int egrat ed int o t he m udbricks yet . Alt hough everyt hing appeared a sea of m ud aft erwards, t he actual brick surfaces had hardly suffered at all ( Figure 18) . This provided a first hint at how am azingly resist ant t his m at erial seem s t o be t o rain, even when st i ll soft and wit h it s surface exposed - som et hing t hat would never happen again in t he fut ure, as long as t he roof would rem ain int act .

Figure 18: Built mudbricks after the first strong rain. The surface shows only some clean-washed straw haulms.

PDF 17

Wit hin less t han a m ont h all t he brick walls were com plet ed and only t he upperm ost layers needed som e ext ra work due t o t he incline we want ed t o give t he roof. I realized t hat m udbrick- const ruct ion wit hout a t im ber fram e is a m uch fast er kind of building t han I had expect ed,despit e having addit ionally produced our own building m at erial pin t andum wit h t he building process ( all in all over 900 m udbricks) .

Figure 19: The basic roof construction - beams bedded in cob with a ringbeam on top. The spaces underneath are later filled with cob.cc

The roof

As st at ed earlier, t here is a paucit y of knowledge as t o how Middle Minoan roofs were const ruct ed. Alt hough t he Lat e Minoan houses of urban Akrot iri on Sant orini are preserved t o an am azing height , archaeologist s know lit t le about t heir original roofs 6 , from t he layering t o t he shape. Usually Minoan houses are supposed t o have been flat , an assum pt ion which m ay have t o do wit h t he shape of t radit ional Cret an houses as m uch as t he t races known from floor const ruct ions in m ult i - st orey Bronze Age urban houses ( Figure 5) . The only cert aint y is t he absence of any kind of t iles or slabs covering t he Minoan m ount ain dwellings, as not a single building of

6 A very good synopsis of what we do know has recent ly been achieved by Lanuzza, 2013, an art icle I unfort unately wasn’t aware of when we built t he experiment al house, alt hough it would not have changed our decisions.

PDF 18

t he 337 ruins exam ined showed t races of t iles. I t seem s as if t he few published cases of roof layering irrespect ive of t he roofs’ slope have a sim ilarit y wit h t radit ional Cret an roofs ( Lenuzza 2013: 80) as visible in Figure 21.

We decided t o follow t he general lines of t radit ional Cret an houses, wit h a slight incline, which would allow rainwat er t o run off easily. This is also support ed by t he exist ence of roof drain spout s in Bronze Age Cret e ( Lenuzza 2013: 84) . These could not have funct ioned wit hout a slight incline t o t he roof, while relat ive flat ness would st ill allow a use of t he roof- space for various activit ies. We chose t im bers sim ilar in size t o t hose

corresponded t o t he t ype of t im ber known from t he voids visible in t he Malia Quart ier Mu houses ( Figure 5) .

The local t im bers would have been rat her crooked, a fact which corresponds wit h t radit ional Cret an vernacular building, where everyt hing wooden available or dispensable due t o a general sparseness of t rees would have been used for const ruct ing roofs. Oft en whole t runks of olive t rees, com plet e wit h branches, had t o serve as t he m ost st able t im ber for beam s as well as for colum ns ( where needed) . A rat her neat exam ple in t he m ount ains, and t herefore likely m ade of kerm es oak, can be seen in Figure 20. Many houses, depending on t heir geographical posit ion, had t o use olive for t im ber, and only old, non - product ive t rees would be sacrificed, gnarled by decades of pruning, oft en producing bizarre roof const ruct ions.

Figure 20: Traditional Cretan roof construction - the timber is here probably oak (soot-blackened).The walls are rubble built with mud mortar, mud plaster and lime paint.

PDF 19

Because of t his crookedness a vernacular house probably would not have had t he beam s as closely set as t hey were in urban Quart ier Mu ( Fig ure 5) . Also, t he ext rem ely heavy kerm es oak beam s in t his dense packing would bring a large ext ra weight t o t he whole const ruct ion. Thus, and for econom ical reasons as m ent ioned above, we used im port ed pine t im ber. Sim ilar t im ber m ight have been used in t he Bronze Age m ount ain houses as t hey are part ly sit uat ed in a pine forest t hat m ay have already exist ed 4000 years ago.

Due t o t he sm all size of our experi m ent al house, colum ns would not be needed, and aft er consult ing wit h m y fat her, I ng. G. Beckm ann , a st ruct ural engineer, concerning t he load capacit y of our t im ber in relat ion t o t he t hickness of roof and it s t ons of weight , t he respect ive t im bers were laid int o a m ud bed. This is one of t he few Bronze Age roofing feat ures act ually known from Cret an archaeology ( Fig ure 5) . The weight of t he roof was t o be dist ribut ed over t he upper ( front - door) and lower walls of t he house via a ring- beam and over t he side walls of t he house via t he 2 m ain beam s addit ionally spanning t he int erior. As a m udbrick wall t hickness from 30cm is load bearing ( Dim iziania and Ort olani2014: 12) , no vert ical colum ns wit hin t he walls are needed. St ill , from a st at ic point of view, vert ical t im bers would probably im prove t he eart hquake resist ance of t he building. This m ay have been one reason why vert ical t im bers were int roduced in Lat e Minoan archit ect ure ( Tsakanika-Theohari 2009) .

Figure 21: Layering in a traditional Cretan roof (currently crumbling). All the visible layers were at some point topped with cement which preserved the stratigraphy usually lost.

PDF 20

Our t wo large m ain beam s were st abilized wit h st ones on bot h sides, while t he ring- beam was int ended t o ascert ain t he horizont al st abilit y of t he roof, always bearing in m ind t hat Cret e has frequent eart hquakes. Anot her t ypical feat ure of m any t radit ional Cret an houses is a slight overhang which we decided t o adopt in order t o give t he walls som e addit ional prot ect ion from rain. As in t he roof const ruct ion we were not t rying t o be t rue t o any Bronze Age prot ot ypes, we did not work wit h wooden pecks or bindings in leat her as an ancient building would have em ployed. I nst ead we resort ed t o t he m odern solut ions of binding wit h wire, nails and screws, which was cheaper in t erm s of working hours.

The act ual layering on t radit ional Cret an roofs varies great ly wit h t he local availabilit y of m at erial. The basics are t hus: load bearing t im ber beam s are covered wit h sev eral increasingly finer layers of plant m at erial ( e.g. branches, t wigs, leaves; or branches, brushwood, sea grass) covered wit h a layer of soil and/ or finally a layer of dom at ochom a region has it s own source, which t he elderly locals st ill always know; it is usually t he m ost wat erproofing clay rich soil available in t he area. I n som e cases t his layer was only t hinly applied ( Figure 21) , a t ask t hat oft en had t o be repeat ed every wint er, which result ed in m any layers. I nt erest ingly som e of t he known fragm ent s of Minoan roofing show exact ly t his kind of layering ( Lenuzza 2013) . We used, as m ent ioned in t he beginning, densely arranged haulm s of t he bam boo- like giant cane on t op of t hin t im ber beam s, t aking t he place of t ree branches in t radit ional archit ect ure ( Figure 22) . Branches are also t he only solut ion when beam s are m ade of knarled t rees, as giant canes need a horizont al base.

Figure 22: Thin beams and giant cane as first roof layers over the main beams. 21

PDF 21

The following layer of t hin woven clot h was added t o prevent dust from falling t hrough t he slit s in t he cane, som et hing t he ancient and t he t radit ional Cret ans m ay have solved wit h a layer of overlapping leaves ( Devolder 2006: 80) . As I int ended t o use t he ex perim ent al house lat er as st orage space or even as an occasional guest room , I decided t o m ake sure in t his case.

Next , several layers of roof- soil were laid on one anot her, first dry ( t o m ake t ransport / application as easy as possible) and t hen wet and com pressed, according t o local cust om ( Figure 23) , unt il it reached a t hickness of ca. 20cm .

Figure 23: Compressing the last roof layer.

By t his t im e t he rainy season had arrived, so unfort unat ely t he roof never really dried off in t he first wint er. St ill it obviously developed cracks albeit not visible on t he surface so t hat each rain caused ext ensive dripping inside t he house. Drip- holes7 also appeared in t he applied basic floor layers. When expressing m y disappoint m ent , and asking locals what we had done wrong, t he first wise old lady looked at m e wit h a grin and

Furt her research int o local roofing cust om s st ill have not led t o a conclusive solut ion. When t he next rainy season com m enced, t he roof leaked again and we changed t act ics. We t ook down som e 5 cm of t he

7 It might be int erest ing t o st udy if any such drip-holes are visible in known archaeological floors.

PDF 22

upperm ost clay, t he layer we had already changed several t im es, applied a t hin polyt hene film and covered t hat again wit h clay. At t he sam e t im e we set up 4 fram es of ca. 1.2x1.2m wit h various roof layering variat ions t o see if it was possible on t his sm aller scale t o find out which roof solut ion would be t he best . This t est is ongoing as t he 2015 - 2016 wint er had very lit t le rain, and we shall cont inue it unt il a possible next experim ent al house is built ( see below) .

From som e Bronze Age roofs we know t hat a drain spout was probably insert ed in t he edge of t he clay const ruct ion ( Shaw 2004, Lenuzza 2013) . Aft er t he first heavy rains, and before t he last roof layer was com pressed, we also insert ed a spout int o t he lowest point of t he roof so t hat t he wall below was not soaked and t he int ense flow speed could not erode t he clay any m ore.

The plast er

When t he roof was nearly finished, wit h t he last layers applied parallel t o working on t he plast er, t he roughcast was applied. I t was based on a sim ilar m ixt ure as t he bricks, alt hough inst ead of st raw we used chaff t o prevent t he surface from becom ing t oo rough. When t hat had dried, t he final rendering was applied, again cont aining chaff, but in t he finest grade available, wit h som e added lim e t o give it a harder surface. The cracks in t he roughcast were welcom e in t his case as t hey provided a bet t er hold for t he final rendering ( Figure 24) . Bot h t he int erior and ext erior of t h e house were t reat ed in t he sam e way.

Figure 24: Roughcast and final rendering of the plaster.

PDF 23

The chaff provided us wit h an ext ra sm ile as aft er about 10 days, t he seeds rem aining in t he fine st raw st art ed t o sprout and covered t he walls in a t hin green cereal field ( visible in Figure 27) . Finally, we paint ed sev eral walls wit h a t radit ional ( also present in t he Bronze Age) m ixt ure of lim e and m ud, t he lat t er m ainly for t he colour, but also for ext ra prot ect ion from hum idit y ( Figure 25) .

We t reat ed t he surface of t he st airs leading up t o t he ent rance in a sim ilar way , as I had not iced t hat in Malia Quart ier Mu som e out door st airs had also been covered in what looked like a clay surface. Subsequent ly, I added sand t o t he surface w hen wet , as t he clay kept t urning int o st icky m ud, but wit h t im e - and sand - t he m ixt ure im proved. St ill I believe now t hat t he Quart ier Mu st eps m ay well have had a plast er surface heavy in lim e and were only coloured t o brownish by t he m any years of use and by being covered deep under brown soil for cent uries.

A set back or was it ?

When we finished t he house we were well int o t he rainy season, and by Christ m as 2014 t he roof and plast ered walls had sprout ed. As m ent ioned t he dripping from t he edge of t he overhang above t he door produced holes in t he clay ent rance plat form , so I im provised wit h a few flat pebbles, set where t he drip- holes were ( Figure 27) , suggest ing how st ones foreign t o t he surroundings of Bronze Age houses m ight t urn up in t he m at erial record of t heir surroundings.

Figure 25: The weather side wall after finishing in fall 2014.

PDF 24

The rainwat er falling from t he spout high above t he ground splat t ered back t o t he wall and left it dirt y and wet , so I applied an old t r ick used in sout h Germ an farm houses, where a heap of sm all st ones would be placed underneat h t he spout t o prevent t he wat er from splat t ering back up anot her feat ure I would not have underst ood during an excavat ion unt il t hen.

The m ost dram at ic dest ruct ion caused by weat her condit ions cam e in January, aft er t he first snow, when heavy, driving rain and hail hit t he st ill slight ly m oist surface of t he west ern side and a whole slab of plast er roughcast and all fell off t he wall. The originally prist ine lim e- paint ed surface was pocked wit h sm all holes from t he hail just as t he unpaint ed ot her half of t he wall. This was t he point where we decided t o put t o use t he ext ra sm all beam s which we had insert ed int o t he t op of t he walls for j ust such an occasion, as I had t heorised t hat som e kind of perishable raft ers would have been applied in t he Bronze Age - not only for prot ect ion of t he walls from t he weat her but also for t he benefit of t he inhabit ant s who could sit and work out side in t he shade. Our raft ers were const ruct ed from giant cane and have funct ioned perfect ly since t hen ( Figure 26) .

Figure 26: The weather side wall since the damage in January 2015 with the giant cane rafters protecting the wall from further harm.

PDF 25

Since t he workers assured m e at t he t im e of t he dam age t hat it would not m ake sense t o repair t he plast er during t he rainy season, we left it as it was and it t urned out t o be one of t he m ost visit ed spot s of t he experim ent al house, as only at t hat spot I could act ually show visit ors in det ail t he m akeup of t he walls from m udbrick t o plast er. Thus it seem s we m anaged t o m ake t he m ost of t he difficult experience, and t here has been no ot her dam age since t hen.

When t he whole const ruct ion had sufficient ly dried out , I al so decided on a m et al and glass door t o be fit t ed int o t he Bronze Age sim ile fram e. No Minoan doors have survived t he cent uries, and t he inside of t he house would have becom e rat her dark wit h a m assive wooden door. I t probably would have only been shut when people went inside t heir houses t o go t o sleep, and darkness would not have bot hered t hem . I opt ed for a clean break from ancient form s, wit h a door t hat is obviously m odern ( Figures 27, 28) .

Figure 27: The experimental house just finished with its tree ladder to get on the roof, the drain, the pebbles under the dripping rafters and the preliminary door, 12/ 2014. On various surfaces cereals can be seen sprouting from the clay- chaff.

PDF 26

Figure 28: The house in 8/ 2016, with its metal door and the “parapet” made of some remaining mudbricks.

Fut ure prospect s

The experim ent al house was adm ired am ong locals as well as foreign visit ors. This produced t he idea t hat an Open - air m useum wit h experim ent al, life- si be an int erest ing proj ect . We drew plans and even found a suit able plot of land owned by t he m unicipalit y, which is current ly an unused wilderness. We im agined how fascinat ing it would be t o have life- size, archeologically accurat e reconst ruct ions not j ust of Minoan houses from t he various Bronze Age periods, but also of houses from t he I ron Age, t he Rom an occupat ion and Medieval t im es.

Furt herm ore weconsidered workshops where visit ors and st udent s could t ake part in t he reconst ruct ions of t he houses and t heir equipm ent . Re- enact m ent s of diachronic activit ies like olive oil product ion, m edicinal herb growing and collect ing, and even perfum e m aking could also be provided.

Several Universit ies from various count ries and EXARC ( I nt ernat ional organisation of Archaeological Open - Air Museum s and Experim ent al 27

PDF 27

Archaeology ) gave us writ t en support . The locals were ready t o house st udent s in t he elem ent ary schools which are em pt y over sum m er. Unfort unat ely, we have been wait ing for an answer from t he m unicipalit y for m ore t han a year now, and even t hough we did not ask for financial support ( an im possible endeavour in t hese current t im es of financial crisis) t he t own fat hers/ m ot hers could not m ake up t heir m inds. Now, by t he end of Oct ober 2016, t hey finally decided t hat t hey do want such a m useum , but on a different plot . While t his is current ly t oo sm all t hey have decided t o buy m ore land t o reach a bet t er size ( hopefully finding an accept able price) . As part of t he plot seem s t o be occupied by a sm all Early Minoan set t lem ent , we m ight be able t o int egrat e t his and possibly it s excavat ion int o our m useum plans however, t im e will t ell.

We are sure t hat if a plot was available wit hout cost t o us, we would be able t o finance t he proj ect t hrough a com binat ion of int ernat ional and local support . This would be a nat ural t ourist at t raction, and could benefit ent erprises such as t he local hospit alit y indust ry as a whole. We have not given up yet , alt hough I wait ed for m ont hs t o writ e t his art icle, hoping I would be able t o report wit h m ore det ail on t his plan and fort hcom ing act ivit ies. As it is, t his shall have t o wait for a hopefully m ore decisive fut ure.

I would like t o cordially t hank here all m y helpers, workers, and support ers in spirit . The experim ent al reconst ruct ion of a Middle Minoan vernacular m ount ain house was a great experience!

Bibliography

Beckm ann, S. ( 2012a) .Dom est icat ing m ount ains in Middle Br onze Age Cret e: Minoan agricult ur al landscaping in t he Agios Nikolaos r egion. Univer sit y of Cret e doct oral t hesis, st able URL: ht t p: / / t hesis.ekt .gr / t hesisBookReader / id/ 29129# page/ 1/ m ode/ 2up .

Beckm ann, S. ( 2012b) . Bronze Age Landscape and Resilience: 4.000 Year s of Tradit ion? I n: J. Kneisel, W. Kir leis, M. Dal Corso, N.Taylor and V. Tiedt ke ( eds.) : Collapse or Continuit y? Envir onm ent and Developm ent of Br onze Age Hum an Landscapes. Proceedings of t he I nt er nat ional Wor kshop ” Socio- Envir onm ent al Dynam ics over t he Last 12,000 Year s: The Creat ion of Landscapes I I ( 14 t h - 18 t h Mar ch 2011) ” Kiel, Volum e 1 Univer sit ät sfor schungen zur prähist or ischen Ar chäologie 205. Bonn: Habelt .

Chaniot is, A. ( 1999) . Milk ing t he Mount ain: Econom ic Act ivit ies on t he Cret an Uplands in t he Classical and Hellenist ic Per iod, in: A.Chaniot is( ed.) , Fr om Minoan Far m er s t o Rom an Trader s. Sidelights on t he Econom y of Ancient Cret e, St ut t gar t : St einer : pp. 181- 220.

Com o, M., and Marazzi, M. ( 2006) . Analysis of t he m asonr y of t he east er n ar ea of t he pr ot opalat ial sit e of Monast iraki, Cret e. Const ruct ion Histor y Societ y New slet t er 28

PDF 28

no. 75: pp. 15- 17.

Devolder, Maud ( 2006) . Fr om t he Gr ound Up. Ear t h in Minoan Const ruct ion Techniques. The Case of Building 5 at Palaikast ro. I n: Aegean Ar chaeology, 8/ 1: pp. 65- 80.

Dim iziani, M. and Or t olani, C. ( 2014) . - Ear t hquake Ear t h const ruction. Online pr esent at ion at ht t p: / / www.doarchitet t iassociat i.com / uploads/ 1/ 3/ 7/ 2/ 13728584/ casa_ant isism ica.pdf . Last accessed 23- 10- 2016

Evans, A. ( 1921) . The Palace of Minos. The Neolit hic and Ear ly and Middle Minoan Ages,. Vol. 1. London : Macm illan.

Giffor d, J. ( 2014) . The Physical Geology of the West er n Mesara and Kom m os. I n J. W. and M. C. Shaw ( eds.) , Kom m os I : The Kom m os Region and t he Houses of t he Minoan Tow n. Par t 1: The Kom m os Region, Ecology, and Minoan I ndust ries. Pr incet on: Pr incet on Univer sit y Pr ess: pp. 30- 90.

Guest- Papam anoli, A. ( 1978) .L’em ploi de la br iquecr uedans le dom aineégéen à l’époquenéolit hique et à l’Âge du Br onze. Bullet in de cor r espondancehellénique 102/ 1: pp. 3- 24.

Lenuzza, V. ( 2013) .Of Roofs and Roof Dr ainage: a Survey of t he Evidence in Br onze Age Cret e. I n: Gr aziadio, G., Guglielm ino R. , Lenuzza V., Vitale S., (eds.) St udies in Medit err anean Archaeology for Mario Benzi, BAR I S 2460: pp. 79- 98.

Nodar ou, E., C. Fr eder ick and A. Hein ( 2008) . Anot her ( m ud) br ick in t he w all: scient ific analysis of Br onze Age ear t hen const ruct ion m at er ials fr om t hr ee sit es in East Cret e. Journal of Ar chaeological Science 35: pp. 2997- 3015

Poursat , J.- C, L. Godar t , J.- P. Olivier ( 1978) .Fouillesexécut ées à Mallia. Le Quar t ier Mu I . I nt r oduct ion générale. Écrit ur ehiér oglyphiquecr ét oise. Ét udes Crét oises23 Librair ieor ient alist e P. Geut hner, Par is.

Poursat , J.- C. and M. Schm id ( 1996) .Fouillesexécut ées à Malia. Le quar t ier Mu I I I . Ar t isans Minoens: Les Maisons- at elier s du Quar t ier Mu. Ét udes cr ét oises, 32 Par is: École Françaised’At hènes.

Soles, J.S., Davaras, C. ( 1994) . Excavat ions at Mochlos 1990 1991. Hesper ia 63, pp. 391 436.

Shaw, J. W. ( 1977) . New evidence for Aegean r oof constr uct ion fr om Br onze Age Thera’, Am er ican Jour nal of Ar chaeology 81: pp. 230- 234.

Shaw,J.W. ( 2004) . Roof Drains and Parapet s in t he Sout her n Aegean. The Annual of t he Br it ish School at At hens. Vol. 99: pp. 173- 188. Thornt on, I an ( 2007) .I sland Colonizat ion. The Or igin and Developm ent of I sland Com m unities. Cam br idge Univer sit y Pr ess, Cam br idge, UK.

Tsakanika- Theohar i, E.( 2009) . The constr uct ional analysis of t im ber load bear ing 29

PDF 29

syst em s as a t ool for int er pr et ing Aegean Bronze Age ar chit ect ur e. I n: Pr oceedings of t he Sym posium ‘Br onze Age Ar chit ect ural Tr ansit ions in t he East er n Medit er ranean: Diffusion and Diver sit y’, München 2008. Weilheim : pp. 127 142.

Vlachopoulos, A. and L. Zorzos ( 2014) . ” Physis and Techne on Thera: Reconst ruct ing Br onze Age envir onm ent and land- use based on new evidence fr om phyt olit hs and t he Akr ot ir i w all paint ings” . I n G. Touchais et al. ( eds) , PHYSI S, Aegaeum 37. Leuven : Peet er s, pp. 183- 197.

Weism ann, A. and K. Br yce ( 2006) .Building w it h Cob : A St ep- by- St ep Guide. Gr eenbooks: Cam br idge.

This paper was published online in: EX ARC Journal 2016/4 http://journal.exarc.net/issue-2016-4/ea/minoan-experimental-house-paying-tribute-middle-bronze-age- cretan-vernacular-architecture (members only).

Бескризисная экономика этногенеза

Explorer

Beckmann S. — A Minoan Experimental House (2016), повний текст