Thursday 25 April 2019


Mesmerised by the marvels of Algeria from above, a programme presented on SBS TV Australia, Saturday 6th April 2019, one is repeatedly astonished by the wonder and beauty of place. Part 2 of this presentation was eagerly anticipated after enjoying the visual delights and intrigues of the first segment. This time, the camera was to travel east and then south, into the limits of the Algerian Sahara, a desert that astonishingly, is, in total, larger than the whole of Australia.

The vistas of the desert opened up. Images of the desert towns appeared as beautiful patterns of tight, protective habitation. As the camera moved south, it passed over a circular structure in the middle of what appeared to be nowhere. Sand stretched from horizon to horizon around these concentric walls as the voice-over described “a casah – perhaps the last Jewish caravanserai in existence.” What was it – a ‘casah’? This was the phonetic, best-guess spelling that was quickly jotted down as something to be followed up. The structure looked just like the brochs scattered around Scotland and its western and northern islands. The similarity in aerial appearance to the most complete broch in existence, that on the island of Mousa off the east coast of Shetalnd mainland, was remarkable. One had to find out more about this desert structure.

Ksar de Draa

Mousa broch

The search started with Googling ‘casah,’ but the corrected suggestion was ‘casbah.’ The video was replayed on ‘Catch Up,’ and the original sounding was confirmed – ‘casah.’ The end was an ‘r’ sound, but the guessed ‘h’ seemed a reasonable proposition given the Arabic source. The word was certainly not ‘casbah.’ Maybe it started with a ‘k’? This was tried and ‘kasah’ turned out to be an old Hebrew word sounding like ‘kaw-saw’ meaning to become sated, be gorged with food. It had been used once in the King James version of the Bible. One seemed to be getting close, but realised the word was not kasah.+

Ksar de Draa, Timimoun   George Steinmetz

The search then moved on to other strategies. Why not search for ‘circular structures in Algerian desert’ and see what images might turn up.? The ‘casah’ would be easy to identify. Nothing like the shape seen previously in the programme appeared in the images. More and more variations in descriptive text were tried; eventually the form appeared in an aerial photograph taken for The New Yorker by George Steinmetz - see; The site was opened – there it was with a caption, the mystery building; the spelling was ‘ksar.’ No more detail other than the name ‘Timimoun’ was offered. Armed with this information, it was back to Google, and yes, more images appeared along with the full name and other references: Ksar de Draa. We now had access to information and images to discover more about this remote 'desert broch.'

The idea arose: if one can discover the purpose of this structure and its functions, one might be able to discover more about the Scottish brochs, to give a different beginning for new theories. There was no necessity here suggesting any parallel other than matching images, but the exercise could perhaps be useful in freeing up understandings. Research on the brochs ironically seems to go around in circles, with nothing ever being resolved. It is as if any final determination would put folk out of research work: might this threat drive perpetual disagreement? There could be nothing to lose by looking at the parallels; so the issues were explored. The immediate similarities that could be observed were: the circular form; the twin walls; what looked like the singular, controlled access; and the prominent location. The Ksar de Draa was positioned high in the desert; Mousa broch was built on a low, rocky promontory, but was taller. The text found in was of interest:

Google Earth image

Mousa broch, Island of Mousa, Shetland Islands

Google Earth image

ksar draa c1
Circular, made of a double wall, the outside of stone, the inner wall in the ground. One door to the north rooms nestled in the double wall on 3 levels, no stairs for, ladders were probably used to access the upper floors. The rooms were not adjoining. No window to the outside. A sentry with a 2 meter wall surrounds the outer wall. the interior has engraved on a wall contours leaves us guessed a star has 6 branch simbole Judaic .de many archaeological index attests to the presence of Jews in the region .ci dessou interview Jacob Oliel "The TOUAT (TOUAT- Gourara-Tidikelt, is now a province
Algerian Sahara forming a quadrilateral of 400 km of 120 km. It is a
thousand kilometers of the Mediterranean, south of Tlemcen. Its population,
Originally consisting of Haratine of African, Berbers and Jews, through
its dynamism, generated the prosperity of this region become the hub
trans-Saharan caravan trade of the twelfth and fourteenth centuries.
Little known, the history of this region was revealed to us especially after the arrival
the French in 1900 and the collection of stories of local Arab chroniclers - no
convenience suspects - by A.G.P. Martini, which allowed to share
some elements of the past resurrected:
- "The natives say that the ksour Tamentit were created by Jews the year
the elephant. Thus the Arabs designate the year during which shipment took place
qu'Abraha, Ethiopian prince undertook against Mecca to overthrow the temple of the Kaaba;
Abraha rode a white elephant. "* Iiiii
- "[The] Arab populations found in this country some of those who had put
culture from the beginning: it was the Beni Israel "
- According to the traveler Helal ben Messaoud, who came from Mosul (Iraq) in the Tuat
Jewish exiles company, (which) was arrested in the year 131 [748-749 AD. BC] to Takhfif,
that the Jews had already evacuated; who settled there he brought with him and Jewish merchants lived there. They found mention on the graves of Jews who had abandoned the country,
that they had arrived in the 4429 years of the release of Adam [...]. It was around the year 5 after J.-
C. that the Jews in question were arrived at Takhfif "iv
- My host, Sidi Youssef Mrabet, had learned from his grandfather that the Jews were the
first inhabitants of Tuat and they existed there as a nation in 260 [905]. I could see MySelf
their synagogues and their arcaded shops ".v
Obviously, these chronicles do not provide information about prior periods

journaliste :

David Bensoussan – Les Éditions Du Lys

The information was cryptic, but it was the only information that could be readily found. A detailed plan was searched, but nothing appeared. By way of summary, the text tells us that:
the twin walls are of stone;
the are no stairs;
the structure has upper levels;
maybe ladders were used for access;
there are rooms in the structure;
the rooms are not connected;
all openings in these rooms are on the inside;
there is a sentry area surrounding the main wall;
this sentry zone is inside a two metre high wall;
there are Jewish markings in the stones.

Google Earth image

Looking at Google Earth, one can see, not only the remoteness of the Ksar de Draa, but also its detail. The inner area of the structure seems to show markings that look like subdividing walls, suggesting habitation. The aerial photographs of the Ksar de Draa show bulbous shapes on the sentry area that remind one of the medieval castle tower plan forms shaped and positioned for defence. The images also highlight the castellated top of the outer circular wall, but these might be from selective erosion. These details seem to point to a defensive purpose. The Ksar de Draa appears to have only one entry point. Was the Ksar only a defensive structure? What was it trying to keep safe?

Ksar de Draa entry

Ksar de Draa interior

Wikipeadia defines ksar as a Berber fort:
Ksar, plural ksour (Maghrebi Arabic:قصر qṣer, plural qṣur; Berber: aghrem or ighrem, plural: igherman) is the North African term for "Berber castle", possibly loaned from Latin castrum. The term generally refers to a Berber fortified village.++

Mousa broch

Turning to the broch form, using Mousa broch as a guide, one can summarise its characteristics similarly:
the twin walls are of stone;
the are stairs (intramural);
the structure has upper levels;
maybe ladders were used for access;
there are spaces in the structure;*
apart from those on the ground level, the ‘rooms’ are not well defined;
all openings in these ‘rooms’ are on the inside;
there is a sentry area at the entrance to the broch;
the broch has only one entry point.

Google Earth image

The similarities are indeed remarkable, so much so that one would be happy to confirm the statement that brochs were primarily defensive structures: forts. The Ksar de Draa had a sentry protective zone as well as the castellated edge, (maybe), that seems to suggest an upper protective perimeter zone as well as the lower fortification. The broch has no such lower zone, but the entry was secured by sentry posts and distinctive reinforcing detailing. Were the broch intramural stairs primarily to provide direct access to the upper perimeter promenade for defence, to allow this role to be fulfilled efficiently and effectively? The stairs do not provide drop-off points to any intermediate, intramural levels, (Mousa), that presumably were reached by ladders or accessed by timber stairs from the interior space, if at all.* #

Lego Mousa

While there are similarities, there is at least one difference: the Ksar de Draa stands alone, as the remarkable desert images reveal: but it was located high, seemingly to give it good supervision of its region. Was this prime position for surveillance? Brochs appear to be strategically located too, and vary in height with the prominence of their geographical location, as if strategic scrutiny of its surroundings was critical – c.f. Mousa and Dun Carloway: but brochs seem to be more social, with the idea being that they were located so that they could see each other; communicate with each other, (Smith^).

Dun Carloway, Isle of Lewis

Mousa broch, Island of Mousa, Shetland Islands

Google Earth image

Might one speculate that brochs, as defensive structures, were also lookout posts, perhaps like the Ksar de Draa, and that the intramural stair access facilitated this function? On the apparent interrelationship of brochs: the idea that brochs communicated with each other as positions in a linked chain seems fanciful: see - One has to ask how messages were transmitted, and what variety and complexity of information might be forwarded beyond the digital yes and no – on and off; fire and no fire? That fire could be used appears a poor argument given weather and distance that would make even a significant blaze look like a candle in the wind. Yet the high, perimeter space would always be there for the lookout and for defence, independent of weather. The idea that the brochs might have been positioned so that the areas being surveilled could overlap, would only ensure a more comprehensive supervision of the whole region if one assumes that brochs were working in concert. This possibility would then at least give some sense to the singularity of any communication, e.g., smoke for danger, if one has to agree to the message idea: but might not smoke be rising form these villages just for cooking and heating? Maybe runners might make more reliable communicators.

Qasr al-Hajj, Libya

On the hypothesis that these structures were defensive, one has to ask: what was being protected? The Ksar de Draa has a similarity to a structure shown by Dan Cruickshank in his TV programme, Around the World in 80 Treasures. ‘Treasure’ number 60 was the Berber granary, Qasr al-Hajj, in Libya. This complex has an unusual top to its circular walls – radial, semi-circular vaults. Did the Ksar de Draa have the same roof form of which the castellations are the remnants? The idea of upper security/supervision may require revision. The Berber granary has a controlled, singular entry, and a small mosque in its centre space. It is a building used for the storage and protection of food. Did the Ksar de Draa have the same purpose, and a synagogue?

Qasr al-Hajj, Libya

Rock paintings, Sefar

The ancient rock paintings in the Sefar wadi of the World Heritage region of Tassili n’ Ajjer, (see: ), suggest that the region was once all savannah with giraffes and other wild animals roaming over open grasslands. While the area might not have been as fertile in the time of the construction of the Ksar de Draa, one could easily assume that the nearby lands were productive, and that the core stone structure was surrounded by a mud brick town that used the Ksar de Draa as a protected store for its produce; c.f. Qasr al-Hajj, in Libya. The Draa region is remote and historically, in such places, marauders cause problems: hence the need for security? Recently the Dakar Rally had to move to South America because of the problem with desert bandits and looters. One could envisage the Ksar de Draa as being the solid stone safe area of the surrounding, less permanent, mud brick village. The traditional method of construction used in the Draa Valley region is mud brick with palm trees being used for timber – see: This study shows lower walls being constructed partially of stone, so there might be some remnants of a settlement surrounding the Ksar de Draa surviving today, but this is unknown.

Ksar de Draa showing the lower sentry surround

Gurness broch

Brochs did have settlements surrounding them, as evidenced by the Broch at Gurness. There is no immediate evidence of a settlement around Mousa, but who knows what stones have been removed for re-use? There is a nearby, more recent laird's house with its surrounding dykes, dry stone walls. The first stones to be selected for this construction work would have been those that were more readily available – the village stones. Maybe brochs were used for purposes similar to the Ksar de Draa functions, for the storage and protection of produce, for the inhabitants to access in the unproductive season - winter?

Mousa broch

There are other serious matters of debate with brochs: Were they roofed? What is the rock ledge, the scarcement, for? (see - Was the Ksar de Draa roofed? One could understand the need for the Ksar de Draa to offer shade. The climate might not have been as harsh as it is today when first erected, but still the need for shelter in these environs is obvious. The region relies on shelter for nearly everything; souks remain an easy example to reference. Roofs may only have provided partial cover inside the circular walls of the Ksar de Draa, but one can assume that with breezes excluded by the high, perimeter walls, the shading of shelter would be useful. Might the central space have been an 'oasis' courtyard area?

Typical souk, Algeria

Brochs do not look as large as the Ksar de Draa, but were they roofed? It is difficult to envisage the broch interior as an open courtyard. The day we visited Mousa, there were several passing showers. One heavy downpour came across when we were inside, on the ground; we had to seek shelter in one of the recesses. It was not very comfortable. Given the effort required to construct these walls, one might assume that brochs were not erected to offer such uncomfortable conditions as we experienced. One could believe that the chance of the walls being roofed was greater than not. Even the option of a circular lean-to inner roof seemed unlikely to offer any adequate, reliable comfort.

Round house

As for the idea that the broch was a long-term refuge from attackers, one has to realise that such a function might almost be self-defeating, with the broch being easily subjected to siege. The Ksar de Draa has a similar problem, but it seems that its stores might have been more extensive; and its defensive perimeter gave a more aggressive, protective prospect than the broch had, even with the broch's suggested upper surveillance zone.

On the fitout of the interior, the Ksar de Draa was of a scale that it could accommodate villagers, visitors, and animals in a ‘caravanseri’ circumstance. It may even have had a synagogue. The surrounding rooms could have been multi-purpose, with lower ground spaces being for uses other than storage; or perhaps for the storage of fodder and water; with the upper levels being accessed by ladder for the storage of grain and oil. It is water, its reliable supply or otherwise, that makes the difference to understanding how these places could have been used – both the Ksar de Draa and the broch. Without a reliable water source, it is difficult to envisage the place as a refuge, other than a temporary one.

Typical plans of brochs

Typical round house plans

The interior of the broch gives rise to much discussion.## The debates have centred around the corbelled course inside the broch, the scarcement. This is a projecting strip of stone that runs all around the inner circumference, about three metres above the ground. It has been argued that it was for scaffolding, (Smith). It seems that, because of its continuity, it was used for the support of closely spaced beams or posts – see: . . . A larger spacing of structure would only require isolated projections. This idea seems to point to a radial pattern of joists that run inwards to an inner frame. Post holes have been discovered in some interiors. The idea that these elements might have been a common feature is further reinforced by the suggested origin of the broch model, the round house, that had an inner structural purlin ring on posts supporting the roof. There might have been a floor with an open centre, inside these posts – possibly for fire and/or a stair? If the idea of projecting stones being provided for a scaffold was indeed the case, one might have expected to see far more projecting stones than just one, low continuous strip, both inside and out – like the Roman Pont du Gard at Nimes.

Round house

Pont du Gard, Nimes - the projecting stones are supports for the scaffolding

The question remains: how were the brochs built? How was the Ksar erected? Without seeing the plan of the Ksar de Draa, one could envisage a progressive spiral strategy for the construction of the set of rooms, set out with a peg, string and plumb bob. There is not much difficulty in the concept of stacking boxes. Brochs are different: see - ... The experience inside of these walls needs more consideration, as does the space between.# Roofed, the brochs would have been dark and unventilated. The Ksar de Draa walls would have cut off any breezes, but the shading might have been more piecemeal and liberal in letting in the bright light into the covered central spaces. Maybe devices were developed to catch the breezes, as in the Emirates – wind towers? With the broch, might removeable skins or roof panels be used to catch breezes, to let out smoke, and let the light in? The quick, direct access to the upper level would have made this possibility easy: “David, could you duck up and close/open the roof please?” We must never think that men of old were idiots who carelessly put up with atrocious, uncomfortable conditions. These solutions were real possibilities. The evidence has all gone: it is only our faith in human nature that can make these points real. The idea of reconstructing life can be useful; see –  This all suggests the broch roof was supported off the posts/purlins and the inner wall, with a rampart perimeter for a lookout defence, and for the servicing of the structure – maybe for maintenance too. Scottish, especially Shetland winds can be damaging.

Was the broch a refuge as the Ksar de Draa appears to be? Did the population and the animals retreat into the broch at the first sign of danger? One might consider it foolish for a group to lock itself up in a place that could be so easily placed under seige. Supplies of food and water have their limits. As for animals, even if one ignores the food, water, and waste removal problems, there is not much space in the broch for any great herd or flock of anything. If one or two animals were of critical importance and value, maybe these were housed in the broch, but not much more might be possible. As for the population, even of a small village, one has to wonder how and where these folk might have been accommodated. How large was a village? - (c.f. Gurness in Orkney). It is simpler to understand the Ksar de Draa as a refuge - (c.f. Berber village).

Gurness broch

Berber village

Were there levels inside the broch? How many? If the scarcement was the base of this inner timber structure, one might assume two levels – one directly on the scarcement; the other propped off it. It would have been structurally difficult to have another scarcement for an upper floor built into the thinner, inner twin wall. The more one ponders the cross section of Mousa, the more it appears that the broch is a base that supported a separate structural cradle, likely to be a timber frame, that had a twin wall enclosure that contained the separate, critical access from ground to roof for security, defence, daily well-being, and maybe maintenance. The Ksar de Draa seems more predictable: daily life carried on at ground level, with the community stores being held in the upper rooms. Indeed, one could even accept other activities in the higher spaces that are a little more remote from things daily and frequent.

Ksar de Draa sentry area

One is really no closer to any definitive resolution on the purpose of the broch. It seems that function of the Ksar de Draa is easier to be more certain about – defence; storage; and possibly a safe retreat, even as a lodge or an inn complete with a place of worship. It has been labelled a Jewish caravanseri, the last one in existence. It does not appear to be an oasis; but might it once have been? Did it have its own wells? Can one start thinking of brochs similarly? It might not be an unusual stance, as the possibilities appear likewise. Architectural forms do have some resonance with purpose and possibilities beyond place.

Mousa broch entry

Mousa broch interior

Can one say that the broch was: a defensive structure; a surveillance location; a store; a refuge; maybe a chapel? Could it have been a permanent residence for the overseer – maybe a priest; a caretaker? One senses a greater communal role for these magnificent structures that must have taken much shared effort and monies/barter to be erected. One could speculate that the building team either involved slavery or enforced labour; or, as with cathedrals, was a committed, cooperative involvement of time and energy for a shared outcome under the management of a master mason and the specialist team. There are different social structures involved here. There appears to be a pattern for brochs that does suggest a master mason type circumstance where specialists moved from location to location as the demand called. One would expect more variation with a different, more fragmented system.

In summary, can one basically describe the broch as a working base, with an accessible, defensive top, with habitation and storage enclosed in between, inside? One has to know more about the society to gauge the possibilities of it being a community church, or a chapel, a sacred place too. It is highly likely that the Ksar de Draa contained a synagogue. Could the broch also have a funereal role – even in part? How did the Ksar de Draa people manage their dead?

One can peruse both the Ksar de Draa and Mousa broch, and be inspired. Both structures are magnificent in their own identities. Architecturally, it would be odd, perhaps unusual, if they had starkly different purposes. As the influential architect, Louis Sullivan, 1856-1924, 'the spiritual father of modern American architecture,' said: “Form always follows function; just as function follows form.”

The KJV Old Testament Hebrew Lexicon
Strong’s Number: 03 Browse Lexicon
Original Word Word Origin
hXk a primitive root
Transliterated Word TDNT Entry
Kasah TWOT – 1049
Phonetic Spelling Parts of Speech
kaw-saw’ Verb
1. (Qal) to become sated, be gorged with food
King James Word Usage – Total; 1

Ksar or Ksour is the North African Meghrebi Arabic term for "castle", possibly loaned from Latin castrum. The Berber original word for "ksar" used in North Africa by the Berber-speaking populations is aghrem or igherman. In the Maghreb, the term has a more general meaning of "fortified village,"or "fort". The Berber word igherman might be a cognate word, with an identical meaning, with the word Garamantes, which is the name of the ancient Berber city-states in modern-day Libya. Ksour in the Maghreb typically consist of attached houses, often having collective granaries and other structures like a mosque, bath, oven, and shops. Ksour / igherman are widespread among the oasis populations of North Africa. Ksars are sometimes situated in mountain locations to make defense easier; they often are entirely within a single, continuous wall. The building material of the entire structure is normally adobe, or cut stone and adobe. The idea of the ksar as a granary is a confused notion of two things, the granary itself, found within a ksar, and the ksar, which is a village, normally with granaries within it. Ksars form one of the main manifestations of Berber architecture.
also see:

Brian Smith, Archivist, Shetlands Museum and Archives

For context and scale of the Ksar de Draa, see YouTube video:

Sectional drawings of brochs are deceptive. Stones illustrated between the twin walls appear to indicate continuous floor levels, suggesting usable spaces between the walls; but this may not be so. A glimpse at the spaces between the walls at Dun Carloway shows stones located as a series of ties, with no continuity. This gives the impression that the spaces between the stone walls were not used for anything other than a practical and structural convenience – to reduce the mass of the high stone; a matter of both quantity – to minimise the number of stones that had to be lifted up to the masons; and structural efficacy – to reduce the mass that had to be held up and stabilised. In order to get the correct impression of what is really happening beyond the hope of habitation or storage, the plan view of the broch at various levels needs to be carefully and accurately documented. These records would have to precisely identify the various relative levels of the bridging stones. At Dun Carloway, the stones seemed to be arranged in an offset pattern as present-day brick ties are; see - ...

On the openings in brochs, (Mousa), it seems that studies using air flow detectors need to be carried out. Such research on an accurate, scaled model of the broch,* would show how air would circulate through the structure in different circumstances: with one floor level; with two, or three floor levels, or none; with and without a roof. This study could indicate the purpose or otherwise of the distinctive openings in the inside walls. It has been suggested previously, (see:, that these slots could be vents for the cavity, to control temperature, aid drying, and encourage air movement to overcome the growth of mould that flourishes in environments that are cool, unventilated, and humid. The study of the air movements could help us understand more about the physics of the structure and its functions.

No comments:

Post a Comment

Note: only a member of this blog may post a comment.