Sunday 4 September 2022


The proposition has been explored previously in The Desert Broch – see: The idea is that, on the basis of Louis Sullivan’s dictum, Form Follows Function, one can assume a certain integrity in matters – that similar forms are likely to have similar functions. While the case for this idea is no doubt able to be challenged, and the exceptions are likely to be many, making the matter perhaps uncertain, the concept could be useful as a guide to possibilities given that we know so little about the Scottish broch. It is an approach with some basis in logic: to date, much broch research has been based on maybes, and best guesses; the parallels are worth investigating.

What is certain with broch matters is that the plan was circular, and that the walls were doubled - concentric. Mousa broch gives us the most complete model of any broch, but, what with destruction, ‘creative’ reuse, and reconstruction over time, we still do not know much about: how the top of the walls might have been completed, and at what level; the roofing details; the interior; and the function/s accommodated in these forms. The desert broch, Ksar Dae Draa, took a similar form to the typical broch, and had many similar features. Although it was larger in size and scale, it was used as a prompt to suggest possible broch functions. When viewing the BBC’s Wild China documentary, a large circular building was shown as a walled village form in southern China. The structure was briefly described as one for defence, having one entry, and various inner levels for differing functions. One was immediately reminded of the broch. Could anything here help us with broch thinking?

The Hakka Tulou is the name given to the earth buildings, ‘tulou,’ built by the Hakka people in the Fujian region of China: see - and These walled villages take various plan forms – square, octagonal, circular, with the cross section of each different plan arrangement remaining the same – but it is the circular form that reminds one of the broch. The building is an array of spaces around a central courtyard that has only one entry. There are no windows in the lower portions of the building that are shaped by a perimeter stone wall about three metres thick. This is a defensive structure. The upper walls are made of rammed earth. The lower level is used for animals; the next level is for storage; the levels above with the window openings are for living spaces, with the bedrooms on the highest level. Access between the levels is by stair or ladder. The dwelling spaces are roofed; the central courtyard is open. This perimeter roof projects out beyond the rammed earth wall and shelters it. The complex is a much larger structure than the broch that accommodates many families.

The comparisons with the broch are interesting. The circular form of the tulou with the one entry and no windows in the lower walls, is for defence. One can suggest that this is also the purpose of the broch given the parallels in purpose that Sullivan suggests. This defensive function has generally been accepted by broch scholars; the problem arises with the use and the interior detailing. If one was to extrapolate, one could assume the tulou functions - animals; storage; living - as also being those of the broch: why not? Scholars do agree that the animals were very likely on the ground floor; but matters get confused and debated on the purpose of the levels above this, or whether they existed at all. There is even a discussion about the enclosure itself, with some arguing that the broch was never roofed - (Smith of Shetland Archives); or maybe roofed only in part, somewhat like the tulou with an internal, circular lean-to structure. The majority believe that the broch was roofed – a position that this writer agrees with after sheltering in Mousa broch during a storm. Who would anyone go to so much trouble to provide such awkward, uncomfortable conditions for a place that is assumed not to be a prison or a sky burial structure?

The interior lean-to roof concept.

That the broch had an inner level or levels also seems to be generally agreed. The scarcement suggests this purpose; but how were these spaces used? If we take our clue from the tulou, the storage level was directly above the animals; and living spaces were above this. Popular illustrations usually show this in the reverse, with living spaces above the animals; and storage above this. There is some sense in the tulou model, because it seems an awkward manoeuvre to transport and access stores through the living spaces. Might the stores, if this was the use, have been on the level above the animals? This means that access to the living areas must have been through the storage spaces, but this might have made more sense than taking stores through the living spaces. The organisation with the living spaces on top, would allow the possibility of light and ventilation for the living areas, an opportunity that would not be available if the living space was located between the animals and the stores. The tulou model makes sense.

In, it was argued that there was an inner structure that was likely to have been propped off the scarcement, a projecting stone ledge, with the upper floor, assuming that there were two, being framed on studs standing on the ledge beams. This concept had the stone walls enclosing the lower level in the same manner as the tulou. This lower, solid portion of the wall could be considered to be a ‘wet’ wall for the animals, with the only timber structure in this space being the set of six inner poles that, it was suggested, was the core setout reference for the circular dry stone enclosure. In the tulou, the lower walls are the stone walls that provided a more durable structure than rammed earth, one able to offer better defence and more resistance to moisture, and animal wear and tear. The upper broch walls were cavity walls that provided a dry inner skin and ventilation; this is the solid rammed earth wall of the tulou, a wall with windows for ventilation, that gets thinner as it rises. The twin-wall ventilation in the broch, which, unlike the tulou, had no windows, would work well for the stores, managing both mould and moisture. The inner ‘dry’ wall continued up to enclose the living spaces above and perhaps carry the roof load. The proposition in this model is that the roof had some form for light and ventilation – or at least the opportunity was there. Might it be the Deeargee shearing shed form? - see: This model of occupancy can be seen as a raised roundhouse or wheelhouse, a structure which has been argued to be the predecessor of the broch. This parallel might assist in providing some concept of roof form for the broch – the conical shape of the roof structure of these neolithic houses which has been frequently illustrated.#

The roundhouse.

This all seems to make some sense. While stores come in and go up one level, and up or down again for use, only waste comes down from the living areas. One has to ask about any access to the intramural stair from this upper living level, that could have been used for waste removal – throwing it over the top; carrying it down below; or might waste have been dropped down to the ‘animal’ level to be managed as byre waste is in the blackhouse -? The logic in this layering of functions would mean that all stores could be accessed without disturbing the living areas. There is good sense in this arrangement if one supposes the broch was used for dwelling. Even ignoring this, whatever function it might have had other than dwelling would still not have been disturbed with the stores being located below.

The 'living' space here is shown typically between the animals and the store area.

An unusual drawing showing the 'living' area at the highest level.

Reviewing this logic with the understanding developed previously – see:, the store area is the inner timber-framed space, with the living areas above. The upper living height would have been the warmest, with the animal heat rising through the stores, and air movement facilitated by the twin walls and the internal slots managing condensation and mould. The thinking now turns to the roof. The living area could have used light and ventilation. Fire would have been on this level, making some sense out of this location: why smoke out the whole of the animal and stores areas? Why not enjoy ventilation and light – although the black house seems to suggest neither was seen as essential, with smoke filtering through the thatch, discouraging vermin. One might note that the roundhouse/wheelhouse, with the conical roof form, also provided no ventilation.

A community.

Inconvenient stores.

Whatever it might have been, dwelling or otherwise on top seems better than in between. Previous commentators have mocked the illustrations that show the high storage areas in the sectional drawings, pointing out how inconvenient this location would be. We have to remember ancient men were no idiots and, unlike us, would not go out of their way to make things awkward on purpose, just for appearance. The broch can be seen as an enclosed, defensive structure used for important animals and stores – and maybe dwelling. It really is of no size, so there is no concept of multiple dwelling here that one sees in the tulou. So who lived here? It is generally assumed that brochs are also centres of power, making it the home of the chief; something like a castle: but might matters have been more communal? Is this idea of a leader, a chief, just too much of an archaeological cliché where ‘big’ means power? There is really very limited space in a broch, so if animals were housed at ground level, as seems to be agreed, it must have been the prized stock only. Might this area have only been for winter food for the animals? If the general stores were above this space, then access was by ladder. It seems very likely that these stores were important foods for unproductive seasons, and valuable products too, making them very likely to have been communal. One needs to note that the Hakka Tulou is a communal structure, just like the Ksar Dae Draa. Could the ‘living’ area have been communal too? Might this space have been a general hall? A chapel? A retreat in danger? - but how many might fit? Was it women and children only, with the men fighting, accessing the defensive heights from the ground level via the intramural stair? The singular entry suggests defence, but how might the building have functioned? A broch does not look as though it could house more than a dozen people in reasonable comfort; and it has been noted that it is a structure that would easily be subjected to a siege. The idea of an upper defence offers some hope of breaking a siege. The upper windows of the tulou have been purposed as locations for archers as well as ventilation, suggesting a similar function as that proposed for the broch roof perimeter space. This is an area that has also been suggested as being useful for roof maintenance.

How can one think about this? One just does not know, but the similarities in form surely offer a gauge worthy of consideration. Shelter is a basic need, like habitation. The broch offers a series of suggestive images that all need to be considered in the broad context of what we know about the times. There seems to be a great gap in our understanding of things cultural in this era. Until we know more about thus subject, it is difficult to say much about dwelling, chief, meeting hall, or chapel options as being the purpose of the broch, even if we might agree on special animals and valuable stores. Might a broch be a communal roundhouse above its community’s valuable animals and stores – a core defensive structure for the community: its heart – the very place that allows the community to be and remain together? People gather together for an advantage. This can be voluntary or enforced.

The suggested model of enforcement of labour by a powerful chief does not seem to find its pattern in the remnant walls, as Armit has nicely suggested in his text; see – and This leaves the preference for an agreeable communal gathering, people coming together to help each other, there being value and purpose in numbers and common interests. The pattern of crofting where neighbours all gather to help each other has not long past. While our era encourages individualism, we cannot and never should interpret other times using our values. So one might say that the most likely model is that the broch is communal, co-operative. One only has to see the produce that is stored for the farm animals for winter feed today to understand how storage is critical for survival, both for animals and man – critical enough to be protected with significant defensive structures.

Perhaps this concept is too idealistic. One has to remember the times of the lords and the serfs, and wonder. We are still guessing; but let us consider the process as being one of conjecture and refutation, where a hypothesis is put forward to be tested: this is the scientific method. While there is value in refutation, one has to remember the importance of tested agreement. Broch research seems to rely just too much on individual spin and personal preferences.

. . . .


If one continues envisaging broch life, one sees, not just the stand-alone circular structure that Mousa is today, but a broch with a village surround like that seen at Gurness, complete with less permanent structures for other purposes too. Following Armit’s reading of the walls, one might see the toing and froing of the villagers from shelter into the fields, and back again. As the seasons change, the summer produce is stored in the broch; the gathering for the community celebration takes part in the broch’s upper level: as winter closes in, the broch stores are accessed both for people and animals. Occasionally there is a feast, where animals are slaughtered and consumed around the communal fire, and celebratory drinks are consumed. The rhythm of the days, the seasons, and the mysteries of light, and time are marked with the celebrations. We must remember that mystery was once an integral part of life, experience, and underdstanding.

The clutter of the village allows individual family groups to shelter, but celebrations need space – the broch. The community needs protection too – the broch. Here food and valuables are stored and defended. In this context, the broch is preferred to be seen, not as personal place, but as the communal heart. It is lookout, store, community hall, chapel, defensive place – safe place, that ensures the survival of the group. When under attack, one could see value in the surrounding village clutter – a buffer zone to discourage siege or easy attack. It is a defensive ring. The broch is the place for retreat, and attack. The village might be ransacked, burnt, but the broch holds the key to existence – the future. The perimeter upper access becomes critical for the broch – for defence; for maintenance; for waste - or is this silly: why might waste be deposited on the village surround? If the roofs below can be burnt, that of the broch could easily be destroyed too if unattended. One has to see a significance in the intramural stair; the effort to construct this stair suggests that it has to hold a vital role; just as the twin walls do too. We might construct follies today just for display, (Gehry, Hadid), but the rigour of the past and its travails surely must ensure that form was purposeful, essential, allowing one to assume that the greatest effort went into the most important matters: the broch. One has to ask: why is the entrance to the intramural stair opposite the entry?

Little wonder that brochs were constructed so boldly – multipurpose place: lookout, shelter, store, communal place – and sacred place? One might hypothesise: the broch was the core of life; it provided life-giving place. Who knows the details of iron age life? It is a shame that iron rusts. One might enjoy using such a structure today. One could say that such a pattern of living is not too different to life in Shetland today, where scattered dwellings on crofts provide for families that come together to celebrate in the local hall for important occasions in life and time. Why might we assume other times would be different? People live, love, laugh, and like with a native joy and necessity that allows one to see farmers and fishermen; fathers, mothers, and children; and communities of the past in the same way as they are seen today; real thinking, feeling, social folk working, living and trading, sheltering, enjoying and mourning, celebrating to mark the mysteries of the passing of life and time for sustenance, both material and spiritual. The broch provided this, as did the Hakka Tulou and the Ksar Dae Draa – for ‘real’ contented people who had the very best that the world could offer. Only our modern world view anchored in the arrogant concept of ‘progress’ likes to see them as ‘backwards.’


To better understand broch matters, one needs to know more of the physics of the form – how air moves through the place in different circumstances; how mould was managed; how fuel was handled; how condensation was controlled. There is already the suggestion that the twin walls have been detailed as cavity brick is today, with one inside dry wall, and one outside wet wall. One also needs to consider waste management, not only from level to level, but from inside to outside. As noted, it doesn’t make any sense to dispose of waste over the surrounding villagers. How is waste managed in the Hakka Tulou; how did the Ksar Dae Draa handle trash? The blackhouse byre had its own drain, with the end wall being demolished annually so as to clean the place out. What happened with the broch which has just one entry/exit? Were animals really housed in the lower level? One has warmth and moisture in these areas that needs to be handled, both as vapour and liquid; there is also the solid animal waste. What happened to this?

It has been suggested that brochs had their own wells for a secure water supply. How was this arranged with all of the other internal activities? There are many contradictions and issues yet to be considered and responded to, answers that will all be useful in determining just how the brochs were used.

A thought on ventilation: if much is to be made of the internal ventilation system of the broch using the wall cavity, one might speculate that the top of the broch had openings to facilitate air extraction. Given the effort to form openings internally, one could find it unusual that the extraction would have been allowed to be through the ad hoc slots between the rocks in the wall or the upper perimeter floor slabs. One could envisage a series of openings just below these floor slabs, with a balustrade parapet above. It has to be realised that air movement would have been critical in the broch, not just for a change in air, but to control condensation. Given the climate, the enclosed space with no windows, and the possibility of occupation by animals and humans, the interior would become warm and moist, with the stone walls very likely to be below dew point, becoming a dripping, wet wall on every level, promoting the growth of mould. Mould is not good for human health or for the items stored. Air movement could manage this situation and make conditions more tolerable. In the living areas, if this is what occurred, one could speculate that wall hangings were used to add insulation, in much the same manner as in castles with tapestries. Here one could suggest skins and fabrics; but who knows what crafts were practised or how dwellings might have been decorated?


The concept might also suggest a staged development. One has to ask if brochs were built in stages, with the first stage being the roundhouse that was raised as extra storage was required. This would allow the village core to develop vertically and minimise disruption in expansion. Maybe the roof was removed, stored and replaced on the higher framing or walls? The building of the broch must have been an important occasion. One could imagine how the energy and costs involved might have been spread out in stages. The suggestions already made in relation to broch building – see: - would allow for a staged process, with the core reference for all dry stone elements being the central timber poles.


There is something about scribbling that records thoughts that assist in sorting matters out. While one might ponder matters for hours, it is scribbling that documents ideas in another way, and allows them to prompt other concepts differently. One might call it ‘doodling’ in order to highlight the distance, dreaming, and discovery involved in the drawing process that is loose and revealing, searching in its own casual manner, becoming a stimulant for parallel visual interpretations that can inform other thinking processes.

The scribbled napkin.

Broch matters have been pondered for years – see:

 - but the questions still linger. Each step seems to raise more questions, that again lead to more scribbling. This process started the other day on the proverbial paper napkin, just because it was there; then the scribbles continued on the reverse side of the napkin until some A4 paper was pulled out to continue expanding on the thinking.#

The napkin sketches related to thoughts about the top of the broch: how was the broch finished off? Ideas on broch building have been recorded previously, with the broad acceptance that the Mousa parapet walkway might have been the way brochs were finished off. While Brian Smith of Shetland Archives has argued that the broch was never roofed, one has put the case for this being wrong, and has suggested how brochs might have been constructed as dry stone enclosures with timber framing inside. Roof or no roof, the matter of how the brochs were completed remains a question.

The more one thinks about the roof and the conditions in the broch, the more essential it is to be explicit about exactly how the broch might have been completed at the top. One knows that Mousa has undergone both destruction and reconstruction, so the current situation cannot be accepted as being the final solution.

When questions are raised about how moisture was managed in the brochs, both as rainwater and condensation, thoughts turn to air movement and detailing. Without air movement it is very likely that brochs would have become very moist and mouldy. Rain water could have easily been excluded, but it is condensation that becomes the problem with enclosed spaces housing animals and people. Even with a fire, the rock walls would be cool, condensing surfaces for airborne moisture that would become the perfect place for the unhealthy growth of mould. If the broch was used for important stores, then mould is the very last thing that one would want.

Drainage and ventilation?

The broch has been previously analysed as a cavity-walled structure similar to the cavity brick walls we have today, with ties sloping outwards to drain the water away from the dry inner wall. These inner walls have carefully formed openings in them that one could speculate might be for the promotion of air movement. The original brick cavity walls used to be vented internally to keep the inner wall dry. Might the cavity be part of a ventilation system that manages moisture and mould?

For this to be so, one might expect that there would have been a structured response to the idea externally at the top of the broch rather than allow on the usual leakage between the rocks for this important function. Such precision in the internal fabrication would seem to need some resolved completion of the concept externally. Might the top of the broch have had vents?

Then one has to ask about the Mousa model: why would the broch builders want to capture rainwater and snow and allow it to pour down into the cavity if the cavity was to keep the inner wall dry? Could there have been no external parapet? The blackhouse offers a prototype for both these options – openings and access without a parapet. Its external ledges on top of the earth-filled, twin exterior walls, are accessible for maintenance, and do allow water and snow to drain to the outside.

The blackhouse model.

The blackhouse model has an opening in the external wall that shows how vents could be formed. The sketches explore the possibilities that then raise the matter of the roof. With the roundhouse being the original model for the broch form, might one speculate that the broch was a staged construction where the roundhouse was lifted, and lifted again in order to accommodate more animals and stores? If this is so, then one could assume the broch roof was the roundhouse roof, even if it was not staged.

Might rocks have anchored the broch roof?

How was this roof supported? Such a roof would be difficult to support on dry stone walls. The windy conditions would suggest that there was a need for a roof that was carefully tied down. Resting structure on dry stone walling does nothing for holding anything down. One could speculate that the timber framing inside continued up as a core set of posts, with a perimeter structure doing likewise. Such an arrangement would allow some detailing for ventilation above the dry stone wall.

Roundhouse stages?

So the sketches continued. Why not have a blackhouse detail at the top of the broch? One might wonder about the security of anyone on this ledge if one considers the broch a defensive structure. So might there have been some castellated form of parapet like the Great Wall of China? This would offer some shelter from exposure to any attackers, and still allow water and snow to drain away to the outside, rather than load the cavity space.

Stair access?

One has to remember that the broch has the intramural stair. One would not want this space to become too wet, mouldy, and mossy, as it would make access dangerous. So the external drainage does make sense; but there is the exit. Most broch models ignore the fact that there is a stair, and make no allowance for any exit at the top of the broch. Given that the builders went to the trouble of providing this access, one has to assume that the exit at the top of the broch was resolved properly; that it was not an awkward hatch or an open void for water to flow into, and for snow to fill up.

What could have happened here? One could propose a sheltered exit area, but what might its form be? Here the sketches guess at possibilities, but they all look too crude. One has to assume some degree of sophistication in outcomes here; the broch has not been built by fools using an ad hoc process. It is not until one returns to the roundhouse concept that a solution arises. Using the raised roundhouse idea, one can postulate a roundhouse roof. Looking at these roofs, one sees a solution to the broch problem of exit: the roundhouse entry itself. The roof of the roundhouse is frequently shown as a thatched swelling that rises from snug, low edges to gain height at the doorway.

Sheltered access?

This model would suit the broch very well. The roof could generally be kept low to the dry stone inner wall until it came to the intramural stair exit, when it could swell up and extend over as a shelter. This would provide a raised portion of wall in the living area – the equivalent of the entry to the roundhouse. This section could become a main vent for the interior; and, with a ladder, provide direct access to the external walkway. Why should the top of a broch not be as inventive and organic as the remainder of the structure?

The roundhouse roof form?

The scribbles explore the options. Could any external vents in the dry stone walling be a part of the castellation forms; or might the openings have been framed by the profiling? The structural detailing seems better resolved by the latter. Still, one is left wondering: does the castellation concept look too much of a ‘castle’ cliché? Is the ‘blackhouse’ detailing preferred as a visual outcome? Why should we even have any preferences here? What causes us to bring these preconceived visions as judgements? Why are we questioning our logic?

There are more matters to test: could the parapets have had drainage outlets as well as ventilation slots? How might the stair exit be detailed with a ‘blackhouse’ edge? Did the stair always go to the roof ledge? Might it have had direct access into the living area? How much ‘restoration’ work did the Victorians undertake? One only has to refer to Maeshowe in the Orkneys to understand that the Victorians were never shy about reconstruction.

Is the castellation form a dry stone detail?

The scribbles will continue.

Alternative drainage, vent, and shelter?



These thoughts and scribbles were developed while writing The Chinese Broch- Hakka Tulou: see -