Howe, Stromness – a hub of Neolithic and Iron Age activity

By Sigurd Towrie

Until the early 1980s, a field near Howe Farm, Stromness, was dominated by a massive mound – the “howe” [1] from which the farm took its name.

Just over half-a-mile to the south-west of the Knowe of Unstan, the Howe mound was prominent on the hillside running down towards the Bay of Ireland and highly visible for miles around, except from the north-west, where the slope hid it from view.

Like many such mounds in Orkney, Howe had long been declared to be an Iron Age broch site.

The landowner’s mid-19^th century explorations had disturbed the upper levels and “produced a considerable number of relics” [2], including a Viking Age glass linen-smoother.

The interior of the mound, however, was untouched – perhaps due, in no small part, to the mound’s notorious reputation as a haunt of the fairy folk. [3]

Other than the drunk fiddlers of folklore, Orcadians tended to avoid such places.

Ahead of planned agricultural improvements, the site was completely excavated from 1978 until 1982 – an operation that revealed a complex series of occupation episodes that spanned the Iron Age and were preceded by two phases of Neolithic activity. [4]

The Iron Age levels dated from around 800BC to AD400, with subsequent use as late as perhaps AD900. Evidence of Bronze Age use, however, was notable by its absence. [4]

Unfortunately, no radiocarbon dates were secured for the Neolithic phases, so we do not know where the site fits into the timeline of Neolithic Orkney, how long the structures were in use or the period between the phases.

In addition, the Neolithic layers produced very few finds. The 26 artefacts recovered made up a mere 0.15 per cent of the total from the site, none of which, including the few sherds of pottery, allowed any date estimations [4].

The first Neolithic construction at Howe was a rectangular structure and perhaps a large standing stone to its north-east.

Described by the excavators as a “mortuary house” [5], it was remodelled at the same time a large, 15-metre-long, structure was built directly opposite its northern entrance. The addition was interpreted as a “stalled tomb”. [4]

After an unknown period, both buildings were dismantled and a Maeshowe-type passage grave erected in their place.

The ‘mortuary house’

Why this building was considered a “mortuary house” is unclear from the excavation report. It makes no mention of any human remains in this, or the second, phase of its use. [4]

It was a sub-rectangular building measuring c4.5 x 5.4 metres (16.4 x 17.7ft) with 1.8-metre-thick (5.9ft) walls. Where evidence of these walls survived, it was generally just the foundations. With no sign of an entrance in the surviving sections it was suggested – based on the orientation of subsequent doorways – access was via a doorway in the missing south wall. [4]

The building’s interior had been seriously disturbed by centuries of later activity, so the original floorplan was not clear. It may have had a clay floor and, in at least one phase of use, a hearth. [4]

The ‘forecourt structure’

The construction of a second building directly to the north – interpreted as a “stalled tomb” [4] – saw the “mortuary house” remodelled and clearly linked to its new neighbour. The excavators proposed that the erection of the “tomb” and the alterations to the “mortuary house” implied the latter’s role had changed to that of a “forecourt structure”. [4]

The remodelling saw a wedge-shaped, northern wall added to the “forecourt structure” to match the alignment of the “stalled tomb’s” south-eastern end. A second entrance was incorporated in this new wall, positioned directly opposite the doorway to the new building.

The inner face of the modified northern wall was lined by orthostats set vertically in a slot cut into the floor – a “decorative feature” [4] closely paralleling the “stone cladding” encountered, some three decades later, in Structure Twenty-Seven at the Ness of Brodgar. [6]

The “forecourt structure” alterations saw its interior reduced in size, the chamber now measuring c4.5m by 4m (16.4 x 13.1ft). Its hearth was slightly off-centre and aligned on the new northern entrance passage and, beyond, the door to the “stalled cairn”. There was, however, no evidence of the hearth being used, although it was not excavated beyond partial cleaning. [4]

The ‘stalled tomb’

Estimated to have been around 15 metres (49ft) long, the “stalled tomb” was aligned SSE-NNW with a narrow, 75cm-wide (29.5in) entrance passage leading into a four-metre-wide (13.1ft) chamber.

Only the southernmost section, which measured 7.85m (25.8ft) wide, was exposed. Like Structure Twenty-Seven at the Ness of Brodgar, the two-metre-thick (6.6ft) east and west walls were slightly bowed [6].

The idea the building was a “stalled tomb” was based on evidence of an orthostat projecting from the eastern internal wall, 2.8m (9.2ft) from the entrance.

Although the stone slab had broken off at ground level, the stump and socket remained. Directly opposite, adjacent to the west wall, was a clay-filled cut that may have held a second orthostat. This, however, was not excavated. [4]

A pair of orthostats jutting from the interior walls was regarded as evidence of a “stall” typical of those found in Orkney-Cromarty chambered cairns.

These structures, generally regarded as being typical of the Early Neolithic, had their internal space divided into compartments, or stalls, by pairs of standing stone-like orthostats. [7]

But was it a stalled cairn?

We now know that it was not only chambered cairns that were divided up by orthostats – other buildings were, including the Knap of Howar, in Papa Westray, and Structures Five and Twenty-Seven at the Ness of Brodgar. [6]

To Davidson and Henshall, the Howe “stalled tomb” was a house.

Not only were “the size and proportions of the building … those of a house” but the “only parallel … is provided by the houses excavated … at Knap of Howar.” [7]

The Howe excavation team, however, considered it unlikely their structure could be anything other than a stalled tomb.

Countering the claims in 1994, they cited the absence of domestic refuse as evidence it was not a house [4].

Not a particularly strong argument considering the paucity of Neolithic artefactual material overall, not to mention “no direct evidence” the “stalled tomb” had been used as a burial place either. [4]

Although the lack of refuse – domestic or otherwise – could relate to re-use of the site in the Iron Age, excavation at the Ness of Brodgar has revealed that building floors were kept incredibly clean. [6]

At the Barnhouse settlement, House Two was also found to have been “kept in a state of cleanliness exceeding that seen in the other dwellings”. [8]

The nearby standing stone, it was also suggested, made it unlikely that the Neolithic phases at Howe were ever “purely domestic”. [4]

While it could be argued that no Neolithic site can ever be considered “purely domestic”, the presence of a standing stone near a structure does not necessarily mean it was a tomb. At the Ness of Brodgar, there is at least one standing stone in the middle of a complex of large, monumental buildings that have yielded no evidence of a purely funerary or mortuary role. [6] 

The link between the Howe “stalled tomb” and the Maeshowe-type passage grave built on top was also considered proof the excavators had encountered a “stalled tomb”. However, the sequence of structural replacement encountered at Howe has since been suggested for other sites in the vicinity.

At Maeshowe we have evidence of an earlier “house” beneath the passage grave [9] and Professor Colin Richards has argued the Stones of Stenness were erected to enclose the remains of large building. [10]

Aside from the parallels between the Ness of Brodgar’s Structure Twenty-Seven already noted, the relationship between the Howe “stalled tomb” and the adjacent “forecourt structure” is remarkably similar to Houses Two and Nine at the Barnhouse settlement [8] and Structures Fourteen and Sixteen at the Ness. [6]

In both these cases, a smaller building stood directly opposite the entrance to a “special” structure.

Key to questioning the “stalled tomb” interpretation at Howe lay two metres (6.6ft) in from the entrance.

Central to the two inner walls was a small hearth (0.8m x 0.6m) – a feature notably absent from all other stalled cairns in Orkney. Like the example in the “mortuary house”, this hearth showed little evidence of burning, leading to the suggestion it was not used often – and possibly only used once before both buildings were deliberately dismantled [4].

To Richards et al, however, the presence of the hearth meant the Howe structures can be “confidently reinterpreted as early Neolithic houses facing one another”. [11]

The Howe excavators got around the problem of a hearth in a suggested funerary structure, and its apparent lack of use, by suggesting it may have been part of the “final deconsecration process prior to the abandonment and careful demolition of the tomb and forecourt structure”. [4]

This raises interesting questions because the hearth in the “stalled tomb” was located exactly at the entrance of the passage grave that replaced it.

We do not know how much time elapsed between the two structures – presumably the construction of the passage grave took place before its predecessors were forgotten – but the position of the hearth seems to have been significant.

If the earlier buildings had been completely levelled when construction began, the hearth’s location must have been visibly marked. Alternately, if the remains were still visible, did the hearth play a role in the early stages of construction? – as has been suggested for the entrance hearth at Barnhouse Structure Eight. [12]

What is certain is that anyone entering or exiting the passage grave passed over a fireplace – a situation similar to the hearth at the entrance to Barnhouse Structure Eight, where “fire appears to have been used to formalise a boundary.” The site selected for Structure Eight was also determined by the location of an earlier hearth. [12]  

We will return to these hearths later.

The lack of dating evidence means we do not know how long the first buildings at Howe were in use, nor when they were founded.

If the so-called “stalled tomb” does represent a “house” that shared architectural features with the Knap of Howar and Structure Five, then we could be looking at c3300BC. On the other hand, Structure Twenty-Seven at the Ness of Brodgar also has similar features and may be later in date. [6]

Whatever the function of the “stalled tomb” it was, along with the associated “forecourt structure”, abandoned after an unknown – but possibly lengthy – period of time. This, it was proposed, was triggered by two episodes of collapse in the latter, one of which blocked the entrance passage. [4]

“Inadequate foundations” were blamed for the collapse [4], a situation reminiscent of the large, piered buildings at the Ness of Brodgar, where appearance seems to have been more important than structural stability. [6]

The passage grave

The two abandoned phase-one buildings were carefully dismantled and their remains covered by a thick layer of clay. This created a 25-metre-diameter (83ft) platform for the construction of a circular, Maeshowe-type passage grave.

With a diameter of 19m (62.3ft), the cairn contained a 2.5m-square (8.2ft) central chamber, three side-cells, an underground cell and a long entrance passage.  Surrounding the platform was a seven-metre-wide (23ft) ditch. [4]

Outside its south-east-facing entrance, a forecourt enhanced “the grandeur of the tomb”. [4]

Access was by the site of the levelled “forecourt structure”, which acted as a causeway across the one-metre-deep (3.3ft) ditch. [4]

The use of this earlier building meant the path of movement through the passage grave’s predecessors was maintained and incorporated into its layout. This brings us back to the apparent connection between thresholds and hearths. Before the “forecourt structure” was dismantled its fireplace was covered over – sealed – by stone slabs. [4]

The end result was that anyone seeking access to the new passage grave was forced to cross two distinct thresholds – the causeway and the entrance – both marked by concealed hearths. At Barnhouse Structure Eight, the entrance passage incorporated a single, concealed hearth. [12]

As with the earlier buildings, the internal features of the passage grave were fragmentary – badly damaged by the later Iron Age re-use of the site.

From the quality of the surviving masonry, however, the excavators had no doubt the structure was “one of the finest Orkney Neolithic tombs yet discovered, possibly the equal of Maeshowe in its quality constructional details.” [4]

They proposed that Howe was later than Maeshowe because “it appears to be a more advanced form of tomb”.

It was, however, “so close in style to that at Maeshowe that it was probably that the same builders were involved.” [4]

The passage grave was the last Neolithic construction, after which it seems the site lay dormant until the Iron Age.

The only evidence of Bronze Age activity was fragments of Beaker pottery outside the passage grave’s entrance [4] – but whether these were in some way related to the closure or decommissioning of the structure is not clear.

The standing stone

The fourth Neolithic feature at Howe was a “substantial standing stone” perhaps 3-4 metres (9.8-13.1ft) tall. [4]

Although the megalith had been removed in prehistory, its one-metre-deep socket hole suggested it was “at least as tall as the largest of those at the Stones of Stenness or Ring of Brodgar”. [4]

The socket hole had also been damaged by later activity meaning the chronology of the standing stone is not clear. We do not know when it was erected nor when it was removed. 

Situated north-east of the earliest building – the “mortuary house” – the excavators suggested the megalith pre-dated or was contemporary to it. The fact the “stalled tomb” was built on a different alignment to the “mortuary house” was, they suggested, because the standing stone was in the way. [4]

The socket lay to the east of the passage grave, but whether the monolith still stood during its lifetime is not known.

The fact the socket was on the inner edge of the ditch could mean the stone was removed before it was cut. That said, the fact the socket was not obliterated by the ditch might suggest the standing stone remained in situ until the early Iron Age, at which point the hole was filled with rubble. [4]

Tomb or house?

Howe was an incredibly complex site and the Neolithic layers just a small, highly fragmented fraction of millennia of activity.

The complete lack of dating evidence from these sections makes it difficult to interpret the earliest Howe structures and their role in Neolithic Orkney. Although there is little doubt the phase two building was a Maeshowe-type passage grave, this does not help us date its predecessors.

Maeshowe-type chambered cairns were long considered representative of the Late Neolithic and therefore later than the Orkney-Cromarty stalled cairns. [7]

We now know this is not necessarily the case.

A re-analysis of Orkney radiocarbon and luminescence dates in 2017 prompted “a radical reassessment” of Neolithic Orkney. [13]

Among the findings of the Times of Their Lives project was that both styles of chambered cairn were first built in the middle of the fourth millennium BC — “although, with current evidence, it is not possible to state which came first”. [13]

So, we now have a situation where it seems both types were used concurrently, with deposition of human remains in Orkney-Cromarty structures ending around 2900BC — four centuries before the practice ceased in the Maeshowe group. [13]

With both styles, however, activity continued after these dates, specifically the deposition of animal remains. [13]

This, together with no secure dates for Maeshowe, means the excavators’ suggestion that the Howe passage grave was a later, superior copy cannot be placed in a timeframe more precise than c3400-2500BC – a period spanning both the Early and Late Neolithic in Orkney.

Much has changed since the Howe excavation in the 1970s and 80s. Back then the excavators had to rely on previously uncovered evidence from Orkney – hence the “mortuary house” and “stalled tomb” interpretations.

Considering the discoveries made in the past four decades, and the new information available to us, it is perhaps time to abandon these interpretations.

But could the “stalled tomb” have been an Early Neolithic house, as suggested by Davidson and Henshall? [7]

It certainly contained a hearth and paired orthostatic dividers, as would be expected in an Early Neolithic house. These, however, tended to be oval shaped with rounded ends. The Howe buildings, with their square ends, are more reminiscent of Late Neolithic architecture – and, to add to the confusion, stalled cairns!

The Howe “stalled tomb” did not have the “piers” found within the “big houses” at Barnhouse and the Ness of Brodgar, which date to c3200BC, but is remarkably similar to Structure Twenty-Seven – a building whose role and date are, as yet, unclear.

How this 17-metre-long building relates to the neighbouring, piered structures at the Ness is also uncertain at present.

Central to the suggestion the pre-passage grave structures at Howe did not have a funerary role is the presence of hearths.

Professor Colin Richards is clear that hearths in the Neolithic were “a potent symbol that embodied the warmth of life” and therefore “not appropriate within the context of death.”

As we have seen, he is confident that the Howe structures represent Early Neolithic houses.

Adding to the argument against the “stalled tomb” interpretation is the fact there was no evidence of a cairn containing the structure [14] – it was, along with its companion, apparently freestanding. This is atypical of other Orcadian examples of Orkney-Cromarty chambered cairns. [7]

The clear pairing of the “stalled tomb” and “forecourt structure” also casts doubt on the tomb interpretation. At present there are more examples of paired buildings in the Orcadian Neolithic than stalled cairns with associated outbuildings.

The lack of domestic refuse, cited as proof the building was not a house, does not stand up when we consider discoveries made in the past four decades. Aside from the complete paucity of any Neolithic material – domestic or otherwise – encountered by the Howe excavators, we now also know that it was not uncommon for the “big houses” at Barnhouse and the Ness of Brodgar to be kept scrupulously clean.

It may be that the poorly preserved Howe structures had roles beyond simple dwellings – and here the presence of the standing stone may become significant – or, like Structure Twenty-Seven, served a purpose we cannot yet recognise.

Even in settlement contexts, differences in architecture and use have been documented.

Beneath Cuween hill in the parish of Firth on the Orkney Mainland was the Stonehall Neolithic settlement.

There, on a glacial knoll, a series of structures were built over a 4,000-year period – from around 3400BC to AD380-550.

The Neolithic structures spanned “several hundred centuries” [15] and activity in and around the final building was markedly different to the rest of the settlement. [15]

The stone tool assemblage was different, the diet of the inhabitants was different and it seems specific tasks took place on the knoll.

Above all, activity in and around the house was clearly visible from the dwellings clustered below.

An interesting element of the Stonehall knoll buildings was their elevation. But siting structures overlooking the lower settlement was not without problems. But despite the challenges of living there occupation persisted, prompting the suggestion that the knoll was clearly “a desirable place to live”. [15]

This persistence in the face of adversity led to the conclusion that “… it is difficult not to translate physical elevation into enhanced social position.” [15].

Regardless of what the phase one buildings at Howe represented, was their hillside position perhaps connected to a settlement on lower ground?

Do we have a situation where a structure/house was set apart from its contemporaries and which went on to be monumentalised and finally contained by a ditch-enclosed passage grave? We can only speculate as it is unlikely we will ever know.

Perhaps Structure Twenty-Seven at the Ness will yet help unlock the secrets of Howe. Only time will tell.

Notes

• [1] From Old Norse “haugr”, meaning “(burial) mound”.
• [2] https://canmore.org.uk/site/1731/the-howe
• [3] “Then there is the story of a young Orphir man who spent a whole year among the fairies in the Hillock of Howe, near Stromness, and of how he danced there all that time with a jar of whisky on his shoulder. When he was finally pulled out, he complained he had not been allowed to finish his jig. It was only when he saw the uppers of his boots dangling about his ankles that he realised he had danced the soles completely off.” Leask 1931. A dialect version of the tale appears in (1910) Old Lore Miscellany of Orkney, Caithness and Sutherland. Volume 3. There are variants of the Howe accounts but they can all generally be found applied to mounds throughout Orkney.
• [4] Ballin-Smith, B. (ed) (1994) Howe: Four millennia of Orkney prehistory. Society of Antiquaries of Scotland.
• [5] This article will use the excavators’ names for the Howe structures for ease of reference. It should be remembered that although these names imply definite roles for the buildings, their roles and functions remain open to interpretation.
• [6] Card, N., Edmonds, M. and Mitchell, A. (eds) (2020) The Ness of Brodgar: As it Stands. The Orcadian: Kirkwall.
• [7] Davidson, J.L. and Henshall, A.S. (1989) The Chambered Cairns of Orkney: an inventory of the structures and their contents. Edinburgh University Press.
• [8] Richards, C. (2005) The Ceremonial House Two. In Richards, C. (ed) Dwelling among the monuments: the Neolithic village of Barnhouse, Maeshowe passage grave and surrounding monuments at Stenness. McDonald Institute for Archaeological Research: Cambridge.
• [9] Challands, A., Muir, T. and Richards, C. (2005) The Great Passage Grave of Maeshowe. In Richards, C. (ed) Dwelling among the monuments: the Neolithic village of Barnhouse, Maeshowe passage grave and surrounding monuments at Stenness. Cambridge: McDonald Institute for Archaeological Research, pp. 229–248.
• [10] Richards, C. (2013) Wrapping the Hearth. In Richards, C. (ed) Building the Great Stone Circles of the North. Windgather Press, Oxford. Pp. 64–89.
• [11] Richards, C., Downes, J., Gee, C. and Carter, S. (2016) Materializing Neolithic house societies in Orkney: introducing Varme Dale and Muckquoy. In Richards, C. and Jones, R. (eds) The Development of Neolithic House Societies in Orkney: Investigations in the Bay of Firth, Mainland, Orkney (1994–2014). Windgather Press. pp.224-53.
• [12] Hill, J. and Richards, C. (2005) Structure Eight: Monumentality at Barnhouse. In Richards, C. (ed) Dwelling among the monuments: the Neolithic village of Barnhouse, Maeshowe passage grave and surrounding monuments at Stenness, Orkney. Cambridge: McDonald Institute for Archaeological Research.
• [13] Bayliss, A., Marshall, P., Richards, C. and Whittle, A. (2017) Islands of history: the Late Neolithic timescape of Orkney. Antiquity, 91(359), 1171-1188.
• [14] Carter, S.P., Haigh, D., Neil, N.R.J. and Smith, B., 1984. Interim report on the structures at Howe, Stromness, Orkney. Glasgow Archaeological Journal, pp.61-73.
• [15] Richards, C., Brophy, K., Carruthers, M., Jones, A., Jones, R. and Jones, S. (2016) Good neighbours: Stonehall Knoll, Stonehall Meadow and Stonehall Farm. The development of Neolithic house societies in Orkney. In Richards, C. and Jones, R. (eds) The Development of Neolithic House Societies in Orkney: Investigations in the Bay of Firth, Mainland, Orkney (1994–2014). Windgather Press, pp. 91-127.