stone rows in their landscape - DARTMOOR

Introduction to the Dartmoor Rows

There are 84 rows on Dartmoor recognised by Gerrard (gazetteer) and 11 more on nearby Bodmin Moor. The Dartmoor concentration is exceptional because 30% of all the rows in Great Britain are found there.

Most of the stones in most rows are medium sized stones, typically 40-100cm high, 20-40cm thick and 40-100cm long (in the direction of the row). Rows of similar sized stones are found elsewhere in South West England and in South Wales.

Several authors have looked for patterns in the locations and the designs of rows to find clues as to their purpose.

The Dartmoor rows tend to be on gentle slopes and to be towards the edge of the moor. The rows are sometimes clustered in a group with two to five others. However, it is also the case that several rows are on steep slopes, some are towards the moor centre, and many stand alone. The lack of rows on the northwest side of the moor has long been noted.

The variation in length is huge - from 3m to over 3300m.

No generally accepted explanation has emerged to explain their orientations, although some have suggested solar or stellar links for some rows e.g. Walker (2005) and Morgan (2016).

Also, no reason has emerged as to why about 45% of the 84 rows have been built as a single line of stones, about 42% as double lines and the remainder are triple lines except for one case of about 7 lines.

Discussion and supporting data on these points, and more, can be found in Emmett (1979), Butler (1997, chapter 5) and more recently Gerrard (research/function).

Analysis of alignment to distant features turned out to be difficult. Most published row orientations are of the whole row and not by section. For those rows with just one section the published orientation is useful, but it is usually only given to the nearest degree which represents a rather wide arc of 520m even at just 30km distant. We persevered by taking row section orientations from published plans. However, here doubt is cast when it is unclear whether the plan is orientated to grid north or true north – a difference of 1.5 degrees on Dartmoor. We therefore know that our method has pitfalls. Nevertheless, our interest is maintained because plausible alignments so frequently present themselves.

Our work to assess alignments to landmarks on the skylines away from the moor continues. We are now refining our results by taking the orientation of row sections from GPS surveys of the individual stones in the row.

Our current estimate is that 75%-80% of row sections plausibly align on a natural landmark or a large barrow – either on the moor or off the moor. Assuming this number is confirmed, it strongly supports the idea that most rows were designed to either align on landmarks, or to work in tandem with a large barrow.  

‘Adjacent’ rows

Rows are found concentrated together at sites such as Merrivale (6 rows), Drizzlecombe (3 or 4 rows), Fernworthy (3) and Shoveldown (6). Stone circles or large barrows are often nearby and so the sites have been thought of as ceremonial complexes.

At such sites some rows converge on the same point, others seem laid end to end, some cross each other and others are just in proximity.

There are also row sections that are set end to end but are not in a ‘ceremonial complex’. Hence, each section of the Upper Erme Row is set end to end with one or two other sections. It is the same in cases where a row seems to comprise two straight sections e.g., Higher White Tor, Holne Moor and, as previously mentioned, Merrivale 2 and Hingston Hill.

In future discussion we will use the term ‘adjacent’ to describe rows which are close to one another in any of the above arrangements.

‘Networked’ rows

Whilst looking for aligned landmarks, we noticed that sometimes the line of a row goes exactly to the end of another row several kilometres away. In other cases, the row line intercepts the row at the point where we suspect row sections meet. A further variation, albeit rare, is that the row line goes to the ends of two different rows, thereby linking three rows together.

A further type of relationship between rows is therefore where two or three rows, several kilometres apart, are ‘networked’ by the row line from one of the rows.

The table shows two sets of examples. The rows on each line of the table directly network with each other. However, it can also be argued that all the rows within each set are networked, some directly but others indirectly, to each other.

Interpretations

One interpretation of these observations is that the line from row to landmark was important because it connected venerated places in the landscape. The row marked out part of the line which was especially important for some reason. This interpretation chimes with what has been observed in various Neolithic / Bronze Age landscapes where views to mountains and intervisibility between monuments seem to have been important (Scarre 2007, p19). It might well follow that burial on such lines was considered beneficial and that is why we find cairns aligned with the rows.

However, that some row lines form a network with each other, that the landmarks are prominent hills, and that some landmarks seem to be a focus of interest of more than one row (either aligned on or suddenly appearing at the end of a row) are all reminiscent of the way modern surveyors fixed locations and used ‘trig’ points prior to GPS.

We further suggest that ‘adjacent’ and ‘networked’ rows are consistent with a surveying purpose because in that context, once a location has been fixed, it makes sense to exploit that fix - by using locations easily measured or fixed from it. One would expect rows to be added at a location, and locations fixed from it, whenever the additional row took the overall survey forward and the landscape allowed.

The row characteristics described so far are consistent with both the ‘venerated landscape’ and ‘survey’ interpretations. Indeed, the two interpretations are not necessarily mutually exclusive if the rows were survey tools used to set straight lines between sacred points in the landscape.

However, in the first interpretation the siting of any one row was determined by the need to embody part of the line between venerated places and, as Gerrard suggests, to create visual links to other key landscape features. The second interpretation implies a broader aim (and a much higher level of organisation) because the goal was to build a network of rows and lines which would work together. In this case the siting of a row was determined by the needs of the overall survey.

A survey interpretation for the rows leads one to consider whether barrows and cairns located on the row line also had a survey purpose. For example, a cairn could provide height to gain a visual link and/or create a landmark for surveyors elsewhere to sight upon.

Rows hypothesis

Our Dartmoor stone rows hypothesis i- which sits within the broader surveys hypothesis - is that:

·        The straight sections of the Dartmoor rows were tools which enabled the laying of a network of survey lines.

·        The lines stretched across Dartmoor. They also went from Dartmoor, across the surrounding land, to hills and ridges generally on the skyline.

·        Barrows and cairns were built to assist the surveying.  

·        Later some survey mounds were reused as burial mounds.

·        The survey barrows also encouraged more barrows and cairns to be built nearby. This time the prime purpose was funerary.

The hypothesis does not explain the ultimate purpose of the surveying and the lines. It leaves open whether it was to produce a map; to manage the land; to navigate; to connect places of spiritual importance and/or other reasons.

 Summary of observations consistent with row hypothesis

Below, we summarise and extend our list of observations that seem consistent with the row hypothesis.

Observations mentioned so far that seem consistent with the row hypothesis are:

1.      Alignment of straight row sections to prominent hills or large barrows.

2.      Rows offering the observer great precision as to whether they are in line or not.

3.      Rows often sited in places with extensive views - so sight lines to landmarks and other survey stations were unhindered wherever possible.

4.      Some rows ending just where a prominent landmark comes into view (not necessarily in line with the row) - creating a sight line to that landmark.

5.      Locations with multiple rows in proximity.

6.      Row sections laid end to end.

7.      The end of some row sections being on the line from another, forming a network over several kilometres.

To these might be added:

1.      The Dartmoor field boundaries known as ‘reaves’ are generally straight lines and can stretch for up to 10km across a hilly landscape. Andrew Fleming investigated the reaves in the 1980s. He pointed out (Fleming 1988 p67) that the ends of some very straight reaves are not intervisible which implies that by the time the reaves were built, in the Bronze Age, good surveying skills existed.

2.      Evidence that the reave builders respected but did not revere the rows (Fleming 1988 p45; Gerrard /Research/Dating). This is consistent with the rows being seen as useful tools to be retained where possible, but they did not hold spiritual significance in themselves, and so could be slighted if necessary.

3.      Some cairns at the end of rows contain charcoal but not human remains e.g., Laughter Tor and Stannon Newtake (Butler 1997 p223) - consistent with fire or smoke being used to help observers see a location or to exchange signals.

4.      Tall standing stones at the end of rows (e.g., Laughter Tor. Langstone, Drizzlecombe) – perhaps so the location can be seen over a rise, from afar and/or to assist alignment along the row.

No doubt some of these observations and interpretations are open to challenge. Also, no single observation is proof of the hypothesis. However, the sheer number of features that are consistent with the hypothesis weigh in favour of considering it further.