The narrative below summarises our current thinking on how European surveys evolved through the Neolithic.
The Anatolian farmers arrived in France about 5000BC and causewayed enclosures (‘enceinte à fossés interrompus’) then start appearing there. Perhaps the farmers brought surveying skills with them - given the sophistication of the society now being excavated at Gobekli Tepe near the Türkiye / Syria border?
These people started settling in Britain around 4200-4000BC, arriving from Brittany, NW France, and the Low Countries. On arrival, some strands of migrants (the ‘strand’ model of Neolithization and dates are taken from Sheridan 2010) surveyed their new landscape. Hence, we see causewayed and tor enclosures being built. They were initially built and used to survey but were then used for other activities. They were used for a relatively brief period – 3800-3500BC (Historic England 2018a). The enclosure survey covered much of the area south of the River Trent. However, there are few enclosures known elsewhere which suggests that the survey did not reach, or was very coarse grained, across much of Ireland, Wales and Scotland. The angles between the enclosures suggest they were comfortable measuring 1/36 and 1/24 of a circle i.e. multiples of 10 and 15 degrees, and possibly 1/72 i.e. multiples of 5 degrees.
In the British Isles, the survey methodology then developed over the next 1000 years. The chronology of the monuments suggests that three further distinct phases of surveying followed, although some techniques were used in more than one phase.
In the second phase, rather than have the survey geometry, and hence survey stations, be constrained by the landscape, they chose to impose a standard geometry across the land. The second survey is characterised by the cursus monuments.
A national geometry of triangles, with sides extending from the south of England to Orkney, was planned. These were laid out and further lines were set at 90 degrees to the national lines, often with the assistance of a cursus. The cursus and the new line were often set to exploit a prominent hill, so that the new line could be easily extended by sighting on that hill. Further lines, this time parallel to the original line could now be set from the new line. Even more lines were generated by setting lines at 60, 45 and 30 degrees to first- and second-generation lines. The survey extended across the Irish Sea. It is likely that the Dunkery Beacon (Somerset) to Leith Hill (Surrey) line was the baseline for the whole scheme.
The exceptionally large 'megalithic' stone rows on Dartmoor relate to the enclosure survey, the cursus survey or both.
The next survey, characterised by circles and henges, extended to a fine level of detail across all of the British Isles. These tools enabled the surveyors to have more choice about where they set lines compared to the rigid cursus grid. The flexibility came from being able to set up as many circles, henges, and circle-henges as were useful and that these systems allowed a greater range of angles to be set. The earlier enclosure and cursus surveys are dominated by angles which are multiples of 10 and 15 degrees i.e. 1/36 and 1/24 of a circle. The stone circles, as evidenced by the remains we find today, supported 1/72, 1/30, 1/20 of a circle i.e. multiples of 5, 12 and 18 degrees as well as of 10 and 15 degrees. It seems likely that the contemporary henges worked to the same angles. Note that in this system, the survey still imposed a geometry on the landscape, but the surveyors had more choice as to how the geometry lay on the landscape. They were very good at reading the landscape to place new circles / henges at locations that they could fix, and from which new lines would enable them to fix as yet unsurveyed natural landmarks.
Stonehenge phase 1 was one of the first henges along with Landygai, Brodgar and Stenness (Historic England 2018b page 3). It lies on or remarkably close to the Dunkery Beacon-Leith Hill line suggesting that line remained a key line in this phase. However, by 2500BC, with the further development of nearby Avebury, Stonehenge had become less important as a survey centre.
Circles and henges are rare in the Midlands/East Anglia region, yet this is where the exceptionally precise Type Bi cursus sites are found. That suggests that in this relatively flat region, the Type Bi cursus survey could not be improved upon. Loveday p115 reports that the Drayton South Bi cursus has a date of 3200BC but that Grooved ware pottery (p193) indicates it was also in use much later. This fits with the third survey, and perhaps the fourth survey, pragmatically reusing the cursus survey where appropriate.
The final, fourth, phase is strongly evidenced in Dorset and Wessex. Here we find the mega-henges and the artificial hills of Silbury Hill and Marlborough Mound– all proven to be late Neolithic. There were numerous innovations, including the introduction of an accurate baseline running between these two artificial hills.
They also substantially increased the size of some henges, not least at Avebury. Given Avebury's close relationship with the Silbury-Marlborough baseline, it was probably the most important measuring centre of the survey. However, the other mega-henges were also important.
We suspect that their method shifted towards measuring angles within the landscape i.e. akin to modern trigonometrical surveying, rather than imposing certain angles. This is suggested by the hundreds of post holes found in the circuit around Mount Pleasant. The post holes are usually interpreted as evidence of a palisaded enclosure. However, the estimate of 1600 post holes hints at a measuring instrument calibrated to about ¼ degree. (1/4 degree would require 1440 posts and might be measured out as 1/20 of 1/72 of a circle). Detailed analysis of the post hole locations will reveal whether they do match an evenly calibrated dial and what the calibration was. The Avebury Outer Stone Circle may have served a similar purpose.
Survey points on hills and ridges were often marked by a large barrow or cairn. These landmarks preserved locations that had been fixed, or aligned, and enabled the survey to be extended further. Whilst many Dartmoor stone rows align to large barrows in north Cornwall and on Exmoor, non-funerary barrows in other regions can also be interpreted as survey markers.
In lowland, round mounds were built to facilitate the survey. Some similarity in form and date between Silbury Hill and the smaller, although still unusually large, barrows at mega-henges Mount Pleasant, Marden and Knowlton is noted by Greaney at al. (2020 page 227). Furthermore, we have found examples of mottes and ‘medieval’ moated houses which are located at the intercepts of key survey lines. That these mounds were part of the fourth survey is suggested by their precise location within the survey and their physical similarity to Silbury Hill.
In South West Britain, stone rows set out a mesh of survey lines. The stones used were typically less than half the height of those used in the earlier 'megalithic' rows. The new design of row worked in tandem with an updated design of stone circle. Such circles comprised carefully cut stone slabs to improve accuracy and enable more angles to be set. On Bodmin Moor and Dartmoor there are clear row/circle relationships and well-preserved circles which demonstrate this.
In Orkney, the Ring of Brodgar was built to replace the Stones of Stenness. The site of the new monument was just 3km to the north west of the old. Given Orkney is about 850km from Avebury, and 875km from Stonehenge, this relatively small adjustment suggests that the third survey was itself effective but they wanted. and believed they could achieve, even more accuracy.
In mainland Western Europe, Causewayed and Tor Enclosure type surveys started over 1000 years earlier than in Britain. Later, a Circle type survey was planned in France. It was started in Brittany but did not extend further.