We knew from our work in the ‘70s and ‘80s that the dry land on either side of the wetter, lower lying parts Flag Fen basin had been settled and farmed from at least 3000BC, when most of the natural tree cover had also been removed. The lowest areas began to flood shortly after that date, as a result of the warmer climate, and rising sea levels, that followed the end of the last ice age, some ten thousand years ago. It seems highly probable that the early Bronze Age inhabitants of the area had already laid-out a route across the lowest land (which would probably have represented a flood hazard in times of heavy rainfall) even if the larger Fens had yet to start forming. Rainwater will always pond in flat, low-lying land, especially if the natural drainage is not too good. We found very slight evidence for such a route when we had removed the last of the posts at the power station site, in 1989. It consisted of a shallow ditch that more-or-less followed the line of the later posts.
The big event took place sometime around 1300BC when a double row of posts was set in the ground in a near-straight line running east-west from Fengate to Northey, across the narrowest stretch of wet ground they could find. These early posts were mainly of willow and alder, wet-loving trees that would have been growing locally. We’re still not certain of the precise chronology of what happened next, but the first two rows (which we labeled Rows 1 and 3) were then augmented by another three, all of which can be seen in the Preservation Hall, on site. There can be no doubt at all that the posts were used to mark-out and stabilize a causeway over the wetland, because not only have we found dozens of carefully arranged planks, all of which had been pegged into place, but the wooden surface has been liberally dusted with sand and fine gravel, which can clearly be seen alongside the central post Row 3 and between rows 1 and 2. Traces of other paths, all running parallel with the posts, were observed during the excavations. We have analyzed the sand and gravel and can state positively that it was not washed there by water, but had either been spread on the ground, or had got there on people’s feet. On the whole we think it was deliberately spread, as it includes some larger pebbles. Wet wood can be very slippery underfoot.
The Preservation Hall
At its Fengate landfall the causeway precisely lines up with a pre-existing ditched droveway. The Fengate Bronze Age landscape was parcelled-up into fields which were separated into individual farms or holding by tracks known in the Fens as droveways. Droveways became even more important in the Middle Ages when they were used to drove or drive large herds of cattle or sheep from remote farms out in the fen, to the main markets, at towns like Boston, March, Wisbech, Holbeach and Spalding. The Fengate droves were smaller in scale – none were longer than a kilometer – but they served much the same purpose: to take livestock safely from A to B, without having them wander off into another person’s farm. We didn’t find evidence for another droveway at the Northey end of the causeway, which argues that animals as such did not use it. Common sense suggests that cattle would soon damage their feet attempting to cross over on so much loose timber. Even though they are mountain animals, sheep, too, are very wary about walking over unstable surfaces. It would seem, then that the Flag Fen causeway was used only by people, and that they were probably on foot.
Various Flag Fen publications describe the posts as being part of a ‘post alignment’ rather than a causeway. The reason for this is that the term causeway conjures up a relatively simple picture in the mind’s eye. Modern causeways are fairly basic: they’re a means of guiding people across wet or boggy land, sometimes (but not always) dry shod. But there was far more to the Flag Fen causeway than that. For a start it seems to have been built on an unnecessarily lavish scale, with no less than five rows of posts. The posts, too, are very large and many are made from oak, which must always have been a very expensive raw material, as it is by far and away the best (i.e. longest lasting and strongest) building material. We have also found one or two posts that either collapsed or were never used and these suggest that at least a proportion of posts could have been very high – and would have stood 10-12ft (over 3.5m) above the surrounding Fen. So they would have looked spectacular. There is also a problem with ‘Row’ 4, whose posts don’t form a tight row, so much as a concentrated spread. This row resembles more a military palisade than a line of guiding markers.
The final problem with the simple ‘causeway’ explanation has been the discovery of hundreds of complete or recently broken objects around or close to the posts. These are not the sort of small, day-to-day objects one might expect to have fallen out of travellers’ bags or pockets – the modern equivalents of Coke bottles, crisp packets and hamburger boxes. They are far more up-market and include about a dozen bronze swords and daggers, bronze spearheads, brooches, pins, rings, fine-turned shale bracelets and complete pottery vessels. There are other strange finds too, including four almost unused corn grinding stones, or querns, together with a number of dog skulls. These finds strongly suggest that this over-elaborate, and possible defended ‘causeway’ also served an important religious or ceremonial role. So we decided to use a new term to describe it: a post alignment.
Study of the growth-pattern of the tree rings of the oaks in the alignment gave us broad dates for phases of activity, but we encountered problems when it came to the linking of these general episodes of building and rebuilding to the actual timbers themselves. This failure to provide a tightly dated structural sequence reflected the fact that most of the dated timbers were posts that had been driven down from the top. Further problems were caused by the fact that many of the posts had lost their bark and much of their sapwood, which made it impossible to arrive at actual dates, rather than date ranges. Even so, we able to make a reasonably well-informed guess at the development of the post alignment: between 1300 and 900BC. In short, after a slow start the structure grows in complexity and a remarkable robust, corduroy-like, layer of timbers, known as the ‘log layer’ was laid down to act as a foundation that was not likely to sink. The layers above the ‘log layer’ were a succession of walkways, mostly on either side of Row 3. As we have seen, the scattering of posts known as Row 4 may actually have formed an irregular wall or palisade of posts. The post alignment reverted to a simple footpath in its latest phase, perhaps around 1000BC. There were also indications for an unexpected partitioning or transverse element in at least one place where the post alignment crossed the platform. This may have been a structural necessity, but other explanations (which I will discuss in the next section) are also possible.
Row 3
It was not until we had completed most of the post-excavation analyses, by 1996, that we were able to plot the types of wood used at Flag Fen. The main conclusion was that fenland species, such as alder and willow, occurred most frequently in the earliest phases of the structure. It was not until later that oak began to be used at all frequently. This is not what one would have expected and may reflect the site’s growing importance, because oak is a valuable structural timber, which could not have been grown in the wetland itself, and must have been transported to Flag Fen, either from the higher ground in the immediate vicinity, or from further afield. The preponderance of fenland species lower down in the sequence doubtless reflects the fact that the alder carr woods around the edges of the Flag Fen basin had to be cleared, before construction of the post alignment and platform could begin.
The last timbers of the post alignment and platform were added shortly after 900BC, in the Late Bronze Age, but the site of the alignment, which would still have been visible during the drier months of the year, continued to be visited, and offerings were made there throughout the Iron Age.
An impression of the post alignment by Rob Fuller
The Bronze Age fields at either end of the post alignment began to be abandoned from about 1400BC, and there was a shift southwards, towards slightly higher ground. At the same time it is now becoming increasingly clear that the wetter parts of the basin were also rapidly growing in importance. Sites around Must Farm, for example, show clear evidence for a growing population and increasing prosperity. But their focus is clearly towards the wetland and the emphasis seems to be more on fishing and wildfowling than on agriculture. By about 700-600BC the process of abandonment was largely complete. The livestock landscape of the Bronze Age was followed, in the Iron Age, by a landscape dominated by mixed farming of both cereals and livestock. By now the emphasis was shifting both towards the drier hinterland, as the presence of a single large farm droveway clearly demonstrates. The settlement pattern altered too: from about 600BC we see the appearance of the first nucleated, village-style settlements that replaced the isolated farmsteads of the Bronze Age.
The transition into Roman times appears to have been smooth and uneventful. Pottery found there shows that the substantial Iron Age farm near the Cat’s Water in Fengate continued to be used into the third century ad. Finally, the area has not produced evidence for post-Roman nor Anglo-Saxon settlement, almost certainly because conditions were too wet underfoot. By now, it was an area reserved for seasonal pasture – hence the name Fengate, which means ‘road to the fen’.