Star Carr

Star Carr lies under the fields of Seamer, in the eastern Vale of Pickering in North Yorkshire. In the Early Mesolithic it stood at the western outflow of a lake that no longer exists. John Moore, who first mapped the early sites on these carr lands, called that water Lake Flixton. One account makes the former lake nearly five kilometres by two, lying mostly to the east of the settlement.

Another places the modern site about five miles, or eight kilometres, south of Scarborough, and a third places it about eight kilometres from the present North Sea coast. The basin itself was cut by ice. At the end of the last cold stage the vale drained west, away from the shorter route to the sea at Filey, and the hollow filled. Islands and peninsulas of sand and gravel stood in the water.

Wetland plants began to colonise the lake in the Late Glacial Interstadial, and marl formed in the deeper hollows. From the start of the Holocene the lake filled with organic mud and peat. By about 5500 BC it had almost gone, leaving reed swamp, fen and carr, with streams and pools between them.

The people who used Star Carr arrived in a landscape that had only just become warm. Less than four hundred years before the settlement was inhabited, the climate of northern Europe had warmed quickly. That warming closed the Late Glacial Stadial and opened the Holocene, the warm interval that is still underway. One later telling of the same change says the banks of the lake were already about four centuries into the Holocene when hunter-gatherers settled there.

Open grassland was turning to birch wood. Horse and reindeer, long dominant, were probably gone from mainland Britain, while roe deer and aurochs were establishing themselves. Birch, aspen and willow grew along the shore. Reeds, water lilies and other water plants filled the shallows, and the lake level rose and fell with rain and snowmelt.

Elk, red deer, wild boar, beaver, hare, wolf, lynx, bear, fox, pine marten, badger and hedgehog are all among the mammals for which the place has produced evidence. Birds were there too.

This was the second settling of Britain in the early Holocene, if the gap argued by later researchers holds. People had already crossed the still-dry North Sea plain from north-west Europe at the start of the warm interval, hunting horses in a mobile way that belongs to the very end of the Palaeolithic. A short cold relapse, the Preboreal Oscillation, seems to fall in the pause between that first arrival and the Early Mesolithic. By the time Star Carr was in use, stone tools had changed.

They were more adaptable, and axes were among the new forms. People also burned wetland and woodland plants on purpose, and more than once. The first evidence for large-scale, controlled carpentry in this part of Europe belongs to the same early post-glacial centuries.

The Mesolithic is the Old World stretch between the Upper Palaeolithic and the Neolithic, the last long period of hunter-gatherer life in Europe and the Middle East between the end of the Last Glacial Maximum and the coming of farming. In Europe that span is put at roughly 15,000 to 5,000 years before the present. In the north, marshland food was rich enough that farming arrived late, about 5,500 years before the present in some accounts. The word itself was added in 1866 by Hodder Westropp, a year after John Lubbock had set out a Palaeolithic and a Neolithic.

John Evans thought the ages blended into one another. Gabriel de Mortillet thought a gap separated them. The name stayed disputed. In northern Europe it is the usual word.

Vere Gordon Childe could still dismiss the period in 1925, and even in 1947, when he accepted that a middle age was a useful idea, many prehistorians had little time for it. In Britain the concept came into general use only in the 1930s, above all through Grahame Clark. The characteristic tools are small flints, microliths, set into larger weapons of wood, bone and antler. Burials of the period are usually simple.

Art is scarcer than in the periods on either side. A few small engraved pendants are known in the north, some of amber in Scandinavia, and one of shale at Star Carr.

Grahame Clark, whose full name was John Grahame Douglas Clark, was born on 28 July 1907 at Shortlands in Kent and died on 12 September 1995. He published often as J. G. D.

Clark. His father, a stockbroker and a reserve officer, died of influenza on the way home from the First World War. The family moved to Seaford, and the boy collected flints on the Sussex Downs. Cambridge took him as a paying student at Peterhouse in 1927.

He began in history, heard the economic historian Michael Postan, and then moved into archaeology, which was taught with anthropology. Miles Burkitt largely left him to himself. In 1929 he travelled to Denmark and Sweden and met Sophus Müller. His thesis, finished in January 1934, and his book The Mesolithic Age in Britain, published in 1932, mapped British Mesolithic flints at a time when the period was little studied and the evidence was mostly scattered stone.

The book still treated tool styles as cultures and cultures as peoples, but it already looked to pollen, then a Scandinavian method, as a way to follow vegetation and climate. In 1936 he published The Mesolithic Settlement of Northern Europe, which put him at the front of the subject, and in 1939 Archaeology and Society. In 1932 he, Harry Godwin and Margaret Godwin founded the Fenland Research Committee. Their digs at Plantation Farm and Peacock's Farm set British prehistory into an environmental sequence instead of a study of objects alone.

Harry Godwin was born on 9 May 1901 and died on 12 August 1985. He was a botanist and ecologist, Professor of Botany at Cambridge from 1960 to 1967, and in 1948 the founder and first director of the Subdepartment of Quaternary Research there, where early work on radiocarbon dating was supervised. In 1981 he coined the phrase peat archives for the layered record of plants, pollen, animals and archaeological remains held in a bog. The discovery account of Star Carr says that John Moore reached Grahame Clark through Harry Godwin and through the curator of the Scarborough Museum.

Clark's own biographical account says that in 1948 he was told of a Mesolithic flint scatter in the peat at Seamer, found by the amateur John Moore. Walker and Godwin's pollen work later placed Clark's occupation at the end of the Preboreal, Godwin's pollen Zone IV, earlier than typologically similar sites in northern Europe. Harry Godwin and Margaret Godwin had already, in 1933, used pollen to show that a barbed point from the North Sea was of the same age as early Maglemosian sites on the continent. That comparison mattered.

Clark had seen how wetland digs in Denmark and northern Germany kept bone, antler, wood and pollen, while Britain still had little more than flints. He looked at Star Carr because it might do for Britain what those wetland sites had done abroad. He drew a parallel between Star Carr and Maglemose in Denmark. Later work has treated that parallel with care: Star Carr is not a typical British Mesolithic site, and the Maglemosian label does not describe everything that happened on this shore.

The curator in the discovery account is named Arthur Clapham. The later excavation volumes say only that the curator of the Scarborough Museum put Moore in touch with Clark, and they do not give the curator a personal name. Arthur Roy Clapham, the botanist that public account points to, was born in Norwich on 24 May 1904 and died on 18 December 1990. At Arthur Tansley's request he coined the word ecosystem in the early 1930s.

He was Professor of Botany at Sheffield from 1944 to 1969, and with T. G. Tutin and E. F.

Warburg he published the Flora of the British Isles in 1952. He was elected a Fellow of the Royal Society in 1959 and appointed a Commander of the Order of the British Empire in 1969. The excavation volumes do not say that this botanist was the man at the museum.

Moore was a founding member of the Scarborough and District Archaeological Society. Through the late 1940s he recorded early prehistoric sites on the low carr lands. One account gives him nine areas of early activity, many of them Mesolithic on the form of the flints, some sealed under peat. Another says he identified ten prehistoric sites around the former lake and on two small islands.

In 1947 he noticed flints in the spoil of newly dredged ditches in the eastern vale. One of those spots, south of Star Carr Farm and beside the New Hertford River, became the site. In 1948 he cut a small trench there, found bone and antler still intact, and the contact with Cambridge followed. Grahame Clark visited in the spring of 1949.

That summer he began the first of three seasons, 1949, 1950 and 1951, on a small budget, under Cambridge and the Prehistoric Society. Botanists were involved from the start. A later biographer, Brian Fagan, judged the digging adequate but not up to the standard of Mortimer Wheeler. Clark published preliminary reports for 1949 and 1950, finished the monograph in December 1952, and saw it out in 1954.

It became a founding text for the British Mesolithic. During the Second World War he had served in Britain with the Royal Air Force Volunteer Reserve on aerial reconnaissance. In 1951 he was elected a Fellow of the British Academy. In 1952 he succeeded Dorothy Garrod in the Disney Chair at Cambridge, which he held until 1974.

He was elected a Fellow of Peterhouse in 1950 and was Master of the college from 1973 to 1980. He was not widely liked. Colleagues found him competitive and remote. He was still one of the important prehistorians of his generation, because he insisted on economy and environment.

He was knighted, and he was later the subject of Fagan's biography.

What Clark found in the lake-edge mud was, for Britain, extraordinary. The peat had kept organic things that normally rot. He recorded almost two hundred barbed projectile points of red deer antler, the antler they were made from, scrapers of aurochs metatarsals, bodkins of elk metapodials, and modified red deer frontlets. With them were animal bone, antler-working waste and flint.

He thought the worked and unworked birch branches were a platform people had built in the reed swamp, and that the objects were the remains of life lived on that platform. The range of tasks suggested a residential camp of a small number of families, used on at least two occasions. Shed red deer antler suggested winter and spring. In 1972 he went further: Star Carr was the winter camp of people who followed red deer up onto the North York Moors in summer.

That picture, of small groups moving with the seasons between lowland and upland, or between inland and coast, became the usual model of Mesolithic Britain. It lasted because the material was rich, the report was good, and no other British site produced a matching assemblage. It also lasted because Clark himself shaped the discipline.

The dates he could use were coarse. Radiocarbon dating was new. Two of the first radiocarbon measurements on British archaeological material were made on wood from the brushwood he recorded, and they fell in the European Preboreal, in line with the pollen. Later attempts to date his objects failed because the conservation treatments had contaminated them.

The site's age had to be rebuilt from new digging.

The rebuilding took decades, and the numbers do not all agree. A heritage summary says the place was occupied seasonally and intermittently for about a couple of centuries around 9000 BC. The public outline of the settlement says it dates to around 9000 BC, a few centuries after the ice. A wider reading, from the two volumes published in 2018, says the site was in use for around 800 years and is much larger and more complex than Clark imagined.

The Bayesian model of the radiocarbon dates, built with Alex Bayliss, puts the start of Mesolithic activity at Star Carr in 9385–9260 cal BC at 95 percent probability, and probably in 9335–9275 cal BC at 68 percent. It puts the end in 8555–8380 cal BC at 95 percent, and probably in 8525–8440 cal BC at 68 percent. The same model makes the span probably 775 to 885 years at 68 percent probability. The modellers warned that later, drier deposits were hard to sample, so the end date may be a little too early.

An earlier summary of the same model says occupation ran between about 9300 and 8500 BC. A 2020 account by Barry Taylor, Nicky Milner and Chantal Conneller shortens the first phase to about 300 years from around 9300 cal BC, marks a rise in scale from around 9000 cal BC, and says that by around 8500 cal BC activity on both wetland and dry ground had stopped. Another telling says the final use is hard to pin down and that occupation seems to have ceased in 8555–8380 BC.

Before that model existed, the chronology was shorter and differently shaped. In 1985 the Vale of Pickering Research Trust cut a trench to the east of Clark's cuttings to study the setting of the site. They found a small collection of animal bone and part of a wooden platform. The dates misbehaved: samples at the base of the pollen profiles looked too old, the timber seemed centuries younger than objects from a similar level, and some objects seemed older than the sediment around them.

Paul Mellars compared the depth of the objects in that trench with the dated peat and suggested the assemblage there had been deposited between about 8700 and about 8400 cal BC. Petra Dark used charcoal bands as a proxy for people. In one profile the first burning began at 8970 cal BC and lasted about 80 years. After a short gap, a second phase began at 8790 cal BC and lasted about 130 years.

That scheme, of two main episodes, structured discussion for years. Later dates on antler further east showed that similar deposition happened in both phases. The new model replaced the two-phase sketch with a longer, more crowded history.

Paul Mellars was born on 29 October 1939 at Swallownest near Sheffield and died of cancer on 7 May 2022 at Hinchingbrooke Hospital, aged 82. He took his degrees at Cambridge, taught at Sheffield for ten years, and returned to Cambridge in 1980 as a fellow of Corpus Christi. His later work was on Neanderthals and on the spread of modern humans, including the question of why people left Africa around 60,000 years ago. He also studied how Mesolithic hunter-gatherers in Britain responded to climate after the ice, and he excavated early Mesolithic sites on Oronsay in the Inner Hebrides.

With Petra Dark he published Star Carr in Context in 1998. He was a Fellow of the British Academy from 1990, received the Grahame Clark Medal in 2006, and was knighted in the 2010 New Year Honours.

Tim Schadla-Hall and the Vale of Pickering Research Trust had been working around the lake since the mid-1970s. In 1977, ahead of a landfill at Seamer Carr just to the north, excavators found Iron Age, Bronze Age and Mesolithic material. That led to a landscape project. The team returned to Star Carr in 1985 and 1989.

Around the lake they recorded many more Early Mesolithic sites. One later count says fourteen. None of the others had the mass of unusual objects. Only three further barbed points are reported from the rest of the lake, two at No Name Hill and one at Flixton Island, and no further antler frontlets.

Animal bone elsewhere was more broken. From that contrast some researchers argued that Star Carr was a special place in its own landscape, perhaps where animal remains were formally put into the water. The excavators of the 2000s thought a single explanation was too small for a site visited over so long a time, but they took the contrast seriously.

The challenge to Clark's season and his camp had already been made in print. Anthony Legge and Peter Rowley-Conwy re-examined the large mammals and published Star Carr Revisited in 1988. Anthony Legge, born on 6 June 1939 and died on 4 February 2013, was a zooarchaeologist. After Cambridge and Eric Higgs's group he worked on the animal bones from Tell Abu Hureyra in Syria, one of the few sites where plant and animal domestication can be followed in detail, and he was Professor of Environmental Archaeology at Birkbeck from 2002.

Their reading of Star Carr moved occupation from winter to summer. A fuller statement of that reanalysis says the evidence points to late spring and summer, that the kill of red deer was weighted toward three-year-old animals and the kill of roe deer toward one-year-olds, that there was no bias toward male red deer, that the amount of meat and the scale of occupation should both be reduced, and that the site was a hunting camp. Peter Rowley-Conwy was born in Copenhagen in 1951, the son of Geoffrey Alexander Rowley-Conwy, 9th Baron Langford, and Grete von Freiesleben. He took his doctorate at Cambridge in 1980 under Grahame Clark, on the Mesolithic to Neolithic transition in Denmark.

From 1982 to 1985 he worked on Abu Hureyra under Anthony Legge. He was Professor of Archaeology at Durham from 2007 until he retired in 2020. He has also written on Christian Jürgensen Thomsen's three-age system and its effect in Denmark, Britain and Ireland. A still later account, from the campaign of the 2000s, says the site was not a seasonal camp at all, and that people were active there through the year.

The three readings stand side by side: winter and spring, late spring and summer, and year-round use. Because the occupation lasted centuries, seasonal habits may themselves have changed.

Other readers had already pulled Clark's platform apart. Some argued that the area he dug would have been under water, and that the bone and the objects were waste thrown from dry ground. Others reopened the economic weight of particular species and the question of whether the shore was a place of ordinary rubbish or of deliberate offerings. The new excavations were designed to settle the size of the site, the work done on dry land, the real extent of the timber, and the time-depth of the visits.

Nicky Milner, whose full name is Nicola Jane Milner, was born on 4 September 1973 in Bridlington. She is an archaeologist of the Mesolithic and of the change to the Neolithic. Her doctorate, finished at Cambridge in 1999, was on seasonality read from the shell of the European oyster at Ertebølle sites in Denmark. She moved to the University of York in 2004, became Professor of Archaeology in 2012, and has led the department since 2019.

She was elected a Fellow of the Society of Antiquaries in 2009 and a Fellow of the British Academy in 2019, and she was appointed an Officer of the Order of the British Empire in the 2023 Birthday Honours. She has directed the excavations at Star Carr since 2004, after working on the site for what her account calls more than fifteen years, and she is the principal investigator of the post-glacial project whose main case is Star Carr and the other sites around Lake Flixton. She has also co-directed work at Howick and at Baylet. Chantal Conneller was born in 1973.

Her Cambridge doctorate of 2002 was on space, time and technology in the Early Mesolithic of the Vale of Pickering. She taught at Manchester, became Professor of Early Prehistory at Newcastle in 2018, directed the Prehistoric Society from 2006 to 2009, and was elected a Fellow of the Society of Antiquaries on 10 October 2016. In 2025 she received the British Academy's Grahame Clark Medal, and in 2026 she was elected a Fellow of the British Academy. She co-directed the Star Carr excavations with Nicky Milner and Barry Taylor.

Barry Taylor is at the University of Chester. The campaign is dated 2003 to 2015 in some accounts and 2004 to 2015 in others. Fieldwalking and test pits came in 2004 and 2005, trenches in the following years, the first recognised structure in 2008, and large open-area digging, funded from 2012, in 2013, 2014 and 2015.

They went back for several reasons. Clark thought he had dug the whole site. The work of the 1980s had already shown activity along a much longer stretch of shore and up onto dry ground he had barely touched. The peat was drying.

A field that had been flat in the 1980s was sloping and uneven as the peat shrank. Early trenches found the water table below the main artefact levels. Bone and antler were desiccated, and some had turned to jelly. Acidity in the groundwater had risen sharply.

Seminars on the decay were held at York in November 2008 and at Cambridge in May 2010. A season in 2010, funded to measure the rate of loss, led to a management study in 2012 which concluded that the decay could not be reversed and that what remained should be dug. In November 2011 the site was designated a scheduled monument and called nationally important. The list entry itself is dated 19 December 2011.

Consent for large-scale excavation followed, and the digging of 2013 to 2015 opened about sixty metres of shoreline, including the ground Clark had already cut. Fieldwalking and test pits put Mesolithic activity across about 19,500 square metres. Less than a tenth of that area has been excavated in detail.

The history those trenches support begins quietly. People seem first to have taken an interest in 9385–9260 BC. Early traces include a mass of worked wood and fallen branches, called the brushwood, placed in the shallows in 9340–9190 BC, with objects left around it through the ninety-third and ninety-second centuries. The detrital wood scatter is a jumble of more than 1,300 pieces in the centre of the site, whole trunks and split timbers among them.

One account says it runs about twenty-five metres out from the shore. Another says it extends eighteen metres, laid down more than once to firm the soft sand so that people could reach water at least half a metre deep. Deposition there is placed at the start in 9315–9245 cal BC, and the end of that deposition in 9115–8915 cal BC. People worked wood, butchered animals, and made tools of antler and flint, and they kept using the water as a place to leave things for six to ten generations.

Toward the south of the scatter, in a gap in the wood, they left articulated limbs and parts of the bodies of at least two red deer, an elk skull, a red deer skull, and two red deer frontlets. At least one animal may have gone into the water as a complete carcass. Most of the other bone is limb bone, split for marrow or fat. Utilised flint blades, dowels and the end of a digging stick were among the finds.

On the dry ground above this water, at least one post-built structure was raised. The 2020 account describes an arrangement of posts around a shallow oval pit, about four metres long and three metres wide, with a second partial arc of posts a few metres to the north and a spread of smaller holes that may be drying racks. Flint inside was sparse, including a small cluster of microliths, which may mean a short use and the repair of a composite tool. Around it, people knapped, cut plants, scraped hides, and worked bone and antler.

Fifteen metres further east, woodchips and split roundwood mark a place on the shore where wood was worked through the same years.

After episodes of burning, possibly to clear ground, in 9070–8945 BC, the investment grew. From around 9000 cal BC the water was shallower. Organic mud had been building for centuries, and the edge may have flooded only seasonally. New wetland plants came in, and the reed beds pushed further into the lake.

Three large timber platforms were built in this shallower peat. They were separate constructions. One account says the whole building period lasted about 175 years, and that each platform went out of use before the next was raised. The dates in the model are 8985–8925 cal BC for the central platform, 8945–8760 cal BC for the eastern, and 8805–8755 cal BC for the western, each at 95 percent.

The eastern date is the least secure, because that platform produced the fewest samples. A tighter reading of the same eastern timber gives 8945–8760 cal BC at 95 percent, and probably either 8915–8895 at 9 percent or 8880–8795 at 59 percent.

The size and build of the platforms are not described the same way in every account. One says the central platform is the earliest, largest and most complex, of timbers up to 3.8 metres, in three layers laid in a single episode with little sediment between them, running at least seventeen metres into the reed swamp and sitting on the same line as the old detrital wood scatter. The same account makes the eastern platform a single layer, eleven metres long, of whole trees and timbers over four metres, and the western platform a return to complexity, five layers, at least 14.5 metres from the shore. Another account, by the excavators in 2020, makes the central platform three layers and at least fourteen metres long, built of twenty-six unmodified trees and fifty-seven large split timbers, some of them over five metres, in parallel arrangements and in a single episode.

In that account the eastern and the western platforms are both smaller and both a single layer, aligned more closely with the shore. The purposes proposed are practical and ceremonial at once, and none is settled. The platforms would have given a way through soft ground and out toward deeper water for hunting or fishing. The central one is consistent with a landing stage or a trackway to a place where boats could be hauled.

No boat has been found. A paddle is taken as a hint that boats were used. The eastern and western platforms, lying along the shore, may belong to tasks at the water's edge. Few objects sat among the timbers, so they do not look like ordinary workshops, unless the debris was kept off them.

An exception on the central platform is a scatter of worn flint tools on either side of a flat piece of wood, read as one person's toolkit, once in a bag that split as it decayed.

On the dry ground the same generations built the structures that are now called the earliest post-built buildings known in Britain. Clark found none. The ground he opened was at least sometimes under water, a poor place for a house. The first to be recognised, in 2008, is the eastern structure: a shallow hollow around a ring of posts, just beyond the Early Mesolithic shore and about thirty-five metres from Clark's trenches.

The published descriptions of its size differ. An early account, from the time it was announced, says a structure about 3.5 metres wide, with posts about twenty centimetres across, perhaps as many as eighteen of them, some replaced during its life, and a floor of moss, reeds and other soft plants twenty to thirty centimetres deep. That account gives the building a use of between two hundred and five hundred years, and says that on its discovery it took from Howick in Northumberland the claim to be the oldest known house in Britain. The report of that house is dated to 2010 in the contemporary news of the find, and to 2013 in the biographical account of Nicky Milner's work.

The fuller structural study revises the measurements. The hollow is about two hundred millimetres deep and just over 2.8 metres across, with eighteen possible postholes. If posts ran the whole way round, the plan was roughly trapezoidal, a little over 4.1 metres long and at least about 3.6 metres wide, with a curving wall at the south-east. If posts were missed, it may have been circular or oval.

The post-pipe in one hole is only sixty millimetres across, and none of the regular holes is more than 190 millimetres. The frame was probably of thin poles, perhaps up to about 100 millimetres thick, set upright. A lattice of flexible stems, bent into a dome and bound at the top, is one reconstruction. An inner set of slightly thicker posts, carrying a ring of beams and an outer conical frame, is another.

Coverings could have been reed thatch, reed mats, birch bark, split timbers or hides. The lower fill of the hollow was rich in organic matter, read as a basal layer of plants such as reeds or bark, and also as possibly ferns. Charcoal of birch and of willow or poplar was recovered from it. A later popular account describes the same building as a three-metre-wide hollow inside a four-metre ring of postholes, and says the holes are barely ten centimetres across, slight for a heavy roof.

The central structure is a shallow hollow, 3,320 millimetres from north to south, 180 millimetres deep, and at least 2,650 millimetres wide, with posts along the western side. Its eastern half was cut away by earlier trenches. The 2024 study of the stone tools calls it the earliest of the dryland structures on the radiocarbon evidence, in use between 9300 and 9200 cal BC, and notes a second episode: a date around 9200 cal BC from the upper fill and a date around 8800 cal BC from one posthole. Geochemistry inside the post ring is depleted by comparison with the ground outside, which implies a wall or some other barrier, and either different work inside or an interior that was kept clean.

Phosphorus is high just to the south-west, which could be waste from an entrance or could belong to a small pit just outside. That pit held twenty-five pieces of worked flint and twenty-four fragments of animal bone, most of them burnt, with no sign that the fire had been in the pit itself. The northern structure is an arc of possible posts. If it was once a full circuit, the footprint was about 3.2 by 2.8 metres.

The holes are too shallow to have held tall free-standing posts, so they may have pinned a frame rather than carried its weight. The western structure is a cluster of small round features, some of them credible postholes, near the middle of a dense scatter of worked flint, much of it burnt. The flint runs about two metres out from the features. They may be internal posts of a larger building whose edge is the flint, or racks for drying and smoking, or the remains of a building later used as a midden.

The 2024 study dates its building to 8945–8760 cal BC at 95 percent probability. Taken together, the eastern, central and northern buildings, if the northern arc was once complete, are oval and of similar size, about 3.6 to 4.2 metres long and 2.6 to 3.8 metres wide. At least three structures are counted in the 2024 study, and at least four in the structural report. A 2024 public account of the same research says the buildings may have been cone-like or domed, and that one is thought to have been a cone of felled wood, covered perhaps with reeds or hides.

These are the earliest known Mesolithic buildings in Britain and Ireland. A structure at Deepcar in South Yorkshire may be close in date. It is about 3.4 by 2.1 metres, focused on a hollow, but it has no postholes and is edged with stones. Around 8000 cal BC, larger circular buildings appear at Mount Sandel in Ireland, Howick in Northumberland, East Barns, Echline, and Ronaldsway on the Isle of Man.

Several of those later buildings have posts angled inward and look more conical than the Star Carr frames. The diversity of Mesolithic architecture is the safer lesson. Many other structures, racks, stores, lean-tos and tents, would leave little or no trace, and living trees may have been used as parts of buildings. Because about nine tenths of the occupied ground has not been dug in detail, more buildings may remain.

What people did in and around the structures is now partly visible in the flint. A study published in 2024, by Jessica Bates, Nicky Milner, Chantal Conneller and Aimée Little, examined microscopic wear on 341 flints from at least three of the buildings. It is the first wear study aimed at the Star Carr structures. The tools had been used on wood, bone, antler, plants, hide, meat and fish.

Within one building, some of those tasks were kept apart. Butchery, the messier work, sat apart from crafting. The authors read that separation as a sign that social habits governed where a task was done. The central building held less material than the others, 407 pieces of flint of which 4.2 percent were burnt, and twelve pieces of animal bone, which fits either regular cleaning or a different use.

The eastern building held a great deal of flint, including forty-two refitting sequences, so much of the work happened where the debris fell, though some rubbish was also cleared out. Use-wear there includes plants and hides, and fish bones and two flints show fish being processed. The western building sits in a midden of flint and bone, much of it burnt. Microwear there includes cutting plants and bone, working hides, and drilling stone, and the spatial pattern suggests debris brought in from elsewhere and built up over time.

An earlier, smaller wear study by Aimée Little had already suggested that some tools were taken to the structures for repair and storage, and that plant-working tools clustered at the lake edge.

Clark's famous dump belongs to this busier middle stretch. The part recorded in 2015, in a baulk his trenches had left, and the objects he had already lifted, formed over a short time, perhaps in a single event, around 8915–8775 BC. He had read it as rubbish on a living floor. The new digging found no platform under it.

The ground was at least partly under water, and the material sank quickly. Hafts and handles are missing from the barbed points and the axes, so they were taken off before the objects went into the swamp. Intact points lie with broken middles, tips and tangs, which suggests that some pieces had been kept and then deposited, not snapped and dropped in the moment. The dump is the densest single mass of antler and bone, more than 560 specimens.

Red deer account for 162 bones, roe deer for 49, aurochs for 41 and elk for 12. It has been suggested that this is the remains of a feast for people gathered to a celebration or to a building project. Across the whole dryland area opened by the recent project, about 581 animal bones were recovered, with another 473 fragments from soil samples. In the wetland, most of the bone lay near the eastern platform and was again dominated by red deer, roe deer and elk.

Long bones were broken for marrow. Split bone and antler also reflect craft.

A near-complete but poorly preserved dog was found in very shallow reed swamp. The estimate of its age at death is six or seven months. Nothing could be said of its size or coat. Isotope work found no marine protein in it.

A dog from Seamer Carr, another site in the vale, did have marine protein in its diet, which has been used, still contentiously, to argue that people moved between the vale and the coast. The Star Carr dog does not support a regular coastal trip, or at least not one on which the dog was taken. Dog bones matter here because human bones of this age are very rare in Britain. A dog tooth from Blick Mead, near Stonehenge, has been read as evidence of a journey of as much as 250 miles from the Vale of York.

That journey is not a Star Carr journey. It shows why a dog is used as a proxy. The dating programme treated domestic dog remains as part of the human activity at the site. Clark's wolf bones were later thought, in one project account, to include domestic dog.

The animal lists also include wild boar. A wild boar jaw and several large flint nodules were placed around the timbers of the western platform. A cache of flint was set at the lake edge, possibly in a small pit, and bone and flint were gathered into a pit beside one of the structures. Deliberate deposition and ordinary butchery and craft sat side by side.

The antler frontlets are the objects for which the site is best known. Clark lifted twenty-one stag frontlets that still held worked antler. He listed five shared traits: the beams reduced by the removal of at least three quarters of their circumference, with the spongy core taken out; the tines reduced by about half; the inner side of the burr removed and the outer side kept; and, where the parietal bones survive, perforations. He sorted them into classes by how much of the beam remained, and he thought they were headdresses.

Holes and the smoothed inside of the braincase would have taken straps and made the cap wearable, and the cut-down antlers would have been lighter. He offered two uses, a hunting disguise for approaching deer, and a costume for ritual. The recent excavations added twelve more modified red deer frontlets, seven of them around Clark's area. The total usually given is now thirty-three.

Only eight of Clark's twenty-one are listed with perforations, and the newer finds bring the total of perforated examples to nine out of thirty-three. The new pieces include males, a large female, and males whose antlers had already been shed. One large male had fought in many ruts and was killed in spring, just after he shed. Two others were young, smaller males with antlers still on.

Experiments at the York experimental centre found that a head could be reduced by skinning with a tranchet adze, packing the part to be kept in damp clay, setting the head in the embers, and then breaking away the exposed bone with a hammerstone. That work belongs on dry ground, where there is a hearth and clay, not in the wetland where the finished objects were placed. Groove-and-splinter work, scraping out the spongy tissue, and grinding the rim with a coarse stone are all evidenced. One perforation has a polish that seems to come from use, from something passed through the hole, rather than from the making of the hole.

The frontlets were not all deposited at once. The earliest in the detrital wood scatter fall between the start of that scatter in 9315–9245 cal BC and its end in 9115–8915 cal BC. The heaviest deposition, including Clark's twenty-one and seven of the newer finds, falls between 8885–8775 BC and 8815–8715 BC. One was placed in the eastern platform while that platform was being built.

Another, poorly preserved and found higher up the slope where reed swamp met wood fen, may be among the last, and it is hard to date. An elk skull had already been put in the water in the first half of the ninety-third century, at the very start of the site's use. Putting elk remains into water is also known in the early Mesolithic of southern Scandinavia. About a century later the emphasis shifted to red deer.

There is no change of technique across that time, but away from Clark's area the range of forms narrows. Frontlets are not known from any other site on Lake Flixton. Modified red deer skulls of a similar idea are known from Bedburg-Königshoven, Hohen Viecheln, Plau and Berlin-Biesdorf, and some Star Carr pieces also resemble material from Friesack. A strict continental definition, requiring two perforations and a fuller set of skull bones, would admit only one of the newer Star Carr finds.

The Bedburg pieces keep more of the nasal bones and little worked antler. The argument now is less about a single type and more about a shared sequence of working.

Clark's choice, disguise or rite, has been pulled apart rather than simply repeated. Chantal Conneller has argued that for many hunter-gatherers the useful and the symbolic are not separate, and that a thing which helps a person take on a deer body could serve hunting luck and ceremony at once. A recent study of the ethnography notes that antler hunting disguises are documented in North America, while ritual use of such clothing is more widely documented in Siberia, and on that basis prefers a ritual reading. The same researchers suggest the word mask rather than headdress, because a mask in that literature is anything that changes who the wearer is, and it need not cover the face.

Variation in antlers, in sex, and in whether the antlers had been shed could be a way to call up different deer identities. That remains an interpretation. What is firmer is that the objects were made with care, deposited on purpose, and deposited again and again across centuries of a changing shore.

The barbed points are the other great series. They are red deer antler, and one account gives 195 of them, between eight and thirty-eight centimetres long, more than 95 percent of all such points known from the British Mesolithic. Another account says almost two hundred, and says the collection is still the largest bone and antler assemblage from a Mesolithic context in Britain. Many of the points and many of the frontlets look deliberately broken.

Flint came from beaches that, at the time, were about ten to twenty kilometres away, and also from the Yorkshire Wolds immediately south of the site. Scrapers, axes and microliths are all present. The microliths were the tips of spears and arrows. There is also worked amber, shale, haematite, iron pyrites, tightly rolled birch bark, and birch resin or pitch.

Bracket fungus, the tinder fungus, survives charred in the largest assemblage of its kind known from Mesolithic Britain. The same fungus was carried by the Alpine iceman found in the high mountains, a reminder of its use in fire-making, not a claim that the two finds are one phenomenon. A shale pendant engraved with lines was found in 2015. When it came out of the ground the lines were barely visible.

Imaging showed parallel lines and smaller marks set at right angles. Experimental cutting suggests the fresh lines would have shown white against the darker shale. The motif has been compared with southern Scandinavia, and the piece is called the earliest known Mesolithic art in Britain. It has been on display in York since 2016.

Readings offered for the marks include stars, a map, a bird, the lines of a palm, a river and its channels, a tree, a leaf, the platforms themselves, or a script compared with later ogham. None of those readings is demonstrated. The pendant seems to have been deliberately placed in the lake. An exhibition of the finds has also presented a complete hunting bow from the excavations as the oldest complete hunting bow known, and has presented the timber as the earliest evidence of carpentry in Europe.

The wooden platforms are the evidence usually cited for that carpentry: aspen and willow split along the grain with wedges, probably of wood and antler, and laid in the bog.

Through the busiest generations, peat kept forming and buried the platforms. By around 8700 cal BC the wetlands were becoming drier land. Fen plants and trees spread over the peat. Knapping, tool use and antler-working, previously kept to the dry ground, moved out onto the fen.

By around 8500 cal BC, on the 2020 telling, both zones had gone quiet and the place was abandoned. The longer model allows activity until the middle of the eighty-sixth century and perhaps, if the missing samples matter, a little later. Nothing in the record says why people stopped. The lake itself went on filling for millennia after they left.

The finds are now divided among five collections: the British Museum, the Yorkshire Museum in York, the Museum of Archaeology and Anthropology at Cambridge, the Rotunda Museum in Scarborough, and Whitby Museum. The physical archive of the recent dig is to be kept at the Yorkshire Museum. A temporary reconstruction of one of the houses was built in 2024 in front of that museum, using stone tools and the kinds of plants available at the time, and it stood there from early August until the start of September. The ancient site itself is buried.

The lake has gone. The peat that made the preservation possible is the same peat whose drainage has destroyed much of what was never dug. What the excavations changed is the picture of the early Mesolithic in Britain. The platforms needed the labour of large groups.

The mass of animals in Clark's dump, and the dryland midden, point the same way. People returned to one shore for centuries, thirty-three generations if the long span is read as human lifetimes, and they kept depositing frontlets, barbed points and skulls while the wetland around them turned from open water to fen. That is a history, not a single unchanging seasonal round. It is also, the excavators argue, an unusual history.

The more ordinary sites around the same lake make Star Carr look special to the people who used it, not only to the people who dig it.