Explore every Emerald Archives article connected with Last Glacial Maximum.
31900–31891 BCE
“31,897 BCE” is a modern calendar label, not a year that can be recovered from Irish evidence. No annals, settlements, graves, tools in secure context or dated human remains allow events in Ireland to be fixed to a particular calendar year this far back. In archaeological terms, the label falls very approximately around 33,800 calibrated years before present (cal BP), within the later Pleistocene and Marine Isotope Stage 3. Such conversion is only a broad orientation: radiocarbon results require calibration, yield ranges rather than single historical dates, and the curve itself is refined as research advances. The responsible history of this moment is therefore a history of environmental phases, surviving animal remains and the limits of what those remains can show.
“31,898 BCE” is a modern conversion, not a date anyone living then could have recognised. No written records, settlements, graves, hearths, tool assemblages or works of art securely belong to this particular point in Irish prehistory. Even the broad chronology needs care. Radiocarbon ages are measured in radiocarbon years before present and then calibrated against independently dated records, including tree rings and other archives, to estimate calendar ages. At around 34,000 calendar years ago, the calibration curve is less precise than it is for recent millennia. A specimen may be assigned a calibrated probability range, while the event that deposited or altered it may remain less securely dated still.
31,899 BCE is a modern calendar label, not a year for which Ireland preserves an event, a ruler, a settlement diary or even an annually resolved archaeological sequence. At this depth of time, the responsible language is approximate. Counted from the conventional radiocarbon “present” of AD 1950, the label falls broadly around 33,800–33,900 calendar years before present (cal BP), within the Late Pleistocene and Marine Isotope Stage 3 (MIS 3), conventionally spanning roughly 57,000–29,000 cal BP. It belongs to the Middle Upper Palaeolithic world of north-west Europe, but there is no basis for assigning a particular Irish human action to this single numbered year.
“31,900 BCE” is a useful modern label, not a year that can be recovered from Ireland’s prehistoric record. No Irish calendar, written account, ruler, settlement sequence or human life can be assigned to that precise point. In radiocarbon and ice-sheet research, dates are normally expressed as ranges of calibrated years before present (cal BP, where “present” is AD 1950), often with sizeable uncertainty. In that convention, 31,900 BCE is roughly 33,850 years before present. It falls within Marine Isotope Stage 3 (MIS 3), a long and climatically unstable part of the last glacial cycle, and just before the great expansion of the British–Irish Ice Sheet that culminated several millennia later.
31910–31901 BCE
Ireland was then part of the wider north-west European world of the last glacial cycle. It was not the uniformly ice-smothered island sometimes imagined from later phases of the Ice Age. The great British-Irish Ice Sheet would eventually expand dramatically, reaching its maximum extent on the Atlantic shelf roughly 26,000–24,000 cal BP and then retreating in stages. Around the equivalent of 31,901 BCE, however, evidence from beneath later glacial sediments indicates that at least some Irish regions had remained ice-free during earlier parts of the cycle. This does not mean that conditions were mild. It means that cold, seasonal and periglacial environments could exist beyond, between or before major ice advances.
Radiocarbon laboratories report ages in years before present, where “present” conventionally means 1950, not the present day. Those measurements must then be calibrated against independently dated records of past atmospheric carbon, using curves such as IntCal20. At great antiquity, a radiocarbon measurement can correspond to a wide calendar range rather than a single year. “31,902 BCE” is thus best read as a notional place within the later Middle Midlandian, the Irish term commonly used for this part of the Last Glacial Period, rather than as an anniversary that archaeology can observe.
The closest important environmental archive is at Derryvree near Maguiresbridge in County Fermanagh. There, freshwater silts and organic layers were preserved between two glacial tills: deposits left by ice advances before and after the quieter interval in which the sediments accumulated. A radiocarbon determination from the organic series, once calibrated, has been published as approximately 36.9–31.7 thousand cal BP. That is a broad range, but it encompasses the notional date of 31,904 BCE. Derryvree consequently provides a more responsible guide to conditions around this point than an attempt to assign a precise seasonal scene to the whole island.
Ireland lay at the north-western edge of the European landmass and climatic system. Sea levels were substantially lower than now because large quantities of water were locked in global ice sheets. The present Irish Sea and Celtic Sea were altered landscapes, with exposed shelf in places and changing routes between Ireland, Britain and continental Europe. Yet lowered sea level did not automatically mean an easy or permanent dry crossing. The configuration of coasts, rivers, estuaries, sea ice and local ice margins changed through time, while Ireland’s relation to Britain was shaped by a complicated, mobile glacial environment.
The most useful evidence near this date comes from Derryvree, close to Maguiresbridge in County Fermanagh. There, freshwater organic silts survived between two layers of glacial till. This is an exceptional piece of geological good fortune: later ice both erased much evidence elsewhere and sealed this small archive below its debris. A radiocarbon determination of 30,500 ± 1,200 radiocarbon years BP from the Derryvree series calibrates broadly to about 36,900–31,700 cal BP. That range encompasses the chronological neighbourhood of 31,906 BCE, though it cannot prove that every fossil in the deposit lived in that nominal year.
“Before present” conventionally means before AD 1950, whereas BCE dates are a later conversion. Around 34,000 years ago, even well-preserved organic material does not yield calendar precision of the kind implied by a numbered year. Atmospheric radiocarbon varied through time, and the calibration curve becomes especially consequential beyond 30,000 calendar years ago. A bone’s laboratory result, its calibrated range, the layer from which it came and the possibility of later disturbance all need separate consideration. Consequently, this article uses “around”, “about” and calibrated ranges where the evidence permits; it does not claim that any animal, group or occurrence can be assigned to 31,907 BCE itself.
Derryvree offers an unusually valuable glimpse because its fine sands and organic silts were sealed between two tills, deposits left by separate incursions of glacier ice. During the period represented by the intervening freshwater sediments, south Fermanagh was not buried beneath an ice sheet. It had shallow water, wet ground and plant-rich hollows in a cold, open landscape. Pollen, moss remains and insects indicate tundra or muskeg rather than woodland: grasses, sedges, herbs and juniper were present, while dry ground was probably patchy and sparsely vegetated. Pools and saturated mossy ground could persist beside exposed mineral soils, shaped by freezing, thawing and seasonal meltwater.
At this point in the last glacial cycle, Ireland was moving towards the much more extensive ice cover that developed later. The most recent British–Irish Ice Sheet reconstructions model substantial growth after about 31,000 years ago and an Ireland almost wholly ice-covered by around 28,000 years ago. That does not allow a simple picture for c. 33,900 calibrated years before present. Conditions varied between regions, seasons and short climatic phases, while the exact timing of ice advance remains an active research question. The island was not yet the uniformly ice-dominated terrain of the later glacial maximum, but neither was it a hospitable temperate landscape.
Ireland stood within the last Ice Age, but it should not be imagined as a simple, permanently frozen island under one unchanging sheet of ice. MIS 3 was climatically unstable, and the growth, extent and behaviour of the British–Irish Ice Sheet varied regionally and through time. Research on southern Ireland indicates that the main expansion of the last Irish ice sheet began from initially ice-free conditions at about 34,000 years ago and reached the south coast only after about 24,000 years ago. Around this notional date, then, parts of Ireland may have remained available to cold-adapted animals even as glacial conditions intensified elsewhere. That is an interpretation from regional geological evidence, not a precise map of conditions on any one day or in every valley.
31920–31911 BCE
Prehistoric chronology needs careful translation. “BP” means before AD 1950, the reference year used by radiocarbon science; a calibrated age converts a radiocarbon measurement into a probability range in calendar years. Neither system licences the apparent precision of a named BCE year. The figure 31,911 BCE is therefore useful as a position on a long timeline, but it should not be mistaken for a securely dated Irish episode.
The numerical label 31,913 BCE is a modern conversion, not an exact prehistoric calendar date. For periods this old, scholars normally use calibrated years before present (cal BP, with “present” conventionally fixed at 1950) or broad geological stages. The calibration curve itself becomes less exact further back in time, while the survival of enough original collagen for radiocarbon dating is a separate challenge. Modern re-dating of Irish cave bones has shown why such caution matters: improved pretreatment has made several Castlepook dates thousands of radiocarbon years older than earlier results.
Modern BCE labels give a misleading impression of precision. Archaeologists date material from this period chiefly through radiocarbon measurement, then convert radiocarbon ages into calibrated ranges using internationally maintained calibration curves. Those ranges may span centuries or millennia; they do not identify an individual season, much less an exact year. Moreover, the relevant evidence is not a single layer sealed in a secure camp site. Castlepook’s cave deposits accumulated over a long time and were disturbed by water, animal activity and early excavation.
At this date Ireland was not the wooded, temperate island familiar today. It belonged to a northern Atlantic world repeatedly reshaped by cold, changing sea levels, snowfields, frozen ground and advancing ice. The great expansion of the British–Irish Ice Sheet lay ahead: the best large-scale reconstruction places its maximum volume at about 23,000 years ago, while Irish evidence shows that ice cover and its limits varied substantially from region to region. Consequently, 31,916 BCE should not be pictured as a moment when all Ireland lay sealed under one uniform ice sheet. It was instead part of a long pre-maximum cold phase in which habitats could alter dramatically over centuries and millennia.
For this period, precision is necessarily limited. Radiocarbon laboratories measure the remaining radioactive carbon in once-living material, such as bone collagen. The result is expressed in radiocarbon years before present, not in BCE. Atmospheric carbon levels changed over millennia, so those results must be compared with calibration curves assembled from independently dated records, including tree rings, corals, cave deposits and lake sediments. At nearly 34,000 years ago, calibrated ranges can be broad, and revisions to methods or samples can alter previous conclusions.
Modern BCE notation is far more precise than the evidence permits. Radiocarbon determinations measure the remaining carbon-14 in once-living material, then must be calibrated against changing atmospheric carbon levels; their results are normally reported as probability ranges, not as a single historical date. Moreover, the relation between a date expressed in “years before present” and a BCE year is not a simple conversion. “Present” in radiocarbon convention is 1950, and calibration can widen or shift the apparent age substantially. For this reason, 31,918 BCE should be read as approximately 34,000 years ago, not as the anniversary of a known occurrence. The relevant archaeological and environmental record spans thousands of years.
The numbered year sits within a long and unstable climatic interval. The Last Glacial Maximum, when the British–Irish Ice Sheet became especially extensive, lay later, broadly between about 26,500 and 19,000 years ago. Around the earlier horizon represented by 31,919 BCE, the eventual expansion of that ice sheet was not yet complete. Geological evidence from north-central Ireland is particularly important. Organic sediments beneath later glacial deposits at Derryvree and Greenagho in County Fermanagh indicate ice-free conditions during parts of the preceding interval. One calibrated range from Derryvree spans approximately 36,900–31,700 cal BP; another from Greenagho spans about 37,200–35,700 cal BP.
31930–31921 BCE
Ireland was entering a decisive climatic transformation. The Last Glacial Maximum, when the British–Irish Ice Sheet covered much of the island, lay ahead rather than behind: broadly, major ice expansion belongs to the period after about 34,000 years ago, with full glacial conditions developing later. Geological evidence from beneath tills in Fermanagh and from the west coast indicates that substantial areas of Ireland had been ice-free during the preceding millennia. That does not mean a mild or stable country. Temperatures fluctuated sharply, seasons were severe, and the growing ice sheet was beginning to reshape the wider region.
Ireland was then part of the long last glacial cycle, but it should not be pictured simply as an island sealed beneath permanent ice. Marine Isotope Stage 3, broadly c. 57,000–29,000 years ago, was climatically restless. Greenland ice-core records and North Atlantic sediments reveal abrupt warmer and colder swings. The British–Irish Ice Sheet was already an important presence, but its precise extent and behaviour around 34,000 calibrated years ago remain debated. Recent synthesis points to earlier growth, retreat and renewed advance before the Last Glacial Maximum, rather than one uncomplicated march towards a frozen peak.
Archaeologists use radiocarbon determinations, stratigraphy, faunal study and geological dating rather than historical chronologies for this period. Radiocarbon years are not identical to calendar years and must be calibrated against changing atmospheric carbon levels; even then, dates are expressed as ranges with statistical uncertainty. The numerical year 31,923 BCE therefore should not be mistaken for an anniversary of a hunt, a migration or a climatic event. It is best understood as an approximate point within a long Upper Palaeolithic world, perhaps roughly 34,000 years before the conventional radiocarbon “present” of 1950.
The evidence is exceptionally uneven. Ireland has no securely identified house, hearth, grave, tool assemblage or settlement surface of this age. Its record comes principally from animal bones, cave deposits, glacial sediments, landforms and changing shorelines. These are powerful sources for reconstructing environments, but they do not provide an annal of a particular hunt, migration or season. Much of the island’s earlier surface archaeology was also vulnerable to erosion, burial and destruction as the last ice sheet expanded. Absence of a camp is not conclusive proof that no person ever entered Ireland; equally, it cannot be converted into proof of a settled population.
Ireland was not yet simply the ice-covered island often imagined from accounts of the Last Glacial Maximum. Research on the British–Irish Ice Sheet indicates that major ice expansion followed this period and reached maximum extent around 27,000 years BP in some reconstructions. Yet conditions were severe, and the record points to rapid environmental change before, during and after the maximum. Ice caps and glaciers occupied uplands and western and northern areas at different times; periglacial processes repeatedly froze, fractured and shifted exposed ground. Any attempt to draw a neat national boundary between “ice-free” and “glaciated” Ireland would be misleading.
There were no states, rulers, written laws or diplomatic relations in Ireland that can be recovered for this time. Nor can archaeologists identify a village, burial ground, hearth or tool-making floor securely belonging to 31,926 BCE. The island’s history is instead approached through geological traces, cave deposits, animal bones and laboratory dating. These sources illuminate changing environments and occasional animal presence far more clearly than they illuminate human lives.
The most important Irish evidence is the Derryvree freshwater sequence near Maguiresbridge in County Fermanagh. Roadworks in 1969 exposed silts and fine sands containing organic material between two distinct sheets of glacial till. The deposit produced a conventional radiocarbon age of 30,500 years before present, with a substantial uncertainty. When recalibrated, the published span is approximately 36,900–31,700 calendar years before present. The nominal date of 31,927 BCE lies within that broad range, but this does not mean that the sediment was laid down in that exact year. It accumulated over time, and the calibrated range expresses the statistical limits and complications of deep radiocarbon dating.
The familiar green, wooded Ireland of later millennia did not exist. Cold-stage conditions supported largely open vegetation: grasses, sedges, herbs and low shrubs rather than dense forest. The terrain would have varied sharply with altitude, exposure, soils and proximity to ice. Frost fractured rock, seasonal thaw moved sediment downslope, and sparse plant cover left ground vulnerable to wind and water erosion. Rivers behaved differently from their modern descendants, carrying greater loads of glacial or periglacial sediment. In limestone districts, caves and fissures offered sheltered traps in which bones, sediments and occasional traces of activity could accumulate.
The Ireland of this broad period was not yet the island familiar today. Its coastlines, sea levels, river valleys and connections with the surrounding continental shelf were all affected by the growth and decay of ice. Yet it would be misleading to picture a single, fixed scene. Geological evidence shows that ice margins and relative sea level changed repeatedly through the later Pleistocene. Research on western Ireland records high relative sea levels between about 40,000 and 19,000 cal BP, interpreted as evidence of substantial loading of the land by ice, while other deposits indicate rapid fluctuations of the ice margin. The point conventionally labelled 31,929 BCE falls within this changing world, not at a securely charted boundary between “ice-free” and “ice-covered” Ireland.
Ireland was not then the green, wooded island familiar today. Climate shifted sharply during the Late Pleistocene, and local conditions altered as ice expanded and contracted across the wider British-Irish region. The great maximum of the last British-Irish Ice Sheet came later, broadly during the period c. 26,000–19,000 calibrated years before present. At the notional date used here, extensive ice was approaching rather than simply forming a fixed, island-wide backdrop. The exact pattern of glacier cover, exposed land and habitable ground remains debated because the later ice sheet eroded, transported and buried much of the earlier evidence.
