Membership is completely free, giving members unlimited access to all available resources, services, features, and benefits without any restrictions or fees.
Welcome back. Sign in to continue enjoying unlimited access to all available resources, services, features, and member benefits at no cost.
Radiocarbon determinations are not calendar dates in themselves. They must be calibrated against changing atmospheric carbon levels, and at more than 30,000 years ago the resulting ranges can be broad or irregular. The current IntCal20 calibration framework extends to 55,000 calendar years before present, but it does not turn a single bone into an exact annual timestamp. The Castlepook femur should therefore be understood as belonging to a period around the mid-30th millennium before present, rather than as proof that butchery occurred in this particular numbered BCE year.
There were no Irish kingdoms, chiefs, borders, treaties or named peoples that can responsibly be assigned to this date. There is no evidence for a permanent population, village, house, burial ground, shrine, tool-making floor, hearth or cultivated field in Ireland in c. 31,838 BCE. Nor can archaeology identify an Irish language, religion, artistic tradition or system of learning at this time. Such absences are historically important. They mean that political affairs, diplomacy, formal conflict, trade and institutions are not merely unknown in detail: they are unsupported categories for this particular Irish horizon.
This was the later part of Marine Isotope Stage 3, a long and climatically unstable interval of the Late Pleistocene. Ireland was approaching, but had not yet reached, the fullest expansion of the last British–Irish Ice Sheet. Modern reconstructions show that ice growth and retreat were uneven between regions rather than one simple island-wide advance. The principal maximum came later, broadly between about 27,000 and 23,000 calibrated years before present, although individual sectors behaved differently. Around the notional year in question, ground in some areas was still available to plants and animals, while glaciers and ice masses were expanding elsewhere.
Derryvree’s pollen, plant macrofossils, mosses and insects provide a rare view of an Irish landscape before the great expansion of the last British-Irish Ice Sheet. They indicate open tundra or muskeg: wet, treeless ground with pools, sedges, grasses, herbs, mossy margins and low shrubs, rather than the oak, hazel and ash woods familiar from much later Irish prehistory. The evidence points to severe winters, short growing seasons and periglacial conditions in which frost was a major force shaping soil and watercourses.
Ireland was not a timeless, uniformly frozen wasteland. MIS 3 was marked by sharp climatic fluctuations, and the growth of the British–Irish Ice Sheet was neither simple nor simultaneous in every region. Recent reconstructions emphasise that ice-sheet expansion and retreat differed around its perimeter. Modelling of conditions around 31,000 years BP has even produced small ice caps in north-west Ireland while not requiring an island-wide marine-based ice sheet at that moment. Other work argues that significant ice may have developed earlier in parts of the wider British–Irish system. The exact extent of ice over Ireland around 33,800 calendar years ago consequently remains debated.
Radiocarbon years are not calendar years. The quantity of carbon-14 in the atmosphere has varied through time, so a laboratory measurement must be calibrated against an internationally maintained curve. At about 34,000 years ago, the conversion may produce broad intervals rather than a precise answer, and the dated material itself may represent only one moment within a longer-lived deposit. A nominal date such as 31,878 BCE must therefore not be treated as the anniversary of a hunt, a migration or a climatic episode.
Archaeologists employ several clocks for this period. Radiocarbon results are conventionally reported in years before present, where “present” means AD 1950; because atmospheric carbon has varied through time, those results must be calibrated against independently dated records. Even calibrated dates are ranges of probability, not single moments. Deposits may also have been disturbed by water, frost, scavengers, excavation and later collecting. A bone can date the death of an animal, but it does not necessarily date the formation of a whole cave layer or the arrival of every other object beside it.
“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.
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.
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.
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.
The numbered year should not suggest false exactness. In archaeological convention, “before present” is measured from 1950, while BCE is a retrospective calendar system. A conversion from 31,932 BCE gives only an approximate position of about 33,900 years before present; it does not identify a single excavated deposit or a particular season. Radiocarbon ages are not calendar ages, and their calibration depends on a curve built from several natural archives, including tree rings, corals, lake sediments and cave deposits. At this antiquity, calibrated results can be broad and sometimes discontinuous.
31,937 BCE cannot be treated as a year in the ordinary historical sense. No Irish written record, named ruler, settlement sequence or securely dated event can be assigned to a single calendar year so remote in time. Indeed, even the conversion between BCE dates and scientific chronologies needs care: archaeologists normally work with radiocarbon years before present (BP, where “present” is 1950) and calibrated calendar ranges, each with margins of uncertainty. The appropriate setting is therefore not a supposed episode in 31,937 BCE, but Ireland in the later part of the Last Glacial Period, approximately 34,000 calendar years ago.
Ireland at this point was not the green, wooded island familiar today. It lay in a cold-world north-western European setting, as climates swung repeatedly between relatively milder interstadials and severe cold phases. The great Last Glacial Maximum had not yet reached its fullest development: the major British–Irish Ice Sheet would become much more extensive later, with its greatest volume reconstructed at about 23,000 years ago. Yet ice growth was already becoming an important force in the region. Its advancing margins, changing sea levels and shifting drainage patterns would eventually erase or bury much of the evidence of earlier landscapes and any traces left by people.
Ireland then stood on the threshold of the last great expansion of the British–Irish Ice Sheet. It was not yet simply the ice-covered island of popular imagination. Geological evidence from beneath later glacial deposits at Derryvree and Greenagho in County Fermanagh indicates ice-free conditions in north-central Ireland until at least about 34,000 years ago. The calibrated ranges are broad: organic material at Derryvree falls within c. 36,900–31,700 cal BP, while material at Greenagho falls within c. 37,200–35,700 cal BP. They establish an environmental sequence, not the weather or shoreline of a particular year.
The date falls in Marine Isotope Stage 3, a cold and climatically unstable part of the last glacial cycle, before the fullest development of the Last Glacial Maximum. The British–Irish Ice Sheet was not a fixed white blanket advancing everywhere at one speed. Ice expanded and contracted in distinct sectors, responding to climate, snowfall, sea level, topography and the behaviour of fast-flowing ice streams. Large reconstructions of the ice sheet identify the period from about 31,000 to 15,000 years before present as one of growth, maximum extent and retreat, with maximum ice volume and many outer margins reached later, broadly during the mid-20,000s years before present.
Ireland lay on Europe’s Atlantic north-western fringe during a climatically unstable phase of the last Ice Age. It was not yet the fully ice-covered landmass of the Last Glacial Maximum, generally placed at about 26,000–19,000 cal BP, but the conditions that would produce that maximum were gathering strength. Research on the British–Irish Ice Sheet indicates that ice expansion was already under way in parts of the wider archipelago by about 31,000 years ago. Its precise limits earlier in the sequence are uncertain, and Ireland’s conditions were not uniform. Uplands, northern districts, Atlantic margins and lower southern landscapes could have had sharply different local histories.
It is important not to turn the number 31,950 BCE into false precision. Radiocarbon laboratories report ages in radiocarbon years before present, with statistical uncertainty; calibration against independently dated records translates these into calendar ranges that can be broad, especially this far back. Geological evidence is often less precise still, dating an advance, retreat, sediment layer or exposed rock surface within centuries or millennia. Consequently, “Ireland in 31,950 BCE” means Ireland in the vicinity of a notional point in time, not a documented year.
This was not yet the peak of the last glaciation. The Last Glacial Maximum is conventionally placed later, broadly about 26,500–19,000 years ago, when the British–Irish Ice Sheet attained its greatest overall extent. Yet the cold world which would produce that maximum was already taking shape. Reconstructions of the British–Irish Ice Sheet indicate substantial expansion after roughly 31,000 years ago, with different sectors advancing at different times. The exact coverage of Ireland at any moment around 31,953 BCE remains uncertain: an ice sheet is not a static white blanket, but a changing system of snow accumulation, flowing ice, meltwater and shifting margins.
About 34,000 calendar years before present, the land now called Ireland stood in a colder, unstable North Atlantic environment. This was not yet the peak of the last great glaciation. Reconstructions differ in local detail, but they agree that the British–Irish Ice Sheet expanded dramatically afterwards, reaching or approaching its greatest extent between roughly 27,000 and 23,000 calibrated years ago. By the height of that episode, much or perhaps nearly all of Ireland was under ice. In 31,978 BCE, therefore, the central historical fact is not an Irish kingdom, migration or monument, but a changing physical landscape on the threshold of severe glaciation.
“31,983 BCE” is a useful modern label, not a date that people or landscapes in prehistoric Ireland could have recognised. At so great a distance, archaeology cannot identify a particular season, event or community in one named calendar year. The label corresponds only approximately to about 34,000 years before AD 1950, the conventional reference point used in radiocarbon work; it must not be treated as a precisely calibrated archaeological timestamp. The responsible scale is therefore a broad one: later Marine Isotope Stage 3, within the Last Glacial Period, and several millennia before the climatic nadir conventionally called the Last Glacial Maximum. The available evidence is uneven, mostly environmental or animal-bone evidence, and comes chiefly from caves in southern Ireland rather than from settlements.
This point falls within Marine Isotope Stage 3, during the wider Late Pleistocene cold stage conventionally called the Midlandian in Ireland and the Late Devensian in Britain. It preceded the fullest expansion of the last British–Irish Ice Sheet, but it belongs to the long climatic deterioration from which that ice sheet grew. Recent reconstructions show that the build-up, maximum extent and retreat of ice across Ireland and Britain were highly asynchronous: different sectors advanced and retreated at different times. The greatest ice volume in the most recent reconstruction occurred around 23,000 years ago, while some sectors had advanced earlier. That pattern is a warning against imagining a single, uniform moment when all of Ireland suddenly vanished beneath ice.
Calendar-year precision disappears rapidly in the Upper Palaeolithic. Radiocarbon measurements are reported as years “BP” — before AD 1950 — and must be calibrated against changing atmospheric carbon levels to estimate calendar ages. Near 34,000 years ago, dates may carry uncertainties of centuries or more; cave deposits can also contain bones accumulated at different times and later disturbed by water, animals or excavation. A specimen dated to a broad interval around this period is evidence for that interval, not for an event occurring on 1 January or any other day in 31,993 BCE.
Around this broad date Ireland lay at Europe’s Atlantic north-western edge. Sea level was much lower than today because immense quantities of water were locked in continental ice sheets. The present coastline, Irish Sea and continental shelf cannot simply be projected backwards: exposed plains, changing channels and ice barriers altered routes between Ireland, Britain and continental Europe. Whether a particular crossing was practicable depended not only on distance, but on season, weather, sea ice, rivers and the advance of glaciers. Later reconstructions show Ireland becoming fully engulfed at the Last Glacial Maximum, conventionally placed around 27,000–23,000 calibrated years BP; that maximum lay in the future of the date considered here.
