In 1985 I presented a  population graph for Britain extending  from the Mesolithic to  recent times, which was characterised  by periodic ups and downs,  the lows being the result of catastrophic  processes in which an overall  loss of the order of 50% in a  century was envisaged. This was  on the level of the historical  population disasters of the 6th and  14th centuries AD.

Just such a  population collapse, I suggested should accompany the contraction and dislocation of settlement in the 12th century BC. What struck me immediately on reading Baillie’s 1988 account of his volcanic “events” was that so many of them matched, more or less, the ups and downs on my 1985 population graph. Here I offer some comments on their possible significance.

In radiocarbon terms the event of  the 4370’s BC corresponds with the  period around 3500 BC when the  first Neolithic colonists arrived  from the Continent; hence the rise  in my curve. Recent fashion has  been to play down the immigrant  contribution to the British  Neolithic, but someone has still to  bring the sheep/goats and seedcorn;  and medically, what would  be the impact of outsiders from the  Continent on an indigenous Mesolithic  population which shows  such scanty signs of having had  any contact with the Continent  since 6500 BC?  The historical analogies  suggest disaster for the indigenes.

How do you date a disaster? Mike Baillie explains

One question which has  received scant attention is why  Britain and Ireland were only settled  from c.3500 BC, when there  were farming groups on the French  and Dutch coasts for some centuries  before this. Baillie’s “event”  perhaps provides the catalyst, the  final straw which launched the  boats – not necessarily in vast  numbers.

The event of the 3190s BC coincides  neatly with the point around  2500 BC at which the Late Neolithic  has for some time been divided  from the earlier Neolithic. No more  causewayed enclosures, long tombs  and the like, and instead henges,  passage tombs, Peterborough and  Grooved wares, and all that goes  with them. Just how a volcanically influenced  climatic downturn  could contribute to such changes  I leave to others.

The answer  presumably lies in knowing more  about Neolithic settlement and  agriculture. The match between  “event” and fall on my graph  should have been closer; my intention  in 1985 was to dip the curve at  a point equivalent to a calibrated  date of c.2500 BC. I erred.

Click here to read more from Colin Burgess on ancient volcanic disasters


The event of the 1620s BC is the  one that got away. I still cannot see  any firm reflection of this in our  Early Bronze Age record. There  may have been a hiccup in the  expansion of settlement in the  marginal lands, especially the uplands,  which had been gathering  strength since cl800/1700 BC. A  temporary setback or retreat would  be very difficult to spot given the  stratigraphic and artefactual exiguities  of our upland sites. But  when an economy is expanding, it  can survive a disaster.

We have already discussed the  event of the 1150s and 1140s; the  next event is that of 207 BC. Somewhere  about this point my graph  begins to show the ever rising  population curve which characterises  the late Iron Age and  Romano-British period. For long,  Iron Age specialists saw the 3rd
century as a time of trouble, characterised  by Marnian invaders and  defence building and refurbishment;  concepts which seem now to  belong to another age. Yet in my  own area, North-East England, there  is good evidence for 3rd century  attention to defence building; and  by the 2nd century settlement had  spread to marginal areas which had  apparently been neglected for more  than a millennium.

Colleagues may  care to scrutinise their own regions  with a 207 BC event in mind.  This leaves the final event of the  540s AD. As Baillie himself has  noted, the 540’s were a time of cold  and plague. In particular this decade  saw the pandemic of Justinian, with  an effect on world population as  catastrophic as the better known  Black Death of the 1340s (see Russell,  in Demography, 1968).  That volcanic “events” could be  linked both to contractions and expansions  of settlement and population  might seem like having one’s  cake and eating it.

The explanation  is perhaps that it depends on the  level of population and settlement at  the time. If there is room for manoeuvre,  as in the 4370’s,  1620’s and  c.207 BC, then settlement can be  shifted and there is opportunity for  expansion both of settlement and  population. But if society is already  at the margins, with nowhere else to  go, as in the 12th century BC, then  the volcanic “event” may well trigger  complete collapse.

Colin Burgess

My thanks to Mike Baillie and Victor  Clube for answering my queries; to  Peter Topping for his comments on  this paper; and especially to Doris
Davies, for coping with my scrawl.  

This feature was part of our ‘Disasters’ special issue,  CA  117


Leave a Reply