The Edge of Memory

by Patrick Nunn

  18 My most recent research into Fiji’s hillforts, conducted in collaboration with the Fiji Museum, has been in the Bua area of Vanua Levu Island. A popular account of this project was published online in SAPIENS on 15 June 2016 and is downloadable from http://tinyurl.com/z9st6eb.

  19 Quotation comes from p. 55 of Henry Britton’s Fiji in 1870, a collection of articles he wrote for the Argus newspaper in Melbourne (1870).

  20 The most comprehensive and accurate assessment of current and recent rates of sea-level change is found in the 5th Assessment Report of the Intergovernmental Panel on Climate Change (IPCC). The science was assessed by Working Group 1. The Summary for Policymakers is an accessible summary for non-specialists – see ipcc.ch.

  21 Closure of the ‘sea-level budget’ for 1993–2010 was a major achievement of the ‘Sea Level Change’ chapter in the 5th assessment report of the IPCC. Until this point, there was uncertainty about exactly which cause/s of sea-level rise were most important. Settling the issue has implications not only for the understanding of past sea-level changes and the modelling of future ones, but also for strategies to mitigate the effects of future climate change.

  22 The original area of the contiguous land mass of Sahul (comprising modern Australia, Papua New Guinea and Irian Jaya, or Indonesian Papua) during the Last Glacial Maximum about 20,000 years ago (when the sea level was about 120m/400ft lower than it is today) was 11,021,024km2 (4,255,241mi2). The modern land area totals 8,473,836km2 (3,271,766mi2), so 2,547,188km2 (983,475mi2), or 23 per cent, was inundated when the sea level rose after the end of the last ice age.

  23 This authoritative compilation of past sea level around Australia was developed by Stephen Lewis and others and published in 2012 in the journal Quaternary Science Reviews.

  24 Delicate bivalves (like Donax and Paphies spp.) have been used in Australian sediment cores to identify former shorelines. Where these shells are found unbroken rather than pulverised, this indicates that they had hardly been moved from the coast they formed.

  25 Six sediment cores collected from the sea floor between the Arafura Sea and Gulf of Carpentaria show that the sill separating the two was covered by seawater most recently about 9,700 years ago (Chivas, 2001, Quaternary International).

  26 David Hopley’s pace-setting 1982 book on the Geomorphology of the Great Barrier Reef made the case for coral-reef coring as a precise way of reconstructing past sea levels with an almost unmatched degree of precision. This book has more recently appeared in an updated version (Hopley, Smithers and Parnell 2007).

  27 See Dixon (1980: 46).

  Chapter 5 : Other Oral Archives of Ancient Coastal Drowning

  1 I am grateful to Marjorie Le Berre for information in these two paragraphs, which is based on conversations we had in Nantes in November 2016 and is used with her permission.

  2 The city of Ys is sometimes placed not in the Baie de Douarnenez, but in the Baie des Trépassés, or in the Baie d’Audierné (Guyot 1979), all along the west-facing coast of Brittany.

  3 English translation on p. xiii of the book by Guyot (1979), from Bertrand d’Argentre’s (1588) L’Histoire de Bretaigne (The History of Brittany).

  4 ‘La ville d’Is fut submergé et presque tous ses habitants périrent’: French text from p. 22 of Sébillot (1899).

  5 ‘Les pêcheurs de Cancale dissent que quand la mer est belle et claire, on voit entre le Mont Saint-Michel et les îles Chausey de debris de murailles. Ce sont les restes d’une ville disparue’: French text from p. 23 of Sébillot (1899).

  6 Quotes from p. 336 of Peacock (1865, Proceedings of the Royal Geographical Society of London), who also concluded that the monastery at Menden is now under 8m (26ft) of ocean, that at Mandan (west of St Pair) now under 10m (33ft), that named after St Moack now under 13–14m (43–46ft), and Taurac (or Caurac) now submerged to a depth of 16m (52ft).

  7 St Guénolé (or Guignolé) is often remembered in English-language texts as St Winwaloe; details of his life were summarised by Doble (1962).

  8 Nowhere is this perhaps more obvious than on the delta of the Brahmaputra-Ganges-Meghna river, occupied mostly by several tens of millions of Bangladeshis, which has over recent decades experienced storm surges that have reached progressively further inland as a result of these being imposed on sea-level rise. As the sea level rises in the next few decades, so storm surges will reach even further inland.

  9 ‘Au temps jadis la Manche n’était pas si grande que maintenant; l’on pouvait aller à Jersey sans rencontrer d’autre obstacle qu’un ruisseau qui n’était pas trés large’ (Sébillot 1899: 23). In addition, a contemporary story about ‘the conjunction of Jersey to Normandy’ was alluded to by Poingdestre (1889: 75).

  10 From Gazette de l’Ile de Jersey, 28 April 1787, quoted by Peacock (op cit, p. 329), who identifies numerous other instances of submerged forests in the Channel Islands.

  11 Reading Alan Seeger’s eponymous 1916 poem, written at a time when the ancient tradition was probably better known than it is today, Lyonesse sounds as though it might have been Ys – as indeed it may:

  In Lyonesse was beauty enough, men say:

  Long Summer loaded the orchards to excess,

  And fertile lowlands lengthening far away,

  In Lyonesse.

  Came a term to that land’s old favouredness:

  Past the sea-walls, crumbled in thundering spray,

  Rolled the green waves, ravening, merciless.

  Through bearded boughs immobile in cool decay,

  Where sea-bloom covers corroding palaces,

  The mermaid glides with a curious glance to-day,

  In Lyonesse.

  12 From an 1854 translation of The Chronicle of Florence of Worcester, quoted in Hunt (op cit) at www.sacred-texts.com/neu/eng/prwe/prwe085.htm, accessed in April 2009.

  13 Quoted by Hunt (op cit). Other authoritative sources that name Lyonesse and place it off the coast of Land’s End include Britannia (Camden 1590) and the 1602 Survey of Cornwall (Carew 1723). Camden was an antiquarian who collected folk tales. He concluded that Lyonesse (the City of Lions) was located off Land’s End where the Seven Stones reef now lies. Camden also reported that sailors could hear the bells of Lyonesse ringing when they crossed the area during heavy seas.

  14 There are examples from the South-west Pacific island groups of the Solomon Islands and Vanuatu in which islands have reportedly vanished as a result of large-wave impact but which must also have been affected, either during the earthquake that produced these (tsunami) waves or a short time after it, by rapid subsidence. It was the latter that caused the islands to disappear. Examples include Teonimanu and the Pororourouhu group in the Solomon Islands, and islands like Malveveng and Tolamp in Vanuatu. These and other examples are in my book Vanished Islands and Hidden Continents of the Pacific (Nunn 2009), available as an audio book at https://tinyurl.com/lbrrgfa.

  15 A contemporary account notes that to the astonishment of onlookers on the south coast of Cornwall, ‘the Sea rose near six feet [1.8m], coming in from the South-East extremely rapid; then it ebbed away with the same rapidity to the Westward for about ten minutes; till it was near six feet lower than before; it then returned again, and fell again … the first and second fluxes and refluxes were not so violent as the third and fourth, at which time the Sea was as rapid as that of a mill-stream descending to an under-shot wheel, and the rebounds of the Sea continued in their full fury for fully two hours’ (Borlase 1758: 53–54). This happened at a distance of around 1,500km (932 miles) from the epicentre. The Reverend Borlase described a more locally centred earthquake on 15 July 1757, during which ‘on the island of St. Mary, Scilly, the shock was violent’.

  16 The Roman count is found in Strabo’s Geography (ad 17?), at a time when the Scilly Isles were called the Cassiterides.

  17 Causeways and house foundations have also been found underwater within the Scilly archipelago (Ashbee 1974).

  18 The story is immortalised in the Welsh-language poem Clychau Cantre�
��r Gwaelod (The Bells of Cantre’r Gwaelod) by J. J. Williams. The extract quoted here was translated by Dyfed Lloyd Evans and accessed at www.celtnet.org.uk in October 2007.

  O dan y môr a’i donnau

  Mae llawer dinas dlos

  Fu’n gwrando ar y clychau

  Yn canu gyda’r mos.

  Trwy ofer esgeulustod

  Y gwiliwr ar y tŵr

  Aeth clychau Cantre’r Gwaelod

  O’r golwg dan y dŵr.

  Beneath the wave-swept ocean

  Are many pretty towns

  That hearkened to the bell-rings

  Set pealing through the night

  Through negligent abandon

  By a watcher on the wall

  The bells of Cantre’r Gwaelod

  Submerged beneath the waves.

  19 A much larger area of more than 6,200km2 (2,400mi2) is envisaged in the analysis of Flemming (1972), who was consequently sceptical that such a land ever existed. The name Cantre’r Gwaelod means ‘the bottom cantref’, a cantref being an area of perhaps 466km2 (about 180mi2 – Flemming 1972).

  20 All quotes paraphrased from accounts quoted by Doan (1981, Folklore).

  21 The quotes in this sentence and the next come from p. 81 of Doan (op cit), who also quotes authoritative sources for the age determination.

  22 ‘It is to representatives drawn from among the famous legendary heroes of the sixth century, the period assigned to the beginning of their national traditions, that medieval cyfarwyddiaid [storytellers] attached the legends of the great inundations’ (from p. 241 in the chapter by Rachel Bromwich in the collection by Fox and Dickens (1950)).

  23 Quoted by Bromwich (op cit, p. 229) from a 1917 account by Richard Fenton.

  24 This information is found on p. 334 of Wilson’s (1870) The Imperial Gazetteer of England and Wales. The idea that human-made structures are visible at Caer Arianrhod was roundly criticised by North (1957) and others, yet they do not address the more subtle point about whether or not the memory of such a place may be authentic … even if its currently favoured physical manifestation may not.

  25 A submerged forest in the Dovey Estuary was dated to 4,700–6,000 years ago when the sea level here may have been 2–4m (6½–13ft) lower.

  26 Another estimate by Kurt Lambeck suggests that there was a wide land bridge between Ireland and Wales from 13,000 to 20,000 years ago, but it is not explained when it may have become impassable. Yet it is noted that about 14,000 years ago, because of the huge volumes of meltwater pouring into the Irish Sea, the shrinking land bridge would have been ‘a swampy and inhospitable region at best’ (1995, Journal of the Geological Society of London, p. 443).

  27 The quotes come from p. 30 of a recent translation of the Mabinogion (Jones and Jones 2001). Renowned Anglo-Saxon scholar Rachel Bromwich commented that ‘this passage may represent the eleventh-century tradition about the submerged lands’ between Wales and Ireland and is a story that appears ‘quite independent’ of that about Cantre’r Gwaelod (op cit, p. 228).

  28 As an aside, it is fascinating to consider the possibility that an oral tradition like that of Brân walking from Wales to Ireland may have seen him posthumously changed to a giant, for how else might anyone be persuaded to believe he had crossed an ocean on foot? It may be that an implausible tradition, preserved for millennia, led to the belief that giants once existed. And of course, once that particular fantastic door had been opened, any number of similar traditions became game for the involvement of giants.

  29 These calculations are based on modelling of postglacial sea-level change by Lambeck (op cit). Cornwall is currently sinking at an estimated 0.6mm each year; assuming this rate applied to the last 10,000 years or so (which it probably did not), then we can ascribe some 6m (20ft) of submergence to it, which still requires the sea level to have been 60–70m (200–230ft) lower for these Lyonesse stories to be based on observation of a land bridge between the Scilly Isles and Cornwall.

  30 In ad 1250, Dunwich housed more than 5,000 people living in at least 800 taxable dwellings, spread out over some 330ha (815 acres) (Sear, 2011, International Journal of Nautical Archaeology).

  31 This effect can be explained by Global Isostatic Adjustment (GIA), which pictures the Earth’s crust as continuing to respond to the removal of terrestrial ice loads during the last ice age. Areas that were thickly ice covered are rising because of the ‘flow’ of crustal (lithospheric) material towards them from areas that were not ice covered, or at least not so thickly. The zone of maximum subsidence in north-west Europe crosses the southern North Sea and can be considered a significant contributor to land sinking along the East Anglian coast.

  32 Not to be confused with the Herakleion on the Mediterranean island of Crete.

  33 These figures are those proposed by Stanley (2007). Founded in the sixth century BC, Herakleion thrived for some 600 years; Eastern Canopus was probably established a little later and remained visible until the mid-eighth century ad.

  34 The most complete account of the Canopic distributary and the rediscovery of its Greek port cities is by Stanley (2007). A comparable explanation refers to the ‘disappearance’ of the city of Vineta in the Oder River Delta on the Baltic coast of modern Poland or Germany.

  35 A contemporary account of the disaster by Pausanias described how ‘the sea flooded a great part of the land and encircled the whole of Helike. Moreover, the flood from the sea so covered the sacred grove of Poseidon that only the tops of the trees remained visible’ (quoted in English in the chapter by Soter, p. 41 in the book by Iain Stewart and Claudio Vita-Finzi, 1998).

  36 This remarkable piece of geological sleuthing was carried out by Soter and Katsonopoulou (2011, Geoarchaeology).

  37 Movements of extremely hot water and steam within the Phlegrean Fields caldera, where these ports were located, account for the sometimes frenetic movements of the Earth’s surface here. Hot water or gas, superheated by its proximity to liquid rock below, forces its way upwards through the Earth’s crust, often pushing into chambers just below the surface and causing them to expand, sometimes only temporarily, and the ground surface to rise. Once these chambers empty, the surface can fall. Alternate rise and fall of the ground surface makes the geography of the Phlegrean Fields region one of the most changeable in the world. In 1982–1984, an increase in such ‘bradyseismic’ effects rendered the city centre unsafe and raised the adjacent sea floor by some 2m (6½ft). Somewhat ominously, the dominant cause of ground-surface changes in the region changed in about 2012 from hydrothermal to magmatic, that is from the subterranean movement of hot liquids to the emplacement of a lava reservoir, as shallow as 3,000m (9,850ft), below the streets half a kilometre (⅓ of a mile) from the seafront in Pozzuoli.

  38 Most of the information about the seismotectonic history of the Rann of Kachchh comes from the work of Bilham (in a chapter in the book by Stewart and Vita-Finzi, cited above), who painstakingly analysed all the qualitative accounts to produce a plausible quantitative scenario that can be used to inform future earthquake planning in this region. Reduced crustal stress in the Kachchh area probably means that another big earthquake here is unlikely in the foreseeable future, although accumulated stress may have been transferred along regional fractures towards Karachi, Pakistan.

  39 Cited in Bilham (op cit, p. 5).

  40 Quotes in this paragraph come from the Griffith translation of Hymn 87 (LXXXVII) accessed online in February 2017 at www.sanskritweb.net/rigveda/griffith.pdf.

  41 The most comprehensive account of Dwaraka is the book by Rao (1999).

  42 This information is found in the Harivamsa, an appendix to the Mahabharata. A yojana is considered to have been a measure of distance, perhaps 12–15km (7–9 miles), not area, so it is not straightforward to estimate how much land was reclaimed at Dwaraka. If we assume the extent of coastal protection structures (bunds and sea walls) to be 12km, then the area reclaimed may have been 5–6km2 (about 2mi2).

  43 Never underestimate people’s at
tachment to place, especially when this is imbued with considerable investment of money and effort. After the devastating impact of Hurricane Katrina on New Orleans, USA, in 2005, there were calls for the most vulnerable parts of this iconic city – those as much as 2m (6½ft) below sea level – to be abandoned and their occupants to be compensated and relocated elsewhere. It did not happen.

  44 From the Mausala Parva (Book of Clubs), the sixteenth book of the Mahabharata.

  45 The key details in stories transmitted orally often become compressed, an effect dubbed ‘memory crunch’ (Barber and Barber 2004).

  46 Thermoluminescence (TL) dates were reported by Vora and colleagues (2002, Current Science). When applied to ceramics, the TL technique can measure the time elapsed since firing, a process that resets the TL clock to zero. By heating crystalline samples (as in bits of pottery), light is produced – the thermoluminescence – and the amount of this light is proportional to the radiation dose that has accumulated over time, something that can be converted to age.

  47 Based on the work of R. N. Iyengar (2005, Journal of the Geological Society of India).

  48 The Harappan cultural tradition appeared about 5,200 years ago (3250 BC) along the Indus River Valley, and spread, based on a system of complex urban centres sustained by agricultural production. The Sausashtra Peninsula, where Dwaraka was located, was on the margins of Harappan culture. It is possible that the Rann of Kachchh, today one of the least hospitable parts of India, was awash and navigable during Harappan times, something that would have transformed the productive and exchange potential of this area.

  49 The story is told in Book Six (Yuddha Khanda) of the Ramayana. In the prose rendering of Sarga (Chapter) 22, Rama spoke harshly to the ocean – ‘O, ocean! I will make you dry up now along with your nethermost subterranean region. A vast stretch of sand will appear, when your water gets consumed by my arrows; when you get dried up and the creatures inhabiting you get destroyed by me. By a gush of arrows released by my bow, our monkeys can proceed to the other shore even by foot. O Sea, the abode of demons! You are not able to recognize my valour or prowess through your intelligence. You will indeed get repentance at my hands’ (from www.valmikiramayan.net/yuddha/sarga22/yuddha_22_prose.htm, accessed in February 2017).