""" Hand-curated input tables for the expansion_analogs study. Everything in this file is DATA ENTRY from published sources, not computation. Each row carries a source key (see SOURCES at the bottom) and a verification grade: V = statement checked against the source text in this study session V2 = checked only against a secondary source (Wikipedia article or another paper's summary) M = standard textbook fact entered from memory, not re-verified in this session run_all.py writes these tables to data/curated/*.csv so that they can be inspected and edited. REVISION (after the referee report): the 19 Pacific rows and 7 culture rows that were graded M in the first pass were checked against the English Wikipedia article for the island or topic (raw wikitext retrieved 2026-10-12, source key WP2); rows where the article confirms the coded status now carry grade V2, and the note quotes or paraphrases the article. Rows that could not be confirmed keep grade M. New tables: PACIFIC_UNIT_EXTRA (flags for doubtful settlements and the later-habitability codes), PACIFIC_ISLANDS (island-level table for the abandonment analysis). Status codes for Pacific units persisted : inhabited at European contact abandoned : published evidence of settlement (habitation sites, gardens, marae) but uninhabited at contact traces : only ambiguous or minor traces of visits (isolated marae, graves, rats); no settlement shown none : no accepted evidence of prehistoric presence """ # ---------------------------------------------------------------------------------------------- # 1. PACIFIC: anchor islands (coordinates come from data/wikidata_geo_cache.json via geo_key, # or from the manual lat/lon given here when geo_key is None) # (island, unit, geo_key, manual_lat, manual_lon) # ---------------------------------------------------------------------------------------------- PACIFIC_ANCHORS = [ # West Polynesia and Fiji (settled in the Lapita period or before the eastern pulse) ("Viti Levu", "Fiji", "Viti Levu", None, None), ("Lakeba (Lau group)", "Fiji", None, -18.20, -178.80), ("Tongatapu", "Tonga", "Tongatapu", None, None), ("Lifuka (Ha'apai)", "Tonga", "Lifuka", None, None), ("Vava'u", "Tonga", "Vavaʻu", None, None), ("Niuatoputapu", "Tonga", "Niuatoputapu", None, None), ("Niuafo'ou", "Tonga", "Niuafoʻou", None, None), ("'Ata", "Tonga", "ʻAta", None, None), ("Savai'i", "Samoa", "Savaiʻi", None, None), ("Upolu", "Samoa", "Upolu", None, None), ("Tutuila", "Samoa", "Tutuila", None, None), ("Ta'u (Manu'a)", "Samoa", "Taʻū", None, None), ("'Uvea (Wallis)", "Uvea", "Wallis (island)", None, None), ("Futuna", "Futuna", "Futuna (Wallis and Futuna)", None, None), ("Rotuma", "Rotuma", "Rotuma", None, None), ("Niue", "Niue", "Niue", None, None), ("Atafu", "Tokelau", "Atafu", None, None), ("Nukunonu", "Tokelau", "Nukunonu", None, None), ("Fakaofo", "Tokelau", "Fakaofo", None, None), ("Swains (Olohega)", "Tokelau", "Swains Island", None, None), ("Funafuti", "Tuvalu", "Funafuti", None, None), ("Vaitupu", "Tuvalu", "Vaitupu", None, None), ("Nanumea", "Tuvalu", "Nanumea", None, None), ("Pukapuka", "Pukapuka", "Pukapuka", None, None), ("Nassau", "Pukapuka", "Nassau (Cook Islands)", None, None), # Cook Islands ("Rarotonga", "Southern Cooks", "Rarotonga", None, None), ("Aitutaki", "Southern Cooks", "Aitutaki", None, None), ("Atiu", "Southern Cooks", "Atiu", None, None), ("Mitiaro", "Southern Cooks", "Mitiaro", None, None), ("Mauke", "Southern Cooks", "Mauke", None, None), ("Mangaia", "Southern Cooks", "Mangaia", None, None), ("Manuae", "Southern Cooks", "Manuae (Cook Islands)", None, None), ("Palmerston", "Palmerston", "Palmerston Island", None, None), ("Suwarrow", "Suwarrow", "Suwarrow", None, None), ("Manihiki", "Manihiki-Rakahanga", "Manihiki", None, None), ("Rakahanga", "Manihiki-Rakahanga", "Rakahanga", None, None), ("Penrhyn (Tongareva)", "Penrhyn", "Penrhyn atoll", None, None), # Society Islands ("Tahiti", "Society Islands", "Tahiti", None, None), ("Raiatea", "Society Islands", "Raiatea", None, None), ("Bora Bora", "Society Islands", "Bora Bora", None, None), ("Maupiti", "Society Islands", "Maupiti", None, None), ("Maupiha'a", "Society Islands", "Maupihaa", None, None), # Tuamotus (anchors spaced so the chain is represented end to end) ("Mataiva", "Tuamotus", "Mataiva", None, None), ("Rangiroa", "Tuamotus", "Rangiroa", None, None), ("Makatea", "Tuamotus", "Makatea", None, None), ("Takaroa", "Tuamotus", "Takaroa", None, None), ("Fakarava", "Tuamotus", "Fakarava", None, None), ("Anaa", "Tuamotus", "Anaa", None, None), ("Makemo", "Tuamotus", "Makemo", None, None), ("Hao", "Tuamotus", "Hao (French Polynesia)", None, None), ("Napuka", "Tuamotus", "Napuka", None, None), ("Puka-Puka (Tuamotu)", "Tuamotus", "Puka-Puka", None, None), ("Tatakoto", "Tuamotus", "Tatakoto", None, None), ("Reao", "Tuamotus", "Reao", None, None), ("Tureia", "Tuamotus", "Tureia", None, None), ("Moruroa", "Tuamotus", "Moruroa", None, None), ("Marutea Sud", "Tuamotus", "Marutea Sud", None, None), # Marquesas ("Nuku Hiva", "Marquesas", "Nuku Hiva", None, None), ("Hiva Oa", "Marquesas", "Hiva Oa", None, None), ("Fatu Hiva", "Marquesas", "Fatu Hiva", None, None), ("Eiao", "Marquesas", "Eiao", None, None), # Australs, Rapa, Gambier, Pitcairn group, Rapa Nui ("Rimatara", "Austral Islands", "Rimatara", None, None), ("Rurutu", "Austral Islands", "Rurutu", None, None), ("Tubuai", "Austral Islands", "Tubuai", None, None), ("Raivavae", "Austral Islands", "Raivavae", None, None), ("Rapa Iti", "Rapa Iti", "Rapa Iti", None, None), ("Mangareva", "Mangareva (Gambier)", "Mangareva", None, None), ("Temoe", "Mangareva (Gambier)", "Temoe", None, None), ("Pitcairn", "Pitcairn", "Pitcairn Island", None, None), ("Henderson", "Henderson", "Henderson Island (Pitcairn Islands)", None, None), ("Oeno", "Oeno", "Oeno Island", None, None), ("Ducie", "Ducie", "Ducie Island", None, None), ("Rapa Nui (Easter Island)", "Rapa Nui", "Easter Island", None, None), ("Salas y Gomez", "Salas y Gomez", "Isla Salas y Gómez", None, None), # Line Islands ("Teraina (Washington)", "Teraina", "Teraina", None, None), ("Tabuaeran (Fanning)", "Tabuaeran", "Tabuaeran", None, None), ("Kiritimati (Christmas)", "Kiritimati", "Kiritimati", None, None), ("Palmyra", "Palmyra", "Palmyra Atoll", None, None), ("Jarvis", "Jarvis", "Jarvis Island", None, None), ("Malden", "Malden", "Malden Island", None, None), ("Starbuck", "Starbuck", "Starbuck Island", None, None), ("Caroline (Millennium)", "Caroline", "Caroline Island", None, None), ("Flint", "Flint", "Flint Island", None, None), ("Vostok", "Vostok", "Vostok Island", None, None), # Phoenix Islands, Howland, Baker ("Manra (Sydney)", "Manra", "Manra", None, None), ("Orona (Hull)", "Orona", "Orona", None, None), ("Nikumaroro (Gardner)", "Nikumaroro", "Nikumaroro", None, None), ("Kanton", "Kanton", "Kanton Island", None, None), ("Enderbury", "Enderbury", "Enderbury Island", None, None), ("Howland", "Howland", "Howland Island", None, None), ("Baker", "Baker", "Baker Island", None, None), # Hawaii ("Hawai'i", "Hawaiian Islands", "Hawaii (island)", None, None), ("Maui", "Hawaiian Islands", "Maui", None, None), ("O'ahu", "Hawaiian Islands", "Oahu", None, None), ("Kaua'i", "Hawaiian Islands", "Kauai", None, None), ("Ni'ihau", "Hawaiian Islands", "Niihau", None, None), ("Nihoa", "Nihoa", "Nihoa", None, None), ("Necker (Mokumanamana)", "Necker", "Necker Island (Hawaii)", None, None), ("Laysan", "NW Hawaiian atolls", "Laysan", None, None), ("Midway", "NW Hawaiian atolls", "Midway Atoll", None, None), ("Johnston", "Johnston", "Johnston Atoll", None, None), # South-west margin ("Raoul", "Raoul (Kermadec)", "Raoul Island", None, None), ("Macauley", "Raoul (Kermadec)", "Macauley Island", None, None), ("Norfolk", "Norfolk", "Norfolk Island", None, None), ("Lord Howe", "Lord Howe", "Lord Howe Island", None, None), ("NZ: Cape Reinga (Northland)", "New Zealand", None, -34.43, 172.68), ("NZ: East Cape", "New Zealand", None, -37.69, 178.55), ("NZ: Wellington", "New Zealand", None, -41.29, 174.78), ("NZ: Otago", "New Zealand", None, -45.88, 170.50), ("NZ: Stewart Island", "New Zealand", "Stewart Island", None, None), ("Chatham", "Chatham Islands", "Chatham Island", None, None), ("Enderby / Auckland Island", "Auckland Islands", "Enderby Island", None, None), ("Snares", "Snares", "The Snares", None, None), ("Campbell", "Campbell", "Campbell Island, New Zealand", None, None), ("Antipodes", "Antipodes", "Antipodes Islands", None, None), ("Bounty", "Bounty", "Bounty Islands", None, None), ("Macquarie", "Macquarie", "Macquarie Island", None, None), # Eastern rim (never settled by Polynesians) ("Galapagos", "Galapagos", "Galápagos Islands", None, None), ("Robinson Crusoe (Juan Fernandez)", "Juan Fernandez", "Robinson Crusoe Island", None, None), ("San Felix (Desventuradas)", "Desventuradas", "San Félix Island", None, None), ("Clipperton", "Clipperton", "Clipperton Island", None, None), ("Socorro (Revillagigedo)", "Revillagigedo", "Socorro Island", None, None), ("Cocos", "Cocos Island", "Cocos Island", None, None), ] # ---------------------------------------------------------------------------------------------- # 2. PACIFIC: unit attributes # fields: unit, region, land_km2, type, status, canonical_mystery, habitable_later, # S_lo, S_hi (short chronology window for first settlement, CE), # A_lo, A_hi (alternative chronology window, CE), # last_occ (approx. CE year of latest dated occupation for abandoned units, or None), # parent (evidence-based source unit, or None), grade, sources, note # region: W = settled before the eastern pulse (West Polynesia + Fiji), E = East Polynesia pulse domain, # X = eastern rim beyond the settled network # habitable_later: 1 if the unit sustained a self-supporting population after European contact # (used only to classify never-settled islands as real targets), 0 if not, None if not applicable # Dates: None means no usable chronology (unit is left out of all timing analyses). # ---------------------------------------------------------------------------------------------- PACIFIC_UNITS = [ # ---- West (pre-pulse) ---- ("Fiji", "W", 18270, "high", "persisted", 0, None, -1100, -1000, -1100, -1000, None, None, "V", "L22", "Lapita settlement c. 1050 BC (Sheppard et al. 2015 as cited by L22)."), ("Tonga", "W", 748, "high/makatea", "persisted", 0, None, -900, -800, -900, -800, None, "Fiji", "V", "B12;L22", "U/Th coral-file date 2838 +/- 8 BP at Nukuleka (B12); L22 gives c. 900 BC."), ("Samoa", "W", 3041, "high", "persisted", 0, None, -850, -750, -850, -750, None, "Tonga", "V", "L22", "Only strictly Lapita site (Mulifanua) c. 800 BC (Petchey 2001 as cited by L22)."), ("Uvea", "W", 78, "high", "persisted", 0, None, None, None, None, None, None, None, "V2", "WP2", "Polynesian population at contact (WP2). Lapita-period settlement in standard accounts; the article only " "mentions sites of about 1400; date not used."), ("Futuna", "W", 64, "high", "persisted", 0, None, None, None, None, None, None, None, "M", "-", "Inhabited at contact. Lapita-period settlement in standard accounts; the article gives no date, so the " "row stays grade M; excluded from timing."), ("Rotuma", "W", 46, "high", "persisted", 0, None, None, None, None, None, None, None, "V2", "WP2", "Indigenous Rotuman population (WP2); first-settlement date poorly known."), ("Niue", "W", 260, "makatea", "persisted", 0, None, None, None, None, None, None, None, "V2", "WP2", "Inhabited at contact. WP2: 'Polynesians from Samoa settled on Niue around 900 CE'; other accounts give " "about 2000 BP. Chronology unresolved, so Niue may belong to the pulse, not the pre-pulse source region " "(see the 'core sources' sensitivity run)."), ("Tokelau", "W", 12, "atoll", "persisted", 0, None, None, None, None, None, None, None, "V2", "WP2", "Inhabited at contact. WP2: settled 'about 1,000 years ago from Samoa and may have been a gateway into " "Eastern Polynesia'. May belong to the pulse (see the 'core sources' sensitivity run)."), ("Tuvalu", "W", 26, "atoll", "persisted", 0, None, None, None, None, None, None, None, "V2", "WP2", "Inhabited at contact; settled from Samoa and Tonga (WP2); chronology poorly known."), ("Pukapuka", "W", 4.3, "atoll", "persisted", 0, None, None, None, None, None, None, None, "V2", "WP2", "Inhabited at contact. WP2: settlement 'can be dated back about 1,000 years'; a hub between West and East " "Polynesia. May belong to the pulse (see the 'core sources' sensitivity run)."), # ---- East Polynesia: persisted ---- ("Southern Cooks", "E", 212, "high/makatea", "persisted", 0, None, 1231, 1290, 800, 950, None, "Samoa", "V", "I21;W11;L22", "Short: mid-1200s (W11), 1231-1290 (Schmid 2018). Alt: genetic 830 CE Rarotonga (I21), " "Atiu about 900 CE (Sear 2020); L22 quotes 1050-1270 (Allen & Morrison 2013)."), ("Manihiki-Rakahanga", "E", 9.5, "atoll", "persisted", 0, None, None, None, None, None, None, None, "V2", "WP2", "Inhabited at contact; one population alternating between the two atolls (WP2). The articles give " "inconsistent dates (Manihiki 'since at least 900 or 1000 AD'; Rakahanga 'about 1350 AD'), so no chronology is used."), ("Penrhyn", "E", 9.8, "atoll", "persisted", 0, None, None, None, None, None, None, None, "V2", "WP2", "Inhabited at contact, estimated pre-European population 2,000 (WP2). Unsourced date in the article " "('since 900 or 1000 AD'); no chronology used."), ("Society Islands", "E", 1590, "high", "persisted", 0, None, 1025, 1121, 1000, 1100, None, "Southern Cooks", "V", "W11;I21;L22", "Short: 1025-1121 (W11). Alt: 1000-1100 (Kahn 2017), genetic 1050 CE Tahiti (I21)."), ("Tuamotus", "E", 850, "atoll", "persisted", 0, None, 1190, 1293, 1070, 1150, None, "Society Islands", "V", "W11;I21", "Short: no archipelago-specific range in W11 text, so the generic pulse window 1190-1293 is used. " "Alt: genetic 1110 CE for Mataiva (I21)."), ("Marquesas", "E", 1049, "high", "persisted", 0, None, 1200, 1277, 1100, 1200, None, "Tuamotus", "V", "W11;I21;L22", "Short: 1200-1277 (W11). Alt: genetic 1140 CE Fatu Hiva (I21); 1166-1258 (Allen 2014); " "L22 gives 1000-1200."), ("Austral Islands", "E", 104, "high", "persisted", 0, None, 1190, 1293, 1100, 1228, None, None, "V", "I21;W11", "Short: generic pulse window. Alt: genetic 1150 CE Tubuai (I21); 1128-1228 (Prebble 2009 in I21)."), ("Rapa Iti", "E", 40.6, "high", "persisted", 0, None, 1100, 1200, 1100, 1200, None, "Southern Cooks", "V", "I21;L22", "1100-1200 (Anderson & Kennett 2012 in I21); genetic 1190 CE (I21); about 1100 (Prebble 2013 in L22)."), ("Mangareva (Gambier)", "E", 31, "high", "persisted", 0, None, 1108, 1275, 950, 1150, None, "Tuamotus", "V", "W11;I21", "Short: 1108-1275 (W11). Alt: genetic 1130 CE (I21); 950 CE (Kirch 2010 in I21)."), ("Rapa Nui", "E", 164, "high", "persisted", 0, None, 1200, 1253, 1150, 1250, None, "Mangareva (Gambier)", "V", "W11;I21", "Short: 1200-1253 (W11). Alt: genetic 1210 CE (I21)."), ("Hawaiian Islands", "E", 16636, "high", "persisted", 0, None, 1219, 1266, 940, 1130, None, None, "V", "W11;L22", "Short: 1219-1266 (W11). Alt: 940-1130, 95% HPD (Athens et al. 2014 in L22)."), ("New Zealand", "E", 268000, "continental", "persisted", 0, None, 1230, 1282, 1270, 1309, None, None, "V", "W11;L22", "Short: 1230-1282 (W11). Alt: 1270-1309, 95% (Dye 2015 in L22)."), ("Chatham Islands", "E", 966, "high", "persisted", 0, None, 1400, 1500, 1400, 1500, None, "New Zealand", "V2", "WS", "15th century to about 1500 (Te Ara; Landcare Research summaries)."), # ---- East Polynesia: abandoned before European contact ---- ("Palmerston", "E", 2.1, "atoll", "abandoned", 1, 1, None, None, None, None, None, None, "V", "L16;WP", "Canonical mystery island (L16). Cook found it uninhabited in 1777 with ancient graves (WP)."), ("Suwarrow", "E", 0.4, "atoll", "abandoned", 1, 0, None, None, None, None, None, None, "V", "L16;WP", "Canonical mystery island (L16)."), ("Pitcairn", "E", 4.5, "high", "abandoned", 1, 1, 1150, 1293, 1000, 1250, 1650, "Mangareva (Gambier)", "V", "L16;L22;WP", "Canonical (L16). Settled around 1200 with Mangareva and Rapa Nui (L22); abandoned in the " "17th or early 18th century (WP)."), ("Henderson", "E", 37.3, "makatea", "abandoned", 1, 0, 1150, 1293, 800, 1250, 1560, "Mangareva (Gambier)", "V", "L16;L22;WP", "Canonical (L16). Latest viable date about 390 bp (L16). Weisler's long chronology allows 800-1050."), ("Teraina", "E", 9.6, "atoll", "abandoned", 1, 1, None, None, None, None, None, None, "V", "L16;WP", "Canonical ('Washington', L16). Stone structures and a canoe; habitation span unknown (WP)."), ("Tabuaeran", "E", 33.7, "atoll", "abandoned", 1, 1, 1100, 1293, 1100, 1293, 1425, None, "V", "L16;WP", "Canonical ('Fanning', L16). Radiometric dates span about 1100-1425 CE (WP)."), ("Kiritimati", "E", 388, "atoll", "abandoned", 1, 1, 1190, 1293, 1190, 1293, 1450, None, "V", "L16;WP", "Canonical (L16). Habitation sites and marae; used about 1250-1450 (WP)."), ("Malden", "E", 39.3, "atoll", "abandoned", 1, 0, None, None, None, None, None, None, "V", "L16;WP", "Canonical (L16). Extensive stone architecture (WP)."), ("Manra", "E", 4.4, "atoll", "abandoned", 0, 0, None, None, None, None, None, None, "V2", "WP", "Not on the canonical list. 'Definite evidence of prehistoric habitation': platforms and enclosures (WP). " "Twentieth-century resettlement failed (1938-1963)."), ("Orona", "E", 3.9, "atoll", "abandoned", 0, 0, None, None, None, None, None, None, "V2", "WP", "Not on the canonical list. Marae, shelters, graves and platforms (WP)."), ("Howland", "E", 1.8, "atoll", "abandoned", 1, 0, None, None, None, None, None, None, "V", "L16;WP", "Canonical (L16), but weak: a 1987 survey found no evidence of prehistoric settlement (WP). " "Recoded as 'traces' in the strict robustness run."), ("Nihoa", "E", 0.7, "high", "abandoned", 1, 0, 1300, 1600, 1000, 1600, 1636, "Hawaiian Islands", "V", "L16;WP", "Canonical (L16). Terraces and house sites; latest possible date about 314 bp (L16)."), ("Necker", "E", 0.18, "high", "abandoned", 1, 0, 1300, 1600, 1000, 1600, None, "Hawaiian Islands", "V", "L16;WP", "Canonical (L16). Shrines; ritual use, long-term settlement unclear (WP)."), ("Raoul (Kermadec)", "E", 32.5, "high", "abandoned", 1, 0, 1230, 1350, 1230, 1350, 1496, "New Zealand", "V", "L16;AW01;WP", "Canonical (L16). Latest date about 454 bp (L16). Settled in the 14th century, from New Zealand " "per Higham & Johnson 1997 (WP)."), ("Norfolk", "E", 36, "high", "abandoned", 1, 1, 1230, 1350, 1230, 1350, 1450, "Raoul (Kermadec)", "V", "L16;AW01", "Canonical (L16). Occupied early 13th to early 15th century, colonised from Raoul (AW01)."), ("Auckland Islands", "E", 570, "subantarctic", "abandoned", 0, 0, 1250, 1350, 1250, 1350, 1390, "New Zealand", "V2", "A05;A24", "Enderby Island: ovens, middens, tools of the 13th-14th century (A05); vacated in the late " "14th century (A24). Not on the canonical list."), # ---- East Polynesia: traces only ---- ("Caroline", "E", 3.8, "atoll", "traces", 0, 0, None, None, None, None, None, None, "V2", "WP", "Marae and graves but no evidence of long-term settlement (WP)."), ("Flint", "E", 3.2, "atoll", "traces", 0, 0, None, None, None, None, None, None, "V2", "WP", "Listed among Line Islands with remains (WP: Line Islands); island article reports none."), # ---- East Polynesia: no evidence of settlement ---- ("Oeno", "E", 0.65, "atoll", "none", 0, 0, None, None, None, None, None, None, "M", "-", "No settlement evidence known; the article (WP2) mentions no prehistoric remains, which is absence of a statement, so grade M stays."), ("Ducie", "E", 0.7, "atoll", "none", 0, 0, None, None, None, None, None, None, "V2", "WP", "Polynesian rat present; no sites."), ("Salas y Gomez", "E", 0.15, "rock", "none", 0, 0, None, None, None, None, None, None, "V2", "WP2", "'No evidence that the island has ever been permanently inhabited'; known to Rapa Nui islanders (WP2)."), ("Palmyra", "E", 3.9, "atoll", "none", 0, 0, None, None, None, None, None, None, "V2", "WP", "No remains reported; never surveyed in depth (WP)."), ("Jarvis", "E", 4.5, "atoll", "none", 0, 0, None, None, None, None, None, None, "V2", "WP", "Unlikely to have hosted occupation (WP)."), ("Starbuck", "E", 16.2, "atoll", "none", 0, 0, None, None, None, None, None, None, "V2", "WP", "No remains reported."), ("Vostok", "E", 0.24, "atoll", "none", 0, 0, None, None, None, None, None, None, "V2", "WP", "No evidence of settlement (WP)."), ("Nikumaroro", "E", 4.1, "atoll", "none", 0, 0, None, None, None, None, None, None, "V2", "WP", "Only speculative mention; 1938-1963 colony failed."), ("Kanton", "E", 9.2, "atoll", "none", 0, 0, None, None, None, None, None, None, "V2", "WP", "Only speculative mention."), ("Enderbury", "E", 5.1, "atoll", "none", 0, 0, None, None, None, None, None, None, "V2", "WP", "Polynesian rat present; no sites reported."), ("Baker", "E", 2.1, "atoll", "none", 0, 0, None, None, None, None, None, None, "V2", "WP", "No remains reported."), ("NW Hawaiian atolls", "E", 11, "atoll", "none", 0, 0, None, None, None, None, None, None, "V2", "WP2", "Midway 'has no indigenous inhabitants and was uninhabited until the 19th century' (WP2). No prehistoric sites beyond Necker."), ("Johnston", "E", 2.7, "atoll", "none", 0, 0, None, None, None, None, None, None, "V2", "WP2", "'Johnston Atoll has never had any indigenous inhabitants' (WP2)."), ("Lord Howe", "E", 14.6, "high", "none", 0, 1, None, None, None, None, None, None, "V2", "WP", "No prehistoric human activity found despite a dedicated 1996 investigation (Anderson 2003, quoted in WP). " "Self-supporting settlement since 1834."), ("Snares", "E", 3.5, "subantarctic", "none", 0, 0, None, None, None, None, None, None, "V2", "WP", "Known to Maori (WP); no settlement."), ("Campbell", "E", 113, "subantarctic", "none", 0, 0, None, None, None, None, None, None, "V2", "A24", "No prehistoric remains south of the Auckland Islands (A24)."), ("Antipodes", "E", 22, "subantarctic", "none", 0, 0, None, None, None, None, None, None, "V2", "A24", "As Campbell."), ("Bounty", "E", 1.35, "rock", "none", 0, 0, None, None, None, None, None, None, "V2", "A24", "As Campbell."), ("Macquarie", "E", 128, "subantarctic", "none", 0, 0, None, None, None, None, None, None, "V2", "A24", "As Campbell."), # ---- Eastern rim, beyond the settled network ---- ("Galapagos", "X", 7880, "high", "none", 0, 1, None, None, None, None, None, None, "V2", "WP2", "'No firm evidence that Polynesians or the Indigenous peoples of South America reached the islands' before 1535; " "a 2008 ANU survey allowed a possible Polynesian visit (WP2). Settled from 1832."), ("Juan Fernandez", "X", 100, "high", "none", 0, 1, None, None, None, None, None, None, "V2", "WP2", "'Unlikely there had been human activity in the islands before Europeans arrived' (2008 ANU survey quoted in WP2). Settled from 1750."), ("Desventuradas", "X", 5.4, "high", "none", 0, 0, None, None, None, None, None, None, "V2", "WP2", "'No signs of prehistoric human activity ... have ever been found on the islands' (WP2)."), ("Clipperton", "X", 6, "atoll", "none", 0, 0, None, None, None, None, None, None, "V2", "WP2", "Uninhabited when discovered, no evidence of prehistoric activity (WP2, Galapagos article). A guano colony of 1906-1917 failed."), ("Revillagigedo", "X", 158, "high", "none", 0, 0, None, None, None, None, None, None, "V2", "WP2", "'No evidence of human habitation on any of the islands exists before their discovery by Spanish explorers' (WP2)."), ("Cocos Island", "X", 23.5, "high", "none", 0, 0, None, None, None, None, None, None, "V2", "WP2", "Uninhabited when discovered by Europeans; little archaeological investigation (WP2)."), ] PACIFIC_UNIT_FIELDS = ["unit", "region", "land_km2", "type", "status", "canonical_mystery", "habitable_later", "S_lo", "S_hi", "A_lo", "A_hi", "last_occ", "parent", "grade", "sources", "note"] # ---------------------------------------------------------------------------------------------- # 2b. PACIFIC: extra unit-level codes added in the revision (East Polynesia and eastern-rim units that were NOT # inhabited at European contact; persisted units need no entry). # unit: (evidence, later_strict, later_lenient, grade, note) # evidence (abandoned units only): quality of the published evidence that the unit held a resident settlement # 'habitation' = habitation sites, gardens or a village documented in the sources read # 'thin' = on the canonical list, but the sources read document only graves or structures of unknown span # 'doubtful' = the sources read speak AGAINST a resident settlement (ritual use, seasonal camp, or a negative survey) # later_strict : 1 if a self-supporting civilian community was founded after European contact and still exists # later_lenient: 1 if, in addition, any post-contact civilian settlement lasted about 20 years or more (even if it failed) # These two codes define habitability by what happened AFTER contact, without reference to the prehistoric outcome. # ---------------------------------------------------------------------------------------------- PACIFIC_UNIT_EXTRA = { # abandoned before contact "Palmerston": ("thin", 1, 1, "V2", "Cook found 'ancient graves' (WP2). Marsters family community since 1863 (WP2)."), "Suwarrow": ("thin", 0, 0, "V2", "About thirty Lever Brothers workers 1903-1914; since then caretakers only (WP2)."), "Pitcairn": ("habitation", 1, 1, "V2", "Community since 1790."), "Henderson": ("habitation", 0, 0, "V2", "Prehistoric settlement sites 500-800 years old (WP2); never resettled."), "Teraina": ("thin", 1, 1, "V2", "Archaeological sites exist but 'the general lack of fresh water makes long-term human habitation " "unlikely' (WP2). About 1,900 inhabitants in 2020 after plantation and resettlement schemes (WP2)."), "Tabuaeran": ("habitation", 1, 1, "V2", "'A single large village being maintained for several hundred years' (WP2). Plantation " "labourers from 1854; about 2,300 inhabitants in 2015 (WP2)."), "Kiritimati": ("habitation", 1, 1, "V2", "Habitation sites and marae; fire pits show alternating use and disuse (WP2). Permanent " "settlement from 1882, abandoned 1905-1912, resettled (WP2)."), "Malden": ("habitation", 0, 0, "V2", "Stone architecture consistent with 100-200 people (WP2). Guano workforce 1860-1927, supplied from outside."), "Manra": ("habitation", 0, 1, "V2", "Phoenix Islands Settlement Scheme 1938-1963: about 130 settlers, failed through drought (WP2)."), "Orona": ("habitation", 0, 1, "V2", "Copra plantation from 1916; settlement scheme village evacuated 1963 (WP2)."), "Howland": ("doubtful", 0, 0, "V2", "'The only modern archaeological survey of Howland ... 1987, found no evidence of prehistoric " "settlement or use' (WP2). Four colonists 1935-1942."), "Nihoa": ("habitation", 0, 0, "V2", "'Extensive prehistoric agricultural terraces and house sites' (WP2)."), "Necker": ("doubtful", 0, 0, "V2", "Ceremonial complexes and stone figures; no habitation evidence in the sources read (WP2)."), "Raoul (Kermadec)": ("habitation", 0, 1, "V2", "Bell family lived on Raoul for thirty-six years until evacuated in 1914; staffed station since 1937 (WP2)."), "Norfolk": ("habitation", 1, 1, "V2", "Settled 1788; continuous community since 1856."), "Auckland Islands": ("doubtful", 0, 1, "V2", "Enderby Island: 'the Polynesians stayed for one or more summers' (WP2), i.e. possibly a seasonal " "camp. Maori and Moriori settlement 1842-1866; Hardwicke 1849-1852 (WP2)."), # traces only / never settled "Caroline": (None, 0, 1, "V2", "A small settlement existed intermittently from 1846 until 1904 (WP2)."), "Flint": (None, 0, 0, "V2", "Guano digging 1875-1880 and copra to 1891 with hired labour (WP2)."), "Oeno": (None, 0, 0, "M", "Uninhabited."), "Ducie": (None, 0, 0, "V2", "Uninhabited (WP2)."), "Salas y Gomez": (None, 0, 0, "V2", "Never permanently inhabited (WP2)."), "Palmyra": (None, 0, 0, "V2", "'No permanent population'; transient staff of 4-25 (WP2)."), "Jarvis": (None, 0, 0, "V2", "'No evidence that the island has ever supported a self-sustaining human population' (WP2)."), "Starbuck": (None, 0, 0, "V2", "Nineteenth-century guano settlement only (WP2)."), "Vostok": (None, 0, 0, "V2", "Uninhabited (WP2)."), "Nikumaroro": (None, 0, 1, "V2", "Colony 1938-1963; failed, 'scarcity of fresh water' (WP2)."), "Kanton": (None, 0, 1, "V2", "'Repopulated with a few people' from the Gilbert Islands; one village, sole inhabited Phoenix island in 2020 (WP2). " "Supplied from outside, so not counted under the strict code."), "Enderbury": (None, 0, 0, "V2", "Four colonists 1938-1942 (WP2)."), "Baker": (None, 0, 0, "V2", "Colonisation-project camp 1935-1942 (WP2)."), "NW Hawaiian atolls": (None, 0, 0, "V2", "Cable and military stations only (WP2)."), "Johnston": (None, 0, 0, "V2", "Military base only (WP2)."), "Lord Howe": (None, 1, 1, "V2", "Permanent settlement since 1834 (WP2)."), "Snares": (None, 0, 0, "V2", "Uninhabited (WP2)."), "Campbell": (None, 0, 0, "M", "Sealing base and (from memory, not confirmed in the article) a sheep farm 1895-1931; coded 0."), "Antipodes": (None, 0, 0, "V2", "Uninhabited."), "Bounty": (None, 0, 0, "V2", "Uninhabited."), "Macquarie": (None, 0, 0, "V2", "Research station only."), "Galapagos": (None, 1, 1, "V2", "Settled from 1832."), "Juan Fernandez": (None, 1, 1, "V2", "Settled from 1750."), "Desventuradas": (None, 0, 0, "V2", "Naval post only."), "Clipperton": (None, 0, 0, "V2", "Guano colony 1906-1917 failed; uninhabited (WP2)."), "Revillagigedo": (None, 0, 0, "V2", "Naval station only (WP2)."), "Cocos Island": (None, 0, 0, "V2", "Park rangers only."), } PACIFIC_UNIT_EXTRA_FIELDS = ["unit", "evidence", "later_strict", "later_lenient", "extra_grade", "extra_note"] # ---------------------------------------------------------------------------------------------- # 2c. PACIFIC: island-level table for the abandonment analysis (added in the revision; the referee showed that the # unit-level table merges small persisted islands into archipelagos while every abandoned island stands alone). # One row per island of East Polynesia with evidence of prehistoric settlement. Abandoned islands are the 16 abandoned # units at island grain. For the 13 persisted units the list holds every island named in the English Wikipedia group # article with an infobox land area of at least about 1 km2, EXCEPT the Tuamotus, where only the 15 anchor atolls of the # unit table are listed (the archipelago has 76 atolls) - a stated incompleteness. # (island, unit, wp_title, land_km2, area_source, contact, grade, note) # land_km2: LAND area. WPI = infobox of the English Wikipedia article (raw wikitext retrieved 2026-10-12; for atolls the # 'above water' figure, not the lagoon); WD = Wikidata P2046; U = the value of the unit table. # contact: 'inhabited' (resident population at European contact), 'doubtful' (the article says or implies that the island # was not, or not continuously, inhabited at contact, or says nothing and the island is uninhabited today), 'abandoned'. # grade refers to the contact status: V2 = stated in the article, M = standard fact not stated in the text read. # ---------------------------------------------------------------------------------------------- PACIFIC_ISLANDS = [ # Southern Cooks ("Rarotonga", "Southern Cooks", "Rarotonga", 67.39, "WPI", "inhabited", "M", ""), ("Aitutaki", "Southern Cooks", "Aitutaki", 18.3, "WPI", "inhabited", "M", ""), ("Atiu", "Southern Cooks", "Atiu", 26.9, "WPI", "inhabited", "M", ""), ("Mitiaro", "Southern Cooks", "Mitiaro", 22.25, "WPI", "inhabited", "M", ""), ("Mauke", "Southern Cooks", "Mauke", 18.4, "WPI", "inhabited", "M", ""), ("Mangaia", "Southern Cooks", "Mangaia", 51.8, "WPI", "inhabited", "M", ""), ("Manuae (Cooks)", "Southern Cooks", "Manuae (Cook Islands)", 6.17, "WPI", "inhabited", "V2", "Cook 'found the island inhabited' in 1777; sixty people in 1823; depopulated later."), # Northern Cooks ("Manihiki", "Manihiki-Rakahanga", "Manihiki", 4.0, "WPI", "inhabited", "V2", "Used as a food supply by the people of Rakahanga; one alternating population."), ("Rakahanga", "Manihiki-Rakahanga", "Rakahanga", 4.0, "WPI", "inhabited", "V2", ""), ("Penrhyn", "Penrhyn", "Penrhyn atoll", 9.8, "WPI", "inhabited", "V2", "Estimated pre-European population 2,000."), # Society Islands ("Tahiti", "Society Islands", "Tahiti", 1044, "WPI", "inhabited", "M", ""), ("Mo'orea", "Society Islands", "Moʼorea", 134, "WPI", "inhabited", "M", ""), ("Huahine", "Society Islands", "Huahine", 75, "WPI", "inhabited", "M", ""), ("Ra'iatea", "Society Islands", "Raʻiātea", 167.7, "WPI", "inhabited", "M", ""), ("Taha'a", "Society Islands", "Tahaʻa", 90.2, "WPI", "inhabited", "M", ""), ("Bora Bora", "Society Islands", "Bora Bora", 30.55, "WPI", "inhabited", "M", ""), ("Maupiti", "Society Islands", "Maupiti", 11, "WPI", "inhabited", "M", ""), ("Mai'ao", "Society Islands", "Maiʻao", 8.8, "WPI", "inhabited", "M", "Inhabited today (343 people); the article gives no contact-era statement."), ("Teti'aroa", "Society Islands", "Tetiʻaroa", 6, "WPI", "inhabited", "M", "Atoll of the Tahitian chiefs; 'many important archaeological sites'."), ("Mehetia", "Society Islands", "Mehetiʻa", 2.3, "WPI", "doubtful", "M", "Uninhabited today; the article gives no contact-era statement."), ("Maupiha'a", "Society Islands", "Maupihaʻa", 2.6, "WPI", "doubtful", "V2", "Archaeological remains; 'inhabited by only three Polynesians' in 1767."), ("Tupai", "Society Islands", "Tupai", 11, "WPI", "doubtful", "M", "Uninhabited today; no contact-era statement."), ("Motu One", "Society Islands", "Motu One (Society Islands)", 2.83, "WPI", "doubtful", "M", "Uninhabited; no contact-era statement."), ("Manua'e (Scilly)", "Society Islands", "Manuaʻe (Society Islands)", 3.5, "WPI", "doubtful", "M", "No contact-era statement."), # Tuamotus: anchor atolls only ("Mataiva", "Tuamotus", "Mataiva", 16, "WD", "doubtful", "M", "First European visit 1820; no contact-era statement. Infobox land/lagoon figures look swapped; Wikidata value used."), ("Rangiroa", "Tuamotus", "Rangiroa", 79, "WPI", "inhabited", "M", ""), ("Makatea", "Tuamotus", "Makatea", 24, "WPI", "inhabited", "M", ""), ("Takaroa", "Tuamotus", "Takaroa", 20, "WPI", "inhabited", "M", ""), ("Fakarava", "Tuamotus", "Fakarava", 24.1, "WPI", "inhabited", "M", ""), ("Anaa", "Tuamotus", "Anaa", 38, "WPI", "inhabited", "M", ""), ("Makemo", "Tuamotus", "Makemo", 56, "WPI", "inhabited", "V2", "About 250 indigenous inhabitants around 1850."), ("Hao", "Tuamotus", "Hao (French Polynesia)", 35, "WPI", "inhabited", "M", ""), ("Napuka", "Tuamotus", "Napuka", 8, "WPI", "inhabited", "M", ""), ("Puka-Puka (Tuamotu)", "Tuamotus", "Puka-Puka", 5, "WPI", "inhabited", "M", ""), ("Tatakoto", "Tuamotus", "Tatakoto", 7.3, "WPI", "inhabited", "M", ""), ("Reao", "Tuamotus", "Reao", 9, "WPI", "inhabited", "M", ""), ("Tureia", "Tuamotus", "Tureia", 8, "WPI", "inhabited", "M", ""), ("Moruroa", "Tuamotus", "Moruroa", 15, "WPI", "doubtful", "V2", "'Early European explorers found that the atoll was not continuously inhabited.'"), ("Marutea Sud", "Tuamotus", "Marutea Sud", 12, "WPI", "doubtful", "V2", "'Once inhabited by the ancient Polynesians' according to Krusenstern."), # Marquesas ("Nuku Hiva", "Marquesas", "Nuku Hiva", 339, "WPI", "inhabited", "M", ""), ("Ua Pou", "Marquesas", "Ua Pou", 105.6, "WPI", "inhabited", "M", ""), ("Ua Huka", "Marquesas", "Ua Huka", 83.4, "WPI", "inhabited", "M", ""), ("Hiva Oa", "Marquesas", "Hiva Oa", 316, "WPI", "inhabited", "M", ""), ("Tahuata", "Marquesas", "Tahuata", 61, "WPI", "inhabited", "M", ""), ("Fatu Hiva", "Marquesas", "Fatu-Hiva", 85, "WPI", "inhabited", "M", ""), ("Mohotani", "Marquesas", "Mohotani", 15, "WPI", "inhabited", "V2", "'The smallest inhabited island in the archipelago'; abandoned after European contact."), ("Eiao", "Marquesas", "Eiao", 43.8, "WPI", "doubtful", "V2", "Adze-quarry island; 'intermittently uninhabited'."), # Australs and Rapa ("Rimatara", "Austral Islands", "Rimatara", 8.6, "WPI", "inhabited", "M", ""), ("Rurutu", "Austral Islands", "Rurutu", 32.7, "WPI", "inhabited", "M", ""), ("Tubuai", "Austral Islands", "Tubuai", 45.0, "WPI", "inhabited", "M", ""), ("Raivavae", "Austral Islands", "Raivavae", 17.9, "WPI", "inhabited", "M", ""), ("Rapa Iti", "Rapa Iti", "Rapa Iti", 40.5, "WPI", "inhabited", "M", ""), # Gambier ("Mangareva", "Mangareva (Gambier)", "Mangareva", 15.4, "WPI", "inhabited", "M", ""), ("Taravai", "Mangareva (Gambier)", "Taravai", 5.7, "WPI", "inhabited", "V2", "'Possible continuous settlement since the 13th century.'"), ("Akamaru", "Mangareva (Gambier)", "Akamaru Island", 2.1, "WPI", "inhabited", "M", ""), ("Aukena", "Mangareva (Gambier)", "Aukena", 1.35, "WPI", "inhabited", "V2", "'Inhabited since' the fourteenth century."), ("Temoe", "Mangareva (Gambier)", "Temoe", 2.1, "WPI", "inhabited", "V2", "'In 1838, Christian missionaries moved all the inhabitants of Temoe to Mangareva.'"), ("Rapa Nui", "Rapa Nui", "Easter Island", 163.6, "WPI", "inhabited", "M", ""), # Hawaii (areas converted from the infobox square miles) ("Hawai'i", "Hawaiian Islands", "Hawaiʻi (island)", 10432, "WPI", "inhabited", "M", ""), ("Maui", "Hawaiian Islands", "Maui", 1883, "WPI", "inhabited", "M", ""), ("O'ahu", "Hawaiian Islands", "Oʻahu", 1545, "WPI", "inhabited", "M", ""), ("Kaua'i", "Hawaiian Islands", "Kauaʻi", 1456, "WPI", "inhabited", "M", ""), ("Moloka'i", "Hawaiian Islands", "Molokaʻi", 673, "WPI", "inhabited", "M", ""), ("Lana'i", "Hawaiian Islands", "Lānaʻi", 364, "WPI", "inhabited", "M", ""), ("Ni'ihau", "Hawaiian Islands", "Niʻihau", 180, "WPI", "inhabited", "M", ""), ("Kaho'olawe", "Hawaiian Islands", "Kahoʻolawe", 115.5, "WPI", "doubtful", "V2", "Settled about 1000; 'from 1778 to the early 19th century, observers on passing ships reported that Kaho'olawe was uninhabited' (flagged citation needed)."), # New Zealand and Chathams ("North Island", "New Zealand", "North Island", 113729, "WPI", "inhabited", "M", ""), ("South Island", "New Zealand", "South Island", 150437, "WPI", "inhabited", "M", ""), ("Stewart Island", "New Zealand", "Stewart Island", 1747.7, "WPI", "inhabited", "M", ""), ("Great Barrier Island", "New Zealand", "Great Barrier Island", 285, "WPI", "inhabited", "M", ""), ("Chatham Island", "Chatham Islands", "Chatham Island", 920, "WPI", "inhabited", "M", ""), ("Pitt Island", "Chatham Islands", "Pitt Island", 65, "WPI", "inhabited", "V2", "'Originally inhabited by the Moriori.'"), # the 16 abandoned units at island grain ("Palmerston", "Palmerston", "Palmerston Island", 2.1, "U", "abandoned", "V", ""), ("Suwarrow", "Suwarrow", "Suwarrow", 0.4, "U", "abandoned", "V", ""), ("Pitcairn", "Pitcairn", "Pitcairn Island", 4.5, "U", "abandoned", "V", ""), ("Henderson", "Henderson", "Henderson Island (Pitcairn Islands)", 37.3, "U", "abandoned", "V", ""), ("Teraina", "Teraina", "Teraina", 9.6, "U", "abandoned", "V", ""), ("Tabuaeran", "Tabuaeran", "Tabuaeran", 33.7, "U", "abandoned", "V", ""), ("Kiritimati", "Kiritimati", "Kiritimati", 388, "U", "abandoned", "V", "Infobox now gives 312 km2; the unit-table value is kept."), ("Malden", "Malden", "Malden Island", 39.3, "U", "abandoned", "V", ""), ("Manra", "Manra", "Manra", 4.4, "U", "abandoned", "V2", ""), ("Orona", "Orona", "Orona", 3.9, "U", "abandoned", "V2", ""), ("Howland", "Howland", "Howland Island", 1.8, "U", "abandoned", "V", "Doubtful settlement."), ("Nihoa", "Nihoa", "Nīhoa", 0.7, "U", "abandoned", "V", ""), ("Necker", "Necker", "Necker Island (Hawaii)", 0.18, "U", "abandoned", "V", "Doubtful settlement."), ("Raoul", "Raoul (Kermadec)", "Raoul Island", 29.38, "WPI", "abandoned", "V", ""), ("Norfolk", "Norfolk", "Norfolk Island", 34.6, "WPI", "abandoned", "V", ""), ("Enderby", "Auckland Islands", "Enderby Island", 7.69, "WD", "abandoned", "V2", "Doubtful settlement. The site is on Enderby Island, not the main island."), ] PACIFIC_ISLAND_FIELDS = ["island", "unit", "wp_title", "land_km2", "area_source", "contact", "grade", "note"] # ---------------------------------------------------------------------------------------------- # 3. NORSE NORTH ATLANTIC # (unit, lat, lon, land_km2, empty_before, settle_lo, settle_hi, end_lo, end_hi, status, parent, grade, sources, note) # empty_before: 1 = no resident population when the Norse arrived, 0.5 = marginal or disputed, 0 = occupied # ---------------------------------------------------------------------------------------------- NORSE_UNITS = [ ("Western Norway (source)", 60.39, 5.32, None, 0, None, None, None, None, "source", None, "M", "-", "Origin region (Bergen as anchor)."), ("Shetland-Orkney", 60.36, -1.26, 2460, 0, 790, 850, None, None, "persisted", "Western Norway (source)", "M", "-", "Norse settlement about 800; islands already inhabited (Picts), so not an empty-land hop."), ("Faroe Islands", 61.97, -6.84, 1399, 0.5, 800, 900, None, None, "persisted", "Shetland-Orkney", "V", "L22", "Norse settlement evident in the 9th century; small-scale human presence 400-600 CE (Church et al. 2013 in L22)."), ("Iceland", 64.5, -19.0, 103000, 1, 830, 880, None, None, "persisted", "Faroe Islands", "V", "L22", "Landnam tephra 877 +/- 1 CE; pre-Landnam cultivation about 830-880; all districts settled within 60 years (L22)."), ("Greenland Eastern Settlement", 61.15, -45.52, None, 0.5, 985, 1000, 1430, 1500, "abandoned", "Iceland", "V", "L22;WS", "Founded about 985 CE (L22). Last written record 1408; latest radiocarbon about 1430; defunct by the late " "15th century (WS). About 560 farms (L22). L22: 'Unlike Iceland, Greenland was already occupied' by circum-polar " "hunter-gatherers, so coded 0.5 (farmland itself unoccupied, island not empty)."), ("Greenland Western Settlement", 64.18, -51.73, None, 0.5, 1000, 1050, 1340, 1370, "abandoned", "Greenland Eastern Settlement", "V", "L22;WS", "Permanent settlement from the early 11th century (L22); abandoned about 1350 (WS). About 75 farms (L22). " "Possible Norse-Dorset contact at Niaquusat (L22), so coded 0.5."), ("Vinland (L'Anse aux Meadows)", 51.595, -55.533, None, 0, 1000, 1021, 1021, 1035, "abandoned", "Greenland Eastern Settlement", "V", "K22;WS", "Wood felled in 1021 CE (K22). Occupation usually put at 3-10 years; 30-160 people; the land was " "already inhabited by Indigenous peoples (WS)."), ] NORSE_FIELDS = ["unit", "lat", "lon", "land_km2", "empty_before", "settle_lo", "settle_hi", "end_lo", "end_hi", "status", "parent", "grade", "sources", "note"] # ---------------------------------------------------------------------------------------------- # 4. MARITIME CULTURES (rebuilt in the revision). # Reference class, stated as a rule: a culture and period with (i) documented open-sea voyages out of sight of land and # (ii) uninhabited land that later carried a self-supporting population lying within about twice its longest documented # crossing. One row per episode; 'lineage' groups episodes that share ancestry and technology, and the lineage-level count # uses each lineage once. The rule was written AFTER the first-pass list existed, the list is not exhaustive, and the # number of capable cultures that did nothing is unknowable, so the counts are descriptive, not a rate with an interval. # (culture, lineage, period, capability, targets_available, outcome_strict, outcome_lenient, counted, grade, sources, note) # outcome: 1 = settled uninhabited land across open sea; 0 = did not; None = excluded under that coding # strict coding counts only waves that founded self-reproducing settlements on more than one landmass; # lenient coding also counts single-target colonisations. # counted: 1 = row enters the counts; 0 = row kept for the record but excluded (reason in the note). # ---------------------------------------------------------------------------------------------- MARITIME_CULTURES = [ ("Lapita (Bismarcks to Samoa)", "Austronesian", "c. 1050-800 BC", "sail; hops to about 850 km", "Remote Oceania", 1, 1, 1, "V", "L22", "Reached Tonga within 150 years of entering Remote Oceania, then stopped for almost 2000 years (L22)."), ("Western Micronesia from Island SE Asia", "Austronesian", "c. 1200-1000 BC", "sail; hop of about 2000 km", "Marianas, Palau", 1, 1, 1, "V", "L22", "Marianas about 1200 BC, Palau by about 1000 BC (L22)."), ("Central-eastern Micronesia", "Austronesian", "c. 50 BC-AD 150", "sail", "Pohnpei, Kosrae, Kiribati, Marshalls", 1, 1, 1, "V", "L22", "Colonised from the south by post-Lapita populations (L22)."), ("East Polynesians", "Austronesian", "c. AD 900-1300", "double canoe; hops of 2000-3800 km", "Polynesian triangle", 1, 1, 1, "V", "W11;I21;L22", "This study, Section 1."), ("Austronesians in the western Indian Ocean", "Austronesian", "1st millennium AD", "outrigger sail; ocean crossing", "Madagascar, Comoros", 0, 1, 1, "V2", "L22", "Madagascar settled in the first millennium AD (L22); no onward wave: the Mascarenes 700-900 km east stayed empty."), ("Norse", "Norse", "c. AD 800-1021", "clinker ships; hops of 300-1400 km", "Faroe, Iceland, Greenland", 1, 1, 1, "V", "L22;K22", "This study, Section 2."), ("Caribbean Ceramic Age colonists", "Caribbean", "c. 500 BC-AD 1000", "paddled canoes", "Antilles", 1, 1, 1, "V2", "L22", "Multi-island colonisation with long pauses; sail apparently absent before 1492 (L22)."), ("Mediterranean Neolithic farmers", "Mediterranean Neolithic", "c. 9000-5000 BC", "paddled craft; short hops", "Cyprus, Crete, and others", 1, 1, 1, "V2", "L22", "Short hops (under about 100 km); included as a low-barrier case."), ("Iberian Atlantic expansion", "Iberian", "AD 1420-1500", "caravel", "Madeira, Azores, Cape Verde, Sao Tome", 1, 1, 1, "V2", "WP2", "Madeira 'uninhabited at the time ... settled after 1420'; Azores 'discovered and populated in the early 1400s' (WP2). " "Start of the global European wave."), ("Canary Islands first settlers", "North African", "1st millennium BC-early AD", "unknown craft", "other Macaronesian islands", 0, 0, 1, "V2", "WS;WP2", "The Guanches were 'the only indigenous people known to have lived in the Macaronesian archipelago region' (WP2); no seafaring is " "recorded at European contact; Madeira (about 400-500 km) and the Azores stayed empty."), ("Ming treasure fleets", "Chinese", "AD 1405-1433", "ocean-going junks; about 300 ships, 27,000 men", "Indian Ocean islands", 0, 0, 1, "V2", "WS;WP2", "Seven state voyages between 1405 and 1433 (WP2), then withdrawal; no settlements founded."), ("Indian Ocean dhow traders (Arab, Persian, Swahili)", "Indian Ocean traders", "c. AD 800-1500", "monsoon sailing across open ocean", "Mascarenes, Seychelles", 0, 0, 1, "V2", "WP2", "Seychelles: 'Arab and Swahili sailors likely knew of the islands ... no evidence of permanent settlement before European " "involvement'; Mauritius possibly known to Arab sailors about 975 (WP2)."), ("Classical Mediterranean sea powers", "Classical Mediterranean", "c. 800 BC-AD 400", "sail and oar; Atlantic coasting", "Madeira, Azores", 0, 0, 1, "V2", "WP2", "Phoenicians, Carthaginians and Romans visited the (inhabited) Canaries (WP2); no settlement of empty Madeira. Whether Madeira lay " "within their demonstrated range is not established, so the row is a weak one."), ("Pacific-coast South American raft traders", "Andean", "c. AD 700-1500", "sailing balsa rafts", "Galapagos, Juan Fernandez", 0, 0, 1, "V2", "WP2", "'No firm evidence that ... the Indigenous peoples of South America reached' the Galapagos (WP2)."), ("Edo-period Japan", "Japanese", "AD 1600-1850", "coastal junks", "Bonin (Ogasawara) Islands", 0, 0, 1, "V2", "WP2", "Bonins 'entirely uninhabited' when reached, 'largely ignored by the ... isolationist Japanese'; first colonists came from " "Hawaii in 1830; Japan colonised them in 1875 (WP2)."), ("Malagasy", "Austronesian", "c. AD 1000-1600", "outrigger canoes", "Mascarenes", 0, None, 0, "M", "-", "NOT COUNTED (revision): the same people as the row 'Austronesians in the western Indian Ocean', which already records that " "the Mascarenes stayed empty; counting it again double-counts one lineage."), ("Aboriginal Australia and Tasmania", "Aboriginal Australian", "Holocene", "bark canoes and rafts", "Lord Howe, Norfolk, New Zealand", 0, None, 0, "M", "-", "NOT COUNTED (revision): open-sea capability is not documented, so the row fails condition (i); also not verified."), ("Post-contraction Polynesians", "Austronesian", "c. AD 1450-1770", "voyaging canoes (range shrinking)", "abandoned mystery islands, Lord Howe", 0, 0, 0, "V", "L16", "NOT COUNTED as a separate culture (revision): it is the end of the East Polynesian wave, discussed as such in the text. " "Long-distance voyaging contracted; abandoned islands were not resettled (L16)."), ] MARITIME_FIELDS = ["culture", "lineage", "period", "capability", "targets_available", "outcome_strict", "outcome_lenient", "counted", "grade", "sources", "note"] # Waves launched (strict coding) and how each one ended: used for the 'did the wave keep going' count. # (wave, stop, note) stop: 'targets exhausted' | 'range limit' | 'inhabited land' | 'retreat' WAVE_ENDINGS = [ ("Lapita", "range limit", "Stopped at Samoa-Tonga for 1,700-1,900 years with the Cook Islands about 1,150 km further on, beyond its longest " "documented hop (about 850 km) but within the range later achieved. Counted as 'filled' under the lenient reading and 'stalled' under the cautious one."), ("Western Micronesia", "targets exhausted", "Marianas and Palau; few further empty targets (L22)."), ("Central-eastern Micronesia", "targets exhausted", "(L22)."), ("East Polynesia", "targets exhausted", "This study: one later-habitable island in reach left empty; long-range voyaging then contracted."), ("Norse", "retreat", "Stopped at inhabited land (Vinland) and retreated from Greenland."), ("Caribbean Ceramic Age", "targets exhausted", "With long pauses (L22)."), ("Mediterranean Neolithic", "targets exhausted", "(L22)."), ("Iberian Atlantic", "targets exhausted", "Merged into the global European expansion."), ] # ---------------------------------------------------------------------------------------------- # 5. Published invasion-biology stage probabilities (literature values, used for comparison only) # (label, taxon, stage, value, lo, hi, grade, source, note) # ---------------------------------------------------------------------------------------------- INVASION_LITERATURE = [ ("Tens rule", "all", "each transition", 0.10, 0.05, 0.20, "M", "WF96", "Williamson & Fitter 1996: roughly 1 in 10 per transition (5-20%)."), ("Jeschke & Strayer 2005", "vertebrates, Europe <-> N America", "introduced -> established", 0.50, None, None, "V2", "JS05", "About 50% per step."), ("Jeschke & Strayer 2005", "vertebrates, Europe <-> N America", "established -> spread", 0.50, None, None, "V2", "JS05", "About 50% per step."), ("Jeschke & Pysek 2018", "plants and invertebrates", "each transition", 0.25, None, None, "V2", "JP18", "Meta-analysis of 102 tests in 65 publications; no support for the tens rule."), ("Jeschke & Pysek 2018", "vertebrates", "each transition", 0.50, None, None, "V2", "JP18", "Same."), ("Moulton & Cropper 2019", "birds, 2-10 released", "released -> established", 0.28, None, None, "V", "MC19", "Abstract; 227 releases."), ("Moulton & Cropper 2019", "birds, 11-100 released", "released -> established", 0.34, None, None, "V", "MC19", "Abstract."), ] INVASION_FIELDS = ["label", "taxon", "stage", "value", "lo", "hi", "grade", "source", "note"] SOURCES = { "W11": "Wilmshurst, Hunt, Lipo & Anderson 2011, PNAS 108, 1815 (text ranges; supplementary table not obtainable from this machine)", "I21": "Ioannidis et al. 2021, Nature 597, 522 (author manuscript; Extended Data Table 1 and Discussion)", "L22": "Leppard, Cochrane, Gaffney, Hofman, Laffoon, Bunbury & Broodbank 2022, J. World Prehistory 35, 163", "L16": "Leppard 2016, J. Pacific Archaeology 7(2), 20 (canonical list of 13 mystery islands, after Kirch 1988)", "AW01": "Anderson & White 2001, Records of the Australian Museum Suppl. 27, 135", "A05": "Anderson 2005, Antiquity 79, 791 (abstract)", "A24": "Anderson et al. 2024, Archaeology in Oceania (abstract)", "B12": "Burley, Weisler & Zhao 2012, PLoS ONE 7, e48769", "K22": "Kuitems et al. 2022, Nature 601, 388", "WP": "English Wikipedia article for the island, retrieved 2026-10-12 (secondary)", "WP2": "English Wikipedia article for the island or topic, raw wikitext retrieved 2026-10-12 during the revision (secondary; " "page list and parsed infobox values in data/island_infobox_cache.json)", "C14": "Cartin 2014, JBIS 67, 119 (arXiv:1404.0204)", "B15": "Bell, Currie, Irwin & Bradbury 2015, American Antiquity 80, 397", "BL20": "Blackburn, Cassey, Lockwood & Duncan 2020, PeerJ 8, e8766", "WS": "Web summaries retrieved 2026-10-12 (Wikipedia, Te Ara, Landcare Research; secondary)", "WF96": "Williamson & Fitter 1996, Ecology 77, 1661", "JS05": "Jeschke & Strayer 2005, PNAS 102, 7198", "JP18": "Jeschke & Pysek 2018, in Invasion Biology: Hypotheses and Evidence (CABI), ch. 13", "MC19": "Moulton & Cropper 2019, PeerJ 7, e7637 (data file S1: 832 introduction records from Sol et al. 2012)", } # ---------------------------------------------------------------------------------------------- # 6. GENETIC SETTLEMENT CHAIN (Ioannidis et al. 2021, Extended Data Table 1 and Discussion text). # Point dates from identity-by-descent tract lengths (generations before 1989, about 29-30 yr per generation); # the paper gives no interval per island in the text read here, so intervals are NOT propagated. # (parent_island, child_island, parent_unit, child_unit, parent_date_CE, child_date_CE, link_stated_in_text) # ---------------------------------------------------------------------------------------------- GENETIC_CHAIN = [ ("Rarotonga", "Tahiti", "Southern Cooks", "Society Islands", 830, 1050, "fanned out from Rarotonga northeast to the Society Islands around 1050 CE"), ("Tahiti", "Mataiva (Palliser group)", "Society Islands", "Tuamotus", 1050, 1110, "thence northeast to the Tuamotu Archipelago by 1110 CE"), ("Mataiva (Palliser group)", "Fatu Hiva", "Tuamotus", "Marquesas", 1110, 1140, "from this hub north to the Marquesas; 1140 CE on Fatu Hiva"), ("Mataiva (Palliser group)", "Mangareva", "Tuamotus", "Mangareva (Gambier)", 1110, 1130, "south to the Gambier Islands; 1130 CE (Extended Data Table 1)"), ("Mangareva", "Rapa Nui", "Mangareva (Gambier)", "Rapa Nui", 1130, 1210, "from Mangareva the expansion reached Rapa Nui around 1210 CE"), ("Rarotonga", "Rapa Iti", "Southern Cooks", "Rapa Iti", 830, 1190, "from this base settlers continued south around 1190 CE to Rapa Iti"), ] GENETIC_FIELDS = ["parent_island", "child_island", "parent_unit", "child_unit", "parent_date_CE", "child_date_CE", "link_stated_in_text"] def write_all(outdir): import csv, os os.makedirs(outdir, exist_ok=True) def w(name, fields, rows): with open(os.path.join(outdir, name), "w", newline="", encoding="utf-8") as f: cw = csv.writer(f) cw.writerow(fields) for r in rows: assert len(r) == len(fields), (name, r) cw.writerow(["" if x is None else x for x in r]) w("pacific_anchor_islands.csv", ["island", "unit", "geo_key", "manual_lat", "manual_lon"], PACIFIC_ANCHORS) ex_fields = PACIFIC_UNIT_EXTRA_FIELDS[1:] rows = [] for r in PACIFIC_UNITS: ex = PACIFIC_UNIT_EXTRA.get(r[0]) if ex is None: # persisted units: inhabited continuously, so habitable under both later-habitability codes ex = (None, 1, 1, r[13], "") if r[4] == "persisted" else (None, 0, 0, "M", "") rows.append(tuple(r) + tuple(ex)) assert set(PACIFIC_UNIT_EXTRA) <= {r[0] for r in PACIFIC_UNITS} w("pacific_units.csv", PACIFIC_UNIT_FIELDS + ex_fields, rows) w("pacific_islands.csv", PACIFIC_ISLAND_FIELDS, PACIFIC_ISLANDS) w("wave_endings.csv", ["wave", "stop", "note"], WAVE_ENDINGS) w("norse_units.csv", NORSE_FIELDS, NORSE_UNITS) w("maritime_cultures.csv", MARITIME_FIELDS, MARITIME_CULTURES) w("invasion_literature_values.csv", INVASION_FIELDS, INVASION_LITERATURE) w("source_keys.csv", ["key", "reference"], sorted(SOURCES.items())) w("genetic_settlement_chain_ioannidis2021.csv", GENETIC_FIELDS, GENETIC_CHAIN) if __name__ == "__main__": import os write_all(os.path.join(os.path.dirname(os.path.abspath(__file__)), "..", "data", "curated"))