Indigenous · History Classes · Grades 10–13 · College · University

The Extermination Census — The Continuations (2026-10-28)

History Class · Level band: Grades 10–13 · College · University

English edition — the language editions of the history classes are outstanding and will follow the site’s translation pattern; none is fabricated here.

The census’s three dated continuation sections

Class: RESEARCH. This class reproduces, in student-reading form, the three continuation sections the Extermination Census received on 2026-10-28: §2.5, the United States and Canada long tail; §4.3, the Caribbean chroniclers read at first hand; and §5.5, the South American catalogue pass, Arawakan first. The census counts by the three missings — a people enters Class A only when people, language and culture are all verified gone; anything less files Class B (ended in the record, descendants continue) or Class C (language gone, people living — never counted), or stays out. Its subtotals are floors, not totals, and its dated statements of what remains owed stand at the close of each section, unaltered.

2.5 The long tail — the 2026-10-28 pass

The §2.4 debt was taken up on 2026-10-28: the tasked candidates from the Florida, Texas and California mission rolls, the Southeast / Gulf / Pacific merger record, the BAE tribal lists and the Canadian record, worked nation by nation with the descendant check run on every line. The full working file, with its per-line audit trail, stands at research/extinction/_tail_working/census-tail-us-canada.md. What it files:

Class A — one new line:

Grouped because the record of their ending is a shared record; counted severally (3), on the §2.1 Florida precedent. No other candidate in this pass met the three-missing test — every other verified candidate failed it at the descendant check, which is the Selk'nam / Karankawa lesson working as designed.

Class B — 23 new counted lines (35 named peoples):

Two entries carried outside the count, in §2.2's manner:

• The Cusabo group — the Lowcountry peoples of the South Carolina coast (Ashepoo, Combahee, Coosa, Edisto, Escamacu, Etiwan, Kiawah, Stono, Wando, Wimbee, and the Sewee at the group's edge), broken by the slave trade (thousands removed, 1670–1715), epidemics and war, no longer existing as tribes by 1750, the survivors most likely merging with the Catawba or the Creek. Outside the counted subtotal because the group's internal count is itself unsettled — an uncertain count is not improved by being added (Waddell line, via the South Carolina ethnohistory record).
• Esselen — CONFLICT, printed whole — the older record declares the Esselen gone: Kroeber pronounced the group "as good as gone" (1925). The correction stands in the same record: Kroeber himself recanted in 1955; Esselen descendants persisted, and today some 460 people identifying as Esselen descendants stand organised as the Esselen Tribe of Monterey County. Filed B on the descendant community's standing, on the Karankawa precedent, and carried with the conflict lines outside the counted subtotal until the owner rules on counting conflict lines.

Class C additions (§2.3's manner — named only to keep them out of the count): Tonkawa — the test resolves C plainly: the language ended (last first-language speakers d. in the 1940s on the standard record; a recent film record names a last fluent speaker still living — the two standings printed, not reconciled), and the people stand as the federally recognised Tonkawa Tribe of Indians of Oklahoma, 700+ citizens. Salinan — declared extinct by the 1930 census; the people survived (681 in the 2000 US census), stand organised as the Salinan Tribe, and are reconstructing a language spoken until the 1950s. Wappo (Mishewal) — last fluent speaker Laura Fish Somersal d. 1990; the people stand organised in the Mishewal Wappo Tribe of Alexander Valley. Yuki — last native speaker Arthur Anderson d. 1983; the people stand inside the Round Valley Indian Tribes, where Yuki language teaching is under way — the Cayuse / Takelma / Molala pattern of §2.3 exactly. And a summary confirmation, severing none: the peoples whose languages ended inside the Grand Ronde, Siletz, Round Valley, Colville and Warm Springs confederations, and the California mission-descendant communities in the recognised rancheria tribes, stand as peoples inside those nations — Class C facts, from which no A or B line is constructed.

Not carried in this pass (working file §5 gives each reason in full): Suma — the descendant check cannot be run on the record reachable (generic absorption only); Grigra — severing it would risk double-counting the filed Natchez dispersal; Pohoy — attestation thinner than its filed Tampa Bay neighbours', held for a first-hand Hann pass; Yscani — probably the Waco under an earlier name (the seam is printed on the Waco line); the Timucua sub-chiefdoms severally — the Timucua precedent; Coast Yuki / Huchnom as separate lines — dialects, not peoples; any California mission-village name — mission rolls name villages, not nations, the Coahuiltecan counting problem in Californian form.

Subtotal delta, §2: Class A 11 → 14 named peoples (8 → 9 lines) · Class B 14 → 49 named peoples (7 → 30 counted lines), with the conflict and grouped lines outside the count now standing at six (Karankawa, Jumano, Esselen; St. Lawrence Iroquoians, Coahuiltecan nations, Cusabo group) · Class C still not a count.

What §2 still does not attempt (2026-10-28 statement, replacing §2.4's open position for the peoples worked): the rest of the BAE tribal list nation by nation (among others: the remaining Carolina piedmont and coastal peoples severally outside the Cusabo group and the Catawba coalescence; the Virginia Siouan mergers beyond Tutelo — Saponi and Occaneechi severally, Manahoac; the Illinois-confederacy peoples severally — Kaskaskia, Peoria proper, Cahokia, Tamaroa, Michigamea); the California mission rolls by mission and the non-mission interior peoples beyond Chimariko, Wappo and Yuki; the Texas mission rolls beyond the Atakapan and Wichita lines; Canada beyond the three lines worked here; and Florida at first hand (Pohoy and the remaining Hann-record small peoples). The honest §2 position is now: Class A floor 14 · Class B floor 49 · the true figure still higher, by an amount this pass does not guess.

4.3 The chroniclers at first hand — the 2026-10-28 audit

Commissioned by §10(iv) and worked 2026-10-28: the six §4.1 lines re-examined against the early chroniclers read directly in public-domain texts — Las Casas, Historia de las Indias Tomos I, IV, V (Madrid 1875–76; Project Gutenberg #49298, #56283, #53131) and the Brevísima relación (1552) in the 1689 English translation (#20321, portions read); Columbus's 1493 letter to Santángel; the Diario of 13 January 1493 as preserved by Las Casas in Historia I ch. LXVII (the only form in which the Diario survives); Ralegh, The Discoverie of the large, rich and beautiful Empire of Guiana (1596), read in full. Pané, Chanca, Federmann, Velázquez 1514 and the Memorial de remedios are quoted through identified scholarly transmissions, flagged as such in the audit, because no standalone PD full texts of Pané or Chanca could be opened. The full audit, with every quotation and every negative search named, stands at _tail_working/caribbean-chroniclers-audit.md (5,408 words). No §4 class changed. What changed is the stated basis of four lines, and the strengthening of two — the audit's proposed evidence-notes are carried on the §4.1 lines themselves, dated.

• Macorix — CONFIRMED, the strongest line in §4. Doubly attested in the 1490s generation. Pané (c. 1498), sent to live among Hispaniola's peoples and learn their languages, records that Columbus told him the province of Macorís "had a language distinct from the other one," and that he answered, in effect, "the only language I know is that of Macorix" — the first European ethnographer of the Americas learned Macorix first, as a language of its own. Las Casas, Historia I ch. LXVII: the coast "era poblada de una gente que se llamaban mazoriges, y otras cyguayos, y tenian diversas lenguas de la universal de toda la isla"; Apologética ch. CXCVII: "Tres lenguas habia en esta Isla distintas, que la una á la otra no se entendia" — Macoríx de Abajo, Macoríx de Arriba, and the universal language. The name is a Taíno exonym ("Macorix quiere decir como lenguaje extraño, cuasi bárbaro" — 'strange speech, almost barbarous'), and the surviving lexicon is one word, baeza "no" (cited from the full Apologética; not located in the 1875–76 excerpt searched — the audit does not pretend otherwise). The extinction is attested by Las Casas in his own person: he had "conversé hartas veces con ambas generaciones," and writing 50+ years after 1493, "no hay hoy uno ni ninguno á quien lo preguntar" — there is not today one, nor anyone, to ask.
• Ciguayo — CONFIRMED, with one qualification carried. The Ciguayo enter the record in the first recorded fight of the conquest: the Diario, 13 January 1493, Gulf of Arrows (Samaná) — "Y esta fué la primera pelea que hobo en todas las Indias" — and in the same entry the language distinction is proven lexically: "Aquí no llaman caona al oro como en la primera parte desta isla, ni nozay como en la isleta de Guanahaní ó Sant Salvador, sino tuob." The name is again a neighbours' label ("se llamaban cyguayos porque traian todos los cabellos muy luengos" — for their long hair); the Apologética (ch. III) gives them a province of more than thirty leagues; the Brevísima knows their province as Guarionex's refuge a generation later. The qualification, printed plainly: in the Apologética's three-languages passage Las Casas counts the two Macoríx speeches and the universal language and does not there name a fourth; he adds elsewhere that he cannot now remember whether Ciguayo speech differed from Macoríx. The chronicler evidence proves a named people, a named province, a speech different from the island's general language, and its disappearance within living memory; whether Ciguayo speech was a language apart from Macoríx or its eastern kin is a question the chronicler himself left open — the census inherits his openness, not a modern certainty.
• Yaio — CONFIRMED for the Ralegh era; NOT a first-chronicler attestation. For the Spanish chroniclers (1492–1552), nothing: the name is absent from every text searched. The first-hand attestation is English and a century later, and it is genuine — Ralegh, Discoverie (1596), from his own 1595 presence on Trinidad: "Those about Parico are called Jajo." The language evidence is de Laet's single wordlist (1640 ed.), already containing a Spanish loan (pero "dog"), which itself dates the contact world the list came from. The line now carries its attestation date (1595/96).
• Caquetío — QUALIFIED. Chronicler-era attested, but on the mainland and by mainland writers: Federmann, who marched the Coro–Barquisimeto country in 1530–31, names the Caquetíos of "Variquicimeto" in his Indianische Historia (1557) as a numerous, trading people; Oviedo and Castellanos (1589) likewise name Caquetíos on the Venezuelan mainland. The name occurs in none of the Antillean chroniclers searched. What no chronicler-era text found does is tie the ethnonym to Aruba, Curaçao and Bonaire: the ABC link rests on the 1513 islas inútiles declaration, the 1515 removal of some 2,000 islanders to Hispaniola's mines (administrative documents, summarised in Haviser 1991), the 17th-century wordlists, and the modern classification (Granberry & Vescelius). "Named at contact" is true of the mainland (Federmann was there in 1530) and unproven, at first hand, for the islands — the line now says exactly that.
• Guanahatabey — NOT IN THE CHRONICLERS at first hand; basis restated. Not a later invention — the name is genuinely early — but its two earliest attestations are both second-hand reports of what Taíno informants said about a people no Spaniard in the record had met. Velázquez's 1514 report: "Guanahatabibes, que son los postreros indios della… no tienen casas ni pueblo, ni labranzas, ni comen otra cosa sino las carnes que toman por los montes y tortugas y pescados." Las Casas, Memorial de remedios (1516) — a reform memorial, not his histories: people "como salvajes… ni tienen casas, sino están en cuevas continuo… se llaman Guanahatabeyes." Against that: the name was searched for and not found in Historia Tomos I or IV, nor in the Brevísima; Las Casas, who lived at Canareo in western Cuba 1513–1515, never claimed to have seen these people. No word of any Guanahatabey language is recorded anywhere — no vocabulary, not even a single gloss — and no encomienda, epidemic or war is documented against them by name. The archaeological "Guanahatabey" (a pre-ceramic western-Cuba population equated with the hearsay cave-dwellers) is a 20th-century construction whose equation with the name Keegan (1989) disputed and Granberry & Vescelius defended; that dispute cannot be settled from the chronicles, because the chronicles contain nothing further to settle it with. The line's basis is now stated as exactly this.
• Shebayo — NOT IN THE FIRST CHRONICLERS, and not in Ralegh either. The negative here can be made unusually precise, because the text to which the attestation is usually credited was read in full. Ralegh's Discoverie lists Trinidad's peoples in his own words — "Those about Parico are called Jajo, those at Punta de Carao are of the Arwacas (Arawaks) and between Carao and Curiapan they are called Salvajos. Between Carao and Punta de Galera are the Nepojos, and those about the Spanish city term themselves Carinepagotes (Carib-people)" — and "Shebaio," in any spelling, does not occur. The standard attribution (Boomert 2000) resolves, on the texts read, to Sparrey's companion narrative of the 1595 voyage (manuscript until 1906; not itself opened in this audit — flagged) and to the surviving ~17-gloss vocabulary in de Laet (1625; wordlists in the 1633 and 1640 editions). The documented fate — flight from Trinidad's south coast to Tobago and the Guiana coast after the 1595 attack, per Boomert — is likewise a late-16th/17th-century record. The line's basis is now stated as late-16th- and 17th-century English and Dutch records, dropping any implication of Ralegh's own text.

The five kingdoms stand inside the held complex — nothing severed. The audit tested whether any severally-attested Hispaniola people exists outside the held Taíno complex besides Macorix and Ciguayo. The chroniclers name five polities again and again — Maguá, Maguana, Marién, Jaragua (Xaraguá) and Higüey — but Las Casas himself gives the reason they cannot be severed: they are kingdoms of the universal-language population. His Apologética counts exactly three Hispaniola languages, and treats Xaraguá's speech not as another language but as the universal language at its most refined (the Brevísima praises Xaraqua's "fluency of Speech and politeness of Idiom" above the island's other kingdoms). Political division, dialect prestige, distinct dynasties: all attested. Distinct peoples with distinct languages: not attested, and contradicted by the chronicler's own linguistic census. Maguá, Maguana, Marién, Jaragua and Higüey therefore stand inside the held Taíno complex (§4.2), are not severed, and no line for any of them is filed. (Oviedo's hearsay of cave-dwelling people in Guacayarima, south-western Hispaniola — unnamed, of the same species as the Guanahatabey reports — is likewise no basis for a line.)

The Lucayan deportations: confirmed at first hand, in mechanism and in outcome. The Brevísima (1689 English): the Lucayan isles "is now laid waste and uninhabited… they are now cut off, some by slaughter, and others ravished away by Force and Violence, to work in the Mines of Hispaniola, which was destitute of Native Inhabitants" — with only eleven persons found, "which I saw with my own Eyes" (the figure is rhetoric — see below; the deportation is not). The Historia (Tomo IV) supplies the machinery: once Hispaniola's own people were failing, the Spaniards "inventaron hacer armadas para saltear los que llamaban Lucayos" — invented the making of fleets to raid those they called Lucayos — and narrates the same fleets working the Guanajes: shiploads of captives, some five hundred then some four hundred souls, carried to Havana, with captives twice retaking their ships. Deportation as an organised, repeated, named practice, attested by the principal chronicler in two works: the texts support carrying the Lucayan deportations at full strength as mechanism and event, while carrying none of their numbers. The Lesser Antilles, at the texts' level: no chronicler read gives any Lesser Antillean people a name of its own, a language note, or a polity — what the record holds is island names and the exonym complex Caniba / Caribe / Carib. The texts support sustained raiding relations across the island arc and Spanish slaving pressure from the first decade; they do not support naming, counting or severally filing any Lesser Antillean people from the chroniclers — and this census does not.

The two corpus exclusions, tested at first hand — both stand. (i) Las Casas's population figures are never counts. Read in his own pages they present themselves as moral magnitudes and do not agree with each other: the Brevísima's Argument holds "above Twelve Millions" perished and, in the same paragraph, "above Fifty Millions in all"; Hispaniola is "of Three Millions of Persons… there is at present but the inconsiderable remnant of scarce Three Hundred"; the Lucayos hold "about Five Hundred Thousand Men" reduced to eleven eyewitnessed survivors. Round millions set against round hundreds, by a writer whose stated purpose is prosecution, with no counting instrument behind any of them — the age's actual administrative count for Hispaniola (the 1514 repartimiento, roughly 26,000–32,000) is a different species of number. Las Casas is a witness to mechanisms, names, languages and events; his figures are testimony about the scale of catastrophe as he wanted it felt. (ii) The "peaceful Arawak / cannibal Carib" binary is the chroniclers' own construction — visible in formation, and Las Casas is its first dismantler. Columbus (1493) reports the cannibal island purely as others' report and half-dissents in the same breath ("I do not consider them in any other way superior to them"). Chanca (1494), on Guadeloupe, shows the identification made from the expectation: the landing party found "four or five bones of the arms and legs of men. When we saw this, we suspected that the islands were those islands of Caribe… For the admiral… had directed his course to discover them" on Taíno indications. Las Casas then takes the mechanism apart explicitly: Columbus suspected the Ciguayo man of being a Carib, "pero no era, porque nunca en esta isla jamás los hobo" — there never were any in this island — and generally, "nunca fueron ni fué comedores de carne humana"; peoples were judged Caribes because they resisted and defended themselves, and the Apologética acquits the whole island on eating human flesh: "fueron limpísimos y exentísimos." A classification its own earliest sources show being applied to people who fought back cannot be carried as the chroniclers' settled finding; the corpus's refusal is now grounded in the texts themselves.

What the audit leaves owed, stated exactly: Sparrey's 1595-voyage companion narrative itself (printed 1906) — the load-bearing Shebayo attestation — was not opened; de Laet's own wordlist pages (1625/1633/1640) were not opened (read via Taylor 1977, Granberry & Vescelius, and Part II of how-they-were-exterminated.md); Pané and Chanca stand read through identified scholarly transmissions (Bourne 1906 via Granberry/Scheper; Hulme & Whitehead 1992), no standalone PD full texts having been opened; Martyr's Eden 1555 translation was located (Oxford Text Archive A20032) but not opened — Martyr carries no load-bearing claim in this audit; absence claims for the Brevísima are scoped to the portions read in the 1689 translation; the Macorix gloss baeza was not located in the 1875–76 Historia volumes searched (I, IV, V), whose Apologética is an excerpt. A future pass that opens Sparrey and de Laet at first hand would close the last transmission gaps in §4's bases; nothing in the audit's verdicts depends on them beyond what is flagged.

5.5 The Arawakan catalogue pass — 2026-10-28

The §10 debt (iii) was taken up family by family, Arawakan first, on 2026-10-28. The full working file stands at research/extinction/_tail_working/census-tail-arawakan.md. The frame it worked against is §1's and §5's: ~150 Arawakan languages attested, at least 100 extinct as languages (Aikhenvald 2024, from Loukotka 1968) — a Class D magnitude, not a nation count. The pass ran David's three-missing test, people first, against each attested name the modern classifications carry (Ramirez 2001; 2019/2020 — 56 languages, 27 marked extinct; Jolkesky 2016; Kaufman 1994 via Loukotka 1968).

The finding of this pass, stated first: ZERO new Class A lines for the Arawakan family — and the zero is a finding, not an omission. It is §3.1's Mexico finding repeated at family scale, and stronger: wherever the descendant check could actually be run on an Arawakan name, it came back positive. The Arawakan pattern of loss is absorption — peoples broken on the main rivers (the record's own explanation: Arawakan communities sat on the great waterways and were therefore "the primary targets of epidemics, slave raids, relocations and warfare" — Philosophical Transactions B survey of Upper Rio Negro diversity, 2022) merged into neighbouring peoples, into mission and caboclo / tapuia populations, and into the mestizo riverine world, leaving descendants everywhere and descendant polities almost nowhere a census can verify name by name. The languages died at a scale unmatched in the hemisphere; the peoples, name by name, overwhelmingly did not — they continued under other names, which is precisely the paperwork fiction this census exists not to repeat. To file Class A lines here on language-extinction lists alone would have manufactured dead nations out of living descendants.

Class B — 9 new counted lines (10 named peoples):

Grouped, outside the subtotal — the Rio Negro / Japurá wordlist peoples: severally named: Yumana (Jumana) · Wainumá · Mariaté · Kauixana (Kayuixana) · Kariai (Cariaí) · Guinao (Guinau) · Bahuana · Yabaana · Anauyá. Attested as peoples of the Negro–Japurá–Içá interfluves and the middle Negro, most from wordlists alone (Spix & Martius 1831; Natterer); their languages stand extinct in every classification, and the regional survey names the mechanism — main-river exposure to epidemics, slave raids, relocations and warfare. But for none of the nine could this pass find a descendant record in either direction: no descendant community stands, and no source states where their people went. A Class A filing requires the people-missing test to be verified, not merely unrefuted; a Class B filing requires descendants to be located, not presumed. They are carried grouped and outside the count, in §2.2's St. Lawrence / Coahuiltecan manner — the largest single block of the Arawakan tail this pass could neither file nor honestly drop.

Class C guard additions (counted nowhere): Baré — the family's textbook Class C: the language's last fluent speaker, Candelário da Silva, d. 2 February 1993, while 16,516 Baré people stood in the 2011–2014 counts in Venezuela and Brazil. Añu (Paraujano) — 21,000 people (2011 census); the language down to one fluent speaker in 2014, with partial speakers and a school revival — a dormant-and-waking line inside a standing people. Kinikinau (Kinikinao) — the guard case of this pass: the classic literature (Cardoso de Oliveira; Ribeiro's list) carried the Kinikinau among the groups that "disappeared," filed extinct in the mid-20th century; ISA's record answers it — some 58 people declared themselves Kinikinau in 1998, about 180 in 2003 and about 250 in 2005, many having earlier been afraid to identify themselves as Kinikinau. A standing warning against every "disappeared" verdict in the older literature. Chamicuro — some 100 people at Pampa Hermosa; the language's fluent line reduced to a handful (8 speakers reported 2008; Ramirez 2020 marks the language extinct — the conflict printed, the people not in dispute). Siusi (Siuci) is removed from candidacy altogether: not a separate ended people but a living phratry of the Baniwa-Kurripako continuum (Aikhenvald). And four names file nowhere at all — not even C — because their languages are not yet gone and their peoples stand: Baure (c. 20 speakers, 2010; revival under way), Paunaka (8 speakers, 2023), Resígaro (one speaker living, in Peru's official BDPI record, within a self-identified Resígaro community, 2017; ethnic Resígaros stand inside Bora and Ocaina communities — the line files only when the speaker line itself ends), and Iñapari (four to five speakers living; catalogues that filed Iñapari extinct — Aikhenvald 1999 and Payne 1991 among them — were wrong while its speakers lived, and this census does not repeat them).

Not carried in this pass (working file §5C gives each reason in full): Igneri / Eyeri — a pre-contact (c. 1450) fusion into the Kalinago world with the language continuing as Iñeri: a §8-type exclusion, and severing it would breach the owner's hold on the Taíno complex; Tegua — unattested, affiliation uncertain, heritage custodianship active; Guayupe — the record verifies neither an ending nor a descendant line; Sae — a name only, a §8 absence; Yavitero — the language ending is the best-documented in this file (last speaker d. 4 January 1984) but no descendant evidence exists in either direction, and a language-extinction date is Class D evidence, not a filing; Marawa · Waraiku and Kuniba · Kanamaré — units unsettled (Kanamaré possibly a dialect of the living Iñapari); Ribeiro's recorded-extinctions list as a class — the same list wrongly files the living Warekena as extinct, and a list demonstrably wrong on a checkable name cannot carry unverifiable names at this standard; the living Chané, Xingu Arawak and Campa-branch peoples — not filed, by the record.

Subtotal delta, §5: Class A 2 → 2 named peoples (delta +0 — a finding) · Class B 8 → 18 named peoples (5 → 14 counted lines), plus the Rio Negro / Japurá grouped entry (9 named peoples) carried outside the count · Class D untouched: the Arawakan ~150 attested / ≥100 language-extinct magnitude still overlaps every line above and must never be added to them.

The Arawakan tail that remains, stated exactly: the grouped nine above, filed nowhere until a descendant record surfaces for at least one name; Loukotka's name-only tail beyond Ramirez's 56 (Amarizana, Guarú, Mepuri, Mandawaka, Mainatari, Wirina, the Trinidad names, Marawan, Paikoneka, Cararí, Morique, the Atorai / Parawana / Aroaqui absorptions inside Wapishana, and their neighbours), each still owed the same people-first run; the Guana-Chané unit questions; the Venezuelan llanos and coast names (Ajagua, Onota, Alile and their neighbours) beyond Añu and Caquetío; and a Glottolog-by-Glottolog reconciliation of the ~150-name inventory, used but not exhausted by this pass. Beyond Arawakan, the non-Arawakan catalogue pass stands owed whole — Tupian, Cariban, Panoan, Tucanoan, Jê, Arawan, Chapacuran, Nambikwaran, Witotoan, the Andean and Chaco fringes, and the isolates — and this pass's central finding (absorption, not verified extermination, wherever the descendant check could be run) is a hypothesis the remaining families must test, not a conclusion they inherit: the Tupian record, where the Juma already filed actually did end as a separate nation inside living memory, will not necessarily answer the same way. Until that work is done, §5's Class A floor of 2 against a Class D field of ~1,100 remains the single largest honest gap in this file.

Back to the Grades 10–13 · College · University classes All History Classes