The History of the World

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

The Code Talkers: The Languages That Won Battles

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

DRAFT — CULTURAL REVIEW PENDING. These history classes are a publisher’s draft blend of the Collection’s Indigenous history research, standing on its own website. Indigenous cultural review precedes any public release of this strand. Readings carry the research labels unchanged: Documented Fact · Tradition — the people’s own account · Debated / Claim. This page carries no advertising.

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

For the teacher — teacher-facing guidance for this class stands in this box only; the reading below is the student reading. This is the full entry; a Grades 7–9 version carrying the service and the ingenuity stands in the middle band, with the suppression record whole reserved to this band. Two places need a teacher's care. First, the Korea and Vietnam section corrects a widely repeated claim in the open — no formal code-talker programme is documented there — and it is part of the lesson that the class hold the documented service (which was enormous) apart from the undocumented programme (which is not in the record). Second, the suppression section names the schools systems; it cross-refers to the Genocide Record class rather than repeating it, and discussion should keep the chronology exact: the punishment of the languages ran before, during and after the wars, not after them only.

The Code Talkers — The Languages That Won Battles

Status: Researched draft — cultural review pending

How to read this entry: Documented Fact = the fact stands in a named source — government and museum records, the nations' own published histories, regimental histories, the men's testimony recorded in their lifetimes. Tradition — the people's own account = what the nations and the men's families teach about them. Debated / Claim = a point the sources dispute, or a claim carried in popular retelling that the record does not hold; both are labelled where they stand. The men's own testimony is treated as a witness family in its own right. Numbers the sources give differently are carried as ranges, flagged, never averaged into a false precision. No restricted or ceremonial knowledge is recorded here; nothing in this entry touches it.

1. How the codes worked

Documented Fact. A code talker was a serviceman who transmitted military messages by radio or field telephone in his own Indigenous language. The United States military's own classification, preserved in the Oklahoma Historical Society's encyclopedia, distinguishes two kinds. In Type Two work, the language itself was the code: ordinary conversation in a language the enemy had never studied, with no dictionary, no grammar in enemy hands, and almost no speakers outside the nation itself. In Type One work, a formal code was built inside the language: chosen words of the language stood for military terms the language had never needed — for aircraft, ships, weapons, ranks, places — and a set of words stood for the letters of the alphabet, so any English word or name could be spelled. The Type One talker therefore spoke a code within a language: an enemy who somehow found a speaker of the language still could not read the traffic, because the traffic was not plain speech in it. The codes were never written into field manuals the men carried; they were memorised, and they lived in the men.

Documented Fact. The advantage was also speed. A message that took a cipher machine roughly thirty minutes to encode, send and decode took trained talkers about twenty seconds end to end — the comparison carried in the Smithsonian's and the National Museum of the American Indian's accounts of the Navajo programme. And the documented verdict of the Second World War is that the Navajo code — the largest of the programmes — was the only spoken military code of that war never deciphered by the enemy; Japan's own postwar intelligence chief, Lieutenant General Seizo Arisue, acknowledged after the war that Japanese intelligence had broken American Army and Air Corps codes but never the Marines' — the talkers' (as carried in Marine Corps histories).

2. The first: the Choctaw Telephone Squad, 1918

Documented Fact. The idea was born in the First World War, in the Choctaw language, in the last autumn of the war. The 36th Infantry Division — a National Guard division of Texans and Oklahomans — was fighting in the Meuse-Argonne offensive in October 1918. German forces had broken the American codes, were tapping the telephone lines, and were capturing, by the regimental account, one in four of the runners who carried messages the wires could no longer safely carry. An American officer overheard two Choctaw soldiers talking in their own language and understood what it could do. A field test followed: a Choctaw soldier at one end of a telephone line, another at the far end, a message passed in Choctaw and written out in English at the receiving end. The Oklahoma Historical Society's account and the 36th Division's regimental record carry what followed: "Telephone Squads" of Choctaw soldiers were posted through the division's regiments — the 141st, 142nd and 143rd Infantry — and began transmitting orders, troop positions and plans in Choctaw.

Documented Fact. Choctaw had no words for the machine age of that war, so the men made the code inside the language — the first Type One work. A machine gun became, in Choctaw, "little gun shoot fast." The squads' transmissions are credited in the record with the American successes at St. Étienne and Forest Ferme in the offensive's final weeks. A written recommendation by an officer of the division, preserved in the regimental record, urged that the practice be extended, remarking that while the enemy might decipher every other code, "he might have all the time in the world and never a chance of catching a Choctaw's 'hello.'" The sentence is carried here as what it is — one officer's contemporary testimony, not a verdict — but the Choctaw traffic of October and November 1918 was never shown to have been read.

Documented Fact, with the counts carried as the sources give them. The first squad numbered eight men; the Oklahoma Historical Society names them: Solomon Louis, Mitchell Bobb, Ben Carterby, Robert Taylor, Jeff Nelson, Pete Maytubby, James Edwards and Calvin Wilson. Eleven more Choctaw talkers joined them — nineteen in the usual count, though the regimental history speaks of fourteen Choctaw men of the 36th Division; the counts are carried here unreconciled, as the sources leave them. Debated / Claim, flagged: the Choctaw are rightly called the first organised code talkers, but the National Museum and Memorial of the First World War records that the record reaches slightly earlier in two places — Ho-Chunk soldiers (Robert Big Thunder and John Longtail are named) used in their language shortly before 21 June 1918 near Château-Thierry, and Eastern Band Cherokee soldiers likewise on 7–8 October 1918. Those uses stand in that museum's record; they were not squads, and they built no code. By the war's end, the record names Cherokee, Comanche, Cheyenne and Osage soldiers also transmitting in their languages.

Documented Fact. One generation's irony opens here, and it does not close again in this entry. The Choctaw code talkers of 1918 belonged to a nation whose children were, in those same years, being punished in government boarding schools for speaking Choctaw. Judy Allen of the Choctaw Nation, whose own family carries this history, stated the irony in the form the Nation's histories preserve: the language was beaten out of children at school and begged for at the front.

3. The largest: the Navajo (Diné) code talkers, 1942–1945

Documented Fact. The Second World War's great programme began with a civilian's memory. Philip Johnston, a First World War veteran who had grown up on the Navajo Reservation as a missionary's son and spoke Navajo, read in early 1942 of Indigenous language use by other forces and carried a proposal to the Marine Corps: the Navajo language — unwritten in any form the enemy could obtain, spoken by no one outside the Diné and a handful of others, and of a complexity that made it, in the phrase the record repeats, a code in itself. He demonstrated it to Major General Clayton Vogel, commanding at Camp Elliott, who recommended the programme. In May 1942 the first twenty-nine Navajo recruits — the original 29 — entered boot camp and then Camp Pendleton, as Platoon 382, and built the code themselves: a dictionary of Navajo words standing for military terms, and Navajo words for the letters of the alphabet — up to three for each letter, so patterns would not repeat — for spelling names and places. Besh-lo, "iron fish," was a submarine. Da-he-tih-hi, "hummingbird," was a fighter plane. Ni-ma-si, "potatoes," were grenades. Ne-he-mah, "Our Mother," was America.

Documented Fact. The programme grew far beyond the twenty-nine. Of roughly fifty thousand Navajo tribal members in 1942, some 540 Navajo men served as Marines, and about 375 to 420 of them were trained as code talkers (the range is the sources' own, carried in the Navajo Nation's and Native Partnership's histories). They served, in pairs and teams, with every Marine division in the Pacific — Guadalcanal, Tarawa, Saipan, Peleliu, Iwo Jima, Okinawa. Thirteen code talkers were killed in the war, on New Britain, Bougainville, Guam and Peleliu (the Navajo Nation's partnership history). At Iwo Jima, six code talkers working around the clock through the battle's first two days sent and received more than eight hundred messages, all without error. Their signal officer, Major Howard Connor of the 5th Marine Division, gave the verdict the record has kept ever since, and this entry gives it in his words: "Were it not for the Navajos, the Marines would never have taken Iwo Jima." The messages that moved the battle — including the traffic of the days when the flag went up on Mount Suribachi — moved in Diné Bizaad, in a code inside it, at twenty seconds a message.

Documented Fact, the men's own testimony. The talkers' testimony records what the code demanded of them beyond fluency: transmission under fire, from exposed positions, with equipment heavier than a rifleman's load, and bodyguards assigned in some units because a talker's capture would have been a catastrophe — testimony carried in the oral histories the Smithsonian's Native Words, Native Warriors project gathered from the men themselves. It also records the other side of the weapon: Navajo Marines who were not talkers were sometimes mistaken by other Americans for Japanese soldiers; and the prisoner-of-war anecdote the talkers themselves told — a captured Navajo soldier, not a talker, forced by his Japanese captors to listen to the transmissions, who told them afterwards, in his fellow talker Chester Nez's retelling, "I never figured out what you guys… were saying" — is carried here as the men's testimony, labelled as testimony, which is exactly the standing it has in their own histories.

4. The Army programmes, and Canada's talker

Documented Fact — the Comanche. In the European war, the United States Army built its programme on the Comanche language. Seventeen Comanche men enlisted for the 4th Signal Company of the 4th Infantry Division; three could not complete the Atlantic crossing with the unit at first, and fourteen Comanche code talkers went into Normandy on D-Day, 6 June 1944, two assigned to each regiment and the rest at division headquarters, transmitting from Utah Beach from the morning of the landing. They had compiled a vocabulary of some 250 code terms in Comanche: wakaree'e, "turtle," for a tank; "pregnant bird" for a bomber; "sewing machine" for a machine gun; puhihwi tekwapu kahni, "metal talking house," for a telephone exchange; Adolf Hitler was Po'sa taivo, "Crazy White Man." The talkers served through France and into Germany; some were wounded; none of the seventeen was killed (the Comanche Nation's record). Germany had prepared for such a weapon and failed at it: before the war it had sent some thirty anthropologists to the United States to learn Indigenous languages against exactly this use, and they could not master them (as carried in the standard code-talker literature). France knighted the Comanche talkers as Chevaliers of the National Order of Merit in 1989; Charles Chibitty, the last surviving Comanche code talker, received the Knowlton Award in 1999 and died in 2005.

Documented Fact — the other nations of the Second World War. The record names far more nations than the two famous programmes. Eight Meskwaki men served as code talkers with the 168th Infantry Regiment of the 34th Division in North Africa. Hopi talkers — some of them Hopi-Tewa bilinguals — served with the 81st Infantry Division in the Pacific; the counts of Hopi talkers vary by source and are carried here without one being chosen. A Chippewa-Oneida group in the Army built code terms from the colours of the military hat cords — blue for infantry, yellow for cavalry, red for artillery — as a 1941 period report in The Masterkey records. Cherokee, Choctaw (again, a generation after 1918), Lakota and Dakota, Kiowa, Menominee, Muscogee (Creek), Seminole, Pawnee, Sac and Fox, Ho-Chunk, Osage, Crow, Pueblo and Apache men, and five Tlingit talkers whose service stayed unknown even to their families until declassification — honoured at last by the Alaska Legislature in March 2019 — stand in the record beside them. Canada's war carries the same line: Charles "Checker" Tomkins, a Métis man from Alberta, translated sensitive radio messages into Cree for the Canadian forces — the service named in the Government of Canada's own Indigenous Veterans Day statements.

Documented Fact — the statute that finally counted them. The Code Talkers Recognition Act of 2008 (H.R. 4544, signed 15 October 2008) directed a Congressional gold medal for each tribe whose members had served as code talkers, with silver duplicates for each talker or his family. The Act as first written named fourteen groupings and opened the door for others to come forward with proof; by the presentation ceremony in Emancipation Hall of the United States Capitol on 20 November 2013, thirty-three Native nations were called forward. Twenty-five received their gold medals that day, the remaining nations' designs still being minted. Of all the men of all the nations, one non-Navajo Second World War code talker was still able to be there in person: Edmund Harjo, Seminole, aged ninety-six. The rest of the honour, for most of the men, came to their families and their nations, decades after the men themselves were gone.

5. Korea and Vietnam: what the record actually shows

This section carries a correction at full strength rather than as a footnote, because the strand's law is that the record is never stretched to fit a sentence.

Documented Fact. No formal code-talker programme is documented for the Korean War or the Vietnam War. The code itself was retired with the Second World War's generation, and the one sentence the reference literature prints on the question is carried here verbatim, because it is the honest state of the record: "Rumors of the deployment of the Navajo code into the Korean War and after have never been proven." What is documented is that the men themselves went on serving. Chester Nez, one of the original 29, re-enlisted and served as a Marine in Korea, retiring in 1974 after twenty-five years in the Corps; William Dean Yazzie (Dean Wilson), another Navajo talker, likewise served through the Korean War. And the service around them was enormous, and is documented at scale: the United States Navy's official accounting counts some 44,000 Native servicemen and women in the Second World War, 15,000 in Korea, and more than 42,000 in Vietnam — the Army's accounting puts the Vietnam figure at over 42,000, more than ninety percent of them volunteers. The part of the common sentence that belongs to Vietnam is therefore real, and it is this: tens of thousands of Indigenous men volunteered and fought bravely in that war, as in Korea — as the companion class, They Signed On — The Long Service, records. The part that belongs to the two World Wars is the part about the codes. History kept whole is worth more than history stretched, and this entry keeps it whole.

6. The silence, and the recognition that came late

Documented Fact. The programmes were classified. When the men came home in 1918 and again in 1945, they came home under orders of secrecy that held for twenty-three years the second time: the Navajo code was not declassified until 1968. Chester Nez's testimony states what that meant in a life: "Until 1968, Nez was unable to tell anyone, including his family." The Choctaw men of 1918 waited longer still for any public accounting. There were no parades for a weapon no one was allowed to name, and the men carried the secret the way they had carried the radios — without putting it down where anyone could see it.

Documented Fact — the recognition, in sequence. It came in stages, most of them after most of the men were dead:

The count of the lateness, stated plainly. Of the original 29 Navajo talkers, 24 were dead before the gold medal existed. Charles Chibitty outlived every other Comanche talker and died in 2005, three years before the statute that would honour his nation passed. Edmund Harjo stood in Emancipation Hall in 2013 as the only man of his whole Army generation of talkers able to stand there. The recognition was real when it came. The record does not permit calling it timely.

7. The bitter line: the suppression surrounded the service

It is sometimes said the governments later tried to remove their language. The chronology is worse than that, and the entry prints the true one: the removal was attempted before the wars, during the wars, and after the wars — the service happened inside the suppression, not before it.

Documented Fact — before. The Carlisle Indian Industrial School, the model school of the United States boarding system, opened in 1879 under Richard Henry Pratt — forty years before the Choctaw squads. Pratt stated the system's aim in his 1892 speech in words the school's own archive preserves: "Kill the Indian in him, and save the man." In the federal systems of the United States and Canada alike, children were punished — in case after documented case, violently — for speaking their mother tongues; the Genocide Record classes carry the systems whole (the United States Interior Department's counts of 408 federal schools in 2022, revised to 417 schools with 973 documented deaths and 74 burial sites in 2024; the Truth and Reconciliation Commission of Canada's record of at least 150,000 children taken and 3,200 deaths identified) and this entry cross-refers rather than repeats it. The men of the Telephone Squads of 1918, and the original 29 of 1942, came out of that world: Charles Chibitty's own record states that he was punished as a schoolboy in a government boarding school for speaking Comanche — the language the Army would later beg him to transmit.

Documented Fact — during. The schools did not pause for the wars. The systems that punished the languages operated through 1918 and through 1942–45, in both countries, while the same governments recruited the languages' speakers for the most secret communications work of both wars. Nor was the contradiction only about language. In Canada, a status Indian could not vote in a federal election without surrendering his status — enfranchisement — until 1960; the men who enlisted in 1914 and 1939 enlisted for a country whose ballot was closed to them (the companion class, They Signed On, carries the record). In the United States, the men of the 36th Division served in 1918 for a country that did not extend citizenship to Native people generally until the Indian Citizenship Act of 1924 — an Act passed, as its own record states, in part in recognition of that service.

Documented Fact — after. Neither system closed when the wars ended. The residential schools of Canada ran on into the late twentieth century; the boarding systems' language regime outlived the code talkers' secrecy itself, and the last of the secrecy lifted only in 1968 — into a continent where the punishment of the languages was still within living memory on every reservation and in every home the men returned to. Judy Allen's irony (§2) is therefore not a First World War footnote. It is the shape of the whole record: the governments punished the languages in the children and depended on them in the men, in the same decades, sometimes in the same families.

The entry's closing line, which the record supports in full: the languages they tried to remove were the languages that could not be broken. German intelligence could not read them; Japanese intelligence could not read them; thirty anthropologists could not learn one of them in time. What survived the schools did what no machine cipher of two world wars could be trusted to do. The men are owed the honour late. The languages are owed the record whole.

Sources for every claim in this reading stand in the research file behind this class. Its weakest and flagged points are carried openly in the reading itself: the Choctaw counts (8 / 14 / 19) are unreconciled across the sources; the "first" claim is bounded by the WWI Museum's earlier Ho-Chunk and Eastern Band Cherokee record; Hopi talker counts vary and none is chosen; and the Korea/Vietnam section states the record's exact level rather than the common claim. Companion classes in this band: They Signed On — The Long Service · The Nations in the War Over Slavery · Chewing Away at the Treaties · The Fight Goes On — the nations' present-day fight over their rights continues the same long line this class records.

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