The Repository Last Known Position

USS Oklahoma: How 394 Pearl Harbor Unknowns Got Names

October 10, 2026 · Runtime 12:07 · Watch on YouTube

On 7 December 1941 the battleship USS Oklahoma capsized at Pearl Harbor, and 429 of her crew died. After the war, 394 of them still had no name: their bones had been sorted by type at Schofield Barracks, and in 1949 they were buried in 62 caskets, in 46 graves at the Punchbowl in Honolulu, under headstones that read Unknown. In 2015 the Defense POW/MIA Accounting Agency dug them all up. We read the file that closed the case: the ship's action report, the 1949 board, the project leader's forensic paper, and the agency's own cameras.

Chapters

  1. 0:00 Pearl Harbor, the morning of December 7th
  2. 0:47 Outboard on Battleship Row
  3. 1:31 135 degrees in eight to ten minutes
  4. 1:56 A survivor on the Maryland
  5. 2:52 Recovery and the 1943 righting
  6. 3:37 Schofield Barracks: sorted by bone
  7. 4:44 62 caskets marked Unknown
  8. 4:57 One casket, 95 people
  9. 5:44 2015: every grave at once
  10. 6:43 13,000 bones, 392 bundles
  11. 8:32 Two matches: bone to bone, man to family
  12. 9:37 One closed case
  13. 10:35 355 of 388
  14. 11:26 What the record supports
  15. 11:46 Next: the 55 boxes

Transcript

Pearl Harbor, the morning of December 7th 0:00

This is Pearl Harbor on the morning of December 7th, 1941, filmed by the Navy from the deck of a ship. Somewhere in this harbor, the battleship Oklahoma is about to roll over.

Four hundred and twenty-nine men died on her. After the war, three hundred and ninety-four of them still had no name. They were buried in Honolulu under headstones that read Unknown, and they stayed there for sixty-six years.

This episode reads the file that closed their case: the ship's action report, a 1949 board, a forensic paper written by the woman who ran the project, and the laboratory's own cameras. And it asks one question. How do you give a name back to a man whose bones were sorted by type, eighty years ago?

Outboard on Battleship Row 0:47

The Oklahoma was a Nevada-class battleship, launched in 1914. On that Sunday morning she was moored on the outboard side of Battleship Row, at Ford Island, with the Maryland inside her.

Outboard is the side that faces open water. It was the side the torpedo planes could reach.

The attack began just before eight. The ship's action report says three torpedoes hit her port side within a few minutes, and two or three more were felt. Her executive officer estimated five to seven.

Two thousand four hundred and three Americans died at Pearl Harbor that day. After the Arizona, the Oklahoma was the second-largest loss of life of any ship.

135 degrees in eight to ten minutes 1:31

She heeled over fast. According to the action report, she turned through about one hundred and thirty-five degrees in eight to ten minutes, and came to rest upside down in the mud, with her keel in the air.

Hundreds of men were still inside, in compartments that were now upside down. A sailor on the Maryland, moored right beside her, described what the rescuers heard.

A survivor on the Maryland 1:56

This is Floyd Welch, a Pearl Harbor survivor, in a Defense Department interview for the seventy-fifth anniversary.

Rescuers cut and chipped their way through the steel to reach them. The action report counts thirty-two men rescued by the salvage party. Four hundred and twenty-nine did not come out.

Recovery and the 1943 righting 2:52

Recovering them took years. Navy crews brought out remains from December 1941 until June 1944, and buried them in two cemeteries on Oahu, Halawa and Nu'uanu.

To bring her back up, the salvage engineers first built a cardboard model of the ship, and worked out how to pull a capsized battleship upright.

The pull began on March 8th, 1943, and the Navy's own photographers recorded it. Twenty-one winches on Ford Island, hauling on cables run over tall frames on her hull, turned the Oklahoma upright by mid-June.

The ship came back. The men inside her, for the most part, did not come back with their names.

Schofield Barracks: sorted by bone 3:37

In September 1947, the Army's Graves Registration Service dug up the Oklahoma dead from both cemeteries and took them to its identification laboratory at Schofield Barracks.

And here is the decision that made this case hard for seventy years. The paper written by the project's leader says the remains were sorted by element, skulls with skulls, long bones with long bones, so they could be buried in as few caskets as possible.

The same paper explains why. The belief at the time was that the Oklahoma dead would be identified, if ever, as a group, and that belief drove how their bones were handled.

The anthropologist at Schofield, Mildred Trotter, certified only twenty-seven separations, skulls and jaws matched to dental records. The Army's Quartermaster General did not approve even those.

In October 1949, a military board classified the men who could not be identified as non-recoverable.

Thirty-five Oklahoma men had been identified. Three hundred and ninety-four had not.

62 caskets marked Unknown 4:44

Their remains went into sixty-two caskets, in forty-six graves, at the National Memorial Cemetery of the Pacific, the crater the people of Honolulu call the Punchbowl. Every headstone said Unknown.

One casket, 95 people 4:57

The case was reopened by a survivor. Ray Emory, a Pearl Harbor veteran, spent years cross-checking burial records and ship's logs, and came to believe that one Punchbowl casket held men who could be named.

In 2003 that casket was dug up. The records suggested five men. DNA testing of nearly two hundred bones from it found ninety-five different mitochondrial sequences.

Ninety-five people, in one casket. One in four of all the Oklahoma dead had at least one bone in that single box.

That told the laboratory something about all the others. If one casket was this mixed, every casket was. Identifying anyone would mean opening every grave at once.

2015: every grave at once 5:44

On April 14th, 2015, the Deputy Secretary of Defense, Robert Work, signed that order. Exhume every Oklahoma unknown, and identify them individually, to the extent practical, within five years.

The new policy set a bar. For mixed remains, the research had to show at least a sixty percent chance of individual identification. For the Oklahoma, the Pentagon said families had already given DNA for eighty-four percent of the men.

On June 15th, 2015, the exhumations began at the Punchbowl.

Over six months, sixty-one caskets came out of forty-five graves, holding the remains of up to three hundred and eighty-eight men. The last four left the cemetery in November.

Each casket left the cemetery under a flag, carried by a joint honor guard, and was driven to the agency's laboratory at Joint Base Pearl Harbor-Hickam.

13,000 bones, 392 bundles 6:43

In Hawaii, every casket was opened and inventoried. What the laboratory found was the 1947 sorting, intact.

Teeth came first. The dental remains were worked in Hawaii, while the rest of each skeleton went to the agency's laboratory in Nebraska, to be sorted again.

Nearly thirteen thousand skeletal elements, in three hundred and ninety-two bundles. Sixty of the sixty-two caskets held at least two bundles.

On average, a casket held about nine bundles, and a bundle held about thirty-two bones. The smallest bundle had four. The largest had one hundred and seventy-five.

The project was run from the agency's laboratory at Offutt Air Force Base, in Nebraska. In the agency's own film about the project, its scientists put the scale plainly: thirteen thousand bones, and five thousand DNA samples.

And here is the measure of the sorting. The project's paper says no man identified so far had come from a single bundle, or even from a single casket.

The team tested the 1940s bundles directly. A left and a right upper arm bone in the same bundle belonged to the same man forty-four percent of the time. An upper arm bone and a shin bone from the same bundle, only six percent.

The inventory outgrew paper, and then a spreadsheet. According to the laboratory, it built a database with the University of Nebraska Omaha to hold it, and that database now serves other mixed cases.

Even counting the dead was a measurement. Counting skulls gives a minimum of three hundred and fifty-seven people. Adding DNA raises it to four hundred, against the three hundred and eighty-eight men the records said were there.

Two matches: bone to bone, man to family 8:32

So the method had two steps. First, DNA joins bone to bone, and rebuilds each man from pieces spread across many caskets. Then DNA joins that man to a living relative.

By April 2016, nearly five thousand samples had been cut and sent to the Armed Forces DNA Identification Laboratory in Delaware, a record for both laboratories.

And bone that had spent seventy years in the ground still answered. The project's paper says nearly ninety-nine percent of the samples sequenced for mitochondrial DNA gave a result.

Each test needs a different relative. Mitochondrial DNA passes down the mother's line. The Y chromosome passes from father to son. And autosomal DNA comes from brothers, sisters and children.

The paper puts reference DNA on file for eighty-five percent of the Oklahoma casualties, and dental records for just under eighty percent. The first identifications, in September 2015, came from teeth.

One closed case 9:37

Here is what one closed case looks like in the record. Fireman Third Class Willard Lawson, twenty-five, of Butler County, Ohio, was accounted for on August 27th, 2018. The release names dental and anthropological analysis, and mitochondrial DNA.

Seaman Second Class Laverne Nigg, twenty-three, of Browns Valley, Minnesota, was accounted for in April 2021, with all three kinds of DNA.

In August 2018, Machinist's Mate First Class Arthur Glenn became the hundredth Oklahoma sailor identified. He was buried with full honors, at the same cemetery where he had lain as an unknown.

Some families lost more than one son on the Oklahoma. In the same film, a scientist remembers three brothers, the Barbers, who all died on the ship, and says the laboratory was able to identify them.

355 of 388 10:35

The project ended in 2021. Of the three hundred and eighty-eight men exhumed in 2015, three hundred and fifty-five were identified. The goal had been eighty percent. Counting every identification since the war, the agency puts the result at ninety-two percent of the men who died on the ship.

Thirty-three could not be separated from the others. On December 7th, 2021, eighty years to the day, their remains were buried together at the Punchbowl, as a group attributed to the whole crew. Their names are on the Courts of the Missing.

So here is the full account. Thirty-five named after the war. Six more before the project. Three hundred and fifty-five during it. Three hundred and ninety-six of four hundred and twenty-nine.

What the record supports 11:26

For seventy years, the last known position of three hundred and ninety-four men was a capsized ship, and then a grave marked Unknown. What reversed the sorting of 1947 was not a new document. It was the bones themselves, read one sample at a time, and families who kept their side of the record.

Next: the 55 boxes 11:46

Next, a case that began with boxes, not graves. On July 27th, 2018, North Korea handed over fifty-five boxes said to hold American remains from the Korean War: five hundred and one bones, from at least two hundred and fifty people. That is the next episode. The fifty-five boxes.

Description and sources

The 1947 sorting is what made the case impossible: nearly 13,000 skeletal elements in 392 bundles, with two bones from the same bundle belonging to the same man as little as 6 percent of the time. One casket opened in 2003 was supposed to hold five men; DNA found 95. So in 2015 every Oklahoma grave was opened at once. DNA did two jobs: first it joined bone to bone across dozens of caskets, then it joined each rebuilt man to a living relative, through the mother's line, the father's line, or brothers, sisters and children. Of the 388 men exhumed, 355 were identified, against a goal of 80 percent. The last 33 were buried together on 7 December 2021, eighty years to the day.

PRIMARY SOURCES

ARCHIVE PLATES — public-domain documents, with author, date, archive and licence.

  • CEM E.D. McKinney, U.S. Navy — 1936-08-18 | Wikimedia Commons | Public domain
  • Department of the Navy. Fourteenth Naval District. — 1941-12-07 | Wikimedia Commons | Public domain
  • Unknown author Unknown author or not provided — 23 February 1943 date QS:P571,+1943-02-23 | Wikimedia Commons | Public domain
  • Unknown author Unknown author or not provided — 13 March 1943 date QS:P571,+1943-03-13 | Wikimedia Commons | Public domain
  • Department of the Navy. Fourteenth Naval District. Naval Air Station, Pearl Harbor (Hawaii). (ca. 1940 - 09/1947), Photographer ( NARA record: 1130789 ) — 19 March 1943 date QS:P571,+1943-03-19 | Wikimedia Commons | Public domain
  • Unknown author Unknown author or not provided — 19 March 1943 date QS:P571,+1943-03-19 | Wikimedia Commons | Public domain

... and 4 more plates, each credited on screen in the video and listed in this project's manifest.

Last Known Position reads a missing-in-action case through the record that closed it: the loss report, the excavation and the laboratory.

About the material. Film: U.S. Navy motion pictures of 7 December 1941 and 1942 salvage work, National Archives Record Group 428 (catalog 76168, 76171, 80358), public domain. Photographs: U.S. Navy, Naval History and Heritage Command, public domain. Video: Defense POW/MIA Accounting Agency, DoD News, Navy and Air Force units, released through DVIDS, public domain. No human remains are shown. Every diagram is drawn in code from the sources listed. Nothing in this video is AI-generated imagery.

The appearance of U.S. Department of Defense visual information does not imply or constitute endorsement. This channel is independent and not affiliated with the Defense POW/MIA Accounting Agency or the U.S. Navy.

#LastKnownPosition #USSOklahoma #PearlHarbor