On the first of June, 1943, three days after the fighting stopped on Attu, an American intelligence section working out of a tent in Massacre Valley finished translating a diary taken from a dead Japanese soldier. The diary belonged to an army doctor, a man licensed to practice medicine in the same country whose soldiers had just killed him. He had begun it on the twelfth of May, the day after the American landings, and kept it for eighteen days. His entries were short, but they became the most complete surviving account of what the last two thousand men on the island were doing while the American ring tightened around them.
The doctor logged the naval gunfire almost every day. On the twentieth of May, shells fell twenty meters from his hospital. A day later, he was strafed while in the middle of amputating a man’s arm. On the twenty-third, a shell brought a patient tent down on its poles and killed two men where they lay.
The following afternoon, a fragment went through a medical orderly’s heart in a foxhole five yards away, and the roof came in. By the twenty-sixth, he was writing that the last line at Umanos had fallen and that no help was coming. Against that grim reality, higher headquarters kept feeding him victories. He recorded a report that Japanese submarines and small craft had sunk an American battleship, thirty-seven destroyers and transports, and sixty other transports off Chichagof Harbor.
Not one of those ships existed. He called it the first favorable news since the fighting in the eastern arm, and he entered it the way a man enters good news. Some of his other entries the Americans corrected in the margins. He believed poison gas had come down on the fourteenth.
The shells were white phosphorus, fired to mark positions in the fog. He believed seventeen Japanese bombers had sunk a cruiser off the coast. American fighters knocked six of those bombers out of the sky, and no ship took a scratch. He walked into the abandoned village church at Chichagof, found blankets covering the floor, and wrote that it felt like somebody’s home.
His left knee had begun hurting after the crossing of the pass, and he mentioned that, too. About sixteen Japanese American linguists of the Military Intelligence Service had come ashore with the assault force to do this work. The army’s own official history names the man who translated the doctor’s diary, Staff Sergeant Yasuo “Sam” Utani, and states that he wept while working through it. Another, Staff Sergeant Shigeo Ito, spent a night sheltering under rocks with a language officer, and had to shift three Japanese bodies aside to lie down.
The next day, he questioned a wounded man whose injuries were full of maggots, got nothing out of him, and said afterward, plainly, that he had failed. A third, Staff Sergeant George Hayashida, crawled into caves with a flashlight and a pistol, and talked about ten men out of them. The army did not fully trust these men. They placed Navy language officers with commissions over the Nisei, most of whom handled the language better.
When one of those officers told a Nisei enlisted man to dig his foxhole for him, the answer was that he could dig it himself. Colonel John Wakerling’s report on the operation, dated the eleventh of June and now in the Fourth Army files at the National Archives, contains a paragraph stating that the command kept some Nisei off the convoy out of fear that their reception aboard ship might disclose the location of the force. Nobody has settled what that sentence means. The failure to hear things ran through the entire operation and started early.
The first American figure for the garrison, the one that persuaded Admiral Kinkaid in March to switch the objective from Kiska to Attu, was five hundred men. By early May, Western Defense Command had more than doubled that estimate. Kinkaid’s own figure was 1,587. On the fifteenth of May, four days into a battle meant to have lasted three, officers carried captured Japanese papers aboard the Pennsylvania showing a garrison of well over two thousand.
The assault force had sailed with about eleven thousand men and no cold weather planning worth the name, and it went on fighting to the original timetable. By the evening of the twenty-eighth of May, the ring was shut. The Japanese had lost Holtz Bay, Fishhook Ridge, and Buffalo Ridge, and now held only the flat ground around Chichagof Harbor with the sea behind them, a position with no depth and no way out. General Landrum ordered a full-force attack for the following morning.
For the first time since the eleventh of May, he held no infantry back in reserve. Colonel Yasuyo Yamazaki had about seven hundred men left. He had lost roughly seventy percent of his force, along with most of his artillery and supplies. The Americans held every height around him.
The 1944 War Department account of the battle sets out the four courses open to him. Surrender, which the Japanese Army code did not allow. A last stand where he was, which would achieve nothing. A withdrawal into the mountains, which would buy him days he could not feed.
Or an attack. He reasoned that the weakest part of the American ring had to be the valley floor, because the American strength sat up on the ridges. If he cut through in the dark and reached the American howitzer batteries behind Engineer Hill, he could turn those guns on the main base at Massacre Bay. At half past midnight on the twenty-ninth of May, he issued the order.
At three in the morning, Company B of the 32nd Infantry, sitting on the valley floor near the southern end of Lake Corries, received an order to pull back. The War Department volume states that the purpose is said to have been letting the men walk to the battalion kitchen for a hot breakfast before the morning attack, and then says, without softening it, that the reasons for the order are obscure in the official files. Nobody has turned up a better answer since. What is certain is that nobody expected enemy action that night, and that a rifle company was walking backward off the one piece of ground the Japanese commander had picked to come through.
They hit its rear guard at about half past three, screaming, bayonets fixed, grenades in hand. The main body drove on toward Clevesy Pass, shedding small parties left and right to keep its flanks clear, and reached the pass weaker for it. In the draws behind the pass, the divisional engineers, the 50th Engineers, and various service units lay bivouacked. They had heard firing up the valley and thought nothing of it, being far out of range.
Their first warning was infantrymen and medics coming down toward them, shouting that the Japanese were through. Within ten minutes of that shout, the engineers had an improvised line on the slope of Engineer Hill and were firing at point-blank range, with cooks, clerks, and drivers in the line beside them holding whatever weapons they had found. Small parties got past them into Clevesy Pass and reached the holdback in Massacre Valley a few hundred yards short of one battery of American howitzers, which was the objective. There the attack died.
At the foot of Engineer Hill stood a clearing station with fifteen American casualties on the floor waiting for daylight to be carried over the hill. Captain Yellen and Lieutenant Friedberg had worked on them all night, had no way to move them, and had no intention of leaving them. The Japanese cut the tent open with a bayonet, threw four grenades through the slit, then put an automatic rifle into the hole and emptied it. Friedberg died.
Every man inside took a bullet or a fragment. Yellen gave morphine to everyone he could reach and passed the syringe to a sergeant for the rest. They lay there under a Japanese machine gun set up a few yards behind the tent until American patrols came through in the early afternoon. When the fog lifted that morning, the machine gunners of Company M up on the tableland could see the whole floor of the valley below them.
The camps in the draws were wrecked. The telephone wire was cut everywhere. Bodies lay in rough circles around the tents. For most of the morning the gunners sat with their weapons idle because they could not tell which of the small groups crawling below were American.
Toward noon, a lieutenant put his glasses on a line of men at the foot of a hill 1,800 yards off, worked out that his own people held the top and the Japanese the bottom, and gave the range. Afterward he counted twenty-six figures at the foot of that hill and in the creek bed. The mopping up ran on for three days, and it is where the material state of the garrison became visible. A tractor driver and his partner, pinned by a sniper at daylight, killed the man and found two paper practice cartridges in his rifle and two more clips of them in his pouch, and nothing else.
In the caves, one American grenade thrown in produced two explosions, because each grenade was answered by a man setting off his own. At Chichagof Village, soldiers fired into three tents, shouted at them, and finally set them alight with tar paper and rags. The men inside killed themselves rather than come out. One man surrendered there in the course of that entire day.
The counting took longer. Burial parties counted roughly 2,350 Japanese dead and put them into mass graves. The National Park Service, which manages the battlefield now, states that several hundred more are thought to lie in the hills where they fell, never recovered. They found the field hospital with its patients dead.
The doctor’s final entry, timed at six that evening, says that all the patients were made to kill themselves, that he dealt with them himself with a grenade, and that thirty-three men of his unit were still alive. He said goodbye to his wife by name and to two daughters, the younger born that February, who would never see him. His last note gives the size of the attack going out into the dark, a little over one thousand men, and their intention of taking the American guns. The prisoner count would not sit still.
General Landrum’s report of the thirty-first of May, the first tally anybody made, gives twenty-four. He wrote it while patrols were still killing men in the draws behind Chichagof, and it went stale the day it left his desk. The taking did not stop on the thirtieth of May. For weeks afterward, observation posts and mopping-up patrols pulled men out of the rocks in ones and twos off ground the histories had already handed over.
A few had hidden since the charge. Others had never joined it. On the eighth of June, an observation post from the 4th Infantry near Clebnikov Point chased a single figure over a ridge, fought a party of Japanese at a range of ten feet in a rocky hollow, and came away with two prisoners, one of whom walked out with his hands up shouting in English that he had no weapon. Asked where he was from, he told them, “San Francisco.
”
The last man on Attu gave himself up in September, more than three months after the army called the island secure. The War Department volume of 1944 and the army’s official campaign history both settle on twenty-nine. The army’s history of the Nisei linguists, published in 2006 and working from the language team’s own papers, gives twenty-eight. All three counts are honest.
Each of them counts a thing that had not yet finished happening. In Tokyo on the thirtieth of May, the Imperial High Command announced the loss of the island, called the charge heroic, and gave the American forces 20,000 men and their casualties as 6,000, roughly double the true strength and roughly four times the real loss. Engineers were grading a fighter strip on Alexai Point within days of the fall of Chichagof, on ground where one of the landings had gone in, and it was operational by the ninth of June. Shemya, thirty-five miles to the east, was occupied on the thirtieth of May, and engineers began a bomber runway there the same week.
In July, aircraft flying from the new Attu field struck the northern Kuril Islands, the first American attack on Japan since the Doolittle Raid. The diary had a longer afterlife than any of that. Copies of the translation went around the island fast, and senior officers who understood what a page of a man missing his children does to troops, who for a year had been told that the enemy was barely human, ordered every copy confiscated. It did not work.
Soldiers mimeographed it, copied it out by hand, added their own notes in the margins, and passed it between themselves anyway. One writer who went looking for it decades later turned up ten separate English versions of the same original. That is why the wording shifts from book to book, and why the copy typed in Massacre Valley on the first of June garbles his commander’s name, his wife’s name, and the name of his God.
