Threshold of War Read online

  Roosevelt was not ready for all-out measures. Admiral King, who was at the White House on July 9 and 17, probably resisted an immediate all-out solution, arguing the unreadiness of his fleet. Time was needed to sort problems and, in King’s words, “get correctly started on this complicated situation.”61 Operational limitations went hand in hand with the president’s inclination to take one step at a time. That step he took, however. Roosevelt’s decision was, as usual, a blend of advice, yet peculiarly his own: American naval protection would be restricted to convoys under American control and formed on American and Icelandic ships, but British merchant ships could join up. This more limited and, as the British said, “cumbersome” approach became Western Hemisphere Defense Plan Four. One advantage of the plan, aside from providing convoys with a figleaf of American nationality, was that it provided for as many convoys as the Americans could escort, not necessarily as many as sailed.62

  Designing an escort system under Plan Four proved to be immensely difficult. With protection limited to convoys on the line of communication to Iceland, planners had to determine how far from a convoy U-boats posed a threat justifying attack by escorts. By June the Admiralty was providing the Navy Department in Washington with daily convoy and U-boat locations.63 How was Washington to keep the American convoys and escorts informed? Who would divert convoys from newly discovered U-boats? How would British ships and American convoys meet up? Who was responsible for the safety of British ships joining a convoy? Who would make British and American convoy schedules jibe and keep a ready supply of escort vessels? Who would deal with a breakout by the Tirpitz? How would Plan Four, restricting protection to convoys, square with American responsibility under ABC-1 for the western Atlantic as a whole, a command arrangement, by the way, which the Canadians bitterly resisted?64

  British, Canadian, and American naval officers labored through the July heat in Washington to find answers to these questions. The fellow traveler idea pressed by Harry Hopkins, providing an American-flag service which coincided with British convoys, proved impossible: the Americans could not cover every convoy. Admiral King, consulting with the president, then devised a plan for alternating convoys with the Canadians, each navy contributing thirty escorts, and agreement along these lines was reached July 22. However, the president reversed himself and through Admiral Stark rejected the agreement on the ground that it was impossible to enter into formal undertakings with foreign authorities for the operation of American naval forces, and so when Plan Four went into effect July 24, provision for combined convoys was withheld.65

  A plan that would meet the president’s requirement of cooperation but not combination proved impractical; a practical plan did not meet the president’s requirement. The Americans were discovering that escort of convoy was so intricate an operation, especially when multinational, that it could only be effected on a basis of political intimacy of the powers involved. Once that was plain, Roosevelt postponed putting into effect an integrated convoy system until his meeting with Churchill at Argentia. Only after he had established common peace aims with Churchill and thereby the political basis for risking and waging war could he move to co-belligerency on the Atlantic.

  Postponement did not seem too harmful because the Battle of the Atlantic turned sharply in Britain’s favor during the summer. U-boats faced short summer nights that curtailed the gathering of the pack and less hunting time the further westward they cruised. ULTRA provided the British and Canadians a supreme advantage by near-current location of most submarines permitting diversion of convoys around them. The Germans ascribed British knowledge of their whereabouts in part to reports from American naval patrols, which, they complained, “greatly hampered” operations. Discouraged by slim pickings from North American convoys and constraints imposed by the Fuehrer, the German navy shifted U-boats closer to Europe and Africa where Luftwaffe planes could spot convoys. The elimination of German supply vessels on the Atlantic in June made further attacks on convoys by German heavy ships “almost impossible.” In addition, BARBAROSSA drew the Luftwaffe away from British ports and, as the summer passed, released British destroyers from invasion guard. With these favoring conditions as well as end-to-end escort, completion of new corvettes, transfer of American Coast Guard cutters, more patrol planes, and fewer independently routed ships, sinkings sharply decreased. Tonnage lost plummeted from 432,025 in May to 120,975 in June. In fact, no ships in North American convoys were lost from the end of June until the second week of September.66

  In the weeks before the Argentia Conference the Atlantic Fleet expanded operations, but slowly, a step at a time. On July 30 the battleship New Mexico, reaching the northern end of the patrol line, kept on to Iceland instead of turning back. Thereafter battleships ceased patrolling and stood guard at the Iceland base of Hvalfjordur and at Argentia, though precisely what they were guarding remained unclear. On August 5, possibly for experimental purposes, the Support Force formed its first unit of old and new destroyers, a mixture required for merchant convoy service so that if old destroyers ran low on fuel and broke off, the convoy would not be left unprotected. On August 9, American ships moved directly into the convoy lanes. Powerful, heavily escorted task forces and small groups of Icelandic ships or navy supply and auxiliary vessels, usually guarded by two destroyers, passed close to British convoys but only seldom and then briefly, almost coincidentally, traveled in company with them.67 Training and preparation for active service continued intensively, but otherwise the fleet was mostly marking time, waiting for the summit.

  Many circumstances inhibited American intervention in the Battle of the Atlantic. Certainly Roosevelt had to move cautiously in the face of opposition in Congress to the extension of Selective Service and the dispatch of troops outside the Western Hemisphere to Iceland. Even so, given his other constraints, the president probably considered that taking a step at a time but moving was a broadly wise and prudent course and one generally congruent with existing public sentiment. Hadley Cantril polls taken July 11 and 19 and sent to the White House showed that the German-Russian war had not decreased the percentage favorable to assisting Britain even at the risk of war. Seventy-two percent wanted Russia to win and 4 percent Germany. Further reason for caution was interception of more Japanese messages indicating an impending move into Indochina. Roosevelt would not want to overreach in the Atlantic while a crisis brewed in Southeast Asia and would have all the more reason to underscore the defensive character of any move.68

  Chapter 5

  July The Containment of Japan

  The outbreak of war between Germany and the Soviet Union faced Japan with a most severe dilemma. The lie of Japanese policy had been increasingly toward the south, for resource acquisition and imperial self-sufficiency, for sealing off China, for Axis collaboration against Britain. The Imperial Army and Navy may not have had the same reasons for supporting the southward trend, but they had reasons running in the same direction, and these jibed with Matsuoka diplomacy, as evidenced by his pact with the Soviet Union. Now in June 1941 the global balance had shifted as rudely as earlier it had, in the opposite direction, with the Nazi-Soviet Pact of 1939. Now Axis collaboration pointed northward. The Soviet Union, instead of being neutralized, would in all likelihood join hands with Britain and the United States, increasing the isolation and encirclement of Japan. Only three months earlier Matsuoka had dreamed of arraying Japan with Germany and the Soviet Union; now the one had attacked the other with only the briefest intimation of its intentions. For ten days after June 22 the Japanese government struggled to regain consensus and adjust its policies to the vastly altered circumstances.

  Important elements swung behind an attack on the Soviet Union. Russia had always been the Japanese army’s traditional foe and most likely antagonist. Leading officers were sorely tempted to strike while the Red Army was beset by invasion from the west. If the Soviet threat were eliminated, the resources of eastern Siberia secured, Moscow’s aid to China cut off, and the Anglo-Americans beca
me preoccupied with an all-powerful Germany, then the southward advance could safely proceed. This seizure of the initiative was termed the policy of the “green persimmon,” from the notion that the persimmon was better secured by shaking the tree while the fruit was green than by waiting for it to ripen and fall. Matsuoka, once he had caught his breath, hastened to join the early harvesters, indeed to lead them, as adherence to the alliance with Germany was the centerpiece of his diplomacy.

  “East Asia—From Which Japan Would Carve a ‘Co-Prosperity Sphere’”: New York Times, August 3, 1941.

  Not all the generals and colonels were so bold. Soviet Far Eastern forces had dealt severely with the Japanese Kwantung Army in border fighting at Changkufeng in 1938 and Nomonhon in 1939. Moreover, in June 1941 twelve Japanese divisions and 800 planes faced thirty Russian divisions (twenty-six by July 2) and 2,800 planes. Even allowing for the larger Japanese division, the Kwantung Army was badly outnumbered and not yet deployed or supplied for attack. Followers of the “ripe persimmon” strategy argued that mobilization and reinforcement must come first. The general staff’s Kantokuen plan called for beefing up the forces in Manchuria to twenty-two divisions, with 850,000 men, requiring 800,000 tons of shipping, for operations beginning August 29. This buildup would still not be enough; conditions would not be ripe for attack unless the Russians also withdrew a major portion of their Far Eastern forces: one-half their divisions and two-thirds of their planes. Accordingly, Japan’s decision for war in the north, which would have to be taken by August 10 to allow completion of operations before winter, would depend on the course of the German-Soviet war and what demands it made on Soviet forces in the Far East. Unless conditions to the north were “extremely favorable” for attack, mainstream opinion in the army held, the southward advance should proceed. Japan should in any case choose its avenue of advance independently of Germany.

  While the Japanese army could visualize an attack on the Soviet Union in certain circumstances and made preparations, the navy was totally opposed, not the least of its motivations being that a land war would enhance the army’s call upon national resources and diminish the navy’s. Of wider import was the navy’s argument that the Soviet Union was likely to coalesce with Britain and the United States and that war with the one would lead to war with all three, whereas a southward advance leading to war with Britain and the United States was not likely to bring in a Soviet Union beset by Germany. The fleet’s carriers would not be available to protect Japanese cities from Soviet air power, naval officers pointed out to their army counterparts.

  A pliant naval leadership strongly influenced by aggressive subordinates on the naval war plans committee insisted on pursuing the southward advance. The navy was determined to break out of what it saw as ever-tightening Anglo-American-Dutch encirclement. It took very seriously the offensive capabilities of the American fleet at Pearl Harbor, which, in war games of 1940, effectively attacked while the Imperial Navy invaded the Dutch East Indies. From this lesson the navy planners moved to the assumption that at some point the southward advance would trigger an American embargo, that this would force Japan to seize the Dutch East Indies for oil, and that war with the Americans as well as the British and Dutch was bound to result. This path of reasoning, always circling back to the presumption of war with the United States, pushed Japan’s navy toward acceptance of general war as a matter of planning and policy. Large threats justified larger allocation of war resources, especially steel for new ships, and the navy was not backward in its claims in policy debates, contributing more than its share to the creation of a crisis mentality. At the same time the navy needed time to prepare and it recognized the risk of war with the United States, so accompanying this fatalistic thinking and feverish argument was a sizable if subdued strain of caution.

  The army, for that matter, was not lacking in enthusiasts for a southern policy. Key generals recognized the danger of a diffusion of Japanese strength between north and south and the necessity of coordinated army and navy action in the south. The next step in the southward advance, acquisition of bases in southern Indochina, would yield critical staging areas for an attack on Malaya and Singapore. Pnompenh, Kompong Trach, and Siem Reap—Cambodian sites for air bases—lay less than 400 miles across the Gulf of Siam from invasion beaches at Singora, Pattani, and Kota Bharu. Beyond the slender isthmus of Kra, which joined Malaya to the Asian mainland, lay the Indian Ocean, Burma and the back door to China. Projecting into the South China Sea like a clenched fist, southern Indochina outflanked the Philippines and would carry Japanese power to the very edge of the resource-rich Indies. In addition to the Cambodian bases, the operation required sites in what is now southern Vietnam: Saigon and Cam Ranh Bay as naval bases and Da Nang, Soc Trang, Nha Trang, and Bienhoa as air bases. The army insisted on moving in troops as well, peacefully if the French acquiesced and by force if they refused. The painstaking process of balancing opposing views and stakes culminated July 2 in an Imperial Conference decision (the emperor present) that Japan would prepare in the north and attack if the Germans were clearly winning. Meanwhile they would secure a final departure line for attack in the south.1

  While the Japanese government deliberated, American officials speculated. They did not doubt that the new tide of war in Europe could have large repercussions in East Asia, but they found it very difficult in spite of MAGIC to fathom what was, after all, a highly conditional and ambiguous Japanese policy. They were aware of conferences to thrash out differences day after day in Tokyo. Writing the day before the July 2 Imperial Conference, President Roosevelt saw what he described as a “real dragdown and knockout fight” going on, the Japanese “trying to decide which way they are going to jump — attack Russia, attack the South Seas (thus throwing their lot definitely with Germany), or whether they will sit on the fence and be more friendly with us.”2 While the president viewed the question broadly as a choice between two camps, his subordinates dwelt on more discrete Japanese choices: attack on the Soviet Union, seizure of bases in southern Indochina, attack on Malaya and the Dutch East Indies, watchful waiting, improvement in relations with the United States, or some combination or sequence of these.3

  Initially, the predominant feeling among military and diplomatic officials was that Japan would attack the Soviets. The opportunity to remove the perpetual menace in the north, if feasible, would prove enormously tempting. This view was strongly reinforced by decryption of a message from Matsuoka to Berlin for Ribbentrop just after the Imperial Conference of July 2. Japan, it said, was “preparing for all possible eventualities” and “keenly watching developments” in eastern Siberia so as to join Germany in “combatting the Communist menace.” British officials, who must have decrypted the same message at the same time, correctly emphasized the “preparing” and “watching” and concluded that Japan would not move northward “for the present.”4

  The Americans, especially the navy, were less conditional.5 The commandant of the Third Naval District in New York reported that a source close to the Japanese business community there expected a war on Russia about July 20. Reports from China also indicated an attack northward.6 In this forecast of northward advance the influence of Admiral Kelly Turner is apparent. The dogmatic and domineering chief of War Plans, who insisted on control of intelligence as well, became all too confident of his knowledge of Japanese ways acquired from a whirlwind ceremonial visit to Japan in 1939 as captain of the cruiser Augusta. Believing that the admirals he met then were not inclined to risk war with the United States, and dedicated to the Europe-first strategy, Turner was easily swayed by evidence of a northward thrust.7 But the Japanese admirals he met in 1939 were not of the same ilk as those in control in 1941.

  This early sense of imminent Japanese-Soviet war diminished as July wore on. British advice that Japan would wait and see and the fact that the Kwantung army was not strong enough yet led to the correct conclusion that the northward advance was dependent on the progress of the German invasion of Russia, thoug
h at what point Japan would feel confident enough to attack was impossible to say. Perhaps the fall of Moscow.8

  At the same time, however, reports from American consuls in Mukden, Dairen, and Harbin provided impressive evidence of the seriousness of Japanese preparations. Beginning July 18, these told of a swelling stream of Japanese infantry, cavalry, and artillery troops arriving in Manchuria and, using coolies and draft animals, passing north to strategic locations within reach of the Soviet border. The embassy in Tokyo reported a large-scale secret mobilization of reserves. Grew asked for information about the progress of the German offensive, as this, the embassy believed, would determine the Japanese attitude.9 Throughout July a large question mark hung over Japan’s intentions in the north.

  On the other hand, British and American officials had no doubt about Japanese intentions toward southern Indochina. MAGIC provided a full account from Tokyo’s messages seeking Vichy’s consent: that an expeditionary force of 40,000 troops was being sent, that Japan would use force if Vichy refused, that Vichy must respond by July 20, that the French did indeed bow to Japanese demands that day, that an agreement was worked out July 23, and that Japanese troops were prepared to disembark July 24—which they did.10

  Nor did officials doubt that southern Indochina was the penultimate and not the ultimate stage of the southward advance. Suggestive of Japanese army thinking was a MAGIC intercept translated July 19 which reported that military authorities in Canton considered that the object of the Indochina occupation was “to launch therefrom a rapid attack” on Singapore with an ultimatum to the Dutch. This telegram especially worried Stimson because it implied that Singapore could be taken by only one division, but he was reassured by his staff that the Malay Peninsula was so narrow that two divisions could hold the Japanese and that they would have to attack Hong Kong and the Philippines at the same time, landings beyond the capabilities of the Japanese navy.