Impermanent alliances: cryptologic cooperation between the United States, Britain, and France on the Western Front, 1917–1918^*

Abstract

The extent of practical cooperation in the business of communications intelligence and communications security between the United States, France, and the United Kingdom on the Western Front has not been documented in depth. This paper will examine cryptologic cooperation between the three allies during the First World War, discuss why the relationships ended after that war, and argue that these impermanent alliances did not shape the cryptologic relationship between the U.S. and the UK that formed during the Second World War.

Notes

* This is an expanded version of a paper presented at the Society for Military History Annual Meeting in April 2014 entitled ‘Precursor to UKUSA? Cryptologic Cooperation on the Western Front.’ The views expressed in this paper are those of the author and do not necessarily represent those of the National Security Agency or the U.S. Government.

1. No attempt will be made to examine the details of bilateral cryptologic relationship between the British and the French. While all parties appear to have had some sort of cryptologic contact with the Italians, no effort will be made to explore that partnership.

2. Warner, Rise and Fall, 59. Many other sources.

3. One could consider that the U.S. entry into the war was partially driven by a massive communications security failure – the inability to protect communications sent via American undersea cable from British cable-tapping allowed the British to intercept the Zimmermann Telegram.

4. For discussion of Navy efforts see Soybel, A Necessary Relationship; for War Department actions see Bidwell, History of the MID. For Van Deman’s visit to the Secretary of War see Gilbert, WWI and Military Intelligence, 28. See as well Woodward, Trial by Friendship for a detailed treatment of the subject of the British relationship and discussion of the Balfour and Joffre missions.

5. Bruce, “A Shadowy Entity,” 6.

6. Ibid., 9–10, 19.

7. Muirhead, “Military Cryptography.” His frequency tables, which he credits to a fellow Army officer and an officer at Leavenworth, would find their way into the 1914 British Field Service Handbook, which heavily influenced the 1917 AEF handbook.

8. “British Naval Officer’s Request for US Aid in Solving Ciphers.” RG 457, Entry 9032, Box 1439. It is not known how this officer knew to contact Moorman.

9. ‘It will be very difficult for them to get a serviceable staff going even in a year’s time.’ Charteris quoted by Trask, AEF and Coalition Warmaking, 17. Frank Moorman would acknowledge later that the G2A6 ‘had the advantage of coming in late and could profit by the mistakes of the British.’ Moorman, “Wireless Intelligence.”

10. Confusingly, the Radio Section of the Radio Division was first named the Radio Intelligence Section. When it became clear that having the same name as the G2A6 was a problem, the name was changed to Radio Section.

11. The Code Compilation Section made the codebooks; however, the G2A6 was responsible for making sure they were properly distributed throughout the AEF.

12. Ferris, “British Army Signals Intelligence,” 27.

13. Nolan, “Final Report,” 1.

14. Krumm, History of the Radio Division. The nature of the training was not specified. Cranwell started life as a Royal Naval Air Training Service facility in 1915 and was given over to the new Royal Air Force in 1918.

15. In this the cryptologic disciplines were not so different from training elsewhere in the AEF. See Hamburger, Learning Lessons, 5.

16. “G2A6 log book.” October 13 1917, RG 120, Entry 105 Box 5769.

17. Moorman to Col. A.L. Conger, 16 October 1917, RG 120, Entry 105, Box 5767.

18. Moorman, 31 October 1917, RG 120, Entry 105, Box 5761. Moorman’s letter details agreement between ‘French and British “Ciphers.”’ This system was likely based on a system used by the British. See also Hitchings to Moorman, 26 October 1917, RG 120, Entry 105, Box 5761, where Hitchings provides a list to Moorman.

19. Russel to Pershing, 12 April 1918, RG 120, Entry 129, Box 5866.

20. Powell to Nolan, 25 February 1918, CCH Series II.F.2.2. This letter was undoubtedly copied from records in RG 120 but has not been located in that source.

21. Barnes, Code Compilation Section, 2. In 1933, well after the war had ended, Nolan told a story about the first U.S., field code being so insecure that the British and French asked that it not be used. This cannot be corroborated in the report of the Code Compilation Section. Nolan may have been referring to the 1915 War Department Telegraph Code, which was agreed to be insecure by all parties. Nolan, “Military Intelligence in the AEF.”

22. “Protocol E.U.F. of August 12th 1918.” RG 120, Entry 765, Box 19 is a communications security agreement between US GHQ and French GHQ from a meeting at Chaumont that day. Signatories were Nolan, Nolan’s deputy Colonel Arthur L. Conger, Moorman, Cartier, Commandant Hue of the French military mission attached to the AEF, and Commandant Soudart, Chief of the Cipher section of French GHQ. It is unknown if similar protocols were signed between the British and Americans or, for that matter, between the British and the French.

23. Colonel Parker Hitt to Nolan, 11 October 1918, RG 120, Entry 105, Box 6699. Hitt (Chief Signal Officer of the American First Army and himself a noted cryptologist) noted that code tables made up for liaison were to be furnished to the French GQG and each French Army Group HQ adjoining the American Army Group.

24. Kahn, Reader of Gentlemen’s Mail, 26. The author has yet to locate material that shows any AEF involvement in the testing of this device.

25. Barnes, Code Compilation Section, 1–2.

26. Moorman to Hay, 20 March 1918, RG 120, Entry 105, Box 5767.

27. Moorman to Hay, 3 September 1918, RG 120, Entry 105, Box 5767. ADFGVX was a complex German cipher system introduced in March 1918 and used primarily between army corps and division headquarters. The French cryptanalyst Captain Georges Jean Painvin achieved a partial solution that relied on key recovery in April; no general solution was found during the war. No September keys were solved until October, explaining Moorman’s plea. NSA, Friedman Legacy, 109–112; and Childs, German Military Ciphers, 13–29.

28. Nolan to Cartier, 3 September 1918; Cartier to Nolan, 10 September 1918; Nolan to Cartier, 13 September 1918, RG 120, Entry 105, Box 5767. Cartier later informed Nolan that the AEF’s intercept was ‘as good as ours’ and that he had no corrections or additions to make to the messages. Cartier to Nolan 15 October 1918, RG 120, Entry 105 Box 5763.

29. Moorman, “Wireless Intelligence.”

30. RG 120, Entry 105 for the correspondence, Entries 105, 2040, 2042 for the visits. After this article was completed, other relevant correspondence was located and is being cataloged and checked for duplication.

31. Cartier (likely) to Hue, 13 October 1917, CCH Series II.F. 2.2.

32. Moorman to G2, 15 October 1917, RG 120, Entry 105, Box 5767.

33. Loghry, History of the Radio Section.

34. “G2A6 logbook.” 13 October 1917, RG 120, Entry 105 Box 5769.

35. Cartier (likely) to Hue, 13 October 1917, CCH Series II.F. 2.2.

36. Ibid.

37. Moorman, “Wireless Intelligence.”

38. “G2A6 logbook.” 1 January 1918, RG 120, Entry 105 Box 5769. The office log provides extensive documentation of this codebreaking exchange until 18 March 1918. At that point, there is a note ‘Log book having not been written up from March 23 to July 1, some of the following was recorded as accurately as could be found out afterwards.’ The content changes drastically at this point and does not contain as much detail about interactions with the French.

39. Loghry, History of the Radio Section; Nolan to Major Hue (rendered as Hugh), 17 December 1917, RG 120, Entry 105, Box 5768.

40. Hulburt, Personal Narrative, RG 120, Entry 2040, Box 130, narrative 197.

41. There were six distinct types of collection stations: field intercept stations which collected radio for intelligence purposes, airplane intercept stations, airplane compass/gonio station, listening stations (telephone and TPS), radio gonio stations, and ‘control,’ or communications security monitoring station which monitored radio. In reality there was some overlap, and listening stations often conducted communications security monitoring.

42. Opcit. Hulburt.

43. Hinrichs, Listening In, 55.

44. Multiple letters to and from Moorman from 18 September to 5 October 1918, RG 120, Entry 105, Box 5767.

45. Moorman to Lt. Charles H. Matz, 14 September 1918, RG 120, Entry 105, Box 5767 and other references mention the use of the Wabash.

46. Barnes, Code Compilation Section, 28.

47. Loghry, History of the Radio Section.

48. Moorman, Report of the Radio Intelligence Section, 18–19.

49. RG 120, Entry 105 for the correspondence and the visits. After this article was completed, other relevant correspondence was located and is being cataloged and checked for duplication.

50. Moorman to Colonel A. L. Conger, 4 March 1918, RG 120, Entry 105, Box 5767.

51. Hitchings, 20 February 1918, RG 120, Entry 105, Box 5761.

52. “G2A6 logbook.” RG 120, Entry 105 Box 5769. The logbook provides extensive detail of the exchange with the French but rarely mentions technical exchange with the British.

53. Many sources, but most recently Beach, Haig’s Intelligence, 163–5.

54. Major Magruder of the G2 to the General Staff Secretary, 8 May 1918, RG 120, Entry 105, Box 5767.

55. Moorman to Nolan, 17 May 1918, RG 120, Entry 105, Box 5761.

56. Hay, Valiant for Truth, 61.

57. Moorman, Report of the Radio Intelligence Section, 18–9.

58. Jenner, “Turning the Hinge of Fate,” 165–205.

59. Richelson and Ball, The Ties that Bind, 137.

60. See http://www.nsa.gov/public_info/declass/ukusa.shtml for the complete set of documents.

61. Captain Abraham Sinkov and Lieutenant Leo Rosen from the Army and Lieutenant (j.g.) Prescott Currier and Lieutenant Robert Weeks from the Navy. Friedman was ill and could not travel.

62. National Security Agency, History of the Signal Security Agency, Volume Two, notes on page 221 that there were four persons who served both in cryptologic units during World War I and in the Signal Security Agency in World War II. The identity of the fourth individual has yet to be determined.

63. National Security Agency, Historical Background of the Signal Security Agency (SRH-1, Volume Two).

64. Friedman, Methods for Solving German Codes.

65. Budiansky, “Difficult Beginnings.”

66. Friedman republished many of the cryptologic reports of WWI as Technical Papers for the Signals Intelligence Service; these were later reprinted in the NSA series of Special Research Histories in the 1970s.

67. Budiansky, “Difficult Beginnings” cites a 15 November 1940 memo from Denniston warning about Yardley whom ‘the Americans sent over at the end of the last war’ for ‘purposes of cooperation.’

68. Ferris, “Road to Bletchley Park.”

69. Mackinnon, “Friedman’s Bletchley Park Diary”; and Erskine, “Friedman’s Diary a Different View.” Friedman’s complete diary is now available here: https://www.nsa.gov/public_info/_files/friedmanDocuments/CorrespondenceMemorandaandPersonnelFileRecords/FOLDER_002/41768289080739.pdf.

70. Denniston, Thirty Secret Years, 76, 78. Denniston and Friedman spent many years discussing the details of the Zimmermann Telegram; Denniston’s son attributes this as a World War I remembrance that bound the two cryptologists, and thus their respective services, together. He makes no mention of any memories of cooperative work on the Western Front although Denniston surely knew of the work of Hay and Hitchings.

71. Smith, “Road to Intelligence Partnership,” 62.

72. Beach, “Origins of the special relationship?”

73. Ibid., 243–345.