Adams Family Correspondence, volume 4


Fleury to Abigail Adams

John Adams' Flawed Key to Lovell's Cipher following 188[unavailable]

Two efforts to decode James Lovell's cipher, a task often resisted and that seemed to Abigail and to John Adams sometimes beyond accomplishment. The cipher was built upon an acceptance by those who would use the cipher that encoding a passage involved substituting numbers (1–27) for equivalent letters, alternately from alphabets (the ampersand was included) in which the initial letters, being agreed upon, provided a key. The letters c and r constituted the key to the cipher Lovell used in correspondence with the Adamses and others, and were thus the equivalents of the number 1 in the two alphabets. Decoding should have been a simple matter in which letters from the two alphabets were substituted alternately for numbers in the coded text. The Appendix to this volume undertakes a full account of ciphers of this type and of the difficulties Lovell's cipher presented to the Adamses.

The first effort is an undated fragmentary sheet, in the hand of Richard Cranch, on both sides of which he has attempted to record horizontally, according to the key provided by Lovell, letter equivalents to the numbers constituting four ciphered passages in Lovell's letter to Abigail Adams of 26 June 1781, below. Cranch has labeled these passages A to D, and having managed the decoding after a fashion, has then written out the transliteration satisfactorily. Although success did crown his efforts, his successive rows of substituted letters reveal one of the sources of the difficulty Lovell's correspondents ex-ixperienced in reading his cipher. Here the difficulty arose from inexact alternation between the two alphabets. As an example: what Cranch calls passage B appears in Lovell's letter as 1-25-10-22-3-11-5-4-3; Cranch's top horizontal row of letters substituted for these numbers reads r-&-&-l-t-a-v-u-t. This substitution, inaccurate as it is, is achieved by beginning from the alphabet in which r is the equivalent of 1 and continuing in alternating sequence from the alphabet in which c is the equivalent of 1. Even with an exact observance of sequence, the result, r-&-&-x-t-m-v-f-t, would have been gibberish, and hence to be discarded. In Cranch's next row the substitution began, properly as it turned out, from the alphabet in which c is the equivalent of i. His reading appears as c-o-l-x-e-m-g-f-e. When he broke the scheme of strict rotation at the fourth character, he was prevented, except by further tinkering, from arriving at the correct reading, c-o-l-l-e-a-g-u-e. Similar carelessness impeded his efforts to decode each of the other passages for Abigail, but it is evident from the fragment that he understood the elements of the cipher. If she shared that understanding, Abigail was perhaps justified in her later statement to her husband: “I have always been fortunate enough to succeed with it” (17 June 1782, below).

The second effort shown is again an undated sheet, this in John Adams' hand. Across the top is an alphabet and at the left is a column of numbers from 1 to 30. From the letters a, c, and r of the horizontal alphabet are hung three vertical alphabets in which those letters are the initial letters. A second column of numbers is hung to parallel the r alphabet. The whole arrangement is one that should have proved helpful in encoding or decoding the cipher. Any success in its use, however, was prevented by a basic flaw stemming from Lovell's failure to explain or John Adams' failure to grasp that the numbers 28, 29, and 30, often appearing in ciphered passages, were baulks or blinds. This misunderstanding is displayed in assigning to numbers 28, 29, and 30, as well as to numbers 1, 2, and 3, the letters c, d, and e in the c alphabet, and the letters r, s, and t in the r alphabet. At any use of one of the baulks, the transliteration would then be fouled beyond correction. Both the frustration felt by John Adams with the cipher from having been able “upon the whole” to make nothing of it, but “able sometimes to decypher Words enough to show, that I have the Letters right,” and his firm conviction after all attempts that “The Cypher is certainly not taken regularly under the two first Letters of that Name [Cranch]” become understandable (John Adams to R. R. Livingston, 21 February 1782, LbC, Adams Papers; printed in The Works of John Adams, ed. Charles Francis Adams, Boston, 1850–1856, 7:521–530; The Revolutionary Diplomatic Correspondence of the United States, ed. Francis Wharton, Washington, 1889, 5:192–199).

From the originals in the Adams Papers.