In the preparation of his farewell address to the American people, GW, as he often did for important documents, sought the assistance of others he thought were superior in literary ability. When he had considered resignation in 1792, he turned to James Madison.1 In 1796 his primary collaborator was Alexander Hamilton.

The starting point for the 1796 address was Madison’s earlier draft. By February 1796, GW had written a new draft that consisted of a brief introduction, a full quotation of Madison’s draft, and a somewhat longer section of additional thoughts inspired by the “considerable changes having taken place both at home & abroad” during his second term. While Hamilton was at Philadelphia in February to argue before the Supreme Court, GW showed the draft to him and mentioned a wish that Hamilton should edit the final document. He did not, however, give the address to Hamilton at that time. On 10 May, Hamilton wrote GW, noting that if he was to have sufficient time to edit, he should have GW’s final draft as soon as it was completed. Upon receiving Hamilton’s letter, GW sent him the draft, by this time slightly altered.2

In his instructions to Hamilton, GW expressed his fondness for the structure of the draft, because he hoped that quoting an address written by Madison, who was now a leading opponent of his policies, might shield him from some criticism. Nonetheless, he authorized Hamilton “to throw the whole into a different form,” requesting, however, that even in that case, Hamilton should revise GW’s original draft to perfect it as well.

After receiving the draft, Hamilton enlisted the assistance of John Jay. The two agreed to meet at Jay’s house to discuss the address. By the time of that meeting, Hamilton had decided to avail himself of GW’s authorization. He used GW’s draft to compile an abstract of points to include3 and then wrote out a new address embodying those points in a new structure without the quotation. Setting GW’s draft aside, Hamilton and Jay considered the new draft “paragraph by paragraph, until the whole met with our mutual approbation. Some amendments were made during the interview, but none of much importance.”4 Hamilton transmitted the result to GW on 30 July.5

As requested, Hamilton then turned to revising GW’s original draft without altering its structure, and he submitted that draft to GW on 10 August.6

After considering the various drafts, GW concluded that he preferred the reorganized document to his first draft or to Hamilton’s revision in that form. He returned the preferred draft to Hamilton on 25 Aug. for further revision, adding, “I shall expunge all that is marked in the paper as unimportant &ca &ca and as you perceive some marginal notes, written with a pencil, I pray you to give the sentiments so noticed mature consideration.” He also requested that any change be clearly marked so as to prevent mistakes “in copying it for the Press.”

Before Hamilton could respond, GW wrote him again on 1 Sept. suggesting the addition of a section on education. Hamilton thought that subject better reserved for GW’s annual address to Congress, and so informed GW in a letter of 4 September. However, when he returned the draft on 5 Sept., “corrected agreeably to your intimations,” it included “a short paragraph added respecting Education.” Feeling too ill to rewrite the address in a fair copy, Hamilton contented himself with marking corrections on the draft he had received from GW.

The draft once again in hand, GW wrote out, in his own hand, the final copy for the printer. On 15 Sept. he submitted that copy to his cabinet, who recommended few, if any, changes and none of substance.7 That same day, he summoned David C. Claypoole, editor of Claypoole’s American Daily Advertiser, and the two made arrangements for the publication of the address in that paper on 19 September. As the final copy was then in Secretary of State Timothy Pickering’s possession, GW promised that it would be delivered to Claypoole the next morning. Claypoole received the copy and later submitted proof sheets and revises to GW, who, he noted, made “but few alterations from the original, except in the punctuation, in which he was very minute.” The address was printed on 19 Sept., but given a date of 17 Sept., for reasons not entirely clear (perhaps GW returned the final revise on that date). After Claypoole published the letter, he expressed a wish, which GW granted, to retain the manuscript original from which the printed version was prepared.8 That document is presented below.

Although the process of composition is quite clear from the existing correspondence, relating that record to the surviving drafts is problematic. The specific problem is with the item that GW sent to Hamilton on 25 Aug. and Hamilton returned on 5 Sept., as there are substantial discrepancies between the surviving version of Hamilton’s reorganized draft and the description of the draft in those letters. Horace Binney, the earliest writer to attempt An Inquiry into the Formation of Washington’s Farewell Address, posited, not unreasonably, that on 30 July Hamilton had sent to GW not the surviving draft, but a fair copy thereof. The surviving draft, Binney thought, was Hamilton’s retained copy.9 He did not know, however, that the surviving draft had on it some pencil notes in GW’s writing.

The existence of those notes (which seemed to fit the reference in GW’s letter of 25 Aug. to marginal notes in pencil) and the large number of corrections on the surviving draft, encouraged Victor H. Paltsits to conclude that Binney was wrong to search for a missing draft when “there is none.”10 Paltsits, however, did not address the discrepancy between those notes and what GW had written in the letter. GW informed Hamilton that there were certain paragraphs marked as “unimportant &ca &ca” that would be omitted and other marginal notes in pencil to be considered. The penciled notes on the surviving draft are present only at paragraphs that were eventually deleted from the address, and they all state only “omitted in address.” Oddly, the final draft that GW sent to the printer includes those same paragraphs (although struck out) with marginal notations (also struck out) that better match the description given to Hamilton.11 Nor did Paltsits address the fact that the many revisions on the surviving draft are purely stylistic not substantive—exactly the sort of changes that Jay remembered being made to the manuscript during his interview with Hamilton.

To believe that there was only a single draft copy sent by Hamilton on 30 July, returned by GW on 25 Aug., and revised by Hamilton in accordance with GW’s suggestions and returned again to GW on 5 Sept. is to believe that GW did not (as he wrote) mark anything on the draft as “unimportant &c &c” although he included such notes in his final copy; that the additional pencil notations were so thoroughly erased that no one who has later examined the surviving draft has reported any traces of them; that in the completely erased notations, GW, who had turned to Hamilton to perfect his language, made suggestions entirely about matters of style (and then made substantive alterations as well as many more stylistic changes in the transition from that draft to the final copy);12 that Hamilton was unable to fit the “short” paragraph on education in the margins of the draft, but wrote it on a separate paper that has now been lost;13 that despite GW’s instructions about marking the draft, Hamilton returned his corrections in a form where their placement is not always clear;14 and that GW, who was concerned enough about the security of the mails to send the draft under cover to Jay rather than to Hamilton, was willing to send by mail his only copy of the preferred text. Surely it is reasonable to consider alternatives.

The hypothesis that best fits the known facts is that on 30 July Hamilton sent the draft that he and Jay had marked up in their interview, without making a fair copy, thus placing that document in GW’s possession. Before responding to Hamilton on 25 Aug., however, GW had a fair copy made on which he wrote the marginal notations he described and which he sent to Hamilton, retaining the original Hamilton draft as his copy.15 This copy, which Hamilton revised and returned on 5 Sept., has not been found. Once GW received Hamilton’s revisions, he made a new copy, incorporating those revisions and any others that he desired but retaining the paragraphs to be deleted with their corresponding marginal notes, so that the cabinet could consider them. When the cabinet suggested no changes, he struck out those paragraphs and submitted that copy to the printer. At that time he also marked the corresponding paragraphs on the surviving draft as “omitted in address.”16

This hypothesis with its assumption of a missing draft does not alter the basic conclusion reached by Binney, Paltsits, and most others who have examined the Farewell Address: the ideas are GW’s, the structure is Hamilton’s, and the language is some indeterminable combination of the two, with contributions from Jay and Madison, all subject to GW’s final approval. It does suggest that the editorial changes on the surviving draft provide no information about GW’s contributions to the address, since they most likely were made before he received that document. More importantly, historians should be reluctant to attribute differences between the final copy and the surviving draft solely to GW. It is more likely that Hamilton altered much of the language in accordance with GW’s intimations on the missing draft. Only the changes marked as corrections on the final copy may be presumed to come from GW alone.