A Treatise of Human Nature: A Critical Edition (review)

Apparently, at an early stage of the planning of the Clarendon Edition of the Works of David Hume, the General Editors of the Philosophical Works adopted the policy to produce clean critical texts on pages which bear no sign of the way those texts evolved. For example, in Tom Beauchamp's critical edition of the Enquiry Concerning Human Understanding there is no indication on the text pages themselves of the numerous changes which Hume made to the 11 editions he prepared. These changes, some of which are philosophically significant, are only to be found in the "Editorial Appendix" and are discussed in the Introduction and the "Editor's Annotations." However, scholars would have gained much if they had been alerted to these changes on the text pages themselves. For example, would it not be useful to know when one reads Hume's famous vis inertiae footnote to Section 7 of the Enquiry (7.25n16; 57–58) that when he first wrote it he praised Newton for putting forward the hypothesis of "an etherial active Matter," and only dropped the reference to an "active Matter" in the third edition of 1756? Certainly one can learn of this change and the reasons for it in Beauchamp's learned discussions elsewhere in the critical edition. However, if a central aim of a critical edition is to provide a convenient way of helping readers understand its history then much would have been gained by making clear the author's substantive emendations in different editions on the text pages. This is done in a transparent way by means of a critical apparatus at the foot of the text pages in P.H. Nidditch's Clarendon Edition of Locke's Essay Concerning Human Understanding—and it is unfortunate that the Editors of the Clarendon Edition of the Works of Hume did not adopt this model. The precedent was set by A. A. Luce and T. E. Jessop's standard edition of The Works of George Berkeley, and followed in The Glasgow Edition of the Works and Correspondence of Adam Smith. Hume scholars know how useful Kemp Smith's notes on the text pages of his edition of the Dialogues Concerning Natural Religion can be in establishing the evolution of the text, in spite of problems shown more recently in Kemp Smith's dating.

Since, unlike the first Enquiry and his other later published writings Hume prepared only one edition of A Treatise of Human Nature, the problems of producing a critical text are far fewer. It is true that Hume did publish a list of additions and corrections for Book 1 in the Appendix which he added to Volume 3, as well as acknowledging (but not correcting) two philosophical errors in his original text. He also made mostly minor alterations to two surviving presentation copies of Books 1 and 2, and some philosophically significant ones in his own copy of Book 3. Unlike the two Enquiries, a usable scholarly text of the Treatise has been available since 1978 when the Clarendon Press published Nidditch's revision of L. A. Selby-Bigge's late nineteenth-century edition of the book. Nidditch worked under severe constraints from the press since he had to retain the original Selby-Bigge pagination. In most cases he called attention to Hume's Appendix revisions to Book 1 by way of asterisks in the texts. Thus, for example, we find an asterisk on page 58, line 8 (T 1.2.5.12) of the text and learn from Nidditch's textual note at the end of the book that we need to consult the Appendix on page 636. Here Hume acknowledges his error in claiming that we determine the distance of objects by "the angles, which the rays of light flowing from the bodies make with each other." On page 500, line 2 (T 3.2.2.24...