The David Hume Library (review)

Hume Studies Volume XXII, Number 2, November 1996 pp. 383-385 DAVID FATE NORTON and MARY J. NORTON. The David Hume Library. Edinburgh: Edinburgh Bibliographical Society with the National Library of Scotland, 1996. Pp. 156 with 8 plates. ISBN 1-872116- 21-3. Paper £16 UK. Whereas catalogues of individuals' libraries commonly are based upon inventories drawn at the owners' or their executors' instigation for purposes of sale, the Nortons have had to reconstruct Hume's collection speculatively from an 1840 inventory of the books of Baron Hume, the philosopher's nephew. Proof that this inventory included Hume's books is not attainable, but the Nortons skillfully build the case for a very high probability. In his will the philosopher left his sister Katherine her choice of one hundred volumes and his brother John the remainder. Thereafter these two left their books to John's son David, the philosopher's favorite nephew, publisher of the Dialogues, and preserver of the MSS calendared by Greig and Beynon. It would seem uncharacteristic of the baron not to have valued his uncle's books. Doubtless the uncle's books could have been lost somehow, but one must imaginatively invent the circumstances that would have prevented them from showing up in the inventory that the bookseller Thomas Stevenson drew up for the baron's will. Moreover, Stevenson's inventory includes ten presentation copies of books given to the uncle and three items with his markings, among them the first two volumes of the Treatise with "a Vast of Corrections and Additions in the handwriting of the Author" (entry 647). Books that we know from letters were sent to Hume are included, like the five-volume miscellaneous works of d'Alembert (Burton, Letters, 209). Numerous other works known to have been familiar to Hume are in the list. Among books currently or formerly bearing Hume bookplates, Malebranche's Recherche de la vérité does not appear in the list, but ten items do. By the time the Nortons have finished laying out the details, painstakingly and lucidly, the case is fairly convincing that a high number of these books had been the philosopher's. I can add one detail to the accumulation. In "Of the Protestant Succession," Hume quotes from Edmund Spenser's A View of the Present State of Ireland, citing "p. 1537. Edit. 1706." The unlikeliness of the page number is compounded by the lack of any record of a 1706 edition of either A View or of Spenser's collected works. If, however, one consults the 1715 edition of Spenser's works listed in the Nortons' catalogue, one finds that the pagination 384 Book Reviews does not start anew in each volume, that page 1537 and A View fall in volume 6, and that the passage quoted is at the bottom of the page. One apparent problem needs discussion. The Nortons endeavor to locate a copy of every item in the list in a source like the Short Title Catalogue or the catalogue of the British Library. This step is made possible by Stevenson's attribution to each item of a format: 4to, 8vo, and so on. Taken at face value, however, Stevenson's attributions create a high number of ghosts in his own list, items, that is, for which there is no record or known example. The awkward answer is that, as the Nortons explain (n. 127), Stevenson's attributions are sometimes wrong, and for good reason. Booksellers and buyers used the terms for format primarily to indicate book sizes, but strictly speaking format pertains rather to the number and pattern of folds in a printed sheet to create the leaves of a book. Given the variability in sheet sizes, format and the dimensions of books are unreliable indicators of each other. Accurate attribution of collational formulae can be difficult, not to mention unimportant to Stevenson's purpose, which was to indicate whether a book was a pocket volume , a deluxe edition, or what have you. Guessing at formats by dimensions would guarantee errors worrying to analytic bibliographers but not to Stevenson and his customers. Consequently some of the Nortons' exemplar copies do not match perfectly with Stevenson's attribution of...