University of Virginia Library

3. The Evidentiary Basis of the Furbank and Owens Canon

Invaluable though the Bibliography is, its structure and layout necessarily make the radical differences in the quality of the attributional evidence hard to appreciate. Users cannot easily see just how many of the items included are anything but solid and secure. In the appendix I have tried to provide a short-form representation of the entire Furbank and Owens canon, abandoning the "certain" and "probable" division and re-categorizing works according to the nature of the evidence we have for associating them with Defoe. The tabular form of the appendix is an attempt to represent the reduced canon in ways that make explicit, item by item, what the basis for inclusion actually is. This table is not a revision but an extension of Furbank and Owens's work: it is meant to represent their Bibliography in a way that should help us more readily grasp the complexities and limitations of the evidence on which it is based. In the next section I will try to offer some analytic conclusions, but first we need to understand exactly what we are dealing with. Let me reiterate an important point: my object is not to determine the authorship of these works. I have supplied additional evidence when I have become aware of it, but mostly the table is meant to illustrate their canon and its foundation. The aim is to make plain how many different kinds of evidence are at issue. I want both to clarify the nature of the basis for attribution in each case and to represent vividly in visual terms just how much of the Furbank and Owens canon comes from each sort of evidence.

The evidence for initial acceptance of a title into the Bibliography falls into six categories. The shorthand category names are listed here with brief explanations of what each one includes.

  • Claimed by Defoe. This includes items published in authorized collections, as well as those acknowledged by name in letters, in the Review, or in An


    Page 140
    Appeal to Honour and Justice. Because the presence of a name on a title page is not necessarily proof of authorship,18 where the strongest piece of evidence is a title-page attribution the work is labeled "signed" rather than "claimed by Defoe."
  • Signed. That is, a work is attributed to Defoe (or one of his known pseudonyms) on the title page or otherwise within the text.
  • Contemporary attribution (made in Defoe's lifetime or in the first decade after his death).
  • Late attribution (made any time between the 1740s and the end of the eighteenth century).
  • Deduced from context. This includes works that are possibly referred to in other "Defoe" writings, or for which there is some other external referent.
  • Deduced from content. Items in this category are generally consonant with Defoe's views, include passages similar to passages in "Defoe" writings, or contain favorite allusions or anecdotes.

Works belonging to the first category—claimed by Defoe—appear in boldface type in the table. They are the only works for which we have anything like certainty. The other categories are meant to be non-judgmental. In other words, the implication is not that items included on the basis of a contemporary attribution are necessarily more probable than works included because of a late attribution, or that either is preferable to one made on the basis of content alone. This is not a spectrum from very certain to highly doubtful and should not be taken as such. All categories except the first inevitably contain works of varying probability.

A work labeled "signed" means that it names Defoe as author (or is attributed to "the Author of the Review" or another pseudonym), or that it has "Defoe" in the title (e.g., no. 98, De Foe's Answer, to Dyer's Scandalous News Letter). Readers need to realize that Furbank and Owens de-attribute some works with such an identifier, as for example Good Advice to the Ladies ... By the Author of The True-Born Englishman (De-Attributions, no. 46).19 They observe that "many poems were saddled illegitimately upon" the author of Defoe's most famous poem, and so "one is thrown back on one's stylistic intuitions."20 Of Good Advice to the Ladies, they conclude that it does not seem stylistically characteristic of Defoe. Whether such a conclusion is accurate or not, the implication is that works signed with one of Defoe's pseudonyms and included as certain in the Bibliography have been admitted on the basis of their plausibility. What Furbank and Owens tend to say in accounts of such works is simply that there "seems no reason to doubt the attribution on the title-page" (e.g., nos. 68, 74, 75). Given that pseudonyms were often fraudulently applied by resourceful publishers, and that mere plausibility is not grounds enough to make an attribution, "signed" is necessarily less definitive than "claimed."

Having complained about Furbank and Owens's seemingly arbitrary cutoff date, I need to explain my distinction between "contemporary" and "late" attri


Page 141
bution with particular care. Col. Jacque is ascribed to Defoe in 1738 (seven years after his death) by a respectable publisher named John Applebee (c.i690-1750). Not to count that as a "contemporary" attribution seems silly.21 That we should extend "contemporary" to include the 1740s is less clear to me, but in this case it would add nothing. None of the works in the Critical Bibliography is included on the basis of an attribution made between 1739 and the early 1750s.22 After Applebee’s Col. Jacque, the next new attributions come from a 1753 "Life of Defoe" (part of "Cibber’s" Lives of the Poets, now thought to be written by Robert Shiels, on which see below). The three works that are first associated with Defoe by Shiels are Religious Courtship (215), A Journal of the Plague Year (216), and The Political History of the Devil (228). I have labeled those "late" rather than "contemporary," simply because twenty years beyond death seems a significant length of time. Broadly speaking, I regard attributions made in the second half of the eighteenth century as "late."

Of course we need to remember that all external evidence—whether "contemporary" or "late"—is suspect, suggestive rather than conclusive.23 An attribution made in 1714 is usually going to be weighted more heavily than one made in 1753, and both will generally inspire more confidence than one made in the mid-1780s. But even the "contemporary attribution" works, those attributed to Defoe in his lifetime or in the decade after his death, vary in the degree to which they inspire faith or doubt. Robinson Crusoe—a virtually certain ascription—belongs to the "contemporary attribution" category, but so does item 63, Queries upon the Bill against Occasional Conformity, attributed in an anonymous pamphlet to "A Dislocated Hosier." The "hosier" is very likely Defoe, which is why the ascription counts for Furbank and Owens as solid external evidence—but it is also less direct and certain than an allusion to "Daniel Defoe" by a known individual who was well-informed about the London publishing scene. I point out that roughly a fifth of the "contemporary attribution" items were also attributed in Defoe’s lifetime to someone other than him and/or were in some fashion disclaimed by him.

Works belonging to the "deduced-content" category are those supported by no external evidence. They are the weakest attributions, and most of them are considered no more than "probable" by Furbank and Owens. Crudely speaking, these items are those included because they have long been ascribed to Defoe and


Page 142
because they are sufficiently consonant with his general thinking to be considered plausible. Sometimes what looks like suggestive evidence turns out to be unhelpful. The case for including 212(P) depends heavily on its use of a verse quotation that had appeared in the Review (from 16 years earlier), but the same passage appears in A Letter from a Gentleman at the Bath (1722), a more contemporaneous work and one never associated with Defoe. The internal evidence in this instance (as in others) is weak. But "deduced-content" should not simply be taken to mean "Defoe probably did not write this work." I tend to think that Defoe is likely to be responsible for (say) A Letter to Mr. Bisset, though the only grounds for thinking so are that it bears some similarities to the polemical device used in The Shortest-Way and the succession pamphlets and that it is not stylistically or thematically implausible. This is not potent evidence, and to consider A Letter to Mr. Bisset more than a probable attribution would be rash. If I could re-write my "Daniel Defoe as Satirist" article, having taken a harder look at the basis for attribution, I would be a good deal more cautious in my handling of A Letter than I was, however plausible Defoe’s authorship.24 To ignore A Letter entirely is unnecessary, but at present we lack adequate grounds for a confident ascription to Defoe. The case made for the attribution of a "deduced-content" work might be compelling or feeble—but such items should never be regarded unskeptically.

The distinction between "deduced-content" and "deduced-context" is often smudgy. Works that contain allusions to other "Defoe" pieces belong to the former category; works that are alluded to in other texts belong to the latter. The attributional evidence for items deduced from context tends to be a bit more convincing than that for those deduced from content because it involves something extrinsic to the text (and to Defoe’s other "known" writings). But again there is wide variation. The strongest evidence for 140(P) and 141(P), Reasons against Fighting and A Further Search into the Conduct of the Allies, comes from a letter. Defoe explains that he is sending to Harley two works, one meant to make "people ... see How The wholl Nation was Forming into One Tribe of Issachar, and Taught to Couch Under The Tyranny of Our Neighbours," and the other written "in Answer to The Dutch Memorialls" with a similar purpose.25 The subtitle of the A Further Search includes the phrase "a Reply to the several Letters and Memorials of the States-General," and the content of the pamphlets suggests that these are in fact the two letters being referred to by Defoe. The authorship of both is deduced from something outside the text but not explicitly claimed by Defoe himself. The attribution is virtually certain without being absolutely so. One could say the same of 125(P). Several of the "deduced-context" items are works to which Defoe "might have" referred in his letters. The epistolary allusions are frustratingly varied in their specificity. In some cases, Defoe mentions a work more or less by name (see 134(P)), and in those instances I have categorized the work as "claimed." Elsewhere he simply describes a piece he has written or anticipates writing; where an item is supported by these vaguer, more ambiguous references, I have categorized it as "deduced-context."


Page 143

This is in some ways the least satisfactory of the categories because of the wide range of evidence involved. A highly probable epistolary reference is strong evidence, but an item included because it is a continuation of a work Defoe only "might have" written seems far from being a secure attribution. Item 182(P), Some Considerations on a Law for Triennial Pamphlets, is a shaky "deduced-context" case. The best Furbank and Owens can say about this piece is that Crossley and Lee believe it to be the one to which Defoe refers in an article in the St. James’s Post, and that this hypothesis "seems plausible" (CB, 164). The argument for attribution is somewhat bolstered by a favorite allusion—but it is nothing like as compelling as the evidence for 140(P) and 141(P). Ideally, one would avoid categories even as broad as "deduced-context" and "deduced-content" (and never mind "probable"), but the evidence involved in establishing the Bibliography is such that dividing the relevant items into sub-groups would require far too many categories to be useful. We would need groups for "two favorite allusions," for "two allusions plus one favorite anecdote," for "one possible external reference plus one quotation of a known Defoe writing," and so on.

In many cases, Furbank and Owens offer two kinds of evidence—for example, both a contemporary attribution and internal evidence. Almost every attribution could be described as having been deduced from content as well as something else—Furbank and Owens often test external evidence of whatever sort against thematic and stylistic plausibility. To label each item for which this is true "contemporary attribution; deduced-content" would result only in clutter. The category assigned is meant to signal the principal basis for attribution to Defoe. The column at the right of the table explains (in shorthand) the further evidentiary basis where one is offered. The reader is of course referred to the Critical Bibliography for fuller details and explanations.


For example, Furbank and Owens exclude from the canon The Fifteen Comforts of a Scotch-Man. Written by Daniel D'Foe in Scotland (no. 143 in Defoe De-Attributions) as well as The Modern Addresses Vindicated, and the Rights of the Addressers Asserted, by D. De Foe (no. 177).


Another instance is The True-Born Britain. Written by the Author of The True-Born Englishman (Defoe De-Attributions, no. 147).


Defoe De-Attributions, 12.


This is a more or less contemporary attribution, though how much faith we should have in it is something else again. Arthur Sherbo’s ODNB entry on Applebee explains that he "is known principally for his connection" with Defoe, and that he had printed four pamphlets of Defoe’s in the 1720s. This would seem to make Applebee an authoritative source for Defoe—except that all four of those pamphlets were convincingly de-attributed by Furbank and Owens.


An apparent exception is The Present State of Jacobitism Considered, which Furbank and Owens include because the preface is signed "D. F." It was attributed by James Ralph in the 1740s (see no. 28 in the table), but the principal basis for its inclusion in the canon is the prefatory signature.


Even Moore, not known for rigorous skepticism in the realm of attribution, warned that he had not found contemporary attributions to Defoe reliable, "except when the ascriptions were obvious." He concluded that "the eighteenth century had no extensive or accurate knowledge of Defoe." See "The Canon of Defoe’s Writings," 161. For some contemporary misattributions, see pp. 161–162.


Huntington Library Quarterly 70 (2007): 553–576.


The Letters of Daniel Defoe, ed. George Harris Healey (Oxford: Oxford University Press, 376.