Text: Burton R. Pollin, “Ends and Means,” Dictionary of Names and Titles in Poe's Collected Works, (1968), vii-xxix (This material is protected by copyright)

[page vii:]

ENDS AND MEANS

Burton R. Pollin

We all have cause for gratitude toward James A. Harrison who brought together so many of Poe's scattered writings and corrected so much of the Griswold text, in THE COMPLETE WORKS OF EDGAR ALLAN POE(New York, 1902), in seventeen volumes. Harrison's purpose was fairly clear in his “General Index to Volumes II-XVI” (16.381-426) — chiefly to list the titles of works that Poe reviewed and produced himself. In a sense it served as a kind of table of contents to the whole set, the first volume being a biographyby Harrison and the last being “Letters of Poe and His Friends.” (Both are excluded from my index.) But Poe's approach to criticism often renders Harrison's approach too limited, especially for the sections called “Marginalia,” where Poe publishes notices of up to four pages next to passing comments of one sentence about old works of literature and personalities in general (e.g., Harrison, WORKS, 16.48-51). These items led Harrison to feel the need for occasionally listing shorter selections and even a few topics, such as “Electricity, Character of” in EUREKA (16.212). Yet an important comment on the “electric light” (16.165) and many references to galvanism are ignored. The basis for Harrison's eventual choice of names is not clear, but the results in the index itself make it. a far from reliable instrument for the students of Poe. For example, Miller, the publisher of Horne's ORION, mentioned only in the headnote to the review (11.249), is indexed, as is the pseudonym Joseph Miller, with a changing middle initial, which Poe humorously devised for his “Autography” series, but missing are Poe's references to the original humorist, Joseph Miller (8.91 and 12.56), to the Reverend Joseph Miller, leader of the Millerites (11.174), to the explorer Joseph Miller (9.226 and 233), to the grammarian Tobias Ham Miller (14.212), to Mr. [page viii:]Robert miller (9.173), and to the Miller in “Mellonta Tauta,” standing for John Stuart Mill. In view of the inclusion of the publisher Miller, about whom Poe says not a word, Harrison's omission of major “Miller” references seems capricious. Equally obscure is the reason for his listing “Old Man” as Governor of Canada (16.411 for 8.319) without identifying him as Francis Head, Lieutenant Governor of Upper Canada; at the same time, he ignores Poe's two references to a poem identical in title — “Old Man” by Oliver Wendell Holmes (12.200 and 15.38). Obviously, Harrison gave little information which was not in Poe's text, save for his filling out a few names for which Poe gave only the surnames.

Other difficulties in Harrison's index soon become apparent to the user. There is no way of distinguishing between a long article or a reference, when Harrison indicates one, since only the first page is given. Thus Horne's ORION is listed as 11.249. In reality it occupied more than twenty-five pages (11.249-75) and was referred to by Poe subsequently four times (12.242, 13.4, 14.233, 15.124), all excluded. Certainly a student of the relationship between Horne and Poe would find the index deficient. A more obvious instance is provided by Homer; there is no listing for the author or for the ODYSSEY and only two references to the ILIAD. In reality, there are thirty-three separate references to Homer or his works, several of great interest and importance to students of Poe's ideas on criticism. This apparent principle, of ignoring passing references or what might almost be called Poe's obiter dicta on literary works and figures, leads to the omission from the index of major names from Poe's body of writings, such as Rabelais (ten references), Fielding (three), Fichte (four), Cervantes (sixteen), and Richardson (three). Several that are included need considerable augmentation; e.g., Beranger's four should be ten, and William Godwin's three should be ten. [page ix:]

Occasionally the index maker for Harrison's edition fell into curious errors as when he attributed MISERRIMUS by Frederic Mansell Reynolds to W. G. Simms, apparently because Poe speaks of Simms's vehement denial that. his own MARTIN FABER was derived from MISERRIMUS. While the text of Harrison seems generally an adequate printing from the originals in periodicals and other sources, an occasional error was allowed to enter; e.g., Poe scorned the Philadelphia charlatan Hague, whom he roundly denounced as a false “weather prophet” in ALEXANDER’S WEEKLY MESSENGER for April 29, 1840 and he takes a passing blow at him in the “Marginalia” in GRAHAM’S MAGAZINE of March, 1846 (16.91), but Harrison's text prints his name as Hugo. The index therefore speaks of this as a “sarcastic allusion to Victor Hugo” and yet omits five references to Victor Hugo actually made by Poe. I do not imply that Harrison's text is to be completely distrusted. All Poe readers must remain grateful to him for establishing so relatively reliable a text so early in the history of Poe studies and for giving a partial key to it.

Some of the errors in the text go back to the original pages of the tales and criticism and some, of course, go back to Poe's errors about names and titles or frequently to his casualness about references, as when he speaks of Tuckerman's SKETCHES OF TRAVEL instead of his RAMBLES AND REVERIES or Sanderson's SKETCHES OF PARIS instead of AN AMERICAN IN PARIS. Occasionally a subtitle becomes the main title in Poe's text — SICILY for ISABEL: OR, SICILY. A PILGRIMAGE or LOUISIANA for HISTORY OF LOUISIANA. The LADIES’ COMPANION becomes LADY’S COMPANION; W. L. Stone's MEMOIR OF RED JACKET becomes LIFE AND TIMES OF RED JACKET and his MARIA MONK AND THE HOTEL DIEU NUNNERY becomes “Pamphlets concerning Maria Monk.” Certainly the public did not demand more accuracy of Poe, and considering the enormous range of his allusions and quantity of his work for so brief a span of productive years, we must have respect for the appositeness and usual accuracy of his [page x:]references. Our purposes now, as students of Poe, from a broader perspective than that of his contemporaries, demand an index which is more detailed and amended with reference to names and titles.

If I may become confessional for a moment, I had no notion at the inception of this project that it would involve so much effort. I envisioned a reading of the complete text of Poe's work, i.e., volumes II through XVI, with red-ink pen in my hand for making various scorings under names and titles, to be subsequently transcribed by a few amanuenses onto cards. These would then be quickly sorted, alphabetized, and typed on sheets for an eventual printing. I was rash enough to believe that this procedure could be fairly easily and quickly executed and that a publisher would be willing to issue the results in what I thought would be a pamphlet of about a hundred pages. The project, finally completed, did follow my initial plan, but required far more time and effort than estimated. My first idea (changed only by the advent of the computer) was to divide the whole into two sections: the names discussed by Poe and the titles mentioned or reviewed. In order to keep the quantity of material manageable, I arbitrarily ruled out geographical and national names, save for a few exceptions. For example, the tribe of Judah (14.2) might be included with reference to the name Judah, but not with reference to its being a national unit. Similarly, to one unfamiliar with the city of Abdera, the name Abderites might look like a sect, such as the Transcendentalists; this possibility justified my including it with a parenthetical note, but it would not justify Parisians or Bostonians. On the other hand, Poe's ingenious coinages, even for inhabitants of cities, such as the Frogpondians (for the despised literary folk of Boston) would require inclusion.

Despite exceptions, my guidelines for names were fairly clear: the names of real persons, [page xi:]including authors; all mythological figures; all characters used by Poe in his own works, including tales, poems, and the play, “Politian”; and all characters mentioned by Poe in the works reviewed. The long and often tedious summaries of plot revolved around characters who, I assumed, might sometimes have given suggestions to Poe for names used in his own work (cf. “Lenore”). In any event, these proper names I regarded as necessary inclusions.

The titles raised several problems. Users of the Harrison index know how irritating and even frustrating it is to find no differentiation between books and poems, everything being included in quotation marks. Moreover, since there is only one continuous index of names, titles, and topics, the eye is bewildered by being forced to jump hack and forth for the initial quotation marks. The virtue of alphabetization is well-nigh cancelled, and many items are overlooked when they are indented one space from the lefthand margin of the text. To avoid this, I decided that the differentiation between books and short-item titles must not interfere with an alignment of all entries. Capitals for book titles offered a convenient way of indicating books differentiated from titles, poems, tales, and articles. By my removing the names of persons to a different section, the distinction would then become clear, for every entry which is printed without quotation marks at the beginning of the line automatically is to be accepted as a short-item entry. However, within the text of the entry itself, if there is cause to refer to another form of the poem, for example, or to another short-item quotation marks are used, since the alignment is not in question.

A major problem in indexing the material, however, has been hinted above, in my reference to Horne's ORION. Many of the items could have inclusive pagination, when the subject of a review article was the whole work or, as in the articles in the Literati Papers, a single person. However, [page xii:]what policy should be followed for references that might be made here and there in the course of a long article? I decided to apply a bibliographic system which I used for my GODWIN CRITICISM: A SYNOPTIC BIBLIOGRAPHY (University of Toronto press, 1967), in which all the 3,400 abstracts which I presented to the reader dealt with William Godwin but were derived from a wide variety of sources. In order to furnish the user of the ten indices included in the large volume with preliminary information about the nature and scope of the material, I added to each entry number a letter designating the type of material originally abstracted; e.g., A for an article, or P for a passage in a book, or M for a significant mention in some work not wholly about Godwin. The reader might then exercise choice about which entries to consult.

In the Poe index, the source of the material was no problem, since it all lay in Harrison's text (with certain exceptions which I shall discuss below). My problem was to set up a system of code letters which would signal the reader about the extent and type of the material relating to that name or to that title. I therefore adapted my letter code from the Godwin bibliography as follows:

	1.		“a” for an article by Poe on a book or person or a subject
	2.		“c” for a comment of at least a sentence or for an implied opinion about the name or title
	3.		“f” for a character in one of Poe's tales or poems
	4.		“fq” for a fictional character in the work under discussion
	5.		“j” for a periodical or journal discussed by Poe [page xiii:]
	6.		“m” for a mere mention of a work or person
	7.		“p” for a passage, usually of a few sentences but occasionally of a few paragraphs
	8.		“q” for any name or title quoted by Poe from another source in its context
	9.		“t” for the title of one of Poe's tales or poems

Combinations with “q” are therefore always possible, for Poe may reprint a mere mention of a name or an extensive discussion. Three letters in combination are possible for a journal or periodical reference, as one may see by looking at the journals beginning with NEW YORK. Most of them are MJ or the mere mention of the journal, but at least one is a discussion of the NEW YORK SUN (15.126-7pj). Since a mention may be briefly made of any title or name in a passage quoted, it is possible also to find a journal which is followed by MJQ, or mention of a journal quoted. Usually, of course, there is only one letter reference to a number, although occasionally one may find that an item appears in a quotation on the page along with Poe's comments, in which case both items are listed for the page, one with a letter such as “p” and the other with a letter plus a “q.” I should, perhaps, indicate now that I preferred using small letters as detracting less from the dominance of the numbers, since the volume and page numbers should be the chief focus for attention in the index. It is obvious to all readers, I believe, that the volume number appears to the left of the period and the page or pages to the right for each reference.

Before embarking on my actual notation, I thought it best to determine which of the articles in Harrison's text are no longer considered to be [page xiv:]Poe's productions. It would be totally misleading to index names that are from another's pen; Professor Thomas Ollive Mabbott had previously let me know about his exclusion from the canon of Poe's works of one or two items, and Jay B. Hubbell mentions several in his chapter on Poe in EIGHT AMERICAN AUTHORS (pp.15-16) I therefore took the problem directly to Professor Mabbott, always so obliging in his aid to students of Poe. He carefully sifted for me the articles which need exclusion, as follows: 7.228-250; 8.1-2, 12-41, 265-275; 10.96-114, 100-162, 171-174; 11.115-123, 127-131. One item I find myself unable to exclude, partly on the basis of Professor Mabbott's earlier statement recorded in Quinn's biography, EDGAR ALLAN POE (p. 354, n.), namely, the review of Griswold's THE POETS AND POETRY OF AMERICA, which appeared in the PHILADELPHIA SATURDAY MUSEUM, 1843 (Harrison, 11.220-243). The indexed material from this very article tends to confirm the Poeian nature of the text, since several of the allusions are repeated in other articles. Moreover, Griswold's animosity toward Poe can be more easily explained on the basis of Poe's writing this highly satirical article than on any other grounds. Lastly, I find that the style of the entire article, which Professor Mabbott now would ascribe largely to Henry B. Hirst, is characteristic of Poe's and is at variance with Hirst's normal style, as I have seen it in brief writings. (I must admit to having made no thorough study of this point, as I hope some day to do.) In any event, since there is a consensus that some if not all of this article sprang from Poe's ingenious brain, I prefer to index it along with all the verified material. The wary reader may wish to make his exclusions in terms of the volume and page numbers of this piece.

I should now indicate the manner by which the actual indexing was performed. In my set of the Harrison edition of Poe I scored in red every name and title to be indexed, using a single underline for names, and a single underline plus a left hand [page xv:]perpendicular for book titles, and two perpendiculars, both right and left, for short-item titles (poems, articles, tales). My reason for the latter was to make them correspond more closely to quotation marks, so that the title, transcribed onto the index cards could approximate this two-fold punctuation mark form. Moreover, as users of Harrison's text know, the close lines of the printing there required a kind of double identification of entries to be extracted by the scribes, as I shall term my aides. The designating letters, such as A for article, or P for passage, or C for comment, I wrote in red ink OVER the name or title, usually over the alphabetizing part of the name where both names were given by Poe. Since much information had to be supplied, I used the margins at the sides and the bottom for inserting additional material, such as the author implied by Poe's reference, e.g., to PARADISE LOST or HAMLET. The margins of course were necessary for corrected titles, although many of these were corrected only after all the material had been typed. I should here explain that implied authors have invariably been added to the list of names, even though Poe did not mention them, and even though Poe could not have known them in the case of anonymous works, such as MEPHISTOPHELES IN ENGLAND by Robert F. Williams or Sir William Henry Sleeman's RAMASEENA (which was misprinted). This perhaps has slightly falsified the picture of the citation of names, by Poe, but it seemed to me a necessary addition, in view of the wide area of overlap here, as when we refer to the book of Isaiah or to Isaiah, to Proverbs or to Solomon, to reading Dante or reading THE INFERNO. For the vast majority of Poe's allusions I feel it safe to assume that he was aware of authorship and implies it in his use of a specific title. I must also explain here that continued reference to a name, for example, an author in an article of which he forms the major figure, such as a review or a paper from the Literati, makes it unnecessary to index that name for more than one place in that article, since the [page xvi:]letter R or A automatically means that the pages are inclusive: similarly for characters (F or FQ). On the other hand, if a reference to a figure is made in the course of several pages, but with varying emphasis by Poe, I have indicated that one as a series of page references with the same or with differing letters, as, for example, with God in EUREKA (16.203-315). (To save space I have listed only the changing page, save when it follows a zero, in which case the changing digits are both indicated.)

There now followed a labor for which I cannot adequately express my gratitude, in view of the speed and accuracy with which it was done. My helpers, under the able and zealous direction of Mr. Redmond Burke, transcribed onto more than 10,000 cards all the indicated names and titles, placing the volume-page-letter in the upper right hand corner. I used pink cards for titles and white for names, as a quick means of differentiation. As my annotations for each book were transcribed onto cards, they were filed into fifteen boxes and prepared for verification. The number of cards would have been almost double this figure had I not relied upon the device of indicating continuations on each page, through a plus mark for a repeated name or title. Thus a name might be listed on a card with the notation; 8.4-5-6-7c. Upon completing the entry, the transcriber would then cross it through and write 8.4-7c for the name mentioned. Titles were written on the cards in two forms, either with or without quotation marks; those without them, of course, were the books, designated by only one perpendicular. While it was still necessary to transfer ambiguous titles to the names section and vice-versa, as late as on the printed sheets, I still regard it as a marvel of devotion and care that the transcriptions were so generally accurate and faithful. Mr. Burke ingeniously ruled a solid black stripe across the top of the whole set of cards in the box, upon completion. The spacing from the margin was different for each of the [page xvii:]fifteen filing boxes, so that we might be able to “place” or locate the correct volume number of any card that had been incorrectly transcribed, since the black dot left on the edge of the card enabled us to identify the volume from which the entry had been copied. A further identification could have been made through a diagonal line ruled along the side of the whole set of cards to enable us to locate approximately in which portion of the book according to its sequence position in the “deck” each card belonged. Three or four “lost” cards or entries were thus located upon our final check, long after all the boxes had been merged.

The next stage was the verification of the transcriptions. All items had to be checked by reading aloud from the book to the cards, preferably by an aide who had not worked on the book being verified. In many cases I used aides entirely new to the project, and always the verifiers worked on boxes different from those which they had transcribed.

The next step consisted of assorting all the cards into two sections of titles and names and then merging these into a continous [[continuous]] alphabetical listing for each type of card. The large number of foreign names used by Poe had rendered the treatment of the alphabetizing name a bit difficult. For titles all preliminary articles were either dropped or, for very short titles and for Poe's titles, retained after the title itself. This meant the need to insert commas and to preserve the capitalization on “a, an, the, le, la, les, etc.” For names, we faced the problem of alphabetizing for such phrases as “de la Rochefoucault” or “Le Brun.” In all, save British names, we tried to follow the rule of using the “Le” or “La” as the alphabetizing word, with “de” left behind the main name: “L’Enclos, Ninon de” or “Audiguier, Vital d’.” German “von” also had to be cut away from the alphabetizing section of the name, unlike the Dutch “van,” but this raised a problem for us in Poe's character of Von Kempelen, [page xviii:]as distinguished from the real “von Kempelen.” Here, I arbitrarily decided to index the fictional character under “Von” and the real one under “K”. I should add, perhaps, at this point, that where the same name is used by more than one author, a parenthetical annotation serves to differentiate one from another so that they are no more merged together than would be two different persons named John Smith. An exception here, of course, would be an historical figure used by two different historical novelists for the same episode from history. Complete consistency, I must admit, was not possible for the many historical figures mentioned or used by Poe and his contemporaries and predecessors, such as Shakespeare. Titles too had to be differentiated, lest all the “Lenores” merge into one. (See “Lenore” for a good example of this differentiation.)

After all the cards were alphabetized, we prepared summary cards of similar items so that the typists could work accurately and rapidly. This entailed our clipping together the same name or title while checking very rigorously all alphabetizations of both titles and names. Here we discovered many discrepancies introduced by Poe or by our own faulty transcription. Poe often cited names and titles from memory without verification. Thus Bryant's poem was given as “Massacre at Scio” (correctly) in one place and “Massacre of Scio” in another. It was my obligation, at this point, to write a new summarizing card that would explain the reason for both forms in some parenthetical manner or to make a correction for the erroneous citation that would neither mislead the reader nor do an injustice to the Harrison text. Problems had to be worked out individually. Sometimes a cross reference card would manage it, as for Poe's common form, “Auncient Mariner.” Names that were markedly different had to be explained, such as LUNAR PLANISPHERES for Poe's LUNAR GLOBE, usually through a parenthesis following Poe's form, which would be first consulted by the reader, I assumed. [page xix:]If Poe's form referred to a major figure or work, that had a different place through alphabetization of first or second words, I had to decide whether to refer the reader through a “see-also” reference in the parenthesis or through a completion of the entry plus a “see-also” reference or merely through inserting the identifying data (volume, page, letter) in both places. Thus William Godwin had many entries, which were altered to include the obvious misprint of “William Goodwin,” on 9.265c. This might be called silent cross referencing. I frequently engaged in this process for an item that had two instances under one variant and two in another, in order to avoid a cross reference, since the parenthetical remark indicated the existence of both forms. On the other hand, considerations of space sometimes demanded “see-also” references for long lists prepared under each form; the best indication of this is God, which also included “Adonai, Almighty, Christ, Deity, Creator, Deus, Jehovah, Lord, Savior.” Apart from the question of the identity of all these terms, it appeared to me useful for many types of students to be referred from one to another of these names. Hence, a “see-also” card was prepared for each of the names, referring the reader to all the other. Reprinting the entire list of pages and volumes under one of these names would be slightly misleading and costly of space. Occasionally I prepared “see-also” cards, where none might be demanded, at first. My own studies in Poe suggested perhaps an otiose inclusion of such matters. For example, I regarded it as useful for an investigator of Poe's variations in his tales, to know the changes that he made in titles. Where the variants are in the subtitle, no further reminder is needed, as with “Loss of Breath.” Where the title changes entirely, I have assumed that the reader might find it useful to have a double or even triple signpost for consideration, as with “The Assignation,” published also as “The Visionary”; or “Shadow — A — Fable,” published also as “Shadow; A Parable” and also as “Siope — A Fable [page xx:](in the manner of the Psychological Autobiographists).” In doing this, I found it necessary to delve into the note section of Harrison's edition of the tales (and also of the poems); no material in the notes that had been indexed in the main text was drawn from this section, but all variants of the titles were included. In addition, I included any textual variations, of which there were many in “Bon-Bon” and the “Duc de l’Omelette.” A reference to the DIAL, for example, was originally a reference to “The Sorrows of Werther,” the latter not appearing in the text of the story given by Harrison. Only in this way, I felt, could the sweep of Poe's allusions be specifically and accurately given.

On the other hand, my rigorous restriction to Poe's material as printed (or written, in a few instances, such as “Introduction to Folio Club Tales”) and prepared for printing made it necessary for me to leave out the well-known names of tales printed before the texts collected by Harrison, e.g., Poe's use of “A Decided Loss” for the 1832 version of “Loss of Breath,” or “The bargain Lost” for the later “Bon-Bon.” Harrison's careless omission of Poe's pseudonym of “Lyttleton Barry” for four out of his five tales, so authored, meant my almost foregoing a very provocative name, but that one inclusion enabled me to list it for the others in a parenthesis. For a few of the notices by Poe in the SOUTHERN LITERARY MESSENGER and in GRAHAM’S MAGAZINE, omitted by Harrison, there must be a complete absence of interesting names in my own index, but I must accept this limitation. My consolation is the availability to all readers of the material indexed, in view of the many reprints of Harrison's text since 1902.

The final stage of the index preparation entailed an even more thorough search into biographical and bibliographical sources than I had previously performed, as well as a more extensive examination of the text, e.g., to [page xxi:]separate the various “Lenores,” the various Smiths, Alberts, Lees, Millers, and even such names as “Ravenswood,” which turned out to refer to two different characters in two books reviewed by Poe. It would have been impossible to record all of these parenthetical additions on the pages of the text in the first place, before transcription onto the cards. Copying the material from page to cards and then to typed copies also would have entailed numerous errors. The sources used for checking were varied, thanks to the collections generously placed at my disposal by the libraries of New York University, Columbia University, and the New York Public Library. I chiefly relied upon Hyamson's DICTIONARY OF UNIVERSAL BIOGRAPHY, THE NEW CENTURY CYCLOPEDIA OF NAMES, Thomas's revision of Lippincott's UNIVERSAL PRONOUNCING DICTIONARY, THE DICTIONARY OF NATIONAL BIOGRAPHY, THE DICTIONARY OF AMERICAN BIOGRAPHY, the eleventh edition of THE ENCYCLOPAEDIA BRITANNICA, Watt's BIBLIOTHECA BRITANNICA, the catalogues of the British Museum and the Bibliotheque Nationale, and THE ENGLISH CATALOGUE. Recourse was had also to Allibone's CRITICAL DICTIONARY, Odell's ANNALS OF THE NEW YORK STAGE, Allardyce Nichols's HISTORY OF ENGLISH DRAMA, Wright's AMERICAN FICTION, 1774 TO 1850, Halkett and Lang's DICTIONARY OF ANONYMOUS AND PSEUDONYMOUS ENGLISH LITERATURE, and other specialized reference works, such as the 1850 CATALOGUE of the Philadelphia Library Company. For the undisturbed, monopolizing use of Hyamson and the NEW CENTURY CYCLOPEDIA I am grateful to the Bronx Community College Library. Despite my wealth of reference aids, I have found it necessary to insert question marks or alternate forms in the parentheses of a few items to indicate my own doubt about names and titles given by Poe in an abridged or vague form such as Gluck, a printer, or Lee, a composer. Rationalization of forms had to be made; e.g., Poe's references to a form of EL MIO CID, usually the French play LE CID but once a German rendering of the Spanish work by Herder, were all listed under THE CID. [page xxii:]

At this point too I had to devote much attention to deciding whether a title was a book or a short-item (poem or short story). Often Poe, in reviewing a book of poems, devotes most of his space to the title poem. Here I indexed the title under both forms, usually adding any existent subtitles to the book for differentiation, as with MELANIE AND OTHER POEMS by N. P. Willis. Poe's context does not always make it clear and the catalogue of the British Museum and THE ENGLISH CATALOGUE often helped to solve the problem. Keeping titles as short-item designations entailed delicate decisions about capitalization, not undertaken for books, which appear as capitals entirely. I tried to follow the punctuation of Poe's original references, with all the vagaries of capitalization. In general, words considered to be important, usually nouns, are capitalized, but the practice was very flexible then and my text attempts to be faithful to Harrison's text. Occasionally, I consulted a standard printing of a poetic title in a source such as Granger's INDEX or else used our modern rules of capitalization for an entry. Once again, perhaps, I should add that quotation marks were used for short-item titles within parentheses even though they were not used at the beginning of any short-item entry where they would interfere with the alignment of the text. Of course, long works, such as “The Narrative of Arthur Gordon Pym,” published as a tale by Harrison, may still be considered as short-item titles. In deference to Poe's estimate of EUREKA and his printing it as a book of 143 pages, which he considered his major opus, I have listed that title as a book.

Finally, after all the cards were ready and verified, I divided them into three sections for the typing, which was generously provided by the Bronx Community College, as I shall indicate in greater detail. Less than a week later, my staff of aides were ready to check the cards against the typed material, prepared in approximately 250 pages. In this form it was easier to discover [page xxiii:]discrepancies in alphabetizing or in listing. Suddenly, for example, we would discover a French name such as “de la Breterie” which had been indexed under “de” instead of “La,” or a “Le Brun” which appeared under “Brun” by confusion with the titles, wherein a L’ETOILE had been listed as ETOILE, L’. Similarly, our underlining of all the book titles took place at this point in the checking, since the typists had been instructed to treat all titles alike, regardless of their bearing quotation marks or not on the pink cards. It was at this stage that we discovered titles which had been recorded as poems instead of books; after all, there was only one short lightly inked red perpendicular line differentiating one from the other in the tight and dense print of the Harrison edition. The ambiguous items once again cropped up, sending me to the various catalogues for a final if not exactly a definitive answer, since authorities are still not in agreement as to when a separate printing of a brief work, such as a poem, constitutes “book” publication; moreover, what should one do with maps or with Brewster's “SELENOGHAPHY,” mentioned by Poe in a note? These problems were, to be sure, rare. At the same time that we underlined the works and referred to the pages of the Harrison edition (now bruised and worn by weeks of almost incessant use), we discovered oversights in the merger of pages and occasionally in the listing of sequential items, such as a mention and a comment occurring on the same page. A little more difficult to correct was the tracing of anonymous works, such as THE CHRISTIAN FLORIST, MEPHISTOPHELES IN ENGLAND or BUBBLES FROM THE BRUNNENS OF NASSAU, for each one required alteration both in the Title pages and in the Names, with the insertion of “new” authors’ names and, sometimes, new cross references. The last check was for all cross references, to guarantee that two forms of names, such as those of Greek mythological figures (Pallas and Pallas Athena, Phoebus and Apollo) were properly cared for. In a few cases, by inserting all inclusive pagination for both forms, cross referencing could [page xxiv:]be minimized (cf. Anacreon and Teian) although generally speaking the insertion of the abbreviation “q.v.” solved many problems of this sort. I attempted to preserve a distinction between “see,” “see-also” and “cf.” used for items of comparative interest but not of identity under a different form. For example, Mr. Gliddon, in Poe's tales (6.118) undoubtedly was suggested to him through his having reviewed ARABIA PETRAEA (10.1-25) in which work he appears as consul at Cairo (10.19-20). “Ramsbottom way” seems to have reference to an item in SKETCHES BY BOZ, to which attention is pointed through a “cf.” The comparison here, of course, is not made with material in the text of my index, as is true always of the “see” and “see also” references.

Here I should perhaps explain my use of “Poe's coinage” and “sic.” In the following index, the latter does not refer to an error, since printing slips or errors by Poe or Harrison have been specifically corrected through a short parenthetical note (see Albano, Aschen-possel, and Hossack). “Sic” means, rather, an ingenious but startling or unusual coinage by Poe of a name or title; it is a short way for pointing attention to Poe's verbal dexterity or whimsicality, e.g., “Melty (sic for a Fury)” or “Snob (sic).”

With the final establishment of the text, revised and corrected, the next decision to be made concerned the form in which the material was to be presented to the publisher. Everyone who has worked with typesetters knows the dangers invited by a transcript of about 250 pages, planned for a book or [[of]] 200 or more pages, when the pages are full of names, titles, parentheses, figures, and inserted letters. It had appeared best from the beginning to plan to prepare typescripts of pages that might be photographed and slightly reduced for offset printing, page by page. Thus the proofreading and revisions would be the responsibility of the author, but the accuracy might be more easily guaranteed if a good [page xxv:] typing service could be used. I had thought that an IBM executive typewriter or selectric machine would yield a copy suitable for printing, but by the time I had finished with my many revisions, insertions, and transpositions on the typed pages, I became apprehensive that two separate (and expensive) typings would be necessary. At this point my expert and knowing wife, Professor Alice M. Pollin, Professor of Spanish at New York University, made a most helpful suggestion — that the whole text be keypunched under the auspices of the Institute for Computer Research in the Humanities at New York University. Her two computerized texts, GUIA PARA LA CONSULTA DE LA REVISTA DE FILOLOGIA ESPANOLA and LAS CONCORDANCIAS DE LA POESIA DE EUGENIO FLORIT (NYU Press, 1964 and 1967) had apprised her of the capacity of computer machines in projects of this nature. Moreover, as Coordinator of Research in the ICRH, she was in exactly the position to interest the Director, Dr. Jack Heller, in arranging for computerizing the text. He most generously looked into the matter and offered the facilities of the Institute for this final stage, which involved keypunching, programming, and printing of the entire text. After the keypunching of the text had been initiated through the experienced fingers of Mrs. Esther Zilberstein, who had prepared several other texts, including the Florit concordance and much of my own GODWIN CRITICISM, of 1967, Mr. Gary Berlind of the Institute offered to relieve Dr. Heller of the task of programming the text as his share of the project. His analysis of this phase of the work will be separately given below.

The advantages of computerizing the text are many. First of all, corrections and revisions of the text can be almost instantly made without recourse to the whole tedious process of telling the typesetter or a typist in a commercial office remote from one's own center of operations what changes should be made, after which one must wait for a time to see the revised copy. The computer [page xxvi:]editing program enables one to correct the disked material at once and see a new printout of any revised lines or sections, in their new guise. The computer, at New York University, now equipped with upper and lower case letters, very much resembling ordinary printed type although still lacking accent marks, can provide pages which are headed, numbered, and given guide letters for quick reference purposes. The material was being keypunched even before the decision had been reached as to the size and exact format of the final pages, since the inserted denotators (providing merely for separation of entries and for capitals for initial letters and for book titles) could take care of any style of printout that we might desire. I had envisioned a revision even of the text in regard to spacing the various volumes apart through “letting out” a double space after all semicolons (between volumes) in the volume-page section of each entry. This would, I knew, be an easy phase of the program for the text, as proved to be the case.

Mr. Berlind and I discussed the needs of the text in a preliminary session and then waited for production of the first half of the text (the names) to enable me to proofread the material when printed out while the titles were being keypunched. Using two keypunch operators would have shortened this phase of the operations, but I knew that Mrs. Zilberstein was best qualified to prepare the cards from such an “interlarded” typescript as mine, and the result has justified this confidence. Mr. Berlind assured me, also, that my initial inference was correct — that very little extra programming was needed to provide features of the text that would have been impossible with ordinary typing. I refer to the two supplementary indices, of Poe's titles and of Poe's fictional characters, both much needed, by students of Poe and to he found in no other source separately printed. Since all his titles bore the letter “t” and his characters the letter “f” the computer could easily be instructed to “search [page xxvii:]out” these entries and print them separately, as has been done for this text. Moreover, statistical data on such matters as the number of reviews, of articles, of separate letters (hence of separate allusions when totaled) could be quickly ascertained by the machine; this has been done, as the last section indicates. Without question, the computer aid, given through the outstanding facilities, cooperative spirit, and expertise of the ICRH, has greatly facilitated the production of the book and enhanced its accuracy, completeness, and breadth of coverage.

It was only at the end of the disking and extraction process that I had more than a vague notion of the specific figures for various aspects of the index, a few of which may be of interest to the reader. The first index (of names) contains 3,409 entries: that of titles, 2,263, for a total of 5,672. The separate clusters of volume-page-letter groups are 6,883 for the first index and 3,512 for the second, or 10,395 in both. This does not present the true picture of Poe references, however, since many of the clusters represent merged references for successive pages. On the other hand, several references are derived from excerpts, quoted at length, and may have less significance. I should judge that the total number of separate references to titles and names in the text of Poe is about 13,000. Assuming an average of two hundred fifty pages per volume in the Harrison edition, we note that there are 3,750 pages of material. The over-all average, therefore, is three and one half references per page. Since volumes II through VI consist of his tales and, VII of his poems, neither being as rich in references as his criticism, we should assume that the great bulk of his references derives from volumes VIII through XVI, or from his criticism. The statistics given at the end verify this assumption. Since Harrison arranges the volumes not only according to the nature of the material but also roughly according to the date, the variations between the volumes of his early and [page xxviii:]later tales, his early and later criticism might repay investigation. Is it simply that Poe's scope of allusion developed in view of his retentive memory? Did he increasingly feel that more allusions gave him a seemingly greater authority? Was his critical method changing toward a greater reliance upon the citation of examples? Many more possibilities may be plumbed, with the evidence or clue of the following text in hand. Other figures, given in the section of statistics, may strike the reader as interesting, such as the rather long list of 310 fictional characters in Poe's works in view of the many stories in which he is the unnamed narrator with a cast of anonymous figures, as in “Shadow” or “The Tell-Tale Heart.” I should warn the reader, at this point, that the number of titles is much greater than the number of works, since not only are the final forms of the titles of tales and poems given but the preliminary versions, in more than one earlier printing by Poe, are also listed. The fictional characters’ names are also weighted or determined, to a certain extent, by my judgment of whether a name is that of a character or simply a personality who “passes through,” one might say, to be labeled as “f” on the one hand or “m” or “p” on the other. These instances, however, do not markedly affect the total. The names section also includes “r” designations, because the authors of books being reviewed are thus most appropriately labeled, it seemed to me. The “j” references, numbering 445, include many repetitions, of course; nonetheless, the high figures reminds us that Poe considered himself to be a “magazinist” whose attention was often devoted to the journalistic productions of his contemporaries. Some meaning may be derived from the proportion of “m” and “p” and “c” entries to each other in each section and between sections; I do not wish to deny the fact, however, that many others would categorize references differently. Yet I hope that there will be, provided, in the index as presented, a series of objective sets of data which will enable us to be more specific in our [page xxix:]views of Poe as critic and creator.

One last but basic problem in the issuing of the book remained to he solved, determining the publisher; and the solution came only when half the text had been keypunched. I should explain that the idea of doing the index came to me first when I had heard about the plan to issue a new publication, THE POE NEWSLETTER, under the auspices of the Washington State University at Pullman, Washington. Professor G. Richard Thompson, the editor, in his enthusiasm and foresight, rose quickly to the suggestion that the index might appear as one or two supplements to this new publication; our exchange of correspondence occurred before I had become fully aware of the size of the finished index. By the time my cards had reached the ten thousand mark, a supplement to a publication was no longer feasible and instead a commercial publisher was eager to undertake the publication. Action on this plan was expedited, and by the end of March the whole matter was settled, entirely to the grateful satisfaction of the author and, I hope, everyone involved in the whole project. My acknowledgments will indicate how many are directly and indirectly included in this large number.