University of Virginia Library

Search this document 


collapse section 
collapse section 
collapse section 
collapse section 
collapse section 
A SYSTEM OF MANUSCRIPT TRANSCRIPTION by David L. Vander Meulen and G. Thomas Tanselle
collapse section 



Page 201

David L. Vander Meulen and G. Thomas Tanselle

There has never been a single standard convention for the transcription of manuscript texts, and it is not likely that there ever will be one, given the great variety of textual complications that manuscripts—from all times and places—can present. Nevertheless, there are certain basic principles and procedures that can reasonably be regarded as essential underpinnings for whatever specific scheme seems best suited to a particular situation. The plan we offer here can easily be adapted to handle details not explicitly mentioned; its fundamental approach is, we believe, applicable to all circumstances. We devised it for use in our edition of Samuel Johnson's Translation of Sallust (1993), and Vander Meulen expanded it in 1997 for his transcription of the Virginia manuscript of Faulkner's Mosquitoes (in his and Thomas L. McHaney's facsimile edition of that manuscript).[1] One of its merits, in our view, is that it is simple and straightforward; it can therefore (unlike many systems that have been proposed) be explained in little space.

First, however, we must make clear that by “transcription” we mean the effort to report—insofar as typography allows—precisely what the textual inscription of a manuscript consists of.[2] Obviously a transcription cannot exactly reproduce the relative precision or carelessness with which handwritten letters are formed, or their relative sizes, or the amount of space between words and lines; but it can aim to record every ink or pencil marking of textual significance on the manuscript—all letters, punctuation, superscripts, canceled matter, lines linking or excising passages, and so on. Judgment is necessarily involved in deciding what is in fact present, as when an ambiguously formed character resembles two different letters; but the transcriber's goal is to make an informed decision about what is actually inscribed at each point. This definition of transcription will seem obvious to many people; the reason we make a point of insisting on it is that numerous discussions of the subject, including some of the most influential, allow for making certain


Page 202
classes of alteration when transcribing manuscript texts, as if a conscious program of alteration is compatible with the concept of transcription.

For example, Samuel Eliot Morison's often-cited comments in the Harvard Guide to American History (1954) describe a so-called “Literal Method” in which silent alterations can be made to correct slips of the pen, unclear punctuation, and idiosyncratic capitalization. Mary-Jo Kline, in A Guide to Documentary Editing (1987; revised 1998), describes “diplomatic transcription” as allowing some standardization, alteration of punctuation, expansion of “ambiguous or archaic” abbreviations, and the like, though she adds that “none of these corrections or emendations is made silently” (p. 155). But neither in this discussion nor in her subsequent treatment of “inclusive texts,” “expanded transcriptions,” and “clear texts” does she suggest that the manner of recording textual evidence constitutes a very different level of consideration from decisions about what words, spellings, and punctuation are to appear in the text itself (that is, the text as distinct from the editor's comments and symbols). Even Michael Hunter's “How to Edit a Seventeenth-Century Manuscript: Principles and Practice” (The Seventeenth Century, 10 [1995], 277-310), which includes many useful observations, does not clearly distinguish these logically distinct matters. After eliminating “type-facsimiles” as an unsatisfactory substitute for photographic or electronic images, he does not consider any other form of unemended transcription and proceeds to discuss various kinds of “editorial intervention,” often aimed at modernization. And Michael E. Stevens and Steven B. Burg's Editing Historical Documents: A Handbook of Practice (1997), after stating that in transcription “editors strive to represent original documents faithfully” (p. 71), outlines the “textual changes made by editors when transcribing,” which include both “silent” and “overt” emendations (p. 72).[3]

Any kind of editorial change may at times be appropriate in a critical text, and it is not our purpose here to discuss principles of emendation, which have already received extensive coverage.[4] Our present point—a fundamental one that seems to be often overlooked in treatments of manuscript transcription—is simply that editorial alterations produce critical texts. All editors have to make a basic decision at the outset of each act of editing: whether to offer their readers a faithful transcription (as faithful as a typographic rendition and accompanying explanation allow) or an emended text (which incorporates alterations that reflect a particular editorial goal). In the latter


Page 203
case, the extent and nature of the emendations may vary considerably from edition to edition, but the result is still a critical text. A single edition might contain both transcriptions and critical texts, but no single text can be both at once, because they are mutually exclusive genres: a text cannot simultaneously be unemended and emended.[5] When one speaks of methods of transcription, therefore, one is talking about the mechanics for reporting what actually appears in a manuscript; one is not considering making changes in the manuscript text. The present discussion takes for granted that the text to be reported is precisely the text of a document; what we are proposing is a way to carry out that task.

When a manuscript is a clean fair copy, with no cancellations, interlineations, marginalia, or other indications of revision, there is virtually nothing for an editor to decide about how to present a transcription (as opposed to the inevitable decisions about how to read certain unclear formations of letters or punctuation). But since manuscripts frequently do show signs of alteration, often very complicated ones, editors commonly must consider how best to make clear the complexities of a documentary text. If an editor chooses not to present a type-facsimile (which attempts to imitate in type the relative positions, sizes, and other visual characteristics of the handwritten words),[6] there are essentially three sets of presentational issues that one must think about (and that provide a means for classifying transcription systems as one encounters them).

(1) Words or symbols. One may use either ordinary language or symbols to explain what is found in a manuscript; one could, for instance, say “interlined” to describe a word inserted above the line, or one could use an upward-pointing arrow (or a caret, or a particular kind of bracket) to mean the same thing. An example of the former is Fredson Bowers's Whitman's Manuscripts (1955) and of the latter is The Journals and Miscellaneous Notebooks of Ralph Waldo Emerson (ed. William H. Gilman et al., 1960-82). The advantage of words, obviously, is that the meaning is immediately clear, without the necessity for learning a system of symbols; the advantage of symbols, on the other hand, is their efficiency in saving space. If the amount of revision is small, there may be no necessity for saving space; in any case, one must always weigh the advantages and disadvantages of each approach in relation to the nature and extent of the alterations.

(2) Inclusive or clear text. Regardless of whether an editor uses words or symbols to express textual complications, there is still the question of whether that information is to be conveyed within a linear text (an “inclusive” text) or whether it is to be presented in an appended list or set of notes, leaving the text free of intrusions (a “clear” text). These do not represent two differ-


Page 204
ent approaches to editing but are rather two ways of reporting identical information. The Emerson edition mentioned above is an example of an inclusive text (though with some textual information in notes), the Whitman of a clear text. Inclusive texts offer the reader an approximation of the experience of reading the manuscript: the difficulties of reading an inclusive text (when there are any difficulties) are analogous to the difficulties of reading the manuscript.[7] Clear texts with appended notes, though they conceal no evidence, do not reflect the original quite as directly; but they may on occasion be desirable nonetheless, especially when the text partakes of the nature of the literary essay or of some other genre that a wide audience may be expected to wish to read only in its final (or some earlier “finished”) form.

(3) Forward or backward chronology. Either choice from the previous two pairs can be joined with a decision either to start the record with the earliest reading (followed by the successive revisions in chronological order) or to start it with the final one (followed by a chronological account from the beginning up to that point). Although any combination is theoretically possible, certain combinations are more common than others in practice. Thus clear texts are likely to contain final readings, with the appended notes recording the earlier ones,[8] whereas inclusive texts frequently give variant readings in straight chronological order, first to last. Indeed, the length to which some inclusive plans go to delineate every stage, beginning with the first, is illustrated by Harrison Hayford and Merton M. Sealts's edition of Billy Budd, Sailor (1962): a canceled word or passage is printed twice, first without any designation of deletion and then with such a designation, since the word or words in question first existed in their undeleted state before they were lined through or otherwise canceled.

The system we propose here uses words rather than symbols and places them in an inclusive text, with the variants at any point normally recorded in chronological order beginning with the first. We have also tried to make the system as simple and convenient as possible, without misrepresenting any complexity that manuscript texts may offer. In order to accommodate simul-


Page 205
taneously those who are interested in the process of revision and those who wish to read only the final text, we have adopted as our fundamental rule that all stages prior to the final one are recorded and described in brackets; as a result, one can read the final text simply by skipping all the bracketed material. This fact, along with our policy of using ordinary prose rather than symbols to explain the placement of alterations, perhaps constitutes all one needs to know in order to implement our system (and one needs no instructions at all to read a text presented this way). But because we have given thought to the application of this approach to a number of specific situations and have worked out some standard wording, it seems potentially useful to offer here a few further details.

* * *

The basic categories of alterations to be recorded are simply stated: material can be canceled, or it can be added—or these processes can occur together. In addition, features such as blots, lines, unusual spacing, anomalous inking, and missing portions of leaves may require identification through editorial comment. It is helpful to differentiate reports of alterations from explanations of what the reviser meant by the change—and to record only the former in the bracketed notes within a transcription. Though even an apparently simple description relies on interpretation and the two activities cannot be neatly separated, the consistency and brevity of the notes will be enhanced, and the final text of the transcribed manuscript easier to read, to the extent that this distinction can be honored. In a similar vein, the editor should proceed cautiously when deciding whether to report multiple changes within a single annotation. Reporting the alterations separately would not be to deny that they may have resulted from a single act of revision, but combining them might go beyond the editor's intention by implying that their origins were linked.

Basic features. In our system, square brackets surround editorial records and comments; braces are used for the next level of parenthesis when commentary is required for alterations within alterations. The editorial statements are italicized and thereby differentiated from words quoted from the document, which consequently do not need quotation marks. (Italics are not common in manuscripts; if they—or, for that matter, brackets—do occur in text being included within the editorial brackets, the system allows the editor to describe the situation.) Generally the basic element in each editorial comment is a participle, such as canceled or inserted, and it normally follows specification of the affected text (as in: [objec canceled] or [next word inserted]). Editorial comments usually refer either to the actual point in the text where they appear or to the text that immediately follows. But sometimes, to reflect the order of inscription, they may refer more appropriately to the immediately preceding text, as when only certain letters of a word need to be discussed and the revision in question occurred after the opening letter(s) of the word had been inscribed; in these cases the word the is used as an adjectival modifier of the cited text in order to make clear that the


Page 206
reference is to something already reported (as in: Mrs [the final s inserted] Weisman). (If, on the other hand, the opening letter or letters of a word are written over a previous false start, the editorial comment precedes the word— as in: [first letter of next word written over s].)

When individual marks of punctuation are cited, they are best referred to by name (as in: [comma canceled]) so that they are not confused with editorial comment. Although indication of line divisions (which would be essential in a type- facsimile) may not always seem necessary, especially when the transcription is accompanied by a facsimile of the document, the editor may decide that the circumstances of a particular text make the practice appropriate (and then would mark each line-end with a bar or solidus). One instance in which such a record can be especially useful (though at the expense of introducing into the transcription a mark not in the manuscript text) is for words that begin at the end of one line (concluding it with a hyphen) and finish on the next; the typographical symbol (as in: pre-|sent) can clarify why a word has been inscribed in fragments. The edition of Sallust notes line divisions, whereas that of Faulkner does not, except for line-end hyphenation. Uncertain readings within the editorial comments can be indicated by a prefatory question mark (as in: [?kind canceled]). Blots, tears, and other damage to the manuscript (whether caused by the inscriber or by the subsequent vicissitudes of the document) need not enter the transcription unless they affect the text; but they would of course be described in a detailed account of the physical features of the manuscript, prepared as an accompaniment to the transcription.

Cancellation. We use the word canceled as the encompassing term for text that has in some way been designated for removal. It comprehends methods such as blotting, lining through, annotating, erasing, and cutting— any of which can be stipulated should the editor deem more elaborate description appropriate (as, for instance, for a procedure that is anomalous within a given manuscript). Canceled seems preferable to deleted because the common use of the latter in computer terminology has meant that deletion is increasingly thought of as removing the physical traces of the original text.

Addition. The word interlined designates one of the most common forms of addition, an inscription made above (or when specified as such, below) the usual line of writing. (An editor could include above in the description, but in most instances doing so would be unlikely to add to the commonly perceived sense of the participle alone.) A related term, inserted, when used by itself means that new characters or words have been added within an existing line (as evidenced by spacing, the shapes of the letterforms, the nature of the ink or lead, and similar characteristics); it may also be combined with descriptive phrases to identify other additions that are not interlineations (as in: inserted in left margin).

Cancellation with addition. Combinations of addition and cancellation tend to offer the most complicated situations to report, though the one that is probably most common—an interlineation above a cancellation—is rela-


Page 207
tively straightforward (as in: [next word interlined above canceled the]). When a passage is moved from one place to another, the transcription should report the presence of the excerpt at both sites. In accord with the goal of presenting the final form of the manuscript text, the complete wording would be given at its new location (with a bracketed note explaining that it had been moved there and identifying its original position), while at its former site another note would record its earlier presence, indicate its new location, and give perhaps the first and last words of the passage (see the examples below). Careful wording can also distinguish two phenomena that are similar to each other: the transformation of one inscription into another (reported by a statement such as: [the final e altered from or]), and what is often called “overwriting,” the superimposition of one inscription on another (recorded as: [first letter of next word written over a quotation mark]).

Examples. All but two of the following editorial descriptions (along with some of the adjacent transcription) have been drawn from the editions of Johnson's Sallust and Faulkner's Mosquitoes to illustrate situations commonly encountered in transcribing manuscripts. The three largest categories have been subdivided (into groupings that are not mutually exclusive) to help in locating particular kinds of changes. A fourth category covers miscellaneous other features, such as blots and tears[9] or matters of spacing and inking. Each example (except the two hypothetical ones at the end) is followed in parentheses by the page number of the corresponding passage in the Johnson or Faulkner edition, which might be usefully consulted for photographic illustrations of the texts being reported on here. The samples also include some complex situations that are unique to Mosquitoes but that show how difficult situations can be handled within the proposed system. It is important to emphasize, however, that the principles informing the system are a more important guide than these examples, not only because the number of possible situations is countless but also because other wording may be equally effective.


Identified characters:

the [speediest canceled] quickest  (Johnson 32) 
the [objec canceled] captious objection  (Johnson 5) 
Or mine [period canceled] either.”  (Faulkner 65) 

Unidentified characters:

I will [letter canceled] put out  (Johnson 9) 
more delightful [one or two letters canceled] than  (Johnson 19) 
time [?w canceled] was  (Johnson 2) 


Page 208

hand, [?fawning canceled] sugaring  (Faulkner 21) 
man / [p and stroke canceled] brass  (Johnson 11) 
design [letter blotted out between the i and g] immediately  (Johnson 35, modified) 

Beginning of a word:

she was [a canceled at beginning of next word] waited  (Faulkner 57) 

End of a word:

The [y canceled at end of word] Allobroges  (Johnson 26) 
Gordon ['s canceled at end of word], without  (Faulkner 11) 

Within a word:

encou [ra- canceled at end of second syllable] / raging the  (Johnson 35) 

Incomplete cancellation:

in [their turn make the same advantages. canceled, with the last two letters and period not lined through] the same  (Johnson 20, modified) 
[“Maybe it dont to some folks,” said canceled, with the opening quotation mark not lined through (Faulkner 91) 



treacherous [the first e interlined] / invasions  (Johnson 3) 
might [the ht interlined at margin] / be  (Johnson 25) 
into a [next word interlined] darkling corridor  (Faulkner 3) 
street, [next two words interlined] Mr T  (Faulkner 29) 
enrich them [next word interlined with caret] in the  (Johnson 19) 
but that, [next four words and comma interlined with caret] as I am told, my  (Johnson 15) 
unhappiness. [next sentence interlined below] Were children really like dogs, do they know it instinctively?  (Faulkner 21) 
life [period canceled] [rest of sentence interlined below] by directing me to Mr F here.”  (Faulkner 23) 
tender, [rest of sentence, and next sentence, interlined below paragraph] making one end of the rope fast to the the bow-ring of the tender as she commanded. He  (Faulkner 73) 


hooks [the s inserted] baited  (Faulkner 63) 
Jenny's [the 's inserted] [came breathing softly canceled] bare feet  (Faulkner 87) 
offended, [the comma inserted] [at your joke, canceled, with the comma not lined through]” Mr T murmured.  (Faulkner 25) 
starlight [a mark, without text, inserted for an interlineation] flat  (Faulkner 41) 
the wet [next word inserted] and [prints of her canceled


Page 209

simple print of her feet on the deck.  (Faulkner 57) 
[next sentence inserted] “Neither do I.  (Faulkner 13) 
fire.” [next two sentences inserted] He stopped suddenly, brooding his cavernous face at them. Suddenly he hurled the bottle crashing into the fireplace.  (Faulkner 35) 
staring [next word inserted in left margin] up at  (Faulkner 79) 
room. [next twenty-three words (then . . . approach) inserted in right margin] then said something to his companions which caused one of them to turn half about in his chair to watch Mr Taliaferro's approach Mr Taliaferro,  (Faulkner 23, modified) 
Wiseman. [next two sentences inserted in text line and in right and bottom margins] “Ah, grape fruit. How Jolly-havent seen a grapefruit since we left New Orleans, hey, Julius?”  (Faulkner 51) 
“Want me to [build a fire here? canceled, with the question mark not lined through] [next three words and question mark interlined and mistakenly marked for insertion before the to] get some firewood?” he asked  (Faulkner 73) 
[next paragraph, with narrower margins, possibly inserted at top of completed page (Faulkner 41) 



Senate [the final e altered from or] letters  (Johnson 4, modified) 
stairs: [the colon altered from a period (Faulkner 3) 


had [first letter of next word written over s] drifted  (Faulkner 17) 
[first letter of next word written over a quotation mark] This  (Faulkner 69) 
But [next letter written over y] T aint  (Faulkner 69) 
together [next comma and word written over two ellipsis points], now . .”  (Faulkner 29) 
supporting [next word written over his] the  (Faulkner 3) 
recall [the c written over m] me?”  (Faulkner 23) 
Cæ [the æ written over other letters] / having  (Johnson 36) 
possessed [the last s written over another letter] at  (Johnson 35) 
trifle, [the comma written over a colon] as  (Faulkner 81) 

Original canceled; interlineation remains:

designation, [the ig interlined above canceled ti] laid  (Johnson 3) 
hibiscus [the first i interlined above canceled y]  (Faulkner 3) 
Sylla [next two words interlined above canceled been] been totally  (Johnson 2) 
with [next word interlined below canceled the] a steadfast  (Faulkner 67) 


Page 210

answered [rest of sentence interlined above canceled quickly.] immediately in a startled tone.  (Faulkner 91) 

Original canceled; interlineation partly canceled:

Ca / taline, [next four words, and intervening cancellation, interlined above canceled who would make of the] a complete [two letters canceled] master of dissimulation  (Johnson 8, modified) 
and [next five words, and intervening cancellation, interlined above canceled fell upon] he now [con canceled] spooned into a a grape  (Faulkner 51) 
seats [also canceled] [rest of sentence interlined above canceled staring at their] and Major Ayers [a canceled] stared at his also.  (Faulkner 59) 

Original canceled; interlineation canceled:

[horror canceled, and interlined carnage canceled (Johnson 31) 
[went {next word interlined} to the {?cam canceled} camp and join all canceled (Johnson 16) 
time, [Mrs M was still trying to catch his eye canceled] [interlined now that Mrs M was no longer there canceled (Faulkner 67) 
she [called David? in a canceled] [interlined said above the called canceled] spoke  (Faulkner 71) 

Original canceled; first interlineation canceled; second interlineation remains:

that [it canceled, and interlined the canceled; next two words interlined above canceled had been] Cinna and / Sylla  (Johnson 38) 
noncommittal [next word interlined below canceled suspense and interlined and canceled caution] alertness.  (Faulkner 55) 
rakish [next word interlined above canceled grace and after interlined and canceled carelessness] impersonality,  (Faulkner 7) 
then [rest of sentence interlined above interlined and canceled it {above canceled them in subsequently canceled phrase led them to the assault.}] brought it on deck again.  (Faulkner 67) 
then [rest of sentence interlined above canceled led it to the assault. (the it having been interlined above canceled them)] brought it on deck again.  (Faulkner 67, modified) 

Original remains; interlineation partly canceled:

intently [rest of sentence, and intervening cancellation, interlined below] beside the other's tall [ghostly canceled] figure.  (Faulkner 29) 

Original remains; interlineation canceled:

but [interlined in canceled] fire,  (Johnson 40) 
trooped [interlined merrily canceled] into  (Faulkner 59) 
sound [interlined noi canceled] of  (Faulkner 85) 


Page 211

other. [interlined (below) The waiter departed. canceled] He  (Faulkner 25) 

Original remains; first interlineation canceled; second interlineation remains:

Like biting [next word interlined above interlined and canceled bark and a tree] bark  (Faulkner 67) 
Pete [next five words interlined below, after interlined and canceled clutching his straw hat] holding his straw hat on clung yet  (Faulkner 49) 

Other combinations:

extreme [ly canceled at end of word, and next word interlined above canceled ?sorry] distress  (Johnson 17) 
there [the ere interlined above canceled ey and next word interlined above canceled had] were  (Johnson 17, modified) 
many [next word interlined above canceled considerations {the final s having been added}] thoughts  (Johnson 9) 
offended, [the comma inserted] [at your joke, canceled, with the comma not lined through]” Mr T murmured.  (Faulkner 25) 
are. [next two sentences and following word inserted before canceled paragraph “I guess so. But then he's got more time to be a genius {as canceled} than I have] But then, he's got more time to be a genius than you have. You spend too much time writing. And  (Faulkner 33) 
cap. [Her niece stood at the after rail beside canceled, with eighteen paragraphs (from the line drawn after the present paragraph to the line drawn across the next leaf) marked to be moved to follow the preceding sentence (Faulkner 47) 
[next sentence, written below the next paragraph, is marked to be moved to this location] “Ah, wretches,” began Mrs. M with flaccid coquetry, shaking her finger at the group.  (Faulkner 51) 
[a sentence (“Ah, . . . group.), present here in the manuscript, is marked to be moved above the preceding paragraph (Faulkner 51) 
Dont feel that [line drawn across page] [rest of paragraph, written in the sixth paragraph below, is marked to be moved to this location] way about it!” He stared at her with his utter yearning. “David, I'm so sorry. What can I do about it?”  (Faulkner 81) 
Dont feel that [rest of paragraph, written in the second paragraph below, is marked to be moved to this location] way about it!” He stared at her with his utter yearning. “David, I'm so sorry. What can I do about it?”  (Faulkner 81, modified) 
as though he [line drawn across page] [a phrase and three sentences (way . . . it?”), present here in the manuscript, are marked to be moved to both the sixth paragraph above and the second paragraph above] [line drawn across page] had heard a sound.  (Faulkner 81, modified) 


Page 212

fiery-cold—[he cannot see it canceled; next five words inserted in left margin, possibly as substitute] does not yet see it; a shape which he will know—[he cannot see it canceled; the previous marginal insertion might also substitute for this passage] The 3 priests  (Faulkner 41) 


threatened / [blot] every one  (Johnson 23) 
hazard [the z blotted] should  (Johnson 23) 
Mr [large space] has got to  (Faulkner 25) 
[extra space between paragraphs (Faulkner 85) 
[line drawn across page (Faulkner 47) 
[rest of sentence inserted in darker ink] opened the door.  (Faulkner 9) 
to foreg [letter(s) following the g torn away
high / [first letter(s) (?r) of next word torn away] espect 



Samuel Johnson's Translation of Sallust: A Facsimile and Transcription of the Hyde Manuscript, ed. David L. Vander Meulen and G. Thomas Tanselle (New York: The Johnsonians; Charlottesville: Bibliographical Society of the University of Virginia, 1993); Mosquitoes: A Facsimile and Transcription of the University of Virginia Holograph Manuscript, ed. Thomas L. McHaney and David L. Vander Meulen (Charlottesville: Bibliographical Society of the University of Virginia, 1997).


Typescripts less often than manuscripts require transcription in addition to photographic reproduction; but typescripts do frequently pose the same kinds of problems as manuscripts when they contain revisions (whether made in handwriting or typewriting). The word “manuscripts,” as used here, is meant to refer to such typescripts as well.


A more detailed historical survey of approaches to transcription (with comments on many specific editions) can be found in Tanselle's “The Editing of Historical Documents,” Studies in Bibliography, 31 (1978), 1-56 (reprinted in Selected Studies in Bibliography [1979], pp. 451-506, and Textual Criticism and Scholarly Editing [1990], pp. 218-272). See also the Kline and Hunter works mentioned above; the Stevens-Burg book gives photographic or xerographic reproductions from a number of editions. Tanselle has also discussed the relation of the European interest in manuscript versions to the Anglo-American tradition of recording manuscript alterations; see “Critical Editions, Hypertexts, and Genetic Criticism,” Romanic Review, 86 (1995), 581-593 (reprinted in Literature and Artifacts [1998], pp. 258-271).


For references, see Tanselle's Introduction to Scholarly Editing: Seminar Syllabus (1998 revision), pp. 16-17 (and longer lists elsewhere in the same volume).


For further discussion of the distinction between critical and documentary editing, see Tanselle's “The Varieties of Scholarly Editing,” in Scholarly Editing: A Guide to Research, ed. D. C. Greetham (1995), pp. 9- 32.


A good example of this approach is Shelley and His Circle (ed. Kenneth Neill Cameron, Donald H. Reiman, et al., 1961-), in which interlinear insertions are printed above the line in smaller type, canceled words have a line through them, and so on.


But when an editor complicates a text by the insertion of textual evidence drawn from more than one document, the result cannot be justified on these grounds (though possibly it can on other grounds). The much discussed “synoptic” text in the 1984 edition of Joyce's Ulysses (ed. Hans Walter Gabler et al.) is an example of this kind of text and is thus not an instance of inclusive manuscript transcription; Tanselle has elaborated this point in “Textual Criticism and Literary Sociology,” SB, 44 (1991), 83-143 (esp. pp. 108-109 and p. 139, note 91), and in “Critical Editions, Hypertexts, and Genetic Criticism” (see note 3 above), pp. 588-590 [265-268].


Bowers's Whitman edition, mentioned above, is an example; and some of his critical editions offer an extension of this approach. In his Hawthorne and William James editions, for example, he presents clear reading texts that are critically emended; these texts, though obviously not intended to be documentary transcriptions, are furnished with appended notes that allow the reader to reconstruct manuscript alterations. In the Hawthorne, these notes explain the alterations in words (see The House of the Seven Gables, 1965); for the James, Bowers created a system that relies on symbols and abbreviations (see Pragmatism, 1975). The latter system, the product of Bowers's long attention to matters of textual apparatus, has had some influence, despite the fact that he required fifty pages for its full exposition; see his “Transcription of Manuscripts: The Record of Variants,” SB, 29 (1976), 212-264.


Although the two examples of tears below suggest a concise form of wording, one might wish to use an even more succinct treatment in cases where two or more lines are affected—such as simply a dash or a short rule in brackets to stand for text lost through the damage that is explained in a separate description of the manuscript. In such a case, the truncated ends of several lines might be transcribed on this pattern:... to foreg[—] /... good examp[—] /