The material to be keyed is taken from books (out-of-copyright editions of Middle English texts), and will be sent to the data conversion firm on CD in the form of 600 dpi bitonal tiff files (one file per page). Pages that need to be considered in pairs (because of text spread across a page opening) will be noted as such in the book-by-book coding instructions (described below), though *all* the books may potentially include run-on notes and similar text that requires consideration of more than one page at a time.
Files will be sequentially numbered within the item. I.e., the files for each book will start at "00000001.tif" and go on from there.
Files will be arranged in directories, one directory for each item; the name of the directory will be the ID number assigned to the book. (see http://www-personal.umich.edu/~pfs/med/cmeitems.html for a list of books in process, with their id numbers).
Not all pages will be keyed. The pages to be keyed will be indicated both in the book-by-book coding instructions and on a spreadsheet printout that accompanies the CD-ROM.
The data-conversion vendor will return keyed and coded text files transcribed from the image files.
Transcriptional accuracy will be 99.995% or better (error rate of 1 character/byte in 20,000). We will test and if necessary reject data by the book (volume), not by the shipment.
Coding will be valid SGML, validated against the supplied "vendor.dtd" or a true subset thereof. Vendor.dtd is an extract from TEI and uses TEI semantics.
Data will be sent on CDs, approximately 15-20 books per shipment
Total quantity will depend on the cost and speed of conversion. The budget is limited and must be 'spent down' within the fiscal year. When we run out of time or money, we stop. If we can avoid both deadlines, assuming an approx. cost of $1/1000 output characters, total output data should be in the range of 60-75MB. Estimated total books: 75+.
NOTE: We recognize the need for consistency, and the expense entailed in changing instructions and procedures midstream; such changes will certainly be minimized. Nevertheless, there is considerable variety in the source material and minor special instructions may be required for some books, or some portions of books, in some cases overriding the instructions given below. See the book-by-book coding instructions (posted at http://www-personal.umich.edu/~pfs/med/mec.html#cme).
|1.1||All of the books contain two chief kinds of content that we wish to preserve:
(1) the Middle English text(s); and (2) the editorial apparatus, which may appear as marginal notes, footnotes,
endnotes, or in other forms. Anything that falls into one of these categories or the other should be recorded. In most
cases this will mean all words on the page, from top to bottom, left to right.
TEXT. The Middle English text and its structures should be encoded using mostly the following structural elements:
|1.2||Some books may include a third type of content that we wish to record: (3) scribal apparatus--notes or additions added in the margins or between the lines of the original manuscript. Since it is not feasible to provide criteria sufficient to distinguish this from the text proper, or from editorial apparatus--depending on how it is presented in the book--this type of content should be treated the same as whichever it most resembles. I.e., if it printed as if it were marginal notes, treat it the same as editorial marginal notes; if it is printed as if part of the text, treat it as part of the text.|
|1.3||Running parallel texts, displayed in a multi-column, multi-row, or
facing-page arrangement, or some combination thereof, need to be treated as separate texts, each one recorded until
its end and not restarted on each page. Editorial apparatus relating to one of the texts on a page needs to be
embedded in that text, not in any of the others. Some specific parallel texts, in some particular books, may be
excluded on a case-by-case basis. If a single heading applies to multiple parallel texts, it may be rekeyed at the
head of each text. A suggested <DIV> structure for handling such parallel texts will be included with the
shipment (see below).
In particular, partial or fragmentary parallel texts will normally be broken primarily at the chapter or section level (<DIV1>), then into parallel versions of that chapter (<DIV2>) when necessary. But full parallel texts will normally be broken primarily into versions first (<DIV1>), then each version into its chapters (<DIV2>).
|1.4||All material should be recorded in the form in which it appears in the book: do not attempt to correct spelling or typographic errors.|
In general, use attributes only when there is specific information to put in them, and the guidelines specifically call for its use. Do not, for example, record attributes of the sort 'TYPE="unknown"'
Record most 'milestone' information (page numbers; folio, leaf, or column numbers of the source manuscript; alternative page numbers derived from some other edition; etc.) as the value of the "n" attribute of the appropriate milestone element, regardless of how they originally appear (e.g. in footnotes, in marginal notes, or in brackets within the text). This requirement overrides the directions belonging to the type of structure in which the information appears. That is, if milestone information appears in a footnote, treat it as milestone information, not as a footnote; if it appears within the text, treat it as milestone information, not as text; etc.
The chief milestone elements are:
Numbers attached to various pieces of the document structure (line numbers; stanza numbers; chapter numbers, etc.) should be recorded as the value of the "N" attribute of the appropriate structural element. If the number appears without additional text, it may usually be recorded simply as an attribute and removed from the text; if it is accompanied by additional material, the whole should usually be treated as a heading (<HEAD>). The most commonly numbered elements include:
In general, do not attempt to record the physical appearance of the page (centering, extra spaces, justification, type face, type size, etc.), though such cues may and should be used to determine the beginning and end of divisions within the text, the distinction between text and notes, etc.
Paragraph breaks should be recorded with <p> in prose and with <lg> (line-group or stanza) in verse.
Record the presence of underlining, bold, and italic text only when it consists of an entire word (or single letter printed as if it were a word, or a mixed-mode word) or more, and only when you are not using its presence as clue to its structural role. That is (1) do not place <I>, <B>, or <U> tags around single letters or groups of letters within words; and (2) as a rule of thumb, if the print mode of the text is being used as a cue to its structural role, do not record it; if not, do. E.g. a milestone or a heading may sometimes be identified by its bold or underlined text: record <HEAD> Chapter 3</HEAD>, not <HEAD><B>Chapter 3</B></HEAD>.
Mixed-mode words (words that contain some letters that are roman, some italic; or some other combination of roman, italic, bold, and underlined) should be recorded as only one of those, the predominant one, Predominance is established not primarily by numerical preponderance, but by context (is the word one of a string of words in italic? If so, it should be treated as italic, even if some of its letters are roman); and by function (are the roman (etc.) letters grouped so that they appear to represent an expansion of an abbreviation? Then the expanded part is not in the predominant mode.).
Record superscripted and subscripted text with <SUP> and <SUB> tags, regardless of how many characters are involved or where they appear.
Because of limitations in the content-model of <NOTE>, passages of verse (especially 2 or more lines, quoted and arranged as verse) within a note will normally be most readily coded as a quotation (<Q>) containing <L>s or <LG>s, embedded within the <NOTE> element.
Record most apparatus material (marginal and foot notes, etc.) at the point in the main ME text to which it relates, set off by appropriate tags, not at the point where it appears on the page. If the place is marked in the text with a flag of some kind (e.g. a footnote reference number, an asterisk (*), etc.), discard this marker both from the text and from the note. If the note (or similar material) is keyed to the text by line number, place the note at the end of the line to which it applies, and discard the literal line number from the note.
Use the "PLACE" attribute of the <NOTE> tag to indicate where the note appears on the page:PLACE="marg" in margin or adjacent to the text PLACE="foot" in a footnote, below the text PLACE="end" in an endnote. PLACE="inline" within the text itself (e.g., in brackets) PLACE="inter" interlinearly (between the lines of text) PLACE="head" in a headnote.If there are multiple distinguishable sets of notes in the same location (two sets of footnotes, for example), distinguish them by appended numbers: PLACE="foot1" and PLACE="foot2" for example.
A note that spills onto the next page needs to be treated as a single note, not two, and should be placed in the text where it applies.
Notes and other ancillary apparatus material that relate to a specified group of lines should be moved into the text at the end of the last line to which they apply, with the line number indication *preserved.* If physical arrangement, rather than explicit line numbers, serve to specify the line number range, supply the line-number range in brackets at the beginning of the note. Notes referenced to a line number followed by "f." ("2365 f." meaning "line 2365 and following") should be treated as notes referenced to a span of two lines (in this case, 2365-66), that is, placed at the end of the second line (2366), with the full line reference preserved in the note: <NOTE PLACE="foot">2365 f.: ... </NOTE>
Apparatus that appears next to a single line or set of lines and seems to relate to that line (or set) should be placed at the end of the line(s) in question.
Notes that apply to two (or more) distinct loci or lines should be reproduced and inserted at *both* (or all) the relevant points.
Apparatus, such as an introductory headnote, that seems to relate to an entire text division should be attached as a note to the heading for that division. A summary of the division should usually be given a tag of its own and tagged as <ARGUMENT>.
Apparatus that relates generally to the material on a page, or for which the appropriate place cannot readily be determined, should be attached to the last line of text at the bottom of the page.
Multiple apparatus applied to the same locus in the text should be recorded in sequence from from most to least specific, i.e., e.g.,
(1) apparatus linked by a flag to that precise point;
(2) apparatus linked to that line generally;
(3) aparratus linked to that specific group of lines;
(4) apparatus linked to that area of the text generally.
Front and back matter will mostly be excluded from keying. When they are included, they are treated much as any prose in the text is treated: divided and subdivided as necessary into <DIV>s, broken up into paragraphs < P>, lists <LIST>, tables <TABLE>, etc. A few examples of typical front matter (title pages, tables of contents, etc.) are included in the appendix, but for the most part you will need to rely on the DTD and TEI guidelines.
Front matter also includes a special set of elements for the title page (both recto and verso are together coded as the title page, separated by a <PB> page break). These are a subset of the usual TEI elements, and are used, together as necessary with <LB> (the line-break tag), chiefly in order to preserve some semblance of the appearance of the original. Use <LB> tags not for every line break, but only where the line break serves as a kind of punctuation and is essential to sense. E.g., <TITLE>Three Middle English Romances <LB>Sir Degare, Sir Launfal, Sir Isumbras</TITLE> (where <LB> marks the beginning of the subtitle).
A typical title page looks like this:
<TITLEPAGE><DOCTITLE>Robert of Brunne's "Handlyng Synne," A.D. 1303, with those parts of the Anglo-French
treatise on which it was founded, William of Wadington's"Manuel des Pechiez,"</DOCTITLE><BYLINE>re-edited
from MSS. in the British Museum and Bodleian Libraries, by Frederick J. Furnivall, M.A. Camb., Hon. Ph.D. Berlin, Hon.
D. Litt. Oxford, Honorary Fellow of Trinity Hall, Cambridge, Founder of the Early English Text and other Societies,
and of the Furnivall Sculling Club.</BYLINE> <IMPRINT><PUBPLACE>London</PUBLACE>: Published
for the Early English Text Society By Kegan Paul, Trench, Trübner & Co., Limited, Paternoster House,
Charing Cross Road, W. C., <DATE>1901</DATE>.<PB> Berlin: Asher & Co., 13, Unter den linden.
New York: C. Scribner & Co.; Leypoldt & Holt. Philadelphia: J. B. Lippincott & Co.</IMPRINT>
<DOCEDITION>Original Series, No. 119.</DOCEDITION></TITLEPAGE>
<DOCEDITION> is used for both edition and series information.
A suggested structure will be supplied (by us) with each book, though the data-conversion vendor is free to modify this if it proves impractical. Special instructions for tagging the book will be included with the suggested DIV structure.
Decisions about structure are not always clear-cut; complete consistency between the coding of one work and that of another will be impossible. Consistency *within* a work, however, is essential.
In general, in prose, anything with a true heading is a <DIV> of one level or another (or, in verse, a line- group). Paragraphs, unless equipped with a heading, are simply <P>. In verse, low-level line groups may be called <LG>, whether or not they have a heading; <DIV>s should be saved for line-groups big enough to have titles, or to appear in tables of contents.
At the lowest level, it is often unclear when a group of lines has enough organization to be called a stanza (line- group <LG>). If in doubt, err on the side of fewer line-groups rather than more. And be consistent throughout a particular poem, so that a particular structure is not sometimes tagged as a <LG> and sometimes left untagged. Clues to look at include, in decreasing order of significance:
A sample suggested/supplied DIV structure (with special instructions) will look like this:
|Text AC = The English Register of Oseney Abbey
Second level Subdivide the second <DIV1>, using:
<DIV2> = the sections denoted editorially by roman numerals and listed in the table of contents on pp. vii- viii. Use the headings in the table of contents as heading for each section, supplying them within brackets. E.g. < DIV2 TYPE="Title" N="VI"><HEAD>[VI. Of the Foundation of Oseney]</HEAD> Sometimes the same heading (or a similar one) is already placed within brackets in the text itself; use the heading from the table of contents and ignore the bracketed one in the text.
Third level Subdivide all the <DIV2>s, using:
<DIV3> = the sections denoted by arabic numerals. E.g.
<DIV3 TYPE="ITEM" N="18">
NOTE: Where <DIV3> breaks coincide with <DIV2> breaks, nest the <DIV3> within the <DIV2> and apply to each the appropriate heading: <DIV2> gets the roman numeral and chapter heading from the table of contents, in brackets; <DIV3> gets the arabic numeral and the bold heading printed in the text itself. E.g.: (p. 10):
<DIV2 TYPE="Title" N="VI"><HEAD>[VI. Of the Foundation of Oseney]</HEAD>
Special instructions for this book.
*Include* the modern English summaries printed in the margins (despite the fact that these are normally excluded from keying); tag them as <ARGUMENT> and place the <ARGUMENT> tag directly after the <HEAD> at the head of the item. Many, but not all, of the arguments will begin with a date. Material in square brackets in the margins is not part of the argument and should be omitted.
An m-dash should be recorded as — or two hyphens, whichever is easier.
End-of-line hyphens should be recorded with the pipe or vertical bar character (|) = decimal 124 = hex 7C, a character which should not be used for any other purpose; if the pipe/bar character itself appears on the page, represent it with an entity reference (|).
Ordinary hyphens in mid-line should be recorded using the regular hyphen character (-).
Decorated characters or large initials should be recorded as if they were ordinary characters; a large decorated "A" should be recorded as "A".
Small caps should be recorded as regular caps, except where convention and appearance require lower-case, e.g. in transcribing a title page.
Superscript characters should be treated as small pieces of superscripted text and recorded using the <SUP> element, not as character entities.
Non-ascii characters should be recorded with SGML character entities, using the standard ISO sets as far as possible. Standard diacritics (umlaut, acute, etc.) should generally be combined with the character to form a single entity (ö à é etc.). Macrons above vowels should be recorded with entities ending in "macr" (ā ō etc.). But macrons above "m", "n", or "p" should be ignored.
Thorn (lower case) : use þ Thorn (upper case) : use Þ: Eth (lower case) : use ð Eth (upper case) : use Ð Yogh (lower case) : use &yogh; Yogh (upper case) : use &YOGH; A-E ligature (lower c.) : use æ A-E ligature (upper case) : use &AELIG; Ampersand & tironian "et" : use & Paragraph/capitulum mark : use ¶ Mid-height dot : use ·
NOTE: There is considerable typographic variety in the printing of some of these characters.
|ellipses||"..." (a string of periods)|
|European quotation marks; inverted quotation marks; baseline quotation marks||" (ordinary quotation marks)|
|opening & closing double quotation marks||" (ordinary quotation marks)|
|opening & closing single quotation marks||' (ordinary apostrophe/quot.mark)|
|d with loop or tick||"d" (ordinary d) (but a barred "d" = ð)|
|f with tick on crossbar||"f" (ordinary f)|
|g with a tick||"g" (ordinary g)|
|k with a tick||"k" (ordinary k)|
|barred-h||"h" (ordinary h)|
|barred-l||"l" (ordinary l)|
|barred-double-l||"ll" (ordinary ll)|
|n with upswept finial||"n" (ordinary n)|
|n with backswept finial||"n" (ordinary n)|
|r with upswept finial||"r" (ordinary r)|
|r with backswept finial||"r" (ordinary r)|
|long-s (see below §7.4.1 for example)||"s" (ordinary s)|
|t with a tick||"t" (ordinary t)|
The common abbreviation "Ihc" or "Jhc" (usually with some kind of suspension mark or macron) should be recorded with the appropriate letters, between dollar signs: $Jhc$ or $Ihc$ or $IHC$ etc. Similar forms should be similarly recorded, e.g. "Jhu" with abbreviation stroke as "$Jhu$", "Jhus" as "$Jhus$, "Jhm" as $Jhm$, etc.
The following character instructions were copied from the book-by-book instructions for individual items, June 2000
Record the left-pointing index finger as &lindex;
the right-pointing index fingers as &rindex;
|Double vertical bars should be recorded with the standard ISO ‖ entity.|
|The strange-looking squiggle that appears almost like a little human figure can be recorded simply by a period (.)|
|The punctuation mark that looks like an upside-down semicolon ("punctus elevatus") can be recorded simply with a semicolon (;)|
|The punctuation mark that looks like an equals sign can be recorded with an equals sign (=).|
|7-shaped "Tironian 'et'" should be recorded with the ampersand entity (&). this normally appears where one would expect to find an ampersand, between words in the middle of sentences.|
|A crossed thorn (which is an abbreviation for "that") should be recorded with the special entity &that;|
|A dotted "y" should be simplified to plain "y"|
|The "long-s" character, should be recorded simply as "s", but is often very difficult to distinguish from "f".||
Punctuation should be retained, including the virgule, period, or mid-line dot marking the caesura or midpoint of a verse line. See also the comments on hyphens above under 7.1.
Strings of dots or asterisks indicating omitted or missing text should be recorded as five of that character in ordinary text, using periods or asterisks separated by spaces: . . . . . * * * * *
Verse lines consisting only of dots (.....), which are numbered as part of the verse, should be treated as lines containing only dots (<L> . . . . . </L>)
Some editions of prose mark extended quotations by placing quotation marks at the beginning of every line. E.g.,
he made reasons...seyenge: " God made alle thynges by reason, and governethe thynges " made by reason; the sterres be movede by reason; and so " oure naturalle lyfe excedynge from reason by slawthe and " ignoraunce awe to be reducede by lawes and reasons. " Wherefore thau&yogh;he there be somme thnges in the rule of " seynte Benedicte, the intellect of whom the dullenesse of my " mynde may not comprehende, y suppose hit be beste to &yogh;iffe " credence to auctorite;
This is OK in verse, but in prose, separated from the printed page, without line breaks, these make no sense. When blocks of beginning-of-line quotation marks occur, EITHER: (1) *include them* but include a <LB> tag too (<LB>"God made alle thynges); or *include them* but record them with a special character entity; whichever is easier.