|
4. Intellectual Content, <msContents> part 1 The <msContents> element encloses the section, or sections, of a manuscript or manuscript part that describes its intellectual content. This wrapper element may in its simplest form simply enclose one or more <p>s; normally, it contains a number of <msItem> elements.� The <overview> element may introduce the set of <msItem> elements if desired (see the documentation on <overview>).� The <msContents> element may be introduced by the <head>, and it may offer the <keywords> element as many times as desired to encode subject headings according to a known scheme, such as that of the Library of Congress Subject Headings (see <keywords>, a standard TEI element with its "scheme" attribute). Two of its attributes are particular to the present context, and are shared with <msItem>:� the "status" and the "incomplete" attributes.� "Status" with its closed list of values (frag | def | unk) speaks to the physical situation of the manuscript as a whole, indicating whether it is fragmentary (missing very large amounts of its support, and thus also of its texts), or defective (missing relatively small amounts of its support and thus of its texts), or whether the physical situation is unknown. The attribute "incomplete," on the other hand, refers to the scribe’s work; the closed list of values encode the lack of completion of the scribe’s work, even though the physical manuscript may be complete with blank leaves reserved for the scribe (incomplete="y");� or the scribe’s work may have been finished, even though the manuscript is defective, when the scribe took advantage of empty margins on already inscribed leaves (incomplete="n"); or the cataloguer / encoder may not be sure of the scribe’s completion (incomplete="u"). Examples:
<msContents>
</msContents> <msItem> contains a unit of bibliographic information within a manuscript or a manuscript part, as determined by the cataloguer. The <msItem> tag encloses untagged as well as tagged materials relating to the intellectual content of a discrete segment of the text of a manuscript.� Different cataloguers will handle the same sections of text in different ways, and the encoding will reflect this variety.� It becomes all the more important, therefore, to use the various subelements correctly and completely, since the searching will focus on these, leaving the wrapper element, <msItem> to serve mainly for formatting purposes.� The wrapping or binding aspect of <msItem> serves to form a unit of what are two discernible parts.� One part focusses on the text as given in the manuscript (and thus employs the elements that encode rubric, incipit, explicit, colophon, and summary).� The other segment of <msItem> cites modern printed editions or studies on the text in question, or allows the cataloguer to comment on the nature of the text (<bibl> with its own subsets of element will be the most usual for this section of <msItem>). The <msItem> element has two attributes of particular application.� The first, "status," refers to the physical situation of the manuscript with its necessary implications for the text, which may be "fragmentary" ("frag") and thus only remaining extent in little more than scraps; or it may be "defective" ("def") and thus proportionately more extant than not, missing only relatively small sections; or its status as "unknown" (which is provided here more for exhaustiveness than for actual use). The second attribute refers to the amount of work completed by the scribe, and it is thus much more closely related to the intellectual content of the manuscript than to its physical situation.� The book may be complete, with all leaves duly present, but with a certain number of them blank because the scribe did not finish his job; the incomplete work is of particular interest for manuscripts copied from rented peciae, as it indicates which peciae were unavailable at a given moment.� The values for this attribute are "yes," "no," or "unknown" of which the one used most frequently will be "yes" (if the cataloguer hasn’t bothered to comment on a manuscript’s completion or not, one assumes that it is complete); "unknown" is provided here more for exhaustiveness than for actual use. Examples:
The <locus> element defines a location within a manuscript or manuscript part. The locations of texts or images or ownership notes (as examples) are normally stated as references to one or more folios, the latter expressed as a span; the locations are sometimes given as pages; the locations may not always be continuous.� The optional <locus> element encodes such locations.� It may contain PCDATA or the <hi> element which is provided for the frequent occurrence of superscript letters in foliation. The element has several attributes:� "scheme" could be used to specify the type of foliation or pagination if deemed necessary and if the cataloguer has not done so explicitly; otherwise, the <foliation> element will prove more appropriate.� Other attributes are "from" and "to" for the leaf/page numbers.� More frequently used will be the "targets" attribute, with IDREFS as its datatyp.� This attribute supplies identifiers of associated images, for example, of transcriptions or of pictorial representations (via <figure> or <xptr> elements). Note:� The number that expresses the quantity of leaves/pages in a manuscript as part of the physical description of the manuscript is encoded with the <extent> element. Examples:
In the last example, the identifier S1 is assumed to reference the section of the manuscript "prayer on fol.182v'' which has been transcribed elsewhere in the current document; the identifiers F33R and F33R reference <figure> elements containing images of the indicated pages. <rubric> contains the string of words that denotes the beginning or the end of a text division, usually set off from the text itself by red ink, or by a different size or type of script, or by lining through, or other such visual device. This element encodes the piece of text in the medieval manuscript that comes closest to the modern world’s title page:� it is the assertion of the book itself as to its author and title.� The assertion may be incorrect; it may have been rejected or corrected by later owners of the book.� If possible, the encoder should recognize with appropriate tagging all names present in the rubric, especially if any of them are different from the author or title of the work as generally accepted today.� Subelements that may prove useful for changes to the rubric made by various readers or owners over time are <add> and <del>.� In addition, see <author> and <title>. On the difficulty of separating a final rubric and the colophon, see <colophon>. Examples:
Other parts of the description would duly encode:� <author>Donatus</author>, <title>De partibus orationis</title>.
<incipit> contains the opening words of a text, or of a division of a text. The incipit functioned in the Middle Ages as a more sure form of textual identifier than the title, and it sometimes still serves that purpose today, justifying the omnipresence of incipit lists in catalogues of medieval manuscripts.� Accurate extraction of incipits from electronic catalogues will depend upon the quality and possibly even the depth of the encoding of this element. The issue of depth arises with the use of the "type" attribute on this element.� When a text begins with a biblical citation, or a legal lemma, or a prologue, or a translator’s dedicatory letter, all its incipits should be tagged and distinguished from one another by means of the appropriate value on the attribute.� No values are pre-set for this attibute. Another common issue with incipits is that they are all too frequently defective, when someone has removed the decorated first letter or first page of a text.� It will remain the decision of the cataloguer whether he wishes to supply the missing text (possibly in brackets or in italics or with some sort of typographical signal), and the markup of the missing text can take advantage of standard TEI elements such as <corr>.�� In any case, the markup should optimally employ the "defective" attribute with the value "yes."� The other two possible values, "no" and "unk[nown]," are included for the sake of completeness, and there seems little value to specifying them.� The default value is in fact, "no." When the incipit is incomplete (i.e. left so by its scribe and/or decorator), as opposed to having become defective through defacing and cutting away, the encoder should continue to use the defective="yes" attibute and value, even though they are an inaccurate description of the situation. Note:� Be sure to avoid encoding as <incipit> the phrase that frequently stands at the very beginning of a medieval text, in red ink, such as "Incipit Liber Psalmorum" or "Here begins the Psalter."� Such phrases are in fact the rubric, while the incipit contains the actual words that open the text itself, in these cases, "Beatus vir qui non abiit" or "Blessed is the man." Examples:
<explicit> contains the closing words of a text, or of
a division of a text.
This element for the explicit or desinit is used to encode the closing words of a text, or of a division of a text.� Although less common than incipitaria, some explicit lists do exist, and accurate encoding would allow the creation of further such lists.� The "type" attribute will allow the encoder to specify which of potentially several explicits in a given <msItem> is being tagged. When the explicit is incomplete (i.e. left so by its scribe and/or decorator), as opposed to having become defective through defacing and cutting away, the encoder should continue to use the defective="yes" attibute and value, even though they are an inaccurate description of the situation. Note:� Be sure to avoid encoding as <explicit> the phrase that frequently stands at the very end of a medieval text, in red ink, such as "Explicit Liber Psalmorum" or "Here ends the Psalter."� Such phrases are in fact the rubric, while the explicit contains the actual words that close the text itself, in these cases, "Omne quod spirat laudet Dominum" or "Let every thing that hath breath praise the Lord." Examples:
<colophon> contains information supplied by the scribe about the production of the manuscript. The colophon may contain various kinds of information:� a ditty about the joys of finishing, the scribe’s name, the name of the person for whom the manuscript is being copied, the place of production, the date of production, the reasons for copying the text, and so on.� The concept of "dated manuscript" as touchstone for all undated and unlocalized codices depends directly upon the accuracy of transcription of the colophon. Note:� Since the colophon occurs at the end of the text or division of a text, it is often difficult to separate exactly the rubric and the colophon, and one may be faced with false results about a given scribe’s normal phrasing, if the encoder attempts the separation.� In such cases, it may be preferable to encode the entire piece of final text as <colophon>.� The preference given to <colophon> over <rubric> is due to the potentially higher value of its information. Examples:
The <summary> element serves to encode the brief prose statement that takes the place of more specifically quoted and encoded textual analysis. When the cataloguer chooses to state or explain the nature of a certain text, rather than quoting its incipit and explicit, he may use the <summary> element to stand in the place of the omitted quotation.� It is the chosen element for the description of texts of medieval archival materials. Note:� While <summary> takes the place of encoded text transcribed from the manuscript, <overview> introduces it. Examples:
The <overview> element introduces a section of the manuscript description by announcing in an abbreviated way what will follow in detail. This element usually occurs between <msContents> and the various subsequent repetitions of <msItem> to which it serves as introduction; it allows the cataloguer a method for commenting on the group of <msItem>s as a whole.� It must include a <p> element. Note:� While <summary> takes the place of encoded text transcribed from the manuscript, <overview> introduces it. Examples:
The <textLang> element encodes the names of the languages that are cited by the cataloguer as present in a manuscript. Although the primary language of a text may be implied by the language in which its title is given (e.g. ‘Horae’, ‘Heures’, ‘Gebetbuch’, ‘Primer’), many catalogues provide uniform titles of works and specify the language of text explicitly, such as "Book of Hours, in Latin and in French."�� It is important for consistent search and retrieval, especially of vernacular languages, that the encoder use the <textLang> element with its "langKey" attribute, and that he derive the attribute’s values from ISO639-2 for the names of the languages.�� If the languages are not cited explicitly in the catalogue, one may usefully employ the "langKey" attribute on <msItem>.� In order to parse successfully, the ISO standard cited for the attributes must also be defined in the TEI header.� Note:� The global attribute "lang" specifies the language of its own element’s content. Note:� The element <langUsage> encodes the language of the description, not that of the manuscript described.�Note:� The element <writingSystem> encodes the manuscript’s primary writing system, independently of its language (e.g. Cyrillic; Glagolitic; Greek; Hebrew), while <scriptTerm>� is used for the names of the scripts that are particular instances of a writing system. Examples:
|