6.  Physical Description

<physDesc> contains the physical description of a manuscript.  This description may be placed under a heading; the description may be expressed in a series of paragraphs or distributed under various specialized elements.

This element and its subelements address the issues that have become paramount in manuscript studies since the second World War, known collectively as codicology or archeology of the book, in which the book is viewed as a historical artifact, worthy of study in its own right.

The architecture of the <physDesc> element is intentionally flexible, so that the element may be used successfully with legacy data from any number of cataloguing traditions.  For example, one may wish to open a first <physDesc> element to contain the basic information, then turn to <msContents> for the texts in the manuscript, and ultimately open a second <physDesc> element for more detailed discussions of the book’s decoration.  If one wanted a heading to each subelement of <physDesc> (when only <msWriting>, <decoration> and <bindingDesc> have content models that make direct provision for this), one could repeatedly open <physDesc> and its one <head> followed by the intended field, close the first <physDesc>, only to open another one and start the procedure afresh as many times as desired.

In its simplest application <physDesc> need only contain a paragraph.  The more complex encoding will further codicological studies by allowing complex searching and computer-generated correlations across large numbers of descriptions.  As always, the specific encoding project will need to establish its own encoding rules to ensure uniform mark-up across a body of material. It is worth bearing in mind the ever-present tension between time and effort invested in encoding vs. the expected result.  Just because an element exists does not mean that it must always be used.

Examples:

<physDesc><p>Parch., 25 f., 21,5 cm x 15 cm, 2 col.</p></physDesc>

<physDesc><support><p>Cart.; </p></support><p>mis. <dimensions type="cm">19 x 13</dimensions>; di </p><extent>ff. 68</extent><bindingDesc><p>Rileg. in membr.</p/></bindingDesc></physDesc>

<physDesc><p>Pap.  </p></physDesc><history><p><origDate>XVI Jahrh.</origDate></p></history><physDesc><p>  117 Blätter.  2<hi rend="superscript">o</hi>.  Brauner Lederband mit Goldpressung und Schnitt.</p></physDesc>

FORM

<form> provides the mechanism for encoding the differing physical shapes in which manuscripts are found, such as codex, roll, amulet, holster, girdle book, accordion-pleat, heart-shaped, tablet.

It is assumed that this element will be used to signal forms other than codices –e.g. rolls, girdle books, etc.— since most manuscripts of the period under consideration are or originated as codices, and cataloguers seldom comment on the form unless it deviates from the norm.  This element must contain <p>.

Note:  The distinction between codices and documents of archival nature, according to which the different kind of text entails considerably different physical appearance, is recognized by means of the "type" attribute (with value:  "document") on <msDescription> and on <msPart>.  See Appendix A, "Encoding Medieval Archival Documents."

Note:  The element <form> is not used to signal fragments; these are signaled by the "status" attribute (with legal values of frag[mentary] or def[ective]) on <msDescription> and on <msPart>.

Examples:

<physDesc><p>Constructed so that the roll could be closed up</p> <form><p>concertina fashion</p></form><p> WWW into a "book," which is read by turning the volume sideways so that the fore-edge faces the reader.</p></physDesc>

<physDesc><form><p>Rouleau</p></form><support><p>de <material>parchemin</material></p></support><p>composé de 8 feuillets de longeur inégale, attachés à un ombilic de bois; longueur <dimensions type="height" units="meters">6.12 </dimensions>m, largeur <dimensions type="width">235 à 245</dimensions>mm.  <scriptTerm>Ecriture bénéventaine</scriptTerm>.</p></physDesc>

<form><p>The 2 outer thirds of each leaf fold vertically inwards over the middle third; the entire leaf is then folded horizontally downward in half.  Tabs extending from the bottom of the middle third of each leaf, together with protective morocco flaps front and back are held by a brass clip with a loop at its other end, 50 x 32 mm, so as to hang the calendar from a belt.</p></form>

SEAL

<seal> supplies information about the seal(s) attached to documents to guarantee their integrity, or to show authentication of the issuer or consent of the participants.

Although the term "seal" is also used to designate the negative matrix of metal from which the positive impression in a softer material is made, it is intended here to describe the resulting impression.

This element may occur within <p>, and itself contain pertinent elements such as <origDate>, <material>, <dimensions> among others.  Its "type" attribute might be used to describe, for example, methods of attaching the seal to the document, such as pendant, affixed, double.  See Appendix A, "Encoding Medieval Archival Documents."

Examples:

<seal type="affixed">sigillo cereo impresso incassato</seal>

<physDesc><p>Parchemin, <dimensions type="height">250</dimensions> x <dimensions type="width">175</dimensions> mm (repli <dimensions type="plica">15</dimensions> mm), scellé d’un <seal type="pendant">sceau de <material>cire verte</material> fragmentaire sur double queue de parchemin</seal>.</p></physDesc>

<seal>Multicoloured silk cord to which the seal was originally attached.  The seal is kept separately in the Coin Room.  It is <material>leaden</material>, <dimensions type="diameter" units="cm">8.3</dimensions> cm in diameter.  Cf. W. De Gray Birch, <title>Catalogue of Seals in the Department of Manuscripts in the British Museum</title> (London, 1900), vol. VI, no. 23082.</seal>

SUPPORT

<support> contains a description of the physical material on which the manuscript’s writing, musical notation, decoration and / or other signifiers are placed, or are intended to be placed.

Within the <support> element, the element <p> is obligatory.  The element <p> may contain an otherwise undifferentiated prose description of the physical material of the manuscript.  If desired, <p> can contain the element <material> in order to ensure more accurate searching on this important aspect of a manuscript (see below for a discussion of <material>).

Examples:

<support><p>Carta di pessima qualitá</p></support>

<support><p>Pergamena</p></support>

<support><p>Written on 166 <material>membrane</material> leaves of average quality, preceded by two 19th-century <material>paper</material> flyleaves and one medieval <material>membrane</material> flyleaf, and followed by one <material>paper</material> flyleaf.</p></support>

WATERMARKS

<watermarks> contains the information regarding the imprint, usually figurative, left on paper during its manufacture by the curved or interwoven slim wire sewn or soldered onto the frame on which paper is made. 

Note that the <watermarks> element may occur inside any other element, of which the most obvious is <p>, while it itself requires an internal <p>.

Examples:

<support><p>Written on 139 leaves of good quality <material>paper</material>, preceded and followed by a membrane flyleaf conjugate respectively with the front and back pastedowns.  <watermarks><p>The watermark of fols. 1-94 is an ox’s head in a circle, not identified in Briquet or Piccard <title>Wasserzeichen Ochsenkopf</title>.</p></watermarks></p></support>

<history><origin><p>Written during the final years of the fifteenth century, as suggested by the identification of the watermarks <watermarks rend="ital"><p>Ange</p></watermarks> and <watermarks rend="ital"><p>Oiseau</p></watermarks>.</p></origin></history>

PALIMPSEST

<palimpsest>  describes reused writing support for a manuscript from which the previous text or set of signs was made more or less to disappear (whether by erasing it or washing it or scraping it) so that a new text could be placed on the same support.

The element <palimpsest> requires, internal to itself, either a <p> or <msWriting>.  Its optional "type" attribute may be applied to the distinction between the earlier underlying text, and the more recent overwriting text.

Examples:

<support><p><material>Parchment</material> (prepared in the southern manner, used for outer bifolia of all quires and for the central bifolia of quires 1-5; <palimpsest><p>at least ff. 6 and 13, palimpsest</p></palimpsest>).</p></support>

<bindingDesc><p>Modern binding retaining the endleaves presumably from the original binding, since they contain a series of <palimpsest type="overlying"><p>pentrials</p></palimpsest> in a fifteenth-century hand and, as palimpsest, the <palimpsest type="underlying"><p>register of a notary</p></palimpsest> located in the same area, given the place names cited.</p></bindingDesc>

<msDescription><msIdentifier><idno>Vatic. 5757</idno></msIdentifier><history><p><palimpsest><p> rescriptus</p></palimpsest>, ut videtur, Bobii, saeculo VII exeunte vel ineunte VIII ita ut libro Ciceronis de republica litteris uncialibus elegantissimis et praegrandibus, fortasse saeculo IV exarato superponeretur enarratio S. Augustini in Psalmos litteris iisdem sed multo humilioribus.</p></history></msDescription>

MATERIAL

<material> denotes the physical substance of which the manuscript itself, or any of its component parts, is composed.

The element <material> may be inserted within any <p> and therefore can be used, for example, within tagging such as  <support><p>, <layout><p>, <decoration><p>, <decoNote><p>, <bindingDesc><p> or <binding><p> to encode substances commonly used in the production of manuscripts, such as paper, lead, gold, morocco, and so on.  Its "type" attribute allows for encoding that can produce very specific search results.

Examples:

<support><p>Cod. en <material type="parchment">perg.</material> y <material type="paper">papel</material>.</p></support>

<layout><p>2 colonne di 36 righe, con rigatura alla <material>mina di piombo</material>.</p></layout>

<decoration><p>The artist has used <material>gold leaf</material> for the haloes and <material>painted gold</material> for the highlights on the robes.</p></decoration>

<bindingDesc><p>Bound, <origDate>s. XV</origDate>, in blind stamped <material>morocco</material> over <material>wooden boards</material>.  Green <material>silk</material> fore edge ties were added at a later date.</p></bindingDesc>

FORMAT

<format> is used to encode two different but related situations:  1) the number of times the writing support has been folded (mainly in reference to paper); and 2) the general designation of size of the codex when its measurements are not expressed by precise dimensions (whether parchment or paper).

The same sets of words  are used by cataloguers to express both situations; therefore, the encoder, when working with legacy data, may feel somewhat uncertain  as to the cataloguer’s intention.  It is because of this ambiguity that we have not attempted to distinguish the two meanings in this dtd.

Examples:

<support><p><material>Paper</material>in <format>4<hi rend="superscript">to</hi></format> with the watermark consequently hidden in the gutter.</p></support> [used here in reference to the folding of a sheet of paper]

<format>Small quarto</format>

<physDesc><p><format>16<hi rend="superscript">o</hi></format>,</p> <extent>573 bln.</extent> <decoration><p>initialen.</p></decoration></physDesc> [physical description of a 13th century bible, clearly of the "pocket" variety, certainly on parchment, and certainly very small.  In addition, the cataloguer’s introduction specifies that he is using the terms sextodecimo, duodecimo, octavo, quarto and folio as general size designations.]

Note:  The element <format> is not used when the dimensions are given with cardinal numbers; <dimensions> is the method for encoding that information, as in this example:  <p><dimensions>267 x 180 mm</dimensions></p>

DIMENSIONS

<dimensions>
encodes the height, width and depth of the manuscript, or of a part of a manuscript.
<height>
encodes the height of a manuscript.
<width>
encodes the width of a manuscript.
<depth>
encodes the depth of a manuscript.

The element <dimensions>  with its subelements for <height>, <width> and <depth> can be inserted within any <p>.  The default unit of measure is millimeters.  If, however, the cataloguer prefers to use a different unit of measure, he should signal it with the attribute "units."  The "type" attribute distinguishes measurements of different parts of the manuscript, such as the book block, or the ruled space or the written space (the decision to measure "ruled" or "written" is usually institutional).  The subelement <depth> is often used in-house by libraries to construct cases for specific codices, or to  estimate amounts of shelf length needed; traditionally the concept of depth is implied by the number of leaves in the book, which is encoded here as <extent>.

Examples:

<p><material>Membr.</material>;<dimensions type="block">mm. 225 x 155</dimensions>; <dimensions type="written">specchio mm. 150 x 95 </dimensions>.</p>

<p>Good quality <material>paper</material>, without watermarks.  Overall size <dimensions type="block">271 x 217 mm.</dimensions> and the column as <dimensions type="ruled">c.210 x 130 mm</dimensions>, with 39-40 lines.</p>

<decoration><p>Miniatures, <dimensions units="inches">5 x 3 inches</dimensions></p></dimensions>

EXTENT

<extent> supplies information about the quantity of leaves, pages or membranes contained in a codex or its parts.

For the purposes of this dtd, it operates internally to <physDesc>, which itself lies inside <msDescription> and inside <msPart>.  If  there is reason to encode of the span of leaves occupied by a text, the location inhabited by a miniature, the placement of the 18th century archivist’s notes, etc. one should use the element <locus>.

Examples:

<extent>pp. 73</extent>

<extent>cc. I + 382 + I</extent><p>numerate 381, in quanto vi sono due numeri 283</p>

<physDesc><support><p><material>Papier</material>,</p></support><extent>337 ff.</extent><layout><p>2 Spalten zu 42-44 Zeilen</p></layout><p><dimensions type="block">225 x 140</dimensions>, Schriftsspiegel<dimensions type="written">170 x 100</dimensions>.  Die ganze Sammelhandschrift von derselben Hand geschrieben.</p> </physDesc>

But:

<msItem><locus>ff. 1-6v</locus><summary>Calendar in red and black ink.</summary></msItem>

<decoNote><p><locus>f. 32v</locus>, miniature of the Nativity of the Virgin.</p></decoNote>

<provenance><p>On the <locus>dorse</locus>, summary of the document’s contents and two sets of collection and box numbers, with the first crossed out.</p></provenance>

COLLATION

<collation> contains the description of the gathering-by-gathering composition of a codex, or segment of a codex, sometimes expressed as a formula, and sometimes in prose.

The structure is assumed to consist of bifolia nested one inside another, already or about to be stitched together as a unit.  There will most often be an even number to express the number of leaves of a given quire, or the number that gradually sums up the leaves in sequential quires (as in the cataloguing system used in Germany).  There will be an odd number when single leaves have been added to or cut from the basic structure.

The element <collation> must contain <p>, which can then contain <hi> to indicate superscripts for formatting.  If more complex encoding is desirable, <collation>’s <p> may contain <formula>, which must have the attribute <notation>.

Note:  This element should not be used for "collation" in the sense of comparison of textual variants.

Examples:

<collation><p>1-7<hi rend="superscript">10</hi></p></collation>

<physDesc><p>Manuscript in </p><collation><p>seven quires of ten leaves.</p></collation></physDesc>

<collation><p>3 V<hi rend="superscript">30</hi> + IV<hi rend="superscript">38</hi> + 4 V<hi rend="superscript">78</hi> + 2 IV<hi rend="superscript">94</hi> + II<hi rend="superscript">98</hi></p></collation>

CATCHWORDS

<catchwords>  provides a description of the system of word(s), written in the lower margin of the last leaf verso of a gathering, as a preview of the first word(s) of the first leaf recto of the successive gathering, to ensure correcting ordering of the quires by the binder.

Although ante litteram, the <catchword> element may also be used to encode the description of pre-twelfth-century ordering of quires by roman numerals on the first leaf recto or the last leaf verso; we have chosen to amalgamate this practice to that of catchwords, since both systems order only the quires (and not the leaves), and since both systems occur once per gathering.

This element may also be used for the mainly post-medieval habit of placing preview words at the foot of each page, as an aid to the reader.

Examples:

<catchwords>Quires signed on the last leaf verso in roman numerals.</catchwords>

<catchwords>The only evidence of the use of catchwords is a tiny trace of one on fol. 144v.</catchwords>

 <collation><p>Hauptsächlich Quinionen</p></collation> <catchwords>mit Reklamanten</catchwords>

<catchwords>Quelques réclames, le plus souvent rognées.</catchwords>

<catchwords>Vertical catchwords in the hand of the scribe placed along the inner bounding line, reading from top to bottom.</catchwords>

SIGNATURES

<signatures>  denotes the system of progressive marking of fascicules and/or of leaves, usually only through the first half of the gathering, so that they may be assembled in correct order by the binder. 

This element is used for either the leaf signatures, or the combination of quire and leaf signatures.  It is used whether the marking is alphabetic, alphanumeric or ad hoc (such as Ave-maria-gratia-plena on the first four leaves of each quire; or increasing numbers of circles, or slashes, etc.).  This element will also encompass mention of the mark that indicates arrival at the center of the gathering, usually in the gutter of  the second recto of the centermost bifolium.

Note:  For the numbering of the quires, one by one, that occurs in pre-twelfth century manuscripts, see <catchwords>.

Examples:

<signatures>Quires 23-32 have no signatures.</signatures>

<signatures>At the bottom of the first four leaves of quires 1-14 are the remains of a series of quire signatures a-o plus roman figures in a cursive hand of the fourteenth century.</signatures>

LAYOUT

<layout> describes the way in which text is arranged on the page, specifying, for example, the disposition and shape of the prick marks, the

number and medium of the ruled or written lines, and the number of columns. 

Although the element is devised for the disposition of the text on the page, it may also contain the description of the way the illustration is integrated with the text, if both occur on regular patterns.

The <layout> element contains <p>. The attributes are provided to allow more intensive markup of some of these codicological features.

Examples:

<layout><p>25-32 long lines ruled in lead.</p></layout>

<layout><p>scr. a due col. di 47 linee ciascuna.</p></layout>

<layout columns="2" writtenLines="47"> p>scr. a due col. di 47 linee ciascuna.</p></layout>

<layout><p>Written in 3 columns, with 8 lines of text and interlinear glosses in the center, and up to 26 lines of gloss in the outer two columns.  Double vertical bounding lines ruled in hard point on hair side.  Text lines ruled faintly in lead.  Remains of prickings in upper, lower, and outer (for 8 lines of text only) margins.</p></layout>

THE WRITING IN A MANUSCRIPT

A number of elements may be employed to describe the writing in a codex; none of them is obligatory; the choice between the various combinations of elements will depend upon the greater or lesser intensity of encoding.

The elements of use in this context are:

<name role="scribe">
<scriptTerm>
<msWriting>
< handDesc>

<writingSystem>
<handShift>

<name role="scribe">
encodes the indexable name of the scribe, omitting descriptive adjectives.
<scriptTerm>
 encodes the indexable name of a script, omitting descriptive adjectives unless the adjectives have become necessary as distinguishers.
<msWriting>
contains the full range of descriptive discussion on paleographic matters of a codex.
<handDesc>
gathers the discussion of a single scribe’s work within the manuscript.
<writingSystem>
is used to designate a manuscript’s writing system or alphabet, of which the script is the particular embodiment.
<handShift>
marks the beginning of a sequence of text written by a new scribe.

The simplest level of useful markup will employ the first two elements in this list:  <name role="scribe"> and <scriptTerm> to encode precisely those words that would appear in a traditional index, the name of the scribe and the name of the script, without any qualifiers that the cataloguer may have used to help the reader visualize the scribe’s hand or the script.  The name of the scribe is sometimes the indication of a persona, rather than an actual name, such as the "Scribe of Bologna University 1452."

If the unqualified script name remains unhelpfully generic ("gothic" being the prime example), and if the particular script is often designated with a specific adjective, the encoder might encapsulate that adjective within the tags as part of the script’s name (e. g. "rounded gothic"; "Burgundian b%acirc;tarde"). 

Particularly when the material being encoded derives from several sources (legacy data from more than one catalogue), the "reg" attribute will help ensure equal retrieval of the same script that has been travelling under different names.  For example, the encoding of a 15th century library inventory might offer reg="humanistic" when the text states "littera antiqua."  The values for the "reg" attribute should be applied consistently across the body of material, or the return in functionality will not be high. 

The element <scriptTerm> also carries a "level" attribute with a closed list of values:  formata, media or currens.  We recognize that a three-way distinction in theory will not always apply satisfactorily in practice, and that the variations in levels of formality of a script are much higher.  At the same time, the option of more levels risks more uneven judgments by numerous cataloguers, and lowers the use of the attribute for retrieval.

Both elements, <name role="scribe"> and <scriptTerm>, may occur inside any <p>, and thus can be used, for example, in discussions of binding or of provenance.

A more complex encoding, useful especially if the material being marked up is of strong paleographic interest, adds the second two elements:  <msWriting> and <handDesc>.  The first, <msWriting> wraps together, potentially under <head>, an entire description of the writing in a given document, potentially by a number of scribes, each of whose individual efforts, potentially comprising several different kinds of scripts, is described under multiple occurrences of <handDesc>.  The <msWriting> element also has an optional attribute, "hands," that serves to count the number of scribes.

The fifth element, <writingSystem>, describes the writing system for a body of written material; its value in descriptions of western manuscripts lies especially in distinguishing those written in the Latin alphabet versus those in a Greek alphabet, but its use is clearly extensible to designation of manuscripts in other alphabets, such as Cyrillic or Hebrew.  The element separates the encoding of writing system from that of language; the latter is encoded with the <textLang> element.

The final element, <handShift>, is an empty marker, a milestone, used to designate the moment of change from one scribe to another internally to a piece of transcribed text.  It would occur in the present context, that of catalogued descriptions of manuscripts, rather infrequently; it is more likely to be used in transcriptions of manuscripts.

Note: The standard TEI element <hand> is used in the TEI header, in much the same way in which <handDesc> is used in this context.

Examples:

<physDesc><p>Manuscrit copié par <name role="scribe">Martin Olivier</name> et <name role="scribe">Raduliaus (?)</name> pour <name role="owner">Gérard de Porte</name>.</p></physDesc>

<msHeading><note><scriptTerm>Karolingische Minuskel</scriptTerm> von einer Hand.</note></msHeading>

<p>Copied in a formal liturgical <scriptTerm level="formata">rounded gothic</scriptTerm> script in two sizes, according to function.</p>

<physDesc><msWriting><handDesc><p>Written in a skilful, upright, late fifteenth century Florentine formal <scriptTerm>humanistic</scriptTerm> hand with some cursive features; rather square in effect.  <msIdentifier type="cited"><institution reg="British Library"> B. M.</institution><idno>Burney 190</idno></msIdentifier>, Juvenal, Florentine, late fifteenth century, with Tedaldi arms, was possibly written by the same scribe, in a rather more formal humanistic script, using an open instead of closed minuscule "a."</p></handDesc></msWriting></physDesc>

<colophon>Laus et gloria christo, deo gratias.  Ego <name role="scribe">Venceslaus de alamania alta</name> complevi sermones suprascriptos.</colophon>

<msWriting><p>Cette bible fut transcrite par </p><handDesc><p><name role="scribe" reg="Sigismund Imthurn">Sigismond In Turri</name> de Sankt Pölten.  </p></handDesc><p>Quelques mois plus tard</p><handDesc><p>un second copiste, <name role="scribe">Etienne Engelsbruck</name> de Haslach, recopie les gloses en marges due texte, gloses rassemblées d’aprés ses notes recueillies durant ses cours de théologie suivis ´ l’Université de Vienne.</p></handDesc></msWriting>

<physDesc><msWriting><p><head rend="bold">Script</head>Fols. 1<hi rend="superscript">r</hi> - 94<hi rend="superscript">v</hi> are in one hand, <scriptTerm>anglicana formata with secretary influence</scriptTerm>.  Rubrics are in a <scriptTerm>bastard anglicana</scriptTerm>.  The same scribe uses a smaller script, pure <scriptTerm>secretary</scriptTerm> for fols. 95<hi rend="superscript">r</hi> – 139<hi rend="superscript">v</hi>.</p></msWriting></physDesc>

<physDesc><p>Copied in a controlled, elegant </p><writingSystem>Greek</writingSystem><p> <scriptTerm>Fettaugenmode</scriptTerm> script.</p></physDesc>

<history><p>On f. 61, above the scribe's signature, his motto <motto>A deo fortuna dependet</motto> copied in</p></history><physDesc><writingSystem>Greek</writingSystem><p>letters</p></physDesc>

<explicit> Et ita sufficiat pro presenti capitulo, et per quos pro toto isto nono libro almansoris.  <handShift>Explicit totalis exposicio Gerardi de solo super novum almansoris rasis.</explicit> [this encoding signals the change of scribe for the final line of the explicit]

PUNCTUATION

<punctuation> incorporates comments about the punctuation in a manuscript.

This element must contain <p>.

Examples:

<punctuation><p>Seven-and-point (?) punctuation.</p></punctuation>

<punctuation><p>Punctuation by <hi rend="ital">virgula suspensiva</hi> for some (but not all) minor medial pauses, colon (with a single hairline between the two points) for major medial pauses, and <hi rend="ital">punctus</hi>  for all other pauses except the <hi rend="ital">interrogatio</hi>.  The ends of texts are indicated occasionally by either <hi rend="ital">virgula suspensiva</hi>  followed by <hi rend="ital">punctus</hi>, or by a 9-shaped stroke (representing the <hi rend="ital">positura</hi>).</p></punctuation>

MUSIC NOTATION

<musicNotation> is the element that encloses a description of the form of musical notation employed.

It must contain <p>.

Examples:

<musicNotation><p>Neumes in campo aperto of the St. Gall type.</p></musicNotation>

<musicNotation><p>Polyphonic, for three voices, with one designated as "contratenore."</p></musicNotation>

<musicNotation><p>Black square notation on 4-line red staves.</p></musicNotation>

THE DECORATION OF A MANUSCRIPT

A number of elements may be employed to describe the decoration of a codex; none of them is obligatory; the choice between the various combinations of elements will depend upon the greater or lesser intensity of encoding.

The elements of use in this context are:

<name role="artist">
<decoration>
<decoNote>

<name role="artist">

encodes the indexable name of the artist or the artistic persona, omitting descriptive adjectives.

<decoration>

gathers together, potentially under a heading, one or more discrete units (whether as <p> or as <decoNote>) that discuss the decoration of a manuscript.

<decoNote>
is the basic unit for describing a specific aspect of a manuscript’s decoration.

The first element listed here, <name role="artist">, may be used inside any <p> in any part of the description of the manuscript.  It should encode only the artist’s name, such as would appear in a traditional index; it may also be used to encode the anonymous naming of an otherwise unknown artistic personage.

The second element occurs only within <physDesc>, since it is the major mechanism for encoding the illumination.  It may contain a heading, marked with the element <head>, to set off its discussion from that of the other large bodies of information in the description; it may incorporate a number of general paragraphs (<p>) or the more specific kind of paragraph that is reserved for discussions of the decoration of a manuscript, <decoNote>.

The third element, <decoNote>, is the one that will allow for the most specific encoding, since it carries a number of attributes.  Other than the standard four global attributes, <decoNote> also proposes:

type
subtype
size
technique
style
quality

all of which have open values, to be determined by the cataloguer/encoder at will.  The remaining two attributes on <decoNote> have a closed list of values:  yes, no, and non-applicable, but without specified default:

figurative (yes | no | na)
illustrative (y | n | n-a)

Definitions of the attributes are:

type
specifies the part of the manuscript that has been decorated, e. g.,  miniature, initial, border, etc.
subtype
permits further limitation on the kind of initial, border, etc.
size
allows for encoding of sizes expressed in relation to the page
technique
refers to the manner of producing the decoration
style
encodes the broad concepts of art-historical periods
quality
allows the cataloguer to express a value judgment
figurative
announces that the decoration contains (or does not contain) representations of actual beings or objects
illustrative
announces that the decoration contains (or does not contain) pedagogical matter such as charts, diagrams, tables, etc.
        

The element <decoNote> may be used internally to <decoration>, <binding>and <msItem>; it requires <p> as subelement. 

Note: The elements, <decoration> and <decoNote>, are not intended to encode specific instances of illumination, but more broadly the discussion about the illumination’s characteristics.  An alternate method for encoding all the miniatures in a manuscript, for example, would be to insert into <decoration> or <decoNote>, a <p> followed by the <list> element with its repeatable subelement <item>.

Examples:

<note>The miniature was attributed by Berenson to <name role="artist">Giovanni di Paolo</name>.</note>

<p>De minaturen van het handschrift kunnen o.i. zeker worden toegeschreven aan de <name role="artist">Meester van het Dresdens Getijdenboek</name>.</p>

<physDesc><decoration><head rend="capitals">Decoration</head><p>Five 7-line miniatures.</p><p>Numerous painted initials in gold and colours.</p></decoration></physDesc>

<physDesc><decoration><head rend="ital">Ausstattung:</head><decoNote><p>Einzeilige Initialen in Gold auf blauem, rotem und grünem Grund mit Goldfiligran, zweizeilige Initialen in Gold mit alternierend roter oder grüner Füllung auf blauem Grund mit goldenem Filigranbesatz.  Miniaturen und Zierseiten allseitig mit Bordüren aus Blüten, Ranken, Putti sowie Medaillons und Felder mit z. T. figürlichen Motiven.  2 grosse Miniaturen auf Einzelblättern.</p></decoNote></decoration></physDesc>

<msItem>Treatise on urines with <decoNote><p>sketches of bottles containing variously colored liquids.</p></decoNote></msItem>

<bindingDesc><binding><p>Polished brown calf over wooden boards, </p><decoNote><p>panel-stamped with an image of St. John the Evangelist.</p></decoNote></binding></bindingDesc>

<decoration><decoNote type="miniatures" size="full page"><p>Full page miniatures at each chapter division, attributed to the <name role="artist"> Maîumflex;tre aux yeux bridés</name>. <note>We are grateful to Prof. James Marrow for the attribution of the artist.</note></p></decoNote><decoNote type="initials" subtype="historiated"><p>Historiated initials by a less capable assistant.</p></decoNote><decoNote type="initials" subtype="painted"><p>Painted initials only in the first three quires, thereafter not completed.</p></decoNote></decoration>

<decoration><decoNote technique="grisaille" figurative="yes"><p>Miniatures on the versos of added singletons.</p></decoNote></decoration>

<decoration><decoNote type="initial" subtype="inhabited" style="romanesque"><p>Inventive initials inhabited by many small climbing men.</p></decoNote></decoration>

<decoration><p><list><item>f. 8, Annunciation</item><item>f. 13, Nativity</item><item>f. 19, Annunciation to the shepherds</item></list></p></decoration>

FOLIATION

<foliation> provides for mark-up of the numbering system(s) applied to the leaves or pages of a manuscript.

It may be used to indicate the scheme, medium or location of folio numbers, page numbers, column numbers or line numbers written in the manuscript, frequently including a statement about the date when the numbering was done.  It must contain at least one <p>.

Examples:

<foliation><p>Neuere Foliierung, die auch das Vorsatzblatt mitgezählt hat.</p></foliation>

<foliation><p>Original foliation in red roman numerals in the middle of the outer margin of each recto; pagination in an early modern hand in the lower outer corner of each page.</p></foliation>

<foliation><p>Lines numbered by 5s in arabic numerals in the space between the columns, in the manner associated with Oxford.</p></foliation>

ADDITIONS

<additions> offers encoding to record any written or drawn additions to the original state of the manuscript, such as marginalia, scribblings, doodles.

The element serves for the miscellaneous and unofficial additions to the manuscript, i.e. those that are not properly handled by the more specific elements such as <msItem>, <decoration> or <provenance>.  The <additions> element is a subelement of <physDesc>; it must contain <p>.

Examples:

<additions><p>Doodles on most leaves, possibly by children, and often quite amusing.</p></additions>

<physDesc><additions><p>Quelques annotations marginales des XVIe et XVIIe s.</p></additions></physDesc>

CONDITION

<condition> summarizes the physical state of a manuscript.

This element will probably be of most use to describe various unpleasant conditions (catalogue descriptions seldom comment on the excellent state of a codex).  In any case, it should not be used to describe technical conservation repairs to a manuscript; these are more appropriately described in <adminInfo>’s subelement <custodialHist>.  The <condition> element requires an internal <p>.

Examples:

<condition><p>Stitching, especially of the first quires, very loose; front cover detached.</p></condition>

<condition><p>Una miniatura a f. 26 è stata tagliata.</p></condition>

<condition><p>The manuscript shows signs of damage from water and mould on its outermost leaves.</p></condition>

THE BINDING OF A MANUSCRIPT

A number of elements may be employed to describe the binding of a codex; none of them is obligatory; the choice between the various combinations of elements will depend upon the greater or lesser intensity of encoding.

The elements of use in this context are:

<name role="binder">
<bindingDesc>
<binding>
<msPart>


<name role="binder">
encodes the indexable name of the binder, omitting descriptive adjectives.
<bindingDesc>
gathers together, under an optional heading, one or more discrete units (whether as <p> or as <binding>) that discuss the binding(s) that a codex may have received over a period of time.
<binding>
is the basic unit for describing the binding of a codex at one point in time
<msPart>
[See definition elsewhere.]

The element <name role="binder"> encodes the name of the binder, whether as an actual name or as the naming of an anonymous persona.  The attribute "reg" may be used to signal the normalized and complete form of the name.

The element <bindingDesc> occurs as a subelement of <physDesc>.  The subelements of <bindingDesc> are an optional heading, paragraph(s) that might enclose other elements such as <decoNote> (itself with its own subelement <p>), or <origDate>, and as many instances of <binding> as are necessary.

The element <binding> requires as subelements either <decoNote> (for discussions of the decoration on the binding), or <p> (more generally).  If more than one binding is known for a given codex, <binding> may be repeated as often as necessary.  The choice of attibutes on this element are intended to offer tighter searching of the date of the binding:

notBefore
notAfter

certainty (high | medium | low)

evidence (internal | external | attributed)

contemporary (yes | no | unk)

Of these, the most relevant is the situation offered by <binding contemporary="yes">.  In other circumstances the <origDate> element is simpler to apply and more specific.

The element <msPart> is recorded here as a reminder that flyleaves in bindings, and the covers of bindings themselves are often parts of manuscripts other than the primary book under consideration, with their own texts, places and dates of origin, and physical characteristics.  For proper searching and retrieval, these fragments should be catalogued under <msPart> with the appropriate subelements.

Examples:

<bindingDesc><p>Bound by <name role="binder" reg="Payne, Roger">Payne</name> in his usual style.</p></bindingDesc>

<bindingDesc><binding contemporary="yes"><p>Bound by the <name type="org" role="binder">Pecking Crow bindery in Paris during the middle years of the sixteenth century.</p></binding></bindingDesc>

<physDesc><bindingDesc><head rend="ital">Einband:</head><p>Weisses Schweinsleder mit reicher, floraler Goldprägung im historisierenden Stil über Pappe; Goldschnitt; 19.Jahrhundert.</p></bindingDesc></physDesc>

<physDesc><bindingDesc><binding><p>Bound, <origDate>s. XVIII (?)</origDate>, in <material>diced russia leather</material> retaining most of the original fifteenth century metal ornaments (but with some replacements) as well as the heavy wooden boards; on each cover:  </p><decoNote><p>alternating circular stamps of the Holy Monogram, a sunburst, and a flower</p></decoNote><p>; on the cornerpieces, one of which is missing, a </p><decoNote><p>rectangular stamp of the Agnus Dei; </p></decoNote><p> rebacked during the nineteenth (?) century.</p></binding></bindingDesc></physDesc>

<bindingDesc><binding><p>Bound in modern vellum. </p></binding><binding><p>Previous binding in tree calf retained in accompanying case.</p></binding></bindingDesc>

<physDesc><bindingDesc><p>Bound in contemporary blind-stamped leather.</p></bindingDesc></physDesc><msPart><history><p>The flyleaves are from <title>missal</title>,<origPlace>France</origPlace>,<origDate>s. XIV</origDate>.</p></history></msPart>