Item Content and Transcription

Abbreviations and Expansions

Common abbreviations: the, that, ampersands

For common abbreviations like "yt" for "that", "ye" for "the" that appear often throughout the text, you do not need to use choice or expan; instead, you can simply tag the abbreviation using the abbr tag with a type value with the correct word:

In most cases, ampersands do not need to be tagged at all--they will be automatically expanded in processing. However, do note that ampersands are special characters in XML and cannot be represented simply by the & character and must be typed in as

&amp;

However, if the ampersand does not expand to "and", then you must tag the abbreviation using the full choice, abbr, and expan method below:

<choice><abbr>&c.</abbr><expan>et cetera</expan></choice>

Abbreviation	Expanded Form	Encoding
Broyr	Brother	<abbr type="Brother">Broyr</abbr>
anoyr	another	<abbr type="another">anoyr</abbr>
fm	fm	<abbr type="fm">fm</abbr>
qre	qre	<abbr type="qre">qre</abbr>
yt	that	<abbr type="that">yt</abbr>
ye	the	<abbr type="the">ye</abbr>
ym	them	<abbr type="them">ym</abbr>
yn	then	<abbr type="then">yn</abbr>
yre	there	<abbr type="there">yre</abbr>
ys	this	<abbr type="this">ys</abbr>
qt	what	<abbr type="what">qt</abbr>
qn	when	<abbr type="when">qn</abbr>
qch	which	<abbr type="which">qch</abbr>
qo	who	<abbr type="who">qo</abbr>
qm	whom	<abbr type="whom">qm</abbr>
wt	with	<abbr type="with">wt</abbr>
yr	your	<abbr type="your">yr</abbr>

Uncommon abbreviations

Uncommon or unfamiliar abbreviations—like Revd—will often need to be expanded in order to provide clarity and to facilitate searching. To do so, first tag the abbreviation with the abbr element, then tag your expansion with an expan, and then wrap both in a choice element. For example:

<title>Letter from the <choice><abbr>Revd.</abbr><expan>Reverend</expan></choice> Mr. <persName>Lyon</persName></title>

Supplying, Regularizing, and Flagging Incorrect Text

Since the LiM represents multiple voices, we need to be explicit in our signalling of our editorial interventions and judicious in our appraisal of what kind of intervention we need to make.

There are three situations where editorial intervention is necessary:

Text that must be supplied due to obscured page images, ink blots, or other obfuscating factors

Non-normative spelling that we want to regularize for searching purposes

Obvious errors

This list is in order of descending preference; in most cases, it is too difficult to determine if something is truly an "error" that can and should be corrected.

Supplying Text

Tag text that has been obscured in some way (usually lost in the margin or the fold of the book) using the supplied tag:

<lb/><supplied>I</supplied>rons upon him, which he looked upon

If you are not certain about the supplied text, then use the cert attribute on supplied with a value of

low

:

provided that he
<lb/><supplied cert="low">w</supplied>ould not sit down, but come out as

Regularizing Text

Tag cases where a word differs from its normative form or spelling and would benefit from regularization to help with searching by tagging the original word using orig, the regularized version using reg, and wrapping both in a choice element:

I had some share the effects of <choice><orig>wich</orig><reg>which</reg></choice>

Note that the names of people, places, organizations, and objects should not be regularized. Since all people, places, and organizations will be tagged and associated with an person, place, or organization entity, it is unnecessary to regularize these proper names.

Corrections

While we do not correct erroneous text, we do flag instances where the word or phrase seems incorrect in some way by using the sic tag:

<lb/>of Distinction, wait <sic>of</sic> General

Additions, Deletions, and Substitutions

Additions and deletions in text may appear in the text either in isolation or as part of a single textual event.

Additions

In all cases, additions should be marked with the add tag with its location encoded using the place attribute. For example:

In this case, text has been added above the line, which would be encoded like so:

<lb/>Donell, <add place="above">George Moir the</add> Laird of Leckie,

Here, we can also tag the "George Moir the Laird of Leckie" as a persName and point to his xml:id:

<lb/>Donell, <persName ref="prs:MOIRG1"><add place="above">George Moir the</add> Laird of Leckie</persName>

Deletions

Tag text that has been marked for deleting using the del element:

<title>Written in <del>in</del> April</title>

If the deletion has obscured the text such that it makes the text difficult (but not impossible) to transcribe, nest the unclear tag within the del; as above, add an editorial note if warrants further explanation or clarification is needed (note, however, that the note is beside the del element):

<title>While I pondered weak and <del><unclear>wary</unclear></del><note type="editorial">Possibly <mentioned>weary</mentioned>.</note></title>

If the text is rendered completely illegible, then use a gap element with an explanatory desc within the del:

<title>While I pondered weak and <del><gap><desc>Illegible</desc></gap></del></title>

Substitutions

When an addition and deletion are related (i.e. the additional text should take the place of the deleted text), then we can encode that relationship by wrapping the add and del with the substitution element subst. For example:

The substitions are highlighted above; in each case, there is a deletion (with illegible content) and an addition above. In the first example, we would encode the additional "some" using the

add

element:

and the deleted word using del with a nested gap that examples the extent of the cancelled text:

To signal that these are a related substitution, we wrap the

add

and the

del

in a subst. So putting this all together:

Unclear Text

Tag text that you are unable to transcribe text that is partially obscured for some reason (illegible writing, scan is missing or incompletel, external damage like ink smudges, etc) using the unclear element:

<unclear>MacDonald</unclear>

If you'd like to expand on what is unclear, or possible other readings, then add a note beside the unclear using the note element with type=editorial

<unclear>MacDonald</unclear> <note type="editorial">Quite likely MacDonald, but could also be McDonald.</note>

If you are unable to transcribe the text at all or the text has been removed completely (i.e. the page is burnt; the text is cut off in the facsimile), use the gap element with an explanatory desc:

<gap><desc>First six words illegible due to scan.</desc></gap>

Elisions and Omissions

Text that has been purposefully omitted/elided by Forbes (e.g. for political or social reasons) should be tagged using the ellipsis element in combination with a metamark and, if the omitted text is known, a supplied. Consider the following line from "Copy of a Letter to Captain Malcolm MacLeod of Castle in Raaza" (v02.0323.01):

First, the gap between the "R" and "on" would be marked as an omission using the ellipsis element like so:

He said, in the Event of a R<ellipsis></ellipsis>on,

This tells us that there is text that has been elided; to encode the space, we use a metamark with a child space element within the ellipsis:

He said, in the Event of a R<ellipsis><metamark><space/></metamark></ellipsis>on,

Note that the metamark element allows you to use either a space element or text content (but not both); if, for instance, the elision was signaled using em-dashes (i.e. "Mr. F———"), then you would put the "———" as the text value of the metamark:

I went to the store with Mr. F<ellipsis><metamark>———</metamark></ellipsis>

To add the supplied text, place a supplied after the metamark like so. Make sure to use a resp that points to your team ID to credit yourself; if you are uncertain about your interpretation, you can use a cert (as described above) with a value of "low":

He said, in the Event of a R<ellipsis><metamark><space/></metamark><supplied resp="team:JT1" cert="low">estoration</supplied></ellipsis>

Elisions: Multiple Choices (Prince vs Pretender)

In cases of an initial P with a set of spaces (which can be either Prince or Pretender), place a choice within the supplied with a seg for each "rince" and "retender" like so:

If The P<ellipsis>
  <metamark><space/></metamark>
  <choice>
     <seg>rince</seg>
     <seg>retender</seg>
  </choice>
</ellipsis> returns...

Foreign Language

To tag that something is in a foreign language, use the xml:lang attribute with an ISO language tag value for that language:

de	German
en	English
es	Spanish
fr	French
ga	Irish
gd	Scottish Gaelic
ghc	Hiberno-Scottish Gaelic
gr	Greek
it	Italian
la	Latin
sco	Scots

How and where you attach that xml:lang depends on whether the segment of foreign language is entirely contained by a single element. For instance, if an entire title in an item is in French, then you can place the xml:lang directly on the item:

<item><title xml:lang="fr">Le petit chien</title></item>

If the foreign language appears has no logical wrapper (i.e. a single foreign phrase in a sentence or a foreign word), then use the foreign element:

<item><title>This example was created <foreign xml:lang="la">ex nihilo</foreign></title></item>