Text editions are manyfold in their approach and can be quite complex. The aim of BPT, and/or a workflow based on it, is to put plain text in central place is to reduce that complexity, and to give different matters a separate place in a workflow. The creation of XML or the rendering of a publication are examples of that.
Complexity can be added by subsequent phases of the workflow, where each phase is handled by scripts. The purpose of the BPT script is to produce a text edition consisting of a base text and one or more sets of annotations in plain text. Its focus is on clarity and consistency.
BPT requires that all text is in Unicode, i.e. uses Unicode fonts, and is encoded in UTF-8.
BPT defines line beginnings and line ends. A line beginning is indicated by a line line. A line end is indicated by the absence of any characters after the last one on a line.
Authors are responsible for clearly indicated line beginnings and line ends.
In many plain text editors, pressing the ENTER key ends a line and starts a new line. This normally results in a U+000A, visible as the beginning of a new line.
For BPT, only the the presence of the Unicode point U+000A is decisive. Any other Unicode point is replaced (using the BPT script) by it.
first line
second line
BPT requires line numbers (if required) to be preprended to all lines. (This does not mean they are rendered as such in the publication).
Authors are responsible for clearly indicated line numbers, for example at every line, or every fifth line. THe BPT script converts these to all lines.
Line numbers are entered by adding a number, followed by a full stop and a space.
The Unicode points are U+0031 (or any other number), U+002E, and U+0020. Any other Unicode point is replaced (using the BPT script) by these.
1. first line
2. second line
There are several types, e.g.
Such marginalia should be attached to (clean, free of anchors and such like) base text via URNs, but these have no place in human-readable BPT. What to do? Treat them like notes.
Ἔστι μὲν καὶ πᾶσιν ὑαῖν τοῖς τὴν ἰατρικὴν αετιοῦσι[^1]
[^1] 71 Merc.
In this example it is unclear if the annotation refers to to the word or the line. Disambiguate as follows:
1.[^1] Ἔστι μὲν καὶ πᾶσιν ὑαῖν τοῖς τὴν ἰατρικὴν αετιοῦσι
[^1] 71 Merc.
There is currently no way to indicate a wider or more narrow span than the line. (But see my BNJ example!!!)
Using different sets of annotation, for example, critical apparatus combined with commentary notes and marginal notes, requires a way to disambiguate, i.e. to mark each system as such. Use simple markers to do so, e.g. letters, like so:
1.[^1] A line[^2] of base text[^3]
[^1] MN A marginal note
[^2] AC A text-critical note
[^3] CN A commentary note
It does not matter which markers are used, as long as they are clear and used consistently.
If we use BPT to digitize print editions, we may be faced with the need to deal with what TEI calls “milestone elements”, i.e. markers of the non-logical structure of a text, such as page numbers of a particular edition.
Leiden+ can help us here: it has markers for blocks of text and divisions like recto/verso (<= and <=.r respectively). (Note that the milestone elements are all closed, whereas the Leiden+ markers open and close). This system can be combined with that of TEI to form the following:
| TEI | explanation | BPT |
|---|---|---|
<milestone> |
a boundary point | <=.ms pindakaas => |
<gb> |
gathering beginning | <=.gb pindakaas => |
<pb> |
page beginning | <=.pb pindakaas => |
<lb> |
line beginning | <=.lb pindakaas => |
<cb> |
column beginning | <=.cb pindakaas => |
The space that is now occupied by the non-sensical term pindakaas can be used to specify the editon or give additional information.
Note, by the way, that the marker for line beginning is superfluous as we already have 1. first line.
1. first<=.pb Migne 456 => line
2. second line
3. third<=.ms GNO page 123 => line
4. fourth line
BPT regards a critical apparatus as a form of annotation, in this case, on the base text, and therefore uses notes to represent text-critical remarks. In particular, stand-off notes are used, that is, the notes are not included in the base text, but separated from it. The separation consists of one blank line.
1. line one of base text[^1]
2. line two of base text[^2]
[^1] first text-critical remark
[^2] second text-critical remark
If the base text is provided by the author in parts, the BPT script will join them. (At least untill the time that DTS and CapiTainS allow for works that don’t consist of a single file). If the critical apparatus is also provided in parts, the BPT script will join them.
BPT uses the vanilla MarkDown system of note numbers. The point in the base text is marked with [^1], for example. This same marker is found at the beginning of the text-critical remark, followed by a space and the body of the remark, like so: [^1] text-critical remark.
BPT requires notes to be uniquely identified per base text.
Te example above shows an annotation marker for a point in the base text. How to mark a span? By the addition of an extra marker that indicates the beginning of the span, like so: ^span of text[^1].` This is a BPT extension of vanilla MarkDown. (Overlapping spans are not allowed).
Note identifiers need not be numbers. They can be anything, except spaces, tabs, or newlines.
How to deal with multiple apparatuses? By adding an identifier for each apparatus to its note numbers.
So, if a text has, for example, both text-critical and historical annotations, i.e. a critical apparatus and “ordinary” notes, then these could be labelled A and B respectively. The first annotation of the apparatus criticus would be marked A1, etc.
1. line one of base text[^A1]
2. line two of ^base text[^A2]
3. line ^three[^B1] of base text[^A3]
4. line four of base text[^B2]
[^A1] first text-critical remark
[^A2] second text-critical remark
[^A3] third text-critical remark
[^B1] first historical remark
[^B2] second historical remark
What does the body of a text-critical remark look like?
BPT expects a scholarly edition of a text to record some or all of the known variations among different witnesses to the text. The critical apparatus contains information about these variant readings of the text in highly structured form.
Each note groups together all readings, i.e. all variations for a word or phrase in the text, per witness.
An example: the first three lines of the Prologue to Chaucer’s Wife of Bath. There are four manuscripts for this text, and for three of them a variant reading of the first line may be recorded like this:
1. Experience, though noon auctoritee[^1]
2. Were in this world, is right ynogh for me
3. To speke of wo that is in mariage;
[^1] El Experience though noon Auctoritee; La Experiment thouh noon Auctoritee; Ra2 Eryment though none auctorite.
The BPT script places the manuscript marker before the variant reading (separated by a space). Any markup (e.g., italics) is removed. Individual variations are separated by a semi-colon.
A note, i.e. an apparatus entry, may contain a lemma, i.e. a reading accepted as that of the original or of the base text. This may look as follows:
1. Experience[^1], though[^2] noon auctoritee
2. Were in this world, is right ynogh for me
3. To speke of wo that is in mariage;
[^1] El Hg **Experience**; La Experiment; Ra2 Eryment
[^2] El Ra **though**; La thouh
The lemma is marked in bold; this can be done by the BPT script because the word or phrase appears exactly like that in the base text.
The above requires a unique identifier for each witness. BPT expects scholarly editions to provide a list of witness and their markers (sometimes known as ‘sigla’).
The BPT script places the list of witnesses between base text and critical apparatus. It may look as follows:
El = Ellesmere, Huntingdon Library 26.C.9
Hg = Hengwrt, National Library of Wales, Aberystwyth, Peniarth 392D
La = British Library Lansdowne 851
Ra2 = Bodleian Library Rawlinson Poetic 149
<=.UNIT NUMBER=>. for example, in ETTO, we use:• Physical structure: Van der Valk: tome <=.vt [tome number] =>; page <=.vp [page number] =>; line <=.vl [page number] =>. These are “closed “elements; place them at the start of the text they structure.
• Physical structure: Rome: tome <=.rt [tome number] =>; page <=.rp [page number] =>; line <=.rl [line number] =>. These are “closed “elements; place them at the start of the text they structure. The markers for the Rome line numbers replace the vertical lines (pipe symbols) that now occur in the base text at every five Rome lines.
<div>, <ab>) and physical (<milestone>, <pb>, <lb>) is not strictly enforced. We really should do this.