| Menota main page List of contents |
 |
|
|
Chapter 7. Altered, corrected and unreadable text7.1 Introduction Version 2.0 (16 May 2008). Links updated 12 July 2016. 7.1 IntroductionThis chapter deals with the encoding of additions, deletions and corrections made in the manuscript by the scribe or later users, or similar changes made in the transcription, e.g. by the transcriber or encoder of the manuscript text. Further, the chapter deals with the encoding of damage to the manuscript that affects the reading of the manuscript text. In ch. 7.2 corrections, deletions and additions made by the scribe or later users of the manuscript are treated. In ch. 7.3 damage to the manuscript that affects the reading of the manuscript text is treated. Ch. 7.4 treats corrections, deletions and additions made by the transcriber of the manuscript text that have been made e.g. from other text witnesses or earlier editions of the text. Ch. 7.5 contains a summary of all elements and attributes discussed in this chapter. 7.1.1 StructureBecause the features described in this chapter are largely non-linguistic, they have more relevance to the textual levels (encoded by the elements facs, dipl and norm) than to the linguistic encoding of texts, parts of texts, words and so on. In many cases, therefore, the encoding of these features will form an incompatible hierarchy with the structure of the text. In such cases, and in cases where it occurs within a word, the encoder will have to encode the features at a particular textual level. These features will, in general, be marked up at the facs level when they correspond to the physical manuscript. Changes to the text made by the editor/transcriber, on the other hand, should be encoded at the dipl level where necessary. The norm level contains the final, corrected text without non-linguistic markup. The following table indicates at what level each feature should be marked up, and/or its contents included, where relevant:
Here 'Element' also refers to the value of the 'category' attribute when the <me:textSpan/> element is used (see ch. 7.5). 7.1.2 ElementsThe encoding recommended here is based on ch. 11 “Representation of Primary Sources” of the TEI P5 Guidelines, where the following elements are defined:
7.1.3 Philological introductionIn a discussion on editorial practice for Old Norse texts Helle Jensen, with reference to Stefán Karlsson 1963, LXVII f., outlines aspects of the manuscript text which should be noted in an edition (Jensen 1988). Jensen's suggestions start with structural markup of e.g. linebreaks in the manuscript. She also gives special signs for each of the features that has to do with scribal or later changes in the manuscript as follows (Jensen 1988, 102 f.):
In addition, Helle Jensen suggests that uncertain readings should be subpunctuated. In editions from the Arnamagnæan Institutes in Reykjavík and Copenhagen these suggestions are in general followed, and in most editions of medieval Scandinavian texts similar systems are used. This gives us a starting point when we are transcribing Old Icelandic and Old Scandinavian manuscripts. The principles presented in this handbook are based on the tradition of producing scholarly editions of texts and individual manuscripts. The system for printed editions outlined by Helle Jensen can therefore very often be translated into the electronic markup language presented in this chapter. Text written in the margins can be of various kinds and of varying interest for our knowledge about the main text and the history of the manuscript. Notes on the main text in the margins are of course valuable when we are interested in the text tradition. Other notes could indicate that someone at a certain stage has used it for example in a transcription of the text. In medieval manuscripts, however, we often also find notes in the margins that have nothing whatsoever to do with the manuscript text. These notes can at first sight seem to be of no value to philological investigation, but in a larger context they can sometimes give information as to where a manuscript has been at a certain stage of its history. If e.g. the same type of scribbles are found in a group of manuscripts where one of the manuscripts can be geographically pin-pointed, this could indicate the whereabouts of the whole group. Information of this kind can also lead to the establishing of new connections between manuscripts that were not previously seen as connected. There are thus good arguments for including information also on this kind of marginal note, but these are more properly contained in the manuscript description in the header (cf. ch. 10.2.2) than within the encoded transcription. The first kind of notes, i.e. comments or additions to the main text, are often treated in foot-notes in printed editions. They are considered relevant to the reading of the text, and are therefore given in relation to the main text. Marginal notes that indicate the owner or user of the manuscript in any obvious way are often treated in the introduction to the edition as they are considered relevant to the history of the text or manuscript. The third category of notes, the ones that do not seem to give any relevant information, is often excluded or treated only briefly in the introduction. This is of course a rational way to handle these scribbles when the printed edition sets the limits, and the information often is obscure and cannot be easily related to parallel information concerning other manuscripts. In the electronic transcription of a manuscript, however, there is no reason to make this limitation. The information can be given in the same way as for the other categories, and thereby give us the possibility to search for all kinds of obscure information. Medieval manuscripts have often become damaged through use, sometimes with relevance for our reading of the text. Pieces of parchment may for example have been torn out, leaving a physical gap in the manuscript. Parts of the text may be illegible because of use or deliberate erasure, or they may be darkened to such an extent that the text is no longer readable. In printed editions, unreadable sections of a text are marked as suggested by Helle Jensen. In the introductions to printed editions problems related to illegible text and damage to the manuscript are often discussed at length. If there are other text witnesses these are often used to replace missing stretches of text. In a diplomatic transcription of a manuscript text, however, the missing or unreadable parts are most often just marked as such. In the following sections the relation between the traditional markup of these kinds of textual and editorial difficulties and electronic encoding will be obvious. It is therefore relevant to take traditional transcription and editing as a starting point for the electronic encoding of transcriptions of manuscript texts. The primary aim of the following sections are to give recommendations for the transcription and encoding of manuscript texts. It does, however, in some instances also give recommendations for editorial encoding, e.g. markup that refers to corrections or additions made by the transcriber or encoder. It is therefore important to keep the transcription and encoding of the manuscript text on the one hand and on the other hand the editorial changes consistently separated, so that the former provides a starting point for the editorial work. 7.2 Scribal features: Additions, deletions and substitutionsIn the manuscript text and in the margins of the manuscript we often find different kinds of corrections, deletions and additions that we want to encode. These changes can be divided into different groups depending on the nature of the change and its relevance for the reading of the manuscript text or our knowledge about the manuscript. The main division is between additions or substitutions to the manuscript text, within the text or in the margins, and deletions made in the manuscript text. The former should be marked with the <add> element while the latter should be marked with the <del> element. Additions and substitutions made by the transcriber or editor are treated in the last section (ch. 7.4). 7.2.1 AdditionsThis section deals with additions made by a scribal hand only. It is a common mistake to use elements designed for this purpose to mark up additions made by an editor - such features are covered below in ch. 7.4. The following elements are recommended for describing additions made by the author of the text, a compiler, scribe, annotator or corrector in the manuscript text. The TEI P5 Guidelines recommend the use of the <add> element to describe additions in the manuscript (ch. 11.3). In the following the use of <add> in relation to our recommended encoding of the individual word within the element <w> and on the three different levels <me:facs>, <me:dipl> and <me:norm> is treated.
Additions which can be ascribed to the author of a text are rare in medieval Nordic manuscripts. The additions being described with the above-mentioned attribute @hand will therefore primarily be ascribed to the values 'scribe' , 'compiler' , 'annotator' or 'corrector' . Scribal additions are probably the most common changes to be recorded in the transcription and encoding of a manuscript text. The list of hands in the header (cf. ch. 10) should identify the individual hand, either as anonymous or, if possible, by name. The main hand in a manuscript will normally be marked as 'mainscribe' . If the addition consists of a series of complete words, the <add> tag should be surrounding the word(s). The following example contains a marginal addition:
Fig. 7.1. AM 748 I b 4to, fol. 18r, ll. 5-7 Here (AM 748 I b 4to, fol. 18r, ll. 5-7) the scribe (identified as 'mainscribe' in the header) has inserted the word 'sigavtr' in the left margin (now partially cut off by the binding) with a mark after 'skollvalldr'. It should be encoded as follows. Because the addition is a whole word, the tag should enclose the word (Note that for the sake of clarity we have limited the use of encoding to the relevant sequence and simplified the orthography): Note that to ensure correct rendering of the addition, no space is included between the <add> and <w> elements. In cases where the addition forms part of a word, the markup should normally be restricted to the facsimile level. In any case, the addition need not be marked up as part of the normalised text. Cf. the following:
Fig. 7.2. AM 748 I b 4to, fol. 1r, l. 15 The location of the addition in the above markup is indicated by the attribute @place; in this case, the addition is made above the line of the manuscript text and therefore uses the value 'supralinear' . The diplomatic and normalised text will normally include the addition if made by a scribal hand. Additions made by later hands will normally be omitted from the diplomatic and normalised text without markup. Additions are sometimes made by an annotator, i.e. comments to the text. This kind of additions could be encoded as the marginal note “vantar ekkert F. J.” by Finnur Jónsson in Codex Wormianus (AM 242 fol. p. 60): <add hand="FJ"> <w><me:facs>vantar</me:facs></w> <w><me:facs>ekkert</me:facs></w> <w><me:facs>F˙</me:facs></w> <w><me:facs>J˙</me:facs></w> </add> It is also possible to indicate with the attribute @place where on the manuscript page the annotation is made. Finnur Jónsson's annotation is made in the bottom margin, and should be encoded as follows: <add hand="FJ" place="bottom"> <w><me:facs>vantar</me:facs></w> <w><me:facs>ekkert</me:facs></w> <w><me:facs>F˙</me:facs></w> <w><me:facs>J˙</me:facs></w> </add> Changes in scribal hands are not considered additions. For the markup of such phenomena, use the empty <handShift/> element. Note that this element is parallel to empty elements like <pb/>, <cb/> and <lb/> in that it only points to a break in the text (cf. the discussion in ch. 4.7 and ch. 4.10 above). See §7.5 below for examples of how to mark up additions which span linguistic (<w>, etc.) boundaries. 7.2.2 DeletionsThis section deals with deletions made by a scribal hand only. Text supressed by the editor is dealt with under 'corrections' (ch. 7.4.2) below. The TEI P5 Guidelines specify the use of the <del> element to describe additions in the manuscript (ch. 11.3). In the following the use of <del> in relation to our recommended encoding of the individual word within the element <w> and on the three different levels <me:facs>, <me:dipl> and <me:norm> is treated.
Deletions that can be ascribed to the author of a manuscript text are rare in medieval Nordic manuscripts. The deletions being described with the above mentioned attribute @hand will therefore primarily be ascribed to 'scribe' or 'corrector' . Deletions should normally only be marked up at the 'facs' level. The text should be removed without indication at the other textual levels. The example in ch. 7.2.1 (Figure 7.1) contains the word 'skolldvalldr' in which the first 'd' is marked with a dot below signifying deletion. It should be marked up as follows: Deletions of one or more words made by the scribe(s) or corrector(s) of a manuscript are encoded as in the passage from the Third Grammatical Treatise below. Note that we for clarity limit the use of encoding to the relevant manuscript line, and that only the <me:facs> and <me:dipl> levels are shown, and a few of the words have been suppressed.
Fig. 7.3. AM 242 fol., p. 94, ll. 11-13 In the TEI P5 Guidelines cited above there are a number of possible types of deletion described with the attribute @type. These could be applied to deletions made both by scribe(s) and corrector(s). If a deletion is made e.g. by 'overstriking' the deleted text it could be encoded as (here only presented on the <me:facs> level): This could then be displayed on the computer screen or in a printed edition in the manner suggested above (ch. 7.1): 14 ...en tuenner flokkar þeirar þioðar er 15 |-liguri hæita er-| traceum hæiter... The text that is marked as deleted must be at least partly legible in the manuscript so that it can be read by the transcriber. If the deleted text is not legible the deletion should be marked up with the <gap/> element, described below (7.3.1). The <gap/> element could be enclosed in the <del> element to indicate that the gap is in some way intentional. Parts of the deleted text that are legible could be indicated by the <unclear> element in combination with the <gap/> element as described below (ch. 7.3.2). As for the markup of additions using <add>, if the deletion is of part of a word, it should normally only be marked up at the <me:facs> level. If the deletion is by a scribal hand, the deleted text will be omitted from the <me:dipl> and <me:norm> levels without markup. In some cases, the deletion will conflict with word boundaries. ch. 7.5 below describes how to mark up such deletions. 7.2.3 SubstitutionsThis section describes the markup of text substituted by a scribe, that is, where a scribe deletes text and replaces it with some other text. In medieval manuscripts a rather common phenomenon is the combination of deleted text and added text. It is not always possible, however, to ascertain the relation between the two. If someone has deleted the originally written text inline this does not automatically mean that a corresponding addition above the line or in the margin is made by the same scribe. It can therefore not be stated as certain whether the correspondence is intentional or not. There is no specific element for this type of feature. We suggest that substitutions made in the manuscript should be marked primarily with the two core tags <del> and <add>. In cases where we can be relatively sure about the agent of the whole substitution this could be indicated with a combination of the <del> and the <add> elements as illustrated below.
Fig. 7.4. AM 748 I b 4to, fol. 13r, l. 25 Here, the seventh word 'barv' has been altered by the scribe, such that the original 'a' at the end of the word is marked as deleted by subpunctuation and a 'u' has been added as the replacement letter above the line. This word should be marked up as follows (facs and dipl levels only): As for the <del> and <add> elements, markup within words should only be included at the facs level, and the dipl and norm levels should include the text which the scribe(s) appears to have intended, as in the example above. The same type of markup can be used for substitutions which span structural boundaries. This type of markup is discussed in more detail in ch. 7.5. 7.3 Damage and illegibilityThe following section deals with text omitted in the transcription or editing of text due to damage or illegibility in the manuscript, and text supplied from other sources such as other text witnesses or earlier editions. The following detail of a manuscript page will serve to illustrate the markup in the following sections.
Fig. 7.5. AM 757 a 4to, fol. 2r, ll. 22-25 7.3.1 Text omitted from the transcriptionWhen the manuscript is illegible we suggest the use of the elements <gap/> and <supplied> to indicate the illegible text, its extension and how it has been supplied (for the <supplied> element see ch. 7.4.1). The <space/> element is used to represent deliberate omissions from the manuscript which have some significance, e.g. spaces left for decorated initials or words.
In medieval manuscripts we often find sections that for some reason are illegible. This can be due to e.g. damage or use. In the transcription we primarily wish to register the sections that are illegible and the extent of the illegibility. We suggest that the illegible sections should be indicated by the <gap/> element. The extent of the illegible section could be encoded as the following text from the manuscript above, where wearing has occurred next to a hole (bottom left of image): With this markup the extent of the illegible section is not defined. It can be presented on the computer screen or in a printed edition in the manner suggested above (ch. 7.1): malsg(re)in(ar) 00...00 nafn If the transcriber or encoder of the text wishes to define the section more accurately it can be done as in the following example. The number of missing signs is given as a value to the attribute @quantity. It should be noted that the number given in the example is not intended as an exact evaluation of the number of signs missing in the present manuscript. This could be represented as follows on the computer screen or in a printed edition. As the accuracy of this kind of evaluation is questionable it should not have the highest priority to display this in e.g. a printed edition. malsg(re)in(ar) 00 nafn See below (ch. 7.4.1) for an example of text supplied in place of the gap by the editor using the <supplied> element. A deliberate space in a manuscript should be indicated by the <space/> element. For example, in the image above, an initial letter has been omitted, presumably for a coloured or decorated initial to be added later. This is encoded as follows: <lb/><w> <!-- initial 'S' is left blank by the scribe here --> <me:facs><space quantity="1" dim="horizontal"/>augn</me:facs> </w> In this example, the text has only been encoded on the facs level, so the <choice> element would be redundant. If the text omitted in the space is supplied using the <supplied> element, the <space/> tag will normally be used at the facs level and the <supplied> element will be used at the dipl level (with the @reason="space" attribute set). See ch. 7.4.1. 7.3.2 Uncertain readings in the manuscriptIn medieval manuscripts we often encounter problems of illegibility due to use or damage. In the following the encoding of such sequences is treated. To some extent this has already been treated in the above section (ch. 7.3.1). In cases where the text is readable to some extent the <gap/> and <supplied> elements should not be used. The TEI P5 Guidelines (ch. 11.5 of the TEI P5 Guidelines) recommend that the <unclear> element is used for encoding damage and illegibility where the text of the damaged or illegible area can be read with some, but not full, certainty.
The example given above from AM 757 a 4to includes a section on the second line which can only be partially read. The following is the encoding for the section, only represented at the facs level: <w> <me:facs>mal&slong;</me:facs> </w>. <w> <me:facs>h<am>&osup;</am></me:facs> </w> <w> <me:facs><gap extent="2"/></me:facs> </w> <unclear reason="worn" resp="TW"> <w> <me:facs>k&ocurl;llut</me:facs> </w> <w> <me:facs>ein<am>&bar;</am></me:facs> </w> <w> <me:facs>hlutr</me:facs> </w> </unclear> <w> <me:facs>af</me:facs> </w> The text may also include fully illegible text, as represented above with the <gap/> element. With this encoding the text could be presented with subpunction or as grey text for all the words that the editor can not read with absolute certainty. 7.4 Editorial interventionsWhen transcribing medieval material we often encounter words or longer sequences of text that we consider corrupt in one way or another. Sometimes it may also be obvious that text is missing in the manuscript we are transcribing or that the scribe has made a mistake. The transcriber of the manuscript text may in these instances wish to indicate the mistake or even correct the text, either directly from other versions of the same text or based on already existing editions of the text. Sometimes the transcriber or editor may also wish to make obvious grammatical corrections in the text without having any other text witness or precedence in an earlier edition. In the following the encoding of corrections made by the transcriber of the text or by an editor are treated. Note that we do not recommend the use of the attribute @hand for the changes made in transcription or encoding of the manuscript text. The attribute @resp should be used consistently for corrections or additions made in the transcription or encoding of the text to distinguish clearly between what is found in the manuscript text and what is made in the transcription and encoding of the text. 7.4.1 Additions made by the transcriber or editorIf text is obviously missing in the manuscript text we may wish to supply it. This could be based on, for example, another text witness or on a earlier edition of the text. The markup of such additions should give information about the source as well as about the responsibility for the addition. To encode additions made in the transcription we recommend the use of the <supplied> element as described in the TEI P5 Guidelines (ch. 11.3.7). In the following the use of <supplied> in relation to our recommended encoding of the individual word within the element <w> and on the three different levels <me:facs>, <me:dipl> and <me:norm> is treated.
If the transcriber or editor wishes to supply text that is missing in the transcribed manuscript text from for example another text witness, this can be handled with the <supplied> element. The interpolated text could be transcribed as in this instance from the detail of AM 757 a 4to above. Note that we for clarity limit the use of encoding to the relevant sequence: This could then be displayed on the computer screen or in a printed edition in the manner suggested above (ch. 7.1): malsg(re)n(ar) <þat> nafn In the above example there are other sources available for the illegible text. This kind of editorial change is, however, not suggested as compulsory. In a primary transcription and encoding the use of the <gap/> element should only give the essential manuscript information. The attributes to <gap/> and <supplied>, such as @source or @resp, can of course be included voluntarily and to the extent that information is available. 7.4.2 Deletions made by the transcriber or editorIf a piece of text obviously should be deleted, e.g. duplicated text in a dittography, the transcriber or editor might want to make a deletion. This is the converse action of adding text, and should be distinguished from similar actions made by the scribe. While the elements <add> and <del> describe actions by the scribe himself or other scribes, the editorial additions and deletions should be singled out by separate elements. For additions, TEI recommends the element <supplied>, but there is no parallel to the <del> element. We suggest the element <me:expunged>, since the noun expunction and the verb expunge are commonly used for editorial deletion.
Note that expunged text will not be deleted from the transcription, but will be contained by the <me:expunged> element. The editor of the text may decide to dsiplay it with no comments, put it in brackets or suppress it. In a multi-level transcription, expunction will typically not be found on the <me:dipl> and <me:norm> levels, but not on the <me:facs> level.
Fig. 7.6. AM 233 a fol, fol. 28vB, l. 39-40 In this example from Niðrstigningar saga, AM 233 a fol, the word “menn” has been supplied and thereafter the duplicated word “mæla” has been expunged: Depending on the style sheet, the diplomatic text might be displayed this way: þuiat þeir villdu ecki vid <menn> mæla -|mæla|- 7.4.3 CorrectionsIn the manuscript it is not always possible to say anything with certainty about the intention of changes in the text. When transcribing the text, however, corrections of obvious mistakes in the manuscript text could be marked with the following tag set recommended in the TEI P5 Guidelines (ch. 11.3). In the following the use of <sic> and <corr> in relation to our recommended encoding of the individual word within the element <w> and on the three different levels <me:facs>, <me:dipl> and <me:norm> is treated.
In a first-level transcription it can be relevant just to mark the obviously corrupted instances in the manuscript text. This could be done with the <sic> element as in this instance from AM 242 fol:
Fig. 7.7. AM 242 fol, p. 98, ll. 3-4 Here the second numeral '.xij.' is written instead of '.ix.', obvious from the context and other manuscripts. The error should be signalled at the facs level and corrected at the dipl level: With this markup it is possible to show the text on the computer screen or in a printed edition in accordance with the suggestions above (ch. 7.1): en(n) .viij. e(ða) *.ix. with the corrected form from the manuscript text underneath the edited text: * .xij. It is also possible to include information about the person responsible for the correction with the attribute @resp and its values. In this example of a multi-level transcription, the <sic> element lies on the facs level, while the <corr> element is introduced on the dipl level, and perhaps silently on the norm level. In a single-level transcription (cf. ch. 3.3), the <choice> element should be used to group the <sic> and <corr> elements. If, for example, a source has the reading Please look left now! in which “left” should be corrected to “right”, this would be the appropriate encoding: Please look <choice><sic>left</sic><corr>right</corr></choice> now! In a single-level transcription, the <choice> element thus groups <sic> and <corr> elements, while in a multi-level transcription it groups readings on different levels, i.e. the <me:facs>, <me:dipl> and <me:norm> elements. The <choice> element is simply a neutral mechanism to group alternative readings. 7.5 Reference: elements and attributesBecause of the large number of elements and attributes introduced in this chapter, the following reference material is supplied. 7.5.1 ElementsThe following is a list of elements to encode non-linguistic features. In the table, 'level' indicates the textual level at which the feature should be encoded, if it occurs within a word. No such features should be encoded at the 'norm' level, as this level should not include non-linguistic aspects of the text. Some elements are included which are not treated in this chapter. Encoders should refer to the TEI P5 guidelines for more information.
7.5.2 Descriptive attributesThe following attributes have similar meanings, regardless of which element they are used with. |
First published 20 May 2003. Last updated 12 July 2016. Webmaster. |
