Edition model

Version: 1.0
Publication: 31.07.2023

Introduction

Some of the principles that are aimed at when designing the edition model are described below. The objective is to ensure that the specific rules relating to text constitution and referencing are as compatible as possible with these principles. Therefore, the specific rules are continuously revised and refined.

Edition model alignment

The edition model should ensure the greatest possible homogeneity of the editions, but at the same time be flexible and easy to use. This ensures appropriate philological processing of the texts and at the same time suitability for use by a wider audience (students, guest editors, cooperation partners).

There is a certain amount of flexibility in dealing with text-specific peculiarities (e.g. irregular punctuation, irregular placement of diacritical marks and apostrophes, etc.), which can be defined in the introduction.

Relation between encoding and representation

TEI-XML is used as the standard for encoding. This ensures long-term connectivity and interoperability in line with the FAIR Data principles.

The encoding consistently distinguishes between content and presentation (rendering). This allows the creation of different forms of output and presentation. Mixing semantic and formal forms of expression should be avoided. For example, such mixing occurs when text-critical phenomena are expressed solely through rendering.

In the case of complex text-critical phenomena that are not expressible with the existing coding options, labelling is avoided and the phenomenon is instead described semantically using a text-critical annotation. This approach also works for editions that are not created with the Oxygen XML Editor (e.g. editions in Word). In the following explanations of the edition model, it is indicated where either an encoding or a semantic description is applicable.

Relation between text criticism and referencing

Text sections that refer to entities (persons, places, institutions, publications, etc.) are not normalised in text-critical form. Normalisation is provided by the name variants in the referenced entity. In letter_00068, for example, Allioni wrote "Nismes". The spelling "Nîmes", which is commonly used today, is evident in the reference to the town. In addition, entering "Nîmes" in the search bar leads to the corresponding letter.

Text structure

The texts in all editions are divided into individual units (e.g. letters, manuscripts, assemblies, etc.) and into pages and paragraphs, a process that can be realised without any knowledge of encoding. The latter also applies to line breaks, although their implementation is optional.

In addition, it is possible to give texts a specific structure depending on the entity group (e.g. letters), which is particularly important for extended types of presentation and analysis. Furthermore, there are certain structural elements such as tables or lists that can be implemented independently of the entity group. Both these structuring approaches generally require TEI-XML markup.

Letter structure

The structure of a letter comprises various elements that are usually assigned according to semantic criteria. Spatial deviations are accepted. The ideal structure is as follows:

• Letter divided into page(s) | <pb/>
• Opening | <opener/>
  • Salutation | <salute/>
• Text body
  • Divided into paragraphs | <p/>
  • Divided into lines | <lb/>
• Closure | <closer/>
  • Salutation | <salute/>
  • Signature | <signed/> [Namen/Unterschrift/Spezifikation o.Ä. i.d.R. bis zur dateline]
  • Date line | <dateline/> [Datum/Ort]
• Postscript | <postscript/>
• Adress field | <closer/>
  • Address | <address/>
  • Lines in the address | <addrLine/>

Possible deviations from the ideal structure are

• A date line or signature in the opening
• No salutation in the opening
• No salutation, signature or date line in the address field
• No postscript, no address field

Special case: salute and signed

The distinction between the elements <salute> and <signed> is not clearly defined in the TEI guidelines. Generally, salutations are labelled with <salute> and not with <signed>.

Example: TEI-XML encoding of a closure with salutation

<closer>
    <salute>I remain<lb/>your most devoted servant.</salute>
    <signed>Munchhausen</signed>
    <signed>Allioni le fils</signed>
</closer>

General structural elements

Tables and lists

Tabular structures and lists are generally represented in a functional, simplified layout according to the editor's interpretation.

Tabular structures and lists are generally not modelled in the same way as the original, but are represented in a functional, simplified layout according to the editor's interpretation. Structures can either be imitated using paragraphs or coded as a table or list using TEI-XML. Indents and indentations are not replicated.

Tables are enclosed with the element <table> and lists with <list>. Entries in lists are marked with <item>. <table> and <list> are to be placed as block elements next to <p> (but not in <p>).

Example: TEI-XML encoding of a table (cf. letter_00287)

<table>
    <row>
        <cell/>
        <cell>44. Unzen, alſ,</cell>
    </row>
    <row>
        <cell>1. Mezen</cell>
        <cell>704. Unzen, und</cell>
    </row>
    <row>
        <cell>1. Sra</cell>
        <cell>11,264. Unzen.</cell>
    </row>
</table>

Example: TEI-XML encoding of a list (cf. letter_00287)

<list>
    <item>32. Mezen, und jede dieſer Mez.</item>
    <item>18. Maas</item>
</list>

Quotations

Quotations from literary texts, letters, reviews, etc. are marked with the element <quote>. Underlining by the authors emphasises are reproduced. In case the origin of a quotation from literary texts, letters, reviews, etc. is known, a free annotation can be added. This annotation may contain a reference to the quoted text or work (possibly only the author) if a database entry is available.

Custodes

Custodes are reproduced in simplified form in accordance with the standard edition model. The corresponding term is only transcribed on the following page (after the page break) for this purpose and is not labelled (cf. letter_03954, letter_03955).

TEI-XML markup of custodes is only conducted by means of extended text constitution.

Dates

Dates are neither annotated nor tagged in the transcribed text. If they refer to the edited entity as a whole, they are included in the structural data of the respective object.

Annotations

The annotation concept specifies the interaction between the edition texts and the structural data. This forms a core element of the data-centred orientation of République des Lettres.

Types of annotation

The République des Lettres platform provides three different types of annotation:

1. Reference annotations
2. Free annotations on content
3. Annotations on text constitution

TEI-XML encoding

1. <rs><note type="annotRef"><p/></note></rs>
2. <note type="annotFree" n="1"><p/></note>
3. <note type="annotText" n="1*"><p/></note>

Numbering of annotations:

Free annotations and annotations on text constitution are numbered automatically. The numbering starts at 1 for each letter and annotation type, whereby annotations for text constitution are additionally marked with an asterisk *. Reference annotations are not numbered. The reference appears on the platform as a hyperlink that opens a modal. Any annotations for the reference are displayed in the modal.

References within annotations:

All three annotation types may contain references to entities in the database. The referenced entities are usually named according to the spelling of their database entry (label), e.g. *preface GGA (1747). The year is mandatory for individual publications, otherwise it is optional, e.g. GdW Göttingen or GGA. Persons are preceded by initials, e.g. G. A. v. Münchhausen.

Text formatting within annotations:

In the annotations, a distinction is made between the editor's text and the original text via a typographical difference between italics and straight font: editor's text always appears in italics (including all punctuation belonging to the editorial annotation), original text appears as straight font.

Reference annotations

Reference annotations describe the relationship between the referenced entity and the original text:

• Identification uncertain: Reference: GdW Göttingen, context info: Possibly GdW Danzig.
• Deviation of the original term from the normative name of the referenced object.
• Information that provides an explanation of the specific situation in the original text beyond the general information in the referenced object. A reference is also suitable as an alternative, e.g. reference: Göttingen, context information: This refers in particular to GdW Göttingen.

Notation in reference annotations:

• Only the name of a referenced person is abbreviated (to the first letter of the surname, e.g.: It is probably T. who was on site at the time). If several people with the same initial letter are mentioned, the surname is fully spelled out. Where several people with the same surname are mentioned, the first names are spelled out.
• Citations follow this scheme: Cf: [page numbers]. (Example: Cf. Sonntag (1983): 983.)

Example: TEI-XML encoding of reference annotation

<rs type="person" key="person_27256" xml:id="hallerNet_letter_00002_ref-1">Johann Scheffield<note n="161" type="annotRef"><p>Probably S.</p></note></rs>

Guidelines for TEI-XML encoding of reference annotations

The <rs> element must contain two attributes:

1. @type. Valid attribute values (or entity types) are person, place, institution, publication, letter, review (plant is deferred).
2. @key. The attribute value corresponds to the ID of the object and is made up of the entity type, an underscore and a five-digit number, e.g. person_01200. Regarding the <note type="annotRef">, please note that

• the <note> is placed before the end tag of the <rs> element;
• no spaces are allowed between the end tag of the <note> and the end tag of the <rs> element;
• the <note> contains exactly one paragraph element <p>.

Free annotation

Free annotations do not refer to a reference, but to a specific passage in the text. They are placed "freely" in the text and can contain comments, refer to further information or explain a particular section of the text. If the free annotation refers to an entire sentence or phrase, it is placed after the punctuation mark, otherwise it is always placed directly next to the word to which it refers.

Example: TEI-XML encoding of a free annotation

Pietra di Dair<note n="28" type="annotFree"><p>Pyrite, sulphur pyrite.</p></note>

Guidelines for TEI-XML encoding of free annotations

Free annotations consist of a <note> with type <annotFree>. When coding, please note that

• the <note> contains exactly one paragraph element <p>;
• the <note> must not be preceded by a space.

Annotation on text constitution

Annotations on text constitution are used for the semantic description of text-critical phenomena. If the text-critical annotation refers to an entire sentence or phrase, it is placed after the punctuation mark, otherwise it is always placed directly next to the word to which it refers. The requirements for semantic description are specified in the text constitution section.

Referencing

A distinction is made between text-based and entity-based referencing. In text-based referencing, a specific source term is provided directly with a reference to a person, a place, a publication, etc. (cf. "Hl. Professor König" in letter_00002). In contrast, with entity-based referencing, the related entity appears in the structural data (cf. "*Artis med. principes (1769-74)" in review_03270).

The following refers to text-based referencing.

In all document types, explicit mentions of entities are only labelled once per unit (letter, supplement, introduction, review, page, etc.), with only the first mention being referenced. An exception is made for letters with supplements, where double labelling of the same entity is possible. For letters, no entities are labelled in the opening and address field and the sender, addressee, senders and address place are stored in the structural data.

Generally, nested labelling should be avoided, which is why entities that are related to each other and occur directly next to each other in the text are subsumed under one reference as far as reasonable (e.g. if the wording/formulation is to be given more weight in case of doubt):

Guideline	Example
Subsume place within a person	my friend in Göttingen
Subsume place within an institution	University of Göttingen
Subsume two places within one	Bümpliz near Bern
Subsume person within a publication	Hallers Flora
Subsume person within a letter	Letter from Haller to Caldani
Subsume two persons within one	Haller's mother
Subsume person and publication within a review	her review of my work in the GGA

In case of ambivalences, e.g. institution vs. place (X studied in Göttingen) or publisher vs. person (printed by Vandenhoeck), the literal citation is referenced. If in doubt, a description can be added in a reference annotation.

If an entity cannot be identified or named, it can nonetheless be provided with a reference. The reference contains the information that it is an unknown entity of an entity group (persons, institutions, etc.).

Example: TEI-XML encoding of unknown entities

<rs type="person" key="person_unknownReference">Madame Ougsburger</rs>
<rs type="publication" key="publication_unknownReference">ce livre</rs>
<rs type="institution" key="institution_unknownReference">Landes Regierung</rs>
<rs type="place" key="place_unknownReference">montagnes de Vinay</rs>

Depending on the entity group, there are different rules for how unknown entities are referenced, as described below.

Persons mentioned

Non-identified persons are generally only added as a new object if at least the surname is available. Otherwise, they are marked as <person_unknownReference>. As special cases, unnamed persons who can be identified as individuals in their function are also included: e.g. Maid of the family Tscharner (see Conventions on structural data).

Persons are either mentioned by name (e.g. Albrecht von Haller) or by indirect reference (e.g. my mother, our friend, the king). Both are marked in the text, with the following exceptions:

• The sender and recipient of the letter are not labelled in the transcription, but instead stored in the structural data.
• In the introduction of a letter edition, the main correspondents (e.g. Haller and Münchhausen in Sonntag 2018) are not marked.

If a reference relates to more than one person (e.g. the Fischers, my two sons), the references are listed separated by spaces, but not more than two persons. Larger groups (especially if indistinct) are not distinguished.

Example: TEI-XML encoding of persons

<!--Proper name, first name, surname and predicate-->
<p>Famous <rs type="person" key="person_01200">Albrecht von Haller</rs> sent me a book.</p>
    
<!--Proper name, title and surname-->
<p>Famous <rs type="person" key="person_01200">Dr Haller</rs> sent me a book.</p> 

<!--Proper name, appellation and surname-->
<p>Famous <rs type="person" key="person_01200">Mr Haller</rs> sent me a book.</p> 

<!--Proper name, only surname-->
<p>Famous <rs type="person" key="person_01200">Haller</rs> sent me a book.</p>
    
<!--Proper name, abbreviated first name, family relation-->
<p><rs type="person" key="person_09090">Haller's sister-in-law Rosi</rs> sent me a book.</p>
    
<!--Alias/pseudonym unrelated to the proper name-->
<p>Our friend <rs type="person" key="person_00703">Maecenas</rs> wrote to me.</p> 

<!--Proper names used as adjectives--> 
<p>c’êst reprocher à l’Ecole <rs type="person" key="person_10394">d’Aristote</rs> de n’avoir pas embrassé la Gravitation <rs type="person" key="person_04708">Neütonniéne</rs></p> 

<!--Indirect denomination-->
<p>The <rs type="person" key="person_01200">author of the Alps</rs> sent me a book.</p>
    
<!--Indirect denomination including place names-->
<p><rs type="person" key="person_01200">Our friend in Göttingen</rs> sent me a book.</p>
    
<!--But: Place does not belong to the person in this example-->
<p><rs type="person" key="person_01200">Our friend</rs> is comfortable in <rs type="place" key="place_00557">Göttingen</rs>.</p>
    
<!--Indirect denomination, only official title-->
<p><rs type="person" key="person_02971">The king</rs> sent me a book.</p>
    
<!--But: proper name, including official title-->
<p>Our King <rs type="person" key="person_02971">George II of England</rs> sent me a book.</p>
 
<!--Indirect denomination of multiple persons-->
<p>The <rs type="person" key="person_01200 person_02971">Hallers</rs> sent me a book.</p>

Places mentioned

Districts are referred to the municipality, e.g. Matte-Quartier is referenced to the city of Bern and included as a name variant (part of). The same applies to villages such as Bümpliz, which today belongs to Bern but was independent until 1919.

When a document mentions Bümpliz and Bern, only the first mention of the current entity is recognised. However, a free annotation with a corresponding note can be added for each subsequent mention of formerly independent municipalities.

Example: TEI-XML of free annotations on places

Nowadays, Bümpliz is part of the municipality of <rs type="place" key="place_ 00309">Bern</rs>

Nouns (e.g. Germans) or adjectives (e.g. german) are also marked if a geographical territory is meant.

Example: TEI-XML encoding of places

<!--Place name-->
<p>The article was written and printed in <rs type="place" key="place_00557">Göttingen</rs>.</p> 

<!--Place as adjective-->
<p><rs type="place" key="place_00557">Göttingen</rs> scholars are fame-seeking.</p> 

<!--But: Adjective describes the language in this example-->
<p>I often read german texts.</p> 

<!--Localisation, place name including description of the location-->
<p>The letter was sent from <rs type="place" key="place_99999">Straussfurt, in Saxony</rs>.</p> 

<p>The letter was sent from <rs type="place" key="place_99999">the largest town in Saxony</rs>.</p>

Institutions mentioned

The types of institutions are listed in the Conventions on structural data.

Non-specific mentions are not recorded (e.g. the Bernese government). Places that belong to an institution name (e.g. Gesellschaft der Wissenschaften zu Göttingen) are not marked.

Example: TEI-XML encoding of institutions

<!--Institution name, including location-->
<p>The <rs type="institution" key="institution_00295">Royal Society of London</rs> was founded in 1660.</p>

<!--But: The place is not part of the proper name in this example-->
<p>In <rs type="place" key="place_00741">London</rs>, a <rs type="institution" key="institution_00295">scholarly society</rs> was founded in 1660.</p> 

<!--Institution name, including location-->
<p>The <rs type="institution" key="institution_00999">Reformed Church in Göttingen</rs> has many members</p>.

Publications mentioned, incl. periodicals

The types of publications are listed in the Conventions on structural data.

Three types are distinguished for publications:

• Single record (concrete characteristics)
• Periodical/collection (abstracted)
• Standardised title (abstracted)

We distinguish between markups in the source text and in the metatext:

• For publications referenced in the source text, volume, page or paragraph references are specified in a reference annotation.
• For publications referenced in the metatext, no volume or page references are included. As no reference annotation is possible there, these specifications appear outside the person tag.

Example: TEI-XML encoding of publications

<!--Publication referenced in the source text-->
dans le troisième tome des <rs type="publication" key="publication_34638">Gazettes de Gottingue<note type="annotRef"><p>Bd. 3.</p></note></rs>.

<!--Publication referenced in the metatext-->
<rs type="publication" key="publication_99999">Müller (1990)</rs>: 324. 

<!--Encapsulation of the proper name-->
<Gessners Flora>

<!--But:-->
sende ich dir die <Flora>, welche <Gessner> in jahrelanger Arbeit erschaffen hat

Lettters mentioned

Mentioned letters (including pro memoria, rescripts, etc.) are marked and referenced if they already exist in the database. Letters that do not yet exist as an entity can be marked with <letter_unknownReference>.

If a letter cannot be assigned to an object referenced in the database, the persons are marked as an exception.

Example: TEI-XML endocing of letters

<!--Mention of a letter-->
<p>I received your <rs type="letter" key="letter_00162">letter dated 1 April</rs>.</p> 
    
<!--Mention of a letter, including persons' names-->
<p>In <rs type="letter" key="letter_00162">Haller's letter to Caldani</rs> it is said that ...</p>

<!--Mention of an unknown letter, including persons' names-->
<p>My <rs type="letter" key="letter_unknownReference">letter </rs> for <rs type="person" key="person_01055">Tscharner</rs>  is ...</p>

Reviews mentioned

Reviews in general

All reviews written by Haller in the Göttingische Zeitung von Gelehrten Sachen (GGA) are catalogued and edited on République des Lettres. Reviews mentioned in the letters (which were not written by Haller) are neither recorded nor referred to (even if they appeared in the GGA). However, the editors are free to provide further information on reviews by means of free annotations.

Example: TEI-XML encoding of reviews with free annotation

<!--Example of correspondence between Haller and J.P. Murray (letter_06158, translated)-->
In the meantime, Lönbom has published a collection of essays on the subject, which I have reviewed in <rs type="publication" key="publication_34638" xml:id="hallerNet_letter_06158_ref-28">our review journal<note type="annotRef"><p>See <rs type="review" key="review_unknownReference" xml:id="hallerNet_letter_06158_ref-29">Murray's review of vol. 1. in <hi rendition="#i">GGA</hi> (21 Jan. 1764)</rs>: 65-70.</p></note></rs>
have.

Haller's reviews in the GGA

Only Haller's reviews in the GGA are referenced, as they are catalogued and edited on République des Lettres. A review must be mentioned in order to be referenced. It is not sufficient to simply mention that a work will be reviewed at a later date in order to place a reference to the review (such a reference is redundant as it is not relevant to the context). Depending on how explicitly a review is mentioned, the reference to the review is made directly in the text or via a reference annotation.

Explicit mention of the review

If one of Haller's reviews is explicitly mentioned in the letter, the review is referenced directly. Explicitly means that the words "review", "review" or similar appear in the letter. The reference to the review may include Haller as the reviewer, the GGA as the review journal as well as the author and name of the publication. As overlong references shall be avoided, only the review may be referenced as an alternative. The other information (publication, author, location, etc.) can be found in the database entry of the review (cf. second example). Similarly to publications, reviews that have not yet appeared at the time of correspondence can also be marked and the status may be described in a free annotation (cf. third example).

Example: TEI-XML encoding of reviews

<!--Explicit mention of a review-->
<p>For <rs type="review" key="review_99999">your review of my work on the origin of plants in the Göttingische Zeitungen</rs> I thank you very much.</p>
    
<!--Explicit mention without naming the review (letter_15537)-->
<p>l’<rs type="publication" key="publication_29278" xml:id="hallerNet_letter_15537_ref-8">ouvrage de M Bergius</rs> que Vous m’avés annoncé. <rs type="review" key="review_02239" xml:id="hallerNet_letter_15537_ref-31">j’en ai parlé</rs> d’aprez une copie.</p>

<!--Explicit mention of a review - with only implicit ref. to public, person and GGA -->
<p>For <rs type="review" key="review_99999">your review</rs> of my work on the origin of plants, which appeared the week before last in the most highly acclaimed and respected Göttingen newspapers, I, Peter Müller, thank you very much.</p>
    
<!--Reference to a review that appeared later-->
<p>I would be delighted <rs type="review" key="review_99999">to receive a review of my work on the origin of plants in the GGA<note n="x" type="annotRef"><p>Haller complied with XY's request. The review appeared in the GGA on 26 July 1753.</p></note></rs>.</p>

Non-explicit mention of the review

If there is no explicit mention of the review in the letter, a reference annotation is used. The reference annotation is placed either at the reviewed publication or at the GGA. It is generally assumed that if both the reviewed publication and the GGA are mentioned in the text, the review is explicitly mentioned (ref. as above). If the reference annotation cannot be set because the GGA or the reviewed publication has already been marked earlier in the text, the review is referenced in a free annotation. Important: It is not necessary to write the entire label of the review in the note. Haller's review + the date is sufficient.

Example: TEI-XML encoding of reviews not explicitly mentioned

<!--Non-explicit mention with reference annotation-->
<p><rs type="publication" key="publication_99999">My publication on the origin of plants<note n="x" type="annotRef"><p><rs type="review" key="review_99999">Haller's review</rs> appeared on 1 January 1760 in the GGA.</p></note></rs> is widely known thanks to you.</p>
    
<!--Non-explicit mention with free annotation (letter_15533)-->
Mon intention n’est jamais <lb/>de desobliger perſonne, et moins <rs type="place" key="place_03334" xml:id="hallerNet_letter_15533_ref-2">Votre Nation</rs>, qu’aucune autre, il me semble que cela ne <lb/>sauroit étre douteux.<note type="annotFree" n="1"><p>This is <rs type="review" key="review_00948" xml:id="hallerNet_letter_15533_ref-17">Haller's review</rs> in the GGA, 14 July 1764.</p></note></p>

Text constitution

While the described form of the text structure, the type of annotation and the reference system are implemented uniformly in all editions, differences can occur in the text constitution. This is due to a distinction in text constitution between Standard and Retroedition. For retroeditions, the focus of the editing is on structuring, annotation and referencing according to the present model. But the transcriptions and text-critical comments are largely taken over from the originally published version; their realisation is described in the introduction to the respective edition.

The standard model also includes optional TEI-XML encodings for special philological requirements. A certain degree of flexibility is important because the editions on the platform include a large number of written documents from a broad temporal and geographical context. Some of these have particular characteristics that need to be addressed in a special way. The following descriptions therefore specify the standard form and an optional variant at certain points. The chosen type of implementation must be applied to the entire unit (e.g. all letters in a correspondence) and made transparent in the introduction.

Non-edited characters and tracings of text carriers

Signs of use and later addenda, such as reference numbers and archival signatures, which may have been corrected and overwritten, as well as contemporary additions, sigla or other annotations, are not transcribed. Additional information on the addresses, such as franking, postal routes etc., is only transcribed if the information visibly originates from the author. Notes made by other people are not transcribed.

Character encoding and punctuation

Greek and Hebrew letters are reproduced in Greek and Hebrew and not transliterated in Latin (occurrences from other writing systems as well). Ligatures and abbreviations can be implicitly normalised.

The transcription of texts in the Latin writing system can also be harmonised. For characters that are no longer in use today (e.g. Lang-s ‘ſ’) or differences in their use over time (e.g. in Latin for ‘u’/‘v’), it is possible to either reproduce them in their original spelling or normalise them. The type of realisation chosen must be consistent in the corresponding edition and be made transparent in the introduction.

Examples of original and normalised implementations:

Character	Original	Normalised
u/v, U/V	Vniuersitas	Universitas
i/j	justitia	iustitia
ij, ÿ	filij/filÿ lachrÿmę	filii lachrymae
æ/Æ, œ/Œ, ę (ligatures, e-caudata)	aulæ, aulę cœpit	aulae coepit
î, â (circumflex for adverbs and nouns in ablative case)	hîc (=hīc) mæstitiâ (=maestitiā)	hic maestitia
è, ò, à (gravis for adverbs and prepositions)	beatè verò à è	beate vero a e

Currency indications and other special characters (e.g. mathematical, alchemical or astronomical notations) are replicated with UNICODE characters wherever possible. If a character does not exist, the tilde in square brackets "[~]" is used as a substitute and its meaning is explained in an annotation on the text constitution.

Author-specific orthography

Punctuation, capitalisation and spelling as well as orthographic peculiarities, errors and inconsistencies are reproduced as precisely as possible and interventions are kept to a minimum. But orthographic or graphematic peculiarities of certain writers (such as the omission of "i" dots or "t" horizontal strokes, umlaut signs, missing "u" strokes or slurs) are generally not imitated. Furthermore, the first word of a sentence and proper nouns are always capitalised and a full stop is added at the end of a sentence if there is no punctuation mark.

If in doubt, harmonisation is undertaken at document level, otherwise according to the writer's standards where a tendency can be identified. In ambiguous cases (e.g. undecidability between "é" and "è"), the present orthography is applied.

Ending syllable elision, nasal strokes and contractions

Duplication strokes or macrons are replaced by doubling the letter (e.g. "nn" instead of "n̄"). Ending syllable elision (e.g. in German for word endings such as "-e", "-em", "-en", "-er'", "-ung" etc.) and suspensions in word endings (such as in Latin "-que" or "-tur") are usually normalised. Similarly, contractions, nasal strokes or special abbreviations are not imitated by bows, hooks or other diacritics. The resolution of these phenomena, which are particularly common in Latin texts, may vary depending on the context.

Examples of latin abbreviations:

Abbreviation	Resolution	Remarks
aīus, aīo, aīa, an̄ę	animus, animo, anima, animae	-
aut	autem	-
Dms, Dmi, Dmo	Dominus, Domini, Domino	-
ē, ee	est, esse	-
Ecclia	Ecclesia	-
epla	epistula, epistola	-
igr	igitur	-
.n.	enim	-
n̄, nō	non	-
noia, noie	nomina, nomine	-
oes, oia, oiu, oib	omnes, omnia, omnium, omnibus	-
ꝑ, ꝕ, ꝓ	per, prae, pro	Often indistinguishable from each other: semꝑ = semper, ꝕsertim = praesertim, ꝓpe = prope
ꝙ, qꝫ	-que	The former often indistinguishable from ꝗ (see below)
ꝗ	qui, quae, quod etc.	Often in combination with other letters or abbreviations: ꝗͥ= qui, ꝗͣ = qua, quia, ꝗͩ = quid, quod, ꝗꝰ = quibus
ꝝ, ꝵ	-rum	-
ꝛ	-tur	-
ꝯ	-us	Often raised or lowered: oibꝰ = omnibus, ejꝯ = eius
ꝟo	vero	A slash or stroke that crosses a consonant: ƀ, ᵬ = ber, bre; đ, ᵭ = der; ŧ, ᵵ = ter, tem, ten; ꝟ = ver, ven

In exceptional cases, certain editorial interventions may be placed in brackets in the transcription. They are intended to support readability by resolving misleading abbreviations or when a forgotten word, letter or punctuation mark is added for reasons of comprehensibility.

Especially for regular abbreviations, it is recommended to transcribe them according to the original and to publish a list of abbreviations in the appendix of the edition (cf. appendix of the Scheuchzer Edition). However, it is also possible to encode the variants (see below).

Typographic annotation

The typographic highlightings underlined and ^superscript are encoded. They are reproduced in the frontend.

The highlighting italics may only be used in annotations on the text constitution and not in the transcription.

Subsequent additions, for example words inserted between the lines, are not reproduced typographically, but described with annotations on text constitution (cf. chapter Annotations on text constitution).

TEI-XML encoding of typographic annotation

Highlighting	TEI encoding
Underline (underline)	<hi rendition="#u"/>
Superscript (superscript)	<hi rendition="#sup"/>
Italics (italic)	<hi rendition="#i">

Text criticism

Text criticism can be expressed either in terms of a semantic description (annotation on text constitution) or by encoding with TEI-XML. The method chosen must be applied to the entire unit (e.g. all letters in a correspondence). The only exceptions are complex phenomena and special cases where the available TEI-XML elements are not sufficient. In this case, a semantic description is applied.

1. Semantic description (standard)

Deletions, insertions and other corrections are described in annotations on the text constitution, which follow a standardised vocabulary. A distinction is made between the editor's text and the original text based on the typographical difference between italics and upright type. Editorial text always appears in italics (including any punctuation relating to editorial annotation), original text appears upright. If there are uncertainties regarding the reading, the word is transcribed according to the editor's assumption and the uncertainty is specified in an annotation on the text constitution.

Illegible elements are transcribed either character by character as "x" (with one "x" representing one illegible character) or, if one or several entire words are illegible, as "[...]". In both cases, a text-critical annotation is supplied.

Multilingual overview of text-critical description

	English	Français	Deutsch
Illegible text	Not legible.	Pas lisible.	Nicht lesbar.
Uncertain reading	see doubtful reading.	see lecture incertaine.	see unsichere Lesung.
Alternative reading	alternate reading: werden.	alternative de lecture : werden.	Alternativlesung: werden.
Uncertain and alternative reading	unreasonable doubtful reading; alternate reading: Unfeasonable.	unreasonable lecture incertaine ; alternative de lecture : Unfeasonable.	unreasonable unsichere Lesung; Alternativlesung: Unfeasonable.
Deletions* (legible)	und follows <auf>.	und après <auf>.	und nach <auf>.
	mathemat above <oeconom>.	mathemat sur <oeconom>.	mathemat über <oeconom>.
	stärker precedes <geworden>.	stärker avant <geworden>.	stärker vor <geworden>.
Deletions (illegible)	die follows <illeg>.	die après <illisible>.	die nach <nicht lesbar>.
Additions between the lines	physical[ische] interl.	physical[ische] ajouté.	physical[ische] eingefügt.
	und seine … retractirt interl.	und seine … retractirt ajouté.	und seine … retractirt eingefügt.
Additions between words / at the beginning or end of the line	ja … hält added in margin.	ja … hält ajouté en marge.	ja … hält am Seitenrand eingefügt.
	und alles … reisen added and interl.	und alles … reisen ajouté.	und alles … reisen eingefügt.
Additions at the side margin	This sentence added sideways in margin.	Phrase ajoutée en marge.	Satz am Seitenrand eingefügt.
Correction (legible)	würde changed from würden.	würde remplacé par würden.	würde ersetzt durch würden.
Correction (illegible)	fuerit changed from ?	fuerit remplacé par ?	fuerit ersetzt durch ?
Text loss	Defect in manuscript.	Manque dans le manuscrit.	Textverlust.
Correction by the editor	Recte fasciculus.	Recte fasciculus.	Recte fasciculus.

* Instead of using angle brackets for deletions, the corresponding word may also be rendered crossed out. This type of representation is only permitted in text-critical annotations, but not directly in the transcription. Example: und follows auf.

---

2. Basic TEI-XML coding (optional variant)

The standard model also allows text-critical encoding using TEI-XML. But it explicitly refers to frequently occurring, basic phenomena such as deletions, corrections or editorial additions. More complex cases, such as the overlapping of several phenomena, are not encoded, but described semantically in text-critical annotations.

TEI-XML encoding of text-critical phenomena

	TEI-XML-encoding	Rendering
Deletion	<del rendition="#lt">	Deletion
Addition	<add/>	\Addition/
Correction	<del/><add/>	Deletion \Addition/
Regularisation	<choice><sic/><reg/></choice>	Mouseover with variant
Abbreviation	<choice><abbr/><expan/></choice>	Mouseover with variant
Completion in case of omission	<supplied reason="omitted-in-original"/>	Com[p]letion