Preliminary, ad hoc schema for encoding Streitberg's Gotisch-Griechisch-Deutsches Wörterbuch (1910). See https://www.wulfila.be/lib/streitberg/1910/#XML for more information. Tom De Herdt, November 2006. Revisions: [2026-03-24] Minor edits. [2026-06-03] Moved attributes out of the xmlns:db namespace; updated documentation. [2026-06-07] Use default namespace, to avoid problems with XSLT processing of embedded dictionary entries. [2026-08-03] Added langage code goh (Old High German). Root element for the dictionary. Timestamp indicating when the original file was generated from the database. Some basic metadata. An entry in the dictionary. Entry identifier, encoding the page number and position on the page, e.g. "P065.15" = 15th entry on page 65. May be used as a sort key (headwords are not sorted alphabetically in Streitberg's dictionary, as compound verbs are grouped together with the main verb). P[0-9]{3}\.[0-9]{2} Index number for homographs that have not been grouped into one entry, corresponding to members of a TEI <superEntry>. The number is printed before the boldface headword in Streitberg's dictionary. However, there may be unnumbered (compound) entries in between numbered entries, which complicates grouping into a superEntry (e.g. the sequence 1.wisan; at-wisan; faura-wisan; fra-wisan; [...] 2.wisan; bi-wisan; 3.wisan: the number is implied on the compound verbs). To be discussed. 1 9 When present, the entry is a reference to another entry; the attribute value defines the type of link. The sense element should probably be encoded using an equivalent of TEI <xr>. link-spelling The entry is a spelling variant, pointing to the main lemma, e.g. ‘birusjos’ referring to ‘berusjos’. link-derived The entry is a derived or inflected form, pointing to the main lemma, e.g. ‘wait’ referring to ‘witan’. link-suppletive The entry is a suppletive form, pointing to the main lemma, e.g. ‘batiza’ referring to ‘goþs’. link-conjecture The entry is a conjecture, pointing to the main lemma. link-hyphen The entry is a form with differing hyphenation, pointing to the main lemma. Temporary flag for internal use. problematic Encoding of this entry is problematic (see notes for information). One out of several homographs with different part of speech, grouped in one entry (equivalent to TEI <hom>). Related entry in the body of another entry, typically a derived or inflected form of the entry word (equivalent to TEI <re>). Definition and translation of the entry. Contains mixed content, optionally followed by one or more nested sense elements (corresponds to TEI <sense>). Headword (corresponds to TEI <orth>) Information on final devoicing, using three-valued logic (true = e.g. hlaifs, hlaiBis; false = e.g. bloþ, bloþis; missing = not applicable). Streitberg indicates final devoicing by highlighting the devoiced phoneme (italics) and adding the stem consonant in parentheses, e.g. hlaiFs (b). To be tagged manually, as he is not always consistent. true false The headword is an inflected form, because the lemma and inflectional class cannot be determined with certainty. Typically proper names, foreign words, or nouns that have only survived in the dative or genitive plural. true Information following the headword, mostly on spelling, phonology or grammar, with references to Streitberg's Gotisches Elementarbuch (1920) or other sources. Alternative form(s), typically spelling variants for the headword. Doesn't fit very well into the current (intermediate) schema, but alternative headwords are rare in the dictionary (about 20-30 occurrences). To be discussed; see also element grammarAlternative. Grammatical information, usually combining part of speech and inflectional class into one label, e.g. ‘Ma’ (noun, masculine a-stem) or ‘st.V.3’ (strong verb, third category). The label for suppletive forms may refer to the main lemma. The grammatical label is implied, i.e. omitted by Streitberg because the headword is a compound. Grammatical information is generally not repeated on derived forms (e.g. verbs with prefix; the reader is referred to the base form). The label should be hidden or rendered in gray when generating an accurate typographical or editorial view. true Normalization of the given grammatical label. The reason to do so is specified in @normReason. May be useful when searching for all instances of a particular class. Specifies the relation between @norm and the label used by Streitberg. variation Alternate spelling or wording, e.g. ‘Akk.Pl.’ for ‘Akk.Plur.’. subclass Subset of the generic class, e.g. ‘st.Adj.a’ for ‘Adj.a’. uncertain Hypothesis, e.g. ‘M?a’ or ‘M(i)’. irregular Irregular instance of the generic class, e.g. ‘unreg.abl.V.1’. The lemma has irregular inflection. true The grammatical class is hypothetical (usually expressed by question mark or parentheses in the label). true Foreign key: identifier of a part-of-speech tag in the database. This information is not in Streitberg. Information that follows the class label in Streitberg's lexicon, mostly on morphology. Alternative reconstructions, having different declension (e.g. skauns* or skauneis*). Doesn't really fit into the current (intermediate) schema, but entries with alternative reconstructions are rare. To be discussed; see also element formAlternative. Note by the encoder, not part of the source text. May be used for additional information (public), to signal encoding problems (internal), etc. Scope or type of the note: internal Temporary note for internal use, e.g. related to the encoding process. public Permanent note that may be rendered as a footnote to the actual data, e.g. related to the content of the entry or corrections to the dictionary text. Name of acronym of the person responsible for the note. List of free-form tags (à la Web 2.0) that may be used internally to flag or categorize the note.
Common elements, attributes and content models. ISO-639 language code, as per RFC 3066: Tags for the Identification of Languages. got Gothic grc Greek, Ancient non Norse, Old de German la Latin en English nl Dutch goh German, Old High Foreign key: numerical identifier of a corresponding lemma in the database. 1 9999 Label used in numbering homographs or nested sense-levels. Either a single digit, letter or roman number less than or equal to 10. [0-9]|[a-zA-Z]|(V?I{1,3}|I?V|I?X) The form is reconstructed, normally marked with an asterisk in the source text. Currently redundant, as the asterisk is retained in the text transcription, but it simplifies queries. Could also be used in future to label reconstructions that Streitberg did not mark as reconstructed; cf. modern practice in Braune/Ebbinghaus 1981, where the asterisk is applied consistently: “Ganz besonders dringlich schien es mir, die belegten von den unbelegten Formen saüberlich zu trennen.” (p. 5) true The form is a conjecture, in square brackets. Streitberg: “Durch Konjektur entstandene Formen stehn in eckigen Klammern [].” true The form is a dubious reading, marked with dagger in front of the word. Streitberg: “ein Kreuz (†) vor einer Form charakterisiert die handschriftliche Überlieferung als besserungsbedürftig.” true Emendable plain text. Mixed content used in block-level elements.
Inline elements, used in mixed content blocks. Short names, in order to improve readability of the densely tagged definition text. Signals apparent errors (equivalent to TEI <sic>). Strictly limited to obvious typographical errors, e.g. “kinnus *Fn” for “Fu”. TEI P4: “Gives a correction for the apparent error in the copy text.” TEI P4: “Signifies the editor or transcriber responsible for suggesting the correction.” Words that are mentioned but not used (equivalent to TEI <mentioned>). Temporary solution for tagging Gothic, Greek, Latin or Old Norse words in the (German) definition text, as indicated by the language identifier (xml:lang), without interpretation of their function. There is some overlap with the translation element for Greek and the quotation element for Gothic. Gothic words followed by a reference to the text are marked up as quotations; only references to headwords or inflected forms mentioned in a grammatical context (e.g. in a sentence like “gaggan has irregular past tense ‘iddja’”) would be tagged as mentioned. The distinction is admittedly problematic. To be discussed. Translation into German or from Greek, as indicated by the language identifier (xml:lang). The distinction between Greek translations and mentioned words is not always clear. The entire dictionary has a problematic concept of ‘multilinguality’: are the German glosses translations of Gothic or Greek words? There is some inconsistency in the way Streitberg handles translations (naturally, since the dictionary predates digital text encoding by many decades). Translations occasionally include grammatical information, as in “einem (Dat.) etwas (Akk.) zeigen” (lemma at-augjan). The labels refer to Gothic case government, not German (even though they mostly coincide). Quotation from the Gothic Bible or minor fragments, i.e. Gothic text (as specified by the language identifier). However, the distinction between quotations and mentioned words is not always clear. As there is only one main source of Gothic text (cf. Mossé 1956, p. 19: “nous ne connaissons pas *le* gotique, nous ne connaissons que la langue de la Bible wulfilienne [...]”), every Gothic word could be considered a quotation, minus reconstructions or inflected forms in a grammatical context (see mentioned element). There are obviously no constructed examples in the dictionary. Generally speaking, quotations are followed by a reference to the text (see bib element). To be discussed. Grammatical label that gives information on case, tense, aspect, usage etc. Stricter typologies (and normalized values) could be defined later on. Manuscript siglum, e.g. ‘CA’ (Codex Argenteus) or ‘AB’ (= 'both in Codex Ambrosianus A and B'), used when quoting spurious readings, conjectures or variants. Unspecified highlighting by means of ‘Sperrdruck’ (i.e. spaced letters, ‘l i k e t h i s’). Used for part-of-speech descriptions in homographs and more generally whenever a distinction is made between (grammatical) items within one entry. Should be replaced by a rendition attribute on logical elements (corresponding to TEI rend). Hardly used at all in the current transcription (and never for grammar elements in homographs). Cross-reference to another lemma, specified by the key in the target attribute. Contains either a Gothic headword (tagged as mentioned) or any other text that could be turned into a hyperlink to another entry (e.g. the abbreviation “s.d.”, “siehe dort”). Foreign key: reference to the key of another entry. P[0-9]{3}\.[0-9]{2} Reference to the Gothic Bible or minor fragments. A normalized target attribute could be supplied automatically in a second stage of encoding, using regex rules. Reference to Streitberg's Gotisches Elementarbuch (1920) or other sources. A normalized target attribute could be supplied automatically in a second stage of encoding, using regex rules. Superscript reference to a numbered note in the Elementarbuch: “Die [...] Zahlen verweisen auf die Paragraphen des Gotischen Elementarbuchs, ihre Exponenten bezeichnen die Anmerkungen.” (Streitberg 1910, Abkürzungen) \d+[a-zA-Z]? Semantically neutral span element, used to tag a segment of inline text for any purpose. Currently groups mixed content preceding nested sense elements (which makes it easier to maintain indentation and readability after applying XSLT transformations). A better purpose would be to bind quotations, translations and text references together, or to tag the grouping of quoted examples according to case.