How tagging works

Each word token in the text can be annotated with four pieces of information:

• cf — citation form: the lemma as listed in a dictionary (nominative singular for nouns, infinitive for verbs)
• gw — guide word: a short English gloss distinguishing this lemma from homophones
• pos — part of speech
• tag — full morphological tag, built from lemma class + morphological features

An optional cg (context gloss) can record the specific translation in context.

Tagging a word

1. Click any word in the text to open the tag panel.
2. If "Possible readings" appear, click one to assign it and move on.
3. Otherwise, fill in pos and cf. The tag builder will appear for the chosen POS.
4. Work through the builder buttons (class, number, case, etc.) — the tag field fills automatically.
5. Press Enter or click Save.

To remove a tag from a word, open its panel and click the chosen reading badge to deselect it.

Keyboard shortcuts

Arrow keys and letter shortcuts work when no text field is focused.

Norse character shortcuts

Type these sequences in any text field to insert special characters:

Parts of speech

Noun proper-name subtypes — PN (person), DN (deity), FN (female), GN (place), EN (ethnic) — are set via the Type row in the noun builder.

Tag format

The tag is automatically assembled by the builder. It combines a lemma class and morphological features, separated by dots.

Nouns

{gender}.{stem-class}[.{name-type}].{number}.{case}[.def]

Examples: m.a-stem.sg.nom, f.ōn-stem.pl.dat.def, m.a-stem.PN.sg.nom

Verbs

{str|wk|irr}[.{class}].{form}[.{tense}[.{person}{number}]]

Examples: str.1.ind.prs.3sg, wk.kasta.inf, wk.telja.subj.pst.3pl, str.1.ptcp.pst.str.m.sg.nom

Adjectives

{gen|dem|poss}[.{pos|comp|sup}[.{str|wk}.{gender}.{number}.{case}]]

Examples: gen.pos.str.m.sg.nom, gen.comp

Pronouns

{pers|indef|other}[.{person}[.{gender}]].{number}.{case}

Examples: pers.1.sg.nom, pers.3.m.sg.gen

Morphological abbreviations

Dictionary view

Click Dictionary in the sidebar to browse the lexicon by headword. Click any entry to see its full paradigm and all corpus instances. The form-search bar lets you look up any word form directly.

The Citation form / Inflected form search box supports wildcards: * matches any number of characters and ? matches a single character, and you can combine several in one query. A query with no wildcards still matches anything starting with what you typed (e.g. kal matches "kalla"), so add * yourself to match elsewhere too (e.g. *son finds anything containing "son").

Export

Click Export data in the sidebar (requires sign-in) to download the annotated corpus in four formats:

• TSV corpus — one row per token with text, section, cf, pos, tag
• TSV lexicon — one row per wordform entry
• CoNLL-U — standard dependency-parsing format
• Vertical — one token per line with structural tags
• JSON — nested by text and section