DoReCo: Language Documentation Reference Corpus

# DoReCo Language Documentation Annotation
# Based on Paschen et al., LREC 2020
# Paper: https://aclanthology.org/2020.lrec-1.324/
# Dataset: https://doreco.huma-num.fr/
#
# Task: Multi-tier language documentation annotation following ELAN conventions.
# Annotators segment field-recorded audio into words and morphemes, provide
# Leipzig-style interlinear glosses, free translations, and clause-type
# classifications -- all aligned to the same audio timeline in parallel tiers.
#
# ELAN-style multi-tier design:
#   Tier 1 (span)  - word_tier: word-level segmentation with word-class labels
#   Tier 2 (span)  - morpheme_tier: morpheme-level segmentation with morpheme types
#   Tier 3 (text)  - gloss: Leipzig-style interlinear gloss for each morpheme
#   Tier 4 (text)  - free_translation: sentence-level free translation
#   Tier 5 (radio) - clause_type: syntactic clause type classification
#
# Guidelines:
#   - Start by listening to the full utterance, then segment words on Tier 1
#   - Subdivide each word into morphemes on Tier 2 where applicable
#   - Use the Leipzig Glossing Rules for Tier 3 (capitalize grammatical glosses)
#   - Provide an idiomatic free translation on Tier 4
#   - Select the clause type on Tier 5 based on the full utterance

annotation_task_name: "DoReCo Language Documentation Annotation"
task_dir: "."

data_files:
  - sample-data.json
item_properties:
  id_key: "id"
  text_key: "audio_url"

output_annotation_dir: "annotation_output/"
output_annotation_format: "json"

annotation_instructions: |
  ## DoReCo Language Documentation Annotation

  You will annotate field-recorded audio from under-described languages using a
  multi-tier ELAN-style annotation scheme. Each tier captures a different level
  of linguistic analysis, and all tiers are time-aligned to the same audio file.

  ### Tier 1 -- Word Tier (word-level segmentation)
  Segment the audio into individual words and classify each:
  - **content-word** -- Nouns, verbs, adjectives, adverbs with lexical meaning
  - **function-word** -- Determiners, prepositions, conjunctions, auxiliaries
  - **interjection** -- Discourse particles, response words (e.g., "mhm", "oh")
  - **hesitation** -- Filled pauses, false starts, self-corrections
  - **foreign-word** -- Words borrowed or code-switched from another language

  ### Tier 2 -- Morpheme Tier (morpheme segmentation)
  Segment each word into its constituent morphemes:
  - **root** -- The lexical root or stem of the word
  - **prefix** -- Morpheme attached before the root
  - **suffix** -- Morpheme attached after the root
  - **infix** -- Morpheme inserted within the root
  - **clitic** -- Phonologically bound but syntactically free element
  - **reduplication** -- Repeated morphological material

  ### Tier 3 -- Gloss
  Provide a Leipzig-style interlinear gloss for each morpheme. Use standard
  abbreviations in CAPS for grammatical categories (e.g., PST, PL, NOM, ERG,
  3SG). Gloss lexical morphemes in lowercase English.

  ### Tier 4 -- Free Translation
  Provide an idiomatic English translation of the full utterance.

  ### Tier 5 -- Clause Type
  Classify the syntactic type of the clause:
  - **declarative** -- Statement
  - **interrogative** -- Question
  - **imperative** -- Command or request
  - **exclamative** -- Exclamation
  - **relative** -- Relative clause
  - **subordinate** -- Subordinate/adverbial clause

annotation_schemes:
  - annotation_type: span
    name: word_tier
    description: "Word-level segmentation. Select each word span and classify it by word type."
    span_mode: temporal
    labels:
      - name: "content-word"
        color: "#3B82F6"
        tooltip: "Lexical word carrying semantic content (nouns, verbs, adjectives, adverbs)"
        key_value: "1"
      - name: "function-word"
        color: "#10B981"
        tooltip: "Grammatical word (determiners, prepositions, conjunctions, auxiliaries)"
        key_value: "2"
      - name: "interjection"
        color: "#F59E0B"
        tooltip: "Discourse particle or response word (e.g., mhm, oh, yeah)"
        key_value: "3"
      - name: "hesitation"
        color: "#EF4444"
        tooltip: "Filled pause, false start, or self-correction"
        key_value: "4"
      - name: "foreign-word"
        color: "#8B5CF6"
        tooltip: "Borrowed word or code-switch from another language"
        key_value: "5"

  - annotation_type: span
    name: morpheme_tier
    description: "Morpheme-level segmentation within words. Identify morpheme boundaries and classify each morpheme."
    span_mode: temporal
    labels:
      - name: "root"
        color: "#1D4ED8"
        tooltip: "Lexical root or stem -- the core meaning-bearing morpheme"
        key_value: "q"
      - name: "prefix"
        color: "#059669"
        tooltip: "Bound morpheme attached before the root"
        key_value: "w"
      - name: "suffix"
        color: "#D97706"
        tooltip: "Bound morpheme attached after the root"
        key_value: "e"
      - name: "infix"
        color: "#DC2626"
        tooltip: "Bound morpheme inserted within the root"
        key_value: "r"
      - name: "clitic"
        color: "#7C3AED"
        tooltip: "Phonologically bound but syntactically independent element"
        key_value: "t"
      - name: "reduplication"
        color: "#DB2777"
        tooltip: "Repeated morphological material (full or partial reduplication)"
        key_value: "y"

  - annotation_type: text
    name: gloss
    description: "Leipzig-style interlinear gloss. Use CAPS for grammatical categories (e.g., PST, PL, NOM) and lowercase for lexical glosses."

  - annotation_type: text
    name: free_translation
    description: "Idiomatic English free translation of the full utterance."

  - annotation_type: radio
    name: clause_type
    description: "Syntactic clause type of the utterance."
    labels:
      - name: "declarative"
        tooltip: "Statement or assertion"
        key_value: "a"
      - name: "interrogative"
        tooltip: "Question (content or polar)"
        key_value: "s"
      - name: "imperative"
        tooltip: "Command, request, or instruction"
        key_value: "d"
      - name: "exclamative"
        tooltip: "Exclamation expressing surprise, emphasis, or emotion"
        key_value: "f"
      - name: "relative"
        tooltip: "Relative clause modifying a noun"
        key_value: "g"
      - name: "subordinate"
        tooltip: "Subordinate or adverbial clause"
        key_value: "h"

html_layout: |
  <div class="annotator-container" style="max-width: 900px; margin: 0 auto; font-family: sans-serif;">
    <h3 style="margin-bottom: 4px;">DoReCo Language Documentation</h3>
    <p style="color: #6B7280; margin-top: 0;">
      Language: <strong>{{language_name}}</strong> ({{language_iso}}) |
      Speaker: <strong>{{speaker_id}}</strong> |
      Genre: <strong>{{genre}}</strong>
    </p>

    <div class="audio-container" style="background: #F3F4F6; border-radius: 8px; padding: 16px; margin-bottom: 16px;">
      <audio controls style="width: 100%;">
        <source src="{{audio_url}}" type="audio/wav">
        Your browser does not support the audio element.
      </audio>
      <div id="waveform" style="width: 100%; height: 128px; margin-top: 8px; background: #E5E7EB; border-radius: 4px;"></div>
      <p style="font-size: 0.85em; color: #9CA3AF; margin: 4px 0 0;">
        Click and drag on the waveform to select time-aligned spans for annotation.
      </p>
    </div>

    <div class="tier-panel" style="border: 1px solid #D1D5DB; border-radius: 8px; padding: 12px; margin-bottom: 12px;">
      <h4 style="margin: 0 0 6px; color: #3B82F6;">Tier 1 -- Word Segmentation</h4>
      <p style="font-size: 0.85em; color: #6B7280; margin: 0 0 8px;">
        Segment the utterance into words and classify each word.
      </p>
      {{word_tier}}
    </div>

    <div class="tier-panel" style="border: 1px solid #D1D5DB; border-radius: 8px; padding: 12px; margin-bottom: 12px;">
      <h4 style="margin: 0 0 6px; color: #059669;">Tier 2 -- Morpheme Segmentation</h4>
      <p style="font-size: 0.85em; color: #6B7280; margin: 0 0 8px;">
        Subdivide words into morphemes and classify each morpheme type.
      </p>
      {{morpheme_tier}}
    </div>

    <div class="tier-panel" style="border: 1px solid #D1D5DB; border-radius: 8px; padding: 12px; margin-bottom: 12px;">
      <h4 style="margin: 0 0 6px; color: #D97706;">Tier 3 -- Interlinear Gloss</h4>
      <p style="font-size: 0.85em; color: #6B7280; margin: 0 0 8px;">
        Provide Leipzig-style glosses (CAPS for grammatical categories, lowercase for lexical items).
      </p>
      {{gloss}}
    </div>

    <div class="tier-panel" style="border: 1px solid #D1D5DB; border-radius: 8px; padding: 12px; margin-bottom: 12px;">
      <h4 style="margin: 0 0 6px; color: #DC2626;">Tier 4 -- Free Translation</h4>
      <p style="font-size: 0.85em; color: #6B7280; margin: 0 0 8px;">
        Provide an idiomatic English translation of the entire utterance.
      </p>
      {{free_translation}}
    </div>

    <div class="tier-panel" style="border: 1px solid #D1D5DB; border-radius: 8px; padding: 12px; margin-bottom: 12px;">
      <h4 style="margin: 0 0 6px; color: #7C3AED;">Tier 5 -- Clause Type</h4>
      <p style="font-size: 0.85em; color: #6B7280; margin: 0 0 8px;">
        Classify the syntactic clause type of this utterance.
      </p>
      {{clause_type}}
    </div>
  </div>

audio_display:
  show_waveform: true
  playback_controls: true
  allow_speed_control: true

allow_all_users: true
instances_per_annotator: 30
annotation_per_instance: 2
allow_skip: true
skip_reason_required: false