format.md

Format

Please insert your fields as you need them (not too much, though) – the format is still up for consideration. Please note that the order shown here should be followed.

• toaq – the Toaq word or phrase in question
• type – part of speech – one of the following:
  • "root" (not necessarily monosyllabic);
  • "compound" for genuine compounds rather than borrowings;
  • "borrowing";
  • "serial predicate";
  • for particles, a descriptive name such as "predicatizer", "text-quote"…
• english – definition in English
• gloss – linguistic gloss in English; use periods instead of spaces
• keywords - a list of English words that one might search for when trying to find the headword; synonyms of the gloss or of the main words found in the sentential definition
• frame – frame definition, if known; "variable" means that the word doesn't have a set frame (possibly anaphoric)
• distribution - a string of the same format as frame where each slot type is replaced with either of ‘d’ or ‘n’ (for ‘distributive’ and ‘non-distributive’); "variable" – as above
• subject - describes how the subject is constrained. Useful for semantics analysis. One of six string values:
  • "event" is for subjects that are necessarily events (like bıe).
  • "proposition" is for subjects that are necessarily propositions (like guosıa).
  • "individual" is for verbs whose subjects can't be events or propositions, and so they should subject-share, like jaı.
  • "agent" implies "individual" and also syntactic ergativity. An example is koı.
  • "free" is for verbs whose subjects can be anything, like gı. This implies no subject-sharing.
  • "shape" is kinda weird. It's for subjects that have a "shape", like sao. I don't know if we want to pretend events have "shapes", but I think it's kind of nice. Then Koı jí sâo means I walk and the event is spatio-temporally big. So if people agree with this I will change all the "shape"s to "free", otherwise I will change them to "individual".
• namesake – whether this word is the name used to refer to its frametype
• notes – array of strings
• examples – array of objects:
  • toaq
  • english
• etymology – etymological information; only present in borrowings
• fields – array of arrays of strings: each element describes one slot; each string is a valid Toaq predicate; those together, when ORed (or rather: «ra»-ed), represent the constraint on that slot. For example:

  [
    ["mıe"],
    ["saı", "lufē"]
  ]

  describes «chıaq»: x₁ must mỉe and x₂ must to ra sảı to lủfē.

tools/normalize.js is for normalizing the dictionary file.

At the moment, oblique words (prepositions and adverbs) should be entered as examples. The normalizer will reject «daı» and «dãı» as duplicates.

Example entry:

{
    "toaq": "cho",
    "type": "root",
    "english": "▯ likes ▯.",
    "gloss": "like",
    "keywords": [
      "like",
      "fond of",
      "enjoy",
      "pleased by"
    ],
    "frame": "c c",
    "distribution": "d d",
    "notes": [],
    "examples": [
      {
        "toaq": "Chỏ da.",
        "english": "There is fondness."
      },
      {
        "toaq": "Chỏ jí da.",
        "english": "I am experiencing fondness."
      },
      {
        "toaq": "Hẻ chỏ déo báq rủa da.",
        "english": "The children are fond of flowers."
      },
      {
        "toaq": "Chỏ súq hı nỉ fủa moq.",
        "english": "Which of these pieces of furniture do you like?"
      },
      {
        "toaq": "chó",
        "english": "those who have fondness"
      }
    ],
    "fields": [
      ["lıe"],
      ["raı"]
    ]
}