Grammar

Introduction

Grammar, specifically morphology, is a big part of the current version of Language Explorer. Rather then dive into to all of the various grammar-related “features”, we suggest that you look at the following page which describes how Language Explorer approaches morphophonology and morphotactics: Conceptual Introduction to FLEx Grammar.

Example Sketch

Another way to see what is included at this point in the way of grammar is to look a one of the automatically generated grammar sketches produced with Language Explorer. Here is one from our made-up test language, Kalaba.

Parsing

The following is taken from Andy Black’s CTC 04 presentation about the parsing abilities of FieldWorks Language Explorer:

Morphological parsing is useful for:

  • Interlinearizing text
    • Break down the words into their morphemes and give each morpheme a gloss, category, etc.
  • Spelling check
    • Word grammar check
    • Can this word be analyzed given the current state of the lexicon and word grammar?
  • A mechanism for language discovery
    • Discovering generalizations
  • Develop a lexicon of morphemes
  • Develop a formalized morphology/word grammar
  • Adaptation from one related language to another.

Andy Black and Gary Simons wrote the following paper in 2006: The SIL FieldWorks Language Explorer Approach to Morphological Parsing.

Distinctives of Current Tool

  • Morphological Sketch
    Can get a write-up describing the current state of the grammar at any time.
  • Notions common to Grammar I:
    The linguistic notions of root, stem, inflection, derivation, compound rules, and clitics are modeled directly.
  • Stealth-to-wealth
    A user doesn’t have to know what a given form is – they can mark it as unknown until they know.
  • Inflection:
    • Has inflectional templates as a construct.
    • Has tools for building inflectional templates.
    • Has a special help tool for glossing inflectional affixes.
    • Allows for inflection classes.
  • Morphological Parsing:
    • Use and maintain a wordform inventory.
    • Distinguish between parses which have been approved by the user and parses predicted by the grammar.
    • When doing interlinear text analysis, parses predicted by the grammar are shown as defaults. The user can choose or override them.
    • Ability to show the trace of a parse.
  • Wordsets:
    • Ability to create sets of wordforms that have a common theme; for example:
      • A given paradigm, including well-formed and ill-formed wordforms;
      • Words with a common phonological process; or
      • Words with compounds.
    • Ability to check the accuracy of the predictions of the current grammar for each such wordset.
  • Residue:
    • Allow for “unmarked” affixes.
    • Allow the user to write ad hoc prohibitions for morphemes.
    • Allow the user to write ad hoc prohibitions for allomorphs.
    • Allow the user to group ad hoc prohibitions to capture some common characteristic.
  • Ability/Potential to integrate with the rest of the FieldWorks tools.

Leave a Reply

(required)

(required)