Quantifying Texts

The list below summarizes typical problems that the Discourse Profiler program can be applied towards solving:

  1. Verify and/or identify the transitivity of a second transitive construction through topic continuity statistics (use the Givón statistical approach).
  2. Identify the selection criteria between two distinct transitive construction types (use the Dryer statistical approach).
  3. Identify the reason for pragmatic word order differences such as between SVO and VOS (see Quick 2006 for VP fronting as a possible explanation for the Pendau language, and which may be applicable to other similar structured languages—also compare Dooley and Levinsohn 2001, Levinsohn 2003).
  4. Verify a transitive construction is not an antipassive (necessary to disqualify an ergative analysis in some cases; use the Givón statistical approach).
  5. Levinsohn’s S1-S4 categories are used for four categories of subject continuity and  provides statistics for a special type of topic continuity analysis that complements Givón’s statistical approach.
  6. Examine the raw profile of the frequency of particular participants according to their NP type, according to their occurrence as a semantic role/macro role, or according to their occurrence as a grammatical relation.
  7. There are many other analyses possible by using the interactive capabilities and making various ‘what if’ hypotheses to test these.  One of the benefits of the Discourse Profiler program is that it allows the user to make rapid shifts and/or refine one’s hypothesis testing or analysis, and to analyze a higher number of texts more efficiently.
  8. An additional benefit for the Discourse Profiler tags in the Toolbox database can be to search for specific examples of text data with a specific code or coding combination.  This can be used for various grammatical analyses and for finding examples to use in your research paper or other research goals (see Pastika 1999, 2006 for examples of an application of my tagging methodology used in his texts).