sets consisting of substantial amounts of actual field data are an
essential part of the quality control process in putting Wordcorr
together. We've found that some of them are also an excellent way
to introduce people to Wordcorr.
You download them from the
Files menu in SourceForge,
under "test data."1
Put them in the same folder where you plan to keep your own
research data. Import them into Wordcorr as explained in Wordcorr
Once a demo file is imported,
- Go to its Views menu
- Click on the Original
- Then click on the Copy
- On the Copy View dialog box,
give it the name Demo.
- Choose the Varieties Only
- Click OK.
Now you have all the data at your
fingertips, but as far as annotations and tabulations are
concerned, you have a clean slate.2
You don't have to enter data, but you are free to try out
any analytical ideas you want. If things get too messy to keep
track of, there's a Delete button on the View panel that will wipe
out your first Demo view and let you start over. Just do what the
preceding paragraph says and try again.
The most productive way to
introduce people to Wordcorr is this: have them
- Click on Entry 1 (the
gloss at the top of the gloss list on Data or Annotate).
- Fiddle around with annotating
it - assigning every form in the entry to one group or another
on the basis of which ones could reasonably go back to the same
- For all the forms in the
largest group, add Indel symbols (/) and maybe Ignore symbols
(.) to line up corresponding segments in all the forms
that are to be compared. Doing that should make all the forms
come out the same length.
Wordcorr Help gives you examples
and explains Indels and Ignores (and metatheses and grapheme
clusters, which we only mention here). Repeat for the other groups.
Once all the forms in the entry
are annotated, go to Tabulate. It shows the correspondence sets
implied by the decisions you made in annotating.
- Assign each correspondence set
to a protosegment. (If they're all p, assign them
to proto p. If they're not all the same, use your
imagination or your linguistic training.)
- Characterize the environment
each correspondence set is in - maybe for that you'll want to
take a look at existing annotations on the other views.
- When every
correspondence set for the group has a protosegment and an
environment assigned, click Save.
Then look at Refine. If it
clarifies things to make changes, make changes. Wordcorr
Help tells you how.
After you finish with Entry 1, do
the same thing for Entry 2, and so on. Take a break for eating and
sleeping every so often. Have fun. Bring in your friends. Start a
new fad. It won't be long until you're comfortable with Wordcorr,
and you'll know what the textbooks
are talking about - in detail.
1You may want to check the install page for additional
information about downloading
2If you want to sneak a peek at how
other people have annotated your demo to get a feel for their
thought processes, choose any other view than your Demo view.
Compare its Data panel, which is for all views, with the Annotate,
Tabulate, and Refine panels. You'll learn a lot that way. You'll
learn even more by thinking your way through your own Demo
|If you have data sets you would like to share as test sets
or training sets for the Wordcorr community, let us know.