Instructions for use:
'Previous Example' and 'Next Example'
refer to corresponding examples in the book, whereas
'Previous Item' and 'Next Item' show the derivation
of individual words.
You can choose any example from the book, and edit them to see alternate
results, by using the menus provided.
The Language menu allows you to select examples by language,
rather than location in the text. You can return to the master list of
examples by selecting Example>Open example list... and then
selecting "Master list of examples...".
For some languages, the Language menu also contains options for selecting
different dialects or micro-dialects of a language. This allows the user
to see how changing one parameter can change the results of harmonization.
(e.g. in Finnish you can select whether the search is relativized to
Contrastive or Marked values.)
Example>Select example from list allows you to navigate directly
to a particular example (rather than using the buttons at the bottom
of the applet). Example>Edit example allows you to edit the root
and to select the affixes or change their order.
The Step Size menu will change the amount of information displayed
for each derivation. In all cases, the underlying algorithm is displayed,
but the amount of detail displayed varies:
- Source
- Every step of the search is displayed.
- Feature
- Only successful copying of features or failure of the search is
displayed (i.e. transparent segments are not displayed).
- Target
- The results of the search algorithm are displayed one needy
segment at a time.
Variables
The following variables represent the parameters of the search:
- δ
- The direction of the search: Right, Left,
or both L&R
- F
- The features needed by the current target segment,
along with τ, the value relativization which defines
"visible" source segments.
- Any
- Any segment with a value for the needy feature is
a valid source.
- Contrastive
- Only segments which are contrastive on the needy feature
are visible to the search.
- Marked
- Only segments with the marked value for the needy feature
(natural or language-specific) are visible to the search.
- R
- A source condition on the segment to be copied from, usually
an orthogonal feature which, if different on an otherwise valid
source will result in defective intervention and default value
insertion.
- χ
- The default values for given features.
- β and γ
- Distance bounds. γ defines whether a distance threshold is
measured in segments or syllables, and β is the number of units
(1 or 2).
- σ
- The number of segments/syllables encountered so far.
- ζ
- The sonority threshold for the search. If a more sonorous segment
is encountered, regardless of its visibility to the search, the search
is terminated and the target copies from that segment.