reduplication of verb extensions

[Jeff and Peg Shrum]

I have a question about how to define things so that the parser will treat a reduplicated morpheme as one morpheme.  As in Greek and many other languages as pointed out in Marlin Leader’s “Flex guide”  reduplication of certain morphemes have the function of indicating perfective aspect.  In TKE when the applicative verb extension is reduplicated then the verb has perfective aspect.  The applicative verb extension has the form “el”, and has the normal meaning of doing something “to, for, for the benefit of, or for the purpose of.”  The meaning of the reduplicated morpheme is very different than the non-reduplicated one.  Even though I have both “-el”, applicative and “-elel” perfective in my lexicon, the parser always chooses to parse using the two smallest possible  morphemes i.e. “-el –el” and never “elel”.  How can I get the parser to choose the bigger morpheme when it is present?

 

Jeff S.

[Andy Black]

On 6/26/2009 1:37 AM, Jeff and Peg Shrum wrote:

I have a question about how to define things so that the parser will treat a reduplicated morpheme as one morpheme.  As in Greek and many other languages as pointed out in Marlin Leader’s “Flex guide”  reduplication of certain morphemes have the function of indicating perfective aspect.  In TKE when the applicative verb extension is reduplicated then the verb has perfective aspect.  The applicative verb extension has the form “el”, and has the normal meaning of doing something “to, for, for the benefit of, or for the purpose of.”  The meaning of the reduplicated morpheme is very different than the non-reduplicated one.  Even though I have both “-el”, applicative and “-elel” perfective in my lexicon, the parser always chooses to parse using the two smallest possible  morphemes i.e. “-el –el” and never “elel”.  How can I get the parser to choose the bigger morpheme when it is present?

First, I'm assuming that when you say the word "parser" you mean the automated parser (invoked via the Parser menu item) and not the manual parser (which can be used when interlinearizing text).

Second, with the FLEx parser, the size of morphemes is actually irrelevant. 

Third, this, then, means that there is some other reason why the "elel" morpheme is not being posited.

The best thing to do is to use the Try a Word tool (Parser menu item / Try a Word), type the word in the "Word to try field", check the box by the "Let me see all the detailed steps the parser tried" item and also check the box by the "Let me break the word into its morphemes" item.  You can then select the morphemes you expect the parser to find (use the same method you do when interlinearizing by hand).  When you've done that and press the Try it button, the parser will try to parse the word using the morphemes you told it to use.  The Try a Word dialog also lets you see the steps the parser took and tries to give a reason for why a particular path it took failed.

Doing this should give you some good hints as to why the parser is not finding the "elel" morpheme.

I'm afraid the best the rest of us can do is only make guesses (e.g. the "elel" suffix is not in the right slot in a template or some such).

--Andy