Difference between revisions of "slinku'i test"

From Lojban
Jump to navigation Jump to search
Line 1: Line 1:
The ''slinku'i test'' is test used to determine whether a potential [[type 4 fu'ivla]] is a morphologically valid. Words that fail the test are called [[slinku'i]], as the word "slinku'i" fails the test.
+
The '''''slinku'i''' test'' is test used to determine whether a potential [[type 4 fu'ivla]] is a morphologically valid. Words that fail the test are called [[slinku'i]], as the word "slinku'i" fails the test.
 +
 
 +
<noinclude>
 +
{{se inspekte/en}}
 +
[[File:Slinky.jpg|thumb|slinkui]]
 +
The "[[slinku'i test]]" ('''le valslinku'i cipra''' in Lojban) refers to the test for words that can become ambiguous if certain syllables come before them. This rule applies mostly to [[type 4 fu'ivla]], which have to be very careful not to cause any morphological conflicts. For example, if you say '''*pa slinku'i''', it becomes the lujvo '''paslinku'i''' ("robe-chain-caretaker").
 +
 
 +
==History==
 +
This word was originally born in [[Loglan]] where it had the somewhat more reasonable meaning "past-line-related" and was glossed "ancestor" (a typical Loglan style lujvo). A conquering hypothesis is that '''*slinku'i''' was originally proposed in Loglan as a fu'ivla to refer to Slinky, a springy toy<ref>https://en.wikipedia.org/wiki/Slinky</ref>. The possible meaning of this word doesn't make any difference since fu'ivla don't <u>have to</u> have a source language, though it helps.
 +
 
 +
 
 +
Currently in Lojban {{jvs|slinkui}} denotes ''slinky toy'' (note that '''slinkui''' is a valid word unlike '''*slinku'i''').
 +
 
 +
==What is the easiest way to determine slinku'i failure?==
 +
* Remove one or more consonants from the beginning of the word. If the result is something that could end a lujvo, then the original word is a slinku'i.
 +
* Put an arbitrary CV sequence (where C is any consonant and V is any vowel) on the front and see if the result looks like a lujvo.
 +
* Ask a [[ilmentufa|parser]] if the word is valid. Parsers that implement the correct morphology will show that the word is invalid.
 +
 
 +
For example, removing the 's' from '''slinku'i''' yields '''linku'i''', which is a valid lujvo. Since a lujvo can be composed of a rafsi plus another lujvo, '''slinku'i''' fails the test. This same process can be applied to determine that the word '''tcalfe''' fails the test: if we remove the 't', then we are left with '''calfe''', an undefined gismu, and since gismu can end lujvo (e.g. the lujvo '''terdatni'''), '''tcalfe''' fails the test.
 +
 
 +
If we insert an arbitrary CV on the front of either '''slinku'i''' or '''tcalfe''', then the result is a lujvo. For example, '''paslinku'i''' and '''metcalfe'''.
 +
 
 +
==Is it certain that CV cmavo are the only prefixes which can cause slinku'i failure?==
 +
*[[John Cowan|John Cowan]]
 +
*:Nothing of this kind is certain until a full morphology algorithm is implemented and exhaustively debugged.  Any takers?
 +
*[[mark]]:
 +
*:It just occurred to me.  Wouldn't '''paslinku'i''' be invalid on the grounds of failing the [[tosmabru|tosmabru]] test?  In which case, wouldn't '''slinku'i''' thus be okay?
 +
**[[rab.spir|rab.spir]]:
 +
**:That would be taking the [[Loglan|Loglan]] approach (if the first consonant cluster of the lujvo is a permissible initial consonant cluster, they stick in a '''y'''). However, extending [[tosmabru|tosmabru]] to include [[fu'ivla|fu'ivla]] makes a very large number of lujvo fail the tosmabru test, and this adds an arbitrary rule you need to memorize in order to make lujvo, especially since exactly which word forms can be fu'ivla is not entirely certain. In Lojban, if there's a conflict between a fu'ivla form and anything else, the fu'ivla loses.
 +
 
 +
==References==
 +
<references />__NOTOC__
 +
</noinclude>

Revision as of 08:58, 29 August 2014

The slinku'i test is test used to determine whether a potential type 4 fu'ivla is a morphologically valid. Words that fail the test are called slinku'i, as the word "slinku'i" fails the test.


slinkui

The "slinku'i test" (le valslinku'i cipra in Lojban) refers to the test for words that can become ambiguous if certain syllables come before them. This rule applies mostly to type 4 fu'ivla, which have to be very careful not to cause any morphological conflicts. For example, if you say *pa slinku'i, it becomes the lujvo paslinku'i ("robe-chain-caretaker").

History

This word was originally born in Loglan where it had the somewhat more reasonable meaning "past-line-related" and was glossed "ancestor" (a typical Loglan style lujvo). A conquering hypothesis is that *slinku'i was originally proposed in Loglan as a fu'ivla to refer to Slinky, a springy toy[1]. The possible meaning of this word doesn't make any difference since fu'ivla don't have to have a source language, though it helps.


Currently in Lojban slinkui denotes slinky toy (note that slinkui is a valid word unlike *slinku'i).

What is the easiest way to determine slinku'i failure?

  • Remove one or more consonants from the beginning of the word. If the result is something that could end a lujvo, then the original word is a slinku'i.
  • Put an arbitrary CV sequence (where C is any consonant and V is any vowel) on the front and see if the result looks like a lujvo.
  • Ask a parser if the word is valid. Parsers that implement the correct morphology will show that the word is invalid.

For example, removing the 's' from slinku'i yields linku'i, which is a valid lujvo. Since a lujvo can be composed of a rafsi plus another lujvo, slinku'i fails the test. This same process can be applied to determine that the word tcalfe fails the test: if we remove the 't', then we are left with calfe, an undefined gismu, and since gismu can end lujvo (e.g. the lujvo terdatni), tcalfe fails the test.

If we insert an arbitrary CV on the front of either slinku'i or tcalfe, then the result is a lujvo. For example, paslinku'i and metcalfe.

Is it certain that CV cmavo are the only prefixes which can cause slinku'i failure?

  • John Cowan
    Nothing of this kind is certain until a full morphology algorithm is implemented and exhaustively debugged. Any takers?
  • mark:
    It just occurred to me. Wouldn't paslinku'i be invalid on the grounds of failing the tosmabru test? In which case, wouldn't slinku'i thus be okay?
    • rab.spir:
      That would be taking the Loglan approach (if the first consonant cluster of the lujvo is a permissible initial consonant cluster, they stick in a y). However, extending tosmabru to include fu'ivla makes a very large number of lujvo fail the tosmabru test, and this adds an arbitrary rule you need to memorize in order to make lujvo, especially since exactly which word forms can be fu'ivla is not entirely certain. In Lojban, if there's a conflict between a fu'ivla form and anything else, the fu'ivla loses.

References