(From Lognet 90/3)
Grammar 75 has now been created and the version of LIP based on it (V.2.15) has been released. If you have an earlier version, you may wish to return it to the Institute for upgrading. Please enclose enough for return postage if you have no balance with the Institute, but updates are still free.
It had been overlooked in the past that kekked sentence predicates could not be tensed. This has been corrected. Kekked sentence predicates now behave like other sentence predicates.
Previous versions of LIP rendered all commas invisible in the humanized parse. Since this obscured some real differences in some parses, commas are now shown if they are lexemic. Commas inserted for phrasing or resolution purposes are not shown. In the past, to see the difference between Mu titci vi le supta = We eat in the soup and Mu titci vi, le supta = We eat here the soup, it was necessary to look at either the tree parse or the machine parse. This is no longer necessary. The linear parse now preserves the comma and shows this difference.
Another major change was made in the handling of pauses. Previously some phrasing commas gave misparses because in certain circumstances they could be lexemic and could not be removed by the preparser. Also it was found that some commas that were intended to be lexemic were removed by the preparser. A new approach has enabled the parser to remove all commas that cannot be lexemic and leave all those that can be. This has highlighted one convention which had not been explicit before: a pause which is required for resolution (before a connective or after a name, for example) can therefore never be lexemic. If a lexemic pause (PAUSE) is desired in such positions, its "louder" allolex gu must be used.
Another change is the addition of guu as a specific closer of termsets. (See "The Many Faces of Gu" below). Gu no longer has any specific meaning of its own, but serves as an allolex of PAUSE, GUE, GUI, GUO, and GUU.
Stephen Rice had proposed to the Academy that case tags (the DIO lexeme) should be permitted in linked predicates, as for example in the sentence Da pa godzi le vedma je cau lo nema dalri. Although not explicitly noted in L1, such usage is already grammatical, and therefore does not require an Academy decision. —RAM