Return-Path: Message-Id: Date: Tue, 11 Jun 91 02:42 EDT From: lojbab (Bob LeChevalier) To: lojban-list Status: RO X-From-Space-Date: Tue Jun 11 02:43:12 1991 X-From-Space-Address: lojbab Subject: Indirect Questions and sumti-raising 4th of 6 related messages The following discussion is based on one of several proposals for the place structure of "djuno", which has been thoroughly beat around for several months, and whose answer is tied in to a lot of debate on these proposals. The official place structure is: x1 knows x2 about x3 For a long time it was felt that this was a cleft place structure, and that it should be simply: x1 knows x2 (with the old x3 accessible through sumti raising) The following discussion of indirect questions uses this place structure proposal, but during the evolution of the cleft structures discussion, a new place structure proposal has evolved which more nearly resembles the original. I decided not to change the discussion on indirect questions since the changes themselves were irrelevant to the specifics being discussed, and would make examples more difficult. I return to the place structure to show its implications at the end of the indirect question discussion. The proposed place structure at this writing recognizes that the 'subject' of knowledge may or may not be at a broader level than the knowledge itself. We say we know "geometry" when in fact we know many facts about and methods of geometry. Those facts do not involve "geometry" itself except in a rudimentary way for which sumti raising is too misleading (as is obvious in English with the parallel statements: I know [that] the sum of the three angles of a planar triangle [is ...] I know geometry. We also have proposed adding a 4th place, an epistemology, as we have to all questions of truth and falsity. Thus the current proposal is: x1 knows that abstract statement x2 about subject x3 is true under epistemology x4 4. Indirect Questions: 2 ways of doing it: when the indirect question word is a form of "ma", just use sumti raising: mi djuno tu'a le klama be le zarci I know [some statement about] the goer to the store, [namely, identity] When the question word is not a sumti, use "kau" marker (memory hook: "pau"): mi djuno le du'u la djan. klama le zarci jikau le zdani I know the predication John goes to the store (Connection?) the house holds. I know it! I know whether John goes to the store or to the house. "kau" belongs to UI. John Cowan wrote the preceding, but Nora adds an additional note here. The term 'indirect question' is somewhat of a misnomer; all Indo-European languages overlap use of relative pronouns with question words, and all use these 'wh- words' in 'indirect questions'. Grammarians could also call them 'indirect restrictive clauses', but this would never sell. Nora prefers to see 'indirect questions' as restrictive relative clauses: mi djuno tu'a zo'ekau poi klama be le zarci I know [something about] the something unspecified that goes to the store, I know it! namely identity. or even more preferably as the simpler abstract bridi: mi djuno le du'u zo'ekau klama le zarci I know the statement the something unspecified goes to the store. I know it! When the question word is not a sumti, you can use "kau" as a marker (memory hook: "pau"): mi djuno le du'u la djan. klama le zarci jakau le zdani I know the predication John goes to the store or the house holds. I know which! I know whether John goes to the store or to the house. John also left out the third case of indirect question, which is the other interpretation of the last English example. This variety is more easily handled: mi djuno tu'a le jei I know [something about] the truth-value of la djan. klama le zarci ja le zdani John goes to the store or the house, [namely the value]. I know whether John goes to the store or to the house. Note that under the current proposed place structure for "djuno", the examples regarding sumti raising would have to be redone, since you could solve the problem more easily using the 'cleft' x3 subject. mi djuno fi le klama be le zarci I know (something) about the goer to the store, [namely, identity]. ---- lojbab = Bob LeChevalier, President, The Logical Language Group, Inc. 2904 Beau Lane, Fairfax VA 22031-1303 USA 703-385-0273 lojbab@snark.thyrsus.com