lojbo karni number 10
Jump to navigation
Jump to search
Copyright, 1989, 1991, by the Logical Language Group, Inc. 2904 Beau Lane, Fairfax VA 22031-1303 USA Phone (703) 385-0273 [email protected] All rights reserved. Permission to copy granted subject to your verification that this is the latest version of this document, that your distribution be for the promotion of Lojban, that there is no charge for the product, and that this copyright notice is included intact in the copy. le lojbo karni Number 10 - August 1989 Published by: The Logical Language Group, Inc. 2904 Beau Lane, Fairfax VA 22031 USA (703)385-0273 Editor and President: Bob LeChevalier This is the quarterly news and product announcements newsletter of The Logical Language Group, Inc., known in these pages as la lojbangirz. We are a non- profit organization formed for the purpose of completing and spreading the logical human language "Lojban". This issue (LK10) is being mailed at publication to about 500 subscribers, about 10% growth since last issue. Press run this issue, 650. Loglan 1 Re-published Book Review on Page 3 This is our most news-filled issue yet. There is a lot of things happening. Are you a part of them? Welcome to New Lojbanists - page 1 Your Mailing Label, About Voluntary Balances, Last Minute Notes - page 2 Book Review of Loglan 1, Editorial on L1 - What Does It Mean to la lojbangirz? - page 3 Other Institute News, On Sailing - and a Correction, la lojbangirz. files to cancel 'Loglan' trademark - page 7 Report on LogFest 89 and the 1989 Annual Business Meeting - page 9 Research and Development - Lojban Parser Status, Grammar Changes Proposed, Baseline Change - page 9 Growth and Publicity - Athelstan's Trip Cancelled; Worldcon Details; New York Visit; Other Visits En Route from Worldcon?; Lojban Video Plans; Other Publicity News - page 11 Education - DC Lojban Class Nears Completion; Blacksburg Class Making Excellent Progress; Other Classes; Self-Teaching Lojbanists; Textbook Status- page 13 International News - page 15 Products and Prices - page 16 Mid-Year Financial Report - Finances Improve, But Your Help is Still Vitally Needed - page 16 Future Plans - Textbook, After the Textbook, Dictionary, Reader, Classes, Grants - page 17 Response to a Letter on Dictionary Plans - from Adrian Bolt- page 18 Request for Historical Anecdotes and Information - page 21 Contents of Ju'i Lobypli #10 - page 21 Note: References to 'Loglan' in this text, unless specifically noted, do not relate to the 'trademark' claimed by The Loglan Institute, Inc., or to products described by that 'trademark'. Welcome to New Lojbanists Over 50 new Lojbanists have been added since the publication of LK9, with most due to our ad in Discover magazine, and to reviews in Factsheet Five and recruiting at the UNICON science fiction convention. Also, word of mouth is finally starting to aid us, as a significant percentage of new people are joining us through individuals telling others about this GREAT, NEW, LANGUAGE! Keep it up; we'll happily supply copies of the brochure on request. Our rate of growth is accelerating; we hope to add more than 100 new people at Worldcon alone. People don't write letters 'the way they used to in them olden days', and therefore, after our first mailing to each of you, we tend to wait weeks, months, and occasionally a couple of years before we hear from you again. In a rapidly evolving project like this one, we'd rather take the chance of mailing this newsletter to all respondents, than to be forever trying to bring people up-to-date on what has happened since they last heard from us. We think that 2 you read our newsletters, and that we are not junk mail that is relegated to the trash upon arrival. If what we are writing is of no interest to you, please let us know and we'll drop you from our mailing list. New people are assigned to level 0 unless we hear otherwise from you; what this means is that we send you the Overview of Lojban and the latest le lojbo karni newsletter, along with a brochure if you haven't received one. We also send an order form and registration form so you can let us know what else you might be interested in. See the descriptions of mailing codes below and reports on ongoing activities, and write to us about any activities you are interested in. We want to hear from you. Your Mailing Label Your mailing label reports to you your current mailing status, and your current voluntary balance including this issue. Please notify us if you wish to be in a different mailing code category. Balances reflect contributions received thru 19 August 1989. Mailing codes (and approximate annual balance needs) are defined as follows: Level B - Product Announcements OnlyLevel 0 - le lojbo karni only - $5 balance requested Level 1 - le lojbo karni and Ju'i Lobypli - $15 balance requested Level 2 - Level 1 materials and baselined/final products - $20 balance requested Level 3 - Level 2 materials and draft textbook lesson materials as developed - $50 balance or more R indicates that you are receiving materials on a review basis pending some publicity we hope you will give us. If your publication can reimburse us for our costs, great, but it is not mandatory. Feel free to call or write to ask about your balance account or mailing code assignment. You are scheduled to receive Ju'i Lobypli if the level code on the top line of your mailing label is "1", "2", or "3", but not if it is "0". If you are listed as mailing code "0": YOU MUST WRITE TO US IF YOU WANT TO RECEIVE Ju'i Lobypli. About Voluntary Balances Our orientation is non-profit. Almost every dollar we receive goes directly into producing the products that we send out, with a surprisingly small overhead and no paid salaries. We subsist entirely on your contributions against our costs of mailing to you. However, a large number of our respondents are college students and others with low incomes, who want our materials but couldn't afford them at our costs. Therefore, we operate on a voluntary balance system. We ask you to contribute what you can towards your balance, and maybe to make a donation to help cover those who can't afford theirs. Only 40% of you are making significant contributions now, and we need to raise this percentage - a small minority of you are supporting the rest. Please contribute, if only a little, against our costs. Ideally, try to bring your balance up to 0, or even to a little surplus. I suspect that most people who receive this newsletter get more reading out of it than they do from a typical news-weekly or other periodicals that require you to pay in advance, and you don't have to wade through advertisements. Some generous donations since last issue staved off the economic wolves for a few more months, but publishing the textbook will be expensive and we haven't found the money yet. Please help! We seek donations in addition to and independent of your contributions to voluntary balances. We have filed with the IRS for Section 501(c)(3) status, approval of which will officially allow your donations (not contributions to your voluntary balance) to be tax-deductible. We hope to have approval by the end of the year. We are operating in accordance with that section, which means that your donations now should be deductible if approval is obtained later, although there is always the possibility of disapproval. We will inform all donors at the end of the year as to the status of deductibility of their gifts. We also note for all potential donors that our bylaws require that no more than 3 30% of our expenses be for administration and legal fees, and that you are welcome to make gifts conditional upon our meeting this requirement. Last Minute Notes There is still a computer bulletin board conference reserved for Loglan and Lojban on the AMRAD BBS here in the DC area. It has, unfortunately, seen little use. I will check it more often if people start leaving messages on it; we at one time had some pretty lively discussions going on the CLBB BBS. The phone number is 703-734-1387, and it is PC PURSUIT-able. On a related matter, USENET, UUCP, and INTERNET mail should now be sent to me via Darren Stalder at [email protected]. Darren will shortly have the brochure and the gismu list available for on-line transmission via the net. Contact him to get either. Mike Gunderloy reported in his new issue of Factsheet Five that there is a cheaper yet reasonably fast way to send international mail. I may not find out soon enough to try it this mailing, but if he's correct, our international prices can drop. Book Review of Loglan 1 by Athelstan, and Bob LeChevalier Earlier editions of Loglan 1 (4th edition, The Loglan Institute, Inc., 1989) have shown us only a shadow of a language promised. In this most recent edition, Dr. James Cooke Brown says of his brainchild, "The long period of patient engineering is over; the [language] is ready to be used." Unfor- tunately, he's wrong. We had hoped to report to Lojbanists and all other members of the Loglan community that Dr. Brown's work was complete and exemplary, a well-worked description of his Loglan language version and source book of ideas for others working with logical languages. Sadly, the errors in this edition of L1 are so variable in type and scope and so great in number that we cannot describe it as anything other than a work in progress. The promise lies unfulfilled; the shadow still remains. Don't misread this as being unabashedly negative. There is a lot of brilliant work on the language embedded in the pages of the new book. But there are so many problems and questions that are not answered; so many contradictions and simply vague explanations, that you already need to understand most of what Brown's telling you in order to find the gemstones and separate them from the dross. As an example of the intermittent genius reflected in the book, there is an entirely new section on testing the Sapir-Whorf hypothesis. The theoretical discussion here is well worth reading, clear and thorough. While there are seeds of a useful approach therein, the resulting experimental design is badly flawed - he doesn't address obvious problems such as introducing bias in the teaching of the language, and the difficulty of encouraging serious, experimen- tally useful, long-term effort on the part of disinterested subjects with no long-term stake in the results. Similarly, the new scientific borrowing system is both comprehensive and impressive. However, it contains pervasive cultural biases in word choice and in the linguistic implications of the phonology, as well as algorithmic errors that reveal incomplete analysis. As described, it won't work. Brown's numerous errors embarrass the Loglan community that is heavily credited with the changes in this edition, as well as the several qualified scholars who purportedly read the text and helped Brown "find the errors". The responsibility for the errors lies with Brown himself; he chose not to publicly debate each proposed change and did not circulate a draft of this book among knowledgeable loglanists, resulting in the draft quality of the published prod- uct. Of course, only Brown knows the entirety of his current language version, making review of the language design by others especially difficult. Unfortunately, the obvious hard work, additions, and improvements that went into this edition do not guarantee overall quality. While we would like to skip over the clerical errors of form and detail and concentrate on problems of 4 concept and method (the rightful ground of criticism), the former are so prolific in form and number as to obscure the latter. Some discussions are so confused or contradictory, that it is impossible to tell whether the errors therein are minor or conceptual. In evaluating a book of this complexity, we must use both internal standards (how well the book meets the goals for which it was intended) and external ones (how the book measures by the common standards of technical prose). As to the book's purpose, let us judge Brown by his own standards: Brown's preface claims that "... Loglan 1 must, as I say, stand alone. It must not only serve the user as a resource book on the grammar, morphology, and usages of the new language ...". Let us first examine large scope errors dealing with the text as a language description. The book is incomplete; it lacks the information needed to evaluate many of his claims, including a formal description of the grammar and complete definitions of each structure word. Such a formal description is promised for a later book, Loglan 6; this does not satisfy, especially in light of Brown's statements in Chapter 6 that he expects the language to continue to change and evolve. One can only respond by asking: At what point will Brown be able to present both a complete formal description of the language and a current corresponding explanation of the same language version? An example of this problem. This version of Loglan is described as a logical language. Yet no proof is offered of its logical completeness, nor of the unambiguity of its grammar (the latter is claimed "conflict-free", a term left undefined). A full grammar description is needed for those who might wish to examine these issues, which are critical to Brown's claims for the language. Another example. Brown claims that his product is a complete language, yet almost all samples of the language are single sentences demonstrating individual concepts. By comparison, other attempts to develop new languages have tried to demonstrate their workmanship and versatility in a 'chrestomathy' - a collection of a wide variety of literary passages from several sources, and often from several cultures, showing how a wide variety of styles can be expressed. Brown's supplemental corpus contains only four 200-250 word extracts from English-language scientific articles, all from the same magazine; his translations consist heavily of scientific word-borrowings and have little vari- ety of grammar. Cultural neutrality is claimed for the language, but several changes since the 3rd edition have significantly weakened this claim. The addition of 'h' as a major consonant might be tolerable, but its ubiquitous use to create consonant clusters in borrowings renders these words subject to severe mispronunciation and misrecognition when used by native speakers of different languages. A Latinate bias is given by using Spanish (a language derived from Latin) as the check language in determining which Latin borrowings are international in nature; a reasonable alternative would have been to use languages from several families to determine international nature of a word, and to verify that some form of the root is recognized in the 'scientific languages': French, German, and Russian (as well as English). Numerous biases remain that were criticized in earlier versions of the language and the book, such as in the use of "madzo" (make...from...) as the ba- sis for causal metaphors, as in "mormao" (dead-make = kill). Also, while Brown has ostentatiously incorporated gender-less pronouns and excluded gender-based grammar, his language examples, are sexist (e.g., he uses the words for "man" and "woman" in examples throughout the book where gender is unimportant, instead of the gender-less words for "human" and "person"; or the gender-less occupa- tions "teacher", "worker" would have served as well). This is to be a language for man-machine interaction, but numerous inconsistencies in its rules and algorithms for name-building, complex-building and pronunciation will prevent progress towards this end. Linguistic errors and descriptive conflicts have made themselves known even in the pronunciation guides. There is no index; for a scientific work of this size, this is a sore omission. In addition, many references in the text are missing from the 5 bibliography, making follow-up difficult. The organization of the book, which appears to build from simple structures to complex ones, is good for teaching or for reading straight through from front to back. For a reference work, it is poor, especially since Brown often diverges from the main point of a section, explaining related concepts that a reference reader would expect to find in different sections. The allocation of material on the word structures of the language between chapter 2 (on pronunciation and word forms) and chapter 6 (word-making) is incomprehensible. Only by reading the two chapters together, skipping the intervening text on the grammar, can one see the complete picture Brown intends to present; but then the contradictions between the two chapters stand out more strongly. Adding to the confused organization, several cross-references were not updated to match the new edition; one is led all over the book trying to find related information. The heavy use of old references and spotty mention of recent linguistic works suggests that Brown has not maintained the current linguistic scholarship needed in a project of this scope; for example, his footnote and bibliographic ref- erences regarding case grammar, language universals, logic, formal semantics, phonetics, and natural language grammar date almost entirely from before 1970. To scholars, the book is a disappointment. Brown often does not defend his ideas, but merely states them. For example, he states his disagreement with Chomsky's transformational grammar, thereby alienating many American linguists, but uses only his own unpublished experience as a basis for his claims, saying that he has not observed evidence supporting Chomsky's positions. Since Brown references none of Chomsky's works from later than 1963, it isn't clear that Brown knows those works enough to criticize them. Brown evidences little scientific detachment. He develops some truly marvelous notions at various points in his exposition, but often appears too self-impressed with their virtuosity. Examples include his discussions of unambiguous syntax within metaphor; his recurring references to his theories about metaphor, thought, and the evolution of language (which may be interesting and even valid, but are not yet generally accepted and in any case are irrelevant to his language description); and the enormous significance he at- taches to having a few people haltingly speaking a now-obsolete version of the language back in 1977. He compares this latter feat favorably with the fluent speech of 30,000 to 100,000 Esperantists (these numbers are disputably low unless they refer only to fully bilingual speakers); Esperantists claim as many as 1 million speakers in China alone. Finally, while Loglan has many features that optionally accommodate structures and ideas from a variety of natural languages, one gets tired of being reminded of this in every section. In a language description alone, Brown's repetitive congratulation of the language design is merely distracting; but since the book also purports to be a scientific work, this tone is unseemly and detracting. Scientists are likely to be less than impressed when someone like Brown seems forced to 'blow his own horn'. Concluding the evaluation of the book as a language description, we note that inconsistencies range from the spelling and place structures of his word lists as compared with each other and the text, to apparent grammatical conflicts that cannot be resolved without the missing formal grammar. Clearly, the book fails to properly describe the language in a manner that makes it useful for the typical reader. By external standards of quality for technical prose, Brown fares badly. Though Brown's enthusiastic though pedantic style (which does at times capture the reader totally) occasionally obscures them, all manner of errors can be easily found in both his language design and his writing. Let us categorize these errors by type with a few examples: - Editorial nits that affect understanding - including duplicate footnote numbers and a type-face that makes it almost impossible to find footnotes in the text, and the cross-reference errors mentioned above. - Major Editorial Flaws - including inconsistencies in Loglan words and place structures, the lack of an index, the omission of bibliographic entries, and an outline that divides closely-related material between Chapters 2 and 6 in a hap- hazard way (leading to many contradictions). 6 - Pervasive Lack of Clarity and Inconsistencies in his Rules and Practices, suggesting errors of fact or analysis - There are numerous errors in his algorithms, and several places where his examples violate rules found elsewhere, while many of his examples are so contrived that they fail to suggest actual usage. His pronunciation guides are vague and self-contradictory, especially his use of /eigh/ as in "mayo" for 'e' before vowels (cf. /ay/ for 'ei'), two contrary pronunciations of 'oi', and a mis-identification of the sound of his 'w' as /eu/ instead of the standard /ue/. - Errors of Omission - We've mentioned the missing formal grammar description, but at least Brown tells you it is not there. Looking at his list of structure words in the Appendix, though, there are several words that we haven't found discussed in the text. A reader constantly is looking for some piece of information or another that can't be found in the section being read. It may be elsewhere, but will go unnoticed until too late without an index. - Errors of Scientific Fact and Linguistic Principle - such as his description of /x/ as a "voiced h", haphazard mixing of voiced and unvoiced consonants in clusters with no allowance for assimilation, and the statistical comparison of Loglan speech to Esperanto. Brown's intermittent linguistic errors throughout the book suggest that his general knowledge of linguistics is uneven and spo- radic, making all of his claims that touch on linguistic theory suspect. - Failure in Currency - Beyond general linguistic error, the use of old references suggests that much of the linguistic knowledge Brown draws on is outdated; as an example, Brown cites only Fillmore's original paper on case grammar as a basis for the development of an optional case system, suggesting that Brown is unaware of the enormous developments in case theory since then. A reviewer of an earlier edition (Zwicky, 1969) called some of Brown's ideas "ante-diluvian". He may have been right. - Failures of Analysis - his theories of potentiality of tense-less usage are either poorly explained or they break down with rudimentary examples (by his explanation, it would appear that "da mrenu" translated as "he is a man" is true for females as well as males of most any age - since this is a potentiality achievable through a specific steps, i.e. a sex-change operation). Of course, Brown ignores his theory in almost every tense-free translation except in the one section in which it is discussed. Brown also inconsistently applies a philosophy for interpreting metaphors that ignores place structures of the components, and commits errors and incon- sistencies in set and mass description. Brown reveals careless change without proper analysis in his non-intuitive (so he admits) right-grouping of vowels which was changed only after looking only at a skewed set of scientific bor- rowings that end in 'i'. Similarly, his new rules for joining borrowings can result in complex words that break up in more than one way ("protonynerji" can mean "proton-energy" or "produce-twist-energy"), and he now permitting "la" at the beginning of names in the face of obvious counter-examples ("Pi,e'r, LaPla's, mrenu" spoken aloud, can mean either "Pierre! LaPlace is a man" or "Pierre LaPlace! Be a man!"). - Bombast - Brown's self-impressed tone, his attitudes towards defending his academic scholarship, his statements about "Institute Policy" regarding word- making and approval, his miniscule scientific corpus as a demonstration a comprehensive and usable language, and the formal grammar that he is maintaining as a trade secret; these reveal a presumption that scientists and other users of his ideas (and those of the other contributors to the language) must (and will) abide by his rules and attitudes. - Major Conceptual Errors - including his culturally biased use of 'h' to solve borrowing problems, his use of a Latinate language for identifying International Scientific Vocabulary roots, and his inadequate analysis of the Sapir-Whorf testing problem. Brown fails one other standard, that of the Loglan Project goals he first set forth in his June, 1960 Scientific American article. He has abandoned audiovisual isomorphism (the use of "rr" and haphazard use of pauses are examples), incorporated major cultural biases, and perhaps most importantly, taught that word-making in a predicate language is primarily a morphological problem, rather than a serious question of place structures and semantics. 7 Although Dr. Brown has put a lot of hard work into this book and into the language, the result bears the signs of an extreme rush to publish. This is sad. Dr. Brown's good ideas are lost amidst the confusion. To summarize: Imagine the Mona Lisa - beautiful hair, eyes and enigmatic smile - but with a crooked nose and blackened teeth, and structures in the background that are out of proportion. It wouldn't be a masterpiece; a scholar would recognize the promise, but the errors would be too glaring to accept the painting as a work of art. In the same way, the L1 errors are too great for the scope he intends. If he had written a small book that didn't intend such comprehensive coverage, some of these could be ignored. But in a work of this scope, they can't be. If an artist can't, or won't, learn perspective, no version of the painting will ever be a great work of art. Dr. Brown clearly intended this work to be his Magnum Opus. Alas, he achieved only a big work. Important! The preceding is an abbreviated analysis. Detailed supporting criticisms are in a longer review in preparation, which will be available on request. We are also soliciting additional and alternative viewpoints; we will collect these and make them publicly available as well. Editorial on L1 - What Does it Mean to la lojbangirz. by Bob LeChevalier The older, 3rd edition of L1, a product of the years of thought and experience that JCB has invested in the Loglan project, has been invaluable in explaining the ideas and philosophies behind the various language features. Regardless of the fact that I've believed the Institute version of Loglan to be obsolete, I've found several people who have benefitted from examining those ideas and philoso- phies as a key to understanding. An updated version, documenting the results of another decade of work by the community, would indeed have been a valuable re- source for furthering the discussion of the language. Unfortunately, the 4th edition is so flawed that I cannot use or recommend it for this purpose. It would obviously have been better for the Institute and its version of Loglan had the 4th edition been well-done, but it would also have been better for Lojban. Both efforts would benefit from a clear statement of underlying principles and philosophy such as L1 was apparently intended to be. It would have allowed the two versions to be compared on their merits. I believe that Lojban has enormous improvements over the Institute's various versions of Loglan in all of the technical areas criticized, and the result would have demonstrated this to everyone concerned. (Incidentally, our two versions of Loglan differ primarily in word choice, which is why the philosophy and ideas of L1 are valuable. Most of the Institute test corpus, by simple word substitution, would parse the same way by our grammar - though possibly not in reverse: we have made some (mostly subtle) changes and improvements. Although I wouldn't willingly choose it, the strife of the past few years has benefitted us by forcing us to rethink many ideas that would have gone un-analyzed, which is why Lojban is so much an improvement over the old language versions, rather than merely a set of word substitutions.) But L1 is now published, and is honestly so bad that I want to cry for JCB. His pride and the undoubtedly large financial investment that he has made in publishing and printing costs probably make it impossible for him to withdraw the book from publication, accept comments from us and others, and re-issue it; this would be the best overall for the project. I want to cry for us, the community, too. Because whether the book sells or fails, JCB will feel forced to press on to vindicate himself, not realizing that his satisfactory place in history is better assured based on the first 20 years of the project, when virtually alone he brought a germ of an idea to flourishing life. I've compared Loglan to a child, which JCB raised to adulthood, but refused to allow to leave the nest. A better analogy is that the Father of Loglan set forth the seed of the language, and the last 15 years have been the gestation 8 leading to the current true birth of the language. If we can work productively with JCB, we will take pride upon completion of a hard labor, and together enjoy watching and helping the child grow up. I feel that la lojbangirz., its members and its correspondents, as members of the Loglan community, have the right and the responsibility to participate in the peer evaluation of such a significant work as L1. We are working separately on a common scientific project, and progress requires a certain degree of impartial cooperation. (We do not subscribe to the idea that la lojbangirz. is a business competitor of The Loglan Institute.) In fact, the need to respond to a work such as L1 is greater because the poor overall quality of the work, if unanswered, will reflect on the community as a whole. Half of our subscribers date from before I split with the Institute, and many of these have sat back and waited. Some have told me that you are waiting to see which language will win out. Others have been waiting, out of respect for JCB, for this long-promised "Going Public Again". We urge those who have been waiting for the 'GPA' to act now; it is time to make your evaluations. If you have access to a copy of L1, and would like to see the full text of our critique (of which the above review is a brief summary), please write or call and we will send it to you. Now is the time to get involved. If you have time and or money, you can support both of our efforts; it doesn't hurt us, especially if you've been inactive. la lojbangirz. (and Lojban) is strong enough to survive and prosper even with two ongoing scientific efforts. There is some consideration of splitting limited resources among these separate efforts - a single united effort would progress more quickly - but any increase in overall loglandic activity helps move the project along. Of course, if you are disposed to choosing between the two efforts, I very strongly urge you to choose la lojbangirz. We are growing in activity and in opportunities to use Lojban to interact with others. We will continue to move into the future, beyond L1. Please help the child grow up. Other Institute News In addition to L1, JCB has issued a price list of other products included as part of his 'GPA', or 'Going Public Again'. The products include several promised for years to be part of GPA, as well as others that appear to have been added based on our offerings or announced plans. (Imitation is the sincerest form of flattery.) As noted above in the review, the details of his design will remain a trade secret until publication of Loglan 6, which he indicates may be at least a year away. (The trade secret claim comes from a letter that went only to 'aficionados' who signed an earlier secrecy agreement.) Institute products include updates to Glen Haydon's old MacTeach programs. (JCB promised perpetual free updates to buyers of this program, but is now charging $10 and you have to send your old disk in.) He is also offering 'MacTeach 2b' for the MacIntosh; from its description, this is believed to be essentially identical to our Mac Lojflash except for the word lists - though JCB is inexplicably charging $5 more in price. Since this program allows you to type your own word lists in, we'll gladly help save you the money. JCB has also announced a new version of LIPtm (a unneeded trademark that we don't object to). The formal grammar is not included with this purchase and is hidden from users, making this product relatively useless; you don't know what might be valid input, and can't easily or usefully interpret failures. (If you, like us, are familiar with YACC parsers and with JCB's earlier grammars, you can probably eventually find all these details by systematic experimentation, but it isn't worth the effort.) A random sentence generator is apparently inspired by Nora's program. No details are given. The Institute is cutting its price for membership in half, to $25/per year. Benefits include the occasional publication 'Lognet'. The Institute also has announced a dictionary for this winter, but indicates that this publication date is heavily dependent on getting volunteers. It is believed that the dictionary format will be similar to that used in the word lists for L1, a minimal format that provides no semantics and abbreviated place structures that are defined only by ambiguous prepositions. 9 Lastly, JCB discusses the possibility of providing 5-day training workshop, for which you would pay air-fare, supply your own local transportation, and pay an additional $300-$500. These are for 'prepared loglanists', so he probably expects you to have previously bought and used L1, MacTeach, LIP, and other products, before attending. (You can save a lot of money if you can come to DC; I suspect that by winter this household will be a full-time Lojban workshop, and we'll be happy to trade teaching for your help.) On Sailing - and a Correction JCB, in a personal note to Nora (not to me), asked for me to correct a 'false impression' that I gave last issue. I reported that I had heard that JCB had gone sailing for a month, and that I had no news of L1's publication. JCB claims to have sent it to the printer before leaving on this trip, and that my statement implied that he had not kept his promise to not sail before the book was done. Nora's order, requested in mid-May, was not shipped until the last day of July. To me, a book that is sold by pre-order isn't 'done' until those orders are being shipped (the product offering indicated a shipping date shortly after the beginning of June). To explain my LK9 announcement, on the day I was doing the final printout of LK9, I received a call from my lawyer. (Those who are new to la lojbangirz. need to know that JCB had threatened us with legal action over our use of the name 'Loglan' in March 1988. As reported in LK8, Jeff Prothero - who was also threatened - and I submitted a proposed settlement of our disputes with the Institute in February.) This call from my lawyer reported that JCB had written to his lawyer, rejecting our proposal without comment or counter-proposal. JCB further said that he was going sailing for at least a month and would not be reachable. I tried to confirm the situation with the #2 person in the Institute, Glen Haydon; he had not heard from JCB in months and knew nothing about L1, (though he is listed as a critical reader). Faced with no solid information, I simply reported what I knew, figuring that those who knew of JCB's promise not to sail before the book was done, would infer, like I did, that he had kept his promise. The following day, I was able to find someone to whom JCB had reported the completion of L1, but LK9 had already gone to press. The actual product announcement came out a week later. To anyone who felt misled by my statements as to JCB's performance of his promise, I apologize. I urge that JCB be more forthright in the future about what is going on, so that future misunderstandings are avoided. We have always committed to publishing news from the Institute in full, and JCB has our continued promise to keep you informed, should he submit such news to us. la lojbangirz. files to cancel 'Loglan' trademark (For those unfamiliar with it, the history of our dispute with The Loglan Institute, Inc. and with JCB, is primarily one of intellectual property and control of the fate of the language. Does the language belong to the inventor who conceived of it, the organization he founded, to the hundreds of people who have contributed ideas, work, and money over the years since, or to the people who use the language? Our position is that all of these share in 'owning' Loglan in its various versions. JCB disagrees. We can supply additional details of our positions upon request, including documentary evidence and legal background.) As indicated in the last article, JCB waited 2 1/2 months to respond, and then gave no evidence of intent to negotiate an end to our disagreements over the use of the word 'Loglan' (which he has claimed as an Institute trademark for 'dictionaries and grammars', not a language.) Since JCB was going sailing for at least a month, and his lawyer was going to be out of the country for a month after that, we interpreted the events of the last few months as an attempt to stall a settlement, possibly until after L1 came out. Faced with no prospect for a quick negotiated settlement, the la lojbangirz. Board of Directors decided that we could afford to delay no longer, and we filed a Petition for Cancellation of the Institute trademark with the US Patent and 10 Trademark Office in late May, claiming grounds that "the registration was fraudulently procured, the mark is the common descriptive mark, or is merely descriptive of the goods, and/or that the mark has been abandoned" through non- continuous commercial use, encouragement of derivative terms such as "loglanist" and "loglandic", failure to police use of the term by others over a several year period, and statements that may have placed the entire language irrevocably in the public domain. The claim of fraud is that JCB knowingly included incomplete and false data on his application, and that he apparently had no authorization from the Institute Board of Directors to file the application. la lojbangirz. has a stronger case that the trademark is damaging to its commercial interests (the basis for trademark rights) than either Prothero or I have as individuals; la lojbangirz. is trying to promote Lojban and occasionally needs to refer to the historical name of the project to describe our efforts and to reach the millions who may have heard of the language under that name. The drawback of filing, though, is that we have directly involved the organization in the dispute, and have subjected our expenses to the 30% adminis- trative/legal limit called for in our bylaws. This isn't expected to be a problem. We will avoid allowing legal fees to draw resources away from pro- ductive uses; the Board also approved a resolution allowing only donations specifically earmarked for legal fees to be used for legal expenses. Jeff Prothero and I have committed to making such donations as are needed, in addition to our regular support of the group. But we also remember that, if this dispute were resolved, these funds could go to much more productive use. We will continue to hope that JCB decides to resume negotiations and to settle this dispute quickly, in the positive and cooperative spirit of our editorial above. The trademark cancellation process is a long, drawn-out one that will distract and drain both of our efforts, though JCB will suffer more from this since he has done less preparatory work than we have. We've avoided discussing the legal issue in these pages; most of you aren't that interested. The issue is vital to our rapid growth, yet doesn't actually threaten to stop Lojban from going forward; there is no threat to us primarily because we have chosen to fight (if we hadn't, JCB could have taken ever more aggressive positions until we either fought or gave up our effort completely; we believe JCB has been using this tactic since before the trademark battle started, including in earlier disputes with his own Board of Directors and with Jeff Prothero, Nora, and I, over copyrights). Moreover, there are principles of intellectual freedom at stake, principles we cannot afford to abandon. We'll report future developments, hopefully positive ones; the legal dispute will, however, take second place to our primary effort to make Lojban successful. The la lojbangirz. Board welcomes your comments and questions on these issues, if you are interested. Otherwise, enough of these distractions! Report on LogFest 89 and the 1989 Annual Business Meeting LogFest 89 was at least as successful as earlier meetings, though we had a drop in attendance. Instead, those who came were more strongly committed to and knowledgeable of the language. In addition, except for the DC class, almost all attendees were from out-of-town, enhancing LogFest's image as a non-local event. In contrast, previous years with higher attendance included more local people who came out of curiosity to find out what was happening (which was good - some of these, such as Athelstan, have taken their place among the la lojbangirz. leadership) Main events included a lengthy discussion of testing the Sapir-Whorf hypothesis, led by Ralph Dumain, Dr. John Parks-Clifford (pc), and Athelstan. We'll report on these discussions in JL10. The DC class put on the 'Cinderelwood' skit written by Nora and published in JL9; it is the longest piece of Lojban ever written (some of Jim Carter's works in old Loglan were longer; we will soon be updating these to the current language with Jim C.'s blessing). The performance was as humorous as the skit; we didn't have much rehearsal time. But everyone had fun, and even those who had not started learning the language were able to follow what was going on between words and actions - finding much of the Lojban understandable in such a context. 11 More significantly, old Loglan apprentice Brad Lowry and new Lojbanist Elliot Deal came down from Philadelphia and recorded much of the activities for the Lojban mini-video, which we hope to first present at Worldcon. Every attendee was given the chance to 'tell the world' what attracted them to Lojban, and we got a lot of good footage. There were several side meetings (including discussions of place structure changes and a couple of baseline changes - see the R&D section below - as well as mini-tutorials, and just plain gab sessions among people with many common interests. Art Wieners brought his Lojban-speaking computer down (I don't think it's named Sam), and people played with teaching programs, the random sentence generator, and the test version of the parser. Athelstan suffered through the weekend with severe poison ivy, but was able to participate, more or less, in all of the activities. We also held the annual business meeting of la lojbangirz. at LogFest, at which Athelstan and fellow DC-class member Albion Zeglin were added to the Board. Athelstan has been frequently mentioned in these pages for his contributory work; Albion joined us in January and wants to teach Lojban to children as one of his goals. We also added John Hodges as a voting member of la lojbangirz. (In a sense, all of you are members, but we have to maintain an official entity.) John, who has single-handedly put together the successful Blacksburg VA class since May, declined to serve on the Board, since he is non- local and can be difficult to reach by telephone. Other Board members and officers were re-elected. I remain as President, pc as Vice President, and Nora as Secretary/Treasurer. Jeff Taylor, who is leading the parser work, and Tommy Whitlock, who did much of the research in building our gismu list, also remain as regular Board members. All of us are interested in your opinions on both technical and organizational matters. We will listen to you! (Letters to me are shared with other Board members, but I will supply address to anyone who wants to write to them directly.) Next year the annual meeting will be moved to Sunday morning, when it will interfere less with the main activities, which are to work with and use the language. I've also committed to improve the air conditioning for next year. Several dozen people generate a lot of heat (hot air?), and Washington DC isn't noted for being cool in June. (I'm also going to move the TV downstairs to give us more distraction-free meeting space.) Mark it on your calendar NOW. The 3rd weekend in June 1990. We expect to have much larger attendance, and some entirely-in-Lojban sessions, next year. We hope to see you there. Research and Development Grammar Changes Proposed - Last issue reported that a number of minor issues had cropped up in the grammar. As we've gone through the exercise of textbook writing, we've discovered that little things that we thought were covered by the machine grammar didn't work. Many of these problems were holdovers caused by our basing part of our machine grammar on ideas in the old Institute grammar. That grammar was developed using a corpus of individual test sentences devised to reflect how JCB thought the language should work. The sentences were 'constructed', and not natural. It has long been easy to find examples from JCB's writings that would not parse the way his grammar suggested. (There is at least one sentence in the new L1 that seems very unlikely to parse, though we of course don't have JCB's grammar to check this with.) As Lojban usage has grown, we've felt constant pressure to expand the realm of possible Lojban sentences. For example, just the other day in a letter from John Hodges reporting on the Blacksburg class, he closed the letter "*?co'o lemi re pendo ku". (Goodbye! My two friends...; the asterisk means that the text is currently ungrammatical, but the question mark indicates that the grammatical status may change.) Under the current grammar, this is invalid - vocatives take only names, incomplete descriptions, and anaphora (pronouns). In the example above, John has used a vocative with a complete description sumti, which we don't now allow. There is only one way to express what he tried under the current grammar, the stilted-to-English-speakers "co'o re pendo po mi" (Goodbye, two friends of 12 mine!). However, given John's example, it was easy to come up with something needed but disallowed by the current grammar: "*co'o. bab. .e noras.". There is no valid way to put an 'and' in a vocative, and especially one based on names. Thus, we added provision that a vocative can take any sumti, although le and la and their cmavo relatives will be left off names and descriptions when simple vocatives are used. The result is "*?co'o la bab. .e la noras." There are other examples of things that needed changing, mostly oversights. We had never added a rule allowing a bare "coi" vocative with no name, although probably every Lojbanist who has used the language has used this vocative. Several months ago, pc had proposed a re-examination of our elidible terminator set in hopes of making it easier to teach and use. When examining the grammar, I came up with some simplifications to the grammar with only a little work. Of the proposed changes, it is these elidible terminator changes that most affect students in the early lessons. Most significant is that "kei" will only be used to terminate lexeme NU abstractions; "ke'e" will be used to terminate "ke" constructs. (This makes the text definition of "kei" in the LogFlash and gismu lists incorrect, and will require us to eventually add an entry for "ke'e" with a rafsi, but these are minor changes that will probably not be implemented until the textbook is done.) Besides the terminator changes, almost every other change is an expansion of what is permissibly said in Lojban. Finally, events demanded a change. Athelstan decided to USE the language creatively, translating a short story by Saki. Trying to capture Saki's literary style, Athelstan quickly found several constructs that seemed perfectly reasonable, but which weren't allowed. They were easy to add, but we now need a new version of the parser to be able to test his translations. It isn't fair to expect people to learn the language if we change things haphazardly, even though the grammar is not yet finalized. We decided last spring that, except for emergency problems and changes that don't really affect use(that haven't happened), we'll use our post-baseline procedures, especially seeking input from active students of the language who would be affected by the change. We decided to seek input from everyone who we know to have finished Lesson 3 of the grammar, based on your actually having USED the language and written something to us. Proposed changes have virtually no effect until Chapter 3, and only a fraction of the areas affected are covered by the end of Lesson 6. A description of the grammar changes, often fairly minor and technical, simply won't mean much to anyone who hasn't studied that far (although we'll send a copy to anyone who asks; those who can read and understand the machine grammar - and have a copy - may be able to work through the changes). Since I was writing up a proposal, I collected together all the fine-tunings of the last few months. The DC class spent a session going over the grammar behind the changes, and the changes themselves, and has agreed the changes are worth the minor relearning effort. They also made some useful suggestions for improving the proposal which were added. By the time you get this, people will already be reviewing the changes. Jeff Taylor is trying to complete a parser with the changes before we go to Worldcon, but in any case before we get back. We are asking for review comments to be returned within a few weeks, and approval of the changes with any required modifications will probably occur by mid-September, which is when we are plan- ning to send out JL10. Upon approval, we will send a description of the changes as they affect the published lessons to all level 3 people (and others who have received textbook lessons that are affected). We will also send updated machine grammars to all those of level 3 who have the current machine grammar, as well as to level 2 people with positive balances and the machine grammar. (The existing machine grammar is quite usable for understanding the basic grammar architecture, so we will continue to send copies to those who order them.) There are corresponding minor changes to the complete cmavo list as well, so we've delayed sending out a 1 or 2 page update to that list until after the approval. Hopefully, this will be the last set of grammar changes before the textbook is completed (early 1990). At that point, we'll probably have a few more, and then the review period will start. We now anticipate a grammar freeze will be 13 decided at the 1990 LogFest, to take effect when the 6 month post-textbook review period ends if no additional problems have surfaced. Parser Status - Other than the grammar changes, the test version of the parser has proven quite useful in testing people's early Lojbanic writings. We haven't solved the problems mentioned last issue, but the grammar change proposal works around the worst problems. Jeff Prothero has done some more work with PLOP, the alternative recursive descent parser. PLOP does not have a YACC parser's problems with certain combinations of elidibles, since it can backtrack and correct its confusion, but recursive descent is much slower on long text. (Jeff Taylor's parser can process a full page of Lojban text in less than 10 seconds, much faster than a human can read it.) We hope to merge Jeff Taylor's lexer with PLOP after the new grammar change is incorporated. I'm hoping that we can start selling one or both parsers by LK11 in November. (PLOP is also public domain, and Jeff Prothero will send a copy to anyone who asks, though it has minimal use until the lexer is added. I'll supply his address to anyone on request.) Baseline Changes - LogFest approved exactly one change to the gismu list baseline, which is now a full year old, and that change is minor. The keyword for "mukti" will be changed from "purpose" to "motive" to eliminate confusion about the meaning of the word. Something can be a 'motive' without being a 'purpose', and quite often the purpose resulting from a motive is expressed as part of the x2 place of "mukti". Since, as may be obvious, the word "mukti" actually derives from "motive", the change should be easy to assimilate. We will not be changing the published gismu lists at this point, nor LogFlash. The approved change will be made in the textbook, and the other publications will be updated in a few months when the textbook is completed. Those with flash cards can make the change manually if you choose, as can those willing to use a text editor on the data file for Mac LojFlash. Other Research News - JL9 suggested a proposed experiment in measuring the actual learnability of Lojban words and whether there is an correlation with recognition scores. We need volunteers who will commit to using a special version of the LogFlash program DAILY (or as close as possible) until they have gone through the words at least once and preferably twice. You should not have used LogFlash before, and we need to know whether you have used any other means to study either the Lojban words or the old Loglan words. On pc's request, we are also planning an experiment with the next few classes that start, asking them to take a standardized test that will allow us to learn more about Lojban's possible effect on thought. This isn't a Sapir-Whorf test, but it is a precursor that will help researchers decide how a Sapir-Whorf effect might be detected. In another arena, Athelstan is making translations from a variety of works in different styles from English, and will shortly be trying some non-English materials - probably Old Norse and ancient Greek. These translations will demonstrate Lojban's versatility as a language. Athelstan is attempting to capture the subtle stylistic variations in each language and author as part of the translation effort, thereby putting our grammar through its paces. A few changes to place structures were discussed and agreed upon at LogFest; we are seeking additional inputs in this area. We have also made a few changes without organized discussion while writing textbook lessons; the LogFest changes were those that involved some question as to which concept actually was intended to be represented by the gismu. I would estimate that about 30-40 place structures have changed in the course of the one year since the gismu list was frozen. Since we have not frozen the place structures, this seems like excellent stability. Nora completed her version of a lujvo-making program just before publication. We've found that the proposed scoring algorithm is not optimal; she doesn't like the choices it makes. We will probably test variations on the algorithm over the next month or two, while ensuring that the program works. DC and Blacksburg students may play with a teaching version of the program. It should be available for others with MS-DOS computers by LK11. Growth and Publicity 14 Athelstan's Trip Cancelled - As mentioned above, Athelstan contracted a severe case of poison ivy the day before LogFest. He was able to attend LogFest, but was laid up in bed for most of the next week. This allowed other problems to crop up, and the result was that he cancelled his trip to Los Angeles and various intermediate stops in June and July. We tried to contact the several people who had volunteered to organize meetings or supply sleeping space, so we hope that noone was left waiting with- out knowing what happened. Your response was great! The stimulus of the trip also served to get some signs of community activity started in Seattle, Portland, LA, Dallas, and Salt Lake City, and we are hoping to see this increase. Worldcon Details - We've finalized our plans for the World Science Fiction Convention - Noreascon III in Boston around Labor Day. Nora, Athelstan, and I will be leaving on the night of 30 August; Tommy Whitlock may also be coming with us, or may travel separately due to work schedule limitations. We hope to stop briefly in Troy NY on the morning of Thursday the 31st to visit Lojbanist Mike Gunderloy, whose Factsheet Five review magazine has been responsible for about 50 new Lojbanists, making him our top recruiter. (Thanks, Mike!) We also may stop to meet with Bob Chassell in western Massachusetts, especially if we can't meet with Mike. In any event, we should arrive in Boston on Thursday afternoon. We will be staying at the '57 Plaza Hotel', and Bostonians and convention attendees should be able to leave messages for us there. The room is in my name (Bob LeChevalier). We are scheduled for four presentations during the weekend, designed to attract a variety of new people to Lojban, and to educate the attendees about our project goals. Athelstan is taking the lead on making these presentations, instead of me, as he is a better public speaker. Nora and I will handle questions; we may also have local Boston Lojbanists and others who are attending help out in the presentations. On Friday night, we will have a general introductory talk. On Saturday night, we will discuss Lojban and Computers, in- cluding machine translation and AI applications, as well as what we've learned about computer-aided instruction. On Sunday morning, Athelstan will conduct a Lojban mini-lesson similar to the one he conducted successfully at UNICON last month (see "Other Publicity News" below). On Sunday afternoon, in the spirit of an internationally-oriented convention, we will discuss International Aspects of Lojban, including the efforts towards cultural neutrality, Sapir-Whorf testing, the international language movement, and our efforts and problems in spreading the word about Lojban in other countries. We hope to recruit many non-USA active Lojbanists at Worldcon. We will have a table in the main traffic hall, and access to side meeting rooms where Athelstan may give additional mini-lessons as required. We hope to keep this table staffed during as much of the convention as possible in order to maximize our educational and recruiting opportunities; if it is ready, we hope to be running the Lojban mini-video to introduce people to the language (and to save our voices). Based on our experience at smaller conventions, we may add more than 100 Lojbanists to our community. As time permits, we may meet outside the convention with local Boston-area Lojbanists. Nothing formal has yet been arranged, but Lojbanist Chris Moriondo is taking the lead on local arrangements. Call him at 508-481-9986 (home) or 508-870-8927 (work) to find out details. Our best hopes for a larger group meeting are on Monday night and Tuesday night, although Monday is Labor Day and we may therefore have attendance problems. We can also hold smaller meetings in various parts of the Boston metro area on Monday and Tuesday with groups or individuals. You can contact us directly at my home number on page 1, by leaving a message at the hotel, or talking to Chris M. We would like to meet as many Boston Lojbanists as possible (and maybe get you started learning the language - Athelstan's mini-lesson takes less than an hour and you will be able to say some things in Lojban on your own afterwards.) A letter will be sent to each Boston-are Lojbanist the week before Worldcon, to remind you, give an update on plans, and hopefully increase your attendance. New York Visit; Other Stops En Route from Worldcon? - We plan to leave the Boston area Tuesday or Wednesday, travelling back via Rhode Island and Con- necticut. Wednesday and Thursday (5 & 6 September) will be spent in the New 15 York City area, where we hope to have meetings with local Lojbanists. Thursday night, we hope to participate in a local science fiction meeting that several Lojbanists attend. We also may spend Friday in the northern New Jersey area. As noted in "Education" below, we believe that interest in the New York City area has risen to the level where a class is possible. As with Boston, we plan on sending out letters a week before we leave, giving our plans. The following people are contacts for our New York visit: Deb Wunder (Richmond Hill) 718-846-2014 Nancy Thalblum (Bronx) 212-543-8295 Art Wieners (NJ) 201-271-1483(h) and 201-949-2784(w) - Art may be hard to reach at home. We plan on spending part of the weekend of 8-9 September with Nora's family, and can possibly meet with Philadelphia-area Lojbanists then. We have no organizer in Philadelphia, but we will try to contact some of you while there. If you are interested in meeting with us, try calling us here before we leave, or leave a message for us with our New York City contacts. Lojban Video Plans - As mentioned in the LogFest report, Lojbanists Brad Lowry and Elliot Deal recorded several hours of LogFest activities. They work professionally in the video field, and hope to have a short loop videotape edited from the recordings in time for LogFest. This is a major boon for us; even a few minutes of quality video would normally cost us several hundred dol- lars, and the video will be very helpful at conventions where we have to say the same things to hundreds of people. Depending on costs, we may be able to have copies for others who want to make presentations at conventions. Contact me after Worldcon for details. Hopefully the recordings will eventually be used to make a longer video suitable for distribution to cable and public TV stations for broadcast. Decisions on this will be made after the shorter video is done, and we know how well it serves our audience. Other Publicity News - We attended the UNICON science fiction convention in Maryland in mid-July. In addition to our normal introductory talk, Athelstan gave an introductory mini-lesson to 8 people who showed up for the 10:30 PM event. This was a nice amount for an experiment at a 500 person convention, especially if you extrapolate to a convention of Worldcon size (6000+ atten- dees. All 8 people completed the session an hour able to speak simple sentences in the language on their own. All 8 also expressed interest in learning the language, and have received lesson materials. Three of these live in the New York City area, which is why we are especially interested in getting a class started now. Thanks to Mike Gunderloy, we have been contacted by a Washington Post reporter who plans a full-length feature about Lojban. This feature will likely be distributed over the Washington Post/LA Times news service to over 300 other newspapers. We are hoping for this article to appear around the end of August. Keep an eye out; if your local paper ever prints articles from this service, you may give them a call and ask them to look for the article - knowing there is a local person interested may increase their interest in including it. Mike G. has also put us in touch with the Whole Earth Review, and we are hoping to be reviewed therein; they contacted me just before this publication. Mike is also publishing an ad for us in the next issue of Factsheet Five, in addition to his regular reviews. See "International News" below for status on non-English versions of the brochure. The only bad news in the publicity front is that Analog was not interested in an article at this time, although Editor Stanley Schmidt suggested that we write an explanatory 'letter to the editor' explaining what we are doing. This will be a priority after Worldcon. All of this going on suggests that we are about to receive a rapid spurt of growth. As classes are offered in more and more areas, this growth will accelerate. It is still reasonable to hope that we will top 1000 people by the end of 1989, 100% growth in 4 months. Education 16 DC Lojban Class Nears Completion (An Invitation to a Party - with Strings Attached) - The first DC-area class is now 7 months old. We intended to finish before LogFest but didn't really come close; I can't write the textbook as fast as I had hoped. Therefore the class took a month-and-a-half off, and has resumed, though in a somewhat reduced form, since a couple of class members can no longer regularly attend, and will finish the course in self-study. We have had active attendance from 4 people besides Nora and myself, who are now only a bit ahead of the class, primarily in vocabulary. There are 2 others who are possibly going to resume attending, making a total active class size of 8. Because the textbook-writing hasn't kept pace with the class, we have moved on ahead, covering large portions of the grammar very quickly. We have found that the basics of the grammar have been covered by the end of Lesson 6, and most of the rest of the class is learning 'bells and whistles' that give additional op- tions, solve problems in special contexts, and of course, giving practice and confidence in using the languages. The class now begins each session with conversation practice, and everyone has been pleasantly surprised by how much they can say and understand. We reorganized the topics being taught, in order to allow the class to effectively review the grammar change proposals. Given two weeks off for the Worldcon trip, the class will finish in mid-Oc- tober, although Sunday night Lojban conversations may continue indefinitely. As a 'final exam', the class is being asked to take at least a full page of some text that interests them, and translate it. The other half of the test is an evening party, tentatively scheduled for the weekend of 5 November, in which ONLY LOJBAN WILL BE SPOKEN. ALL LOJBANISTS WHO ARE WILLING TO ABIDE BY THE CONSTRAINTS ARE INVITED, even if you aren't in the class. We are especially inviting the Blacksburg class, if they can come up, and any other self-studiers or class members. (We may have the party on the holiday weekend of 15 October to enable people from out-of-town to come, but you must let us know soon.) Con- tact us in late September or early October for final details. People coming from out-of-town can stay with us. There will be some time for socializing - optionally in English - before the party officially starts. Blacksburg Class Making Excellent Progress - The Blacksburg class may well be ready to participate in the party. 5 students, led by John Hodges, are meeting on a when-convenient basis, and working through the lessons together. At last report they were finishing Lesson 3 and should be into Lesson 4 by now, although some students are already using more advanced concepts. Their efforts have demonstrated that the self-study class concept works. Already, every member of the class is writing perfectly grammatical Lojban sentences of a sophistication at least sufficient for normal conversation on simple subjects. When their vocabulary grows, they may outpace the learning rate of the DC class. John H., the nominal 'class leader', has had trouble staying ahead of the others, but is challenged rather than bothered. The Blacksburg group is making a major contribution to the textbook writing effort; their problems have told us what needs to be changed to make the book usable in this type of situation. With the DC class ahead of the textbook, they are now in the primary feedback role for future lessons. Other Classes - As indicated above, we are hoping to kick off one or more classes in the New York area, following the emergence of a team of people who can collectively lead the class and an infrastructure that could help the Lojban community grow quickly. It turns out that, of the 40 people in NYC, as many as half are associated with each other through a local science fiction fan group; Lojbanists Marc Glasser and Donna Camp are among the leaders of this group. Classes still haven't started elsewhere, though Boston should get a big boost towards doing so from new recruits at Worldcon, coupled with our visit. Brooke Albert is no longer able to lead a class, at least for a while. Linguist and language instructor Chuck Barton remains willing to help in teaching a class, but not by himself. We're hoping that Chris Moriondo, listed under the Worldcon plans above, will have the time to help in this way, but other volunteers are welcome to step forward. There are over 50 Lojbanists in the Boston metro area. Los Angeles, San Francisco, Portland, and Seattle all seem just on the point of a class, but haven't started to organize. Contact your local leader, listed in LK9 (or write us and we'll put you in touch). A correction on our LA organizer, Rory Hinnen: his phone number is actually 818-796-8096 (Pasadena). 17 It takes only two to have a self-study class together, and there are two such in Salt Lake City. John Hodges even built the Blacksburg class by himself, recruiting the other 4 to study with him. Lojban sells itself, and there are lots of people who will respond if you tell them what it is. We are already talking about the next DC-area class. Athelstan will probably be teaching it, with us helping him when needed. It may start as early as a couple of weeks after the 1st class ends, in November, or we may wait until after the New Year, when the holidays won't interfere - though we'll ask people who know they are going to participate to get a head start on learning the vocabulary. Self-Teaching Lojbanists - Jamie Bechtel, among many self-studiers, has completed Lesson 6, and has sent his Lesson 6 writing to us - the first original Lojban science fiction story. We hope to have it in JL10. He also attended LogFest, where he was approximately at the same skill level as the DC-class. Some 55 people total now have received textbook lessons; it is presumed that many, if not all of these, are actually studying them. We'd like more feedback than we've gotten. The best feedback is usage, and 14 Lojbanists have given us reasonably grammatical 'running' Lojban text (connected thoughts longer than one sentence). A few who haven't have gotten commentary to help them in usage with which they are having difficulty. Over 100 people now have a version of LogFlash, and another 22 have Mac LojFlash. This is not counting people who have gotten copies from others and haven't told us. About 20 people have flash cards. Our self-studiers range across the globe, from Anchorage AK to Austria, and shortly to Italy as well. We have 3 non-native English speakers learning the language, including one German, one Italian, and one Estonian (which is non- Indo-European like Finnish, although heavily Europeanized). Textbook Status - Only 6 lessons of the textbook have been completed, as I've had to handle never-ending distractions for the last few months, and was ill for a while from the stress. Things are moving again and I'm well into Lesson 7. There will be an update lesson written, based on the grammar change proposal, when it is approved. Now that people have been learning the language, we are thinking about the final draft of the textbook. We've had enough feedback to know a few things: - many more examples and exercises are needed, especially in the early lessons; - there are too many unconnected points taught in Lesson 1, confusing people; - while the basic idea of the Lojban bridi gets across to people in Lesson 3, it is supposed to be understood in Lesson 1 and 2. - the final textbook will start out like Athelstan's mini-lesson, and get you saying things in the language early. The delay of conversation until Lesson 10 was a mistake. People want to be able to say things earlier; it helps their motivation, and in a classroom setting speeds up their learning the vocabulary. Unlike the theorists who say that it is easier to get people to write since there is no pressure, the DC class reminded us that Americans are so out-of-the- habit of writing that they find it harder to write a paragraph on their own than to produce a sentence aloud; - discussion of terminators, especially "cu", needs a lot of clarification; so does the entire discussion of elision in Lesson 4; - an embarassing error - we recently noticed that sometime just before creating the gismu baseline, we changed the meaning of rafsi "kun". We must either change it back, or the word "kunbri" that has been used in many of our teaching materials must be changed - as it is, it doeesn't mean the intended "empty-predicate". Most likely, it will become the x2 place of bridi, and the word for the concept that has been labelled "kunbri" will become "selbri". (Comments welcome!) - a few topics taught in later lessons need to be moved earlier, including specification with "be" and "bei", simple possessives, number brivla, using "e" between sumti, and using "je" in tanru (the last derives from observations made by JCB in the revised L1, which point out a common misunderstanding about tanru that cannot be discussed without "je"; - most importantly, people either aren't able or willing to maintain the pace we asked for in vocabulary-building. This is unfortunate because it is the critical path to fluent speech. Even now, most of the time spent in conversation in the DC class is spent in looking up words because no one has 18 made it into Maintenance mode in the vocabulary. At even 10 words per day, they should be through the words twice, but people aren't doing vocabulary drill daily. You cannot ever really learn a language without at least this minimal level of daily activity, but we can't allow the student/reader to be held up by this. We will reorder the gismu list to fit the new outline and to set a slower pace. The next version of LogFlash will have other changes, cutting the error practices from 6 repetitions to 3; related changes will speed up the first Gaining Control pass through the words. We've learned that for most people, learning the words really starts when you get into Maintenance mode. When you've seen all the words once, you no longer have the long New Word lessons, and can go through the words twice as quickly. The first time through, it seems the best approach is to get exposed to as many words as possible as quickly as possible, learning the ones you easily can. More important though is to learn the range of the vocabulary so that you can use the English to Lojban gismu lists effectively; it is embarrassing to have to say 'I didn't know that was in the gismu list'. For most people, that first time through Gaining Control simply takes too long to be effective. Serious students can do 40-50 words a day, but it takes at least an hour a day; few people have made that kind of commitment, even though the time to master the words reduces to only a couple of months as a result. There may be additional changes as we hear from people who are using the draft lessons. Thus, the next version of the textbook will be much different than the current one. As our way of showing thanks to you who have started early, Nora and I are committing to donating funds to return half of what you've spent on draft lesson materials as a credit on the textbook provided that 1) you have sent some writings or comments to us showing that you've actually used the mate- rials, and 2) you have a positive voluntary balance (or are reasonably close) at the time when we start taking orders. International News As indicated above, we now have people actively learning the language in Austria, and shortly in Italy. Patrick Bar, a native German speaker, recently moved to Austria. He is attempting to organize European Lojbanists and may have contacted you, if you are one. New Lojbanist Silvia Romanelli, who lives in northwestern Italy has also indicated that she wants to learn the language, as has said that she will try to find others nearby to study with her. Her mate- rials are en route. These two are translating the 'What is Lojban? la lojban. mo' brochure into their respective native languages. Quebecois French Andr Bilodeau is well along towards a French translation of the brochure. We have others who are helping check the translation and making sure that it is equally usable in France; there have been few problems thus far. We have 41 people from outside the US on our mailing list, including 17 in Canada, and 20 in Europe. Farthest away in both geographic and language distance is a new Lojbanist from Thailand who we're hoping will be able to start learning the language soon. We hope to expand these numbers at Worldcon, which has several hundred international attendees. The commitments demonstrated by those who have started learning the language will add to our credibility in gaining new people, as will our efforts to develop non-English materials like the brochure translations. Athelstan and Tommy Whitlock are already making plans to attend the 1990 Worldcon in Amsterdam, by which time we may have several European speakers of the language. Another class member, Sylvia Rutiser, may be visiting the Netherlands in the near future, and will try to contact our Dutch Lojbanists. She doesn't yet know the dates for her trip. DC class student Paul Francis O'Sullivan wants to translate the Lojban teaching materials (not just the brochure) into Esperanto. For any of you overseas who have trouble with the language in any of our publications and teaching materials, feel free to write and let us know, and tell us of what languages you can read better than English. We have access to speakers of many languages who can help in explaining technical or other 19 confusing terms. We can probably accept questions in your native language, if you let us arrange for a translator here ahead of time. We will do whatever we can to help you with our materials. We are investigating the possibility of grants to support our European efforts from a foundation sponsored by the West German government. We may be able to get full German teaching materials after the textbook is completed, especially if those of you in Germany let us know of your interest. On other matters, we are trying to obtain the ability to accept payment via Master Card and Visa (US people will be able to use this too). This apparently is the best way to send money internationally. Most checks sent internationally (though apparently not Canadian checks) can cost as much as US$30 to cash, which isn't worth it for the amounts you are sending. Most Lojbanists in other countries send US currency, wrapped in extra paper so it isn't visible to someone looking. We can also get your country's currency converted to US money, if this is easier. We know that it is difficult to send money, and we are committed to supporting you even if you cannot afford to do so, or if the procedures are too difficult. We want people outside the US to learn Lojban, and will not let a lack of money prevent a serious student from getting teaching materials. NOTE: An International Postal Reply Coupon (IRC) is only worth US$0.40 (40 cents). We've been told that it costs much more overseas, so that this is NOT a good way to send money. Products and Prices We have no new products to announce this time, although there are or will be updates to the cmavo list and the machine grammar shortly, as described above. There also will be several new products around the end of the year, but lets wait till LK11 for them. Most of you getting this haven't read what we've put out; you don't need more stuff. Lojban tapes have been put off yet again, since I'm redoing lesson 1. We've got some others who can speak the language besides Nora and me, but no script. I hope to be able to announce at least one sample tape by the time of next issue. What we do have now is a new order form, one that hopefully summarizes our offerings in a way which is useful to you. We want you to know where to begin, what to order first, and the order in which to read things. Our experience has shown us for that some things I've written, if you read them without the background that I expect, are unintelligible. Lojbanists have such a broad range of backgrounds that nothing I write is going to satisfy everyone; I therefore need to bring people to a common point before proceeding. Hence the suggested order on the order form. The price list shows our new pricing policy; I've tried to reflect our actual costs better, yet make our order system as automatic as possible. The more time I spend on each of your orders, the less time I have available to spend on the newsletters, the textbook, and in answering your letters. Note that the draft cmavo list and machine grammar are lower in price now. These can be useful, but we recommend that you at least study the first 6 lessons first; as these other documents have less explanation and almost no ex- amples. We still would like more feedback on our products. Are we giving you your money's worth (or ours)? Can we be doing anything any better? We want your opinion. Mid-Year Financial Report Time for our quarterly beg for money. Our finances have improved since last issue, but not nearly enough. Your help is still vitally needed if we are to keep our heads above water. We received income of well over $1000 in May, and about $2000 in June. We've grown, however, and our expenses are now running at $1100 per month. We seem to consistently spend about $20.00 per person per year, no matter what we try to do about expenses. As the project has grown, I've spent more telephone time 20 consulting with pc on grammar points, and with others attempting to coordinate activities around the US. We haven't yet gotten to textbook publishing, which will cost us several thousands of dollars. We need to be in better financial state in order to publish. Do you believe in us? Are you getting as much out of our publications as we're spending on producing them. If so, then please help, at least by keeping your balance positive! Here is an interim report: a summary of our finances this year up until 15 August. Voluntary Balance Contributions $5828.55 Donations - Legal Fund$928.07 Unrestricted Donations$2328.10 Total Donations $3256.17 Total Income $9084.72 Non-Administrative Expenses$7037.00 Legal Expenses $828.07 Non-Legal Administrative Expenses $471.52 Total Expenses $8336.59 Administrative/Legal Percentage 16% 1989 Net Income $748.14 Miscellaneous Totals: Positive Balances Refundable$2567.53 Positive Balances $4226.73 Actual Cash on Hand $2045.82 You can see from these totals that we have less money than we owe. Either the deficit must be made up from our inventory, which is of uncertain value, or we are technically bankrupt. Only donations, and people paying off negative bal- ances, can improve this total. People with positive balances who loan us additional money improve our cash flow, which is important in the short term; however, it doesn't help our long term solvency, since we have to pay them back eventually. Negative Balances ($7845.63) Write-offs due to Dropouts($620.02) Donations Since Formation$8534.80 Our donations serve almost entirely to cover our negative balances, and therefore aren't free to be used to finance new endeavors. Totals By Level, % With Positive Balances, % With Some Balance Contributions, % Donating : Level # %Posit %Some$ %Donate 3 54 42% 77% 33% 2 150 44% 77% 17% 1 27 30% 59% 11% 0 261 3% 8% 1% R 5 - - Of the Level 0 people, 77 are recent new additions who have never responded, and 115 are people who have been receiving our publications for over a year and have never responded. Another 15 of you sent in a couple of dollars when you first contacted us, but have never heard from you again. Are any of you out there and interested? We need to hear from you! As you can see, those of you in Level 0, getting only LK, are rarely contributing anything towards your balances. Yet yours is the least that we ex- 21 pend - less than $10 per year. We aren't asking that much of you: to cover this expenditure on your behalf. Please help! Future Plans Here is the quarterly speculation about our schedule and plans for the next year or two, based on what we know, and what we want to do. I haven't proven especially accurate; Nora calls me an incurable optimist. But I still believe these plans are achievable, if you let us know you want them, and keep helping out. But, in deference to the fact that she's usually been right, I'm going to spend more time on what we plan to do, and less time on when we plan to do it. Textbook - As described above, I'm going to have to do a significant amount of rewriting of the first several lessons. My productivity will probably be low until the first class ends. By the volume that I'm writing per lesson, I expect that the book will be divided into 3 volumes of 20 total lessons. The first volume will be about equivalent to the first 8-9 lessons, the second will have the remaining lessons. The third will contain Appendices like word lists of various types: gismu lists, rafsi lists, cmavo lists, and perhaps some lujvo and borrowings (le'avla), as well as the machine grammar, synopsis, handy tables for reference of the most frequently used constructs, a glossary, and an index for all 3 volumes. In short, the 3rd volume will be the one you carry around for reference when you've finished studying the lessons. When? - probably in late winter or early spring, allowing time for another round of reviews; we don't intend to be rushed into repeating the L1 mistake. We'd like the first volume to be a complete draft by January, for use by the 2nd DC-area class. Tapes - I hope to start producing tapes around Christmas, designed to go with the new draft of the textbook. Beyond that, no predictions as to how many or when. Reader - By summer, we should have a collection of readings in Lojban - some original, and some translations - to give people a chance to practice what they've learned. Teaching Programs - We have people working on converting LogFlash PC to the MAC, the Amiga, to the Apple II, and to UNIX-compatible C. Of course, different people have committed to these conversions several times over the past few years, so any promise of when has no credibility. I'd like to think that our current wave of volunteers is more likely to come through, but no promises. We should have an updated LogFlash PC around January to go with new word lists, and incorporating the changes we described above to make it more likely that people will keep going once they start, at least until they've gotten through the words a couple of times. The place structures in the new program data will be much more readable. The instrumentation required for testing learnability must be added to a version of the program. We need volunteers for this experiment; the main advantage you will get is that your LogFlash copy is free. I think this version only volunteers for that experiment will get this version, which may be slightly slower in operating than our commercial version. A version of LogFlash that teaches place structures and lujvo may be available in the spring, as well as a version that teaches the cmavo and grammar. Constructing data files takes a lot of time, and we also want to minimize what you have to type in for each drill item. We will have a new update of the random sentence generator reflecting the latest change proposal to the grammar and cmavo list changes in a couple of months. We need to revise the textbook outline first, so that we know what is to be included at each lesson level. As mentioned above, we may shortly have a program that teaches lujvo-making and recognition. It will not be like LogFlash, but may well be usable in tandem with LogFlash 2, which teaches rafsi. Flash cards to reflect the new word lists may take longer; we will not reprint flash cards until the current sets are sold; the changes will primarily be in the order, not in the cards themselves, so people getting cards after the word lists are redone will get a list of changes that they can make to the lesson numbers on each card in order to order them correctly. Word Lists - The word lists in volume 3 of the textbook will NOT be a dictionary. There simply won't have been enough usage to justify a dictionary 22 effort. Dictionaries should be descriptive of actual usage, and not prescriptive in telling you what you have to do, especially in a language de- signed for experiments in linguistics. The description of "klama" in the Overview of Lojban is only a part of what I would expect to include for each gismu in a dictionary. We will, though, by mid-year, have built up a large set of vocabulary word proposals, which we will be able to sort into lists of various types. They will be available on computer media, or in print, and will be put out as much for comment as for actual reference. Dictionary - We'll discuss this more in the next section, but I don't even plan to start on the dictionary for another year. We'll be expanding the word lists for a while, and distributing them in English and Lojban order, but these will be more of a source book, they won't be 'official', 'baselined', or even very complete. We will expect some level of comments and usage before we try the more ambitious dictionary writing project, for which we'll probably seek grant money. Growth - My target is to hit 1000 people by the new year, 2000 by next LogFest, and 5000 or more by the end of 1990. This may be ambitious, and will take your help in recruiting; but this should become easier as more people actually are speaking and writing the language, and as word-of-mouth spreads. If we actually accomplish these goals, some of the dates mentioned above, but probably for the better. Why? At the 1000-person level, our prices will drop as we go to offset printing instead of copying; at the 2000-person level, we may be able to justify part-time help. (There is a lot of overhead in filling mail orders, and we have to reach a much larger size before we can hire a part-time employee to handle them for me, leaving me free to do technical work.) A related note; the number of people we have when the textbook is done will significantly affect the price we can sell it for. If we can justify a printing of 5000 instead of 1000, it will probably save several dollars per copy. 5000 copies is possibly a $50,000 investment, and I'm NOT rich. We'll need commit- ments from you to buy the books and to use the language afterwards, and we'll need a higher balance contribution rate, in order to get a loan to finance the printing. Classes - I'd like to see a dozen classes starting at the beginning of the new year to give the 2nd draft textbook an acid test. I foresee classes in every major metropolitan area by the end of next year, and perhaps even some univer- sity accredited courses. Grants and Research - I'm already looking at the possibility of an academic paper telling the linguistic world why they should be interested in Lojban. It will have to be good to pass peer review; the people who write it won't be aca- demic linguists (though we may be able to get help and credential credibility from pc as a co-author.) Such a paper is critical to getting research grants, as is getting our 501(c)(3) status (which is still under review by the IRS). I'd like to have a paper done this winter, which means publication sometime late next year. This will position us to start applying for grants late next spring (the cycle of grant applications takes 6 months or more). Serious 'funded' academic research (as opposed to the trial research projects proposed in JL9 that will merely back our credibility as a scientific effort) may com- mence around the end of 1990. A couple of our efforts may be financed by grants before then. I mentioned trying to get help for international lesson materials development; I'll start on this shortly. I may also send out feelers after Worldcon to a couple of private foundations that might support us based on our goals, rather than on our academic standing. AI Applications - These are funded primarily by governments, occasionally by industry (but only if a profit-making result seems likely.) The process for government funding is only slightly quicker than the grant process. We have identified an AI-application defense research project that might justify using Lojban, but the US defense budget is being cut, and this is not a fancy weapons system that is likely to attract congressional support. We're certainly not counting our chickens before their hatched. More likely, unless we can find an ongoing project that will take Lojban under its wing, real grant funding or commercial AI research support will take at 23 least a year, and will be dependent on our establishing through YOUR usage of the language that Lojban deserve to be taken seriously as a tool for this research. Response to a Letter on Dictionary Plans Normally, we would answer letters in Ju'i Lobypli, but this letter came from a level 0 subscriber, and addresses the plans covered in the discussion above, so I'm going to presume my answer to be of enough general interest to include here: "le lojbo karni #9, page 12: '... language dictionary, probably taking several years to achieve a final form'. I hadn't realized the gap between the textbook and the dictionary would be so long. Isn't this going to drastically curtail the learning of the language? What happens after I've read the textbook - am I to be stuck at that level of achievement/with that vocabulary until the dictio- nary? If we get to coin our own words from the language description then its going to have to be 100% unambiguous - no mean achievement; I don't envy you. ..." Adrian Bolt It may be difficult to convince English speakers that you don't need a dictionary to learn a language, but after all, you did it as a child before you could read a dictionary. But this begs the question; you are an adult and have more sophisticated needs than a child does. If you look at other languages that build much of their vocabulary via an 'agglutinating' approach, such as German and Chinese, you will find that dictionary usage is much lower among speakers of these languages than among English speakers. German has a secondary problem as a result, that many of the words therefore end up acquiring multiple meanings. Chinese also has this problem but more seriously. Each of its root pronunciations, which correspond to our rafsi, can have several meanings; there may be as many as a couple of dozen different Chinese roots, each represented by a different symbol in print, but pronounced identically. They rely on context to sort out the confusion in spoken Chinese, but they also do a very Lojbanic thing, they add extra roots on as type markers, just as we will add an extra word on to a tanru or lujvo to make our meaning more specifically clear. Thus Chinese "ma" for "horse" will have the root for "animal" attached if there is any likelihood of confusion. The Chinese do not generally use dictionaries at all. When we had a Chinese student visiting us, she looked at our dictionary and found many words that she had never heard or seen before; each local dialect has its own variations of words for a concept, each built from a preferred metaphor. During the first few years of Lojban use, our vocabulary will grow in this manner. The problem will never be too serious; from the start we will be collecting and recording words that are used to establish usage patterns. The Chinese had several thousand years to build their vocabulary, before people first compiled dictionaries. Lojbanists aren't going to be completely at a loss for words before the dictionary is finished. You will have 1300 gismu that have baselined, fixed meanings (a small number of place structures change, but too few to be noticeable). Athelstan estimates that another 7000 concepts of more or less usefulness can be expressed as simple conversions or abstractions of the gismu, and the lujvo made from these unambiguous manipulations will be as firm in meaning as the gismu themselves. Given multiple English synonyms for these concepts, you will probably have 10000 or more English words covered unambiguously, although not to as detailed a specificity of meaning as in English. You will use tanru to restrict those meanings; this is ambiguous, but less so than the English words to which they equate. If tanru are insuffi- ciently unambiguous for your needs, you can use longer, more exact grammar and place structure expressions to convey any tanru relationship unambiguously using only the gismu, or so we believe. We also will have initial lists of proposed lujvo, probably hundreds at the time of the textbook publication, but thousands within a short while. I have 24 some 5000-10000 lujvo proposals on paper, some of which are redundant or contradictory, and I can make up new lujvo at dozens per hour myself. These lists will be proposals only - they may have only tentative place structures, though we will try to address the issue of systematizing the place structure of the typical lujvo; there will always be exceptions that don't fit such a system, but most will, as Jim Carter demonstrated back in 1984. We will publish these lists for review and comment, and for use when needed in translation. For normal conversation and composition, you won't likely bother to use the lists unless you are completely at a loss for a tanru. Perhaps not then - you'll just talk around the subject until your point is made. 'Natural' language use seems to work that way. The lists will be used in the manner of a thesaurus rather than a dictionary. You probably never use a thesaurus in normal conversation (even if you use a dictionary). Why use one in speaking Lojban? When people learn foreign languages, they do not use a dictionary for every word. They guess, they invent, and they learn from the usage of others. A typical English-speaking new college graduate probably has a vocabulary of 100,000 words or more. This averages to learning over a dozen new words a day for every day of your life, with growth in the last few years probably double that rate or more. And I'll bet that few people look up that many words in the dictionary every day. In Lojban, as those who have written in it already have discovered, new words come to you almost as fast as you need them even when you first start. Fluent Lojbanists will be unlikely to exceed their capability to make the words they need. This will probably take a leap of faith on your part, but it works. But I can answer a question I suspect is in your mind, especially given your mention of ambiguity. Yes these words are going to be ambiguous, just as tanru are, in the early days. That is the only way that it could be. Even if we defined every word exactly before we started, people wouldn't use those exact definitions simply because they can't use dictionaries in real time. Lojban is syntactically unambiguous - its grammar does not lead to multiple interpretations - but Lojban's semantics is NOT unambiguous. We will minimize the ambiguity by eventually assigning the most popular meaning of a lujvo as the only meaning, making some checks and allowance for a limited set of rules. This will work ONLY because the choice IS the most popular usage, and is thus em- bedded in actual usage. Language inertia of actual usage is the most powerful force and the only effective one for stability and restricted meanings that we have. As for the effects of this ambiguity on Lojban's claim to allow 'speaking nonsense' that cannot be extrapolated simply from context - for the first few years this will not be easy - you'll have to be long-winded to speak unambiguous nonsense. But few people will have the command of the language to need or want to do this extemporaneously in speech. In writing you'll have the time to word sentences carefully to meet your expressive needs. I haven't yet mentioned borrowings from other languages, which form a large part of the technical vocabulary that a person will use. These words are going to be no more ambiguous than the words they are borrowed from, and possibly less if you use my proposed rafsi-labelling convention so that people know that "charm" and "strangeness" are subatomic concepts and not psychological ones. What I suspect will actually happen in Lojban writing, by the way, is what already happens. People who write in Lojban will add a little mini-glossary, often written in Lojban itself, which defines at least the tanru (or perhaps a longer definition if the tanru isn't obvious) for each lujvo they use for which the meaning isn't obvious from context. From these glossaries, your vocabulary will grow; and from these, so will we gather our dictionary data. Probably by the end of 1990, when work on a dictionary can probably begin, we will have word lists of thousands of words. Actually compiling a dictionary is mainly an editing job, requiring much time and care, luckily aided nowadays by computers. But a dictionary of 50,000 Lojban entries, small by English standards, will probably a few million words of text, several times the length of the textbook. Small wonder that I predict years to finish the dictionary. JCB took a couple of years for the 8000 word 1975 dictionary, and most of that work was keypunching and card sorting; the amount of editing done was minimal (and it showed). 25 Remember though, it will take years to get the dictionary in final form. There will be intermediate forms and word lists, starting with the textbook itself, but as these get large, they will be best used on computers - Lojban word lists can quickly become huge. If we get grants, we can put more bodies to work on the project, and the dictionary will be done more quickly. But no dictionary is ever truly done. the moment that we finish publishing a dictionary, we will have to add the new words from another year of usage. My own reference, Webster's New World Dictionary, Second College Edition has 160,000 entries and is updated about every two years with about 20,000 changes. This is the price of a living lan- guage. It will be much easier with Lojban; we have clear routes to making new words, and the choice of meanings, if not unambiguous, is fairly limited. If you remain unconvinced by my arguments on the nature and use of a dictionary, I recommend the introductory discussion in the Webster's above; it gives a linguistic view on the making of dictionaries that defines the problem better than I can. Other editions probably have similar discussions. So don't let the lack of a final dictionary stop you from learning and using Lojban. There will be plenty that you can read and write, do and say, without using one. Your learning of Lojban will not stop with the textbook; it will only begin. P. S. Adrian also mentions the Icon programming language and MacIntosh hypercard. We're looking at both Icon and SNOBOL for improving the random sentence generator and for the eventual translation program. pc is spending his sabbatical this year working on a logic teaching program written in SNOBOL. But neither Nora nor I has time right now to tackle a new programming language. We stopped in the middle of learning Turbo-Prolog when JCB forced us to break away from the Institute with his legal threats. At least two Lojbanists are looking at Hypercard applications with Lojban, Karen Stein in the Blacksburg class and Board Member and DC-class member Albion Zeglin; possibly others as well. Others have talked of adding sound to Mac Loj- Flash, but no one has ever done any work on it - we have the pronunciation algorithms. I don't know much about the capabilities of the MAC, since we don't have one; I've had enough trouble just trying to get data transferred to the MAC and getting copies of the MAC programs needed to fill orders. This may improve, now that a couple of local people have MACs, but I won't promise great new products for the MAC because I can't personally fill those promises, and I won't commit for others. Request for Historical Anecdotes and Information As I contact people from outside the community, I am finding occasional interest in the history of the Loglan project. I suspect this interest will grow as Lojban becomes more well-known and successful; witness that many discussions of Esperanto in the last few years have dealt with its early history (though this may be due in part to that language's recent centennial). The linguistics staff of the Library of Congress is collecting and maintaining copies of all of our publications as a chance to document the development of an artificial language while it happens. I joined the project in the early 1980's and have relatively small amounts of information about the pre-1980 period, except what is recorded or implied in The Loglanist (the Institute journal put out by pc from 1976-83) and JCB's NSF proposals. My personal recollections are dominated by the recent politics, and as a result, alas, so are my descriptions of the project history. I'd like to have a more neutral, historically balanced perspective. The new L1 gives credit to dozens of people, and summarizes the history briefly. Some of the people credited are among you readers; some of you may have heard of the language from others who were involved in the earlier years. (We wouldn't mind getting them interested again, by the way, but that isn't the point here.) Therefore, I am seeking all varieties of historical facts, anecdotes, first person accounts, and even rumors about Loglan (and Lojban) that you may have 26 heard through the years (Try to indicate whether you think the information is true or not, but rumors, misconceptions, and false data, if identified as such, are also valuable.) Also interesting are perspectives as to why and how you became interested in Loglan and/or Lojban; we of course keep all correspondence we've received in order to retain this type of information. I am slowly building a chronological history from the information I gather. When it is presentable, I will add it to the product offerings. An actual book about the history of the Loglan project will have to wait years, until we see how the story comes out. By that time, the project founder and early workers may not be around to tell their stories. We urge you all to preserve your memories on paper; the book can then be written when the world is ready for it. Contents of Ju'i Lobypli #10 JL10 will mark a major threshold in Lojban history. We expect to have more writings in Lojban than ever before, and of a wider variety. For the first time, the torch will be passed from Nora and myself to those of you whom we have helped teach. The Lojban text in JL10 will have been written by others, not by us (though Nora may find time for another comic strip). Athelstan has already committed to a Lojban article on Sapir-Whorf, and a translation of a short story by Saki. He also wants to finish sample translations from other languages, per- haps ancient Greek and Old Norse that he is competent to translate from, thus helping to prove the language's versatility. We also hope to have writings from other Lojban students, including Jamie Bechtel's short story, which is the first Lojban science fiction. Athelstan also plans to write about an idea he's had: the resurrection of 'skaldic poetry', the heroic poetry of the Norse and Icelandic sagas, in Lojban. He believes that Lojban is especially suited to this poetic form, as well as similar forms that rely on metaphorical and figurative expression coupled with extremely rigid rhyme, meter, and alliteration rules. Skaldic poetry forms dominated northern Europe for over 400 years, incidentally, during the time period when modern languages first developed; it would be ironic if skaldic poetry led the evolution of the first post-modern tongue, as well. Form and manner of expression is relatively more important in skaldic poetry than is the abstractness of idea that dominates modern poetry. These features are relatively easy to control in Lojban, with its regular pronunciation and stress patterns and its optional terminators. New Lojbanists may thus find it easier to write skaldic poetry than other forms. Athelstan is noted locally for his skaldic compositions, so his discussion should be informed and interesting. He may even be able to include a Lojban sample by the time JL10 comes out. (Anyone want to try Japanese haiku poetry in Lojban; this poetry is similarly form-oriented?) Poet Michael Helsem of Dallas is also learning Lojban, and his efforts are getting better as he gains a handle on the grammar. Preston Maxwell has been translating short stories from other cultures into Lojban. We hope that we can present some of their works next issue as well. 4 people who attended LogFest and participated in the discussion of testing Sapir-Whorf made commitments to write up their perspectives on the discussion. Tommy Whitlock, Athelstan, pc, and Ralph Dumain - it's in print now, so I'd better get something from you. We may have some additional notes based on JCB's discussions of the subject in the new L1 edition. JL10 won't be put together until we get back from Worldcon. We'll have a current report on what happened, though we'll also include a report in LK11. And who knows what else ...? So, until then: co'o.