me lu ju'i lobypli li'u 13 moi

From Lojban
Jump to navigation Jump to search

For a full list of issues, see zo'ei la'e "lu ju'i lobypli li'u".
Previous issue: me lu ju'i lobypli li'u 12 moi.
Next issue: me lu ju'i lobypli li'u 14 moi.

Copyright, 1990-1, by the Logical Language Group, Inc. 2904 Beau Lane,
Fairfax VA 22031-1303 USA Phone (703) 385-0273
lojbab@grebyn.com

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.

Number 13 - August 1990
Copyright 1990, The Logical Language Group, Inc.
2904 Beau Lane, Fairfax VA 22031 USA (703)385-0273
Permission granted to copy, without charge to recipient, when for purpose of promotion of Lojban.
Lojban Grammar Baselined

cmavo list Completed

Details Inside, and More.

Ju'i Lobypli (JL) is the quarterly journal of The Logical Language Group, Inc., known in these pages as la lojbangirz. la lojbangirz. is a non-profit organization formed for the purpose of completing and spreading the logical human language "Lojban", and informing the community about logical languages in general. la lojbangirz. is a non-profit organization under Section 501(c)(3) of the U.S. Internal Revenue Code. Your donations (not contributions to your voluntary balance) are tax-deductible on U.S. and most state income taxes. Donors are notified at the end of each year of their total deductible donations. We note for all potential donors that our bylaws require us to spend no more than 30% of our receipts on administrative expenses, and that you are welcome to make you gifts conditional upon our meeting this requirement.

Press run for this issue of Ju'i Lobypli: 360. We now have over 680 people on our active mailing list.

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 identified by that 'trademark'.

Your Mailing Label

Your mailing label reports 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 15 August 1990. Mailing codes (and approximate annual balance needs) are defined as follows:

Level B - Product Announcements Only
Level R - This is a Review Copy for Publications
Level 0 - le lojbo karni - $4 initially + $5/year balance requested
Level 1 - Ju'i Lobypli - $20 initially + $20/year balance requested
Level 2 - Level 1 materials and baselined products - $25 initially + $25/year balance requested
Level 3 - Level 2 materials and lesson materials - $50 initially + $40/year balance requested

OUR FINANCES HAVE REACHED A CRISIS; we will have to take action. Please read the news on Finances inside. If you have a negative balance over -$50, we must reduce your mailing level to level 0 unless you contact us with a very special reason. If you have a balance between $-25 and $-50, we are also likely to reduce you to level 0, but can be more flexible, especially if we hear from you. Those with lower negative balances will be reduced on a case-by-case basis. If you have a negative balance, write to us, and try to contribute something towards your balance; even if you only send $5 or $10. Those who contribute some money will not be cut back yet, but we need to hear from you by 20 October. Note that, due to size, the base price for this issue of JL is just short of $10.00.

Contents of This Issue

This issue marks major milestones, but reports our serious financial situation. The issue was delayed almost a month because we didn't have enough money to pay for it. This issue includes the new grammar baseline and the cmavo list as separate enclosures to an over-50 page issue. However, if our finances do not improve significantly, future issues will be much shorter. The choice is up to all of you.

Bob LeChevalier starts a regular 'column' written directly in Lojban, and without translation. As of this issue, all the materials needed to read Lojban text have been distributed - its time to set an example of using the language. We also include Lojban writings from Michael Helsem, who has written some 40 poems in Lojban, and from Athelstan and John Cowan.

Our regular news section discusses results from LogFest 90, an update on the textbook plans, and our financial woes. We also include two items from the editor of the Esperanto League for North America newsletter, Don Harlow, a response to JL11's Esperanto article and an article by Don on artificial languages, reprinted from his newsletter with permission. We have some writings from Andy Hilgartner, outlining his several decades of General Semantics research that have recently been affected by his contact with Lojban.

Finally, LogFest attendees chose to have the community vote on the logo and another issue. Please respond by 20 October. If you are level 3, or have used LogFlash, we are also asking you to report on where you are at in learning the language by then. There will be another weekend get-together at Bob and Nora's place that weekend and everyone is invited. Let us know you are coming.


						   Table of Contents

News
  LogFest Results - Lojban Grammar Baselined, gismu List Baseline Changes  --3
  cmavo	Dictionary Progress				       --8
  Finances						       --9
  Language Definition Status Summary			      --11
  Products Status (including Textbook Update)		      --13
  International	News - Athelstan's Travel Plans		      --15
  News From the	Institute				      --16
JL11 Esperanto Discussion - A Response from Don	Harlow	      --16
Masters	of Tongue Fu, by Don Harlow			      --24
Two Papers by Andy Hilgartner				      --29
Letters, Comments, and Responses			      --32
le lojbo se ciska					      --40
Enclosures - Baselined Machine Grammar,	Abbreviated cmavo List,	Ballot/Questionnaire

Computer Net Information

If you have access to Usenet/UUCP/Internet, you can send messages and text files (including things for JL publication) to Bob at: lojbab@snark.thyrsus.com

Join the Lojban news-group.
Send your mailing address to: lojban-list-request@snark.thyrsus.com
Send traffic for the news-group to: lojban-list@snark.thyrsus.com

Please keep us informed if your network mailing address changes.

Compuserve subscribers can participate. Precede any of the above addresses with INTERNET: and use your normal Compuserve mail facility (its possible that you can send only to addresses in the '@' format). Usenet/Internet people can send to Compuserve addresses by changing the comma in the Compuserve address to a period: nnnnn.mmmm@compuserve.com

Whether you wish to participate in the news-group or not, it is useful for us to know your Compuserve address. For example, any decision for la lojbangirz. to obtain a Compuserve account will be based on a need to serve a goodly number of you that want to exchange information.

We've been requested to more explicitly identify people who are referred to by initials in JL, and will regularly do so in this spot, immediately before the news section. Note that 'Athelstan' is that person's real name, used in his public life, and is not a pseudonym.

'pc' - Dr. John Parks-Clifford, Professor of Logic and Philosophy at the University of Missouri - St. Louis and Vice-President of la lojbangirz.; he is usually addressed as 'pc' by the community.

'Bob', 'lojbab' - Bob LeChevalier - President of la lojbangirz., and editor of Ju'i Lobypli and le lojbo karni.

'JCB', 'Jim Brown'- Dr. James Cooke Brown, inventor of the language, and founder of the Loglan project.

'The Institute' - The Loglan Institute, Inc., JCB's organization for spreading his version of Loglan, which we call 'Institute Loglan'.

News

LogFest Results

Lojban Grammar Baselined - By consensus, the attendees at LogFest approved a preliminary baseline of the Lojban grammar, subject to minor corrections and typos that were expected to show up before publishing in this issue. (An awful lot of work was accomplished in the last two weeks before LogFest, and the copies printed for LogFest were rather hurriedly prepared.)

A final baseline is expected in about 6-9 months, after the textbook is completed. However, given the extremely slow rate of change to the grammar up until now, few if any substantive changes are expected before the final baseline. All major subsections of the grammar have now been examined at least twice, and Bob was able to give lectures on the attitudinal, tense/modal, negation, MEX, and lerfu systems at LogFest with minimal preparation, demonstrating that these new developments are teachable.

The grammar has of course been verified through usage, with translations and original materials in a variety of styles and on a variety of subjects, as well as in conversation. The design has proven solid and robust.

The grammar still requires formal usage testing (as opposed to disambiguity testing with the computer tool 'YACC'), and no one has stepped forward to update the parser to the new grammar. In lieu of this, we will be updating the random sentence generator to the new grammar, and examining its outputs manually. We also will be using the older parser to examine Michael Helsem's poetry and any other Lojban text that is submitted during the preliminary baseline period, and will ensure that discrepancies are accounted for by known changes to the grammar.

A copy of the approved grammar is included with this issue. The grammar is a 'machine' grammar, and designed primarily for computer use. However, people can understand and use it, too, with explanation. We've provided such an explanation as front matter for the grammar copy, and would appreciate comments, especially from non-computer people, as to its understandability and usefulness.

Several people are working on condensed versions of the grammar that are more understandable to everyday people. These versions, written in a format called 'Extended BNF' are only two to three pages long. The primary weakness of E-BNF grammars is that they cannot be verified by YACC; we therefore want to ensure that the versions are thoroughly checked before publishing them.

Anyone interested in working on the E-BNF effort should contact us. John Cowan and Carl Burke are leading the effort, and will respond to you. Include a Compuserve or Usenet/Internet/ Bitnet address if you have one, and want to communicate via net.


gismu List Baseline Changes - The list of proposed new gismu swelled just before LogFest from the list presented in the last JL issue, making for a lively, if long, session at LogFest devoted to going over the proposals. About 3/4 of the changes were adopted, with a couple of issues left up in the air for further comment from the community.

Almost all of the changes are additions to the list of gismu, the type of change having least impact on someone learning the language. There was one word change accepted, the change from "ckamu" to "mleca" discussed in the last issue.

There are three changes to keywords for existing gismu, specifically the gismu for some of our basic Lojban language concepts ("gismu", "tanru", and "lujvo"); these word changes are not changes to the meaning of the gismu, but are instead corrections to eliminate misconceptions. The existing keywords are the words that Jim Brown used in his writings for the concepts, and have historically been criticized as being 'the wrong word'. We've stuck with them for tradition, but reactions from the community when the proposals were circulated on the computer network indicated that the keywords were causing confusion and that it was time for a change.

The problem is, of course, that these concepts are not commonly used in English, and no English word accurately reflects their meaning. LogFest came up with the phrase "root word" to serve for gismu, but was unable to agree on new phrases for "tanru" and "lujvo". Some of the proposals are described below. For purposes of discussion at LogFest, we limited keyword proposals to 15 characters or less, the existing LogFlash limit. Since we are increasing that limit to 20 characters to support the cmavo keywords, some of you may be able to devise clearer phrases than we did at LogFest; for example, some of Bob's preferences for "tanru", "modification pair", "modified concept", or "modified relation", are now permitted. We need your suggestions within the next month or two.

For the new words that have been approved, and for the open issues that have obvious new word implications, we have already started the word-making process. For the first time, we are able to use a native speaker for the primary research in a non-English source language. Vijay Vaidyanathan of Albany NY, a native Tamil speaker who is fluent in Hindi (and whose wife is a native Hindi speaker) is providing the original Hindi research. Mimi Herrmann, a DC-area language aficionado, is researching the Arabic words. We turned to Russian translator Gary Burgess for aid in Russian, and Bob did the other three languages using his shelf of dictionaries. When all of the contributors get finished, hopefully in a few weeks, Bob's new 386 computer should make short work of building the possible words, and we'll be able to tell you the results next issue.

Lojban gismu are made by searching for sound patterns that most match the words from the source languages. A score, called a recognition score, is calculated based on weighted averages using the speaker population for each source language. The weighting system for language populations takes the number of native speakers and adds half of the number of second language speakers.

Bob researched how the last several years affected the language population weights. We had expected a significant growth in speakers of other languages, but was surprised to find that English has receded, and rather rapidly. While it is still the language of choice in international business and in science, there are now considered to be fewer speakers of the language worldwide than there were just a few years ago. If we use the new data in making the gismu, English drops to 3rd place well behind Hindi, and is being threatened by Spanish.

The most striking cause of this change is in India. When India first became independent in 1948, English was often the language of choice as a lingua franca for communication between speakers of the several hundred different languages spoken in India (there were some 15 languages cited as 'official' at the time of independence). Now, however, only English and Hindi are considered 'official', and there is a decided bent towards Hindi. There are some 230 million native Hindi speakers (and perhaps another 60 million Urdu speakers - Urdu is considered by linguists to be the same language as Hindi, but is written in Arabic script). This is about the same number of native speakers as English has world-wide. In second language speakers, the real difference shows up. Hindi is the lingua franca of around 400 million speakers in India, and this number is rapidly increasing under the Indian literacy program. English is now spoken by only 21 million Indians, only 2% of the population.

English is also receding in Africa, although not as dramatically.

The Russian numbers have also changed, as greater recognition has been placed on the non-Russian nationalities in the Soviet Union.

The numbers are confirmed in another source, a survey description of the world's languages edited by B. Comrie, an expert on language universals. The section on English estimates the number of speakers as 350 million, whereas quotes as high as a billion were easy to find when we started surveying languages in 1987 before the initial gismu-making effort.

Following are the numbers of speakers in millions derived from the 1990 Britannica Book of the Year. The final column gives the weights normalized to 100. We've given numbers for the next two languages behind Arabic, which are rapidly gaining. However, we are unlikely to replace Arabic with Bengali until the numbers are solidly in its favor, since Bengali is Indo-European and confined almost entirely to two countries.

Language  Native 2nd  Net	1990  1987                     
    		 Score	 Wt.   Wt.                     
Chinese    754M  339M 923M	35.5  33.5                     
Hindi      288M  478M 528M	20.5  16.5                     
English    299M  228M 413M	16    18                       
Spanish    300M   38M 319M	12    12.5                     
Russian    170M  118M 229M	 9    12                       
Arabic     181M   22M 192M	 7    7.5                      
Bengali    176M   26M 189M	 -                             
Portuguese 156M   12M 161M	 -                             

We will experiment with both sets of weights in making the 20 new gismu to be constructed; there is no policy on whether to update weights periodically, since we don't anticipate making many more gismu.

Now, here is the summary of decisions made by the LogFest attendees:

The following proposals were approved with little controversy:

  1. Change "ckamu" to "mleca" for rafsi considerations;
  2. Add "daytime", change keyword for "day" ("full day"?, "24hr day"?);
  3. Add "virtue", as distinct from "good", to parallel with "evil";
  4. Add "citrus";
  5. Add "cabbage", to include broccoli, cauliflower, and perhaps lettuce;
  6. Add "hemp", to include natural rope, burlap, marijuana, and hashish;
  7. Add "protein";
  8. Add "buckwheat";
  9. Add "cassava", to include taro and yam, and other starchy roots (not tubers);
  10. Add "sorghum";
  11. Add "magenta" and "cyan" as the missing two subtractive primary colors;
  12. Change the keyword of "gismu" from "primitive" to "root word";
  13. Add "North America", the continent, as distinct from "merko", referring to the U.S.;
  14. Add "South America", the continent, as distinct from "xispo", referring to Latin America;
  15. Add "Antarctica".

The following proposals were added with considerable debate and discussion:

  1. Add "glimmering" to cover the concepts of morning and evening twilight, as well as the phenomena of "astronomical terminator" and "penumbra"; the poetic usefulness and the astronomical extension of the concept led to passage;
    1. "dawn"/"morning twilight" and "evening twilight" were voted down.
  2. Define "morning" and "evening" symmetrically;
    1. The specific symmetry required much debate; consensus was finally built around a culture-dependent definition, wherein morning is the time between sleep and work, and evening is the time between work and sleep, according to the cultural norm. In a tanru this could be modified to a personal norm.
  3. In a discussion of "decrease" as an opposite of "increase", initial sentiment for adding it was weak;
    1. It was noticed that the existing place structure of increase was transitive; it was proposed that by changing the place structure to the intransitive "x1 increases in property x2 by amount x3", an opposite gismu for "decrease" would be better justified. Without the change, the semantic difference from "adjust" and "add" was felt to be too small.
    2. The vote to add "decrease" in parallel to the new meaning of "increase" was then successful.

The following changes were voted down:

  1. Add "text";
  2. Add "tears";
  3. Add "ugly", the opposite of "beautiful";
  4. Add "diffuse", the opposite of "concentrated";
  5. Add "deficient", (after discussion, it was decided that "deficient" is the opposite of "excess" with "sufficient" as middle ground. The opposite of "sufficient" then, encompasses both excess or deficient;
  6. Add "alfalfa";
  7. Add a common term for the Western Hemisphere continents.

The following are still open issues:

1. The definition of "arm" was not discussed; we forgot. The issue was based on our learning that different cultures include/exclude the hand and the shoulder from the concept of "arm". Tentatively, we will define an "arm" as a non-supporting limb, without specifically excluding or including the extremity. 2. It was decided to change the keywords for "tanru" and "lujvo". Unfortunately, there has been no consensus on what to change them to. The clear sense is to avoid linguistic jargon and words that have multiple meanings in English such as "compound". There is some sentiment for keywords that show a parallelism in definitions of the two concepts, though alternatively the parallelism could be made clear in the extended definition. The proposed choices, in roughly chronological order are:

tanru		     lujvo                               
open compoundclosed compound                               
relation phraseaffix compound                              
relation phraserelation compound                           
word cluster    cluster word                               
word cluster   affix cluster                               
word grouping affix grouping                               
grouped words  affix word(s)                               
modified phrasemodified word                               
phrase relationaffix relation                              


As noted earlier, some slightly longer keywords can be considered, since we have to create a version of LogFlash supporting the 20-character keywords for cmavo data. Bob is adding the following based on this new option:

modification pairmodification word

3. The familial relationships never quite seem to satisfy. It was agreed to add "sire" and "dam" to the definitions of "patfu" and "mamta". Later it was suggested that we retain some unsatisfactory holes and combinations, which are of uncertain importance. The fact that American culture is shifting away from traditional family structures makes it unlikely that we (who are almost all Americans) can decide on a culturally neutral solution. The choices are then to be maximally inclusive of the possible relationships, or to pare the list in ways that ignore American sensibilities. The general preference seems to be for the former. Thus, we can make the following matrix:

Gender-neutral	Male	 Female	  Gender-neutral	   
but genetic			 not-necessarily-          
				     genetic		   
							   
panzi		bersa	  tixnu	      se rirni		   
offspring	son	  daughter    reared		   
							   
				      verba		   
				      child		   
							   
				      cifnu		   
				      infant		   
							   
		bruna	   mensi      tunba		   
		brother	   sister     sibling		   
							   
se panzi	patfu	   mamta      rirni		   
   ?		father	   mother     rearer		   
		sire	   dam				   
							   
se jbena						   
mother/father						   

Note that as currently defined, "patfu" and "mamta" are defined biologically, whereas their counterparts (except "se jbena") need not be.

An obvious suggestion is to make "patfu" and "mamta" non-biological. However, it can be argued that with animal breeding and genetics, and in some less transitional cultures, the biological parents have a uniquely important role enough to be considered 'primitive'. If so, the tanru "mamta se panzi" and "patfu se panzi" may be too long to be satisfactory. (We asked Vijay, our Hindi/Tamil expert, though, and at least for human parents, the biological aspect is secondary to the social relationship - non-biological parents are called by the same 'mother' and 'father' terms as biological parents. If this is the case, the longer tanru for specifying biological parents may be acceptable.)

Other options:
a. Change "panzi" to be its inverse, making "se panzi" into "offspring";
b. Add a different gismu to be the inverse of "panzi", with a keyword something like "engender";
c. Add two gismu to specifically represent the genetic relationships "sire" and "dam" as distinct from the social terms "mother" and "father";
Other options are possible. Note also that "rirni" is not quite the same as "mother/father", so we may need another genderless general term here.

For now, we are making up gismu for "engender" as one of the 20 words. The actual word that results may affect the decision.

In addition to the above, there is also the question of the extended family, which we have long ignored. We can be very specific about "mother-mother", "father-sister", and other extended family relationships, but we cannot be general. Most cultures either use very general terms or very specific ones (Hindi and Chinese distinguish between a father's older brother and younger brother as different word-concepts); in American culture, of course, divorce and remarriage is causing extended family relationships to become so complex that specific terms will not suffice. For discussion purposes then, Bob is proposing (and making):
a. "elder/ancestor" for family members of generations preceding the parents (including non-direct line, the relationship is more social/ethnic than biological). Gender would be added via tanru, as would explicit biological lineage (or a place could be used for specifying lineage, with specific names used in alternation with properties of the lineage of relation). The conversion would give "descendant" as well as "grandkids" in the broadest sense. This was independently proposed by John Cowan as "x1 is an ancestor of x2 of degree x3".
b. "aunt/uncle/godparent" for non-lineal (socio-ethnic) family members of the parental generation. The conversion would give "niece/ nephew".
c. "cousin" for non-immediate (socio-ethnic) family members of the same generation.

The generalized family relationship is still expressed by "lanzu", which can be modified via tanru.

Related Discussion - Days of the week were dis-cussed prior to the gismu baseline discussion, and it was decided to add color-and continent-based names as alternatives to the number based names that have been standard (it was at this point that we realized that we were missing 3 of the 7 continents). In addition, the number based names will be set to run from 0 to 7, with Sunday serving as both 0 and 7, depending on speaker preference and cultural orientation.

John Cowan expressed great skepticism that any alternate system would catch on. They seem too much like crackpot 'calendar reform' efforts, and also aren't well supported in numerical date representations. He also noted that not all cultures have a 7-day week. The generic concept of a week is the time between successive market days, which ranges from 4 to 9 days in agrarian non-Western cultures. He thus suggested that "jeftu" add a place to indicate the culture (it was later realized that a 'standard' is already in the place structure). He is using days of the week based on the International names of the classically known heavenly bodies, thus allowing him to parallel Romance languages that did likewise. (English and the Germanic languages loan-translated the Roman gods used in planet names to the corresponding gods' names from German mythology.)

Athelstan notes that in Israel, the days are number-named from 1st-day to 6th-day, followed by the Sabbath; this is similar to our adopted system.

Other LogFest Results - LogFest only had about 18 attendees, but they were a truly exceptional crowd that demonstrated by their commitment to the language that Loglan/Lojban is going to continue to grow and prosper. More than half were from out of town; three graduates of the Blacksburg class came. Almost all were level 3 active language students; thus we never really got to exercise our plans to support activities for newcomers at the same time as a main track.

Unfortunately, a lot of key people we thought were going to come didn't. For example, Dr. Yorke, scheduled to lead a discussion of his proposal (printed in JL12), didn't make it, but won an award for the best excuse ever for not coming to LogFest. He called Friday evening to tell us that his house was surrounded by fire trucks. (The lengths people will go to avoid coming! Luckily, the fire was not too serious.)

We discussed Dr. Yorke's proposal anyway. There was little support - the people working on Lojban are interested in a full language, not a hybrid form. la lojbangirz. thus will not change its goals and plans to match his ideas, although we are willing to support and encourage anyone who wants to develop the idea further and give it a proper test. Most participants, though, said that a hybrid LogEnglish might as well use straight English words for the predicates. The loss of audiovisual isomorphism is not significant until someone comes up with a speech recognizer. It was also suggested, and we will consider, using such a hybrid form in teaching materials for beginners, so that they do not need to master the vocabulary to get a start on the unique grammar features of Lojban. This will not be used for the initial textbook, however. If someone wants to try this technique in teaching a class, let us know.

Saturday was primarily devoted to technical presentations on the major areas of the grammar that had been reviewed for the baseline. Athelstan gave his mini-lesson during one gap in this, and we had a short period of Lojban conversation during another gap (Bob was able to make do without his word lists. But especially gratifying was that several others participated, including some who had not been part of a class.)

We also discussed John Hodges' proposals on time and typography. Bob's counter-proposals on these issues were adopted, although with no change to the period before the "i" at the beginning of the sentence (see JL12 for both the original proposals and the counters). All typography symbols are optional, and do not affect grammar or pronunciation - symbols of punctuation are used IN ADDITION to the Lojban punctuation words. We won't promise to use them in most JL writings. We type Lojban text slowly enough now, without trying to master some new conventions. However, we will accept and print Lojban text that uses the optional conventions.


Annual Member's Meeting - On Sunday was the annual meeting of Members of la lojbangirz., which ran over as usual, ending rather hurriedly at 3pm after 5 hours of most lively discussion.

While everyone in the community is considered a member of la lojbangirz., we must maintain a formal membership for election of Board members, changes to Bylaws, and as it turns out, whatever else the membership decides to act on. We added 4 new formal members, making a total of 12.

One major item of business was adoption of the grammar baseline, and this was pro forma, almost a letdown after the intensity of the previous day's work. We also moved LogFest back a week, effective in 1991. Next year it will be on the weekend of June 23, 1991. Make your plans early.

If June is a bad time of year for you, there may be another chance. A second get-together will be held this fall at Bob and Nora's house (so save your maps to LogFest), tentatively the weekend of October 20-21. Because the next JL issue will come out too close to then to allow meaningful publicity, this is your only notice. This year's 'Log-Fair' will be an experiment that will be repeated if there is interest. This year's meeting may also serve as a textbook review party, if Bob has gotten sufficient done to justify this.

Electronic mail distribution of our materials was also discussed in the meeting, and a committee was formed to devise a policy on what things will be released in this form. We will use the Planned Languages File Server (discussed in JL12), as well as similar forums such as the Compuserve Foreign Languages Forum, as official repositories. All electronic media distributions will have some type of header giving its status as a draft or baseline, 'copyleft' (similar to Free Software Foundation policies) or public domain status or full copyright if appropriate. In general, our books and major publications that we need to get income from to survive will be protected - unless events transpire to eliminate need for that income. All material that defines the language will be public domain, along with some other things like the brochure that we want widely distributed. Most of our stuff will be 'copyleft', allowing distribution without charge as long as various notices attached are retained.

One fear of public domain status before the language has a solid community was that there is no way that electronic media readers can be certain that they have gotten the 'real thing'; someone could modify the official documents, attach any labels we put on to certify 'officialness', and redistribute them. The 'correct' solution lies in trademark law, but we are reluctant to go in that direction given our disputes with the Institute. Since we are committed to a public domain language, we will probably just say that if you want to be absolutely certain you have the latest and greatest version of something, you will have to order it from la lojbangirz.

The electronic media committee was also charged with developing an introductory lesson for distribution on electronic (and paper) media. This will be tied in with our policy on who gets materials without paying for them (see Finances below). The lesson may have some similarity to the Esperanto postal mail course. Athelstan has prepared a draft mini-lesson for this purpose, which is being tried by a few people. The mini-lesson will also partially replace the current, rather unsatisfactory, Overview that we send to new people.

A second committee was charged with developing a plan to restore our fiscal integrity. The details will be discussed in the next section.

Most surprising about these two committees is that the membership insisted that Bob not be actively involved. "These are NOT Bob jobs!" said Karen Stein. The membership seemed concerned that Bob has been spread too thin, and these committees could handle things without his active involvement. So all of the preceding is subject to change by the committees; your Editor doesn't know much more than you beyond what has been said. The formal membership will hopefully have a report from each committee in August, in order to allow decisions before Athelstan leaves for Europe (he is on both committees).

Bob comments: It is gratifying that the members feel enough commitment to volunteer significantly firmly to relieve me of what has become too much responsibility. This is your language, not mine. Finally, a group of people is acting like they believe me, and acting like it is THEIR language, and taking the responsibility needed to make it succeed.

cmavo Dictionary Progress

John Cowan has assumed responsibility for producing a draft version of the cmavo portion of the dictionary, taking over from earlier work done by Jeff Taylor. His first product for us is the abbreviated definition cmavo list included in this issue.

John has proven extremely productive in the last few weeks, helping with the grammar baseline and the E-BNF effort, compiling the cmavo index included with this issue, and volunteering for major work responsibilities on the dictionary, while contributing valuable suggestions on a variety of other topics. He also has started writing a 'daily' Lojban journal, believed to be a first in that category (although Jim Carter reported doing likewise with an earlier version of the language in the early 1980's). John first learned about Lojban only a couple of months ago, thus showing how quickly a motivated person can master enough of the language to take a leading role.

The cmavo index included with this issue is the first complete list with definitions since the draft list of October 1988. The definitions are short, and not very explanatory in some cases; John is working on expanding them. You can use the partial cmavo list from JL10 to make additional notes on many of the cmavo; the exercise will also help you to learn those cmavo that were on that list.

In many cases, you will need to look up the lexeme for each cmavo in the machine grammar in order to know how it is used; some 100 cmavo are each 'in a class by itself' - a lexeme with only 1 member whose meaning is solely grammatical.

Most of the remaining cmavo are found in a very few lexemes. Members of lexeme UI are the attitudinals, discursives, and observationals used to comment metalinguistically on what you are saying. Lexeme BAI contains the modals; these can be used somewhat like adverbs to modify the selbri, or can be used somewhat like prepositions to tag sumti that are not part of the place structure, thus tying them into the bridi relationship being expressed in the sentence. Lexeme KOhA has the 'pro-sumti', which act somewhat like English pronouns. Lexeme COI has vocatives, which are like attitudinals but are addressed specifically to a listener. Lexeme BY is the lerfu - the letters and symbols of the alphabet, and various auxiliary shifts and markers. Lexeme PA are the numbers and related symbols used to read off numerical strings.

The English 'keywords' intended for use in LogFlash 3 (which will teach the cmavo), all 20 characters or less, are included in the abbreviated cmavo list. Bob and Nora reviewed John's list, adding these keywords (needed for LogFlash 3 - Bob has learned most of the cmavo now in testing that program. See the products section below). We want comments from the community on the keyword choices before releasing LogFlash 3. Do the keywords make sense to you (assuming that the cmavo definitions do)?

The shortened definitions in this list will be expanded into full text definitions for the dictionary, and John intends to devise examples for each grammatical usage of each cmavo. This effort is only a goal, and may not be practical for all words.

Finances

LogFest attendees recognized that our finances have reached the point of crisis. The 18 attendees, in exceptional gestures of support, donated over $1000 to la lojbangirz., but this was merely enough to keep our bank balance positive, allowing this issue to be printed.

Even with these donations, our income for this year is running well behind the income for last year, in spite of the fact that our audience/customer base is some 50% larger. We've kept expenses down; they are also running behind last year, but this is primarily due to firm controls on non-vital expenditures. Our new 501(c)(3) status makes U.S. donations tax deductible, but donations are also well below last year even with the recent influx from LogFest.

The problem is that, while most people who order stuff pay for it, not all of you do. Our prices have been set at a non-profit level, so each non-paid sale is a pure loss that has to be made up from donations. Sales of our software, the only products with a price that makes us some profit, are down, and probably will continue to be until the textbook is published.

Less than 100 people had positive balances at a recent audit. Only a couple of dozen have the balance amounts corresponding to each level that we request on our registration forms and in the mailing label summary in each JL issue. Most magazine publishers require you to pay for subscriptions in advance - we cannot afford to have people paying for issues after their balance goes negative.

Several of our largest contributors, who are footing the bill, are starting to get angry about all this, and have demanded changes in policy. They want their money going to promote growth in the language, and not to finance 'deadbeats'. (Harsh words, but that is a direct quote from one major contributor.)

The most extreme position is that everyone should be able to afford to contribute something towards their balances - the price of an order of french fries every couple of weeks would fully pay for a level 1 subscription. Anyone not willing to contribute this amount simply isn't interested in supporting us.

The counter to this position recognizes that some people don't have the money for french fries, or for Lojban. We don't want to exclude prisoners, students, and people from other countries where incomes are a fraction of those in the U.S. Lojban is not supposed to be a language for the 'monied' class, but for everyone.

An intermediate position, which may win out in the long run, is that we are going to have to insist on payment from most people, and make exceptions on a case-by-case basis, only by special request, and often with strings attached such as required volunteer work or the submission of writings to show that the beneficiaries are actually using what we're giving them.

Regardless of the final decision, something must be done, and now. Less than 20% of our people have positive balances, and we must get the rest of you to pay some amount towards our expenses on your behalf or we'll have to cut you back, or cut you off.

Some decisions have been made. If your balance is more than $50 negative as of 20 October, you will be cut to level 0. Only a special pleading will change this. Contributing some money, even a little, will probably put off a cutback, if we are convinced that you are likely to continue using the material and contributing some money. If you can pay off your balance, or make a commitment to do so by a specific time, all the better.

If your balance is more than $20 negative as of 20 October, you being put on notice that we may have to drop you to level 0 as well. You should write to us, and try to send some money, even if it is only $5 or $10. Ideally, you will bring your balance positive.

If your balance is negative at all, and we haven't heard from you in over a year, we need to hear from you, or you also may be dropped in level. Even a short note to let us know you are still interested and reading the material we send you will help; if you send some money to bring your balance positive, all the better. The Ballot/Questionnaire enclosed with this issue is an opportunity to respond without writing a letter.

Of course, we would strongly prefer that you bring your balance up to the desired support level listed on the front of each issue for your level ($40 for level 3, $25 for level 2, $20 for level 1, and $5 for level 0); that way we don't have to keep asking for money each issue. We'll accept whatever you can contribute. But the bottom line is that those who want the language to succeed, and can afford it, should have balances over $20, or even $50, for our use as operating funds, and are contributing a little extra on the side to help us with those who do not pay.

If you feel that your balance charges have been unfair, that we've charged you for something you haven't gotten, let us know, and we will make the situation right. If you question whether our material to you isn't worth what we are charging you, let us know what you think it is worth, and we may be willing to negotiate - but be aware that we are spending what we charge you.

Your negative voluntary balance is not a 'bill'. We have no legal way to collect it if you don't want to pay it. But if we can't get much higher percentages of people paying for the stuff we send them, we will be in serious trouble. The question is whether you want us to succeed. If so, you'll contribute.

Even if most people bring their balances positive, this won't be enough; our expenditures will be rising as we get ready to publish the textbook, and as our numbers continue to grow. We actually need to convince many more of you to keep the specified positive balances to support your subscription level. Those positive balances are our operating funds that allow us to keep day-to-day business going; we currently spend about $1000 a month.

We've also learned more about the realities of the book publishing business, and why book prices have gone up so much in recent years; bookstores and distributors demand and get 40-55% off list price, and even then don't pay for 3 months or more and expect to return the covers of any books they don't sell (they discard the books - postage and labor is too high to send the whole book back) for credit.

Based on this, we need enough income from sales within the existing community to pay the entire pre-publication print and marketing bill, or we'll be bankrupt. Either that or get a lot of donations, which seems unlikely.

As a result of our financial situation and this future reality, all existing discounts are terminated as of 1 August 1990 pending any decisions from the finance committee and the membership. Overseas people (excluding US/Canada/Mexico) will be charged a flat 20% surcharge to handle the increased postage. The primary impact of this will be that JL prices inside the US will no longer be discounted for bulk mail.

A 20% discount will be given on any order over $20 which is prepaid - i.e. you have enough in your balance, or contribute enough to cover the price. (This discount will cancel the surcharge for people outside North America.)

At textbook time, there will probably be a further carrot for those with positive balances. Everyone in the community will get an offer to buy the textbook in advance of the official publication date at some discount from list price. We expect that for prepaid orders from people with positive balances, we will be discounting the textbook price as much as $10. This of course means that the price will be that much higher for everyone else. Those with negative balances will likely receive little or no discount.

In addition to all of the above, the finance committee will probably be recommending that we send a fund-raising letter to each of you, independent of JL, pointing out how much we've accomplished on a mere shoestring and asking you to contribute so that we can continue to produce your language. We still need donations above and beyond paid sales in order to finance our continued growth (and to pay off existing debts).

With our finances as they are, we are at a disadvantage in seeking outside grants; donors look for financial stability and community support, and we just aren't getting enough. It also costs money to prepare grant proposals, and we don't have it.

The finance committee will be considering how to value volunteer efforts on our behalf, and we may offer volunteer work as a way to ameliorate a negative balance. The offerings will be slim; there isn't much volunteer work that we can place value on, and we can't afford to 'pay' much.

Volunteer activities include translating materials into foreign languages, giving talks and recruiting new people, and writing significant amounts of Lojban text. In the short term, we're looking for volunteers in the DC area to come over to Bob's house and type several hundred addresses into the computer - these are addresses of book dealers and reviewers that we will have to contact before the textbook is published.

Your response to all these measures will determine where we go from here. The 350 JL subscribers are Lojban's best supporters but also the biggest drain on our resources. If we can go from 100 people with positive balances (the current number) to 250, we will probably survive. If we can get to 300 or more with balances over $10, the Lojban project may again be fiscally healthy.

We welcome all suggestions for other ways to raise money, to gain donations, and get more balance contributions.

Language Definition Status Summary

With all that hand-wringing about finances, it is worth reviewing what your contributions have bought so far in terms of products and services, and most important, in terms of the language itself.

The design of the language is basically complete; we await various write-ups before final baseline of the design, because we need to have a clear written statement of what the design is in order to protect that design against change.

The language is stable. Preliminary baseline changes have been minimal, and have almost entirely been additions to the language that have no impact on people who have already started learning. Let us look at each design area to see where it stands:

Phonology - Lojban pronunciation has not changed since being baselined over 2 years ago. Few questions have been raised about the design. There is no reason to expect changes for the indefinite future, although we will be trying to make the Synopsis discussion more readable to the average person before incorporating it into the dictionary.

Morphology - The Lojban morphology has also been baselined for over 2 years without significant questions being raised. There is a remaining open issue on the exact rules for borrowing words from other languages. The required form - the essence of the design - is firm, but whether there needs to be additional constraints on borrowing forms remains a matter to be decided by the people who use the language.

There has been a proposal, supported most notably by Michael Helsem, that would allow either 'r' or 'n' as the 'hyphen' after a CVV rafsi, unless the following consonant is the same letter. This is a simple change that would only increase options (currently, you use only 'r', unless the following letter is an 'r', in which case you use 'n'), but we are reluctant to make the change. One reason is that it seems too unimportant to justify a baseline change. Another is that it further increases the number of possible forms for a lujvo, making it harder to produce the dictionary (and taking up more space). Finally, the change would constrain le'avla (borrowing) space, something we should avoid unless we have a good reason.

Orthography - The Lojban alphabet and required writing conventions are unchanged since the baseline over 2 years ago. In JL12, Bob proposed some additional optional conventions, which were adopted at LogFest. We've been given a proposal to use different letters to represent the 'i' and 'u' glides in diphthongs, since these technically are different sounds than those in the non-diphthong vowels; we'll probably discuss this proposal prior to next issue, but expect no changes.

gismu - The gismu list was baselined just about two years ago. Specifically baselined were the gismu themselves and the corresponding English keywords. (The rafsi and place structures are discussed below.) Exactly one gismu was changed at the recent LogFest; no other changes have been even proposed. There have been 4 changes to keywords, including the three for "gismu", "tanru", and "lujvo" mentioned above as just approved. All 4 changes were instigated by Lojban learners who expressed confusion about the meaning of the word based on the keyword, and suggested a clearer word; none of the gismu meanings, as expressed by the keyword, have changed. Further changes are not expected.

The recent LogFest provided the first additions to the gismu list since the baseline. There are looser controls against adding words, since these cause no relearning; we've been surprised to go two full years (until the recently approved changes) with no additions at all. The couple of open issues mentioned above may lead to further additions, and there is an ongoing re-review of the gismu in thesaurus fashion that may reveal one or two possible words, but we don't expect many.

Two noted Lojbanists, Athelstan and Michael Helsem, have voiced the opinion that we should be much more willing to add new gismu to the existing set. Tommy Whitlock, on the other hand, is adamantly opposed to adding new gismu, since he thinks there are too many already. This balance of opinions among the most senior Lojbanists, and the increasing number of active students, means that additions will occur slowly if at all, and with extensive review from several members of the community.

rafsi - The rafsi list has proven as stable as the gismu list, even though it was not baselined. The reluctance to change anything that has been released via LogFlash lists has been enough to defer change proposals indefinitely. There was one change early on, when we changed the rafsi of "narge" to "-nag-", freeing "-nar-".

The recently adopted negation proposal split the negation cmavo "na" into three words, two of which needed rafsi. "- nar-" was assigned as the rafsi for "na", while "-nal-" was assigned to "na'e". All of the other grammar changes led to the addition of perhaps a half dozen rafsi assigned to cmavo, and a couple of the rafsi being freed. (See the summary with the JL12 attitudinal proposal for a list.)

There are two change proposals to be considered prior to the next LogFlash release, to make more 'hyphen-friendly' rafsi for the abstractors "ka" and "ni" which are being used more commonly in lujvo that require hyphen 'y' than originally estimated. Again, no other changes are planned, although we will be looking at the rafsi more thoroughly at the time the dictionary is written, when the rafsi list will be formally baselined.

We don't want to freeze the rafsi list until we've analyzed a lot of lujvo made by as many different Lojbanists as possible, thus minimizing the likelihood that usage statistics (applied to 'fine-tuning' the selections) are skewed towards words favored by one or two Lojbanists who do not know all the gismu equally well.

Place structures - Place structures will also not be baselined until the dictionary is written. A very slow review and rewrite of all of the place structures is in progress, and will be completed for the next release of LogFlash and the textbook. The new versions of place structures being examined will be expressed in greater detail than those included with the current gismu list, and should be easier to understand. (The maximum definition size will be 96 characters instead of the current 40.)

A change to the place structure is inherently a meaning change, and we try to avoid them. Almost all changes have been additions and deletions to the last trailing places that few have learned or used. More commonly, the changes are clarifications to better let a reader know what type of sumti value is expected in a sumti place.

Of all parts of the language design, the place structures are the least stable and finalized, but Bob can testify from his experience that only a few changes have ever affected anything he's written. (Most significant of these is the place structure of "bridi", which was modified to put the selbri in the 2nd place; the x2 place of the place structure for "bridi", as printed in the gismu list, makes no sense - a bridi is a sentence/relationship and has no 'meaning' independent of context. The change, of course, rendered "selbri" as the appropriate word for the concept, since the word relates by definition to the x2 place of "bridi". ("kunbri", for those who were not aware or have forgotten, is an artifact-word from before the gismu baseline, and should be replaced by "selbri" where ever you see it.)

Grammar - The Lojban grammar is defined by a computer-verified definition called the machine grammar. As reported above, this grammar has now been baselined. Even before the baseline, there had been only 3 or 4 changes in the last year as we wrote the draft textbook lessons and re-examined entire areas of the grammar. These changes were in the more esoteric portions of the grammar that were seldom being used, and text written in Lojban since around JL8 (including all of the draft lessons) is nearly as grammatical now as it was when written.

The new baseline is a 'soft' baseline, that will allow us to make minor corrections that show up in textbook writing. Few such changes are expected, but we may find that we can allow some constructs that are currently forbidden, as well as rules that we thought were in the grammar that were omitted typographically. These changes will not 'enter the language' officially until the textbook is published.

The strength of the grammar is that each of the special areas like negation, tense, MEX, and attitudinal indicators, has received thorough 'end-of-development' analyses that make it unlikely that the language will prove inconsistent or incomplete in these areas. As recently as last issue, we were uncertain that our MEX design would stand up to analysis; the final design proved quite versatile, and we believe that we have met the goal of being able to readily express any mathematical expression 'reading it off the notation'.

The weakness of the grammar is that we do not have a parser that reflects the final grammar; such a parser is needed to test the grammar against a corpus of prepared sentences, ensuring that the grammar breaks sentences up the way we think it does. We are thus forced to use the random sentence generator as a 'backwards test tool', looking at the sentences it generates from the baseline grammar, and seeing that they match our intended grammar.

Still, we don't anticipate that the grammar will change much before its final baseline when the dictionary is published.

cmavo - The cmavo list in this issue is the first complete one in two years, and reflects a lot of changes. There could be a more cmavo added prior to final baseline, but probably very few (there isn't a lot of spare words available for adding); we expect almost no changes other than additions. The fact that the list has been published will serve as a stabilizing factor.

The keywords for the cmavo are likely to change. Simply put, they haven't been looked at by many people, and are inherently less valid than the gismu keywords. Since cmavo have little semantic meaning, we have to use short phrases that don't say a lot to try to convey the memory hook that an English keyword is supposed to provide. We'd like feedback on the keywords, while recognizing that most readers don't know that many of the cmavo.

All in all, we've accomplished a lot in just 3 years. With your support, imagine what the next 3 years will bring.

Products Status

First a reminder that the discount policy has been drastically changed, effective 1 August. A flat 20% surcharge outside of North America; a 20% discount for a paid order (positive balance exceeding the price at the time of shipment) over $20. (The discount will cancel the overseas surcharge.) Virginia orders should add 4.5% sales tax. Note also that for software, there is no surcharge for MS-DOS 3 1/2" diskettes, but you must specify in your order if you want them.

Remember that we cannot promise to fill your order unless it is prepaid; our finances are too thin right now.


Textbook Status - Believe it or not, the textbook is finally started (again). Spaced around 3 issues of JL and one issue of LK in 3 months, planning and conducting LogFest, researching MEX and tense grammar in time for the grammar baseline, and assisting John Cowan in assembling the gismu list, along with some major work on our legal battle, Bob finally sat down at the keyboard shortly after midnight on July 4.

The new version of the textbook is already unrecognizable as compared to the first, even though only 20 odd pages are written so far. The first lesson, which will serve as a language overview, is divided into short sections only a page or two long, with many examples and exercises in each section to help you see whether you understand. Using a much smaller vocabulary (perhaps only 25 gismu in lesson 1), we will examine much more of the basic grammatical features of the language. By page 20 you will be making Lojban sentences, hopefully with little trouble; in the draft lessons you did not make sentences until late in lesson 2 - about 80 pages along.

The textbook will be more interesting to read. We are trying to put more interesting examples in (difficult with very little vocabulary). Several pages have boxed and highlighted recaps of the key points of the text. Many of the most significant and unique features of the language will be touched on by the end of the first lesson. You will know that Lojban is a truly different language quite quickly. The text also ties back to English examples, helping you understand better how English works, based on a comparison with Lojban. Thus, even if you never find a practical use for Lojban, you will receive benefits in terms of expressing yourself better in English (and any other language you learn).

Under the current outline, the text is divided into three parts. The first part, which will consist of an introduction and the single overview lesson, will tell you about Lojban, lightly introducing the basic concepts and giving you the 'big picture' of the language. Some areas are treated very lightly - pronunciation is conveyed only by guides that tell you how to say each word and sentence. This is because pronunciation is a 'big subject' and a very boring one to start off with. We want you to be motivated to speak the language, not bored. Some of the topics are: the concepts of bridi, cmavo, selbri, tanru, sumti, place structures, conversion, ellipsis, elision, descriptions, abstractions, questions in Lojban. We include a brief summary of several other unique features that are too difficult to cover in the first lesson.

Part II will be about ten lessons long, each much shorter than the draft lesson size. This part of the book will build depth on the basic concepts presented in the introduction and explain many of the secondary structures that you need to say what you want in Lojban. The Part II lessons continue to use a much smaller vocabulary than the draft lessons (perhaps 300 words), but expect you to look up some words that are not formally part of the vocabulary to be learned.

A major change is that we will not expect you to learn most of the gismu vocabulary within the first 8 or 9 lessons. While some people have demonstrated that it can be done, most students in the classes have not kept up with the expected pace. The current plan is to add an extra stage in LogFlash, before Gaining Control, that exposes you to a lot more words quickly, but does not expect you to master them. There will be only 'New Word lessons', 'error practices' and some brief reviews in this introductory mode - the object is to have you quickly able to recognize more words in Lojban text and to learn the scope of the vocabulary. Many who have learned part of the vocabulary have tried to write sentences, but have not been able to find the right word because they didn't know it was there. You should do better using this modified technique.

The goal is that by the end of Part II, you will have completed this initial review of the words, and will be started in 'Gaining Control', which will hopefully go much smoother for you as you work through Part III. We are hoping to shrink the huge demoralizing bubble of error words in the 'Failure Pile' that seems to afflict many people.

Following are the topics to be covered in Part II, according to the current outline. This is of course subject to change as the book is written:

Lesson 2 -     Pronunciation and Word Forms Some Classroom 
       Expressions					   
Lesson 3 -     Learning	Vocabulary; Simple tanru           
Lesson 4 -     Making Names				   
Lesson 5 -     Numbers					   
Lesson 6 -     tanru and lujvo;	selbri Structure	   
Lesson 7 -     sumti and Place Structures; Relative Clauses
Lesson 8 -     Tenses and Modals; se tcita sumti	   
Lesson 9 -     Logical Connectives and Negation		   
Lesson 10 -    Discursives and Vocatives		   
Lesson 11 -    Keeping Lojban Unambiguous and Clear	   

By the time you start on Part III, you should know some 300 gismu and 50 cmavo by actually having used them in sentences. You will not be expected to produce longer text than single sentences. In Part II, grammatical features will be pretty much covered in isolation to help you recognize the key point of each section; this is a bit like Jim Brown's technique in Loglan 1, but the earlier introductory lesson will allow us to keep the concepts tied together much better than he was able to - you'll know the destination while travelling a most interesting journey through the language.

In Part III, we will start presenting longer texts and dialogues, which will have enough vocabulary available to be meaningful and adult. Exercises will require you to more spontaneously produce original Lojban sentences, especially in a classroom or study group. Unlike the draft lessons, though, you should have the knowledge and confidence you need to make up sentences by the time we ask you to do so. In classroom use, bits of Lojban conversation should start occurring.

After the first few lessons in Part III, the remaining lessons will be less oriented around specific concepts in the language than Part II. Instead, we will explore the vocabulary associated with some topic, present some of the more esoteric grammar points that are useful for talking about that subject, and then use the language to do just that. Some problems in translation and original composition in Lojban will be covered.

Whereas Part II is called 'Learning Lojban', Part III is called 'Using Lojban'. You will be expected to write and/or converse in the language throughout Part III, and should be comfortable doing so by the end of the book, with vocabulary limits as your main constraint.

There is a lot of writing ahead, but the book is off to a good start. Moreover, with the baselined grammar and the compiled cmavo list, we are much more confident that what gets written will not have to be continually rewritten.

We're not going to promise a publication date. As mentioned above, we've learned a bit about the lengthy process of publishing a book if you want to make money at it (and we can't afford to lose money).

What we can promise is that the book will be available to the Lojban community (you) in advance of the official publication date at a substantial discount. This helps us in that we pay for the printing bill without waiting for distributors and vendors to pay us, and it rewards you for sticking with us up until now. But to receive the best discount you must have a positive voluntary balance.

Another progress report will be given next issue, and perhaps we'll be able to guess at a date by then.


Other Products - With the baseline of the grammar and the preparation of the cmavo list, we are moving forward on a variety of teaching products.

Most directly dependent on the grammar and the cmavo list is the random sentence generator, which will also be used to test the grammar. It takes only a few days to incorporate the new standards, and we will probably have an update available by 1 October. The update price will be $10; due to our financial situation, we can no longer provide updates any cheaper than this. The original price will be $12.

The lujvo-maker has now been completed, providing drills and demonstrations of both lujvo-making and decomposition. Updates are available for $10; the original price will be $12. The only future enhancement to the lujvo-maker that seems to make sense would be a feature that builds and tests le'avla (borrowings) for proper structure. Because the lujvo-maker does not take a lot of space on a 360K floppy disk, we will include computerized text copies of various word lists and the grammar on the disk.

The only stalled product based on the grammar is the Lojban parser; we have had no volunteers to complete the work. There is a possibility, however, that Jeff Prothero will soon have a new version of 'PLoP', his "Public Domain Loglan Parser" updated to the Lojban grammar. This would be an 'unofficial' parser, using the YACC grammar, but not the YACC algorithm. PLoP is of a type called a 'recursive descent' parser, which is more flexible than a YACC parser, but can be much slower. It will work fine on individual sentences up to some length, taking at most a few seconds, but it is very slow on blocks of text. By comparison, the last version of Jeff Taylor's parser processes a full page of text in a couple of seconds, albeit based on an older grammar.

A new version of LogFlash for MS-DOS machines will be prepared and hopefully released this fall, and it will be significantly enhanced. First of all, we are already testing 'LogFlash 3', which teaches the cmavo (we can provide this test version now to people who are ready to learn the cmavo and don't want to wait for the full release, but please don't ask for it unless you are ready to use it).

Second of all, we are adding an initial review stage prior to 'Gaining Control' that will quickly expose you to a lot of words in 'New Word Recognition mode', hopefully allowing you to read language text earlier while enhancing scores when you advance to the more difficult stages.

We are also adding several user-friendly features. First will be an installation program that will ensure that floppy disk users have 'COMMAND.COM' available, and unpack any packed data on disk automatically. LogFlash will ask you for confirmation before overwriting an existing file. It will allow you to tune the program by changing the default of 6 repetitions in error practices and/or the number of review words from the 'Under Control' pile that are presented in each session. It will also allow you to more easily switch between learning modes. We will also be try to make large error practices more friendly, giving you a sense of progress by telling you how many times you have done a word successfully. An option we are considering will allow you to look at the entire list of words for a lesson prior to taking the test. We may also allow you to add or update your own memory hook data during the review portion of a lesson. (We welcome additional suggestions on how we can make the program more friendly, but we need your responses quickly.)

We want to increase the speed of the program by reducing the amount that it has to read files from disk. We aren't sure how much work this will be, and are not making specific promises.

Finally, we will also be adding instrumentation that will allow LogFlash to be used for scientific research into how different people learn words, most especially to see if recognition scores used to make Lojban gismu have any correlation with actual learning rates. Depending on finances, we may be offering a volunteer credit to anyone who (learns the words and) returns their instrumentation files within a specified time after we send you your order (instrumentation data is only useful to us if you work at LogFlash consistently - the time limit will help motivate you to keep at it).

Finally, the new LogFlash version will support the updated gismu list. The gismu list as it is being revised will allow 20-character keywords instead of 15, and the definition field will be 96 characters, instead of the current 40 character limit. As a result, the definitions will be much clearer and you should have a better idea what a Lojban word really means. We may be providing an editable hint field that you can use to add mnemonic aids, and we may also allow you to change the keywords from their official values to something more memorable (most useful to British users who have suffered American spellings for too long).

Not included in this update are LogFlash versions that will teach place structures and grammar (although learning the modal cmavo of lexeme BAI will teach you a lot of important place structures). Maybe next year.

We are hoping to have Eric Raymond's UNIX version of LogFlash available by the end of the year. Unfortunately, due to the lag in development time, both the UNIX version and the Mac LojFlash version will be stuck with the current file formats for a while. Dave Cortesi plans to update his Hypercard flash program for the MAC to use the new file formats; if so, this program will also be available at approximately the same time as the new version of PC LogFlash.

We are undecided about whether to produce and distribute the revised gismu list separately, as well as John Cowan's cmavo dictionary, or wait a few extra months and try to get it into a first edition dictionary format. The textbook effort will probably be the determining factor. Publishing these revisions will be lower priority, since you can use copies of the existing lists for most purposes. The revised lists would also be very expensive - the bi-directional gismu list might run up to 60 pages, and 100 pages if we add a thesaurus-sorted list. At those sizes, it would be cheaper to produce the dictionary; we don't have the money to maintain such large documents in inventory.

We also want to publish a 'tiny gismu and cmavo list', having only keywords on the English side, that would be small enough to carry in your pocket or purse.

Of course, the textbook and dictionary will be the centerpiece of our product line. We hopefully will be able to follow up these two books with a 'Best of JL' and a first book of Lojban writings during 1991 - but this is again dependent on money.


International News

We don't have a lot of news this time that is specific to the international Lojban community. We have now successfully processed Master Card/ Visa orders from overseas, as well as one larger Canadian-denominated check. We ask that those of you sending checks from Canada clearly indicate whether the check amount is in Canadian or American dollars; apparently some of your banks will issue U.S. dollar checks, and we do not need to use the more expensive service to process them. We can't be sure, however, that our banks here will process your check correctly if you are non-specific as to the currency.

Our major international event in the next few months is Athelstan's planned visit to Europe. On that, we have worse than no news. Just as we were going to press with this issue, some problems came up that threaten whether the trip will take place as planned. It is possible that the trip may be delayed (causing Athelstan to miss Worldcon at the end of August), and possibly even cancelled.

Assuming the trip does come off as planned, we have few additional itinerary details - only two of you have written to us letting us know you want Athelstan to visit. Athelstan does now have a point of contact in Europe. He plans to visit Peter and Mary Lynn in Goettingen, West Germany around the first week of September. If you haven't made contact with Athelstan by writing to us here, you can contact the Lynns:

Peter and Mary Lynn
Schopenhauer Weg 13
Goettingen D3400 FDR
WEST GERMANY

home telephone: (49)-551-706485

Peter is also working on a German translation of the brochure, which may or may not be ready by Athelstan's visit.

Athelstan also plans to be in northern Italy around 21-23 September, visiting Lojbanist Silvia Romanelli (who also reports having translated some of the draft textbook lessons into Italian).

News From the Institute

The Loglan Institute published another Lognet around mid-July. Dr. Brown has apparently assumed editorship with Rex May's resignation reported last issue.

There is little news in the issue. The issue was dominated by half of a paper by Rex May challenging some of the basic design points of the language. The second half of the paper is supposed to be in the next Lognet issue, and will present Rex's proposal for radical changes in the Loglan morphology. (Rex sent a copy of his paper to Bob for comment independent of submitting it to Dr. Brown.)

Dr. Brown discusses and responds to each of Rex's major points. Bob observes that Dr. Brown's discussion is an excellent defense of the basic language design, providing a few previously unknown historical details about the language design process. Most important of these is the revelation that Dr. Brown did conduct some 'engineering tests' of the recognition scores algorithm used to make gismu, something we had no evidence of when we responded to Sheldon Linker's questions on the subject several issues ago. While we could wish that such tests were better documented, it is reassuring to be able to say to critics that they were conducted. All in all, our plaudits to Dr. Brown.

Dr. Brown also reports on solutions to two outstanding morphological issues in Institute Loglan; these problems were discovered by Nora and raised in Bob and Athelstan's review of the 4th edition of Loglan 1 last year.

Nora is skeptical that the solution devised for names containing "la" will work universally, but with the Institute grammar a 'trade secret', it is impossible to analyze the current design. (Unlike Institute Loglan, Lojban forbids the various name markers from being embedded in names to prevent such problems.)

The solution to the other problem, that of hyphenating borrowings, is similar to our own solution for Lojban.

There is brief mention that the Institute plans to revise The Loglanist under a new name, possibly by the end of the year. There also is a report that one of the Institute's software packages had a bug that is now fixed, but there were no details given.


JL11 Esperanto Discussion

A Response from Don Harlow

[Don Harlow is editor of the Esperanto League of North

America newsletter. His position makes him a natural spokesperson for the Esperanto community in responding to our essays in JL11. However, see also Ralph Dumain and John Hodges in the 'Letters' section below for more comments on Lojban and Esperanto.]

Thanks for the latest copy of Ju'i Lobypli. I was particularly interested in Athelstan's comparison of Esperanto's "16 rules" with a similar set of rules for Lojban.

Athelstan is quite right in suggesting that "the rule set is incomplete." In fact, the "16 rules" are largely a heuristic device created to introduce Esperanto to persons with a late 19th-century European education, by describing Esperanto in very simple terms relating the language to something more familiar to the student -- i.e., the Indo-European languages. This can be seen by the reference in rule 2 to the "two cases of Esperanto" (Esperanto has as many cases as any other language), the reference in rule 6 to the passive voice of verbs formed by compounding (there are no compound verbs in Esperanto), by the reference to the "imperative mood" in the same rule (the -U ending subsumes, but is hardly restricted to, the traditional IE imperative), and particularly by rule 8; logically, prepositions (which are basically case-forming morphemes) should govern an unmodified noun form, and it is only because of the contrast with the Indo-European languages, where they usually do not, that this rule is necessary.

The so-called "Fundamento de Esperanto" is, in fact, about 200 pages long, and includes the "16 rules" (repeated in five different languages), a complete dictionary of some four thousand roots -- an additional four thousand or so have been added to the canon since that time, plus between eight and sixteen thousand unofficial roots that need not be considered part of the language -- and a series of some 42 exercises designed by Zamenhof to demonstrate aspects of syntax and the Esperanto word-formation system. The "16 rules" themselves are, as I say, a heuristic device, and a convenient skeleton on which to hang the language's "flesh." Most of the material in these rules would, today, be better presented in tabular form.

A few points about Athelstan's presentation:

  1. Athelstan does "not describe word or sentence order...." This seems a bit ingenuous to me, since as far as I can tell word and sentence order play a more significant role in Lojban than they do in Esperanto, and so to describe a "set" of Esperanto rules and equate them to a single Lojban "rule" that is at a much higher level is not quite cricket. An example is rule 3. The Esperanto presentation of the morphology of the adjective is quite complete in four lines; the Lojban presentation says only that "any selbri may modify any other selbri by position," but does not define how this is done (do selbri modify other selbri preceding them? by following them? by sitting in the next line up?) This is like saying that Lojban code is more concise simply because the reader is presented only with a subroutine call, while in the Esperanto code the reader is shown the entire content of the subroutine. The content is there in Lojban; Athelstan has merely found it convenient to overlook it.[1]
  2. Granting Athelstan's contention that several of Esperanto's "single rules" contain other rules, he does himself the favor of counting some of those sub-rules more than once, if they are referred to in another "super-rule." For instance, he counts the rule that the direct object is shown by affixing the -N ending at least three times (rule 2, rule 3, rule 5). The computer equivalent would be rewriting the subroutine each time it was called -- at which the compiler would, no doubt, burp.
  3. Given that Esperanto's "16 rules" are a heuristic device, they are certainly more complete and successful than those presented by Athelstan for Lojban. Speaking "quantitatively," they are accessible to a much wider range of people than the Lojban rules. The Esperanto rules refer largely to nouns, verbs, adjectives, past tenses, etc., which are terms that are generally recognizable to graduates of the seventh grade, or equivalent (my ten year old daughter is familiar with them, from school). Athelstan's Lojban rules, on the other hand, use unglossed terminology that might confound a college graduate -- anaphora, non-veridical, place tags, etc. (I consider myself moderately well educated, but I had to look up "anaphora" in a dictionary -- and was not much wiser for the experience.)
  4. Speaking "qualitatively," Athelstan in many places describes his Lojban rules using Lojban terms that will have no meaning to the casual reader -- a rather recursive sort of action, if you ask me. "Lujvo are formed by simple junction of the gismu or rafsi??? The definition of each one of those terms should be counted as a separate rule (axiom, if you will).
  5. Some comments on individual rules:
    1. The description of participles in Esperanto rule 6 is not properly part of this rule but belongs in the hidden (also for Esperanto!) working of word-building (rule 11 see below); the description of the passive voice properly belongs to the Ekzercaro. I do not, however, fault Athelstan for taking these items as he found them.
    2. "Every word is pronounced as it is spelt." Pardon me for referring to Loglan rather than Lojban -- and if this is not also true for Lojban, you need not pay attention to this comment -- but this is not completely true for the language. Loglan treats the sound written in English as "CH" as a stop "t" followed by a fricative "sh", written "tc," rather than as, more correctly, a single harsh fricative halfway between the stop and the fricative. Brown was here apparently influenced by the (not invariably phonetic) International Phonetic Alphabet, which in this case appears to have been heavily influenced by French. Esperanto more correctly treats this single sound with a single letter. I am not sure whether Loglan and Lojban treat the single sound written in English as "ts" as two sounds (again as a stop followed by a sibilant, rather than as a single harsh sibilant) or as a single sound/letter ("c") as in Esperanto. (A similar use of two letters to designate an intermediate sound is the occasional use of "kh" in English to describe the Esperanto "h^", a sound intermediate between "k" and "h".)
    3. Esperanto's Rule 11, of course, refers to the Ekzercaro -- see particularly Exercise 42. Athelstan refers to some sort of "variant compounding rules"; I would be interested in seeing these. The actual rules describing the word-formation system are neat but complex; they were first formulated as late as 1910 by de Saussure, writing under the pen-name "Antido", and expanded by Kalocsay in the 1920's in a well-known essay. The latest set appear in the Plen Analiza Gramatiko de Esperanto (1985 edition), where they fill some 148 pages and differ little form Kalocsay's earlier rules. That these rules are of little use and less interest to the practicing Esperantist can be seen from the fact that their earliest codification occurred some 23 years after the language began to be spoken; most people can figure the system out after looking at a page or so of examples, and never bother to refer to the rules, to which they don't have access anyway.[2]

Unfortunately, a couple of Athelstan's comments suggest that he isn't really qualified to comment on Esperanto in general, any more than I am on Lojban (which is why I keep correcting you on Esperanto rather than commenting on various points of Lojban grammar, syntax, etc.). For instance, on p. 20 he refers to "Esperanto's dependency on case declensions." There are no declensions in the traditional/IE sense in Esperanto. The -N ending, to which he is probably referring, defines the target of an action (direct object) or, if no action is committed, the destination of a movement[3]; it can be applied to adverbs as easily as to nouns and their accompanying adjectives. Again, the terms "nominative case" and "accusative case" in this sense are sops to Indo-European sensibilities; Esperanto has neither one in the narrow sense of a declension. In the broader sense, of course, it does have nominative and accusative cases, as do English, Chinese, or -- one presumes -- Lojban; it also has genitive, dative, instrumental, ellative, terminative, sociative, etc. cases, as do English, Chinese, and -- I again presume -- Lojban.

Regarding your own essay "On Comparing Lojban and Esperanto" let me make several short (I hope, as, I am sure you do) comments:

  1. Under "aesthetics" you mention a couple of sentences that "are longer than the colloquial English translation"; and in an earlier issue you begged off translating a song from English into Lojban because the translation would be longer than the original. This seems to me to be an acceptance of the old saw that "any translation into any other language will average about 25% longer than the English original" -- and (a word to the wise) it seems to be a very dangerous attitude to take.[4] Every translation I make into Esperanto from English comes out significantly shorter then the original. More than that, so far as I know a competent translator can get the same results in just about any language going. I would hope, for the sake of Lojban, that this "expansion effect" is a function of the translator rather than a function of the language. If not, it is a strike against Lojban.
  2. You have again quoted the "like it is done in your own language" comment, which was not made by Zamenhof, but in the basic Interlingua textbook of 1950!!! Esperanto is extremely well-defined, partly through the 16 rules as described above, but mainly through the Ekzercaro, which also appeared in the Unua libro in 1887. No reference to outside languages was or is necessary. I thought we'd been over that ground before! As to the Europeanness of Esperanto ... proof of the pudding. Esperanto's greatest successes in the past few years have been outside of the Indo-European language area. (From May to October of this year, a nationwide Esperanto course is running on Chinese television -- a more significant matter, I think, in a country with only one national TV network instead of four or five, and no more than two or three channels in even the largest cities.)
  3. The comment that "Lojban took 35 years to reach a point of development where it was speakable" might perhaps have been avoided. Esperanto took some 12-14 years to reach the point (1887) at which Zamenhof considered it optimal; but the Ur-Esperanto of 1878 was already speakable, at least according to the anecdotal information. That it took Lojban (I presume you mean Loglan) 35 years to reach the point at which it was speakable is not, I think, a point in its favor as a means of communication.
    The rapid growth of Esperanto in its first years after public release was a spontaneous affair. You quote a figure of 8% a month growth in the number of Lojban students. Based on Zamenhof's published address lists -- and making a conservative assumption that only ten percent of those who claimed to be able to speak Esperanto could actually do so -- in the first half year after Esperanto's publication, the number of Esperanto speakers grew at a rate of more than 100% per month. (This high figure, of course, like your own, comes from starting with such a small base; and it dropped considerably by the early 1890's)
  4. You attribute some significance to the fact that you "NEVER [HEARD] A SINGLE CONVERSATION IN ESPERANTO" at the Esperanto table at Worldcon. I personally have met only one of the people who worked at that table (and he was there for only an hour or so), and I know that he speaks fluent Esperanto; I can't answer for the others. But when you've sat at a few more tables at conventions, and have carried on a few conversations in Lojban under such circumstances, you will learn an interesting fact: more people -- or at least Americans -- are repelled when they hear a conversation they don't understand than are attracted.[5] When possible, I always use English under such circumstances. (This is not always possible; at the last three conferences of the Foreign Language Association of Northern California that I've attended as an exhibitor, my co-exhibitor and I have spoken nothing but Esperanto -- because he's a Rumanian, and not terribly comfortable in English.)

Hope that you have found all this of some interest.


Bob responds - That the 16 rules are intended only a heuristic device seems to be lost on many Esperantists, who often try compare the 16 rules to our set of YACC rules, which number about 550; Athelstan's effort was an answer to those critics. See Ralph Dumain's discussion and my response in the letters section below for more on this.

Don effectively supports our assertion that the 16 rules have as a subtext the entire grammar of European languages. "The Esperanto rules refer largely to nouns, verbs, adjectives, past tenses, etc., which are terms that are generally recognizable to graduates of the seventh grade, or equivalent". But these terms are only recognizable to students of European languages.

The emphasis should be on 'student', by the way. While Don's 10-year old may find the terms familiar, we have found college graduate English speakers who have long since forgotten the terminology of grammar classes. To many of our audience, 'noun' is as bad as 'anaphora' (maybe worse, since no one feels guilty that they don't know what anaphora are. Anaphora are, by the way, the superset of 'pronouns' - the things that stand for and refer to earlier referents in the discussion; 'cataphora', the opposite term, cover variable words that refer to things in future discussion, but 'anaphora' also is used as the general term covering both sets of variable reference words. Based on Don's comment, however, we will start using a Lojban lujvo "ba'ivla" - /bah,HEE,vlah/ for the general concept of 'anaphora'; the source metaphor 'replacer-word' should help people remember what the word means).

Athelstan intentionally used specialized Lojban terms that were as opaque to a European language speaker as they would be to a speaker of a non-European language. This may help point out what a Chinese or Swahili speaker suffers reading the Esperanto rules. We don't seriously intend using the 11 Lojban rules as a heuristic device; as Don says, they just aren't very understandable. Furthermore, they cover no more of the Lojban grammar than the Esperanto rules cover of its grammar. However, they do help point out some ways in which Lojban is similar to European languages, including Esperanto.

I remain unconvinced that Esperanto's grammar is unlike Indo-European languages. As an example, contrary to what Don implies, the number and specific cases in a language are not universals, and are significant aids to classifying them. That a language has 'nouns' and 'verbs' and 'adjectives' that work in ways familiar to us, that most sentences have a 'nominative' agent case as the subject, usually appearing before the verb, and an 'accusative' object case that usually appears right after the verb. These are anything but universal, though they are found in most, if not all Indo-European languages. Many languages have no nominative or accusative cases, being organized around cases called 'ergative' and 'passive'. Some languages do not even have a clearly identifiable subject, and Japanese has both 'subjects' and 'topics' that each serve some of the purposes of the Indo-European 'subject'.

Now what Don says later about the "-N" ending could be used to argue that Esperanto's cases are different from the Indo-European ones, but by standard linguistic terminology, that ending is a 'declension' that marks its word as being in a case (grammatical role) which differs from the grammatical role it would be in if the declension were not present.

Lojban has NO grammatical cases. Linguists and artificial intelligence people can assign 'case labels' to the various sumti places in the structure, but these are not grammatical cases. They are semantic cases that indicate the semantic relationship between the place and the rest of the sentence. In Lojban there are as many potential semantic cases as there are words in the language - an infinite number. The places defined in the place structure are merely those most essential to conveying a relationship. We list the places in the definitions of the words partly to remind people that Lojban bridi express relationships, and to remind them of the essentials of the concept to be related.

In one sense, Lojban doesn't even have a 'subject'. Technically, all of the sumti places are 'objects' that are related by the selbri. However, in at least two ways, the 1st (x1) place of any given bridi predicate, whichever of the sumti it happens to be in a given arrangement, has a unique role among the places which might as well be labelled as 'subject', for consistency with the terminology of linguistics. We'll let linguists determine if the x1 sumti really is a 'subject' in the traditional sense, or whether another term better applies.

Now it turns out that many of our relations resemble European languages in that the first place is often an agent and the second place is an object. This may represent a European bias, albeit unintentional. The intent is to include places in approximate order of frequency of use in discourse; our model for usage frequency is unfortunately the English language we hear most often. The desire to bring in a broader perspective before finalizing the structures is one reason why we are avoiding baselining the place structures until the last possible minute, and why place structures will be among the first things to be re-evaluated after the 5-year freeze.

In any event, the resemblance does not give Lojban the Indo-European cases of Esperanto. There are no case endings, no grammatical requirements such as that adjectives must 'agree' with a particular case. We have 'case tags' in Lojban, but these are optional and even frowned upon for 'cases' in the place structure, and anyway resembles a combination of 'prepositions' and 'adverbs' more than case inflections on words. (They also resemble what Don calls 'case-forming morphemes'; however, in Lojban they are separate words that do not 'govern the form' of any other word.)

Lojban has no 'passive voice' either - a 'passive voice' is an artifact of Indo-European grammar which is used less in English and Germanic languages than in other European languages. In Lojban, there are various methods of rearranging the sumti places of a predicate. One might label any arrangement that doesn't have an active agent in the x1 position 'passive', but again, this isn't the same as the European 'passive voice'. (See B. Comrie's books The World's Major Languages and Language Typology and Linguistic Universals for excellent discussions of the typological features of language.)

Lojban is distinctly different from any natural language in several ways. The first step in learning Lojban, therefore, involves stepping out of the constraining ideas of natural language to learn these new concepts. Once that is accomplished, then for European speakers, Lojban is probably comparable in learning difficulty to Esperanto; Lojban has a somewhat simpler grammar, but Esperanto's roots are more highly recognizable to Europeans (and English speakers). For Chinese speakers, Lojban may actually be easier, since many features of Lojban's grammar at least superficially resemble Chinese features.


"Athelstan ... describes his Lojban rules using Lojban terms ... The definition of each one of those terms should be counted as a separate rule (axiom, if you will)." - Should the definition of each of the Indo-European grammatical terms used in the Esperanto rules have also been counted as 'axioms'? If so, I think Esperanto comes out far the worse for the added criteria. The number of specialized Lojban words we need to discuss the grammar is fewer than the number of words needed to discuss a European language.

"Athelstan does 'not describe word or sentence order....' This seems a bit ingenuous to me..." - There are two types of word order that can be talked about. The order of words of particular grammatical type in a sentence is specified by the entire set of rules of the grammar. There is no meaningful 'rule' or 'rules' that govern this kind of word order. The order of the places for a given brivla, on the other hand, is not a grammatical issue in Lojban at all, unlike European languages and Esperanto (I understand that Chinese is also relatively free in word order).

Thus, Athelstan did not discuss word order because it is not part of the Lojban grammar. The order of the places is part of the semantic meaning of each word, just as the meanings of 'subject' and 'object' for each Esperanto verb are part of the meaning of that verb. From our perspective, such semantic rules are at a lower level of the language than grammatical rules. Lojban has no higher level rule that can be said to govern the order of places. There may be some patterns, but we haven't really tried to find them.

"The Esperanto presentation of the morphology of the adjective is quite complete in four lines; the Lojban presentation ... does not define how this is done (do selbri modify other selbri preceding them? by following them? by sitting in the next line up?)" - The Lojban 'morphology of the adjective' is complete in zero lines, since we don't have adjectives. selbri modify other selbri in many ways, some of which are adjective-like. The modification can be left-modifies-right or right-modifies-left, logical connection, or non-logical connection. In all but the simplest left-to-right modification, there are cmavo that can be translated literally into English or other languages, revealing the order, and we believe that all possible orders and groupings can be represented in some way. Athelstan simply didn't find anything to say about Lojban that corresponded to what was being said in the Esperanto rule. What he said was complete and accurate - position in a Lojban sentence totally determines what modifies what.

As for Don's facetious suggestions on how selbri might modify each other by position, I reply in kind: do Esperanto adjectives get written on the line before?

Interestingly, in other places, Don excuses his 16 rules for non-specificity: "the description of the passive voice properly belongs to the Ekzercaro" and talking about word-formation rules "they fill some 148 pages". Again, our purpose was to compare what was present in the Esperanto rules with a corresponding level of detail about the Lojban rules. We recognize that neither set of rules is complete; we want to be able to point this out to Esperantists that cite the 16 rules as a statement of Esperanto's simplicity. So Don has made our point for us.


"Most Esperantists ... definitely obey a particular rule of word-formation ... -- one that, so far as I know, has never been written down, and would be difficult to codify in a few simple sentences." - Hopefully Lojban is sufficiently regular that no one ever will have to say this about the language. Our word compounding rules are quite rigid, and yet fairly unrestricted. We don't constrain any word from modifying another, and provide some fairly esoteric grammatical conversions to allow you to combine concepts that are grammatically incompatible.

"Athelstan refers to some sort of 'variant compounding rules'" - I believe Athelstan was referring to the extensive set of additional rules, not conveyed in the set of 16, that take 148 pages to describe, as well as rules such as the ones Don describes as not written down.


"... he does himself the favor of counting some of those sub-rules more than once, if they are referred to in another "super-rule." - Athelstan was merely trying to show that the 'super-rule' grouping concealed the true rule count. The exact number of rules, I'd hoped we had demonstrated, was quite irrelevant. Lojban's 550-odd stated rules, by the way, are expanded by YACC into about 800 unique computer-labelled 'states' which correspond to expanding and repeating each of the 'subroutines' Don refers to as often as is necessary.

A Lojban-based computer process does not choke on such expansion, since the expansion is a natural product of YACC. When we say Lojban is grammatically unambiguous, it is because in each of these 800 states, by looking at the next word only, a Lojban processor knows what state to go to next. The grammar process consists simply of jumping from state to state until the end is reached.


"Loglan treats the sound written in English as 'CH' as a stop 't' followed by a fricative 'sh', written 'tc,' rather than as, more correctly, a single harsh fricative halfway between the stop and the fricative. Brown was here apparently influenced by the (not invariably phonetic) International Phonetic Alphabet, which in this case appears to have been heavily influenced by French. Esperanto more correctly treats this single sound with a single letter..." - Correct by whose standard? (Correctness always has a standard, as any Lojbanist knows from the place structure of "drani"). The IPA is the standard alphabet of linguistic phonology, and hence is the way that one must describe sounds when talking to a linguist. To claim that the linguistic standard phonetic alphabet is wrong because it doesn't agree with Esperanto seems a bit backwards.

The combination of a stop and a fricative is called an 'affricate' and can be treated as either one sound or as two. In Lojban, we treat all affricates, including 'tc' and 'ts', as two sounds; so do most linguists.

This is due to the simple reason that if you say the stop and the fricative together, they phonetically blend to form the affricate in a way indistinguishable to most listeners. Thus, if we were to write the affricates as a single letter, we would have to forbid the two-letter combinations that are equivalent. Since no other single letter sound in Lojban can alternatively be expressed as two sounds, to match the Esperanto distinction in only a couple of cases would be inconsistent. (Does Esperanto forbid the two-letter equivalent combinations of the affricates to prevent confusion?)

Esperanto's approach causes untold heartache to typists, forcing the addition of non-standard diacritical marks to several letters to fit the language within the Roman alphabet. (There is at least one typo in the Esperanto rules because of this - I forgot to manually go back and add an Esperanto diacritical mark that is not supported by my word processor or printer.)

Esperanto is not consistent on the matter of the affricates, by the way. While representing the affricate sounds that are expressed by Lojban 'tc' and 'ts' with a single letter, as well as the voiced equivalent of the first ('dj' = English 'j'), Esperanto does not have the voiced equivalent of 'ts' as a single letter as consistency would require. The sound of 'dz' in it is expressed using two letters in Esperanto words (an example is found in one of Don's footnotes), even though it is a 'single sound' by the identical logic as the other three.

In Comrie's book on the languages of the world, similar comments to mine are made in explaining why 'ts' and others are not considered as one in Germanic languages. It is pointed out that linguistically, any stop can be combined with any fricative, and each such 'affricate' combination could be treated as one sound or as two. Examples include 'ps', which will be recognized from Greek, and 'pf' from German. But neither Esperanto nor English nor Lojban treat 'ps' or 'pf' as a single sound.

Don is wrong in equating the 'kh'/Lojban 'x' sound with the two affricates. 'x' is a pure fricative - called an 'unvoiced velar fricative' or an 'unvoiced palato-velar' fricative depending on exactly where the tongue is placed (these are the sounds of German 'doch' and 'ich', respectively). The 'x' sound linguistically has nothing to do with an 'h' sound, which is actually formed in the epiglottal region. That we represent 'x' as 'kh' in English is a convention; it has nothing to do with sounds (notwithstanding this, trying to combine a 'k' with an 'h' will give a reasonable 'x' sound).

Unlike English and German, IPA does use a single letter for this sound. (The true velar affricates - combinations of stops and fricatives - aren't pronounceable either as single or double sounds for English speakers - in Lojban, they would be expressed as 'kx' and 'gq', if 'q' is defined as the voiced equivalent of 'x' - found in Arabic as the sound at the beginning of Libyan leader Qaddafi's name.)


"... the old saw ... 'any translation into any other language will average about 25% longer than the English original' ..." - Almost any literal translation will take longer than the original. Translating Lojban to English literally is usually even more expansive than 25%, often 2- to-1 or greater; just look at any of our translations here in JL. On the other hand, the reverse direction gives the same result.

The translator's art involves producing idiomatic non-literal translations that capture the approximate sense of the original. This will sometimes be shorter, sometimes longer, since the source language may be using an idiom that has no counterpart in the target language (which is always the case with Lojban at this point). Also, almost any culturally-based word has to be expanded into a phrase in another language if meaning is to be preserved. If Don is 'always shorter' as he claims, he is undoubtedly omitting subtleties of the source language version that he considers either obvious or irrelevant given the context. If he is correct, he is a true artist; otherwise, his readers are missing useful and perhaps important information.

In Lojban, there are other factors, based on its unusual grammar. Where logical structure is always explicit, the convoluted logic of some English sentences has to be expanded to great length; on the other hand the English "it is not the case that" is expressed briefly as Lojban "na". When Athelstan translated Saki (see JL10) he found the resulting text was about the same length or shorter. (There are actually more words, since Lojban words seem to average about 30% shorter than English words; there are also more syllables - Lojban words seldom have syllables more than 3 letters and certainly not as long as 'strengths'.)

I doubt that Don's objection to the old saw proves true for all languages, by the way. I suspect that regardless of the translator, most Romanized Chinese (where most words are one or two syllables) translates to Russian (with inflectional suffixes that are one or two syllables long on most words) resulting in a longer text.


"That it took Lojban 35 years to reach the point at which it was speakable is not, I think, a point in its favor as a means of communication." - Wrong. It shows that we were diligent in our research. And with good reason; we know much more about language now than in Zamenhof's time, and we have a tougher and more skeptical audience (the academic world) to please. We also had a bigger job to do, since Lojban was designed from scratch.

Whether or not Don is right about the Indo-European-ness of Zamenhof's grammar, there is no doubt that Zamenhof started with European grammar and simplified. We (originally Brown and later others as well) started with nothing except a goal of matching predicate logic structures, and the vague notion of speakability. Because we had no working language to emulate, there were undoubtedly going to be false starts and re-engineering of major features. I suspect that much of Zamenhof's development period was used to select the root word stock; only a small fraction of Loglan/Lojban development time has gone into word-making.

In a sense, Esperanto took the entire evolutionary period of Indo-European grammar to be developed. (Of course, by the same logic, Lojban took 2500 years, since predicate logic was invented, to be developed).

(You can also compare the actual Esperanto development period with the time that we've taken to redevelop the Lojban version of Loglan from scratch to avoid copyright - less than 3 1/2 years so far, and I suspect that our design is far more intricately specified than Zamenhof's was when he published. By Don's histories that I've read, I gather that Esperanto was not complete in a sense of being standardized until sometime after 1900. Depending on your definitions, we will be comparably standardized either when the textbook and dictionary are done or after the 5 year baseline period proves the language is stable.)

I've been told that a major milestone occurred as late as 1905 when the annual Esperanto meeting was first conducted in Esperanto; at this meeting it could first truly be said that Esperanto was a 'living language'. Lojban should achieve that status in a much shorter time, although possibly with a smaller speaker base.

I note that Jim Brown considered his language speakable in 1977, or possibly even earlier (there are reports that a group called the 'Loglan Sogrun' conversed to a minimal extent in the 60's). Brown actually tried to teach the language to college students in the 50's - though with no particular success - and sold books teaching the language starting in 1966.

Brown's books of the 60's were probably as complete as Zamenhof's 1888 book, but Brown did not have the follow-through that Zamenhof did, nor the 'market' ripe for the language that Zamenhof had with the simultaneous collapse of Volapk. Also, to put it simply, Brown's books, while they explained things in considerable detail, had no text longer than individual sentences. They were thus at best mediocre in teaching the language for actual use. But this was not a flaw in the language or its design, but rather in its inventor's teaching and writing style.

Loglan/Lojban has had an added handicap over Esperanto - a changing plural set of goals which is more than mere 'speakability', and rising standards on what it takes to achieve those goals. The standard of unambiguity changed with the development of computer tools like YACC, and a language thought to be unambiguous suddenly wasn't. I believe I've done more work researching language universals than Brown did.

The whole point of the JL11 discussion, of course, was that comparison of development periods just isn't practical, and the various numbers in the above discussion should prove this. But Athelstan and I were trying to respond to comments and questions that have been frequently raised by Esperantists. If the '35-year' development effort can be claimed as a strike against us, we have the right to argue it as a virtue instead.


"... more people -- or at least Americans -- are repelled when they hear a conversation they don't understand than are attracted. When possible, I always use English under such circumstances." - I was merely observing that at a convention table 'selling' a language, it seemed strange not to hear the language. I would expect that Americans are not much repelled to hear a 'strange' language if they expect to hear one, and one would expect to hear one at an Esperanto table, which is not a BART train. I certainly did, which is why I made the comment.

(On the other hand, Americans are often offended to hear a language other than English when visiting a foreign country, but this is the Americans' problem, not the natives. In the US these days, perhaps 10-20% of the people have a native language other than English, so Americans will have to get used to hearing things other than English.)

I also have a different philosophy as to what it takes to sell a new language to Americans. If you use English whenever that is a possibility because it is a common language, you merely support the argument that 'we don't need Esperanto (or Lojban) because English is already spoken by most everyone who wants to talk to people from another culture'. Regardless of whether it is true or not, the average American is going to think that you are speaking English because it is easier or more convenient than Esperanto. And if it IS easier for you to speak English than Esperanto to another Esperantist, you are missing out on a prime opportunity to learn to speak it better, while demonstrating that the language is useful to passers-by (something most of them are probably unconvinced of).

When I can speak Lojban fluently I will try to speak Lojban at convention tables promoting the language, if the other people manning the table also speak comparably well. If I have problems with people who seem repelled, I'll add a sign inviting them to ask us what we're saying.

This will entice people and cause them to see that we think the language is worth speaking when we could be speaking English instead; they will also be curious as to what we are saying, and we'll happily explain. This may not be how it works out in reality, but this is our goal, and our limited experience so far is that using the language in public prompts curiosity and not repulsion. (We've done nicely at conventions with people who notice our buttons with the slogan "e'osai ko sarji la lojban.")

If we're wrong, Don can say "I told you so". But if this turns out to be the case, then I am most pessimistic that any language will be acceptable as an International language to Americans. At any given time on the path to acceptance, there will be Americans who don't know the language. If a foreigner is not going to learn English (in which case English is the international language), then the American must learn Esperanto or whatever before the need arises where it must be used, or she/he won't be fluent when that need arises. And this means speaking the language extensively with English-speaking cohorts before then, by definition.

In any event, to go from a few thousand to 250 million Americans speaking a particular foreign language will take some aggressive (and skillful) marketing which may be offensive to some people. Possibly as offensive as the USEnglish people are in promoting English (whether one agrees with their opinions or not, their words and tactics are pushy and offensive). The trick is to market aggressively while minimizing offense.

I should note that I while I disagree with Don on this point, I find many of the Esperanto marketing techniques quite skillful, and hope that we Lojbanists can learn from them. This is only practical under a cooperative, as opposed to competitive relationship between the two communities.

Footnotes

  1. Your example on p. 25, "X1 is good for X2 by standard X3," which I presume is written in Lojban -- from your past references to Prolog -- as something like "Good X1 X2 X3" - - would indicate that the position rules in Lojban are much more complex than those in English, and vary from property to property. With regard to my later comments on case, the descriptive rule for speakers of Indo-European languages would be: "The property good relates a noun in the nominative case in the immediately subsequent position, a noun in the dative case in the third position, and a noun in the "standardize" case in the fourth position." Hopefully Lojban's rules are more consistent than some of those of English, in which, for instance, the accusative succeeds a positional dative but precedes a prepositional one...
  2. Some of these rules have not yet been codified. For instance, Kalocsay and Waringhien, the authors of PAG, recognize that Esperantists regularly use adjective roots as prefixes for noun roots -- novedzino, dikfingro are common examples -- but do not admit that this usage is grammatically justified. Most Esperantists go on doing this anyway, and they definitely obey a particular rule of word-formation in doing so -- one that, so far as I know, has never been written down, and would be difficult to codify in a few simple sentences.
  3. Which, if we suppose the -N ending to mark the accusative case in the traditional Indo-European sense, makes vers such as "to go" transitive in Esperanto -- something most IE languages would not allow.
  4. When I was young I read -- in a number of places -- that no other language is nearly as good as English for swearing. In fact, English is a rather pale language in this regard; compare it with any Eastern European language, for instance.
  5. I was carrying on a private conversation in Esperanto on a BART train a week ago, and was excoriated for this by the middle-aged lady sitting next to me.

Masters of Tongue Fu

by Donald J. Harlow

originally published in The ELNA Newsletter
reprinted with permission

When people say "International Language" today, they are probably talking about Esperanto. In China, in fact, the language is better known as shi jie yu, which simply means "international language," than as "Esperanto." In those parts of the world where "interlinguistics" is an accepted part of the science of linguistics, articles on the subject -- if they are not purely historical in nature -- will almost certainly refer almost exclusively to Esperanto. Discussions of the literature of artificial languages will concentrate totally on that of Esperanto, since only very underdeveloped literatures exist for other artificial languages, and for most of them, don't exist at all. Any study of the sociology of an artificial language, too, will concern itself only with Esperanto, since only two other artificial languages ever had populations of adherents even remotely comparable to that of Esperanto, and then only for very short periods of time.

But Esperanto is neither the first not the only "international language." Attempts to create such a language go back at least to the thirteenth century, when the Abbess Hildegarde of Rupertzberg, a lady more recently exhumed -- and justly so! -- by the women's movement, the gnostics, and various musical organizations (how refreshing it is that Hildegarde, one of the earliest of the "Renaissance Men," was a woman!), created her "Lingua Ignota." The philosophers Comensky, Leibniz, and Descartes all wrote about the international language; Bishop Berkeley worked at developing one. In the last century, the Frenchman Sudre created Solresol, a language meant to be whistled or trumpeted, and it enjoyed a very long period of popularity in some circles in France; at one point the French military even considered adopting it, possible because trumpets can be heard over greater distances than shouted commands. Who knows? Had the French followed through with this idea, their defeat in the Franco-Prussian War in 1871 might not have occurred, and all later history would have been different.

No one knows how many "international languages" have actually been proposed. The figure certainly exceeds a thousand. These range from genuine a priori languages, all of whose material is invented out of whole cloth, to slightly modified ethnic languages, such as Basic English. But of the thousand or so such languages, only a few have ever attained any degree of popularity, and most of that has been spurious -- a creation of the news media, ever in search of some new and interesting story. Chronologically, these most famous of international languages have been: Volapk, Esperanto, Ido, Occidental, Basic English, and Interlingua. For those who know little or nothing about the origins and fates of these languages, I would like to give an introduction to them.

Volapk was invented in 1880 by a German priest, Monsignor Johann Martin Schleyer. Schleyer, a polyglot, recognized among his less talented parishioners the need for a language to communicate across national boundaries, and set our to create on. The result was Volapk. The language enjoyed tremendous popularity over the next decade, but, because of certain aspects of its grammar and vocabulary, it generated a strong movement for reforms among many of its speakers; and Schleyer, who saw himself as the language's Pope, so to speak, refused to even consider such reforms. The language's most vocal adherents split into two factions, one supporting Schleyer and on supporting his chief opponent, a French professor named Auguste Kerckhoffs. The resulting struggle destroyed the language, many of whose proponents in any case were shifting their allegiance to the rising (green) star of Esperanto by the end of the eighties. By the beginning of the new century, Volapk was all but dead, though at least one (very small, very irregular) bulletin in the language seems to have appeared as late as 1960. When Bernard Golden went in search of speakers of Volapk on the language's 100th birthday, he found a total of ten -- all of whom also spoke Esperanto.

It is worth noting, however, that at its peak Volapk boasted perhaps 100,000 adherents -- though how many of them could actually speak the language is open to question. In this regard, it is interesting that it shared several characteristics with Esperanto. The two of these that are perhaps most important, in my view, are: (1) an agglutinative system of word-formation, in contrast to the standard Indo-European system (more correctly: lack of a system); and (2) the desire of the inventor to solve the problem of communication between people of different languages, not just to invent an artificial language.

I don't want to go into Esperanto's history in any detail here. If you want to read a good book about the early period, get a copy of Edmond Privat's Historio de la Lingvo Esperanto, or his Vivo de Zamenhof. I would only wish to say that, more than a hundred years into its existence, Esperanto's eventual fate has not yet been decided. Given that over its history the language has had few friends, except for a (relatively few) far-sighted and courageous souls who have actually gone out and learned it, while it has succeeded in gaining for itself a notable array of enemies -- Adolf Hitler and Josef Stalin spring immediately to mind -- the staying power that the language has demonstrated is quite encouraging.

Let me only add here that Zamenhof, like Schleyer, was interested not in creating and artificial language but in finding some viable solution to the problem of communication between different peoples. And in Zamenhof's case -- he was a Jew living in late 19th century Russia -- the problem was far from a theoretical one.

Zamenhof stated (in his First book) that Esperanto was not our typical European language. Arguments over Esperanto's Europeanness go on even today. Certainly, despite recent modest accretions from Japanese and other non-European languages, Esperanto's lexical material remains primarily European, chiefly Romance, in origin. Other aspects of the language's structure are less convincingly European. Certain tendencies in popular use of the language -- for instance, the occasional doubling of short adjective roots to show emphasis, rather than through use of the -EG suffix -- show a pattern of thought in the language reminiscent of Chinese.

I don't intend to argue here over whether Esperanto is fundamentally European or non-European; but certainly many early speakers of the language in Western Europe found it less European (more particularly, less West-European) than they would have liked. This was particularly true in France, where many early leaders of the national Esperanto movement would have preferred a more Francophone, or even Anglophone, tone to the language. A few of these gentlemen, in fact through a rather underhanded process, set themselves up as "reformers" of Esperanto, and in 1907 produced a version of Esperanto that appeared much more in tune with the linguistic norms of the world -- i.e., French and English. For a while, they expected that their new language would replace classical Esperanto, but when this did not happen -- a vast majority of ordinary speakers of the language refused to make the necessary changes in their habits -- the "reformed Esperanto" split off and became an artificial language in its own right, Ido.

While Ido shows a decided shift away from Esperanto's agglutinative word-formation system, back towards a more Western European orientation, it does not represent a complete break with the linguistic ideas expressed first in Volapk and then more clearly in Esperanto. The real difference between the two languages lay in the motivations of the men who developed them. It is fairly apparent that the problem of communication was of little interest to Prof. Louis Couturat, Louis de Beaugront, and Major Charles Lemaire, the primary motors behind the development of Ido; they were more concerned with what they saw as Esperanto's linguistic blemishes. This is hardly surprising; the pleasant little conspiracy into which they entered for the purpose of replacing that Russian Jewish eye-doctor as the guiding force in the international language movement shows in them an ethical blind spot that would not fit well with a genuine concern for the communications needs of ordinary people. Insofar as I do did prosper -- and it prospered, in fact, much more than did any other "international language" except Volapk and Esperanto -- it did so, I believe, despite the people behind it, not because of them.

Ido, in fact, appears to have attained a maximum population of about 10,000 adherents by the early 1920's -- not all that far behind Esperanto in that period. But as the ranks of Esperanto swelled through the twenties, to reach more than a hundred thousand by 1930, those of Ido appear to have declined. It nevertheless remains extant even today, though in what seems to be a basically moribund state. Ido, like Esperanto, has actually produced a small original literature -- though, strangely enough, so far as I know the only genuine literary work ever published in Ido, a collection of original poetry, was published by the Kultura Centro Esperantista in Switzerland.

A recent newspaper article about another constructed language project referred to Esperantists as "verbal hobbyists." As a matter of fact, Ido did much to cull the verbal hobbyists out of the Esperanto movement very early on. One result of this is that, for many years, the Esperanto movement has been remarkably free of individuals who see the language only as an interesting project, whose main purpose in existing is to improve itself by adopting their recommended reforms. Another result is that the Ido movement ended up consisting mainly of just such people. It is hardly surprising, then, that when yet another "improved" international language came along, it would skim off a far greater percentage of members from the Ido movement than from the Esperanto movement. This language was Occidental, proposed in 1922 by the Estonian Edgar De Wahl.

The language's very name gives away De Wahl's motivation. An early Esperantist, he also abandoned the language early on, apparently in protest against its non-traditional structure. Whether he was ever a practicing Idist, I don't know, but suspect that from the time he left Esperanto he followed a very different and more radical route. Occidental, built upon the basis of an earlier project, Julius Lott's Mundolingue, can best be described, I think, as a late and very highly rationalized Romance dialect, with noticeable German accretions. It was, in fact, nothing less than an attempt to codify West European thought processes in a constructed language. Supporters of Occidental justified this by asserting that civilization, being essentially European in nature, should be represented by an essentially European language. In this way, the language would help make the blessings of European thought available to the rest of the world -- or help keep the rest of the world under the European thumb, as the more cynical might tend to think.

The nineteen thirties were, in some ways, the apogee of language construction; Occidental was merely the most successful and best known of a series of attempts to create a new international language. The famous Danish linguist Otto Jespersen, for instance, a long-time mainstay of the Ido movement, abandoned the language in favor of his own project, Novial, which was largely a clone of Occidental. But the best-known project of this period probably remains Basic English.

Basic English, invented in 1930 by the Englishman C. K. Ogden, was an attempt to simplify English and make it more suitable for international use. Ogden claimed to have reduced the entire vocabulary of the language to 850 words. The problem was that his claims were spurious; the language included far more than 850 words (Ogden did not count "international" words such as alcohol in his 850 word vocabulary, though they were considered part of the language; and he added several 1000-word technical vocabularies). Also, many people felt that Basic English was merely a "Trojan horse" for a more standard brand of the language. The event proved this latter group correct; in the 1960's, the British Council, a government-sponsored organization devoted to spreading English among the heathens, bought the rights to Basic English, and since that time it has been used only as in introduction to standard (read: British) English. Though several famous English-speakers supported the language from time to time, among them Winston Churchill and H. G. Wells (who, in The Shape of Things to Come, had the whole world speaking Basic English), no popular movement for this language was ever generated.

Because of the growing number of language projects, there was some confusion as to which one would be, or even should be, the ultimate international language. This confusion had begun when Volapk, which had offered such high hopes to the world, fell apart and was replaced by Esperanto; and it had become endemic when the Ido schism occurred in 1907. By the late twenties, with Esperanto and Ido and Occidental and who knew how many other projects vying for attention, it was understandable that the ordinary individual would throw up his hands in disgust. An American Esperantist, Mrs. Alice Vanderbilt Morris -- of the New York Vanderbilts, I believe -- funded the establishment of a new organization to do research into the problem and find some sort of acceptable solution, for instance a compromise between the different language projects. The organization was called the International Auxiliary Language Association, or IALA for short.

IALA, located in England, though it did valuable research work, had little luck in convincing anyone to compromise. The Romance-based "naturalistic" languages such as Occidental and Novial would not be ready to yield in the direction of "schematic" Esperanto; and Esperantists at that time were not yet ready to forgive the Idists for the dirty work at the 1907 crossroads. In any case, the Esperantists, who even then made up between 80 and 95% of the entire International Language movement, felt that they had no need to compromise. Furthermore, by the mid thirties they had other and more pressing problems to attract their attention -- proscriptions in Germany and the USSR, for instance.

Eventually, IALA, after moving to the United States at the outbreak of war, came under the directorship of Dr. Alexander Gode, and set out to create its own language, which was published in 1950 and given the name Interlingua.

Interlingua is even more quintessentially Romance that Occidental, and in its turn attracted away many of the remaining adherents of Occidental, which tried to stave off the inevitable by renaming itself "Interlingue." But again its creator really had no interest in resolving communications problems; he himself stated that his real purpose was to provide the world with a "standard average European" vocabulary, culled from the Romance languages. Interlingua made modest inroads in the American press's coverage of attempts to solve the language problem through the fifties and early sixties, and there exists a small Interlingua movement, mainly in Europe, even today; but the language never had the widespread support that Esperanto developed even in its earliest years. Its one notable success was in giving the coup de grace to Occidental, whose last magazine bit the dust in 1985.

To recap the situations of these various languages today:

  1. Volapk is a dead issue and has been for the better part of a century. It is not and has not ever been represented by any kind of corpus of literature.
  2. Esperanto continues to grow, and today boasts at least two million speakers, perhaps more, of whom some one hundred thousand actively use the language and participate in the movement to promote the language. Some 150 to 200 periodicals appear regularly in the language, not counting local club bulletins. It has a large and growing body of literature, both original and translated.
  3. Ido retains a small movement and several periodicals to link that movement, though none of them seem to appear more often than quarterly. It has a very small body of original and translated literature.
  4. Occidental is dead.
  5. Basic English as a separate language is dead.
  6. Interlingua has a small relict supporting movement, mainly in Europe. It has few if any periodicals, and no body of original literature to speak of.

Although Interlingua is not the only postwar entry into the international language competition, it is the only one to receive any publicity and to generate a supporting movement of any size. And it is a product of the year 1950. It appears that, to a great extent, the production of such languages peaked in the 1930's, and went largely out of style after the Second World War. Why?

I would tend to blame the apparent "success" of English for this. The War gave French, already in decline, a deathblow, and by about 1950 it was apparent that English was destined to become the international language, by default. So what need for Esperanto, Interlingua, Ido, and other entries into the competition? The outcome was already decided. The other postwar projects -- the Romanids, Neos, Intals, Loglans, etc. -- were doomed to obscurity. Esperanto survived this period, and even prospered to some degree, not because people saw it as the coming world language (though there were those who never lost this hope) but because (a) it had already developed an independent infrastructure that could keep it going even through the most difficult periods -- as Soviet Esperantists proved during the period from 1937 to 1956 -- and (b) it had already developed other reasons for existence besides as a solution to the world language problem.

But the success of English has always been more apparent than real. The growth of English in the intervening period carried the language from 11% of the world's population to about 8.5% -- not the most inspiring rate of growth. Where English has failed, of course, we have tended to blame local conditions for this, or to assume that this failure is non-representative of the world as a whole -- as when, for instance, after a hundred years of concentrated English teaching has not produced a nation of English-speakers in Japan, we insist that "improved teaching methods" would no doubt resolve this problem, or when columnist Neal Peirce, supporting California's English-only initiative, insists that we tend to retreat from English in this country "while the rest of the world stampedes to English."

Forty five years after the end of World War II it is, I think, apparent to anyone that if English has not failed as THE international language, it has certainly come nowhere near fulfilling all those promises that were made for it at that time. Nor is it likely to do so in the foreseeable future, even granting continued U.S. military and economic primacy in the world -- a very unlikely possibility.

Which means that the whole question of the international language is open again. It means that the Esperanto movement, barring the sort of deliberate repression we've seen from time to time in Russia and China and Rumania and Germany and elsewhere, will prosper anew. Indeed, it has been doing so since the mid-seventies.

And it means that, in the field of artificial languages, Esperanto may begin to see some aspiring competitors spring up. In fact, those competitors are already here. In 1972, an Englishman, Leslie Jones, published his Eurolengo, a basically Romance language based on English and Spanish. A young French teacher made the pages of the Guardian in Britain (favorably) with his Uropi. Two summers ago, several Esperanto clubs in this country received letters from a young man developing a project he called Linguos. Loglan, a product of the late fifties which made the pages of Scientific American in June, 1960, has recently been revived in two different forms. And just the other day the ELNA Central Office received a booklet, mostly in German, about a new Romance-based project called Unitario.

None of these projects has, at least in this country, received the sort of publicity that panicked Esperantists in the early fifties when Interlingua appeared. A recent article on Lojban (a schismatic variant of Loglan) that was picked up by the wire services and published in many newspapers around the country, appears to have been less than enthusiastic about the language; with the exception of Uropi, none of the others listed above have even been mentioned in the American press.

But I think that we will hear more of them -- and others like them -- in the future. And much of what we hear, as was the case with Ido and Occidental and Interlingua, will not be why they are ideal solutions to the problem of communication between different peoples, but why they are superior to Esperanto.

Are they superior to Esperanto? Probably so, at least on their own terms. Ido was superior to Esperanto in its adherence to West European linguistic norms. Occidental was superior to Esperanto in its similarity to other Western languages. Interlingua was certainly superior to Esperanto as a quintessential Romance language. And if what you wanted was a watered-down form of English, Basic English certainly filled the bill better than Esperanto.

In Esperanto's own terms -- facility of learning, cultural and political neutrality -- none of these languages was in any way superior to Esperanto, nor even equal to it. The same can be said, I think, about recent and future projects.

The mentioned projects fall basically into two categories, from what I have seen of them, Eurolengo, Uropi, Linguos and Unitario appear to be fundamentally what we may call Euroclones, like Occidental and Interlingua. The designers of these languages, apparently unfamiliar with the work of De Wahl, Jespersen and Gode, are making the same mistakes again -- assuming that the world will best be served, and will let itself be served, by an artificial language with nothing to recommend it but its Europeanness. They don't realize that if this is what the world wants, it is more likely to learn Spanish.

Loglan and its offshoot Lojban fall into quite a different category. Of the mentioned languages, they have been getting the most publicity. But it should be noted that no language as a priori in its origins as Loglan has ever succeeded in generating a body of speakers. To add to Loglan's difficulties, it was originally created as a means of testing the Sapir-Whorf hypothesis (now largely discredited), and for this reason its author claims to have made it as far from ordinary linguistic patterns as he could. This may be a fine way of establishing an experiment, but for purposes of communication it's a non-starter. Loglan will most likely go the way of Barnett's Suma; a few years from now, if you want to learn about it, you will probably be able to get a book out the state library, but nowhere else will any information be available.

Esperanto remains the only truly viable artificial international language: easy to learn, relatively neutral, with a wide base of cultural and practical services for the user to call on around the world. Of all the artificial languages extant today, only Esperanto is, not the result of an attempt to create a language, but the result of an attempt to solve a problem.

And only Esperanto lives.


Bob responds (actually not very much): Funny, I thought Don Oldenburg's article was quite favorable towards the language (and so did he), though I'll admit that the headlines used in some newspapers could be taken as satirical. Certainly the amount of print space given the language was quite significant. But a good news story reports facts rather than conveys enthusiasm, so I can understand Don not finding much enthusiasm therein.


"In Esperanto's own terms -- facility of learning, cultural and political neutrality -- none of these languages was in any way superior to Esperanto, nor even equal to it." - This invites all kinds of disagreement. Facility of learning is of course an open question. Esperanto probably has better teaching materials at the moment because of 100 years to develop them; probably many of the other languages proposed would be equally easy to learn. As to cultural neutrality, Don admits early on that Esperanto derives its lexical materials from European languages. Even if Sapir-Whorf is true, it is likely that a language's word-stock has far more overt ties to culture than does grammar. Don has (in letters to us) written about the ideology held by Esperantists - a language with an ideology is the antithesis of politically neutrality. The goal of being a world language is itself inherently political; some cultures will view such a concept as a threat. Lojban's goals as a whole are basically non-political; international language aspects are a side-benefit rather than a primary goal.

(In one letter to Dr. Brown, Don actually criticizes us for not having an underlying ethic other than ensuring clear communication - a purely linguistic goal. Apparently Don doesn't realize that a non-linguistic ethic is inherently a cultural bias. If Esperanto has such an underlying ethic, it is false to claim that it is culturally neutral without demonstrating that the ethic is universally accepted in all cultures - an unlikely prospect.)


"Of all the artificial languages extant today, only Esperanto is, not the result of an attempt to create a language, but the result of an attempt to solve a problem." - Don says this right after saying that Lojban was designed to test the 'untestable' Sapir-Whorf hypothesis, which was 'discredited' primarily because it was untestable (which is obviously bad science). Testing an untestable hypothesis as important as Sapir-Whorf sounds like an attempt to solve a problem to me. So is developing a speakable language with an unambiguous syntax, as well as developing an a priori language that removes constraints on thought rather than imposing them (as all other attempts I know of tried). There are other problems in the world relevant to language besides the one Esperanto is associated with.

"And only Esperanto lives." - A nice slogan, but questionable at best. Don seems to base this on the existence of an original literature. (By most other standards of 'living' Don mentions in his article, at least Interlingua would be considered 'alive', if sick-a-bed.) Michael Helsem seems to be on the road to matching the entire original literary production of Ido and Volapk before Lojban has a single fluent speaker, and I know of at least two or three others that have more than contemplated literary efforts in Lojban, but either want to acquire more skill before trying or (in at least one case) are waiting for people who can read the language without translating it first.


Two Essays by Andy Hilgartner

[Andy is interested in anybody's comments on these papers, which are tangential to Lojban, but are definitely tied to logic and language. You can also write to him for more information; indicate whether you have any familiarity with the theory called 'General Semantics'. We also have copies of a couple of his papers, including one inspired by his contact with Lojban, available at the special-order price of 15c/page. Andy's address is: 254 Kensington Place, Marion OH 43302, and his phone is (614) 389-4595.]

STATEMENT OF INTENT

I ASSERT that the patterns by which we and our ancestors for the past several thousand years have lived have now failed us.

In my view, these patterns center on the lived assumption that "I already know how things REALLY ARE" -- the perhaps unspoken, even unrecognized, pretense to "absolute certainty," with a consequent unwillingness to re-examine and revise "what I thought I knew." From that lived assumption follows a certain quarrelsomeness: If I already know "how things really are," and you express a view which differs from mine, I will find your PROVOCATIVE BEHAVIOR a THREAT to my AUTHORITY. And if I cannot persuade, manipulate or coerce you to revise what you say so it matches what I ALREADY KNOW, I may take steps to defend my own TRUTHS by suppressing your mistaken OPINIONS -- or may even set out to suppress YOU. And the means of suppressing YOU range from verbal putdowns, to fisticuffs, to murder, war and genocide.

Today, we know how to use cosmic forces (such as nuclear fission and fusion -- A-bombs and H-bombs) to defend our "absolute certainties." Under these conditions, I assert, those lived assumptions which lead us to pretend to "absolute certainty" have outlived their usefulness. To persist in relying on them will, I predict, lead us into species suicide and extinction.

I DECLARE the possibility of intentionally revising our lived assumptions. We can re-build these patterns we live by so that we live not from the practice of defending our presumed "absolute certainties" (or from what I call self-defending), but rather, from the practice of testing our own guesses (which I call self-correcting).

In other words, I declare that we humans now have it within our grasp to produce a fundamental, principled, conscious, and deliberate revision of the structure of human social transacting. For example, we can build up social patterns with which to replace our non-viable social institutions: the self-defeating or self-eliminating aspects of our dealings with ourselves, and the self-defending patterns in what we now call "nuclear family," "friendship," "social group," "corporation," "local government," and "nation-state." We can come to recognize our patterned dealings with the-human-species-as-a-whole. And, having built up these newer, potentially viable patterns, we have it within our grasp to replace the older, non-viable patterns with the newer ones.


I PROMISE to catalyze this revision of the fundamental lived assumptions of the humans species so that, by the year 2007, we have put the new patterns into use planet-wide.

And I REQUEST your direct and immediate participation in this project.

C. A. Hilgartner, MD

Course of Development of a Theory

The theoretical system developed by our research group demonstrably and verifiably opens up a new domain of human knowledge. As of this date, it amounts to approximately 50 person-years of innovative work. In the following four pages, let me tell you something about the background, the increasing rigor, and the further promise of this inquiry.

Since 1963, the work has gone through at least three distinct developmental stages:

A) My earliest paper on this topic (1963) presents a theory of human behavior, stated in ordinary scientific English, but based on known premises which no one else had successfully used in this way. In a paper presented before the International Conference on General Semantics in 1965, I extended this theory of human behavior into the arena of large social institutions. Also, I made logical claims for the doctrine: self-consistency, and parsimony. When I gave the paper, I found myself applauded rather than shot down.

But, while still at that conference, I came to an uncomfortable insight: I recognized that at that time, no one had yet specified the relations between logical assumptions and grammar for even one discursive language. That means that any discursive language remains in the role of "a language of unknown structure." Further, I recognized that one cannot know a doctrine better than one knows the language in which one states the doctrine.

Therefore, my theory lacked rigor -- as long as I left it stated only in a discursive language such as English, I could not back up logical claims made for it.

I left that conference determined to perform a logical analysis of my doctrine, and to state it as an axiomatic system in a mathematical language of known structure. I intended to satisfy myself as to whether one could in fact back up the logical claims I had made for it.

B) Working with John F. Randolph, then Fayerweather Professor of Mathematics at the University of Rochester, I succeeded in doing the required logical analysis. We wrote four long papers which utilized an algebraic set theory notation -- the very paradigm of "a mathematical language of known structure" -- to put the doctrine into the form of an axiomatic system.

At that point, we had something really new: a logically rigorous and empirically testable theory which comprehensively accounts for how a human deals with himself, with his non-living environment, with other humans and (in principle) with other species

To call the theory comprehensive means that one can use it to study "happenings" on any level of interest, from that of molecular structure -- e.g. the structure of heme molecules with various possible side chains, only a few of which have a shape that will allow the ring to combine with divalent iron so as to form the active center of a hemoglobin molecule -- up to that of how the human species as a whole gains its living in the biosphere.

In 1969, the Journal of Theoretical Biology printed three of these four papers. That, too, created something of a stir -- we received more than 1200 requests for reprints of one or more of these papers.

Meanwhile, I had what mathematicians refer to as a "new toy" -- an empty form composed of empty set theory symbols, devised originally to account for the behaving-and-experiencing of individuals -- and I set about finding out what else it could do. I successfully applied it to a number of topics: small group phenomena; large social systems; biological theory. I even made some trespassing ventures into the physical sciences. Eventually, I began using it to focus on the topic of the foundations of logic and mathematics. And at that point (Fall 1971), I began developing another uncomfortable insight.

C) The difficulty centered about a possible contradiction between the "content" of the theory and the notation in which I expressed this "content" -- a collision between central premises. To express this difficulty, I will need to state the setting ("universe of discourse") for the developing theory, its central tenet, and the contrary of this central tenet.

  1. I can express the setting for this developing theory by means of a run-on phrase such as an-organism-as-a-whole-dealing-with-its-environment-at-a-date. When one defines a notational theory on a setting, one restricts discussion to the topics which fit onto that setting -- thereby preventing oneself from unknowingly getting off the subject.
  2. I can express the central tenet of the developing theory in terms of the construct of an organism making a distinction or discrimination -- expressible by a sentence such as "This IS NOT that!" In a world of ceaseless change ("at-a-date"), this sentence appears valid in general.
  3. To express the contrary of this tenet requires the construct of our organism not-making a distinction, expressible by a sentence such as "This IS that!" In a world of change, this sentence appears never valid. Indeed, by the common definition of the term mistake (Old Norse, "to take wrongly"), whenever an organism non-verbally TAKES some non-verbal this as if it WERE some other non-verbal that, he makes a mistake.

Let me paraphrase these simple-sounding phrases into more pretentious logical terminology: Where, in dealing with his environment, an organism non-verbally TREATS this as if it were that, in effect he posits the identity of this and that -- he errs fundamentally; where he non-verbally distinguishes between them, he posits their non-identity -- in that respect, he does not err.

Then the central postulate of the developing theory requires that, on this setting, we disallow the construct of identity (or the binary relation of identical with) in any guise of form, explicit or tacit. In the developing frame of reference, the construct of identical with has no usage except to designate situations in which somebody makes a mistake. In discrediting the construct of identity, I explicitly extend the designated realm of error to include the case in which our organism posits the identity of this with this (or of A with itself). The construct of self-identity conceals the claim that we KNOW what we have perceived and designated as A -- knowledge we do not and cannot have.

To take the rejection of identity as one's central postulate does not lead to paralysis or aphasia. Instead, it strips away the pretense to delusional "knowledge," leaving us ready to act on our assumptions.

However, this central tenet MIGHT contradict the modern logical axiom of identity, which states, "For all x that belong to the delimited domain D, x is identical with x." Thus, in the mathematical theory of sets, one cannot dispense with the construct of identity: for, by postulate, every set qualifies as identical with itself.

Hence, I feared, there might exist or arise a contradiction between what my theory SAYS and the notation in which it says it. At this point, I can prove that such a contradiction does arise; then, however, I could only sense it as possible and feel sick to the stomach over it.

Eventually (Christmas 1971) I concluded that, so long as I continued using the mathematical theory of sets, I left myself no way of avoiding or otherwise handling that possible contradiction. So I resolved to abandon set theory, and all other formalized or discursive languages from the Western Indo-European (WIE) tradition, and to devise my own.

To shorten an already-lengthy story, in the spring of 1972, I made a fundamental discovery. It concerns the assumptions encoded, within WIE languages such as English or set theory, in the grammatical distinction between noun and verb. Briefly, we tell the nouns from the verbs by regarding any noun as identical with itself, and regarding no verb as identical with itself. By the same token, we regard that which we designate by a noun as also self-identical (really existing, persisting, static-and-unchanging), and that which we designate by a verb as also not-self-identical (somehow transient). In the WIE pattern, one obtains a "complete sentence" or a "well-formed formula" by placing at least one noun or noun-phrase next to at least one verb or verb-phrase. Thus, regardless of our intentions, regardless of whether we noticed or not, every time we form a complete sentence in a WIE discursive language or a well-formed formula from a notational language, we posit at least one static-and-unchanging "thing" which enters into more or less transient "relations." In other words, by utilizing the grammar of the WIE languages, we ACT as if, independent of any observer, that which exists independent of any observer has a structure identical with that of the grammar of the WIE languages.

This discovery opened the way toward the development of the desired non-WIE formalized language. I found a way to disallow the hidden assumption I had disclosed, and by means of a small number of explicit logical steps, to derive a grammar from by chosen premises.

This too constituted a new development. Humans had never before had a DERIVED grammar to play with, only inherited, traditional ones; although the works of Edward Sapir and Benjamin Lee Whorf predict or foreshadow this development.

About then I started collaborating with the linguist Ronald V. Harrington of the University of Rochester. On this derived grammar we developed a "Let's keep track of what we say" language, analogous to set theory but fundamentally different in structure. As one way it differs from set theory and other traditional WIE notational languages, the developing notation systematically takes into account the observer. In this notation, one finds it impossible to make a statement except from the point of view of "an-observer-observing-the-observed."

Subsequently, we extended the notation, and

  1. translated the findings of the set theory calculus of human behaving-and-experiencing into the new notation, obtaining a general theory of social systems;
  2. developed a "numbering theory," a "personal geometry", and a "notational physics with physicists in it";
  3. provided evidence suggesting that we humans can now encompass the physical, biological and human psycho-social sciences within a single frame of reference, based on a single set of postulates. This possibility seems to me to exceed the dreams of the seekers after a unified field theory in physics.

At the very least, the new frame of reference gives us an unfamiliar standpoint from which to view, and re-think, human concerns. That alone warrants studying it with care.

Letters, Comments, and Responses

from Ralph Dumain
on Sapir-Whorf Discussions at LogFest 89, and other topics

I don't recall having taken detailed notes at last year's Logfest, hence I was worried at not being able to recollect the discussion sufficiently to be able to write up my own version of the Sapir-Whorf debate. I was hoping that the other participants' reports in the newsletter would either suffice or help to jog my memory. Of course, these reports were published last year, so my memory now requires drastic jogging for me to be able to remember the discussion. At the moment my memory is extremely vague, so forgive factual errors on my part.

I can only remember that I saw the light bulb go off in at least one person's head -- Athelstan's perhaps -- meaning that I convinced one or more of you that you have to refine your conception of exactly what "Whorfian effects" you anticipate finding and how to construct an adequately defined and controlled test. I was disappointed by pc, however, who stuck fast to his ill-informed notion that metaphysical bias resides in the grammatical categories, and that one should commence a Sapir-Whorf experiment based on that assumption. I believe I stressed then, and remind you now, that meaningful experiments require that the variables be controlled. Not only that, one must know what the variables are. In comparing Lojban to any natural language, there are not only several variables involved, but the significance of those variables and their relation to hidden variables may be problematic. Suppose it turns out, for example, that languages vary greatly in a given set of surface-level grammatical features but prove not to have any significant difference at the deep-structural level. In that case, the surface differences would be deceptive and should not even be compared one-to-one, but rather the structures of the entire grammatical systems and their semantic interpretations would have to be investigated, understood, and compared. The level at which pc is thinking is but a shallow caricature of scientific method.

Aside from reconstructing my own version of last year's discussion, which I could still attempt to do with the prompting of others, I have nothing further to contribute on the Sapir-Whorf hypothesis at this point. It is your move now, and if someone in the group can refine the Lojbanist conception of Sapir-Whorf and Lojban's relevance to it, then I am open to further discussion.

I don't think you will get very far based upon your current thinking about Sapir-Whorf. Rather than worrying yourself about metaphysical bias and the ways different languages supposedly limit what can be thought, it would be better simply to focus on what a more precise language can do. As with formal logic and mathematics, or Arabic numerals as compared to Roman numerals, it is possible to extend one's thinking beyond what one can normally do otherwise. This is not exactly Sapir-Whorf, but it is Lojban's only convincing selling point. The creation of a cognitive community that demands precise expression and provides a language for it is intriguing even if it does not attract research dollars. Precision and explicitness in communication, assuming that Lojban is workable in everyday social interaction, can further the expression of and maybe even the formulation of thoughts, and remove significant interpretive ambiguities and other difficulties on the part of the hearer. I say significant, because a lot of expressions that would confound a computer are easily interpreted without mistakes by humans. You once mentioned the issue of forcing assumptions on the listener. In those cases where such "assumptions" actually imply misinterpretation of meaning or intent, they could be removed by more exact expression. It should be understood, though, that those "assumptions" are not metaphysical except when the utterance itself involves ideological issues, in which case the conceptual bias is located in the terminology used in the utterance.

Thanks for finally publishing my bibliography, after stalling for a year using the lamest excuses. The intent and viewpoint of my annotated bibliography were clearly stated, hence there never was a question of misleading the reader. Both the references and my comments help to communicate to the reader just how many factors and subdisciplines are involved in dealing with the issue of the relationship between language and thought. Now that the readers know the different things that need to be considered, they can take it from there. Since I lack the time now to take care of it, it would help you to find a linguist who could find up-to-date information on advances in linguistic theory since the early 1970s that bear upon clarification of the relation of language to cognition.

The closest I have come to dealing with linguistics in a long time was attending a linguistics conference here in December. I queried a couple of friends about the current state of linguistic theory, who were rather cynical. They did not feel, however, that any given school of thought was being discriminated against in terms of research funding; the politics is more personal than doctrinal. The book exhibit was overwhelming; there is more going on than anyone can assimilate -- books on syntax, discourse analysis, you name it -- it's hard to get a grip on. I saw the new book that Mouton has published on interlinguistics (i.e. international planned languages like Esperanto), but it was too expensive to buy even at a discount. There were 6 books in a series on the DLT project, the machine translation system that uses Esperanto as the interlanguage. With two other books I know about, that makes 8 books in all. One of those books includes articles about other machine translation projects, including one that uses the purportedly logical Indian language Aymara as its interlanguage!

A few comments on articles in your recent newsletters. The lengthy article that compares Lojban to Esperanto struck me as much to-do about nothing, as no Esperantist today believes that his language only has 16 rules. That was used at one time as a propaganda device by careless people, but I think people are more thoughtful nowadays, at least on that point. Anyway, it is necessary to understand the historical origin of the "16 rules." They are not descriptive but prescriptive. They came from the effort to put and end to the constant attempts at reforming the grammar that people who are never satisfied with the form of Esperanto or any other planned language kept attempting to make. Adopted as part of the "Fundamento," the 16 rules declared those easily describable, non-negotiable, mandatory features of the language. Together with a basic lexicon and a set of examples illustrating the language in use (including syntactic features not explicitly described elsewhere), the "16 rules" formed the Fundamento. Of course, Esperanto like all other languages contains thousands of syntactic rules, some of which are captured in prescriptive grammars, and many more of which the speakers are unconscious. Esperanto is learned as other languages are learned, without complete formal grammars at hand, and non-Europeans do not have to learn an Indo-European language before they learn Esperanto, any more than they would have to learn French before they could learn English. Also, Esperanto can borrow words from any language, not just European ones.

On the alleged non-competition between Esperanto and Lojban. They are non-competitive if Lojbanists refrain from pushing Lojban as an international language, since the Sapir-Whorf hypothesis is of no concern to Esperanto. However, the minute someone makes a claim for a new international language, several issues arise. Anyone coming forth with a new language who looks like a crackpot automatically discredits the international language movement in the eyes of the public, hence Esperantists have a stake in the matter. In the past, this means that somebody hides in their attic for 15 years creating a language, self-publishes a little book describing his new language, and announces in a press conference that he has just created the now world language. It is one thing to have a hobby, it is another to make bombastic proclamations that one's creation (whether of a language, a new monetary system, or any utopian scheme) will change the world when the lack of social realism is so obvious to all. Those kind of people are obvious cranks, and hence they compromise Esperanto whenever they claim that they have concocted a new world language, as if the adoption of an international language were some kind of magic. Hence Esperantists have justifiably reacted negatively.

Now, I do not claim that Lojban/Lojban is guilty of this extreme behavior. The Washington Post article did not cast Lojban in such a light. You have not yet claimed Lojban to be the future international language. But you have already resorted to dubious propaganda in order to make yourself look good and Esperanto bad.

You suggest that, as Lojban is a superior engineering effort than Esperanto, it can quickly catch up even though Esperanto has a century-long head start. The creators of Ido also thought they were superior language engineers, and where are they today? There are social, political and economic reasons why no planned language, Esperanto or otherwise, has been universally adopted, and those obstacles cannot be surmounted by the most able of engineers. Here the narrow, blinkered mentality of the computer specialist is so painfully evident.

There is also the supposed cultural neutrality of Lojban that makes it superior to Esperanto. But Lojban has not only neutrality, but cultural nullity. Esperanto had social roots (and still does today) in the circumstances of late 19th century Eastern Europe, and in spite of the provinciality of the Warsaw Ghetto, Zamenhof and Esperanto still managed to attract the admiration and loyalty of people throughout the world. The European "bias" of Esperanto's grammar is a non-issue, as that is the part of the Sapir-Whorf hypothesis that has been thoroughly discredited. The European lexicon of the Esperanto language is advantageous to technologically oriented non-Europeans, though it may ideologically repulse others. But the speech community of Esperanto is most diverse, whereas the community of Lojban is extremely uniform and narrow -- computer nerds, sci-fi buffs, people interested in logic and semantics -- not much of a basis for an international culture, and certainly not an ideologically neutral or even divers culture. Esperantists, in spite of the European bias of their language's lexicon, have risked and even sacrificed their lives in fighting racism and fascism; no Lojbanist I know would ever make such a sacrifice.

What is most irritating is misusing facts in order to support misleading generalizations. I accept as truthful the statement that while sitting next to the Esperanto booth at a sci-fi convention, you did not overhear the Esperantists speaking Esperanto to one another. You dishonestly suggest by that example that Esperantists are not even accustomed to speaking the very language they are advertising to others. What hypocrisy, in light of the fact that Esperanto conversation has been going on for a century in the most diverse of circumstances, while no Loglan/ Lojban conversation in the context of any normal social interaction has ever taken place! I too have staffed an Esperanto booth upon occasion, and I too have only used English to speak to my fellow American Esperantist booth-mates, because it is basically an English-speaking environment, and I do not generally speak Esperanto in an English-speaking setting although I am perfectly capable of speaking the language.

So it seems that in spite of your lip service to non-competition, you are already pitting Lojban against Esperanto in a competitive fashion, and you have also resorted to duplicity in doing so. Under those circumstances, you cannot realistically expect amicable relations between Lojban and the Esperanto movement. You know that I do not tolerate dishonest propaganda on the part of Esperantists, as evidenced by my disagreements with Don Harlow. I surely am not going to let the young upstarts of Lojban get away with any nonsense, especially when they are highly educated people who claim to be able to use their language in order to improve their thinking and their world view.

I enclose a photocopy of a commentary on Loglan/Lojban from Rick Harrison's The Alembic. I pass this along for the completeness of your archives, not to torment you. Mark Tierisch's reasoning leaves something to be desired in many parts of this article. Although this article makes Esperanto look good in comparison to Loglan, its reasoning doesn't hold up, especially since Esperanto like all other languages has a lot more than 30 grammatical rules, let alone 16. The only place where I unequivocally agree with Tierisch is where he refers to Loglan as not culturally neutral but as a reflection of the "culture of nerds." The disparaging term "nerd" is hardly necessary, but the description accurately pinpoints the subcultural basis (and hence metaphysical bias) of Lojban: science fiction and computer buffs and the like.


Bob responds on a couple of items - I'm not going to discuss all of Ralph's points - I'll leave that to the community, especially the discussion on Sapir-Whorf. Suffice it to say that I think Ralph makes some different assumptions than we do about what kind of useful information can be obtained by studying Lojban in a Sapir-Whorf context. For example, he makes comments about the possibility that all languages have a common 'deep structure'. This may be so, but even if only surface structures are directly related to culture, it would be useful to confirm it. (If the only 'important' features about a language's structure are in its 'deep structure' and all language 'deep structures' are the same, then Sapir-Whorf is claiming the nonsensical idea that all cultures are the same.)

We of course do not consider that the Sapir-Whorf claim of a relationship between culture and a language's grammar has been 'discredited', as both Ralph and Don describe it. It is precisely that claim that Lojban is designed to test. Thus, a clear separation from European language structures is vital to Lojban's goals, and Esperanto's lack in this area is a primary reason for its unsuitability for our purposes.

I think Ralph has an incorrect view of the Lojban community. You are far more diverse than he claims. A large percentage are computer-literate, and many read science fiction, but not all; in any case, even those two categories define widely varied audiences. I can see that education is inherently a potential bias, but I challenge Ralph or anyone else to state actual metaphysical biases that are common to all members of either group, or to the Lojban community.

To tie back to something I said regarding Don Harlow's writings, Lojban's metaphysical diversity can be shown by a wide diversity in political beliefs among the community. Within the Lojban community are sizeable numbers of libertarians, socialists, and anarchists, extremes of both the right and left, along with more mainstream political philosophies. It is an incomplete argument to infer metaphysics from politics, but I think it is a reasonable idea.

Whether most Lojbanists (the majority of whom probably oppose both racism and fascism) would die for their beliefs, I cannot say. At least some of our supporters are in the Armed Forces and are committed to die for their country if necessary. Ralph impugns the honor of these and other Lojbanists with his statements.

I recognize that Esperanto has had its martyrs. One would hope that martyrdom is not a vital prerequisite to achieving an international language. One 'problem' with martyrdom, is that, while it draws together the community associated with those who have died, that same strong feeling alienates those outside of the community, and causes them to misunderstand. Some may be drawn to a movement that people are willing to die for; others are repelled by the 'fanaticism' that they perceive in such an attitude.

In any event, fighting racism and fascism is not what Lojban is about, although I personally would hope that with increased understanding of other cultures that is possible through learning Lojban, people would find it more difficult to persecute those who differ from them.

I will admit that any discussion of Esperanto and Lojban will lead to some comparisons. Our purpose in the articles was to blunt the validity of such comparisons. My statements about Esperanto do not claim that anything is 'wrong' with it; I merely feel that Lojban is better designed for the purposes it is meant for than Esperanto is. But those purposes are different from Esperanto.

Only where we talk about the potential for Lojban as an international language is there even a basis for comparison. In this area, though, I stated that Lojban would have no significant role unless both a) Esperanto clearly fails as an international language and b) Lojban's other uses make it attractive as an international language. The international language goal is an incidental one for Lojban (though important to some among Lojbanists, including some who are also Esperantists). There is plenty of room for both languages to successfully achieve their goals.

My point is that both languages can gain by cooperation rather than competition. An Esperantist is already more open to the possibilities that make Lojban interesting than a typical member of the non-Esperanto public. Similarly, a higher percentage of Lojbanists are aware of and interested in Esperanto than of the general public. If this commonalty can be harnessed, positive synergistic effects are likely.

In this light, my comments about Esperanto being spoken at convention tables should be taken much more positively. I did not and do not claim that Esperantists cannot speak their language. Rather, I believe that the outside image of Esperanto as 'useful' and 'important' suffers when they do not and they can; Lojban will similarly suffer if Lojbanists do not use their language. My calling this situation to peoples' attention, and saying that I plan to do differently, says nothing at all about the relative merits of the two languages. It was a friendly, and I thought constructive, criticism. (As an aside, Ralph is incorrect in stating that Lojban has not been used in 'normal' social conversation. Extensive use, not yet - but surely within a year even this will have changed.)

As a final note, Ralph's last reference is to an letter in The Alembic that was a diatribe against Lojban. In it, writer Tierisch (who hasn't ever been on our mailing list and is unlikely to know much about the language) compares Lojban's rules to Esperanto's 16 that we discussed last issue. Ralph mentions this, but just a few paragraphs earlier said "no Esperantist today believes that his language only has 16 rules. That was used at one time as a propaganda device by careless people, but I think people are more thoughtful nowadays, at least on that point." Don Harlow said something similar. Apparently they are wrong. Perhaps the leaders of the Esperanto movement know the significance of the 16 rules, but the community of Esperantists as a whole may not.

Tierisch's letter made many incorrect claims about the language and suggested that he felt threatened in some way by Lojban's ideas (perhaps in the way Ralph suggests Esperantists feel about 'crackpot' language inventors). We wrote a reply, but The Alembic folded without printing another issue.

My own feeling is that people should not feel threatened by ideas that differ from their own. I can understand that Esperantists dislike the 'guilt by association' that comes from association with 'crackpots'. But this is just part of the territory. People like playing with language and new invented 'languages' will crop up all the time. Reacting by disparaging the inventor merely offends the inventor; it doesn't stop other inventors, nor helps Esperanto's image. I think there are better approaches.

My main point here is that the positive effects possible if both of our efforts worked at promoting created languages in general, as well as our specific versions, instead of knocking at each other. The potential benefits of cooperation far exceed the benefits we can gain at each other's expense.


from John Hodges:

...

I took to heart your essay in JL11 that "there is no competition between E. and L., because their goals are different." But I'm not sure your argument succeeds.

The goal of E. is to be an international language, to be "everybody's second language". Notice that this is a global ambition, and implies that any other "second" or "international" language is a competitor. They have an established claim to this role, with 100 years of experience, 10,000 books, and 2,000,000 speakers (1990 World Almanac figure). Also some martyrs, persecuted by the Nazis and other militant nationalists.

Lojban has three major goals: 1) to be a research tool for scientific study into the relationships between language, thought, and culture - we hope that studies will prove that people think more flexibly and/or more logically in Lojban than in any other language; 2) to find computer applications, e.g. in artificial intelligence, human/machine interface, and machine translation; 3) to be an international language. (We welcome anyone to use it for anything, but these are the goals we had in mind during all those years of development.)

Goals 1) and 2) are less-than-global ambitions, which genuinely do not challenge Esperanto. But your essay in JL11 keeps goal 3), which does. You soften it by saying that the challenge will not be a serious one for many years, and people should have their own choice on it, anyway. But it is still there, and there may be a practical conflict between goals 3) and 1).

Goal 1) is to be a research tool for learning about language, and the relationships between language and thinking. To achieve our scientific goals, we want/need to gather a body of at least several hundred fluent L. speakers from a wide variety of linguistic and cultural backgrounds, who can participate in controlled studies. No doubt to get these, we will have to recruit and teach thousands.

To gather such a varied body of speakers, we could translate our teaching materials into Chinese AND Hindi AND French AND Spanish AND so forth, or we could translate them into Esperanto, a relatively simple task. If we CAN recruit Esperantists, we should try to do so. But CAN we?

It is clear what the Esperanto community can do for Lojban. But what does Lojban offer the Esperanto community? Why should they host our experiment, given the conflict implied by our goal 3)?

One line of thought I have explored in recent days (many of my lines of thought are half-baked - I am asking for feedback). Perhaps we should explicitly make goal 3) conditional on prior success in goals 1) and/or 2), and commit to cooperation with the Esperanto community should the event arise.

We begin by describing Lojban as "an experimental human language". (I think this is true, anyway. I expect that our first five years of use will show us changes we want to make, when our 5-year baseline expires.) We point out that carrying a Lojban textbook written in Esperanto book in E. book services will hardly threaten the spread of E.; it is just one more cultural opportunity that opens up if you learn E.

If the L. experiments to test Sapir-Whorf show that, as we hope, people think more flexibly and/or more logically in Lojban than in any other language, OR IF future computers still find transcribing/parsing/translating Esperanto to be beyond them, while Lojban is translated with ease, THEN AND ONLY THEN will the question arise of whether to trade in Esperanto for a newer model. By hosting our experiment, the E. movement will have stuffed the ranks of L-speakers with Esperantists, assuring them the loudest possible voice in the future development of Lojban Mark II.

Carrying the thought further... How much good would such a voice do them? English speakers have suggested deriving the gismu from English alone; this is still a rotten idea if the favored language is Esperanto. Lojbanized gismu do not resemble their source words closely enough. Even if pure source words are used for gismu, they float in a sea of cmavo that makes the result incomprehensible to speakers of the source language. Lojban is just too radically different. But current Lojban has rules of spelling and word-formation designed so that today's computers, with PRIMITIVE abilities at pattern-recognition, could transcribe and parse spoken Lojban correctly. The abilities of future computers may allow us to relax those rules. (neural nets, optoelctronics, etc.)

For the rationale above to work, Lojban Mark II would have to be rebuilt from the ground up. SO- hypothetical question, for 10 to 20 years hence - if we wished to make a language with predicate grammar, and accommodating the limits of computers of the time. and as compatible as possible with Esperanto, how close could we come? Could, e.g., the prefixes and suffixes of Esperanto substantially replace to cmavo? Could all brivla have only one or two places, mimicking conventional parts of speech, with the remaining places added using lexeme BAI? Could we seriously imagine making a Lojban Mark II that was a superset of Esperanto, so that existing E. speakers and books would remain compatible? Or perhaps one that was designed to make conversion from Esperanto as easy as possible?

I am out of my depth here. But if we are to seriously recruit among Esperantists, we may have to commit in advance to something like this, provided the experiments show Lojban Mark II to be a worthwhile effort. Then again, perhaps these thoughts are way off base, and the future of Lojban does not lie in recruiting Esperantists. Are we, ultimately, competitors after all?


Bob responds - John stated the 'flaw' in my competition argument better than anyone else, but I still stand by what I said above, that both languages can co-exist without competition between them. There are some hidden assumptions behind a deduced 'unavoidable' competition based on John's logic.

The most important flawed assumption is natural for most Americans: that for one language to be 'everyone's second language', there can be no other international language. For monolingual Americans, learning a second language seems onerous enough - why would anyone want to learn two 'international languages'?

Simple. One learns different languages for different purposes. Languages are tools for communication; you use the best tool available for the communications job at hand. By this argument, of course, Don Harlow is right in using English to talk to another English speaker, and Esperanto when he wants to talk to someone who doesn't know English as well as they know Esperanto. (Which I agree with in general, making exceptions for the times when the language is on display for outsiders, or when a particular educational purpose would be served.)

Especially if Lojban proves Sapir-Whorf true to any extent at all, someone learning Lojban will think differently (and perhaps 'better' by some standard) than if they know only Esperanto or their native language.

We have no problem recruiting Esperantists. They have the same range of interests as any other group of similar size. In fact, Esperantists are a fertile recruiting ground because they are already interested in language.

Some Esperantists will find the design goals of Lojban, or specific design features, worthy enough for them to further study the language. Then, when they know more, they can decide to study both languages or to just study one. John's argument is flawed here; he assumes that, because the goal for Esperanto is to become "everyone's second language", every Esperantist holds that goal as a nirvana that they cannot turn away from.

But Esperanto is not likely to achieve its purpose within our lifetimes. So many Esperantists will be interested in the language that offers them more personal gratification within their lifetime. Some will find this in Lojban; possibly others in some other language. Many, perhaps even most, will concentrate on Esperanto, or will work with Esperanto and Lojban. For these, Esperanto provides the immediate satisfaction of a large speaker population with which to communicate, while Lojban presents a peculiar intellectual challenge that may at some later time prove more rewarding. There is no competition implicit in our existence for such people.

An Esperantist who denies the value of learning other languages is as close-minded as the nationalists that oppose Esperanto. Some will be this way, and that is their right. But far more valuable to both Esperanto and Lojban would be cooperation between the two groups. Undoubtedly, Lojban will attract a lot of people that would not be interested in Esperanto (as Ralph says, computer people and other scientists, and science fiction readers, are a natural audience for Lojban). Some of these may not find Lojban to their liking (too different, too small a speaker base, etc.), and may proceed onward to discover Esperanto. The reverse will be true among Esperanto recruits. By having information on both languages available, people can make an informed choice as to which language serves their interests.

A side benefit results. A cooperative, open, attitude is presented to the public. This attitude ameliorates the impression that international linguists are fanatical idealists, an impression that turns off a lot of people. Our relaxed attitude towards international language success has not only reduced Lojban's 'threat' to Esperanto, it has calmed the portion of the Lojban community that opposes the idealistic 'world language' effort.

Incidentally, one member of our original class here in the DC-area, Paul Francis O'Sullivan, is a lifetime member of the local Esperanto chapter. He finds no conflict in working with both languages and is translating the brochure into Esperanto for us. (Reviewers are welcome to volunteer.) Jamie Bechtel, our first Lojban 'creative writer', is also an Esperantist, as is poet Michael Helsem. Numerous others, too.

We are gaining cooperation from Esperantists. Bruce Arne Sherwood, a 'big name' in Esperanto, taught courses and wrote articles comparing Loglan and Esperanto in the early 1980's, and carried on a lengthy correspondence with pc to ensure that the facts were right. No animosity or competition was evidenced. Mike Urban, an Esperantist known for developing MacIntosh Hypercard teaching software for Esperanto (and one of the Worldcon table representatives), has advised us on some technical points of Esperanto, as well as on teaching software. Etc.

We cannot test Sapir-Whorf based on teaching the language through Esperanto. If all of the target population spoke Esperanto as well as Lojban, there would be no way to separate effects of the two languages from each other. We must use monolingual speakers who learn Lojban as their first non-native language, or better, bilinguals raised speaking Lojban and their native language from birth. In this way, Sapir-Whorf effects would be least hidden by uncontrolled variables (a problem mentioned by Ralph that we are indeed concerned with).

As for Lojban Mark II, I doubt if it will happen. If there are changes after 5 years, they will be minor, evolutionary ones. That is why we are forcing the 5-year period, to ensure that inertia keeps the language stable.

People in the Loglan community are tired of learning a changing target. Regardless of how flawed Dr. Brown's versions of Loglan are, the Lojban development would never have been conceived of, much less completed, if not for Brown's intellectual property claims that forced us to work from outside rather than within the Institute.

If Lojban does evolve in new ways, the speakers will be the ones who decide, as John suggests. If the speakers are Esperantists, some of the underlying concepts of Esperanto will find their way into the language.

However, as John points out, Lojban and any natural language are too different. Lojban is also too different from Esperanto to offer significant pattern matching. A predicate language is too unlike an Indo-European grammar, or anything that can even be described like an Indo-European grammar. If you rule out changing all the words once again (a relearning burden that would be unacceptably high - as anyone who has used LogFlash with both Institute Loglan and Lojban words can testify), there simply isn't that much that is worth changing. (It is also possible that to make such changes would destroy whatever there is about Lojban that makes it worth 'trading in' for.

No. Lojban will stand on its own, and will gain support from Esperantists on its own merits, or not at all. As long as I have influence, I will resist attempts to make there be an 'exclusive or' choice between Lojban and Esperanto among potential speakers. If we do this, there will be no competition. (Hmmm! Could increased competitiveness be a fallout of linguistic confusion between 'inclusive or' and 'exclusive or'? A Sapir-Whorf effect that we might find negated among Lojbanists!)


Let's turn to one more letter on Esperanto (and a few other subjects), from Paul Doudna. Bob's responses to some of them are embedded:

...

The discussions of the Sapir-Whorf hypothesis have been interesting. However, I'm inclined to think that testing this hypothesis may be like testing the Shroud of Turin. That is, no matter how it is tested and no matter what the results are, people will continue to believe as they did before.

I am sending you an article on E-Prime. If you haven't seen this already, I think you will find it interesting. The article is written in E-Prime, which is almost the same as English. The author obviously believes that language does have a significant effect on our thinking.

[Bob: E-Prime is a variation on English devised by General Semanticists that avoids using the verb 'to be' in all its forms. Andy Hilgartner's articles above are written in his version of E-Prime, and Andy, also, apparently believes that language has a significant effect on thinking. Score one against Ralph's and Don's claim that S-W has been totally discredited.]

Concerning the Lojban logo, I notice that most of the suggestions involve some kind of visual ambiguity. Considering that one of the primary claims of Lojban is its freedom from ambiguity and also the fact that some critics claim that Lojban is actually more ambiguous than English, the use of an ambiguous design for a logo would be quite ironic.

I showed the articles comparing Esperanto and Lojban to a friend of mine who is an Esperantist. His reactions were very negative. I must agree with him that many of the points of comparison were not valid. The articles themselves contained some disclaimers, implying that the comparison should not be taken too seriously. In particular, the attempt to compare "rules" I don't think really works. The meaning of the word "rules" is used quite differently in [discussing] the two languages.

Here are two suggestions for a more meaningful comparison:

  1. Translate some sample sentences in English (chosen equally by Lojbanists and Esperantists) into both languages. Include relevant comments on any peculiar features of the translations.
  2. Compare the underlying assumptions behind the two languages. Zamenhof and Brown had in mind quite different concepts of what constituted an ideal language. These concepts of course determined the way the resulting language should be constructed. This type of comparison might be very difficult since in many cases these underlying assumptions are not made explicit.

When I heard a talk on Esperanto about a year ago, it sounded almost like the speaker was talking about Loglan/Lojban. There is no ambiguity in Esperanto, it was claimed. (But the two languages mean something different by "ambiguity".) It was further claimed that Esperanto is culturally neutral. (Again, the meaning of "cultural neutrality" is not quite the same in both languages.) Esperanto is completely "logical". (Meaning that the grammar is free of irregularities typical of most languages, not that it is based on a system of logic as Lojban attempts to do.) And of course the spelling is completely phonetic. (Both languages are alike in this respect, although Esperanto doesn't have spoken punctuation.)

[Bob: A good response and some good suggestions. Any volunteers among the Esperantists to devise some sentences to translate and/or some lists of assumptions and ideals. We may need Paul to serve as a moderator to point out where our definitions don't jibe.]

Have the 600 rules of Lojban been published? I suspect that no matter how many rules are stated explicitly, that there will be a potentially unlimited number of implicit semantic rules that are used in any language to actually understand what any given sentence means.

[Bob: On the first: Yes, this issue! Though the number is now closer to 550, depending on how you count. Every word has a 'rule' defining its semantic meaning. If you count those as rules, than a language with fewer words has fewer rules. However, you can turn this around. The universe of discourse for 'all of language' is approximately the same for all languages. A language that divides up semantic space into fewer words tends to end up with words being used for multiple meanings. Lojban has one advantage in that Lojbanists generally try to avoid unnecessary figurative extensions of meaning and to explicitly mark those extension where accurate interpretation is important.]


The article on negation was enlightening, as far as I could follow it. I'm not familiar enough with Lojban grammar to fully appreciate its significance for Lojban. I would have two observations:

  1. I'm not sure how this will work in Lojban, but in English, we typically form negatives which theoretically represent contradictory statements or complement classes. In practice, in spite of the language representation, we tend to think in terms of contrary stereotypes. To treat these contrary stereotypes as if they were jointly exhaustive concepts (as the language form would seem to indicate) can lead to illogical conclusions. For example, "Un-American" and "unchristian", to give two extreme examples.
  2. It is pointed out, quite correctly, that negation in English is very complex in practice. It is only roughly analogous to the vastly simplified negation of formal logic. Basically, the same conclusion applies to the logical connectives AND, OR, and IF. That is, these words (along with NOT) in the context of formal logic do not mean what their counterparts mean in English. It will be interesting to see if those who use Lojban will adhere to the theoretical meanings of such words (thus resulting in a logical form of thinking far beyond what is normally found among human beings), or will the meanings of these words merely shift, losing their precise logical definition (in terms of truth tables) and evolving to something much closer to English.

[Bob: On the first, we ended up adding two cmavo to lexeme NAhE, the contrary negation lexeme. In addition to na'e, which refers to the generalized scalar contrary, we've added no'e as a scalar 'middle' or neutral, and to'e as a polar opposite. Thus Lojban allows explicit distinctions in "un-" that are not possible for English.

On the second, Jim Brown and the Loglan community in general have looked at AND, OR, and IF, much more thoroughly than they had NOT. In addition, with the possible exception of IF, these do not have the complex questions of 'scope' that negation has. We're fairly confident that no problems remain here. When I write the textbook on those parts of the language, we'll be more certain.

There should not be much backsliding to English versions of the logical connectives if we've properly taught them, because the various non-logical English versions of these connectives are also built into Lojban. For example, we have causal connectives for causal IF, a wide variety of ANDs, and a flexible restatement of OR in terms of set membership. The availability of the English non-logical combining forms should keep the pure logical connectives 'pure'. We'll certainly find out. (Note that most of the English-like forms are more highly marked in Lojban than the logical connectives, so there will be some caution necessary in this area.)


In the process of trying to sort out and file the material I have accumulated over the years about Loglan/Lojban, I notice that in many cases material (such as the article on Lojban Negation) does not contain the date and does not contain the name of the author. I realize that in the past I have written some things myself that are not properly identified. However, I think it would be useful if all material that is distributed was identified with a date and name.


[Bob: Good idea, and we'll try to do better. We've put out JL with enclosures on the assumption that most people who save the material put it in notebooks as they get it, and hence have stuff associated together in a 'useful' order, which is usually not by date. For next issue, I'll try to come up with a list of our publications of the past, which ones are worth keeping (other than for historical interest), and explain how Nora and I have set up notebooks to keep the 'useful' information at hand.

la lojbangirz. publications often do not bear an author's name if we want the article to be seen as a product of the organization rather than of an individual. In such a case, we'll try to make sure the organization name is on it.]

Another thing that I think would be useful would be a short glossary of often used Lojban terms, to be updated at frequent intervals. This might be published as a separate sheet to be kept as reference, or if short enough, included on the back page of each newsletter. There are certain terms used frequently which I don't always remember (and new readers of the newsletter will not know at all). For example, I am trying to find the difference in meaning between the definitions of bridi and selbri. They are defined somewhere, I'm sure, but I can't locate the definitions.

[Bob: I explained selbri above, but in case you missed it, selbri corresponds to the logic term 'predicate', while bridi (in the gismu list) corresponds to the term 'predication'. In other words, the bridi is the whole sentence, while the selbri is the thing in the middle that determines the relationship among the sumti.

We used to have a glossary in every issue, up to around JL4 or JL5, but it got too big. The Overview was written to be a textual glossary for newcomers, figuring that old-timers were tired of seeing (and paying for) the definitions for words they already knew. The Overview, of course, is getting old, and has kunbri for what is now selbri and a couple of other minor errors. But the errors have thus far been seen as too minor to justify a revision sent to everybody.

The textbook will contain a glossary of both Lojban words of this sort and English linguistic terms that are used a lot in describing the language. In addition, Lojbanist Nancy Thalblum has started putting a glossary together as she studies the language. We may publish this separately, or use it as the basis for the textbook glossary. People are encouraged to write in with lists of words they would like to see defined in a glossary (other than the obvious Lojban ones).]


le lojbo se ciska

First this issue is a little ditty written by Athelstan. It's cute, if not profound. Sing it to the tune of 'Oscar Meyer' jingle.

    My Lojban has a first name                             
       it's 'logical', you see                             
    My Lojban has a second name                            
       the 'language' that's for me                        
    I like to use it ev'ry day                             
       and if you ask me why, I'll say                     
    la lojban. cu se nelci mi                              
       gi'e ve tavla do fo mi                              


The last line is pronounced:

/lah,LOHZH,bahn. shoo,seh NEHL,shee,mee/
/gee,heh veh,TAH,vlah doh,foh,mee/

and means, roughly:

"Lojban is liked by me, 
and is spoken to you by me."

Also on a light note, John Cowan made his first Lojban effort a real humdinger. Here's his introduction and the text. See the translation section if you can't figure it out. (If you haven't tried a translation before, you may want to try John's journal entries, which follow this monstrosity, first.)

In this month's writings, we've tried to use the added writing conventions to make the text easier to read. Hope it helps. Hope we are reasonably consistent in our usage of the conventions.

John:

And (ta-dah!) my first piece of Lojban. I'm using a format similar to JL10's, so you can see what I said, what I meant to say, and what I was trying to say. The original, by the way, is the 1984 winner of the Bulwer-Lytton Fiction Contest, an international bad-writing contest where the contestants submitted the opening paragraph from "the worst of all possible novels". I found it in a book called "It Was A Dark And Stormy Night", a collection of entries.

Herewith the Lojban:

ni'oni'o le nimcitno goi ko'a po'u la ka,as. noi ko'a melbi bai jorne le sramudri noi ke'a kusru zi'e po le dabli'e goi ko'e po'u la danlu ku secau le kecti .i le'i prenu poi ke'a se saptutklu zi'e po ko'e ca zbasu lo derxi be loi mudri bei leko'a jamfu noi ke'a cisyselpli .ibabo le voksa noi ke'a cladu je klina zi'e po'e la nebnauzag. noi ke'a te pemci je tajnau cu fegcru <<lu ganai le do fagryrinka cu rinka lei fagri ja'e le nimcitno se jukpa gi do lifri lenu mi setca le xarci le do se citka poi ko'a romoi ko'u li'u>>

As mentioned earlier, John Cowan has started writing a Lojban journal. Here are the first 3 days of his journal entries. Only minor corrections were needed. John's pleasant rhyme (you'll see), used an incorrect word. We corrected the word, calling the result a near-rhyme.

Most new Lojbanists should be able to understand this text using only the gismu list, rafsi list and cmavo list, and making literal translations and educated guesses as to how the result goes together. Try it and see:

ni'oni'o de'i le djedi po la lunas. vau

ni'o le mutce cmalu jukni cu cpare le djacu tubnu .i carvi ja'e le nu ko'a farlu .i le solri cu febri'a le carvi .i le jukni cu rapli cpare

ni'oni'o de'i le djedi po la mars. vau

ni'o le mi tixnu po'u la airin. verba .i ko'a nanca li ci .i ko'a kelci je krixa .i je ko'a sipna .i ko'a bacru fi la gliban. .i mi bacru fi la gliban. .e la lojban. .i mi na'e drani bacru fi la lojban. .i mi zutse .i je mi viska le nu le mi tixnu cu kelci .i glare
ni'o mi jdice le nu mi te cmene le'i djedi le'i plini
ni'o cusku fa la viktorias <<lu mi ga'i ba'e na'e se zdile li'u>> .i di'u xajmi mi .i zo ga'i jibni rimni zo ba'e la lojban. .i zo na'e go'i .i le nu le'i rimni cu porsi cu xajmi mi

ni'oni'o de'i le djedi po la saturn. vau

ni'o cusku fa la daisets. suzukis. <<lu mi po la zen. skicu lo kilycna fo lo na'e kilycna li'u>>
ni'o mi tirna lo verba zgike .i ri mutce pluka mi
ni'o mi pu tcidu lo cukta be le nu ke pamoi penmi le na'e terdi prenu po'u la tranx. bei la alan. din. fostr.

Next we have a letter and another poem by Michael Helsem, who has been extremely prolific in the last few months (we are pleasantly drowning in his Lojban poetry). Due to our tight schedule, we've made some changes to what he wrote that we haven't been able to run by him, so hopefully he'll forgive us if we messed something up. Michael does follow the excellent practice of sending us fairly literal English translations of everything he writes so we can (usually) figure out his intent.

Michael's text is very complex, indeed probably too complex either for him or for a typical reader; it appears that Michael was trying to paraphrase his natural English idiom, throwing in plenty of attitudinals and discursives. However, the result is worth studying; the letter exemplifies several more esoteric features of Lojban grammar as well as the problems in translating idiom. At least look up the words and make your best guess what he's trying to say. Then check the translation to see what he intended.

A far better practice in writing your own letters, by the way, is to figure out what you want to say, then express it in short Lojban sentences as John Cowan did in his journal. You will find after writing only a few pages that you will develop a more Lojbanic idiom.

Michael actually made very few grammatical errors. There are, however, some 'wrong words', some unfortunate Americanisms that translate poorly when he tries to do them literally, and some equivalents to English pronouns, that in Lojban are frustratingly vague or misleading. All in all, though, this is an excellent effort.

I editorially embedded the original of the poem included in the letter within the non-grammatical quotes "lo'u ... le'u"; the poem had one 'grammatical error' that could not be corrected without destroying the sense of his introductory comments regardin the poem.

di'e zirjbo po'a xatra de'i la xav. po'e la mumast.

ni'o coido'u mi dopeza ledo nu jarco le mi pemci cu ckire ra'u .i puvi'eku mi mutce cutyzu'e pi'o gi'e pu'i ninpemci jmina la'u lo mo'amei .u'u ne ki'u lemi cabi'ibajenairu'i .ia li'anai zmadu nu mutce cutyzu'e .i ta'onai pu ra ku mi pinka so'u lepu selja'o ti'u


.i pamai sera'a le pemci po'u se tcita <<lu leka sarcu kei vau li'u>> ku'o mi na jimpe le krinu be ledo nu punji zo <<cu>> le crastu po zo <<nu>> mu'inai lemi pu nu dunda (sei zo <<sabji>> lu'anai cu drani se'u) lo temge'a tcita (to te'i zo <<ba>> vau toi) pe vi le trixystu ne seba'i .i ti jufra <<lu le nunsti ne sekai leka pratci cu batci mu'anai li'u>> vau ?xu

.i remai .uocai sera'a le seltcita be <<lu le firgai mu'anai vau li'u>> ku vi le da'amoi vlali'i ku zo <<co>> cu se setca fi <<lu fasnu cictcima li'u>> ja'e <<lo'u lo nalsti nu cictcima fasnu le'u>> (sei zo <<za'i>> basti ?xu be zo <<nu>> se'u) .i zo <<fasnu>> ca nalsarcu .i'a .iku'i ta rinka lenu mi ninzga lo puze'u nandu .i mi su'oroi pilno zo <<nu>> lo paroi tortei bo fasnu gi'e drata go'i fi lo ranji clatei bo fasnu .i re frica valsi cu sarcu vau pe'i .ije .ie mi pujeca luzypli (to tai zirjbo cai vau toi) zo <<go'i>> .iku'i .ei zasti fa le nuncumki be lo naldikni se spicru poi kakne lenu sisti vi lo crastu mu'anai

.i ta'onai cimai sera'a le romoi pemci ku le da'aremoi vlali'i cu binxo <<lo'u lo pamei seizga le'u>> vau lu'anai .io .i banzu fa ta

ni'o levi pemci cu pilno pa leimi terga'i be fe le sumti tcita purste (sei mi camdji djuno be leri romoi tarmi se'u) .i mi pilno <<lo'u sexebe'i le'u>> zo <<be'i>> vi le pemci noi se tcita <<lu le te pemci .e le se binxo vau li'u>> .i [<<lo'u sa'a]

ko mi zasyspo
.ije mi ba vuzyvuzyxru
sexebe'i do
.i makfa
.i roroi ku
mi'o zukte ra


le'u>> vau sa'a

Finally, Bob promised to try writing some of his 'natural' Lojban (almost everything else he's written in the language has been constrained to a limited vocabulary, such as the short readings in the draft textbook lessons). The issue is so delayed that he promised to keep it short this time (relatively). Remember. No translation is provided. You can write and ask specific questions if you make the attempt and get lost. I can't promise perfection at this stuff yet.

.uo .uonai .i dukti .i na go'i .i mi ciksi
ni'o zu'u zo <<.uo>> noi smuni lenu mulno cilmo ku'o cu mapti leza'i da lojbo gerna .e leza'i de lojbo gismu .i so'u cilre girzu zu'o tadni cu mulno .i so'u drata ke cilre girzu caze'a penmi joi tadni vijevajevuku .i cabi'iba le karni ke cabna selci po'u <<lu ju'i lobypli li'u>> cu se pagbu loi lojbo selsku .i ri krasi selci'a bau la lojban. (to mi bau la gliban. na pensi ja ciska [sei .ue .o'a lo pluja logji selsku po'u dei cu frili se'u] loi mulno selpei pu lenu mi samci'a dei pe ve'a toi) .i la maiky'elsym. cu te pemci bau la lojban gi'e ca mansa ke gerna drani .i la djan. kau,n. cu dikni skuci'a bau la lojban vi levo'a seirkarni .i .ua la lojbangirz. cu se ganzu fo loi lojbo .i ri mutce snada ke mulno selzu'e mi'o pe ve'u
ni'o zu'unai zo <<.uo>> .ebazibo zo <<.nai>> cu mapti le lojbo pu'u farvi .i ri ba'e na mulno .i .ei .uonai le ctucku vau .i mi dunku ri ki'u lenu mi nalbanzu mulri'a .i mi xaksu loi dukse temci le zu'o ganzu joi si'orcanja joi flagau joi dintro .ianai .au .i le vlacku gunka cu balvi .i .ei la lojbangirz. ba banro zukte .i .ei loi zmadu be loi ca lobypre ba cilre .a'o le bangu .i .ei piro lei cempre ba se ganzu fo lenu pejri'a le cecmu
ni'o .einai mi zukte pamei .i .ei roko zukte .i ko cilre la lojban. .i ko dunda piro lei jdini poi te marde pagbu be le selfai poi dinselxaksu .i .uuse'i loi nalbanzu jdini cu seldu'u mi gi'e rinka lenu mi gunka dirsno
ni'oku'i .ia ba'a le tcini cu xagmaubi'o .i piso'ado ba sidju .i so'ida ba cilre je pilno la lojban. .i .ai .au .eicai ko morji je tinbe le selsku cnimu'i be la lojbangirz. be'o po'u <<lu .e'osai ko sarji la lojban li'u>> .i ganai ko go'i gi la lojban. snada gi'e bangu do
ki'emi'e la lojbab. po'u la bab. leceval,ier.
fe'o

Translations

Our first translation is John Cowan's first paragraph, which we'll repeat so you have it handy:

ni'oni'o le nimcitno goi ko'a po'u la ka,as. noi ko'a melbi bai jorne le sramudri noi ke'a kusru zi'e po le dabli'e goi ko'e po'u la danlu ku secau le kecti .i lei prenu poi ke'a se saptutklu zi'e po ko'e ca zbasu lo derxi be loi mudri bei leko'a jamfu noi ke'a cisyselpli .ibabo le voksa noi ke'a cladu je klina zi'e po'e la nebnauzag. noi ke'a te pemci je tajnau cu fegcru <<lu ganai le do fagryrinka cu rinka lei fagri ja'e le nimcitno se jukpa gi do lifri lenu mi setca le xarci le do se citka poi ke'a romoi ko'u li'u>>

Pidgin translation:

Totally new subject. The woman-youth known-as she1 who-is-identically the-one-called Kaa who-incidentally-is beautiful, forcedly is-joined-to the upright-wood such-that it is cruel and such that it is specific-to the fight-leader known-as he2 who-is-identically the-one-called Animal without the mercy. The-mass-of persons such-that they are-simple-tool-culture-people which-are-specific-to him2 simultaneously make a heap-of some wood at her1 feet such-that they are-sexually-used. And-then the voice such-that it is-loud and is-clear and is inalienably-possessed-by the-one-called Beautiful-male-buttocks who-incidentally-is a poem-maker and a superlative-man angrily-utters, quote, if your fire-causer causes some fire with-result the woman-youth is-something-cooked then you experience that I insert the weapon into your something-eaten such-that it is last, unquote.

English original:

The lovely woman-child Kaa was mercilessly chained to the cruel post of the warrior-chief Beast, with his barbarian tribe now stacking wood at her nubile feet, when the strong clear voice of the poetic and heroic Handsomas roared, "Flick your Bic, crisp that chick, and you'll feel my steel through your last meal."

John's notes on the translation:

There doesn't seem to be any way to say "mercy" in Lojban. I substituted "pity", which is adequate in the negative, because "pitiless" = "merciless", but pity and mercy are quite distinct in the positive. [Bob: "mercy" can be derived as a scalar negative either from the gismu for "cruel" or for "severe". Since he's already used "kusru" in the text, "naljursa" or "to'e jursa" seem like better choices.]

I attempted to avoid the excessive use (to my way of thinking) of tanru in much of the Lojban I've seen to date, in favor of relative clauses. On the other hand, I'm far from sure that I have their syntax correct. [Bob: He was close, omitting only the zi'e connectives for multiple relative clauses and phrases attached to the same sumti.]

I'm rather proud of "la nebnauzag." = "Handsomas".

The clumsiness of "fagryrinka rinka lei fagri" is a deliberate attempt to translate "Flick your Bic" into something that sounds equally idiotic. Unfortunately Handsomas loses his terseness (strong, silent male type) in Lojban, at least in my Lojban.

[Bob adds: The Bulwar-Lytton competition is actually a contest for the worst first-line of a novel. John's translation breaks it up into multiple sentences, but it is possible to combine all the separate Lojban pieces into a single sentence. (To make elidable terminators easier to track, I'm removing John's excess relative clauses; I'm also making a couple of other minor changes that I think are improvements):

ni'oni'o ba lenu le melbi nimcitno po'u la ka,as. goi ko'a bai jorne binxo le kusru sramudri po le dabli'e po'u la danlu goi ko'e secau loka naljursa kei kei ca lenu lei se saptutklu po ko'e cu zbasu lo derxi be loi mudri bei leko'a cisyselpli jamfu kei le cladu je klina voksa be la nebnauzag. noi te pemci je tajnau cu fegcru <<lu ganai ledo fagryrinka cu rinka lei fagri ja'e le nimcitno se jukpa gi ko lifri leli'i mi setca le xarci ledo se citka poi seri'a romoi li'u>>


The translation is fairly close to John's:

Totally new subject. After the beautiful woman-youth who-is-identically the-one-called Kaa (known-as she1) forcedly joined-becomes-to the cruel upright-wood which-is-specific-to the fight-leader who-is-identically the-one-called Animal (known-as he2) without mercy, and while the-mass-of simple-tool-culture-people which-are-specific-to him2 make a heap-of wood at her1 sexually-used feet, the loud and clear voice of the-one-called Beautiful-male-buttocks who-incidentally-is a poet and a superlative-man angrily-utters, "if your fire-causer causes some fire with-result of the woman-youth cooked-thing then (imperative) experience the experience of my inserting the weapon into your something-eaten which is the causedly-therefore last".

If that was too hard, you may or may not benefit from the Lojban parser output for this sample. In this case, the parser sufficiently matches the current grammar that I had to make no changes to parse the text. (The parser version I used was written last October, showing how stable the core grammar has been for the last year or so.) The capitalized words are all the elidable terminators that were 'left out', each identified by lexeme. The parentheses correspond to multi-part rules in the machine grammar, as explained in the text material accompanying the machine grammar.

({ni'o ni'o} {<[({ba <le [nu ({<le [melbi nimcitno] KU> <po'u [la ka,as. (goi {ko'a} GEhU)] GEhU>} {bai <[jorne binxo] [({<le [kusru sramudri] KU> <po [(le dabli'e KU) (po'u {la danlu KU <goi [ko'e] GEhU>} GEhU)] GEhU>} {<se cau> <lo [ka (naljursa VAU) kei] KU>}) VAU]>}) kei] KU>} {ca <le [nu ({<lei [se saptutklu] KU> <po ko'e GEhU>} cu {zbasu <[lo (derxi {be <loi mudri KU> <bei [le (ko'a {cisyselpli jamfu}) KU]> BEhO}) KU] VAU>}) kei] KU>}) (le {<cladu je klina> <voksa [be ({la nebnauzag.} {noi <[(te pemci) je tajnau] VAU> KUhO}) BEhO]>} KU)] cu [fegcru ({lu <[(ga nai) ({le <do fagryrinka> KU} cu {rinka <[(lei fagri KU) (ja'e {le <nimcitno [se jukpa]> KU})] VAU>}) gi (ko CU {lifri <[le (li'i {mi CU <setca [({le xarci KU} {<le [do (se citka)] KU> <poi [({se ri'a} {ro moi}) VAU] KUhO>}) VAU]>} KEI) KU] VAU>})] FAhO> li'u} VAU)]> FAhO})


Next, we have John Cowan's journal entries. John didn't supply a translation, and I don't know Suzuki's non-Lojbanized first name, but otherwise his writing is quite straightforward. Unlike Michael's writing, and unlike John's first Lojban work above, John is choosing to write in simple, methodical sentences. The result is that he makes few and minor errors. This is the type of sentence I think most anyone can write with ease after the first few draft textbook lessons or some disciplined sentence building using the machine grammar as described in the introduction to that document.

ni'oni'o de'i le djedi po la lunas. vau

ni'o le mutce cmalu jukni cu cpare le djacu tubnu .i carvi ja'e le nu ko'a farlu .i le solri cu febri'a le carvi .i le jukni cu rapli cpare

Dated the day of the Moon.

The much-small-spider climbs-on-surface the water-tube (John has proposed a place structure change matching this usage). Rain, with result therefore of it1 falling. The sun boil-causes the rain. The spider repeatedly-climbs.

[This is a translation of a familiar nursery rhyme. John had used "febvi" alone in the next-to-last sentence, with a strange place structure result. In the last sentence, "krefu" might be preferable to "rapli", being a more literal translation, but given the way nursery rhymes are so often repeated, "rapli" may be a more correct term. John's use of "ko'a" is non-standard, since he never assigns it. In some usages, it isn't necessary, since there is only one possible referent. This isn't the case here, although it is in the following.]

ni'oni'o de'i le djedi po la mars. vau

ni'o le mi tixnu po'u la airin. verba .i ko'a nanca li ci .i ko'a kelci je krixa .i je ko'a sipna .i ko'a bacru fi la gliban. .i mi bacru fi la gliban. .e la lojban. .i mi na'e drani bacru fi la lojban. .i mi zutse .i je mi viska le nu le mi tixnu cu kelci .i glare
ni'o mi jdice le nu mi te cmene le'i djedi le'i plini
ni'o cusku fa la viktorias <<lu mi ga'i ba'e na'e se zdile li'u>> .i di'u xajmi mi .i zo ga'i jibni rimni zo ba'e la lojban. .i zo na'e go'i .i le nu le'i rimni cu porsi cu xajmi mi

Dated the day of Mars.

My daughter, the one named Irene, is a child. She1 is measured in years the number 3. She1 plays-and-cries. And she1 sleeps. She1 utters in English. I utter in English and in Lojban. I non-correctly utter in Lojban. I sit. And I see my daughter playing. Warm! (observative)

I decide I will name the set of days, the set of planets.

Expressed, by Victoria "We (Hauteur!) are not amused." The last sentence is funny to me. "ga'i" approximately rhymes with "ba'e" in Lojban. "na'e", too. The event of the set of rhymes being sequential is funny to me.


[I suspect that John wanted a specific attitudinal where he said "glare", given the context. A lot of possible meanings, otherwise.

I'm wary about naming "the set of days, the set of planets". I think he wanted "lu'e" (lexeme LAhE) in front of the second set, making it that he was naming "the set of days, the symbols for the set of planets". But this is a fairly advanced usage that John probably hadn't seen yet.]

ni'oni'o de'i le djedi po la saturn. vau

ni'o cusku fa la daisets. suzukis. <<lu mi po la zen. skicu lo kilycna fo lo na'e kilycna li'u>>

ni'o mi tirna lo verba zgike .i ri mutce pluka mi

ni'o mi pu tcidu lo cukta be le nu ke pamoi penmi le na'e terdi prenu po'u la tranx. bei la alan. din. fostr.

Dated the day of Saturn.

Expressed by ?Daisets Suzuki "We of Zen describe a sharp-shovel (spade) as a non-sharp-shovel (non-spade)."

I hear some child-music. It (the music) pleases me.

I did-read a book about the-event-of first-meeting with-the non-earth-person who is Tranx, by Alan Dean Foster.

[No problems, and the last sentence was both excellently grammatical, and about as complex as I could expect from most students of the draft lessons studying on their own. The "ke" in the last sentence is legal but unneeded, possibly a remnant learned from something I wrote more than 1 1/2 years ago, when "nu" unnecessarily required the "ke" following it.]


Now, here is the translation and commentary for Michael Helsem's letter. We'll present the text and translation as a whole so you have some idea what the letter is talking about, then go back and look at each sentence. The italicized text is a literal translation, which in places will be difficult to understand. The normal English text is what Michael was apparently trying to say, paraphrased from his own translation and somewhat simplified. It is not always exactly what he said.

di'e zirjbo po'a xatra de'i la xav. po'e la mumast.
The following utterances are a "purplish-lojbanic" (figuratively) letter dated that named 6 which is possessed by Five-month (6th of May).
The following is a letter in "purple Lojban" dated 6 May.

ni'o coido'u mi dopeza ledo nu jarco le mi pemci cu ckire ra'u .i puvi'eku mi mutce cutyzu'e pi'o gi'e pu'i ninpemci jmina la'u lo mo'amei .u'u ne ki'u lemi cabi'ibajenairu'i .ia li'anai zmadu nu mutce cutyzu'e .i ta'onai pu ra ku mi pinka so'u lepu selja'o ti'u
Greetings! I, to you-all, for your showing my poems, am grateful (Most important point!). Earlier, throughout this space-time, I'm very affair-active used-ly, and could-and-did poetry-quantity add (to something) in-amount a too-few-some (Repentance!) which (the-too-few-some is justified by my then-until-later-but-not-continuously (Belief but Obscurity!) greater state of affair-activity. Anyway, before that (an unclear reference), I comment(ed?) on a few of the earlier shown-things, associated with time unspecified.
Hi, y'all. Thanks for showing my poems. Earlier (all over) I was very busy, and can add too-few poems because of my from-now-on, but not continuously, even more busy state. Anyway, before that (the poems), I'll comment on a couple of the previously printed ones, while I'm at it.

.i pamai sera'a le pemci po'u se tcita <<lu leka sarcu kei vau li'u>> ku'o mi na jimpe le krinu be ledo nu punji zo <<cu>> le crastu po zo <<nu>> mu'inai lemi pu nu dunda (sei zo <<sabji>> lu'anai cu drani se'u) lo temge'a tcita (to te'i zo <<ba>> vau toi) pe vi le trixystu ne seba'i .i ti jufra <<lu le nunsti ne sekai leka pratci cu batci mu'anai li'u>> vau ?xu
First, about the poem which is tagged "The Necessity", I don't understand the reason for: your putting "cu" on the front-site of "nu" despite motive of my previously giving ["providing" approximately is correct] an interval grammar tag, specifically "ba" which is at the back-site, incidentally with a motive. This (?) is a sentence about (the quote) "the stopping, characterized by produce-tool-ness, is biting", isn't it?
First, about the poem "The Necessity" (the 3rd limerick in JL11). I don't understand why you put "cu" before "nu", and omitted the "ba" that I had further back. I intended the equivalent of "the stopping, characterized by produce-tool-ness, is biting". (The answer, explained below, is that his original English translation was ambiguous and did not convey this meaning to me.)

.i remai .uocai sera'a le seltcita be <<lu le firgai mu'anai vau li'u>> ku vi le da'amoi vlali'i ku zo <<co>> cu se setca fi <<lu fasnu cictcima li'u>> ja'e <<lo'u lo nalsti nu cictcima fasnu le'u>> (sei zo <<za'i>> basti ?xu be zo <<nu>> se'u) .i zo <<fasnu>> ca nalsarcu .i'a
Second (whew!) about the be-labelled-by "the face-cover ...", in the next-to-last word-line, "co" is inserted into "event-ish wild-weather" therefore with result "a non-stop act of wild-weather event" ("za'i" replaces, doesn't it, "nu"). "fasnu" is unnecessary (Acceptance!).
Second. About the poem "The Unmasking" (also in JL11). If "co" were inserted into the selbri of the next-to-last line, would the result be "a non-stop act of wild-weather event"? Also, "za'i" should replace "nu", shouldn't it? I accept that "fasnu" is unnecessary (responding to one of Bob's comments in JL11).

.iku'i ta rinka lenu mi ninzga lo puze'u nandu .i mi su'oroi pilno zo <<nu>> lo paroi tortei bo fasnu gi'e drata go'i fi lo ranji clatei bo fasnu .i re frica valsi cu sarcu vau pe'i .ije .ie mi pujeca luzypli (to tai zirjbo cai vau toi) zo <<go'i>> .iku'i .ei zasti fa le nuncumki be lo naldikni se spicru poi kakne lenu sisti vi lo crastu mu'anai
But, that causes that I now-observe an in-the-past, for-a-long-time, difficulty. I a-few-times use "nu" for a one-time, short-time-interval, event, and otherly use ... for a continuing, long-time-interval event. Two different words are necessary, I opine. And (I agree) I then-and-now loose-use, (by method of purple Lojban!), "go'i". But (Obligation!) existing is the possibility of non-regular piece-utterances which are able at ceasing at-the in-front-of-site (End of example).
But this raises a difficulty I've had for a while. Sometimes I use "nu" for short, one-time events, and sometimes for long, continuous events. I think two words are necessary. And I agree that I loosely-use "go'i" after a "purple Lojban" fashion. But there ought to exist the possibility of an unsystematic ellipsis which is able to stop short.

.i ta'onai cimai sera'a le romoi pemci ku le da'aremoi vlali'i cu binxo <<lo'u lo pamei seizga le'u>> vau lu'anai .io .i banzu fa ta
(Returning to subject), thirdly, about the last poem, the antepenultimate word-line becomes "a single self-observer", precisely, respectfully. Enough, is that.
Thirdly, in the last poem (in JL11 actually, he meant the next-to last one), the antepenultimate word-line should indeed become "lo pamei seizga". But enough of that.

ni'o levi pemci cu pilno pa leimi terga'i be fe le sumti tcita purste (sei mi camdji djuno be leri romoi tarmi se'u) .i mi pilno <<lo'u sexebe'i le'u>> zo <<be'i>> vi le pemci noi se tcita <<lu le te pemci .e le se binxo vau li'u>>
This poem uses one of my modifications of the argument-label previous-list [I intensely want to know its (the list's) last form.] I use ??"sexebe'i" for "be'i" at the poem, which incidentally is labelled as "The Poet and the Thing Become".
This poem uses an extrapolation from one of the words in the old lexeme BAI list. I sure want to know what the final lexeme BAI list is. I use "sexebe'i" for "be'i" at the poem, which I call "The Poet and the Thing Become".

.i [<<lo'u sa'a]

ko mi zasyspo
.ije mi ba vuzyvuzyxru
sexebe'i do
.i makfa
.i roroi ku
mi'o zukte ra

le'u>> vau sa'a

[" editorially inserted quote for ungrammatical text]

Temporarily-destroy me;
and I will yonder-yonder return
through the medium of you.
It's magic.
All the time,
we do it (with purpose)

[" editorially inserted end-quote and "vau" to make the non-grammatical quoted poem a valid 'incomplete sentence'.


Now lets go over it in detail, answering Michael's questions, and raising a couple of points about the language not covered in draft textbook lessons (or at least not clearly). Don't let the extensive comments imply otherwise - Michael did an outstanding job considering the complex grammar he tried to use. He made fewer errors than his previous attempts, and they were more of the nature of not knowing the right word and/or depending on English idiom, than they were grammatical flaws.

I've broken the text into individual sentences. A parser output is also included for each line. The parentheses match over the whole text, but will not match perfectly for each line (there will usually be an extra right parenthesis at the end of a sentence corresponding to a left parenthesis near the beginning of the text). We'd like feedback on how useful the parser outputs are to you. Do they tell you anything, or are they just so much gibberish?


di'e zirjbo po'a xatra de'i la xav. po'e la mumast.
({di'e CU <[zirjbo po'a xatra] [(de'i {<la xav.> <po'e [la mumast.] GEhU>}) VAU]>}
The following utterances are a "purplish-lojbanic" (figuratively) letter dated that named 6 which is possessed by Five-month (6th of May).

Comment: To date the letter, we used "de'i" (associated with date ...) and not "ca". Using "ca" claims that it was a letter at the time of the date given, without saying anything about what it is now. In addition to "de'i", "tu'i" and "ti'u" perform the same function for letters, respectively with regard to location and time-of-day. None of these three serve a 'tense' function; they attach a date, location or time as a labelling relation.


ni'o coido'u mi dopeza ledo nu jarco lemi pemci cu ckire ra'u
{ni'o <coi do'u>} {<[({<[({<[({<[( mi {do <pe [za (le {do <nu [jarco ({le <mi pemci> KU} VAU)] KEI>} KU)] GEhU>}) cu ({ckire ra'u} VAU)]
Greetings! I, to you-all, for your showing my poems, am grateful (Most important point!).

Comments: In his original, Michael omitted the "do'u" after "coi". This had strange (and humorous) effects: in "... coi mi dopeza ..." the "coi" grabs hold of the next sumti ("mi"), causing Michael to greet himself and imply that the rest of the letter is addressed to himself. The rest of the sentence then said (to himself) roughly that "You-all are grateful to the showing of my poems for something unspecified." Moral: either be specific who you are greeting, remember to terminate the vocative, or immediately start a new sentence. Michael also had attached "lemi pemci" to "jarco" with the "be" link. This is grammatical, but wasn't necessary. "nu" and other abstraction clauses presume that an entire sentence follows, and you can include all of the sumti you want without attaching them. Remember that you have to terminate abstractions (with "kei") if another sumti follows (it doesn't in this case), or you'll have the same kind of strange effects as the vocative "coi" caused.

Mike translated "ra'u" in the above sentence as an incidental "(of course)"; his Lojban actually indicates that his gratefulness was his most important point of the letter. Note that Michael's "ra'u" attaches specifically to "ckire", and not to the whole sentence. To apply a discursive or attitudinal to the sentence as a whole, you must put it at the beginning of the sentence, or express the normally elided "vau" at the end of the sentence and attach the "ra'u" to that.


.i puvi'eku mi mutce cutyzu'e pi'o gi'e pu'i ninpemci jmina la'u lo mo'amei .u'u ne ki'u lemi cabi'ibajenairu'i .ia li'anai zmadu nu mutce cutyzu'e
i [({<pu vi'e> ku} mi) CU ({<[mutce cutyzu'e] [(pi'o KU) VAU]> gi'e pu'i <[ninpemci jmina] [(la'u {<lo [mo'a mei u'u] KU> <ne [ki'u (le {mi <[(ca bi'i ba) (je nai) (ru'i ia li'a nai)] [zmadu (nu {<mutce cutyzu'e> VAU} KEI)]>} KU)] GEhU>}) VAU]>} VAU)]>
Earlier, throughout this space-time, I'm very affair-active used-ly, and could-and-did poetry-quantity add (to something) in-amount a too-few-some (Repentance!) which (the-too-few-some is justified by my then-until-later-but-not-continuously (Belief but Obscurity!) greater state of affair-activity.

Comment: Some complicated but cute tense constructions that come close to what Michael intended. What he said, "puvi'eku", is kind of self-contradictory; "vi'e" means throughout an interval of space-time (which includes a time dimension), but Michael has already specified a specific time as "pu". "vi'u" is the corresponding space-only interval 'tense', so "puvi'uku" might accomplish what he intends. Another possibility is "pufe'eroroi". But Michael has also relied incorrectly on English idiom, or he wouldn't have said this at all (Michael isn't really everywhere doing anything, though he might feel it sometimes), so I didn't change what he wrote.

"puvi'uku" sets the time in the past. Michael then adds a second tense, "cabi'ibajenairu'i", apparently not realizing that this second complicated tense of the sentence is an offset from the first. Compare the English "I realized three weeks ago that I would shortly be very busy."

Note that the parser output breaks the compound "cabi'ibajenairu'i" into individual words. "ca" means "at the same time as" the tense implied by "puvi'eku", which is of course sometime in the past. The "bi'i" indicates that there is a time interval involved, in this case starting at "ca" and ending later at "ba". But "ba" is still an offset from the initially set (past) tense of the sentence; thus the sentence refers to a situation that might be entirely in the past from NOW, the time when he wrote the letter, though this seems not to be what Michael intended. Moral: use multiple tenses in a sentence only when you are fairly sure you know what you are doing.

One error is as much our fault as his: Michael used "pi'o", intending "instrumentally" (which normally should be "sepi'o", but was erroneously listed as "pi'o" in the October 1988 cmavo list). I left "pi'o" unchanged, though, since it isn't clear what Michael really meant by "instrumentally". In this context, I would guess that "instrumentally" would mean that he was busy 'using something'. "pi'o", on the other hand, means that his busy-ness was 'being used by someone/something'; i.e. possibly, he was busy in his employment. Hopefully, this is what he meant.

Nora recommended adding the "ne" in "... .u'u ne ki'u"... . Without "ne", the causal "ki'u" phrase modifies the whole bridi (in this case, the 2nd half of a compound bridi centered on selbri "ninpemci jmina"); with the "ne", the phrase is an explanation of the 'too-few-some'.

Based on my best guess as to what Michael intended, I changed his causal from "ri'a" to "ki'u". He seems to be presenting a justification for his too-few-some of poems. With "ri'a", he is saying that his then-until-later-busy-ness physically caused the too-few-some. There is a second interpretation of Michael's English such that, using "ri'a" and no "ne", Michael's busy-ness caused the adding of too-few-some; this alternative wasn't noticed until just before press time.

.i ta'onai pu ra mi pinka so'u lepu selja'o ti'u
<i ta'o nai> <[(pu ra) mi] CU [pinka ({<[so'u BOI] [le (pu selja'o) KU]> <ti'u KU>} VAU)]>}
Anyway, before that (an unclear reference), I comment(ed?) on a few of the earlier shown-things, associated with time unspecified.

Comments: "pu ra" is unclear here, the previous sentence is just too complex to figure out what 'that' is. My best guess is that he intends the "too-few poems", but only by assuming English idiom; Michael isn't commenting before "the too-few-some poems (that) were added", since the poems were already written; he is commenting before he includes them. Avoid vague references - they are unfair to the reader/listener. My 'lojbanic' preference would be to say something like "le pemci kansa deipeza" ("the poems accompany this writing"), then followed by (in this sentence) "pu la'edi'u" ("before this accompaniment") instead of "pu ra". Michael also could also have been tantalizing and correct with "pu da'u" ("before what I will eventually express")

Michael originally had "pu ra ku", which became illegal about a year ago when we increased the varieties of elidable terminators, so it isn't Michael's fault. "ku" closes a description, or it early-terminates a sumti tcita that isn't accompanied by a sumti such as "ra" thus preventing it from absorbing a following sumti unintentionally; e.g. "puku mi klama" ("Earlier, I came.") vs. "pu mi klama" ("Before me, (something) comes")

Michael translated "ti'u" here as "while I'm at it". "ti'u", as mentioned above, is used to tag letters and events ('the 6 o'clock news'). Using "ti'u" without a time is merely confusing. There may be a way to say what he wants briefly, but it isn't apparent to me; Michael already used one vague tense marker ("pu ra") in the sentence, and Lojban gets really nebulous when you pile vagueness upon vagueness. Given my alternative "pu la'edi'u", described above, the simple tense "ca" on the selbri, or "caku" at the end would convey "while I'm at it". Other possibilities exist, perhaps using the sumti tcita "po'i".


.i pamai sera'a le pemci poi se tcita <<lu leka sarcu kei vau li'u>> ku'o mi na jimpe le krinu be ledo nu punji zo <<cu>> le crastu po zo <<nu>> mu'inai lemi pu nu dunda (sei zo <<sabji>> lu'anai cu drani se'u) lo temge'a tcita (to te'i zo <<ba>> vau toi) pe vi le trixystu ne seba'i
{i pa mai} {<[(se ra'a) ({le pemci KU} {poi<[se tcita] [(lu {<[le (ka {sarcu VAU} kei) KU] vau> FAhO} li'u) ku'o]>})] mi> CU <[na jimpe] [(le {krinu <be [le (do {nu <punji [({<zo cu> <[le crastu KU] [po (zo nu) GEhU]>} {<mu'i nai> <le [mi (pu {nu <[dunda (sei {zo sabji lu'a nai} cu drani se'u)] [({lo <temge'a [tcita ({to te'i} {<[zo ba] vau> FAhO} toi)]> KU} {pe <vi [(le trixystu KU) (ne {se bai KU} GEhU)]> GEhU}) VAU]> KEI})] KU>}) VAU]> KEI}) KU] BEhO>} KU) VAU]>})
First, about the poem which is tagged "The Necessity", I don't understand the reason for: your putting "cu" on the front-site of "nu" despite motive of my previously giving ["providing" approximately is correct] an interval grammar tag, specifically "ba" which is at the back-site, incidentally with a motive.

Comment: Though very complex, this one made sense to me. I took out extraneous "be..bei..." constructs and replaced "lu...li'u" by "zo" when he quoted a single word. The only grammatical error was minor he originally had *"... po'u se tcita ...". "po'u" indicates a relative phrase (a sumti or tagged sumti) follows. You need "poi" for a relative clause (a full bridi).

Michael's question relates to the 3rd limerick in JL11. His original third and fourth lines read ".iku'i le pratci / nu sisti ba batci ...", which I changed to ".iku'i le pratci / cu nu sisti kei batci", and he is asking why I left the "ba" tense off my rewritten version, using "cu" instead.

Answer: His original English translation of the limerick is a superb example of ambiguous English. He said: "But the produce-tool / cessation bites ...". This can mean "But the produce-tool is a cessation biter ..." or "But the produce-tool cessation is a biter ...". One always assumes non-figurative usage in Lojban, so only the first is plausible - a tool can conceivably 'bite', a 'cessation' isn't a physical thing.

Note that his English makes no use of future tense; I probably ignored the "ba" since it didn't match his English; it may have looked like a typo. However, his original is grammatical, and means "But the produce-tool cessation will bite ..." which matches closely the second, figurative, usage that Michael intended.

Given my interpretation of his English, my "cu" made the "ba" ungrammatical. Furthermore, if I removed the "ba", a "kei" was required where the "ba" was to close off the "nu", else the event clause would include the "batci": "nu sisti batci" ("event of cease-r biting ...").

The peril of trying to write Lojban poetry at this early stage is that the people who read it may not understand it, even if it IS correct. Figurative metaphors are sometimes chancy in English poetry - in Lojban, with no idiom, they are nonsense (and in most cases make illegitimate cultural references, a no-no in a culturally neutral language). Can good poetry be written using only 'analytic' metaphors? I'd like to see Michael try.

The answer, then, seems to be that my change was incorrect. Hopefully, we'll remember this when it comes time for Michael's first Lojban poetry book.


.i ti jufra <<lu le nunsti ne sekai leka pratci cu batci mu'anai li'u>> vau ?xu
i (ti CU {jufra <[lu ({<[le nunsti KU] [ne ({se kai} {le <ka [pratci VAU] KEI> KU}) GEhU]> cu <[batci mu'a nai] VAU>} FAhO) li'u] [vau xu]>})]
This (?) is a sentence about (the quote) "the stopping, characterized by produce-tool-ness, is biting", isn't it?

Comments: Michael's sentence was grammatical, but malglico (very 'English').

Warning! The demonstratives "ti", "ta", and "tu" are unconscionably vague in written text! They presume that the listener has a way to tell what the speaker is indicating. In writing, the only way to do this would be to draw an arrow to the referenced item. ("ti" might be used to 'obviously' refer to the letter itself, but "deipeza" is much clearer; better to avoid these by rule in writing.)

English pronouns are extremely tricky to translate into Lojban, since we use them so sloppily in English. It will be the mark of malglico translation (as opposed to original Lojban expression) to see sloppy pronouns when Lojban has so many specific mechanisms for 'pronoun' anaphora (oops! I promised to use "ba'ivla".)

In this case, no ba'ivla could be clear, since Michael intended to refer to what he had originally written in his letter to me, something I had to dig out of the files to check. Michael needed to be explicit: "lemi selsku" ("my expression", and possibly even "le ba'emi selsku" to emphasize that it was his version that he wanted me to check). Even using his English translation of the letter, I would not have assumed 'this' to refer to his original letter; I only dug it out because there was no "ba" in the version printed in JL. Moral: an unambiguous language requires that its speakers remember to check for possible misunderstandings of vague references.

Having gotten past the first word of the sentence, there is a major logic problem of label/reference confusion, a problem of the sort made famous in Through the Looking Glass, when Alice talks with Humpty Dumpty about what the song is, what the song is called, what the name is called, etc. Michael is saying in this bridi that something ("ti") is a sentence about the quoted text given. The sentence he is referring to is not about the quoted text. He means that his sentence was intended to say the same thing as the quoted text. The term for this is 'indirect discourse', and is most familiar in English in the form "He said that he was going to the store." The cmavo "la'e" accomplishes indirect reference in Lojban; it refers to the referent of any sumti it precedes. So Michael should have had "la'elu" at the start of the quote.

I added the "vau" to have the true/false question "xu" clearly apply to the whole sentence and not just to the quote; I wasn't sure what he was asking. Colloquial Lojban will almost certainly express yes/no questions with the "xu" at the beginning - it is simplest grammatically. If the question is about a specific part of the sentence, something we indicate in English with emphatic stress ("You went to WORK, didn't you?"), then the question will probably still be pre-marked with "pau", warning the listener to expect a question.


.i remai .uocai sera'a le seltcita be <<lu le firgai mu'anai vau li'u>> ku vi le da'amoi vlali'i ku zo <<co>> cu se setca fi <<lu fasnu cictcima li'u>> ja'e <<lo'u lo nalsti nu cictcima fasnu le'u>> (sei zo <<za'i>> basti ?xu be zo <<nu>> se'u)
[i re mai uo cai] [({<[se ra'a] [le (seltcita {be <lu [({le <firgai mu'a nai> KU} vau) FAhO] li'u> BEhO}) ku]> <vi [le ({da'a moi} vlali'i) ku]>} {zo co}) cu ({se setca} {<[fi (lu {<[fasnu cictcima] VAU> FAhO} li'u)] [ja'e ({lo'u lo nalsti nu cictcima fasnu le'u} {sei <zo za'i> CU <[basti xu] [be (zo nu) BEhO]> se'u})]> VAU})]>
Second (whew!) about the be-labelled-by "the face-cover ...", in the next-to-last word-line, "co" is inserted into "event-ish wild-weather therefore with result "a non-stop act of wild-weather event" ("za'i" replaces, doesn't it, "nu").


Comments: If the "whew" applies to the stuff that preceded being complete, it should have gone before this sentence, which starts a new topic; Michael's usage suggests that this second point is his concluding one, which is not the case. The attitudinal would be clearest in a separate 'sentence': "... .i .uocai .i remai ...".

In an unambiguous language, quotation can be a problem, especially, as in Michael's letter, when he wants only a few words out of context. Lojban's grammar requires that a quotation be fully grammatical standing on its own, and most out-of-context quotes are not. The quote must also match the original, of course - you can't normally add words inside the quote to make the text grammatical without telling the listener/reader. In English print, editors abbreviate quotes and correct grammar by putting interpolated text in brackets and using ellipsis marks to indicate omitted text (e.g. "editors abbreviate ... [with] brackets ..." as a shortening of the last sentence). In Lojban, all punctuation must be spoken, so this isn't easy to do while remaining clear (we do have a way of expressing this when appropriate, though, which I'll get to in a moment).

Thus, in Lojban, if what you want to quote is less than a complete sentence, or has omitted text, the easiest way to do this is with the 'error-quotes' "lo'u ... le'u". I changed Michael's second out-of-context quote to use these (as well as some others further on in the letter).

The other way is to make the result grammatical and mark all the changes to the original quote. The Lojban words for "etc." are "vau" and "mu'anai", The former is a grammatical particle, of course, marking that there are no more sumti being expressed. The latter is a discursive having no grammar, and meaning roughly "concluding my examples"; it thus does not serve the all-purpose role of English "...". Lojban simply cannot do the latter with all the flexibility of English and still be grammatically unambiguous. I'll discuss this further below when I talk about Michael's comments on "go'i".

First an interlude on editorial manipulation of quotes (you probably won't mind a break from plowing line by line through the letter.)

We've put several features in the language to make manipulating quotes easier, some of which I've added to Michael's letter. However, for extensive use of out-of-context quotes, only "lo'u... le'u" is practical if you want to keep clear what was quoted and what wasn't.

To summarize the quote manipulation features briefly:

  • the word "sa'a" is the most important. Attaching like a discursive to a word or a construct, it says that word or construct wasn't actually part of the original quote. This allows you to insert text in an ungrammatical text to make it grammatical, while making it clear what you added. The result is similar to an editor's bracketing text that was added (usually after omitting a longer text in a quote) to allow the remainder to make sense (e.g. "The ... [woman] sat next to me"). (By the way, for sticklers who ask how to quote text with the word "sa'a" embedded, a second "sa'a" after the first says that the first is really there. We think we've provided adequately for the infinite series of complications that can be concocted along this line of analysis.)
  • the word "li'o" followed by "sa'a" to show that the word "li'o" isn't really part of the quote can be used to replace the "..." ellipsis in a quote provided that the result is grammatical. This usually requires some other modifications to the quote, each marked with "sa'a" as well.
  • the 'unquote parentheses' "to'a...to'i", again with "to'a" marked with "sa'a" to show they are to be taken metalinguistically, are used to interrupt a quote to get back to the "quoter's" level of the sentence. The classic use of this is for conversation quotations ("That window", he said, pointing down the hallway, "is dirty"). In such a split quote, the two pieces are each ungrammatical, but the whole is grammatical. The division in the text is purely stylistic, providing a certain emphasis. Note that the material between the two quotes is also fully grammatical. In Lojban we express this by having the quote be the basic text, jumping out-of-quote with "to'asa'a", and then returning to the quote with "to'i" when the commentary is complete. Again, both the inside of the parentheses and the outside have to be completely grammatical as separate units. Furthermore, the 'unquote' has the grammar of a parenthetical remark - a 'free modifier', and is mildly constrained on where it can be inserted (not in the middle of a digit string representing a number, for example).

For those who ask, interrupted quotes are important to Lojban for literary usage (they are seldom used in spoken language); we added them after Athelstan tried to translate Saki (see JL10). They are used in all natural languages that we've investigated, with few constraints on where the interruption can occur. The alternative is the stylistically 'boring' "He said '...'. She said '...'. He said '...'." which is the 'natural grammar' form for both "cusku" and "bacru".

Others ask why have all these unfriendly limits and constraints, as Michael does later in his letter. The answer is 'unambiguity', of course. For these people, there is the 'error quote' that tells the listener to treat the quote as a literal quote, but to not worry about the grammar of the content. (For out-of-context quotes, though, putting a "li'o" discursively after the ending quote would show that text was omitted, a useful notice that makes it clear that the 'error quotes' were used for out-of-context ellipsis, and not to mark truly ungrammatical expression.)

Now, finally answering Michael's question as I understood it, which is based on his JL11 poem entitled "The Unmasking", if you insert "co" into "nu fasnu cictcima", you get "nu fasnu co cictcima". The former means "event of being an occurring type of wild-weather", the latter means "event of being an event of-type wild weather", and is equivalent to "nu cictcima fasnu" ("event of being a wild-weather event"). The difference between the two forms is that the final brivla determines the place structure for any attached sumti. Since this selbri is bare within a "nu" abstraction, that criterion does not apply here (the place structure of a "nu" abstraction is constant, and there are no expressed sumti anyway). The choice then is based on whether the event that Michael wants to describe (it is inherently an event because of the "nu") is a "(weather type of event)-event" or an "(occurring-type of-weather)-event". The "fasnu" still seems redundant.

Michael also asks about "za'i" in a metalinguistic parenthesis (he originally had this parenthesis in the middle of the quote - this would have required "sa'a" as mentioned above, and made the quote grammatically difficult to follow, where in this case the exact grammar was relevant to his question).

"za'i" could indeed be used instead of "nu". "za'i" refers to a subset of event/state abstractions, for which the general word is "nu". "za'i" states are those event situations which are being seen as essentially stable and uniform within the event, which have either a finite or an infinite duration, as opposed to being a point event, and which, if either starting or stopping boundary exists, it is effectively a single point. Thus we normally think of 'being awake' as a 'state', whereas being hit by a car is a point event, expressed by the achievement abstractor "mu'e". (Obviously, we don't necessarily think of being hit by a car as an achievement, but by Aristotelian logic it is, based on the 'shape' of the event, called the "event contour".)

I may write more on abstractions for either JL14 or JL15, since they relate also to the changes made to the tense design. Writing them up will help me in creating the textbook sections on these topics. But no promises.


.i zo <<fasnu>> ca nalsarcu .i'a
i <[zo fasnu] [ca ({nalsarcu i'a} VAU)]>}
"fasnu" is unnecessary (Acceptance!).


.iku'i ta rinka lenu mi ninzga lo puze'u nandu
{i ku'i} {ta CU <rinka [(le {nu <mi CU [ninzga ({lo <[pu ze'u] nandu> KU} VAU)]> KEI} KU) VAU]>})
But, that causes that I now-observe an in-the-past, for-a-long-time, difficulty.


.i mi su'oroi pilno zo <<nu>> lo paroi tortei bo fasnu gi'e drata pilno fi lo ranji clatei bo fasnu
i (mi {<su'o roi> <[(pilno {<[zo nu] [lo ({pa roi} {tortei bo fasnu}) KU]> VAU}) gi'e ({drata pilno} {<fi [lo (ranji {clatei bo fasnu}) KU]> VAU})] VAU>})]
I a-few-times use "nu" for a one-time, short-time-interval, event, and otherly use ... for a continuing, long-time-interval event.

Comment: Michael had used "go'i" here where the second "pilno" appears ("... gi'e drata go'i fi ..."). "go'i", however, refers to the bridi in the previous sentence, not to the first half of this compound bridi. One of "go'i"'s major purposes, is to enable answering "xu" questions. If "go'i" counted bridi that were pieces of sentences, you would have trouble answering a "xu" at the beginning of this compound sentence asking whether the entirety were true.

There is no member of lexeme GOhA for Michael's purpose, either - there are several other ways to rephrase what he wrote. For example, in a compound bridi where there is a common selbri like this, you can rearrange the sentence. I've used subscripted tenses to deal with the dichotomy of times, and added in the dichotomy discursives "zu'u" and "zu'unai" to show their usage:


.i mi pilno zo <<nu>> lo zu'u paroi tortei bo fasnu ne su'oroikuxipa .e lo zu'unai ranji clatei bo fasnu ne su'oroikuxire
I use "nu" for on-the-one-hand a one-time, short-time-interval, sometimes1, AND, on the other hand a continuing, long-time-interval event, sometimes2.

Other ways to express the time dichotomy include 'termset' constructions that allow logical connections between sets of sumti.


.i re frica valsi cu sarcu vau pe'i
i [({re BOI} {frica valsi} KU) cu (sarcu {vau pe'i})]>
Two different words are necessary, I opine.

Response: Indeed. And we have five (or more). As mentioned above, "nu" is the general, abstract event/state term, while we have for words for 4 various kinds of events/states, also within lexeme "nu". In most communications, it is not necessary that you be specific about the 'event contour'; it is irrelevant or obvious from the context. On the other hand, if you wish to make a distinction in which the kind of event is important, you can do this as well.

The Lojban concept is to minimize metaphysical assumptions. In this case, by this, we mean that we don't require the speaker to assume that being awake is a 'state', while being hit by a car is an 'achievement'. You can talk about the 'state' of being hit by a car or the 'achievement' point-event of being awake. The former is a bit mind-boggling, I'll admit; the latter occurs for me sometimes when Nora gets up in the morning to go to work, at the inhuman hour of 6AM. Of course, you can choose not to decide how to think of the event, and use "nu", since Lojban doesn't require you to specify the type of event in order for it to be an event. This is the same argument that justifies optional tense in Lojban.


.ije .ie mi pujeca luzypli (to tai zirjbo cai vau toi) zo <<go'i>>
<i je ie> <mi [(pu je ca) ({luzypli <to [(tai {<zirjbo cai> vau}) FAhO] toi>} {<zo go'i> VAU})]>}
And (I agree) I then-and-now loose-use, (by method of purple Lojban!), "go'i".

Comment: The logical connective between sentences here seems unnecessary. The ".i" at the beginning of most Lojban sentences sentence can be translated as the run-on "and" that often occurs at the beginning of English sentences like Michael's; in Lojban you have to run-on, unless you are done talking or changing subjects. malglico!

(You use the logical connectives between sentences when there is logical import to the joining. For example, in "da blanu .i da klama", because there is no logical connective, the two "da"s could refer to different 'something's. In "da blanu .ije da klama", both "da"s are 'bound' under the same logical scope, and hence must refer to the same thing.)

Michael loosely-uses the English meaning of the word "loose" in the lujvo "luzypli". We want Lojban words to have a simple, pure meaning, without importing all the connotations of English. I can think of several meanings for the tanru 'loose-user' that keep us from wanting to import this phrase from English. What I think Michael intended is more clearly rendered as "jbige'a pilno" ("approximate-grammar use") or "jbidra pilno" ("approximately-correct use") With these clearer tanru, Michael's comment practically answers itself.

Michael also probably wanted "ta'i" instead of "tai" in the parenthesis. Both words convey some senses of English "in manner". "ta'i" refers to a form, while "tai" refers to a method. The context indicates that he expresses in the form of purple-Lojban.


.iku'i .ei zasti fa le nuncumki be lo naldikni se spicru poi kakne lenu sisti vi lo crastu mu'anai
{i ku'i ei} {zasti <[fa (le {nuncumki <be [(lo {naldikni <se spicru>} KU) (poi {kakne <[le (nu {sisti <[vi (lo {crastu mu'a nai} KU)] VAU>} KEI) KU] VAU>} KUhO)] BEhO>} KU)] VAU>})
But (Obligation!) existing is the possibility of non-regular piece-utterances which are able at ceasing at-the in-front-of-site (End of example).

Comment: Good sentence, with sumti rearrangement to liven it up. The relative clause gives a pleasantly Lojbanic active sense in talking about the 'abilities' of an utterance. The clause could more briefly be expressed with a 'potential tense', lexeme CAhA: "poi ka'e sisti vi lo crastu" ("is-innately-capable-of ceasing at-the in-front-of-site"). Of course, in a sentence based on "cumki", the tense can be optionally omitted entirely, since 'possibility' implies 'timeless capability' in most circumstances.

I unfortunately can't agree with Michael's statement. As I stated above, the grammar cannot provide total flexibility to just stop any old place and have the listener able to unambiguously determine what you meant. Pragmatically, this may be desirable, but most incomplete sentences are grammatically ambiguous. My favorite example is the English "I went to the window quickly and ...", which has innumerable possible follow-ons, using several different grammatical forms. In English literary usage, ambiguity is permissible and sometimes desirable. In Lojban, it cannot be, or we don't have any special claim as a language. Lojban, on the other hand, has expressive flexibility that isn't possible in English, derived from the increased varieties of vagueness (semantic ambiguity) that Lojban gains by not having an ambiguous grammar.

Pragmatically, whether people will speak Lojban grammatically, or will 'loose-use' Lojban is unknown. Michael seems to have enough command of the language to help set the example for the former, though this will undoubtedly cramp some of his poetic style. But then, Lojban poetry that uses Lojban's unique distinction between grammatical and semantic ambiguity will truly be a new art form.

Incidentally, as a aside that seems relevant here, John Hodges queries Michael for comments on the following (with input sought from anyone else with an opinion):

"I've heard that poetry gains beauty from a disciplined structure, that precision is not the enemy of art. If so, great. I have noticed that the first thing many people try to do with a 'logical' language is write poetry in it. This may be only because it is new and therefore exotic."

Presuming John's intent: on the other hand, Lojban may impose a discipline that may make it possible to achieve 'beauty' that isn't possible in a language as unstructured as English or indeed in any natural language. What do people think on this subject?


.i ta'onai cimai sera'a le romoi pemci ku le da'aremoi vlali'i cu binxo <<lo'u lo pamei seizga le'u>> vau lu'anai .io
(i ta'o nai ci mai) ({<[se ra'a] [le ({ro moi} pemci) ku]> <le [(da'a re moi) vlali'i] KU>} cu {binxo <[lo'u lo pa mei seizga le'u] [vau lu'anai io]>})]
(Returning to subject), thirdly, about the last poem, the antepenultimate word-line becomes "a single self-observer", precisely, respectfully.

Comment: Some confusion here as to which poem Michael meant. Michael really means the next-to-last poem in JL11; the last one pertains to Sapir-Whorf. From examining my comments, Michael is choosing one of my alternate expressions for the line in question.


.i banzu fa ta
i [banzu ({fa ta} VAU)]>
Enough, is that.

Comment: Another dangling demonstrative. Who knows what "ta" is? The bare ".i banzu" would have been more clear, and Lojbanic, although this sentence is the truly appropriate place for the ".uocai" he put in an earlier paragraph, as anyone who has plowed through all this can agree.

ni'o levi pemci cu pilno pa leimi terga'i be fe le sumti tcita purste (sei mi camdji djuno be leri romoi tarmi se'u)
ni'o <[({<le [vi pemci] KU> cu <pilno [({pa BOI} {lei <mi [terga'i (be {fe <le [(sumti tcita) (purste {sei mi CU <camdji [djuno (be {le <ri [(ro moi) tarmi]> KU} BEhO)]> se'u})] KU>} BEhO)]> KU}) VAU]>}
This poem uses one of my modifications of the argument-label previous-list [I intensely want to know its (the list's) last form.]

Comment: Excellent. Hopefully, Michael will be at least temporarily satisfied by the cmavo list enclosed.


.i mi pilno <<lo'u sexebe'i le'u>> zo <<be'i>> vi le pemci noi se tcita <<lu le te pemci .e le se binxo vau li'u>>
i {mi CU <pilno [({<lo'u se xe be'i le'u> <zo be'i>} {vi <[le pemci KU] [noi ({se tcita} {<lu [({<le [te pemci] KU> e <le [se binxo] KU>} vau) FAhO] li'u> VAU}) KUhO]>}) VAU]>})
I use ??"sexebe'i" for "be'i" at the poem, which incidentally is labelled as "The Poet and the Thing Become".

Comment: Michael used *"mi pilno zo sexebe'i zo be'i ..." in his original. "zo" is a very picky single word quote - it truly wants a single Lojban word. "sexebe'i" is a 3-word compound, and the "zo" quoted only the first word of it. Lest you think me to be picky, Michael's result was actually grammatical, if nonsense, parsing as (I use "se" with transmission media "be'i"...), since "xebe'i" is a valid sumti tcita, and in fact was probably the one he wanted based on his English for the poem. A double conversion is meaningless for a modal, which can only specify one place the im-plied bridi relationship; using lexeme SE before a modal indicates that the indicated non-first place of the associated brivla describes the relationship of the attached sumti.

The cmavo list enclosed with this issue exhaustively interprets the modals for first two places associated with each lexeme BAI member, and gives the identified 'useful' ones that access at-least-third places (some of the 2nd place conversions seem redundant to the unmarked 1st-place modals, such as "du'i" and "sedu'i").

.i [*<<lo'u sa'a]

ko mi zasyspo
.ije mi ba vuzyvuzyxru
sexebe'i do
.i makfa
.i roroi ku
mi'o zukte ra

le'u>> sa'a vau sa'a

i ({lo'u sa'a ko mi zasyspo i je mi ba vuzyvuzyxru se xe be'i do i makfa i ro roi ku mi'o zukte ra le'u sa'a} vau sa'a)] FAhO>})

[" editorially inserted quote for ungrammatical text]

Temporarily-destroy me;
and I will yonder-yonder return
through the medium of you.
It's magic.
All the time,
we do it (with purpose)

[" editorially inserted end-quote and "vau" to make the non-grammatical quoted poem a valid 'incomplete sentence'.

Comment: Note that the parser makes no attempt to figure out what is inside the quotes, which is why "lo'u...le'u" exist. There were several minor problems with this succinct poem that totalled to an overall bad effect; one of the problems I had to correct to even legally put it in error quotes instead of non-Lojban quotes, which seemed unfair to Michael, who made a game attempt.

  1. Of course the erroneous sumti tcita made the text ungrammatical. This is actually the only grammar error in the whole poem, but warranted the error quotes.
  2. The lujvo in the second word is properly formed, but Michael's original word *"vuzvuzyxru" was not. The latter breaks up morphologically, with the CV at the beginning falling off in spoken form, resulting in *"vu zvuzyxru", which is based on the long-rafsi form for a hypothetical Lojban gismu *"zvuza" (or some other final vowel). The test described in the lujvo-making algorithm applies here: if you start off with a CVC rafsi, and the first consonant cluster is a permissible initial, the CV will fall off. You have to break up the permissible initial with a hyphen "y" to keep this from happening.
  3. The only other serious problem is "ra" which has no legal referent, since all of the preceding sumti are personal pronouns which cannot be anaphorized (oops! se ba'ivla). In any case, from Michael's English translation, he probably wants "lenu go'i" or some other member of GOhA, depending on what "it" is - the English is too ambiguous for me to guess.
    On the other hand, another possibility is "le go'i", an option only because he uses "It's magic" as the translation for the observative "makfa", which has no sumti. This suggests that the "it" in the last line is the same it that is ellipsized in the observative. This, if it was what Michael intended, would be an new and unforeseen use of Lojban's unique capability for grammatically unambiguous ellipsis, a truly Lojbanic poetic usage.
  4. Not quite as bad are some curious and possibly malglico word choices. However, it is possible that Michael actually intended what his word choices mean:
    "zasyspo" expands literally to "temporary-destroyer"; the underlying tanru suggests to me that the brivla relates to temporarily performing acts of destruction, as opposed to destroying with temporary effect, which is what the rest of the poem implies. Of course, both are possible and Michael may have intended the semantic ambiguity. But I don't like it 'enshrined' as a lujvo when I can't be sure what it means.
    "roroi" is a version of "always" that means literally "at all times" or constantly over the unspecified interval of time. The colloquial English "all-the-time" suggests one of the aspect tenses like "habitually", "regularly", and "typically", all found in lexeme TAhE, might have been what Michael intended.
  5. I suspect that Michael made the bad lujvo discussed in 2. above partially because it looks and sounds interesting. It makes me uncomfortable though, to make a lujvo solely for that purpose, when the lujvo-based bridi has no grammatical advantage (in terms of grouping or terminators) or Zipfean advantage (length) over the shorter and simpler "mi bavuvu xruti", which I believe is both grammatical and means what Michael intended, keeping the tenses in the tense place. I gave rafsi to the locators of lexeme VA to allow potential for relative lujvo such as "the here-sitter and "the there-sitter", say, for use by a teacher in a classroom, where situations might frequently arise for using such relative words.

I won't pose an alternative for Michael's poem. He needs to re-express what he meant. I ended up not doing more of his poetry in this issue (including a few more poems in the same letter) because a lot of his poems have word-choice questions of this sort that I can't quickly resolve. (This means, Michael, that I'd like you to go through your poems one more time, looking for the types of problems I've discussed here. Also, try back-translating your poems 'literally' without your English handy.) Michael's doing an excellent job of exercising the language though. I just wish he had a parser so he could see how some of his sentences are taken. Perhaps the parser outputs above will help him and others.

Michael's expressions are very complex, and sometimes I suspect, very English in phrasing. I wrote the instructions for using the machine grammar specifically with people like him in mind. Several of you, like Michael, want to express very complicated thoughts in Lojban right from the start. The only way I can see to accomplish this, without taking the extensive analysis that this letter required, is for people to exercise building up sentences from smaller, simpler pieces, as described in the instructions for the machine grammar, then seeing what they've built. If John Hodges is right about discipline and art, then this kind of disciplined grammar development may be far more useful to an artist than creative tanru and lujvo-making, at which Michael is clearly exceptional.

Now its your turn. I'd like to report next issue that we're pleasantly drowning in Lojban text from all 110+ level 3 students and a few of the rest of you.

co'o