# me lu ju'i lobypli li'u 2 moi

For a full list of ju'i lobypli publications, see ju'i lobypli.
Previous issue: me lu ju'i lobypli li'u 1 moi.
Next issue: me lu ju'i lobypli li'u 3 moi.

# Me la Uacintyn Loglytuan *    Number 2 - August 1986

(Notes on the title -

1. My use of me apparently causes this to translate to 'Be a Washington Loglan worker'. Which wasn't what I intended, but sounds better. I want to encourage you. But since this newsletter is quickly becoming national in scope, I expect future issues to drop the Uacintyn. After all, you needn't move here to be a Loglan worker.
2. I've also been told that my name is improper for another reason - having used me to turn la Uacintyn into a predicate, I need another me la to turn Loglytuan into a name predicate. Otherwise, I would have to write it as me la Uacintynloglytuan, which I won't inflict on anyone (compare la Iunaitedsteits, a name which I've improved upon with my borrowings proposal described below). The double predicate is also needed to make it clear that whatever is named by la Uacintyn modifies la Loglytuan. Alternatively, I could have left the second word as a complex: loglytua, which would have made my grammar correct, but left us with a sentence rather than a title. Oh well.
3. As noted below, the problems in the prims for languages will hopefully result in changes. So I expect next month's title to be me la Loglentuan. Which brings us very close to the old Loglan name for TL, but I'm not volunteering to edit that.)

### YES JIM, THERE IS A LOGLAN COMMUNITY

Furthermore, it is alive, and if not completely well, regaining its health quickly. We have people waiting to go to work, and progress should accelerate after jcb's return, since several questions need his clarification. I am setting a personal target of the end of next summer to complete those contributions I am involved with for GPA (Going Public Again - the acronym for republication of Loglan materials). The actual decision on GPA, of course, is JCB's. But I think the dictionary is the critical path, and with my accelerating progress, and lots of help from all you out there, we can make it.

### ARE YOU ALIVE?

I've had $95 in contributions so far, interestingly from 3 invitees – Bob Chassell, Jeff Prothero, and Mike Sargent. Much thanks to you three. They and all others who write or call will keep getting the full newsletter, and the back issues that they haven't gotten. No contribution is necessary, although all are welcome. Last month's issue cost about$145.00 in printing and postage, due to its size, plus my phone bills which are unrepeatable. This month's issue is shorter, but will go to more people, so I expect the cost to be about the same. Note to anyone reviewing things for me. Please try to get me comments by the weekend of 13-14 September.

### LOGFEST

The first step is going to be Logfest 3, to be held at my house from Friday 12 September to Sunday 14 September. All Loglanists are invited, especially if you bring a sleeping bag. I have bed space for perhaps as many as 6 guests, and plenty of additional space for sleeping bags. I am 2 blocks from the Washington Metro, and can therefore be reached from either Union Station (rail) or Washington National Airport, for a low cost, no driving (or rental car) weekend vacation. If you can only spare a day - fine, come anyway - this will be very informal. I am interested in anyone who can bring a PC/compatible or CP/M machine, since we will have all working versions of MACTEACH available for use (see below). Directions are given in the following pages. I can always be called for more details:

                            Bob LeChevalier
2904 Beau Lane
Fairfax VA 22031
(home) 703-385-0273  (work) 703-847-4465


FLASH - While I have no commitments, I have invited jcb, who may be back from Europe by then, to attend. A definite enhancement to any activities we may have.

### NEWS

On with a status update - much has happened in a month and a half. I'll cover things project by project. Then I'll summarize by highlighting the contributions that those of you who volunteered are making.

1. The newsletter and SIG formation. Well, most of you know about this newsletter by now, and I've summarized the distribution. So far, there has been one significant effort to organize a SIG as a result. Ken Dickey has been contacting the seven Portland area Loglanists. There are also stirrings in Boston, which has the most active SIG members in any metropolitan area. San Francisco, of course, has more invitees. Because of the response and phone and mail contacts since the last newsletter, my target areas have expanded to include: DC, Boston, England, Ontario (both Ottawa and Toronto), Philadelphia/Wilmington, San Francisco, Portland, Raleigh NC, St. Louis, and New York/ New Jersey. I also am contacting people in Houston, Chicago, and LA who I am hoping will volunteer to be center contacts. I also need a new volunteer for San Francisco, since Birrell Walsh says he doesn't have the time to coordinate activities.
2. MACTEACH. Glen Haydon is to be commended. He has completed his MACTEACH 1 program based on jcb's word list included in the last newsletter, and has delivered it to PC/Z-100 users. He is preparing to send CP/M and Apple versions out shortly. He deserves credit as the only person in the past couple of years to complete a Loglan project (I exclude jcb, whose activities are ongoing and therefore have transcended completion).

But this is only half the story. Glen's program has some problems. One is size. Since it is written in FORTH, the word list and program are in memory all the time, which enhances speed. But the word list has exceeded available space, and he has had to split the program/word list in half. A second problem is that there are few FORTH programmers who can build on Glen's efforts to create the needed series of MACTEACH versions. A final problem is that the program is more of an aid to learning than it is a teacher, although what he built is apparently what jcb asked him to build.

To solve these problems, I contacted Nora Tansky, who had a similar concept implemented on her TRS-80 several years ago. After her visit last week, we can now announce an improved MACTEACH 1. This version is written in TURBO-PASCAL, and has been tested on both PC and CP/M machines. It implements the full flash card algorithm described by jcb in TL1/5, and can therefore be truly called Computer-Aided Instruction. We await jcb's return to bless our version, and to supply a needed cassette with spoken Loglan to match the word lists. But meanwhile, we are hoping to have MACTEACH 2, which teaches affixes, working for the Logfest next month.

We have laid out a series of MACTEACH programs (described below), which Nora will be implementing over the next few months. These programs will build on each other, and eventually teach you the entire language. The major effort will be creating word-lists in the proper form, and of course, we are awaiting the final GPA grammar.

It is our intention, subject to jcb's approval, to supply free upgrades to the new version when it is hopefully deliverable with updated word list and cassette (October?). This is to repay those of you who invested in advance in MACTEACH without a completed product. Other pricing considerations are described below. Those who helped in user-testing will get free copies in a few weeks when I get time to make them.

Incidentally, as part of testing the program, I have worked through the entire word list. I had never managed to use the flash cards successfully, and my initial knowledge was probably less than 30% of the words. After one month of about an hour a day (with occasional skipped days), my accuracy is up to around 90% and increasing daily - personal testimony to the value of the improved MACTEACH as a teacher.

3. Notebook 3 (NB3). We of course are awaiting jcb's return to complete this. I have had the existing material reviewed by several people, and will be assembling a single comment set for jcb. The opinions are unanimous - jcb is working on an excellent summary of the language. We have lots of questions and comments, both technical and editorial, but these will only serve to improve what jcb hath wrought. The primary technical issues so far appear to relate to the phonetic values of Loglan vowels.
4. LYCES/LIP/MacGram. Our weak point. I finally reached Scott Layson, who should be sending me the latest LYCES. While only jcb knows for sure, it appears that grammar work stopped with Trial 24 in 1984 due to the limits of space in the old LYCES. This was during the time when the so-called 'Carter changes' were being tested. I have a copy of Trial 24 for any of you who know the MacGram work well enough to use it, and I should be able to give an accurate status of the grammar as soon as jcb returns. Jeff Taylor has expressed interest in helping with completing any YACCing, and in enhancing the Parser(LIP) to support larger blocks of text. Since he got his Master's degree doing Yacc-related work, we have identified a qualified worker.
5. Primer. Chuck Barton has sent me the first 3 lessons of his Primer revision. They date from before TL6, though and need to be greatly revised to make them current. Chuck did a lot of work, and I believe a good foundation has been laid to complete the Primer. I'd like to find someone who is a good writer to help Chuck. He shouldn't have to do it all, and he has the language training expertise to work with someone who can help in generating text. Perhaps another in the Boston area? As Chuck says, he writes concisely. Others without linguistic training can therefore be useful by helping identify areas where expansion is needed to clarify his writing for the novice.
6. Other computer aids. Nora Tansky has a Loglan speaking program. Art Wieners, a recovered 'lost Loglanist' who is conveniently a couple of hours from her, wants to build upon her work. We may eventually have a MACTEACH that tests you directly from computer-spoken Loglan. Jack Waugh is undertaking similar work with the Amiga speech generator, but has had problems because it was designed to support English phonemes which do not exactly match those of Loglan. Nora is also, on a lower priority basis, updating her sentence generators and translation program to reflect trial 24. Perhaps by the Logfest, but she is doing an awesome amount of MACTEACH work, and I hate to distract her.
7. Dictionary Update/Eaton Interface. Now for the biggies.

The two efforts are now effectively combined. Kieran has somewhat passed the baton to me on Eaton. He approves of the ideas I proposed last issue, but more than that, he does not have time to serve as an effective foreman now. Jeff Taylor has sent me copies of computerized Eaton material, but apparently jcb has the rest, so any attempt to build word lists or dictionary entries must wait.

Actually, my proposal is on hold for now. This is because there is a loose end in the complex-making algorithm. Coupled with this has been lagging software development. I cannot at this time stress the complex- making experiment described last month while still pursuing higher priority dictionary efforts. Perhaps in a few months.

But the Eaton work is proceeding as part of my dictionary work in two ways. As I go through the dictionary line by line, I am looking up the Eaton numbers for each E-trans line where one is applicable, and noting it in the text (line type 9 - see below). Secondly, I am using jcb's charter of making 'words of opportunity' as a fallout of the latter effort, to come up with complexes that express some of the fine distinctions between various Eaton words for the same or different concepts. I'm adding these as new word proposals, also with Eaton numbers.

The result is that when I have completed my effort, we will know just how good our Eaton coverage is, and what words need work. We could also easily produce an Eaton-correlated word list showing the words for each concept in Eaton order. Meanwhile, my total of Eaton numbers entered serves as a measure of my progress on the dictionary (only 1% Eaton coverage so far, but that is just due to lagging word-processing - I have a couple of dozen pages of unentered text/changes, including over 100 new word proposals).

I'm still awaiting volunteers for dictionary review. In hopes of convincing the masses that you can contribute, I am enclosing as an Appendix, the three short files that I've completed. Open season for kibbitzing!

Chuck Barton has volunteered to review the etymologies for c-Prims in the dictionary. I have extracted them (an easy effort using the new format), but am awaiting his further time availability. That's a big job. Anyone else want to help? You need to own or have access to a good dictionary for each of the 8 languages, and be familiar with phonetic values for each language's alphabet. Chuck wrote material in TL2 on this matter which omits only the 'h'. If you do not have those issues, we can supply copies of the data. Chuck, can you write a quick summary of the impact of 'h' on your phoneme sets.

8. jcb's trip. I've talked to Evie, jcb's wife, who returned last week from Europe. jcb has had a lively summer. The boat he bought had major problems and they spent a while in Venice waiting for repairs, then had to speed their progress thereafter to make it to their goal. Meanwhile, they were shot at by the Italian navy (nothing jcb did, apparently they entered an unmarked practice range and weren't seen). When Evie left, they had made it to Ibetha (near Majorca), where jcb lived for several years. jcb still had about an 8 day sail to Gibraltar, followed by a couple of weeks contracting for repairs. He should thus be back between Labor day and 1 October, hopefully with appropriate timing to join the Logfest. Meanwhile, Evie reports that he has lost weight, and recovered his strength completely from his operation. He is reported to be in high- spirits, relaxed and mellow. We can thus expect very positive outflows when he returns to work this fall.

### A Summary of Contributions

Rebecca Bach - editorial assistance on these newsletters. She also helped on element word remaking, as well as encouraging me to start this venture. Helped review MACTEACH (both new and old versions).

Chuck Barton - linguist; has worked on the Primer, phonetic values for alphabets, etymologies for C-prims, new C-prims, notes on dictionary problems, and is reviewing NB3 and most of my work. Serving as a SIG center in Boston (17 people).

Arthur Brown - DC area participant in SIG meeting, testing of the new MACTEACH.

James Cooke Brown (jcb) - gave us the reason to be working, and hosted me in June (along with his wife Evie). As a result, I've had the information I needed to proceed so ambitiously this summer. Helped me on the element words, supplied the word lists, and worked with me to firm up the complex-making algorithm. (And a lot more!)

Gary Burgess - Russian linguist in England for the US Air Force. Serving as SIG center there (8 people), saving me much postage. Gary introduced me to Loglan 6 years ago.

Kieran Carroll - currently inactive Eaton Interface foreman, and serving as SIG center for Ontario (11 people).

Bob Chassell - financial contribution

Ken Dickey - SIG center for Portland OR (7 people). He is the only other SIG center besides myself to actively contact and organize his group.

William Harrington - ex-Loglanist who responded to the first newsletter with a good letter, and by donating all of his TLs, his Primer, flash cards, Dictionary, and L1 to the cause. By doing so, we have a couple of spares of some of these out-of-print items.

Glen Haydon - completed the original MACTEACH, and delivered the PC version. CP/M and Apple versions to go out soon.

Scott Layson - LYCES/LIP foreman and the only other member (I think?) of the Loglan Academy besides Jim. He is hopefully sending me the latest version of those programs, and will be consulting on modifications that are needed to LIP.

Bob LeChevalier - you've heard enough about what I've been doing.

John and Anita Lees - supplied me with the address list that got the SIG going nationwide. Attempting to test the new MACTEACH on the Z-100.

John Parks-Clifford (pc) - former editor of TL, he has extensive files on proposed changes, as well as being one of the most knowledgeable people about the language besides jcb. He is supplying what he has, and answering tough questions about old TL issues. SIG center for St. Louis (5 people).

Ed and Julie Prentice - Ed volunteered to edit Lognet this summer, so I contacted him to give him information about our DC SIG. This led to my calling John Lees and Nora Tansky, and hence to the nationwide SIG. Ed is co-SIG center for Boston with Chuck Barton, and advising me on doing a Logfest. He is also reviewing stuff, and attempted to test an early version of the new MACTEACH, revealing some compatibility problems.

Jeff Prothero - financial contribution

Lawrence Proksch - jcb had him call me, which first gave me the idea that people outside the DC area might be interested in SIG activities.

Michael Sargent - financial contribution

Joel Shprentz - local FORTH expert; I originally asked him to help Glen finish MACTEACH. With the new MACTEACH under development, he has switched and is writing the complex-remaking program for the dictionary (and eventually the related programs for complexing). He also has supplied a capability to test the new MACTEACH on a CP/M machine. Participated in local SIG meeting.

Nora Tansky - one of the few who kept working the last few years. She has a Loglan speaking program, a translation and sentence generation program, and wrote the flash card program that was the basis for the new MACTEACH. She is now working on MACTEACH 2, while reviewing NB3 and dictionary stuff, and serving as Philadelphia/Wilmington center (8 people).

Jeff Taylor - helped on Eaton work and had kept reasonably current on the grammar, thus giving me good data in both areas. Jeff has volunteered to work on the LYCES/LIP work, and is reviewing NB3 and dictionary stuff. He also supplied missing pages form Eaton. (All those with copies from jcb are missing pages 8-9 and 190-191). Did a good job testing the new MACTEACH, submitting several suggestions which we took in full.

Birrell Walsh - temporary center for the San Francisco area (23 people). Birrell hasn't much time now due to work situation, but has offered to review material (including NB3) as time comes available. At one time, Birrell was one of the best Loglan writers - I'm hoping that jcb and SF area SIG members will help encourage him towards becoming more involved when the schedule improves. Birrell has also volunteered to write an article for the Whole Earth people to be published at GPA time.

Jack Waugh - local SIG member. reviewed NB3, participated in the SIG meeting, and helped test the new MACTEACH (as well as the old). He is trying to get some Loglan support capability on the Amiga.

Tommy Whitlock - center for Raleigh NC (4 people) and German/French/Hebrew linguist (among others).

Art Wiener - just contacted Art, and he has volunteered to be center for New York City and northern NJ (12 people).

All together, I estimate that about 800 hours have been put in on Loglan this summer since jcb left. This is almost what the Half-Time Graduate Assistant (HAGA) will do in a full year. Thanks to all of you for helping me show jcb what his Loglan supporters can do for him. And let's keep it up!

### More on LOGFEST 3

I'm taking off that Friday to allow people who so wish to arrive on Friday and be fresh on Saturday. (if anyone wants to come earlier and sight-see in the DC area, no problem - just let me know). I also will be going in late to work on Monday for people who want to leave on Monday morning.

Nora reports that the Amtrak ride (at least from Philadelphia) was very convenient, and she had no trouble getting to my Metro station.

Other than Nora, I have no committed attendees, but Chuck Barton, Kieran Carroll, Ed and Julie Prentice, Bob Chassell, and Art Wiener have all expressed interest. By the way, if it is necessary or desirable to bring (non-Loglanist) family along, I may be able to oblige with space, if not babysitting or entertainment. But sleeping bags will definitely be necessary in that case. (I have 2 guest rooms, and 4 entertaining rooms as well as a large back yard if someone wants to pitch a tent. I also have a tent and 2 bags, myself).

How to get here -

• by car
• (from North) take I-95 to the Capital Beltway I-495. Head west into Virginia, about 25 miles to I-66. Take I-66 westbound to the first exit - which is labelled Vienna, about 2 miles. Take the exit, turn left and go across the freeway. You are on Nutley St. Continue thru one light (for the Metro station), then turn left at the next block (Hermosa Dr.) It is 3 houses long and ends in my street. Turn left, and my house is the second driveway on the left.
• (from South) take I-95 to I-495, then it is about 7 miles to I- 66. Then follow directions as above.
• by train or bus - you will end up at Union Station (or right next door if by bus). There is a Metro Station there. It will cost $1.10 in non-rush hour to my station. Board and transfer to the westbound Orange Line at the Metro Center station (about 2 stations). Then a leisurely 25 minutes to the Vienna station, which is the end of the line and my stop. Exit the train, go upstairs to the south exit and telephone (or call from Union Station or Metro Center). It's about a 5 minute walk, or I can pick you up. • by plane - Washington National has a Metro stop. Board a Blue Line train heading north - about 5 stops to Rosslyn, where you transfer to the Orange line westbound as above. By rent-a-car get directions to either I-66 (more confusing) or I-395 (longer). From I-66 head west 13 miles to I-495, then as by car above. From I-395, about ten miles to I-495, then west as if you were coming from the south by car. • from Washington Dulles it is about a$17.00 cab fare, or I could possibly pick you up. It is about 25 minutes to drive. By rent-a-car, take the Dulles Access Road 16 miles to I-495, then south/west on I-495 5 miles to I-66, then follow directions above.
• I don't recommend coming into Baltimore Airport. It is 50 miles away. But it has an Amtrak station, and you could ride into Union station and continue from there. Amtrak schedules may not match well with flight schedules.

Activities - Nora has suggested an all Loglan conversation session. This will no doubt be difficult, as no one knows the vocabulary well, and there are uncertainties about the grammar. But we can try, dictionaries and word lists in hand. If I can get a microphone, we can tape the conversation, and go over it afterwards and comment/learn from our mistakes.

I will have both my Z-150 and CP/M machines for use with MACTEACH, and I hope a couple of others (primarily locals) will bring machines. I will try to have enough copies of Eaton, the latest grammar, NB3, LYCES and LIP, and whatever has been completed of the dictionary work for review. If Chuck Barton is there, he would like discussion of Loglan phonetics (just what is a Loglan o, etc.). Bring dictionaries, L1s, and TL4/3. I have several dictionaries, and 2 copies of most TLs, as well as a complete set of Lognets, but only 2 L1s and TL4/3s. This will be informal - we'll decide a detailed agenda when people arrive.

And feel free, if you are burned out on Loglan, to take a day to see the Smithsonian or other sights.

### A Proposed Algorithm for Generic Borrowings

There are many problems with borrowings - it isn't easily clear when making one that it is valid. (The algorithm is worse than the complex- making one, since the Slinkui test is more slippery to define, and there are some loose ends that lead me to be uncertain whether iglu type words are legal borrowings). Then there is the problem that the listener is unlikely to know the words at first, and there are few clues. And the dictionary/Loglan Academy will seldom have even a nearly complete list of borrowings once there are many speakers. Then, borrowing length is independent of Zipfean principles. And lastly, no valid scheme has been proposed for complex making with borrowed prims.

Last month, I printed my submission on borrowings for Lognet that was never published. In it, I suggested that what I called pseudo-complexes should be the preferred form for new borrowings, with possible shortening to shorter allowable forms when usage and Zipf demanded. I have elaborated this after some 'usage', and this entry describes the result.

When I visited jcb, we worked on remaking element words, and this planted the seed for the idea. The replacement of y for r/l/n as the C/C juncture hyphen provided the fertilizer. jcb told me that he had observed that a -CVCV ending and an impermissible initial consonant cluster in the first five characters seemed to invariably give a valid borrowing, even though that did not exhaust the set of possible borrowings.

The result - this algorithm, which is still stated loosely for discussion purposes.

1. Most borrowings fall into categories. They are jargon or specialty words for a field, or types of some category of thing that itself is a valid primitive or complex. The Loglan Academy will maintain a list of approved prim/complex categories for borrowings. Prim-based categories would have a -CVCV final borrowing suffix made by dropping one of the consonants (usually a middle one - but attempting to preserve similarity to the 3 letter affixes, and/or to keep a high recognition score). Complexes could have a -CVCV, a -CVCVCV, etc. depending on usefulness. I've proposed an initial set in an Appendix. For examples, see the nationality words above as prim categories. Sciences each would have a CVCV or CVCVCV ending with -se final, such as -lise for biology, -fise for physics, -tase for astronomy, -nuse for mathematics, and -likese for biochemistry. Titles would include -binami for military ranks/titles, -ganami for nobility titles, or -nami for a generic title.
2. Not necessarily as standardized at first would be a set of borrowing prefixes. In some cases prims or affixes could serve as the basis of these; if Zipf demanded a shorter or longer word for a concept, or if it would fit better into a category set with other borrowings, then it would be easily accomplished. An example is the use of basni in basnrfubo, in the last section.
3. The (hopefully past) existence of I-prims (see above) suggests a standardization method for formally devising prefixes, which can also take care of n-prims as well, while retaining the 8-languages flavor of Loglan prims. Normally people will have a tendency to invent borrowings as they need them, probably using as a basis for the prefix an international word if they know one, a Loglan prim or metaphor if applicable, or likely a word from their own language. At this stage, an auditor or reader easily can tell it is a borrowing, and the suffix tells him the general subject matter of the word. The remainder is possibly unknown, but may be deducible from the clues of context and category.

When a borrowing is submitted to the Academy for formalizing, however, the 8 languages technique will be used, attempting to maximize multilingual recognition scores. Note that more flexibility exists due to the variable length possible. For a word that qualifies for consideration as an n-prim, use the native language as a 9th (if necessary when not one of the 8 basic languages) with a weight of .50. (This would normalize to a 1/3 weight for a 9th language, but the normalized value would vary if one of the 8 were the native tongue increased to .50; I am suggesting that normalization not be bothered with; we want relative scores for a single word more than comparison with other words). The result might be that Swahili might dictate simbrnima (lion-animal), if it were not already a prim. An Innuit might propose iglrhasa for igloo, but the Academy might weight the construction by the 8 languages (which I'll represent by bisli, their composite) into bisgluhasa.

Of course, when the Academy decides that a word is to be formalized, if it is one which is frequently used and conceptually understood by the entire range of the Loglan community, it could utilize the untouched region of borrowing space to devise a shorter word, without suffixes. If it is used metaphorically, of course, it could be turned into a C-prim form, with an affix.

4. As I've implied in my examples, we have a handy way to attach the two parts together, and incidentally solve many problems of creating an impermissible initial and render the resulting occasional mish-mash of consonants pronounceable - our now free r/l/n vocalics inserted at a C/C juncture. They aren't always needed, but can serve as a handy reminder that one has a borrowing - the consonant masses in the middle look daunting at first, but as taste tests showed, are both pronounceable and understandable. And they occur nowhere else in Loglan except names.

Every complex suffix starts with a -C, and every prefix either ends with a vocalic, in which case no hyphen is needed, a consonant, in which case the vocalic serves as both buffer and hyphen, or a vowel, in which case one must (optionally?) choose between the recognizability a final vowel might give, and possible confusion of xCV- -CVCV with any possible x- -CVCVCV forms. I do not suggest formalization of rules at this point. I think usage will give an answer. I've used examples of each elsewhere in this newsletter for people to unscientifically taste test. Your opinions, please.

My personal preference is to use the vocalic hyphen both when required to give a consonant pair, to make otherwise illegal joints possible, and also whenever it may be useful as a buffer. But, as those who look closely at the examples in this newsletter will see, I try to preserve the recognizability of the prefix, often choosing an n or l hyphen, or even letting the end of the borrowing form merge with the suffix, if it seems to make sense. The way that borrowings are likely to be coined - on the fly in conversation - supports such informality, although it may be desirable after a few years experience to impose some more rigid rules.

5. A final note, not yet thoroughly tested. It seems that this approach gives a solution to the problem of how to form borrowings into complexes. I do not know just how complete the solution is:
1. It seems that you can always take the unreduced affix CVCCy and CCVCy and prefix into onto a pseudocomplex. The y, never found in a prim or a borrowing, tells the parser that the beginning is not yet complete. The parser then sucks up the borrowing whole; we already know that the borrowing doesn't fall apart. Similarly, though less desirable, one can replace the final vowel of a borrowing by y and swallow up a following prim or borrowing into a complex. There could be some morphological ambiguity in the latter, since the borrowing suffixes are not unique without final vowels, but the root is mostly intact.

An alternative is to require any -Cy- to be appended to link to a following prim/borrowing. This leaves the root undisturbed at the expense of an extra, if short, syllable. I suggest using the missing consonant from the preceding borrowing suffix as an arbitrary solution. That is what I did in the one example below.
2. CVC affixes also prefix well with a -y hyphen, although with vowel initial borrowings, some pronunciation problems may exist. These problem also apply to case a.
3. CCV affixes are uncertain. It looks like they prefix without a hyphen, but then it looks just like a longer borrowing, if one that might be interpretable from multiple roots. (See French-Canadian in the nationalities in the MACTEACH word-list article below). If a -Cy- is needed to join, one might as well use the unreduced form. CCV final seems ok with either rule from case a.
4. CVV/Cvv also join well as prefixes, but blend, using the r/l/n hyphen. Final, they work like all others, similarly to case a.

My personal taste is for the unreduced form with y as a prefix and -Cy- and whatever works as a suffix. The longer word forms shouldn't be too bad. These are, after all, complexes of borrowings; they are generally not heavily used in themselves or they would be prims. So a borrowing complex is inherently less likely to be a frequent concept.

I also like the idea of longer borrowings. It puts a Zipfean prejudice against jargon in conversation, and heightens to emphasis in Loglan on the prims and LWs. This will tend to separate Loglan further form English for English-speaking Loglanists, and will also tend to heighten Whorfian effects, since the impact of other languages through borrowing will tend to be selected against. English jargonizes by shortening, and results in efficient specialized talk among those who know the vocabulary, but poor communication skills with those outside of specialty groups. I know - I've made my mark professionally as a rare computer specialist who can write coherently - and most of you have tried to read either a computer manual or a legal document. The proposal biases away from jargon, and our tendency as primarily English speakers who work in specialized fields like computers and linguistics. And until we have a body of non-English speaking Loglanists, we will need to be constantly aware of the potential for English bias.

Examples abound, but I'll remake a couple of food prims to wrap this up. Banla, our old friend the banana, can be redone as:

banlrcidi      for generic types of banana food (mashed, chopped, sliced, dried into trail mix or the fruit after you peel it).

banlrticu      the tree
banlrfuta      the fruit
banlrlifa      the leaves


livbu, the apparently deleted prim (not in the MT1 list) becomes

livbrticu      olive tree
livbrcidi      olives, pitted, green, black, chopped, ground, or sliced
livbrsida      the pit
livbrfuta      the fruit, unpitted
livbrkolo      olive color


or to make a new (old) metaphor

norjrticu      orange tree
norjrfuta      orange fruit
norjrcidi      as you eat it
norjrsida      the seeds
norjrsapi      the peel


and the old joke about what the chicken said when a fruit was found in the nest becomes

Vizka le norjrsapinymarmlcidi, which loses something in the translation.

### More on MACTEACH

As mentioned above, the upgraded MACTEACH will have a cassette tape which may be reviewed with computer display to match. jcb plans to make such a tape upon his return. We will try to have such a tape to go with each version of MACTEACH, although we may not be able to reissue the tape each time the word list changes. MACTEACH will organize lessons and learning sessions for you, and will drill you on the words that you have trouble learning. It supports several users working on different word lists to proceed independently. When word lists need update, except for the problem of the cassette tape, a simple text editor can make changes, so we can keep you current without much difficulty.

Nora and I have laid out a series of successively more advanced MACTEACH programs which we have the know-how to design. Eventually, with the grammar, we will need jcb's help to refine the details, but we intend even that to use as input the latest trial grammar in a usable raw form. The following summarizes the current plan:

MT1       Primitives - Recognition/Recall, Loglan/English
MT2       Affixes - Recognition/Recall, Affix/Loglan and Affix/English
MT3       Complex Making – Recognition/Recall
a) Complex/component affixes
b) Complex/English metaphor
c) Complex/English keyword
MT4       Conversion/Abstraction (Argument Structure) – Primitives
a) nu fu ju
b) po pu zo
c) lo lopo lopu
MT5       Conversion/Abstraction – Complexes
MT6       Little Words
a) quantifiers
b) letterals
c) attitudes
d) auxiliary words
e) grammatical supports
f) connectives
g) compounds
h) phrases
MT7       Borrowings
a) Borrowing Affixes (subject to approval of my proposal below)
b) Recognition/Recall of significant words
MT8       Formal Argument Structures – Primitives
a) using pua, pue, pui, puo, puu to label positions
b) using case tags to label argument type
MT9       Formal Argument Structures – Complexes
MT10      Predicates - All concepts from L1 Chapter 3 tested collectively
MT11      Arguments - All concepts from L1 Chapter 4 tested collectively
MT12      Simple Utterances - From planned NB3 section and revised Teaching Corpus.
MT13      Parsing - You learn to do what LIP does instinctively.
MT14      Complex Utterances - From planned NB3 section and revised Teaching Corpus.
MT15      Translation - Randomly generated grammatical Loglan text will be generated which you can translate and check. Based on Nora's sentence generator, modified to give more meaningful sentences.


The above is of course subject to your comments. We're still only working on MT2. But we believe that when you finish the 15 levels of MACTEACH, you will know the language as well as it can be taught without having someone to practice with.

Pricing - I am suggesting the following as a pricing structure for the MACTEACH series. Do you think it is reasonable?

$50 for the initial program (old MT1 and new thrown in when available)$25/program up to MT9, then $35/program (grammar programming will be harder)$10/updated list or significant enhancement (except MT1 for those who preordered it - thus new orders will be $60 for both programs, dropping back to$50 after the new program is delivered.) \$5/minor update, basically to cover diskette and shipping costs.

If you use a machine for which MACTEACH is not supplied, I am recommending that you be able to buy another version, receiving it either via modem or diskette. You can then make any necessary changes in language, screen protocol, etc. and user's manual. Upon returning an updated copy to the Institute, you would get a 50% rebate, and receive a share of any royalties on sales for your machine. If more than 5 copies are sold, the remaining 50% would be rebated, as well.

### Problems with the MACTEACH 1 Word List

When I started learning the words using MACTEACH, I discovered, much to my chagrin, that several words I knew (from post-GMR lists) had been changed. I started checking the list against the master list I had confirmed with jcb in June, and found that the master list isn't master any more. I made a list of differences, and found that 90% of them could be easily explained by thinking about what jcb must have run into during the Scientific American Translation project. He apparently had never noted them in his master list. But there are a few places where mistakes in the MT1 list are likely. Also, as I started doing dictionary work, I noticed some new prims, some changes (especially to I-prims), and many changes to argument structures. I gave up analyzing the latter, the abbreviated structures given in the MT1 list are insufficient to compare accurately against the old dictionary. But there are either changes, or jcb made some mistakes.

When Nora visited last week, we went through each word as time permitted. My annotated list should be an Appendix. Meanwhile, use the MT1 list, it is better than any set of mark-up set of changes to the TL6 lists that I can easily give you, the changes are so numerous. Thus Appendix 3 of last month's issue is woefully incomplete (although I don't think any of those changes are wrong). Following are the major problems I've noted, categorized, and my proposals to fix them. Please note that I went through the full learning process with MT1 using the existing list prior to making these proposals, so you know I'm paying at least as high a relearning penalty as anyone - I've certainly avoided changes that are purely aesthetic.

1. Problems with the keywords vice clue words: In many cases jcb has given clue words in the English expansion that are not identical in meaning with the Loglan primitive, such as 'mortal' for morto. In many of these cases, including the latter, the clue word was the one used in the etymology to enhance recognition scores. However, in some cases, he has taken a good clue word, often with multiple meanings, as the keyword; this can lead to ambiguities or misinterpretations of meaning. Examples are 'crook' for gokru, 'stuff' for ctifu, harbor for harko, order for korji, post for posta. 'Pad' is used as the keyword for spadi and the clue word for padzi. The most common use of pad in English is closer to padzi in meaning. 'Pad' is a good clue word for both, but 'mattress' and 'cushion' are more clear for someone learning the primitives. Likewise, 'crush' is the clue word for kraco and the keyword for zakra; I have proposed 'pulverize (to bits)' to be the latter keyword, with 'crush' kept as a clue word. In our proposed changes, we added clue words for about 50 words based on the etymology, and in some cases changed the key word while keeping the old key word as a clue word. We also added the English metrical prefixes as clue words.
2. Many I-prims have been changed, in some cases without obvious unpacking necessity. Some were more recognizable as a result (hotle uses h), but others were made less so. For example, perna in L4 was changed to perli in NB2, and now to persa. Of the three, perli gives the best recognition, since it captures some Chinese. I've now done some dictionary work to support my analysis. As a result, I am now of the opinion that the I-prims should be justified with 8 language etymologies to be considered truly international (in which case they are indistinguishable from C-prims). Many I-prims do not appear to translate as cognates in the oriental languages, and Russian also tends to have a low score. I suspect that the recognition scores for most I-prims as they are currently made, would be no higher than blanu or matma.

In addition, many of the I-prims are not low Eaton numbers, so that it is even questionable whether they should be short prims rather than longer borrowings. jcb seems to have seen this in some cases; about 1/4 of the I-prims were left out of the MT1 list, including spera, pasli, kalko (which I also thought was calico rather than cocoa). On the other hand, konse was left out, and it is a useful word. To unpack it and improve its recognition score, I suggest konce. And londa ('x is fair in complexion/hair color') is difficult to make a metaphor for.

Barcu, ckafe, and kabre should also go away, among others. Based on the L4 definition, a bar is a tavern or taberna in Spanish and Wirthaus in German. But in French it is either cabaret (the source of kabre) or cafe (the source of ckafe). And I would apparently have trouble defining a cafe internationally; in the U.S. it usually is a small restaurant (cmalo resra), but often refers to an all-night diner/coffee shop; in Paris, the meaning is apparently much more specific and limited.

jcb apparently went to a great deal of trouble (see below) to make room for volta as a prim. It, and nearly all other measurement units, should be borrowings - they are seldom used in non-technical communication, and even more rarely occur in complexes. (I am proposing voltlekmeli as a replacement.)

3. Articles of clothing are another matter, especially since some are C- prims. Blusa(blouse), an I-prim, has a totally different meaning in the U.S. than in Europe. In the U.S., only women wear blouses. In other countries, and in the major English definition of the word, only men wear a shirt (a collared, button-down formal dress top for men), but in the 1980's unisex fashions have modified that definition, at least in this country. After discussion, Nora and I are proposing that all of the garment prims be retained, but rendered genderless in definition. This actually fits international usages of the concepts quite well, and covers most major forms of dress. The new definitions are:
resfu     generic garment prim, and root of any complex not covered by the other prims.

cadre     a full- length garment (top and bottom half combined); the best genderless English equivalent is 'robe'.

pantu     a lower-half (of a) garment with legs vice open-bottomed; pantu cadre would mean either a pant-dress
(usually worn by women), or a jumpsuit (worn by both sexes, with only minor differences).

skara     a lower-half (of a) garment which is skirted (open-bottomed) an English 'dress' would be skara cadre.

blusa     a fullness in the top-half (of a) garment (blousing in English) It can thus be used as blusa curta
(the American use of blouse), blusa cadre skara (a blouson, in current American fashion), or even

curta     a top-half (of a) garment, genderless. Kalra curta would refer to the formal shirt/blouse.

kapma     a garment for the top of the head.

cutci     low-topped leg-less outer footwear.

butpa     high-topped legged outer footwear.

hozda     stockings/socks, inner footwear with or without legs.


A suit is durna resfu, a gown is durna cadre, and an evening gown is durna cadre skara; a tie is durna nanda, with a bow tie being a kalra durna nanda; a brooch would lose its prim and become durna pinda, while a medal would be barda pinda or briga pinda, depending on its purpose; a belt or girdle (classic definition) is litnu resfu (the ladies undergarment 'girdle' is ninri litnu resfu, and brassiere is titfa resfu);

I think the above examples demonstrate that we have proposed truly primitive, useful, and culturally neutral concepts for each clothing prim.

4. There are serious problems with jcb's concept of nationality trios such as logla, logli, loglo for Loglan language, people, and culture, respectively. The problems are:
1. jcb has included the 8 source languages, Loglan, and also Italian, Scottish, Amerind, and Roman in the MT1 list. The reasons for including the latter four as prims are unclear. It could be because they have low Eaton numbers, but this is a case where Eaton is definitely obsolete, as international politics has evolved. In any case, there are far too many languages, cultures, and peoples of the world to use C-prim space for each. And in some cases, there are languages without distinctive peoples, peoples without distinctive languages, etc.
2. A more fatal flaw comes from the concept of these words as primitives. They cannot be made into unique complexes. For example, a student of the Japanese language, and a student of the Japanese culture would both be ponjystude.

By the way, this leads to the implication that the final vowel in Loglan prims is important only for recognizability and affix generation, something jcb (intuitively?) noticed when he arbitrarily assigned the -i ending when remaking gotso to godzi, because it gave the best affix set. This may have long term implications, as I can see the language evolving to attach meanings. Meanwhile, as long as you do not pronounce a schwa (which would make it a hyphen), a computer, if not a human, should be able to resolve the word, based on the first four letters.

3. Other related words, the monetary units, are ambiguous. It seems clear that dalra was intended to mean the American dollar. However, Kieran might justifiably disagree (and we have no word yet for Canadian). For some reason, Chinese and Japanese monetary units are not included - even though the yen is one of the four most important currencies today.

This problem has a ready solution, especially given my proposal for making borrowings described above. For the 8 root languages and Loglan, a prim meaning 'pertaining to the ... people, culture, or language' is needed. I also added a word for Hispanic, pertaining to Latin America, since the Spanish culture and Latino culture are quite different, in the same way that Americans and British are different, while still both English. I have selected the -i ending, except for Loglan which uses -a for traditional reasons associated with the language name. Each such word is defined with an affix for complex-making. This required a little remaking. We also need a C-prim for 'culture', certainly a primitive concept (I tried hard to come up with a clear metaphor without success). It turns out kultu gives a very high recognizability (49% or more), though there are no affixes available in that region. We thus have:

dotci     doc       Germanic, Teutonic
frasi     fra       French
gleci     gle       English (especially British)
hespi     hes       Spanish (from Espana, replaces span- which has no affixes)
hindi     hin       Hindu
hispi     his       Hispanic
jungi     jug       Chinese (jug is taken from jugra which uses it in only one complex)
logla     log       Loglandic
merki     mer       American (taken from merji, which can use mej instead)
nipni     nip       Japanese (replaces ponj- which has no convenient affixes)
ruski     ruk       Russian


To these, the affix -leu is added for lengu(language), -peu for pernu(people), or -kultu for culture. In addition -cme can be used for the primary monetary unit, and -gui for the country (Unfortunately, poldi has no final affix, though it might be preferable when associated with a culture to describe the country of a people).

Thus we end up with:

docleu    docpeu    dockultu  docrcme   docgui
fraleu    frapeu    frakultu  fracme    fragui
gleleu    glepeu    glekultu  glecme    glegui
hesleu    hespeu    heskultu  hesrcme   hesgui
hinleu    hinpeu    hinkultu  hincme    hingui
hisleu    hispeu    hiskultu  hisrcme   hisgui


(Latin America is lo hisgui; Hispanic-money is probably meaningless without lo)

jugleu    jugpeu    jugkultu  jugcme    jugrgui
logleu    logpeu    logkultu  logcme    logrgui
merleu    merpeu    merkultu  mercme    mergui
nipleu    niprpeu   nipkultu  nipcme    nipgui
rukleu    rukpeu    rukrkultu rukcme    rukrgui


For other countries, peoples, monetary units, the complex making proposal included elsewhere should be used. Some examples that probably need no translation since they are obvious when pronounced, if not seen (parenthetical words have limited use except possibly with lo):

italrlegu      italrpenu      italrkulu      italrmeni      italrguni
kandalegu      kandapenu      kandakulu      kandameni      kandaguni


(also kandafralegu, frasykandapenu - showing how borrowings of this form can be complexed, etc. - this, by the way, is an example where I chose kanda- over the hyphenated kandr- for recognizability. Any preference?)

skatrlegu      skatrpenu      skatrkulu      (skatrmeni)    skatrguni
dansklegu      danskrpenu     danskulu       danskrmeni     danskrguni
ijeptrlegu     ijeptrpenu     ijeptrkulu     ijeptrmeni     ijeptrguni
brazlegu       brazlpenu      brazlkulu      brazlmeni      brazlguni
mekhiklegu     mekhikpenu     mekhikulu      mekhikmeni     mekhikoguni
polsklegu      polskapenu     polskulu       polskameni     polskaguni


(or use polskr- to include an r hyphen)

vietnamlegu    vietnampenu    vietnamkulu    vietnameni     vietnamguni
sualhilegu     swalhipenu     swalhikulu     (swalhimeni)   (swalhiguni)
(or suarhil-)
inhuitlegu     inhuitpenu     inhuitkulu     (inhuitmeni)   (inhuitguni)

5. I have problems with the element words in the MT1 list. Since jcb assigned me to propose remade words, my response is detailed in an Appendix. In short, I believe an element word should not occupy C-prim space or be assigned affixes, unless it really is likely to be used metaphorically, in which case the primitive really should mean x-like, or x-pertaining, where x is an element which has a separate chemical borrowing as per the Appendix. Only carbon, iron, silver, gold, lead, and copper, have been used in a good metaphor in L4 (tin has been used in tinny, which has many different meanings in English - but usually means metallic; I can't find it in other languages' dictionaries). Chromium and radium and sulfur have obvious metaphorical use, and possibly helium, calcium, neon, phosphorus, sodium, hydrogen, mercury, oxygen, chlorine, and perhaps potassium, titanium, or aluminum have less obvious, but definable metaphorical properties, but have never been used as such. Zinc, nickel, and tin have few such uses, even though they were easy to make C-prim forms of. In any case, with chemical compounds written in acronym form most of the time, the only justification for a C-prim form is for ease in making metaphorical complexes, and that alone should be the basis for an inclusion decision. Furthermore, while many of the proposed prims could have affixes out of the remaining space, I personally feel they should not, unless a few metaphors of likely non- technical use are conceived. The affixes proposed in the Appendix can always be used in borrowings, but let's leave room in the prim and affix- space for growth in the non-technical language.
6. This next problem has slowed the complex-making. jcb and I never discussed how to automatically select between two CVC affixes for a word when two are available. There are a couple of dozen words with two CVC affixes; for most of these, one could be easily eliminated without causing more than one or two hyphenated complexes. And doing so would free up some affixes which would enhance affix space to allow coverage of the several prims which may have complexes when the Eaton work is done. In most cases, deleting a CVC has minimal effect on non-hyphenated coverage, and in a couple of cases (e.g. godzi - dzi), I was able to find a CCV that would relieve hyphenation pressure. But from my taste-testing, I don't find the hyphenated CVCyCxx form that bad, especially when the other choice requires learning a more complicated complex-making algorithm. I've made recommendations in the Appendix based on which affix deletions cause least hyphenation. It seldom is more than 1 or 2 words.

Meanwhile MT2 is being designed to test the current 2 CVC possibilities. Believe me, it will be much tougher to learn in those circumstances, as we've discovered just in trying to devise a good program design. The program will be simplified, of course, if dual CVC affixes are eliminated. I've deferred remaking the dictionary complexes until jcb gives me some direction. The complex-remaking program has been delayed, anyway, due to time conflicts.

7. There are a few apparent 'true errors', where it seems that jcb may have made a mistake or lost a change he had made to his master list. The most significant is the reversal of nervi and nurvo. The former used to be nerve, and was still that way as of the master list. Nervous had been remade to the latter. The MT1 list has them backwards, which leads to major losses in recognition scores. (Nervous actually gets a nice C-prim recognition score as nurvo). Since both have only CVC affixes, re- exchanging them should have little effect.

blako and marpi were omitted, as was dapli. (The latter overlaps ponda and retpi in meaning, and can be deleted, if the e-trans for the others are modified). Dertu and detra were changed on jcb's Master list to deu/der and det/dea respectively; they still have the old affixes in the MT1 list. dic was omitted from ditca, hea from helba, meo from metro, mou from mordu, and sno was snola in the summary tables in NB2, but has never been on any list. Surdi was changed to surla for no apparent reason - it sounds closer now to surna/surva than it did to sundi, so unpacking seems unlikely. And it destroys the symmetry and learnability of nordi/surdi and lesta/lusta. There are also the few inevitable typos that survive even multiple edit passes (rolgu, malbi, and a missing argument '..' in nutra, etc.) jcb also used American spellings in all key words, until I got to gray/grey. For MACTEACH, we obviously have to choose one spelling or another, but since the new MACTEACH allows manual editing of the raw word file, the British user can easily change back the few words that our barbaric dialect has modified.

8. There were several prim changes for unpacking, or possibly from the SA translation work. jcb also has added purpu, apparently the only proposed color word from TL1 that he accepted. Affixes were added for the metrical prims, as well as for ficli and nerji. Pei was moved from petci to perti, and foi was moved from fosli to fotli. A few prims were remade, apparently to get new affixes. In general, all of these changes are reasonable. A few cause recognition losses. My main complaint is that jcb hadn't updated his Master list, which means an extra manual editing pass over the whole dictionary for me to find all instances of each change - which is time-consuming. But such changes are to be expected.

The most drastic affix change was the shuffling/remaking of vXX affixes and their prims, with the only apparent effect of making room for volta. I am not in favor of this type of change, partly from the relearning (and editing) load. But recognition scores have been changed, and generally are lower, and the whole process lends an aura of instability to the language. In this one case, the effects are relatively small, and I don't propose undoing the changes. But in the future, I hope that such major changes are brought before the community, before we have to relearn them again.

9. There are many argument changes, some vague. The changes are numerous enough that questions could be raised (especially by critics of the language) as to whether our argument structures are arbitrary. I think this may currently be true, although jcb's L2 arguments indicate an underlying rationale existed at one time. The most common change has the addition of a 2nd argument 'from ...' to many former single-argument nouns, such as rozme, banla, and grunu. I'm not sure what the 'from' means, though, the original source location or a giver or a more recent source. And some such noun-prims did not add the argument, even though there was no apparent difference from the others that were changed.

Also, krani reversed the order of its from and to arguments. Cninu has a completely new meaning given. Kamla added a via.. argument, while the same L4 argument for kanla was omitted. Kutra reversed the y and w arguments. Kruli was also changed to a different concept since L4. In this case the concept is the more likely use of the word, but all metaphors need to be checked and a new one formed to deal with the old meaning (kruli mordu?). In any case, either jcb has made extensive changes which will have to be applied to the dictionary, or these changes were mistakes. For dictionary purposes, I am presuming the former. There is, of course, nothing in the MT1 list that says what jcb's intentions or rationale was in these changes, and he may have notes indicating all the dictionary changes that need to be made for each. I have therefore attempted to analyze the patterns in jcb's arguments and am proposing changes as I edit, where necessary to be consistent with those patterns. I am erring on the side of over-specification of arguments, since jcb seems to be adding arguments rather than deleting them in his changes. If an argument is not used, it will be dropped through non-use.

10. Nora and I are proposing the following additional changes based on our discussions:
1. Lasti should be changed to mean plastic, more common than it was in Eaton days. Elastic can become tetcu klabu or spori tetcu. Plastic has comparable recognition scores. Note the MT1 definition for spori - either it or elastic must be a complex of the other given that definition.
2. Other materials need primitives - clay and pottery are good examples. We didn't have good enough dictionaries handy to make a proposal.
3. Furniture as a concept needs a prim. Bed and table both are primitive concepts independent of their existence as furniture.
4. Stire is defined to mean stairs. I believe it should refer to a single step, which has more primitive application (tiers, stepstool, stepladder, and the front step(s) to one's porch. To make stairs, a prim is needed for ramp (step-ramp), which is also primitive - and raises the question of lever and pulley, the other two elementary machine types. All three have many metaphorical uses and are hard to describe as metaphors themselves.
5. Pundo and marli, and the new volta are examples of either non-metric or uncommon measurement units, and the first two are not useful in complexes (kilomile? micropound?). There simply are too many measurement units to occupy c-prim space with (and expect people to memorize). Measurement units beside time and metric distance should become complexes or borrowings based on merli or skalu. Volt would then become lenki fosli merli (lekfoimei), or as a borrowing pseudocomplex lekfoimeli, or keep the international root with voltlekmeli.
6. Four prims have definition problems. Bitsa (between) is defined to have a one-part form which is said to be 'beside'. However, between as a concept can involve more than two objects; e.g. among the trees. And even if one is between two trees, this does not mean that one is beside either. Or better, if one is located between Boston and Atlanta, is one beside either if the second location argument is left off. Nedza seems a better primitive for the concept 'beside'.

Balma is unclear in that it equates ...is a ball with 'spherical'. The earth is really an oblate spheroid. A football is not spherical, but is still a ball, and if I wad paper into a ball, it will certainly not be spherical. I suggest a redefinition as ...more spherical object composed of material...as compared to...(in degrees of symmetry) Since even a football has a high degree of symmetry, this definition fits all the primitive uses of ball I can think of.

Then there is larte, which seems to overlap between art and skill (see agriculturalist for an example of the latter). These seem to be two very different concepts. The latter could be process-know, but it could appear in so many complexes that a prim, preferably with affixes seems desirable.

Futbo is the last. It is defined as football/field hockey. Since this covers at least 3 or 4 sports (Canadian and American football vs. soccer, and field hockey has variants such as rugby), I have suggested that the word be the primitive for a outdoor/field sport (played on a defined area or field, and originally, though not always played outdoors). Various specific sports than can be defined as pseudocomplexes of -fubo, such as sokrfubo, merknfubo, rugbifubo, to more esoteric variants such as skizrfubo (competitive skiing), tensifubo (tennis), volrfubo (volleyball), and yes, basnrfubo (baseball).

7. Exchange affixes klu and kla for kutla and klabu, respectively. This eliminates the un-naturalness of the former.
8. Add leu for lengu, sni to sensi (which formerly was used by snire - which in turn gets sre and/or sie.
9. The most serious unpacking problem remaining is takna/takma/tokna. The former two are serious problems, since in a noisy battle environment, if da heard an opponent de shout T(a/o)k(m/n)a! vocatively, da's reaction would vary drastically, depending on da's interpretation, a potentially fatal misunderstanding. We came up with a couple of possible remakes which preserve or increase affix coverage. The best was changing takma to takci, which picks up some Chinese (kung chi - I don't know the phonetics well enough to tell how much). Takna can be remade to something like tanka, but again I don't know the phonetics well enough to evaluate it (it seems to me that both the English and Chinese should be 3/3 currently, which only heightens my uncertainty.) In any case, if recognition scores are high enough, this also gives tak to takci and tan for tanka.
10. Breko, as defined, should not be a prim; stop-device is a logical metaphor, but there is no prim for stop/halt with regards to motion. One could build metaphors like move-cease, move-cease-device, and move-cease- cause, for stop/halt, brake(n), and brake(v).

But I suggest a separate C-prim for stop/halt which is not exactly the same as cease. (Webster's shows cease as referring to immediate termination of individual da's activity; stop is not always immediate in its effect, and implies an action of its own, to actively terminate action of ba or even ra ba performing the action unless the scope is limited. Thus cease requests passivity, while stop requests activity.

11. Djale can be remade to rinje giving a significant recognition score increase, a more natural affix (rij), and reducing the pack of dj- words. (Br-, tr-, tc-, and st- are the other highly packed CC initials).
12. Dilko is usually given delicious as a meaning, and that is the key word in the MT1 list. But in looking over the metaphors it is used in, the aspect that is most important is subtleness or delicacy (a 'little bit'). I'm suggesting delicacy as the key word.
13. Pulso - Although the major languages all have this word, it is a metaphor in all of them. Impulse/impel comes from the roots im- and -pull and it means sudden-pull, as straightforward as any natural language metaphor. Delete this prim. It should be a complex, or rather two. For physics, it should be sudden-force. For most other usages (generally behavioral), it can remain sudden-pull, or even be sudden-cause.
14. Sente is proposed as a C-prim for holy/sacred as per Jeff Taylor's request printed in Loglan. I needed one, too. I get 4/8C shen sheng te 4/5E saint 2/2F saint 4/5S santo (no G or R and I don't know about H or J). The score is at least 48%.
15. Change tcidi to tcudi, losing Spanish but reducing confusion with titci and tcina, and giving it cud as an affix.

This sounds like a lot of changes, and an implied criticism of jcb's work. But Jim has done an awful lot of work in the past couple of years and I, for one don't expect perfection or all-accomplishment. It would give the rest of us nothing to do except learn Loglan. But seriously, as the rest of us start getting more involved, I hope this newsletter, and perhaps a revived Institute publication scheme, will cause ideas to flow again. It is only through our shared knowledge of the language and the reasons for all the little decisions that went into it (and the big ones as well), that GPA will be a positive experience. Each of us should have the knowledge to explain, and even defend, that which we've helped Jim do over the years, as well as that which he has done without us.

### Summary of Appendices

Issue #1

1. Comma-Delimited Format
2. rjl submission to Lognet
3. Affix and Etymology Changes Since NB2
4. Complex-Making Algorithm, Parts 2 and 3
5. MT1 List of Prims and Affixes

This Issue

1. Proposed Borrowing Suffixes
2. Dictionary Data for Review and Comment
3. Element Words
4. Proposed Revision to MT1 List (if done in time - I may mail it separately).

## Appendix 1

### Proposed Borrowing Suffixes

-meni     cmeni          monetary units
-dugi     dugri          other units
-meli     merli          scales of measurement
-timeli   tidjo merli    weight measurements
-kemeli   ckemo merli    time measurements

-fici     ficli          types of fish
-kopi     korti parti    lesser body parts
-mari     marpi          types of reptiles
-gamari   grada marpi    types of dinosaurs
-nima     nimla          types of animals/mammals
-nida     nirda          types of birds
-seka     sekta          types of insects
-senima   sekta nimla    types of crustaceans

-futa     fruta          types of fruit
-gudu     gusto donsu    types of seasonings
-niki     nikri          types of cheese
-hopi     hompi          types of drinks
-lihopi   likro hompi    types of alcoholic drinks
-vino     vinjo          types of wine
-cidi     tcidi          types of food (could become -cudi/tcudi

-minu     minku          types of minerals
-toku     troku          types of rocks

-fora     flora          types of flowers
-heba     herba          types of plants
-sida     sidza          types of seeds
-ticu     tricu          types of trees
-sapi     skapi          peel, skin, shell, outer covering

-kulu     kultu          cultures
-legu     lengu          languages
-penu     pernu          peoples

-nami     namci          titles
-binami   bilca namci    military titles
-ganami   ganta namci    nobility titles
-ginami   garni namci    government titles

-hasa     hasfa          types of housing
-teko     tekto          types of architecture
-kemi     kemdi          elements/chemicals
-kolo     kolro          colors
-refu     resfu          forms of dress
-sedi     sente denli    holidays

-petuka   petci turka    jobs/professions
-pituka   pleci turka    hobbies
-pesi     pleci sisto    games (primarily indoor)
-fubo     futbo          sports (primarily outdoor)

-maci     matci          types of machines
-pace     patce          types of devices
-tope     tomki patce    types of appliances
-mupe     muzgi patce    musical instruments
-time     trime          types of tools
-timi     trime matci    types of mechanized/power tools

-sesi     sensi          science
-fise     fizdi sensi    physics
-hese     herba sensi    botany
-huse     humni sensi    anthropology
-kese     kemdi sensi    chemistry
-likese   clivi kemdi sensi   biochemistry
-lise     clivi sensi    biology
-loji     lodji          logic
-mise     smina sensi    psychology
-nise     nimla sensi    zoology
-nuse     numcu sensi    mathematics
-pase     prase sensi    engineering
-pejo     penso djano    philosophy
-rose     rodja sensi    agronomy
-sose     socli sensi    sociology
-tase     tarci sensi    astronomy
-tese     tetri sensi    meteorology
-topese   tomki penso sensi   computer science
-tose     troku sensi    geology
-topo     tomki penso    computer terminology


## Appendix 2

### Dictionary Review Pages

The following are the files I have completed to date. Some minor notes on changes to the Universal format from Issue #1 Appendix 2 are also included.

line type 8 comments to be extracted prior to dictionary production. line type 9 Eaton numbers, a/b/c refers to the multiple terms for a concept. p means that the concept is represented only partially by the Loglan word. line type 2 I have added the English metaphor after a left-paren at the end.

I have replaced the ',' after the line type on many lines with codes indicating selection criteria or future actions/review needed:

n    new word proposal (any line)
m    metaphor change (generally line 2)
d    definition change (line 4)
e    E-trans change (line 5)


after a decision, the above will be changed back to ','

p    proposed alternate change (consensus or jcb will decide, so far hese are only my preferences)(any line)
x    delete this line as part of a change; retained for comparison


After a decision, the above will be deleted, though a comment may be left in their place.

!    action item for future work, often for a decision by jcb (line 8)
c    comment (line 8)

4CPX.1

bleklimadma     8! ensure hypothesis arguments consistent, also theory
cliransonma     1,4-Cpx\6.0\'75\\
cliransonma     4x(3v)\x rhymes y with w
cliransonma     4d(4v)\x rhymes y with w in language h
cliransonma     9,17421
cliransonma     4,(1v)\x makes rhymes/a rhyme
cliransonma     5,rhymer\(n)\- one who makes rhymes\cliransonma\
cliransonma     5erhyme\(vt)\- to make a rhyme\cliransonma\
cliransonma     5,rhyme\(vt)\- make...rhyme with...\cliransonma\
djadamsenma     8c the old metaphor is invalid - bottom in gravity field doesn't apply, basis is better
djadamsenma     8c but I then used the dictionary definition to derive the suggested metaphor
djadamsenma     8c in any case ensure consistency between metaphysics/metaphysician
djadamsenma     8! ensure consistency with metaphysics
fredirgotma     1,4-Cpx\2.2\'75\\
fredirgotma     4x(4v)\x makes y advance from w to h
fredirgotma     4d(6v)\x makes y advance from w to h in process q under conditions x1
fredirgotma     9,06704a
gunponkrisi     1,4-Cpx\\'75\\
gunponkrisi     2,gunti\ponsu\krido\sisto\(imperialist-creed
gunponkrisi     4x(1n)\x is an instance of imperialism
gunponkrisi     4d(2n)\x is an instance of imperialism as practiced/proposed/favored by y
gunponkrisi     5,imperialism\(n)\- a specific creed\gunponkrisi\
gunponkrisi     5,imperialistic\(a)\- pertaining to imperialisms\gunponkrisi\
gunponkrisi     5,imperialism\(n)\- mass term of creeds\lo gunponkrisi\
klimatsitma     1,4-Cpx\\'75\\
klimatsitma     4d(4v)\x theorizes y about w under conditions h
klimatsitma     8c to make consistent with hypothesize, they should be kept consistent if changed
klimatsitma     8! ensure consistency with  hypothesize, theory
klimatsitma     4,(1n)\x is a theoretician
klimatsitma     5,theoretician\(n)\- one who makes theories\klimatsitma\
nirpatretka     1,4-Cpx\\'75\\
nirpatretka     4d(2a)\x is more seasonable/typical of the current season than y
nirpatretka     8c clarification
nirpatretka     2,nirne\parti\retca\katli/season-distinctive/
nirpatretka     5,in season\(av)\- seasonable\nirpatretka\
nirpatretka     5,seasonable\(a)\- characteristic of a season\nirpatretka\
nirpatretka     5,seasonableness\(n)\- a specific property\ pu nirpatretka\
nodjanfirma     8! make dictionary entries for these
nodjanfirma     4d(4v)\x worries y by doing/being w under conditions h
nodjanfirma     8c5,alarm\(vt)\- worry...by...\nodjanfirma\
nodjanfirma     8c5,alarmed\(a)\- of one made anxious\nu nodjanfirma\
nodjanfirma     8c5,alarm\(n)\- a spec. state of being alarmed\po nu nodjanfirma\
nodjanfirma     8c5,alarming\(a)\- of alarming acts/people\fu nodjanfirma\
nodjanfirma     8c5,anxious by..., make...\(vt)\\nodjanfirma\
nodjanfirma     8c5,make...anxious by...\(vt)\\nodjanfirma\
nodjanfirma     5eworry\(vt)\- causes to be worried...by...\nodjanfirma\
nodjanfirma     5,worried\(a)\- of one made worried\nu nodjanfirma\
nodjanfirma     5,worry\(n)\- a state of having been made worried\po nu nodjanfirma\
nodjanfirma     5,worrying\(a)\- of worrying acts/people\fu nodjanfirma\
normutfujdo     1,4-Cpx\2.4\'75\\
normutfujdo     2xno\muvdo\futci\donsu\(fixed-future-give
normutfujdo     2,no\muvdo\futci\ckozu\(fixed-future-cause
normutfujdo     4x(3v)\x destines y to fate/destiny w
normutfujdo     4,(4v)\x destines/dooms y to fate/destiny w for reason h
normutfujdo     5,doom...to...\(vt)\\normutfujdo\
normutfujdo     5,destine...to...\(vt)\\normutfujdo\
normutfujdo     9,07609
normutfujdo     5,destined\(a)\- of one destined by some agent\nu normutfujdo\
normutfujdo     5,fated\(a)\- of one destined by some agent\nu normutfujdo\
normutfujdo     5,doomed\(a)\- of one destined by some agent\nu normutfujdo\
normutfujdo     5,destiny\(n)\- as given by some agent (converse)\fu normutfujdo\
normutfujdo     9,04311c
normutfujdo     5,fate\(n)\- as given by some agent (converse)\fu normutfujdo\
normutfujdo     9,04311a
normutfujdo     5,doom\(n)\- as given by some agent (converse)\fu normutfujdo\
normutfujdo     9,04311b
norsetfalcu     1,4-Cpx\6.1\'75\\
norsetfalcu     4x(2v)\x is consistent with/in saying y
norsetfalcu     8c what one is consistent with must be separate from what one says
norsetfalcu     8c also, both speech and actions can be inconsistent, possibly with each other
norsetfalcu     4d(4v)\x is consistent with y in saying/doing w under conditions h
norsetfalcu     8c the possible alternate is 'x is cons..with y. when said by.. under..
norsetfalcu     5econsistent\(a)\- with...in saying, of a speaker/actor\norsetfalcu\
norsetfalcu     9,17516
norsetfalcu     5econsistency\(n)\- a property of acting/speaking consistently\pu norsetfalcu\
norsetfalcu     5econsistency\(n)\- mass term of action/speech consistency properties\lo pu norsetfalcu\
norsetfalcu     5econsistent\(a)\- with...of a statement/action\fu norsetfalcu\
norsetfalcu     5econsistency\(n)\- a property of a statement/action\pu fu norsetfalcu\
norsetfalcu     5econsistency\(n)\- mass term of action/statement consistency properties\lo pu fu norsetfalcu\
ponkamkrisi     1,4-Cpx\\'75\\
ponkamkrisi     2xponsu\kumtu\krido\sisto\(communal-creed
ponkamkrisi     2mmunce\ponsu\krido\sisto\(communal-creed
ponkamkrisi     8c to be consistent with ponkamkrido in 3cpx
ponkamkrisi     8! ensure consistency maintained, do commune to match
ponkamkrisi     4,(1n)\is communism/an instance of communism
ponkamkrisi     5,communism\(n)\- an instance of communism\ponkamkrisi\
ponkamkrisi     5,communistic\(a)\- pertaining to communal creeds\ponkamkrisi\
ponkamkrisi     5,communism\(n)\- the mass of all communisms\lo ponkamkrisi\
ponlilfurle     1,4-Cpx\1.8\'75\\
ponlilfurle     2,ponsu\lilfa\fu\letci\(own-right
ponlilfurle     4x(2n)\x is a title/right of ownership to/in y
ponlilfurle     4,(3n)\x is a title/right of ownership to/in y under system/laws w
ponlilfurle     5,right of ownership\(n)\- title to...under...\ponlilfurle\
ponlilfurle     5,title\(n)\- right of ownership to...\ponlilfurle\
ponlilfurle     9,05512b
rodmadlardu     4d(4v)\x practices agriculture at location y using methods w under conditions h
rodmatsenma     1,4-Cpx\\'75\\
rodmatsenma     4,(1n)\x is an agronomist/agricultural scientist
rodmatsenma     5,agronomist\(n)\\rodmatsenma\
rorperkrisi     1,4-Cpx\6.9\jlt,'85\\
rorperkrisi     2xro\pernu\krido\sisto\(people-creed
rorperkrisi     2mpernu\garni\krido\sisto
rorperkrisi     8c consistency with roperkrido democrat
rorperkrisi     8! keep it consistent
rorperkrisi     4d(2n)\is an instance of democracy/a democratic creed as practiced/believed by y
rorperkrisi     5,democracy\(n)\- a specific system\rorperkrisi\
rorperkrisi     9,18912
rorperkrisi     5,democratic\(a)\- pertaining to systems\rorperkrisi\
rorperkrisi     5,democracy\(n)\- mass term of systems\lo rorperkrisi\
tordurcutpa     1x4-Cpx\2.3\'75\\
tordurcutpa     1,2-Cpx\2.3\'75\\
tordurcutpa     2xto\durzo\cutse\papre\(contract-paper
tordurcutpa     2mtogri\papre\(agreement-paper
tordurcutpa     8c shorter metaphor for common word, also often more than 2 involved
tordurcutpa     8! make tordurcu consistent
tordurcutpa     4x(3n)\x is a written contract y between parties w
tordurcutpa     4d(5n)\x is a written contract to do y under conditions w between parties h under legal system q
tordurcutpa     9,07206a
tordurcutpa     5econtract\(n)\- written, to do...under...made by...\tordurcutpa\
tordurcutpa     5ecovenant\(n)\- written, to do...under...made by...\tordurcutpa\
tordurcutpa     9,07206b
tordurcutpa     5ecompact\(n)\- written, to do...under...made by...\tordurcutpa\
tordurcutpa     9,07206d
tordurcutpa     5eparties\(n)\- legal, pertaining to contracts\ju tordurcutpa\
tordurcutpa     5eparty\(n)\- specific term, legal, pertaining to contracts\po ju tordurcutpa\
tordurcutpa     9,06303a
tordurcutpa     5eexpiration\(n)\- terminating condition of a contract\fu tordurcutpa\
tordurcutpa     9,13705

3CPX.1

djadamsensi     8c damni is the wrong kind of 'below'
djadamsensi     4,(1n)\is metaphysics/is a metaphysical law or principle
durfomporli     1,3-Cpx\2.3\'75\  %\
durfomporli     2,durzo\forma\porli\(manner-master
durfomporli     4x(2n)\x is the conductor of orchestra/performance y
durfomporli     4d(3n)\x is the conductor of orchestra y in performance w
durfomporli     9,07311b
durfomporli     5,conductor\(n)\- as of a band\durfomporli\
durfomporli     5,conduct\(vt)\- control the manner/style of...\durfomporli\
durfomporli     5,conducting\(n)\- an act of style-controlling\po durfomporli\
grozbumarmu     1,3-Cpx\1.7\'75\  %\
grozbumarmu     2,groda\zbuma\tarmu\(big-gun
grozbumarmu     4x(1n)\is a cannon/field-piece/a piece of artillery
grozbumarmu     4d(1n)\is a cannon/field-piece
grozbumarmu     9,05022b
grozbumarmu     8c artillery is a different concept in Eaton
grozbumarmu     5,cannon\(n)\- artillery piece\grozbumarmu\
daryreotau      1n3-Cpx\3.8\rjl,'86\ %\
daryreotau      2ndarli\renro\tarmu\(far-throw-weapon
daryreotau      4n(2n)\x is an artillery piece from event/period y
daryreotau      4n(1n)\x is an artillery piece
daryreotau      9n12224
daryreotau      5,artillery\(n)\- a spec. piece\daryreotau\
daryreotau      5,artillery\(n)\- mass term\lo daryreotau\
gunponkrido     1,3-Cpx\\'75\  %\
gunponkrido     2,gunti\ponsu\krido\(empire-believe
gunponkrido     4,(2a)\x is more imperialistic than y
gunponkrido     4,(1n)\x is an imperialist
gunponkrido     5,imperialist\(n)\\gunponkrido\
gunponkrido     5,imperialistic\(a)\- pert. to imperialists\gunponkrido\
kandurmatci     1x3-Cpx\\'75\  %\
kandurmatci     8c delete this, obsolete and non-Eaton
kandurmatci     2xkonte\durzo\matci\(counting-machine
kandurmatci     4x(1n)\x is a counter/a machine for counting y
kandurmatci     5xcounter\(n)\- a machine\kandurmatci\
norcenjalti     1,3-Cpx\\'75\  %\
norcenjalti     2,no\cenja\jalti\(constant-product
norcenjalti     4x(3a)\x is inversely proportional to y with product w
norcenjalti     4d(4a)\x is inversely proportional to y with product w for property h
norcenjalti     5,constant of inverse proportionality\(n)\\fu norcenjalti\
norcenjalti     5,inversely proportional to...\(a)\\norcenjalti\
norcenjalti     5,inversely, varies\(av)\- with fixed product...\norcenjalti\
norcenjalti     5,the less...the more...\(ph)\- with fixed product\norcenjalti\
norcenjalti     5,the more...the less...\(ph)\- with fixed product\norcenjalti\
norcenjalti     5,varies inversely with...\(v)\\norcenjalti\
ponkamkrido     1,3-Cpx\\'75\  %\
ponkamkrido     2xponsu\kumtu\krido\(communal-believe
ponkamkrido     2mmunce\ponsu\krido\(community-own-believe
ponkamkrido     8c if munce is not proper, kumtu\ponsu seems better than ponsu\kumtu
ponkamkrido     8! if it is ok must change commune,communism to match
ponkamkrido     4,(2v)\x believes in communism y
ponkamkrido     4,(1n)\x is a communist
ponkamkrido     5,communistic\(a)\- of persons\ponkamkrido\
ponkamkrido     5,communist\(n)\- one who believes in communism\ponkamkrido\
kumgotponkri    1n4-Cpx\6.1\rjl,'86\  %\
kumgotponkri    2nkumtu\gotri\ponsu\krido\(common-industry-own-believe
kumgotponkri    4n(2v)\x believes in socialism y
kumgotponkri    4n(1n)\x is a socialist/social democrat
kumgotponkri    9n17624
kumgotponkri    5nsocialistic\(a)\- of persons\kumgotponkri\
kumgotponkri    5nsocialist\(n)\- one who believes in socialism\kumgotponkri\
kumgotponkri    5nsocial democrat\(n)\- one who believes in social democracy\kumgotponkri\
kumgotponkrisio 1n5-Cpx\6.1\rjl,'86\  %\
kumgotponkrisio 2nkumtu\gotri\ponsu\krido\sisto\(common-industry-own-creed
kumgotponkrisio 8c alternately the international. metaphor socli\re\pernu\krido\sisto
kumgotponkrisio 8c which is social democracy
kumgotponkrisio 4n(1n)\x is socialism or social democracy/an instance of socialism or social democracy
kumgotponkrisio 9n17624
kumgotponkrisio 5nsocialism\(n)\- an instance of socialism\kumgotponkrisio\
kumgotponkrisio 5nsocialistic\(a)\- pertaining to socialist creeds\kumgotponkrisio\
kumgotponkrisio 5nsocialism\(n)\- the mass of all socialisms\lo kumgotponkrisio\
kumgotponkrisio 5nsocial democracy\(n)\- an instance of social democracy\kumgotponkrisio\
kumgotponkrisio 5nsocial democracy\(n)\- the mass of all social democracies\lo kumgotponkrisio\
rodmatsensi     1,3-Cpx\\'75\  %\
rodmatsensi     4,(1n)\x is an instance of agronomy
rodmatsensi     4,(1a)\x is agronomical
rodmatsensi     5,agronomy\(n)\- an instance\rodmatsensi\
rodmatsensi     5,agronomical\(a)\- pertaining to agronomy\rodmatsensi\
rodmatsensi     5,agronomy\(n)\- mass term\lo rodmatsensi\
rorperkrido     1x3-Cpx\5.0\'75\  %\
rorperkrido     1p4-Cpx\5.0\'75\  %\
rorperkrido     1m3-Cpx\5.0\jlt,'85\  %\
rorperkrido     2xro\pernu\krido\(people-believer
rorperkrido     2pre\pernu\garni\krido\(most-people-govern-believer
rorperkrido     2mpernu\garni\krido\(people-govern-believer
rorperkrido     4,(2a)\x is more democratic than y
rorperkrido     4,(1n)\x is a democrat
rorperkrido     9,15324
rorperkrido     5,democrat\(n)\\rorperkrido\
rorperkrido     5,democratic\(a)\- believes in peoples rule\rorperkrido\
rorperkrido     5edemocratic\(a)\- pertaining to democrats\rorperkrido\

LPRIM.1

lesta           1,L-prim\2.0\'75\\les
lesta           4d(2a)\x is further east than y
lesta           4d(2pp)\x is east/eastward of y
lesta           4d(1n)\x is east/in the east
lesta           5,eastern\(a)\- more east than...\lesta\
lesta           9,08803a
lesta           5,east of...\(pp)\- toward the east\lesta\
lesta           5eeasterly\(a)\- pertaining to (from/to/in) the east\lesta\
lesta           5eeast, ...in the\(pp)\- a place to the east of here\lesta\
lesta           5eeast\(n)\- a place to the east of here\lesta\
lesta           9,06129a
lesta           5,eastward\(av)\- in an easterly way\lesta\
lesta           9,06129c
lesta           6,norlesta\nordi\lesta\\
lesta           6,surlesta\surdi\lesta\\
logla           1,L-prim\\'75\\log
logla           8c verify affix, the log- series is the prototype for my proposals re culture-specific terminology
logla           8c I have also included the MCT1 prim-series for comparison labelled as p lines
logla           4p(1n)\x is an element/feature/expression of loglan
logla           5ploglan\(a)\- pertaining to the loglan language\logla\
logla           5ploglan\(n)\- an element of loglan\logla\
logla           5ploglan\(n)\- the mass of such elements\lo logla\
logla           4d(2a)\ x is loglandic in manner y
logla           5eloglandic\(a)\- pertains to the hypothetical culture\logla\
logleu          1n2-Cpx\\rjl,'86\\
logleu          3nlogla\lengu\\(loglandic-language
logleu          8c I have proposed leu as an affix for lengu - it is free and useful
logleu          4n(1n)\x is an element/feature/expression of loglan
logleu          5nloglan\(a)\- pertaining to the loglan language\logleu\
logleu          5nloglan\(n)\- an element of loglan\logleu\
logleu          5nloglan\(n)\- the mass of such elements\lo logleu\
logpeu          1n2-Cpx\\rjl,'86\\
logpeu          3nlogla\pernu\loglandic-person
logpeu          4n(1n)\x is a Loglandian
logpeu          5nloglandian\(n)\- a citizen of the hypothetical Loglandia\logpeu\
logpeu          5nloglandians\(n)\- the populace of Loglandia\lo logpeu\
logykultu       1n2-Cpx\\rjl,'86\\
logykultu       3nlogla\grupa\loglandic-group
logykultu       8c what is really needed is a c-prim for culture - it is not easily subject to metaphorizing
logykultu       4n(1n)\x is the Loglandic culture
logykultu       5nloglandic\(n)\- the hypothetical culture of Loglandia\logykultu\
logykultu       5nloglandia\(n)\- an element\instance of Loglandian culture\po logykultu\
logcme          1n2-Cpx\\rjl,'86\\
logcme          3nlogla\cmeni\loglandic-money
logcme          8c the basic monetary unit of any country would be built on a parallel metaphor
logcme          8c there would be no prim for pound, mark, dollar, etc.
logcme          8c the p-prims are ambiguous anyway - there are several countries with pound, dollar & mark units
logcme          4n(1n)\x is a Loglandian monetary unit, la Logcmer
logcme          5nla logcmer\(n)\- the monetary unit of the hypothetical Loglandia\logcme\
Logyguis, la    1nName\\rjl,'86\\
Logyguis, la    3nlogla\gunti\loglandic-country
Logyguis, la    4n(na)\x is Loglandia, the hypothetical country where Loglan is spoken culturally
Logyguis, la    5nloglandia\(na)\- the hypothetical country\la Logyguis\
logli           1pL-Prim\\'86\\
logli           4p(1n)\x is a Loglandian
logli           5ploglandian\(n)\- a citizen of the hypothetical Loglandia\logli\
logli           5ploglandians\(n)\- the populace of Loglandia\lo logli\
loglo           1pL-Prim\\'86\\
loglo           4p(1n)\x is the Loglandic culture
loglo           5ploglandic\(n)\- the hypothetical culture of Loglandia\loglo\
loglo           5ploglandia\(n)\- an element\instance of Loglandian culture\po loglo\
lunra           1,L-prim\\'75\\lun
lunra           4,(1a)\is lunar/moon, pertaining to the natural satellite of the earth, la Lun
lunra           5,lunar\(a)\- pertaining to the satellite of the earth\lunra\
lunra           5emoon\(a)\- pertaining to the satellite of the earth\lunra\
lunra           8c until NASA and sci. fic. 'moon' represented the Eaton concept e.g. moon men
lunra           9,04812
lusta           1,L-prim\1.9\'75\\lus lut
lusta           4d(2a)\x is further west than y
lusta           4d(2pp)\x is west/westward of y
lusta           4d(1n)\x is west/in the west
lusta           5,western\(a)\- more west than...\lusta\
lusta           9,07903
lusta           5,west of...\(pp)\- toward the west\lusta\
lusta           5ewesterly\(a)\- pertaining to (from/to/in) the west\lusta\
lusta           5ewest, ...in the\(pp)\- a place to the west of here\lusta\
lusta           5ewest\(n)\- a place to the west of here\lusta\
lusta           9,06019a
lusta           5,westward\(av)\- in an westerly way\lusta\
lusta           9,06019b
lusta           6,norlusta\nordi\lusta\\
lusta           6,surlusta\surdi\lusta\\
nordi           1,L-prim\1.6\'75\\nod
nordi           4d(2a)\x is further north than y
nordi           4d(2pp)\x is north/northward of y
nordi           4d(1n)\x is north/in the north
nordi           5,northern\(a)\- more north than...\nordi\
nordi           9,06301
nordi           5,north of...\(pp)\- toward the north\nordi\
nordi           5enortherly\(a)\- pertaining to (from/to/in) the north\nordi\
nordi           5enorth, ...in the\(pp)\- a place to the north of here\nordi\
nordi           5enorth\(n)\- a place to the north of here\nordi\
nordi           9,04813
nordi           5,northward\(av)\- in an northerly way\nordi\
nordi           6,norlesta \nordi\lesta\\
nordi           6,norlusta \nordi\lusta\\
snola           1,L-prim\5.7\'75\\sno
snola           8!verify whether sno still affix for snola as per nb2
snola           4d(4v)\x entails/infers/logically implies y under assumptions w by rules h.
snola           5,entail\(vt)\- imply...under...by logic...\snola\
snola           5,entailment\(n)\- a specific property\pu snola\
snola           5,entailment\(n)\- mass term\lo pu snola\
snola           5,imply\(vt)\- entail...under...by logic...\snola\
snola           5einfer\(vt)\- entail...under...by logic...\snola\
snola           9,16803
snola           5eassumptions\(n)\- the basis for an inference\fu snola\
snola           5eassumption\(n)\- a specific base for an inference\po fu snola\
snola           5esupposition\(n)\- a specific base for an inference\po fu snola\
snola           5erules\(n)\- a logical basis for reasoning\ju snola\
snola           5einference\(n)\- the product of logical reasoning\nu snola\
snola           5eimplication\(n)\- the product of logical reasoning\nu snola\
solra           1,L-prim\1.0\'75\\
solra           4,(1a)\x is solar/sun-, pertaining to earth's sun, la Sol.
solra           5,solar\(a)\- pertaining to earth's sun\solra\
solra           5,sun...\(a)\- pertaining to earth's sun\solra\
solra           9,02303
surdi           1,L-prim\2.3\'75\\sur
surdi           8c MCT1 list has surla, which is a change. but doesn't seem necessary, at least for affixes.
surdi           8c recommend keep old, for lesta/lusta nordi/surdi parallelism, since these are created L-prims.
surdi           8c in general, I oppose prim changes without good reason, since they make a lot of work for me.
surdi           8! check on surla/surdi since it is a prim
surdi           4d(2a)\x is further south than y
surdi           4d(2pp)\x is south/southward of y
surdi           4d(1n)\x is south/in the south
surdi           5,southern\(a)\- more south than...\surdi\
surdi           9,07425
surdi           5,south of...\(pp)\- toward the south\surdi\
surdi           5esoutherly\(a)\- pertaining to (from/to/in) the south\surdi\
surdi           5esouth, ...in the\(pp)\- a place to the south of here\surdi\
surdi           5esouth\(n)\- a place to the south of here\surdi\
surdi           9,07015a
surdi           5,southward\(av)\- in an southerly way\surdi\
surdi           9,07015c
surdi           6,surlesta \surdi\lesta\\
surdi           6,surlusta \surdi\lusta\\
terla           1,L-prim\3.1\'75\\
terla           4,(1a)\x is terrestrial/earthbound/earthly, pertaining to the planet Earth, la Ter
terla           9,00816ap
terla           5,earthbound\(a)\- terrestrial, of the earth\terla\
terla           5,terrestrial\(a)\- pertaining to the planet earth\terla\
terla           5eearthly\(a)\- pertaining to the planet earth\terla\
terla           9,10122
terla           5eearth...\(a)\- pertaining to the earth\terla\
terla           6,terke\terla\ckemo\\
terla           6,tersensi\terla\sensi\\
telyloi         1n2-Cpx\\rjl,'86\\
telyloi         2nterla\lokti\\(earth-confined to locality
telyloi         4n(2n)\x is earthbound for reason y
telyloi         5nearthbound\(n)\- confined to the earth\telyloi\
telyloi         8c the more common meaning of earthbound vice terla


## Appendix 3

### Element Words

This Appendix documents the proposed changes in Element Words to be made in the dictionary. Of the several forms listed, I prefer those listed in section 3., and those in column 5 of section 4. I will also be updating the names to the preferred listing in Section 4, and replacing the short form name currently in the dictionary by the Acronym word. Thus, at this point, I will not have reduced the total number of dictionary element words. Which are leat useful/desireable in the dictionary can be determined later, or lists can be added in dictionary appendices.

1. Element names shall most commonly be referred to as names. The name form (la ...) shall use the best aurally-similar morphology while:
1. ending with a consonant;
2. preserving visual similarity, especially for the initial letters and for those letters that are used in the symbol derivation;
3. for element words whose symbol is derived from another word than the International name, use the symbol-derivation word as the basis for the name.
2. Acronymic names will be used in discussions involving chemistry, where chemical names are used frequently. Acronyms will generally be used in naming compouns in such text.
3. All elements shall have an S-prim which refers to the element/chemical. As per my borrowing proposal, these will be pseudo- complexes consisting of an affix-like prefix, followed by the affix-like suffix -kemi which is clearly related to the affix for kemdi (chemical). The prefix shall be based on the chemical symbol, and will end in a consonant to ensure a consonant pair. If the consonant combines with 'k' to form a permissible initial, causing *slinkui problems, insert an 'r' as a pseudo-hyphen.

Column 1 is the current Loglan dictionary primitive for the element. Column 2 is the dictionary Name form for the element. If there is more than one possible name form, the one I think is better is listed first. Column 3 is the Acronym word, with symbol in parens. Column 4 is the primitive jcb, rb, and I came up with prior to jcb's departure to Europe - only part of the list was done. jcb expressed the goal of preserving the symbol near the front of the word foremost, retaining aural similarity to the International (English) name secondary, and visual similarity tertiary. When I returned to DC, I redid the list in an attempt to attempt to standardize names, while keeping consistent with jcb's philosophy. In this case a second possibility is shown in column 4. In typing this up after rereading my UL1 comments on borrowings, I came up with a new idea on how to do element prims. The results are shown in column 5.

Old Word  Name-Form      Acronym-Form   jcb-form       Pseudo-Complex
(S-Prim)                                (SB-Prim)      (LB-Prim)

barmi     la Barium      Baica(Ba)      barni          barkemi
barnumi


(The availability of y in names makes la Bariym possible, but the aural gain is less important than the visual recognition.)

berki     la Berkelium   Baikei(Bk)     berkeli        bekrkemi
bekerkeli
berli     la Berilium    Baice(Be)      berlili        berkemi
berlilumi
bismu     la Bismut      Baici(Bi)      bismu          bisrkemi
la Bismuq                     bimsumi
bormi     la Bromin      Bairei(Br)     bromini        brokemi
borno     la Boron       Bai(B)         borno          borkemi
bornomi


(C initial preserves the visual similarity to the symbol in element names, and is therefore preserved over K initial which keeps aural similarity but hides the symbol).

calfo     la Californium Caifei(Cf)     calfornumi     calfokemi
calsi     la Calcium     Caica(Ca)      calci          calkemi
carbo     la Carbon      Cai(C)         carbo          carkemi
cermi     la Cerium      Caice(Ce)      cermi          cerkemi
cermumi
cesmi     la Cesium      Caisei(Cs)     cesmi          cesrkemi
cesmiumi
cobla     la Cobalt      Caico(Co)      cobaltumi      cobrkemi
colri     la Clorin      Cailei(Cl)     clori          clorkemi
cormi     la Cromium     Cairei(Cr)     cromi          crokemi
cromhumi
cumri     la Curium      Caimei(Cm)     cmuri          cumrkemi
cmurumi
cupri     la Cuprium     Caicu(Cu)      cupri          cuprkemi
(copper)
dispo     la Dysprosium  Daicy(Dy)      dirsposumi     disrpokemi
la Disprosium


(y = schwa is now permitted in names and acronyms (only), so the Di- form, which preserves aural similarity but forces remake of the symbol, as per NB2, is less preferred.)

farni     la Francium    Fairei(Fr)     francumi       frakemi
ferno     la Ferum       Faice(Fe)      fernu          ferkemi
(iron)
fulro     la Fluorin     Failei(Fl)     fluorini       florkemi
la Florin


(I have to pronounce la Fluorin with a di-syllabic vowel, but the visual- similarity makes it preferred and a Loglandian tendency to try to penultimately stress unknown words would make the di-syllable aurally recognizeable, even though most people pronounce the International Name like la Florin.)

gadlo     la Gadolinium  Gaidei(Gd)     gadlinumi      gadkemi
galmi     la Galium      Gaica(Ga)      galmi          galkemi
galmumi
galru     la Aurum       Acu(Au)        aurmi          aurmukemi
(gold)                        aurmui
germa     la Germanium   Gaice(Ge)      germanumi      gerkemi


(As with hard c initial, the soft g initial must be changed to improve visual similarity. la Djyrmeiniym would be required to make it aurally similar.)

kirpo     la Krypton     Kairei(Kr)     kriptoni       krikemi
la Kripton


(The name choice is hard. The former name preserves visual recognition. But the word comes from the root krypto-. The latter must be remade in Loglan borrowings because y is not permitted, and the logical remake is kripto. Still, the roots are being lost in many element names - or Germanium would use the N-prim root for Germany and be la Dotcium.)

kotsa     la Kalium      Kai(K)         kalmi          kalkemi
(potassium)
lacti     la Actinium    Acei(Ac)       actnumi        actkemi
lafni     la Hafnium     Haifei(Hf)     hafnumi        hafkemi
lalmi     la Aluminium   Alei(Al)       almuni         almukemi
lamri     la Americium   Amei(Am)       amceriumi      amcrkemi
lanta     la Lantanum    Laica(La)      lantanumi      lankemi
la Lanqanum                   lantanhumi


(In names where visual recognition is important, the availability of q for /th/ in names seems superfluous. Many languages just use 't' for 'th' in such cases. This says nothing about the usefulness of q in Linnean binomials (the reason jcb added it), since the latter are usually not visually recognized, but rather associated with the Greek roots and are aurally recognized.)

largo     la Argon       Arei(Ar)       argoni         arkemi
larse     la Arsenik     Asei(As)       arsenumi       asrkemi
la Arsenic


(Several languages use the k ending to preserve the hard k sound.)

lasta     la Astatin     Atei(At)       atstatini      atsrkemi


(Difficult to come up with a jcb-form borrowing. 'ts' is a permissible initial.)

lelmi     la Helium      Haice(He)      helmiumi       helkemi
lerbi     la Erbium      Erei(Er)       erbiumi        erkemi
lerpu     la Europium    Ecu(Eu)        eurorpini      eurkemi
lidro     la Hidrojen    Hai(H)         hidro          hidkemi
la Hydrogen
la Hidrodjen


(It is possible to preserve the visual recognition using y, but the remade prim has had an affix assigned, and the first listed name form is fairly recognizeable if one presumes i/y and g/j will be common letter changes in Loglan names. I prefer 'j' to 'dj' in name conversion where visual recognition has any significance.)

lindo     la Indium      Inei(In)       indiumi        indrkemi
lirdi     la Iridium     Irei(Ir)       irdiumi        irdnkemi


(irdrkumi, with two 'r's close together is hard to say. Probably harder still for Japanese. This is why its nice to have 'n' and 'l' available as pseudo-hyphens in borrowings.)

litmi     la Litium      Laici(Li)      lithiumi       litkemi
la Liqium
la Litmium
loksi     la Oksijen     Oma(O)         oksigumi       oksrkemi
la Oxygen                     oksijumi       lokrkemi
la Oksidjen


(The visually similar la Oxygen is almost unpronouncable, and any primitive forms will have to use the 'ksi' combination, anyway. The pseudocomplex is given two possible ways. The first preserves the symbol, and is best. The second correlates to the lok affix listed above for the oxygen-like L-prim.)

lolmi     la Holmium     Haico(Ho)      holmiumi       holkemi
losmi     la Osmium      Osei(Os)       osrmumi        osmkemi
osmsumi


(Another tough one for deriving a jcb-type borrowing. There is nothing but the 'sm' consonant pair to work with. Note that this is a rare case where 'm' can be used in a pseudo-hyphen in a borrowing.)

lutme     la Lutetium    Laicu(Lu)      lumtetumi      lutkemi
magne     la Magnesium   Maigei(Mg)     magnesumi      magrkemi
manga     la Manganes    Mainei(Mn)     manganumi      mankemi
megru     la Hygrajirium Haigei(Hg)     higdrargumi    higrkemi
(hydrargyrum = mercury)       higdrarjumi


(Taking a little liberty with the International root, since it is seldom used in a recognizable form. The result clearly uses the symbol, and is relatively pronunciable.)

molbi     la Molybdenium Maico(Mo)      molbidumi      molkemi
natri     la Natrium     Naica(Na)      nadri          natrkemi
(sodium)


(jcb remade natri to nadri to unpack. Since no affix results, and the pseudo-complex works, this remake need not occur. Leave natri for the rare L-prim for sodium-like.)

nedmi     la Neodymium   Naidei(Nd)     nedmoumi       nedkemi
nedmeodumi
nenlo     la Neon        Naice(Ne)      nenloni        nenkemi
neptu     la Neptunium   Naipei(Np)     neptunumi      nepkemi
nikle     la Nickel      Naici(Ni)      nikceli        nicklkemi
la Nikel                                     niklkemi
nitro     la Nitrojen    Nai(N)         nitro          nitrkemi
nobmi     la Niobium     Naibei(Nb)     nibmoumi       nibkemi
pacti     la Protactinium Paica(Pa)     protactinumi   pratackemi
parse     la Praseodymium Pairei(Pr)    praseodumi     praseodkemi
patni     la Platinium   Pailei(Pl)     platinumi      patkemi
patlinumi
polno     la Polonium    Paico(Po)      polnonumi      polkemi
pomte     la Prometium   Paimei(Pm)     prometeumi     pomrkemi
posfo     la Posforus    Pai(P)         posforumi      posrkemi
pubmu     la Plumbium    Paibei(Pb)     pubmumi        pubkemi
pulto     la Plutonium   Paicu(Pu)      pultonumi      pulkemi
renmi     la Rehenium    Raice(Re)      renhiumi       renkemi
la Renium


(The 'h' in the name is silent, but significant for visual recognition.)

ribte     la Ybterbium   Ybei(Yb)       ibtertumi      ibtrkemi
ridno     la Iodin       Ima(I)         iodnini        iodkemi
ritri     la Ytrium      Yma(Y)         itmetrumi      itrkemi
rodmi     la Ryhodium    Raihei(Rh)     rodhiumi       rohodkemi


(Uses a schwa as a pseudo-buffer in a name. The typical Loglanist will filter it out, and it allows the early 'h' with no supporting consonant.)

rubdi     la Rubidium    Raibei(Rb)     rubdimumi      rubkemi
rurna     la Uranium     Uma(U)         urniumi        urkemi
rutne     la Rutenium    Raicu(Ru)      rutheumi       rutkemi
sacni     la Sycandium   Saicei(Sc)     sacnadiumi     sacnkemi
sagla     la Argentium   Agei(Ag)       argentumi      agrkemi
(silver)
samra     la Samarium    Saimei(Sm)     smaraumi       samkemi
selni     la Selenium    Saice(Se)      selneumi       selkemi
sibmo     la Stibium     Saibei(Sb)     sibtiumi       sibtrkemi
(antimony)                                   sibkemi
silko     la Silikon     Saici(Si)      silconi        silkemi
sinta     la Stanium     Sainei(Sn)     santhumi       stankemi
(tin)                         santanhumi
sorni     la Srontium    Sairei(Sr)     surtontumi     sorkemi
la Strontium                  surtrontumi
la Syrontium


('str' is a difficult triple, and the 't' interferes with the symbol. Y can be used as a pseudo-buffer; it improves the sound but cuts visual recognition.)

sulfa     la Sulfur      Sai(S)         sulfuri        sulkemi
talmi     la Talium      Tailei(Tl)     talhumi        talkemi
tanta     la Tantalium   Taica(Ta)      tantalumi      tankemi
tecne     la Tecnetium   Taicei(Tc)     techenetumi    tecnkemi
tecnetumi


(ch is listed as a legitimate permissible medial, but I'm not sure I believe it. Any opinions?)

telru     la Telurium    Taice(Te)      telrulumi      telkemi
terbi     la Terbium     Taibei(Tb)     tibterbumi     tibkemi


(Remembering that Ytterbium, Yttrium, and Terbium are all name after the same place, any International Name, or S-prim, should make them somewhat similar.)

titna     la Titanium    Taici(Ti)      titnanumi      titkemi
tormi     la Tyhorium    Taihei(Th)     torhiumi       torkemi
la Torium
la Qorium


(The preferred choice uses the schwa pseudo-buffer, which interferes visually, but sounds good and preserves the symbol. The other two options lose the h in the symbol.)

tulmi     la Tulmium     Taimei(Tm)     tumluhi        tumkemi
la Tyhulium                   tumhulumi
la Tulium
la Qulium


(The first name brings the 'm' in the symbol towards the front with as little as possible visual or aural change. The second is the closest possible to the International spelling. The third and fourth do not support either the symbol or visual recognition, but are aurally correct.)

tunse     la Wlfram      Wma(W)         uulframi       uulkemi
(wolfram = tungsten)          volfraumi      volkemi


(Visually recognizable, though the missing 'o' looks funny since we do not think of 'w' as a vowel. But try pronouncing it; it sounds surprisingly similar to Wolfram. For prims, I like the Vol for its visual and aural similarities to Wol. But uu is closer aurally, and the uu reminds us of the vowel 'w'.)

vanda     la Vanadium    Vai(V)         vandaumi       vankemi
zenlo     la Xenon       Xaice(Xe)      zenloni        zenkemi


(Not easy to pronounce, but no other way preserves the symbol, and it uses the new 'x'.)

zinko     la Zink        Zainei(Zn)     zinkumi        zinkemi
zirko     la Zirconium   Zairei(Zr)     zirconumi      zirkemi
-         la Einsteinium Esei(Es)       ensteinumi     esnteinkemi


(Watch the pronunciation.)

-         la Fermium     Faimei(Fm)     fermumi        fermikemi
-         la Mendelevium Maidei(Md)     mendelevumi    mednekemi
-         la Nobelium    Naico(No)      nobmelumi      nobkemi
-         la Larencium   Lairei(Lr)     larncumi       larkemi


### Lessons Learned about Pseudo-Complex Type Borrowings from the S-Prim Element-Word Remaking

The element words currently take a disproportionate amount of dictionary space, given their usage frequency, and not much time should be spent on them. It is hoped that the new dictionary has enough non- scientific words to remedy the unnatural scientific (vice literary) bent of the language, without having to delete such words. But they do illustrate the usefulness of my pseudo-complex method of borrowing, and incidently reveal some priciples that can be used generically in making borrowings.

Now that 'l' is never a hyphen, and 'r' and 'n' only occur following CVV/Cvv affixes in complexes, the use of the vocalic pseudo-hyphen following a consonant in a borrowing can be a general solution to *slinkui problems, which are due to the initial consonant pair being a permissible initial - hence srlinkui is a valid borrowing, and not too difficult to pronounce).

Similarly, the appendage of a vowel to a CVC affix (as in -kemi) can be used to make pseudo-complexes that preserve the Loglan affix-root, while never being confused with true complexes. CCV affixes can also be used in this way, but do not automatically prevent confusion with complexes (-CCVV/CCvv can break into -C'CVV/-C'Cvv, and the prefix might combine with the dangling -C so as to form a CVC-affix. There are many ways around this, but the check must be made, whereas CVCV can never be at the end of a true complex. CVV and Cvv affixes can be used in pseudo- complexes in this way without possibility of creating a true complex, but the appended vowel causes a very vowel-rich CVVV or CvvV, which is a rare ending form in the natural languages.