(From Lognet 96/2. Used with the permission of The Loglan Institute, Inc.)

Lo Lerci


Letters policy: Unless otherwise stated, letters addressed to logli in general, to The Institute, JCB, or any editor of Lognet will be considered as offered for publication. But it would be good if the writer explicitly offers. We reserve the right to edit letters, mostly just to drop material that has to do with ordering books, etc. If you are writing us by paper mail and your letter is a long one, we’d be grateful if you’d enclose a soft copy on a diskette. We can read most word-processors and having your letters on disk saves us a lot of typing.

The first letter is from Steven Lytle, a new member. It includes a biographical sketch, something I try to remember to ask all new members for.

Dear Mr. Brown:

I’m 40 years old, married, with three children (and a grandson). I was born and raised in central Indiana, but have also lived in Arizona and other places. As a child I was very interested in science and math, but in late high school I developed an enduring interest in languages. I joined the U.S. Army upon graduation from high school to become a Russian linguist. After a four-year tour and spending a year-and-a-half in Germany, I got out of the Army, got my degree in Russian, along with lots of language, math, and computer courses, and rejoined the army. I spent another 14 years in the army as a Russian linguist, in Germany, Ft. Hood, Texas, and Ft. Riley, Kansas.

I’m out of the army now, working as an accounts-payable clerk at a large local firm, and freelancing as a translator/proofreader (so far, all from German to English).

I became interested in Loglan in college, between 1977 and 1981, when I found a copy of L3 [sic] in the local library. [Probably you mean L1, as the original L3,  a “programmed textbook”—as books using operant conditioning were then called—never got beyond microfilm. The name Loglan 3 is now being reused as the principal title (there is also a subtitle, of course) of Steve Rice’s new textbook on Loglan, the first real textbook on the language ever to be written. (L1, of course, is not one.) The first seven lessons of the new L3 will soon be published as the second issue of La Logli. Watch for it.] The library also had L4&5. Then I read the Scientific American article. It was a most intriguing idea. I had been learning Esperanto for a few years altready, and Loglan seemed like another good plaything. But I couldn’t get it out of my head. Every now and then I would find my notes about Loglan again, and wonder what became of it.

Just a few months ago, I located and read a copy of The Troika Incident. That was really an eye-opening! Now I want to be actively involved in the language.


Steven W. Lytle

Well, this is one of the rarer routes to Loglan! “Back from the future”, it might be called. The Troika Incident, for those who don’t know it, was—is still!—my 1970, science-fictionish, utopian novel, published by Doubleday. It had some vogue around American university campuses in the early ‘70s, being used in a wide variety of courses. And Loglan (which is nothing more, but then nothing less, than a piece of utopian mental furniture) figures in it. T, as it is known around here, is out of print now. But a great many U.S. public libraries still shelve T. If you can’t find one at your friendly local, why I just might be able to find a loan copy for you, soi crano. (Utopian novels, being laid in the future—at least most of the modern ones are—do not soon go out of date! So you might still encounter some useful ideas in T.) 

The next letter is from Phil Driscoll. Phil is another new logli, and I had also asked him for “some biographical details”. He obliged us with this brief but fascinating note:

Dear Logli:

I’m a 40-year-old computer programmer. I’m a fluent Esperanto speaker with an interest in the structure of the various languages around the world and in the various writing systems used across the world. (I’ve even generated a couple of TrueType fonts for the Shaw alphabet for English!) I also collect machines to cast old-fashibed metal printing type. I have a Linotype and four machines to make type for it and [for] hand-setting. Thanks for the back issues of Lognet!

Phil Driscoll

The next letter is from Roy Bigelow, and in it he continues his running dialogue with The Institute. Other installments of that dialogue may be found in the Lo Lerci of LN 95/2 and earlier.

Dear Jim,

... Again, I read with great interest, and appreciate, your comments concerning my letter ... in Lognet 95/2. This letter is a continuation of the dialogue begun there. I agree with nearly everything, but...

Concerning the [English] word no: To answer your remark, I believe that it is used in a “non-logical” way when used to mean the opposite of yes. However, I appreciate your efforts to make sense of that nonsense sentence concerning semeiotics. It should have read:

‘I remembered that, in relationship to semeiotics, I (perhaps foolishly) employ the concept of “context-free” to distinguish between syntax and semantics ...’

To expound on the application of semeiotics to the Loglan language, it is my view that a goal to be achieved is a Loglan syntax that is entirely “context-free”, Loglan semantics that are entirely “context-dependent”, and Loglan pragmatics that, while generous, are obvious to the user. [As to syntax, the grammar of “machine Loglan” is context-free; that of “human Loglan” (the language that is translated into machine Loglan by the Preparser) is, of course, context-dependent in that there are a fair number of human lexemes—among them, and most spectacularly, NO—that must be “subscripted” as machine subspecies of that busy human word (e.g., NO1, NO2, NO3, etc.) before the Parser can uniquely parse the utterances in which it occurs. As to semantics, I like to think that Loglan semantics is not entirely context-dependent. In some respects yes, but in most respects no. That is, in L, too, reba ji nu purda liu rozme ga rozme. The pragmatics of a sentence—that is, the use to which it is being put—is of course always context-dependent...as it must be.—JCB]

Concerning the [Loglan word] ni [(zero)]: I do not agree that [English] null is another [word] for zero. My connotations (and my dictionary) tell me that null indicates that something is lacking a significant quality [sic], while zero, when considered as a number, is the opposite of infinity and,when considered as a digit, is the first digit in the sequence of digits. I note that null does not appear in mathematical expressions if one considers a mathematical expression to be the result of one or more mathematical operations and that they should be represented (as indeed they are) by different Loglan words. Also, I believe you misspoke in talking about “counts” in that you should have said ‘any unsigned integer can be a count’. [Indeed I should have!]

Concerning your Notes 3 and 4 about equivalence and identity: You may be right, but I’d be very careful in espousing views concerning mathematics which are at variance or in conflict with the arguments presented in the Principia Mathematica. Also, while there is no specifically mathematical symbol for equivalence, the symbols that do exist have a rather precise meaning when appearing in a mathematical context. I agree that to pio to bi fo provides a “correct expansion”, but if and only if: the word to is understood to be the third, and fo is understood to be the fifth, numeral in a sequence of numerals at least five numerals long; both words are understood to be integers; the word pio is understood to represent the mathematical operation of addition; and the word bi is understood to be, at least in arithmetic and algebraic ... contexts, a word meaning numerically identical to. [Do you mean by the word numeral in the above what is usually meant by digit? Your further use of it in the following paragraph suggests you do.] 

Note that if the sequence of numerals [sic] is three or four numerals [sic] in length, to pio to bi nenini or to pio to bi nene also provide “correct expansion[s]” and all three “correct expansions” can be said to be (and I am suggesting that they be considered to be) mathematically equivalent, and, if and only if properly qualified by the appropriate understandings, logically identical. A detail.

Thanks again for your attention and patience ... .


Roy V. Bigelow

The next letter is from Jeremy Dunn, an Old Member (Nurcnu Djori, not Laldo Djori, for I believe J is one of our younger members). Jeremy has some interesting thoughts about how we could use zi za zu in ways that could turn many predicate-pairs into scales, and how those usages might also facilitate subjunctive speech. 

Dear Logli,

It was in the course of reading about the subjunctive mood in Lognet 95/2 that I realized that that topic fits in with thinking that I had been doing related to other parts of the language. Rather than get into the subjunctive mood directly I will first give you my thoughts on the handling of antonyms in Loglan and how this got me to the subjunctive mood.

One aspect of Loglan that I never really became satisfied with is how it handles antonyms. Of course JCB did away with such distinctions when he smashed everything into predicates [Heavens! This sounds like L predicates were made on an  anvil! And all the while I’ve been thinking that they are the very soul of semantic delicacy, soi clafo!—JCB] but I think this is one area that needs some elucidation for poor linguistic amateurs like myself.

One of the features that I liked about Esperanto was its ability to create the antonym of a word in a simple fashion. Creating antonyms simply reduces the number of words in this region by a half, I would think that any so-called logical [sensible?] language would want to do this.

There are three possible systems for creating antonyms that we might employ. Let us use some [English words] for demonstrat[ional] purposes. Let us say that the term anti- produces the opposite meaning of a given word. If we have the antonyms of happy and sad we could set happy as the root and define sad as anti-happy, or we could do it the other way round and set sad as the root and define happy as anti-sad. 

[One trouble with this scheme, as all social psychologists who try to build attitude scales soon learn, is that very few of the many pairs of so-called antonyms in a language mark the ends of true scales for the humans who use them. That is, as humans become less happy they don’t necessarily become more sad, or vice versa; and so on, through a long list of such pseudo-scales (though some pairs, usually physical measurement words, do mark real scales: heavy-light and short-long are two that do). In a certain sense, each comparative word in a language invokes its own scale: kind-unkind; happy-unhappy; sad-“unsad”. That’s the work that nu does for us in L.] 

There is a third method we could use that I haven’t seen in any natural language but seems more fundamental and symmetrical to me. Suppose we create a word that combines the idea of “happiness/sadness” into a single concept; let us say that the word hapsad represents this construct. We could now represent happy and sad as anti-hapsad and hapsad-anti respectively where we understand that we are now using word order to designate which end of the spectrum we are talking about. Loglan does nothing along any of these lines, the words for happy and sad are gacpi and kec[ri] (Loglan 4&5) [the first is now hapci, since the addition of /h/], which bear no visible relationship [to] one another although they are clearly related as antonyms.

In addition to the antonym issue the same types of words are related in terms of being of different degree of intensity along either end of the scale, as in the series happy, elated and ecstatic or the series miffed, angry and livid. All of these word series are simply different degrees of a more fundamental concept. Loglan does the same thing in creating a series of words that are visually and verbally unrelated.

How then to get differences of degree while making antonymic relationships apparent? We need a series of intensity markers to get the differences of degree, I suggest expanding the scope of the usage of zi, za and zu from temporal and spatial degree to [mark the] degree of intensity of any [predicate].

Now let us create a concrete example of what I am suggesting. Let us create the word [hacri] from the [hapci]/kec[ri] pair to represent all the degrees of happiness/sadness combined generally. The series happy, elated, ecstatic would be zi [hacri] , za [hacri] and zu [hacri] . Naturally the antonymic counterpart to this series would be [hacri] zi, [hacri] za and [hacri] zu. We need another marker X that simply means that one is on one side of the fence or the other without specifying a particular degree. Thus X [hacri] would mean happiness in general, [hacri] X would mean sadness of some degree in general. One would also have the option of using [hacri] by itself to indicate that one is neither happy nor sad about something; even in Loglan this would be normally be more complexly expressed.

A quick perusal of the Loglan dictionary shows me that Loglan is becoming almost as filled as English with dozens of words that are really antonyms or merely differences of degree. This is wasteful and unnecessary. By using my method we can get rid of at least six to eight words for every combin[ed] concept that we create. Surely this would further clarify our thoughts about things.

There is one more useful distinction that we might make about antonyms. There are two types of antonym, which I call absolute antonyms and relative antonyms. What I call absolute antonyms are true opposites such as love/hate or happy/sad. Relative antonyms are pairs of concepts that we think of as opposites but that in a strict sense are not. The pair smart/stupid are not true opposites because stupidity is not the negative of brain capacity but the negative region of a non-zero dividing line that we call average intelligence. Similarly the pair hot/cold have the same type of relation. To make this relation slightly [more] observable in the language, I suggest that the words that are coined for absolute antonyms use the CVCCV form and those for the relative antonyms use the CCVCV form.

So much for antonyms; let us consider some possible side effects of expanding the scope of zi, za, and zu. We might use these intensity markers to strengthen or weaken the association between the predicates on either side of a logical connection, and this is where we can solve some of our subjunctive mood problems! Consider the example of A meteorite impact might have led to the extinction of the dinosaurs. Let us separate this statement into two statements that are logically connected such as If there was an ancient meteorite impact then this caused the extinction of the dinosaurs. You rightly point out that these two statements do not [embody] the same logical relationship, but let us rewrite it in the form A za noa zi B where the za is indicating that we are attaching a moderate degree of truth to the A statement and a lesser degree to the B statement. I adopt the convention that if there is no intensity marker on one side of a logical connective, then the statement on that side is assumed to be a virtual certainty. This usage of the intensity markers gives us fuzzy logical relationships which many subjunctive mood statements can be reduced to. 

Another example was The blood test was inconclusive, so he could be the father of the child. We could rewrite this as His blood test was positive and he is the father of the child. This would have the form A zi e zi B. We use zi on the left side because the test was inconclusive and therefore the statement of the blood test being positive has a low truth value; similarly there is a low truth value [for his] being the father  ... simply [because his being male] makes [it] a low but positive possibility [that he is the] father.

There is another sense in which we might use these intensity markers with a logical connective. Suppose a woman’s child skipped school and didn’t mow the lawn like he was supposed to. The next day at breakfast the mother tells him You will go to school today and mow the lawn! Now this use of and is not simply an abstract relation of truth values because the mother vocally is emphasizing that this and is a strong and and better be paid attention to. In Loglan she might use the form zu e zu to mark this kind of subtlety.

What of fantastic relationships such as If I were president of the United States, then I would have Ming the Merciless bomb Canada? We need another little word Y that would act as a fantasy marker to indicate a predicate without factual basis, our previous statement would then logically connect as zi noa Y. We [use] zi on the left side because there is at least a low probablility that I might become president but use Y on the other side because Ming the Merciless is a fictional character; at least we’ll hope so!

Of course if I was delusional and actually believed what I was saying, then noa would be my expected choice of words. Maybe if we taught everyone this kind of Loglan, we could tell when they were going nuts.

This approach may not handle all cases of the subjunctive mood but surely has applications for expanded uses of logical connections, improved antonyms and improved synonyms. As always I am interested to hear what the founding fathers have to say on these issues.

             Jeremy Dunn

This “founding father” thanks you for your contribution to our discussion of the subjunctive. The Keugru will certainly take your ideas into account in making its decision. A suggestion: quite apart from the possible use of zi/za/zu in marking subjunctives, you have made quite a number of proposals for changing the language in this letter. Why don’t you disentangle them from one another and submit them to the Keugru? Either one at a time or as a set of related changes?

The next letter is one made long by my extended answers to its questions. It is from Bill Rambo, another Nurcnu Djori, who had assumed, first, that it was our work on MacGram (the machine grammar project) that precipitated the 1987 schism with the Lojban folk. It didn’t; and in telling Bill what did—and also what the “more powerful disambiguation method” alluded to in my paper with Bill Greenhood is all about—it occurred to me that my ordinarily private answers to such questions might deserve publishing again, as many Cninu Djori (New Members) have come aboard since the last time—it was around 1992, I think—that either the MacGram Story or the Schism Story has been told. Naturally, people are very curious about both these dramatic episodes in our history, the one a Loglan-enhancing event, the other, a Loglan-diminishing one.

Dear Dr. Brown:

... [Thank you for] your and Greenhood’s 1991 article “Paternity, Jokes, and Song: A Possible Evolutionary Scenario for the Origin of Language and Mind” ... Thank you [too] for the names and addresses relating to Lojban ... 

I’m afraid I protrayed myself as more knowledgeable in the language and logic area than I am. I have only a layman’s knowledge of either and I’m afraid I wouldn’t know “a more powerful disambiguation method” if it came up and tapped me on the shoulder. Actually, what I supposed [about the schism with the Lojban people] was that either a part of the Loglan commmunity had become dismayed at the grammatical changes that were taking place and had rebelled against the fully formal [disambiguation] process, or else the reverse: that they wanted to take more time and work up a more “user friendly” implementation of the disambiguation process. 

[Member Rambo made a reasonable guess; but as I had explained in an earlier letter to him, his guess about the origin of the Lojban schism was wide of the mark. The 1987 schism that led to Lojban had nothing to do with grammar or disambiguation. The work that finally led to the syntactical disambiguation of Loglan—the 1977-82 project that we called “MacGram”—had been completed five years earlier. To the best of my knowledge that grammatical work had been happily received throughout the Loglan community. At least no voices were ever publicly raised against the remarkably few changes in L’s grammar rules—all confined to what might be called its “punctuation system”—that turned out to be necessary in this new, syntactically unambiguous dialect of the language.

As an aside we can now say that 1975 Loglan was already remarkably unambiguous. We know that to be true because the 1977-82 grammar work, while exceedingly subtle and productive, had no massive effects on pre-MacGram L. Apparently the original, heuristic work on disambiguation—the computer work undertaken by me at the University of Florida in 1962-64 by which syntactic ambiguities in the-then embryonic language were located one by one and removed—had to a remarkable extent succeeded. So when, in 1975, the Aho-Johnson-Ullman constructive-proof algorithm for identifying parsing conflicts in programming languages became available—thus enabling the formal removal of all such syntactic ambiguity from such languages—we were ready for it. Our adaptation of their methods for our own, speakable, human language rapidly succeeded...or at least it seemed pretty rapid to me! (I confess that there were other, less patient workers who not only found our progress slow, but even came up with “proofs”, from time to time, that our goal was “unreachable”! Ah, the seductions of mathematics.)

Still, we reached it. The first conflict-free grammar of L became available in February 1982; and, though L has grown hugely in richness and completeness since that time, L grammar has remained continuously in that conflict-free state. (Maintaining our grammar in a conflict-free condition is one of the many things that our present Grammarian, Dr. Robert McIvor, does for us today.)

Please forgive me, Bill, for this long historical aside. By the way, the “CACM paper”, the paper that was originally intended for the Communications of the ACM but never made it, is about to be published by Kirk Sattley and Alex Leith in the first issue of our longer journal, La Logli. It describes in full technical detail how we built that first, conflict-free grammar of Loglan. It also describes the software tools we used to do that with and with which we still maintain it.

What did produce the 1987 schism, then? I’m afraid it was a rather scruffily commercial issue that the LeChevaliers raised with The Institute in 1986, and then used our unwillingness to yield to them on as a reason for breaking away from The Institute in 1987. The question they raised with us was this: Was The Institute to retain exclusive commercial control over its publications, through its trademarks and copyrights, and so maintain the uniformity of a single, albeit rapidly growing language? Or was Loglan and its documentation to be put into the public domain, where the LeChevaliers and other commercially interested persons like them could freely use any parts of it to develop both software products of their own and the language itself in any direction that they chose?

My natural fear was that, under the freewheeling, commercial circumstances that the LeChevaliers envisaged for themselves, and which they wanted us to create for them and others by abandoning our trademarks and copyrights—indeed, even our mailing list was to be given away to them!—as many different “loglans” would be likely to develop as there were partisans of differing points of view about it. About three-quarters of our then-membership supported the Board and me in resisting the “balkanization” of our language in this way. But, sadly, about a quarter of our people did follow Bob and Nora LeChevalier out into this undiscovered country where the bell still tolls.

There, they adopted the role of the commercial competitor. They not only sued The Institute for the cancellation of its trademark—the unique applicability of the word ‘Loglan’ to all The Institute’s products—but they actually succeeded in that project after about three years of costly litigation. (Costly to them, that is, for, because of the pro bono efforts of our attorney and now-president, Wes Parsons, their suit cost us nothing but time.) But now, because of the success of that suit, anyone who wishes to call anything at all a “loglan” may do so...and also sell it as such!

But, more to the point, the lojbi—my word for the inhabitants of that other imaginary country—have evidently developed a slightly different but still loglanoid language, using virtually the same morphology and a grammar that is very similar to our own—but an entirely different vocabulary, it should be noted, for we still retain our copyrights!—which they still call “Lojban”. Of course! To do otherwise would be like calling all the children in a family by the same name! And they manage to keep Lojban from being balkanized by—guess what?—virtually the same sort of mild copyright protection that we have been using for years!

That ironic story is, in a nutshell, the history of the Loglan-Lojban schism. (But it is odd how history repeats itself. The history of the Esperanto-Ido split is very similar...and a similar record of human folly.)]

... Thanks for translating the entire [disambiguation] riddle for me. It is a help to see something I puzzled over done right. You know, I think what Loglan needs more than a primer to precede L1, is an elementary reader to accompany it: a “Dick and Jane” sort of thing that would let the student see how Loglan should be used in a wide variety of everyday language situations. The point I wished to make with the riddle pun, or, better, wished to celebrate, was that Loglan makes the basic error on which the riddle is based (in addition to the pun, of course) impossible, because Loglan distinguishes between the two meanings of and. (Of course, I had never realized there were two meanings before reading L1!) But hey! What point is there to translating a pun unless you translate it in the sense in which it is meant to be misunderstood? [Sorry! I have mislaid the pun to which Djori Rambo refers.]

And that third sentence in the above paragraph could stand some “disambiguation”, but I can’t think just how to do it. “Loglan makes the basic error ...”, uo! [Anyone care to help Bill say this?] 

I’m afraid I haven’t spent much time on Loglan lately ... it sort of “started to come out of my ears” and there are other things, too many other things I wish I had the capacity to do, too. I never really got into M1-3, but what I feel I need now rather than drill and memorization is practice in using the language (the few parts of it I’ve covered...say a third of the book and that not thoroughly understood). [You’ll find that Steve’s new textbook—the first third of which is going to appear as the second issue of La Logli—will do just that for you.] I’ve tried writing simple sentences but of course I don’t know whether they are correct [LIP can help you answer that]. Also the vocabulary available in L4&5 isn’t sufficient to enumerate the small animals I see everyday, or to describe the goings and comings of my goats, or to make out a grocery list. [No, of course it isn’t. For that you need LOD, the new dictionary, the one that’s being constantly updated by the users themselves. But LOD is, I’m afraid, available only as software, not in print. I think it takes about 4 meg on a hard disk. Can you not run a program that big?] In fact, almost the only area [L4&5] seems to cover well is sex! I was forming a very dubious impression of that until I came across your explanation [in the Preface to the Second Edition, p. iii, col. 2] of how [this came about].

One minor item in a host of others. I was gratified to see the word gotca for goat in L4&5, with gotfe and gotma for the genders; this seems like a logical system [one since considerably improved, however; see below]. But then the handling of names for some other animals, say cat and dog, are less regular. Why is this? [This was a defect of the 1975 vocabulary, which has since been improved by Bob McIvor’s “animal inflection”, adopted by The Institute in 1991, in which final -a indicates the female of that species, -o the male, -i the infant or juvenile, -u the generic animal, and -e the comparative adjective meaning like <that animal>. Thus now we have fifteen words for goat, dog, and cat: gotca for nannygoat, gotco for billy, gotci for kid, gotcu for goat-in-general, and gotce for goatlike. Similarly, katma/o/i/u/e are the five words for female cat, tomcat, kitten, cat-in-general, and catlike, while kanga/o/i/u/e are bitch, male dog, puppy, dog-in-general, and doglike. All animal words—even borrowed ones like kangaru—are now subject to this handy inflection.]

Living in Little Dominguez Canyon as I do, I was gratified to see two definitions for canyon; as there are two types of canyon, more different from each other than the one is from a valley. D[ominguez] is the other type, however, and I think “rock walls between” would be a better metaphor for this type than “broken mountains between”. Just a suggestion...I’m not up to word-making yet. [Oh, but I think you are! Why not propose ?trokycalbia? Or even ?vrelanbia from “between cliffs” for your type of canyon? Just write any member of the Purmaogru (word-maker-group), who are now Alex Leith, Bob McIvor, Steve Rice, and myself, with your idea.]

Loglan 1 is formidable until one gets into it. But I wish my college texts had been as clear (and as somehow entertaining) as it is. [Siasia. Clarity has become, through Loglan, the great love of my literary life. And having fun with words has been one of its lesser loves from the beginning.] However, there is a host of small items and one or two major topics that bother me as a newcomer to Loglan and precise logical thought. Some of these are my misunderstandings, I’m sure (since they disappear on rereading); some I suspect are the result of editing and rearranging in the various editions (as the missing Appendix B, which is referenced several times...by the way, I really miss this one appendix, as it is frustrating to be told how to say “a train that is coming or going or both”, but not how to leave the “or both” off!).

Now I see a letter by James P. Salsman in Lognet 94/2:23 concerning an errata list “that might indicate some error, type, or misprint in Loglan 1”. Perhaps I should reread L1 from the start and send him a list of my surviving margin[al] notes; at least they may be helpful in showing what is and what is not clear to a rank beginner.

I hope my last letter did not sound flippant in tone, addressed as it was to “Dear Jim”. Somehow I failed to connect plain “Jim Brown” of the two postcards with Dr. James Cooke Brown, and supposed my correspondence was perhaps being handled by one of the “volunteers” mentioned in...was it the Autumn Bulletin? I even worried a little that this “Jim Brown” was exceeding his authority! 

Yours truly,

Bill Rambo

Lo Narzvo! hue la Djim, ze la Djim Braon, ze la Djimbraon, ze la Djeimz Kuk Braon.              —JCB