From Lognet, issue 91/4.

Sau La Sacdonsu

(From the Start-Giver = Founder)

A Parallel Perception: Talking to Fellow Kejgrudjo Rice over the phone the other day, I became aware for the first time of a remarkable shared perception. He'd encountered Loglan in 1982, as he told us in LN89/1:11, as a result of a library search for new constructed languages. His first contact with the language, therefore, was with its 1975 form. He tried to contact The Institute then but failed. So he had studied the 1975 language pretty thoroughly before making contact, which he did in 1988 or 1989. Steve has been an active member of TLI ever since...very active, one might say, having "risen meteorically" to become one of the Academists (Kejgrudjo) of the language. (Question for the sticky-minded: Do meteors ever rise?)

Steve tells us in that early letter that he was "astonished" at the changes that had taken place in the language between 1975 and 1989, having difficulty at first in seeing the new ways of speech he encountered as anything other than "improper". But over the phone the other day he put a new spin on the difference between the 1975 and the 1989 states of the language that he's now perceived.

It turned out that, like any careful scholar, Steve had made some notes about the '75 language. While it charmed him as a linguistical construction, it didn't seem to him to be a langauge...not in the sense, for example, that Esperanto does seem to be a language even at first glance. So he made some notes about the features Loglan lacked that intuition told him "real languages" should have, as well as the features that Loglan had that seemed to make it not a language. Having done this, he put the notes aside, and, I gather, in the intensity of his later involvement with post-GPA Loglan he didn't think about or look at those notes again until a couple of weeks ago, when he chanced upon them again.

Then he did examine them. And guess what? All of the lacks or defects in '75 Loglan, as Steve Rice then perceived them, had been either supplied or removed—i.e., corrected—in 1989 Loglan!

I was stunned by this observation...happily stunned. Because it matches in a remarkable way the perception I was having from the inside—as the designer/engineer of the language—of the changes that were being forged in it over the same interval. I don't remember how many times between, say, 1977 and 1988 I have pushed back from the work, gazed at what was taking place on the workbench before me, and said reverently, 'My god! It's becoming more and more like an ordinary language!'

Now it's not that I expected—or wanted—Loglan to leap off the table and start roaming over the countryside like Dr. Frankenstein's monster when I shot the current to it. But things were taking place in my creation between 1975 and 1989 which, willy-nilly—that is, without my ever intending them to—did look suspiciously like features of a natural language. Let me repeat: it was never my conscious intent to bring this result about. The things taking shape on my workbench were the product of my consciously pursuing internal design criteria—non-ambiguity, minimality, resolvability, transformability, learnability—all of which, in the extreme manifestations in which I pursued them, were unnatural features of himan language. Those criteria, plus the more concretely quantitative goals—like high coverage for the affix set—which, as an engineer, I was using to measure my proximation to them, were what drove the work.

These being my conscious aims—to push the language toward extreme positions on the various evolutionary dimensions of language which I believed I had glimpsed—it was always surprising to me (and pleasing, too, there is no denying) to see the creature on my workbench taking on the form of an ordinary natural language as well. I might be entitled to hope, it occurred to me, that if it looked like a language, it might actually be one!

That's why Steve's observation, made about the two states of Loglan that marked the beginning and the end of this long "engineering period," was such a welcome one to me. It could be the inside and outside views of the same phenomenon.

That was so promising a thought that I felt I must share it with you.

Secret Thoughts: My earlier thoughts about Loglan's possibly not being a "real language" (in Steve Rice's intuitive sense) were secret thoughts, of course. I have been saying for years that, in 1977, we found out that Loglan actually is a language because we discovered then that people could actually speak it. And setting aside for a moment the rather atypical character of these "people"—my apprentices—they actually could. But I would have been the first to admit even then that speaking Loglan the way my apprentices and I spoke it in 1977 and 1978—about trivial topics and with simple, and at times faltering syntax—was not a test of the full capacity of the language. It was not a test of its "logical engines" at full bore. It did not assess the usability by humans of its"higher syntax" which I meant the full logico-syntactical apparatus by which the most intricate and even nonsensical claims are supposed to pass from Loglan speaker to Loglan hearer intact. No one doubts that Loglan would indeed be a remarkable language if we could prove that that was happening!

The reservations I still have on this score—well, not so much reservations, as the doomed impatience of the aging engineer who would like to see even the "higher syntax" finally tested and quietly at work in spontaneous speech—are typified by my remarks in Lo Nable last time. You recall I asked there (LN91/3:15-16), and set as one of our problems to find out, if we can actually use the righthand closure mechanisms we have formally provided? Computers can, we know that. But will human speakers be able to use them "properly"? Or will we instead relapse into our old dependence on the human hearer's gift for sensing plausibilities, and so feel free to use any of these closers that come to mind, or none at all, saying as we do so the Loglan version of 'Well, you know what I mean!'?

Steve has consoling views on this...or at least on these kinds of problems and their eventual solution. He says that what we are doing to Loglan now, in all the small ways in which we are perfecting and changing the language—and which we have been doing for the last two years—have been moving it even further toward being a "real language." I'm not sure how he knows this. But his intuitions on this matter strike me as worth paying attention to.

Funnily enough, he says that the lojbanists are moving in exactly the opposite direction! They are apparently making Lojban less and less like a natural language. Let me hasten to say that this is Steve's judgement; I have not studied Lojban myself, nor followed its development. (Mea culpa, but I am frankly repelled by even the surface features of this ersatz version of my work.)

Yet to at least one scholarly observer who has looked closely at them both, apparently Loglan is becoming more and more like a natural language, while Lojban, its offspring, is becoming less and less like one. Interesting. I thought you'd like to know.

A Loglan-Only Retreat: Jim Smith was in town last weekend to learn the fine art of "putting together a Lognet", and among other things we talked about was the desirability of repeating one day soon those early Loglan-only group experiences that have proved so fruitful for the language and its people in the past. We decided to call these experiences "retreats".

The first time we held such a retreat was in 1972. That was when 5 or 6 members of a class I was teaching in Loglan at the University of Florida—a not-for-credit class, of course—decided to have a "Loglan weekend" together at my house...if I was willing; and of course I was.

It was quite a weekend. I can remember to this day nearly all the names and faces of this group of early logli who became the first Loglan Sogrun; and some of them displayed an impressive talent even then for the language. I haven't been able to keep track of all of them, unfortunately; but one of them, Mike Pique, is still a member of The Institute, and thus the oldest member of it (shouldn't that be "longest member", soi crano?) apart from myself. (What it should probably be is "djolao", "memberishly old" or "old as a member"; whence 'the oldest member' would be le djolao je raba = 'the memberishly-older-than-everyone-else (person)'.)

We invented a number of kitchen and dining- room usages that weekend. I remember one of them quite particularly. Someone had asked how to say 'Where are the knives?' and after thinking a moment I'd proposed Ba najda vi ie? (That was Old Loglan, of course. We would now say vihu? instead of vi ie? and it would be a single word.) Anyway, I can still remember how the answer, Ba najda va (given by someone who knew where they were, obviously), sent waves of eeriness coursing up my spine. 'Something is (being) a knife there' indeed!

Anyway, such Loglan-only experiences are immensely fruitful...for the logli who participate in them as well as for the language. We know that. (So do the Lojban people, by the way. They've had two such retreats so far...maybe more.) During The Institute's apprenticeship program, back in 1977 and '78, I and whoever was on-board as my apprentice at the time would hold at least one Loglan-only session a day, and often more; and this lasted for about a month of sessions with each apprentice.

That kind of intensive apprenticing seems to be beyond our reach today. TLI couldn't afford to pick up the tab for the turcia's (work-learner's) board-and-room as it did in those days (although, if memory serves, each turcia did pay t's own airfare to and from Palm Springs, or, later, San Diego), and with my heavy other current duties, I couldn't afford the time. But long weekends together, as we shared in G'ville in '72, might serve the same function as it did before: turn spoflo logli into real ones, and beef up the usages of the language at the same time.

What do you think? The venue, of course, would again be San Diego. My wife and I can put up four logli at one time...and not necessarily in married pairs as we have four single beds in our guest rooms. Other participants could bunk out in one of the motels that are reasonably close by; or perhaps by that time, other SD logli with beds to spare could manage the overflow.

As to nahu, what do you say to next Spring? Give me a month or so to get both my land-legs and my land-brain back in my body after their long sea-trip, and we'll do it.

Jim Smith will organize it. Send your applications to participate in our Spring '92 San Diego Retreat to him, and we'll do it.

A Flurry of Case-Tags: We're having a wonderful argument about how to assign case tags properly, on the Loglan e-mail circle. (You might like to join this circle of logli ge lerci srite. Let me join with Jim Smith in urging you to sign up with some e-mail facility. There are about a dozen of us logli now on either CompuServe, Internet, or Bitnet—all of which are connected to one another for the purpose of transmitting e-mail—and the interaction is so fast it's almost like talking. In some ways it's even better than talking... in that you have a chance to think out privately, and in the peace and quiet of your study, what you're going to say in response to some question or proposal. (If only more real conversations were punctuated with such reflections!) Anyway, it's proving very fruitful, and I invite you to join one of these nets and then our Circle. Any one of us with an e-mail address on the backcover of this issue can "let you in", that is, introduce you to the rest of us.)

James Jenning, Bill Gober, Kirk Sattley, Bob McIvor and I have been the main participants in the case tags segment of our Circle. We've been mulling over various rules for assigning case-tags to new predicates, starting with the rules which my daughter, Jenny Brown, discovered and then used in 1987 to make her assignments of cases to the primitives, a work we published that year in NB3. The Case Tag Segment is trying to make improvements on those rules. In doing this, we've discovered that some of Jenny's old NB3 assignments probably need to be changed, but that most of the assignments are very solid. The main need now is to provide our dictionary workers with a handbook—all but one of us is a dictionary worker—something they can use to assign tags to the places of the huge number of new predicates they are finding in the dictionary. The goal of course is to achieve uniformity of assignment. As things stand, assigning a set of tags to a predicate is still a pretty idiosyncratic business.

Dr. McIvor, the Chief Editor of the forthcoming 3rd Edition of our dictionary, has offered Kirk Sattley the job of preparing the Case Tag Handbook, and I'm glad to say that Tisra Sattley has accepted. So we have yet another useful document in the works. I imagine we'll all want a copy of Kirk's handbook when he's finished it. Contact Kirk, by the way, if you've made any studies of this matter that you think might be useful to him.

A Loglan Correspondence Course: Talking of "useful documents in the works" reminded me of yet another exciting project that has gone into orbit since I last sat down with you. That's a new "Correspondence Course in Loglan" that Steve Rice has put together for people on the electronic networks. He's already begun to offer it. (For the time being, anyway, Steve's course is free. So that's still another reason to join one of the networks. It takes a modem, a phone, and a computer to join one, and of course a small monthly fee. The particular network you join will supply you with its software. Many of us find the new CompuServe software pretty impressive.)

I've read the first three lessons in Steve's Correspondence Course, and I can say without hesitation that they're the best Loglan teaching texts I've seen so far...better than the 4 or 5 primer lessons I produced myself back in 1978, for instance. (That was before the logli in my head "started muttering", with the result that I couldn't "hear them any longer." GMR had got to them. They weren't certain any longer how Loglan words were constructed; and so they went mute and remained mute until that engineering job was done. ) Unfortunately, I never got back to the primer-writing task myself. So I am especially grateful to Steve Rice for taking it on now...especially as he seems to have more talent at it than I had! By the way, Steve's Correspondence Course is not the same work as his Loglan Zero, but is related to it.

I think this may be a real breakthrough, Logli. The day Steve Rice's "Correspondence Course in Loglan" hit the electronic networks may turn out to be a red letter day in the history of Loglan. Goodness knows how many computerized logli it may yet inspire to actually learn the language...not to mention the new ones it may drag in from the electronic boondocks.

Djecea go Kukra! Talking of Steve Rice, as I have several times in this column, reminds me that you should be told of a crisis that he and his family are undergoing just now. Two of his grandparents, a grandmother and a grandfather, have recently become quite ill. One is still in hospital and the other one is now at home. Steve, as he puts it, is the "principal caregiver" of his grandfather, the one at home.

Until Steve's family crisis is over, we should not expect the usual high participation of this djipo logli in our community affairs; and it is for that reason that both Lo Cninu Purda and Sau La Keugru have been suspended in this issue. We trust they will be back next time.

In the meantime, Dr. McIvor and I—indeed all of us here at The Institute—wish Steve and his family all the best that man and nature can provide for his grandparents, and, in particular, their speedy recovery.

Kukra djecea, Hoi To Peerpenre je la Stiv!

A New Meaning of "Public Domain": Apparently the LeChevaliers are changing their minds about what the phrase 'public domain' means. Wife Evy came across this snippet in one of their publications recently:

Our publications and software [say the LeChevaliers] are generally copyrighted, with unrestricted permission to copy and distribute any publication or program so long as you retain our copyright notices, and do not charge for the product. [Italics ours.] Any derivative works must subscribe to this license. [Italics again ours.]

The question raised by this new spin on their copyright policy is this: If you can't make money from copying their products, in what sense are those products "in the public domain"? Elsewhere, they say they still are. But what sense does that make now?

Moreover, if you base a work on someone else's work—by incorporating some part of it, for example, in your own (as the LeChevaliers incorporated Institute-provided word-lists in their teaching program back in 1986), which is what a "derivative work" is—but if in order to publish it you must first "subscribe" to that other person's "license" (whatever that means, but presumably it means get one from him), what good is this "public domain" work to you? (a) You can't make money from it, and (b) you must get the other guy's license to publish it. In short, the LeChevaliers' published work on Lojban doesn't appear to be very public domainish any longer!

Clearly their present conception of what constitutes a "public domain" work has been modified from their original conviction, the one that caused them to leave The Institute in a huff. They did that, they said, because they wished to do, and let others do, what they are now apparently trying to prevent their own communicants from doing, namely making and selling "derivative works"!

Ah well, times change...and with them, human minds, soi crano, that is to say, human constructions of reality.

Please don't misunderstand this gentle gybe. We welcome this change of heart (if that's what it is). In any case, we welcome even the apparent convergence of the LeChevaliers' thinking with our own. It may be that this brief snippet heralds the eventual dissolution of our "business policy" difference with the LeChevaliers, the cause celebre that caused them to leave us in the first place. If that difference does finally melt away, that could lead, if not to an all-points rapprochement between our two organizations, at least to an improvement of our public relations with one another; and that is eminently to be desired.

"Predicate Langauges" vs. "Logical Languages": "Recently"—it first came to my attention last Spring in discussions with Steve Rice, but that's pretty recent for a septuagenarian!—a distinction has been drawn by logli between "predicate languages" and "logical languages" along with an associated thesis that Loglan, Lojban, and Guaspi—the latter being yet another linguistic invention by a former loglanist, James Carter—or what might be called "Loglan and its offspring", are better described as predicate languages than logical ones.

Both expressions are of course metaphors. Conceived literally, every language is a logical one; for there will always be some machinery in it for performing some of the logical transformations. Equally literally interpreted, every language is a predicate language, in that all languages have predicates. But these pallid senses are not, of course, what are meant by these metaphors by those who coined and first used them.

The sense in which I first used the metaphor 'logical language' back in 1956 is the Leibnizian one—or at least it is the sense that I associate with the name of the 17th Century German philosopher Leibniz—namely, that a contrived language, a "calculus of thought", is philosophically conceivable, and ought to be humanly contrivable, which would improve its speakers performances of the "ratiocinative arts." Such languages —such a language, in fact, for at the time, I conceived of only one—could be called "logical" both because logic is arguably the primary ratiocinative art (and it would be its speakers' performance of that art which it would be the duty of the new language to improve), and because it was the formal advances that had taken place in the study of logic over the preceding century and a half that seemed to me to justify, in 1960 (when I first proposed it), the revival of the Leibnizian dream. In this original sense, Loglan, Lojban and Guaspi—at least as far as I understand the purposes and functions of these other two—are still logical languages. And so far as I know, there are no others. No computer language, for example, is "logical" in quite this sense...although all are logical in another, humbler sense: they use logic for getting around.

What then is the metaphorical sense of the phrase 'predicate languages' in which it is a better designation of this set of three languages? It was, I think, Steve Rice who first used the phrase; and so we must ultimately ask him. But I have discussed the matter with him, and listened to him characterize the "predicate languages", and I have derived from these discussions the strong impression that Steve believes that no natural language does what each of these three does, namely assign exactly one lexeme to all of its predicates.

This is, indeed, a remarkable move; and it lies at the very heart of the distinction between our logical languages and the rest of the speakable ones. The unilexemic predicate is, in fact, the most powerful simplifying principle in the grammar of Loglan, and the other two languages have apparently followed us in adopting it. So far so good. All three "loglanoid" languages are apparently predicate languages. But is this a better characterization than the Leibnizian one? I do not think so. Because in the end even this most novel of grammatical feature exists in these languages only to make their logicality possible.

How do I know this? Well, I found the unilexemic predicate in formal logic and I put it in Loglan for that reason. Read Quine on the backcover of your L1 for at least one logician's unqualified approval of this move.

The Reciprocity of "Schismatic Surprise": Talking about the LeChevaliers and the schism their "principled" departure precipitated, as we were a moment ago, reminds me of something else you might find illuminating. They were surprised that we even had a copyright policy. I have reported that in LN91/3:18. But so were we at their reaction to this news.

The plain and still surprising fact about them is that the LeChevaliers were the first creative logli we'd ever had dealings with who were the least bit put off by our copyright policy. Everybody else, both before and since, who has made something of value for TLI has taken it for granted that in doing so da was cooperating with TLI—was in some sense serving it, not competing with it—and that TLI would naturally publish, or see to the publishing of, da's derivative work on a royalty-sharing basis. This includes the long string of people who worked on the formal grammar, the makers of LIP, the writers of all our diverse software, the writer of our forthcoming Loglan Zero book and our new correspondence course, and all the editors of our new dictionary and other publications. All these people, without exception, have found TLI's copyright policy not only acceptable but honorable and just. They regard their work as either a gift to TLI or something to be shared with it.

I leave you to figure out why the LeChevaliers were so appalled by this arrangment.—JCB