(Originally appeared in Lognet 91/2)

Don't Shoot the Engineer

By Mark W. Zacharias

This article is adapted from a letter to JCB, and is accompanied by JCB's responses. These occur both interlinearly and at the end.

First, allow me to describe my background and interest in Loglan, not only as an introduction, but also so that you may see from where the following discussions and comments are coming. I have a Master’s degree in chemical engineering, but have been involved with computers throughout by career. My background is not in computer science, logic or linguistics. Thus, my viewpoint may be considered as technically naive, although because of this, it may be closer to the typical Loglan speaker of the future than it is to the current logli. The current small, elite group seems to be dominated by those three professions.

I own a piece of a small process controls company, which loads me with responsibilities that require enormous amounts of my time. I was initially attracted to Loglan because it sounded like an interesting and worthwhile endeavor. I had planned to be just one of the thousands you were to recruit during GPA, paying my dues, reading Lognet and slowly learning the language. I have managed to find enough time to read L1 and learn all the primitives. But I could not resist making the time to write this letter, which is the result of a year’s worth of stimulus from Lognet.

I have several reasons in particular for writing. I would like to make some comments regarding past Lognets and Loglan in general. I would also like to make some currently informal suggestions based on these comments, and finally, pose some questions which JCB or other expert logli might be able to answer. I will start with my comments.

I was fascinated by Rex May’s grand discussion of Loglan in LN90/2 and LN90/3. While most of what he had to say was anywhere from debatable to clearly wrong, I found the whole exercise to be very stimulating, and I learned a lot by reading the rebuttal. One point that Rex May raised that bothered me also was the letter combination used for the “ow” sound. Given a choice without any prior knowledge, I would have picked the ‘a’ sound from “bat” plus the Loglan ‘u’. Of course, Loglan does not have the ‘a’ sound from “bat”, so the Loglan ‘a’ was correctly chosen. [Yes, but for a different reason.JCB]

The question at hand, then, is what sound is best used for the second half of the “ow” sound, the Loglan ‘o’ or the Loglan ‘u’? Rex May claimed it was ‘u’, but I have decided that ‘o’ is indeed correct. The reason is that the difference between the ‘a’ in bat and the Loglan ‘a’ is that the latter sound comes from deeper in the throat. Likewise, the ‘o’ sound comes from deeper in the throat than the ‘u’ sound. Thus, while the ‘u’ sound would correctly follow the ‘a’ sound from “bat”, forming what might be called an American “ow”, the ‘o’ sound correctly follows the Loglan ‘a’, forming an Oxford English “ow”. [I think that what Mark is hearing as an “American [ow]” is the triphthong formed of Loglan /a + o + u/, a sound that is often heard in America—especially from its childrenas a sign of pain. To make it, we start with [ah], pass through [oh], and end up with a prolonged [oo]. Loglan has no triphthongs; so we just use the first two-thirds of this one as our /ao/ sound. It’s not easy to glide from [ah] to [oo] directly, that is, without passing through [oh]; but it is easy to pronounce [ah + oo] as a disyllable...which is exactly what we logli do to make our au sound. The point is we have both ao- and au-words in Loglan, and we need good, distinctive ways of pronouncing them. We get those ways by assigning the monosyllable [ow] to ao and the disyllable [ah-oo] to au.—JCB]

My next comment regards Loglan on a much grander scale. I, like all other logli, would like to see Loglan adopted for use by a very large number of people. I do not believe, for reasons that I will give below, that this will happen soon. Until much wider acceptance and usage of Loglan becomes reality, I recommend that we be prepared to still accept changes to Loglan, even in a fundamental way, if needed. I know that JCB and other logli have been working on the language for decades. I know that GPA was supposed to be the end of all changes. [Hardly. If the language stopped changing, it would be dead! Our post-GPA policy has been to let the language grow and change, but in ways that do not jeopardize the collectively large investment that has already been made by those of us who have learned large parts of the current language.—JCB] I know that in LN91/1 there was an article by Bill Gober arguing against this (see “Shoot the Engineer...”, page 4ff). But I still believe that changes, if there are good reason to change, should be allowed.

The “Shoot the engineer...” saying is usually applied to situations in which you have a market that is ready for a product. The point is to ship the product now, and make any further refinements and enhancements later. I do not believe that Loglan now has that market, and I believe that there are some compelling reasons for change, which I will also discuss below. In the mean time, I suggest that there should be a good mechanism for disseminating changes among current logli. (I am hazy on these procedures as they exist now.) [Did you receive, or buy later, the 1989 “Autumn Bulletin”? If neither, you would probably find it interesting. It describes the Grammarian, the Academy, and the Word-Makers Council, and their roles in managing orderly yet continuous change. Copies of this bulletin are still available.—JCB]

Why have I offered up these dire opinions about Loglan not becoming a widely accepted language in the near future? [I don’t think they’re particularly “dire,” just realistic.—JCB] Unlike current logli, for whom Loglan is an intellectual [challenge], the average person would view learning Loglan as an effort. In order to get such a person to do so, Loglan would have to offer da some benefit to be gained. Indeed, da’s return would have to exceed da’s investment. Thus, as part of its long-term strategy, The Loglan Institute should strive to maximize the learner’s return and minimize da’s investment.

How can the return from learning Loglan be maximized? In my opinion, the only real solution is to find a commercial use for Loglan. One possible use might be as the basis for a continuous-speech interface for computers. Loglan would be better than natural languages for this because of its low number of phonemes and lack of syntactic ambiguity (which would make high-level processing easier). Another, more useful idea would be as the basis for constructing knowledge bases. Imagine the power of a system which could read articles (in Loglan) for you, and then answer your questions (posed, of course in Loglan) verbally and interactively! This is the type of benefit that would give people a real incentive to learn the language. If I am ever to become heavily involved in the activities of the Loglan effort, it would most likely be in developing applications like these.

How can one's investment in learning Loglan be minimized? The obvious answer is to make the language as easy as possible to learn. I understand that JCB's studies have shown that this is already the case. My personal experience has shown that the primitives are indeed very easy to learn, and I suspect it is because of their designed-in recognizability. I believe, however, that there are other steps that may be taken to improve the learnability of the little words and grammar; some ideas to this end will be described below. But first, I would like to make a final comment about projects like the Loglan development effort, which I believe is germane to my discussion.

Although Loglan was invented by one man, its ongoing development is clearly the work of a collective: a committee. The greatest danger to the success of a collective effort is the lack of a common, clearly defined goal. Therefore, I believe that all logli should be made aware of the explicit goals of the Loglan development effort. In addition to the goals which you have already adopted, such as lack of syntactic ambiguity, I would add:

1) Minimization of arbitrariness

2) Minimization of the number of rules

Cold, hard discipline with respect to these two goals will insure the maximum possible learnability of the language. Incidentally, I voted against the use of any new declensions in the recent survey, because adding them would add new rules. Although declensions allow for smaller words, what may be done with declensions may be done by forming a complex.

Given these comments, I would now like to offer some suggestions. They will range from minor suggestions to major heresies. I must stress that they are suggestions being made by a non-professional, and should only be treated as such. However, I feel that they should at least be heard, on the off chance that those of you with professional expertise and years of experience with designing a language might actually feel that some of them are worthwhile. You will notice, as you read these suggestions, that each one deals with either the elimination of arbitrariness or a reduction in the number of rules.

My first suggestion concerns the various choices offered by Loglan [grammar], such as the choice of using mutce ci sadji ce instead of cui mutce sadji ce. While choices do increase the flexibility of a language, they also make it harder to learn, as the novice must know the meaning of both choices. [The criterion that is governing the design of Loglan grammar at this point is “logical completeness”. Ci is swift and easy; and in nearly all cases, it suffices. But a small residue of the logical possibilities cannot be expressed with an infix. It takes a leading parenthesis, like cui, and a trailing one, like cue, to express them at all. Yet to abandon the "logically defective" infix and confine the options of the speaker to these two parentheses would lead to an almost unspeakable heaviness in the language. It would lead to “symbolic logic made unspeakable” again!—JCB]

I don’t advocate the elimination of choices, but I believe that the Institute should consider designating a subset of Loglan, which contains no such choices, that is a recommended starting point for novice logli. This subset could be given a name such as “Basic Loglan”. [Not a bad idea. Ci would indeed suffice in Basic Loglan.—JCB]

My second suggestion involves adopting a strict one-vowel, one-syllable rule. By doing so, we would eliminate the four rules that specify the pronunciation of “ai”, “ao”, “ei”, and “oi” as single syllables. This will not have such a great impact on pronunciation as one might expect, as the difference between [bey] and [beigh-ee] as a pronunciation for bei is really not that great. The only serious impact this would have would be in the designation of stress, as nerdei would now be pronounced /ner-DEIGH-ee/ rather than /NER-dey/. And this change would eliminate the choice between the disyllabic and the monosyllabic pronunciation of vowel-pairs containing i and u (a good thing, with respect to learnability).

I would also like to suggest some changes for two-vowel words. Below I have a table of two possible sets of modifications. The main purpose of the modifications is to remove arbitrariness from the attitude indicators. The more radical of the two scenarios goes beyond this, as I will describe below the table.

Category Current New 1 New 2 Meaning

Intention ai aa ea Will

ao ai ee Want to

ae ao eo Hope to

au au eu Indifferent

Conviction ia ia ia Certainly

io ii ie Probably

ii io io Perhaps

iu iu iu Don’t know    

Obligation oa oa oa Must

oe oi oe Should oi oo oo May

ou ou ou Doesn’t   Matter

Requests ea ea ai Action

ei ei ei Truth

eo eo ii Permission

eu eu oi Hypothetical

Pure ua ua ua Satisfaction

Expression ue ue ue Surprise

ui ui ui Pleasure

uo uo uo Anger

uu uu uu Sorrow

Change #1 would involve six word changes and would eliminate the arbitrariness in the attitude indicators, replacing it with the following rule:

The strength of the attitude will be indicated by the second vowel and will decrease with alphabetical order (‘e’ being excluded for change #1,’i’ for change #2).

With two additional changes, hi could be made the identity interrogative, thus making all the interrogatives begin with ‘h’; ie would then become the close comma.

Change #2 would involve twelve word changes (including hi as the identity interrogative, as ie would mean “probably”), and would use ao as the close comma. This more radical change would eliminate the arbitrariness of the attitude indicators and make all interrogatives begin with ‘h’, as did proposed change #1. It would also provide a stark contrast between the attitude indicators and the requests, and would be more esthetically pleasing, in particular by avoiding the two difficult combinations “aa” and “ae”. Finally, the “odd-ball” two-vowel word (currently ie), would be more distinct, as only it and ai would begin with ‘a’.

Since I am an advocate of change for the better, I would recommend the second, more radical proposal. I have included the first because I understand that other logli would be more resistant to change.

My next suggestion is to minimize the number of primitives, as this would maximize the learnability of the language from the point of view of the listener. The first change would be to eliminate the cultural declensions, which would have the additional effect of eliminating the three declension rules. A single “culture” primitive, such as loglo would be retained. Complexes would then be used for the “person” and “language” words, such as ?logleu instead of logla and ?logpeu instead of logli. Some examples of other primitives which I believe could be done away with are farfu (male-parent), matma (female-parent), brudi (male-sibling), sorme (female-sibling), dotra (first-year-part) and cimra (third-year-part).

My limited experience with Loglan has led me to believe that certain features, such as pronunciation, formation of tenses, logical operations, primitive and affix vocabulary, etc., are easy to understand and learn. On the other hand, I have found much of the rest of Loglan grammar to be baffling. This may be due to the fact that L1 is not a formal teaching guide, and that once I learn the formal structure of the language, it will make more sense. In the mean time, it would seem that the grammar could be modified to maximize learnability.

While I don’t know the grammar well enough to make any concrete recommendations, I can say that the flavor of the modifications would be along the lines of the suggestions I have made above. For example, I know that there are various little words that serve the function of left and right boundary markers under particular circumstances, such as cue, cui, gu, kui, etc. Is there any rule or pattern which allows the novice to recognize these words as belonging to a class (such as left boundary marker) without having to resort to rote memorization? If there is, I am currently unable to recognize it.

My final suggestion does not concern Loglan itself, but is a suggestion for Lognet. It would seem to me that Lognet should have a column with the specific purpose of answering questions posed by struggling Loglan students. Letters to the editor, of course, may now ask questions, but a dedicated column might be more instructive, and encourage budding logli to ask the questions that, if unanswered, would hinder their progress.

I have two such questions, which I will pose now, but which might be saved for the inaugural column. [Instead, they are answered below.—JCB] My first question concerns the optional one vs. two syllable pronunciation of “i-” and “u-” two-vowel forms. In words that end with such combinations, where is the stress? For example, is vemsia pronounced [VEM-syah] or [vem-SEE-ah], or does it depend on which method of pronunciation of the “ia” I choose?

My second question is a bit more trivial. I came across the sentence No mi totco tu in MacTeach 1, translated by the program as “I am not touching you.” Wouldn’t a more accurate translation be “I don’t touch you”, which would imply a lack of tense? If not, why?

Thank you for reading my ideas; I look forward to seeing JCB's responses to them.

Ae tu gudbi clivi.

JCB Responds:

1. As my in-line comment suggested, you’ve got our post-GPA policy slightly wrong, Mark. (For other readers, ‘GPA’ stands for ‘Going Public Again’, which is what we did in July 1989.) It is not that we’re going to refuse to make any changes in the post-GPA language; obviously that’s not true. We not only haven’t done that, but Loglan would soon be dead if we did. However, the Academy is firmly committed to managing the flow of changes in such a way that it will not jeopardize the collectively huge investment that many people—somewhere between fifty and a hundred of us, I reckon—have already made in learning the post-GPA language. Observing our implied contract with our "investors" is something that we take to be our solemn duty. There’ll be more on our policy toward change in my SLS next time.

2. You say the The Institute doesn’t have a market that is “ready for its product”. But we do. It’s a small one; but it’s real. So the “shoot the engineer” policy does apply. And I’m confident we can make our small market grow into a big one as more and more of our people learn the language, write text in it, attend conferences, start local study groups, contribute words to our dictionaries and finished works to our literature, and in general become skillful users, teachers and exemplars of the language. Not only that, but as our numbers grow, so will The Institute's bank account, and with it, the size of the advertising nets we can cast. Without either risk capital or public funding,we have to start small, reinvesting our profits as we make them. There’s no other way. But a small market is still a market; and some of the buyers in this year’s market can become the exemplars and teachers of the buyers who enter our larger market next year. For example, our membership has tripled in the first two years of our GPA. That's market growth. So yes, Bill Gober is right (LN91/1:4-6): we have a market and the “shoot the engineer” policy does apply.

3. You say that you’re "hazy“ on the Institute’s procedures for change. I strongly recommend that you buy the 1989 “Autumn Bulletin”, which describes them in some detail.

4. The role of the current group of logli—who are probably helplessly (but I trust not hopelessly!) "elite"—is to prepare ourselves to teach the language in the simplest possible terms to men-in-the-street. We can do that; the man-in-the-street probably can’t do it for himself. At least he's not very likely to hoist himself into the language by his own philosophical bootstraps, as some of us clearly can. But once we have established good logical usages, solved the philosophical problems that arise in handling a parsimonious metaphysics—No nouns, no verbs! What sort of language is that!—then we can easily demonstrate to others how this parsimonious language can be used with telling simplicty for the ordinary purposes of speech.

I am reminded in this connection of an experience we had in the Loglan Sogrun back in 1974. It was the first Loglan class I'd taught, and we were having an all-Loglan weekend. A group of us were preparing a meal together when someone asked—in Loglan, of course—how to say “Are there any knives?” I thought for a moment and came up with Ei ba najda vi? And it brought a burst of laughter from my fellow logli. For (as all you 1991 logli immediately recognize, soi crano) the absurdly simple thing I had said back-translates into “logical English” as ‘Is anything (being) a knife here?’

Ok. So perhaps it does take a logical/analytical/linguistical/poetical side to one's brain to get things started: to invent these usages, to see their roles as models for other usages once they have been invented. But it is not going to take that or any other kind of elite to learn that Ei ba _____vi? is how you say ‘Are/Is there any ____ s?’ in Loglan. That’s no more complicated than learning to ask for ‘De l’eau’ and not ‘Eau’ when you want a drink of water in France; and a few tens of thousands of quite ordinary people learn that each year.

So our job—perhaps necessarily a job for a logico-linguistical elite, when the language to be taught is both a logical and a metaphysically parsimonious one, and its usages are still few—is to learn how to teach these simple, elegant, indeed beautiful things to children, newcomers, and pernu-in-the-street.

5. Your formula for growth—”minimize investment; maximize utility”—is of course a sound one. As to its first clause, I do think we have already satisfied it. Loglan is easy to learn...once you’ve got a leading learner, teacher, or examplar who can lead you around the metaphysics, or you learn to do that for yourself. In my opinion, handling the "metaphysics" of Loglan, which means dealing with the “absurd simplicity” of most Loglan usages and the extraordinary elasticity of the Loglan predicate, is the biggest stumbling block to learning Loglan. It is not the number of rules. The size of Loglan grammar is already almost unbelievably small—around 200 rules—while the number of its lexemes or "speech parts" is, at around 50, about a quarter to a sixth of what we find in natural languages. As Mark apparently agrees, the number of rules and lexemes cannot be unrelated to ease of learning. And, as he's already apparently experienced, mastering the primitive set is ridiculously easy. The complexes will, I believe, go in even more swiftly than that.

So much for “minimizing the learning investment”. I feel that, at the level of language design, we’ve already carried that process about as far as it can currently be carried. Loglan may turn out to be the easiest language ever to learn. But we haven’t perfected our teaching kit yet, and we really don’t know how to teach Loglan at a distance...especially to solitary learners, and these are at present nearly all we’ve got. And we don’t have either the money or the volunteer staff yet that would allow us to explore the breakthroughs in second language learning that students of “language acquisition” have recently made. The plain truth is that in our present financial condition, we can’t afford either to spend the money or to divert the energies of our still-small group of volunteers to develop, say, the “computer conferencing” techniques for teaching-Loglan-at-a-distance that Alan Gaynor told us about, or, what may prove to be an even more promising development, the new video techniques for teaching language-at-a-distance by both showing and eliciting the learner’s “total response”. Believe me, these two projects will be very high on The Institute’s agenda when we do have the money and/or the staff to tackle them. In the meantime we’re apparently doing a pretty good job with our CAI (“Computer-Aided Instruction”) techniques for teaching Loglan ...although I have yet to actually talk to a logli who has learned da’s Loglan from a computer!

As to the second half of Mark's growth formula— “maximize the utility of learning Loglan”—I’m afraid that’s largely out of our hands. Many of us agree with Mark that the very brightest star on the horizon of Loglan's prospects is applying it to the man/machine interface, perhaps even ending up, as Mark suggests, with applications that can actually translate, read, study, and report back on truly massive bodies of stored text. But while we can invite people and companies to start developing these and other commercially valuable applications of Loglan—and make our invitation attractive to them by having a sane and non-greedy policy for sharing our intellectual property rights with them (see LN90/1:16-17 and LN90/3:26-27)—we can’t hope to actually develop this heroic software ourselves. To do that, we're talking about millions of dollars...a bit more, soi crano, than the few thousands we still need for our next magazine ad! In short, we’re much too poor, small, and busy with higher priority work to be spending much time on that interface ourselves.

So we rely on others—Mark Zacharias, perhaps!—to keep that fire burning. In sober truth, however, that interface could give Loglan its historical raison d’etre. (I sometimes think that we're surviving now, not for any good historical reason, but through a simple act of will. I certainly hope I'm wrong, of course.)

6. I’ve commented in the text on your ci-suggestion, Mark. It's a good one. It's always useful to master the most commonly used forms in a system first. Almost invariably they will also be the shorter and simpler forms. So by all means let us teach and learn ci first. The cui/cue-pair and their rarer uses can come along in good time. In particular, they can come along when the student wants to know how things that cannot be said with ci can be said at all. Showing yourself, or your student, that such rare but logically completing moves actually exist in the language is often an "ah-hah!" experience.

7. I agree that freedom to choose between the two ways of pronouncing the “optional disyllables”—between [oo-ee] and [wee] as the pronunciation of ui, for example—should not be retained. It’s time that we had definite rules for pronouncing all vowel-pairs, including these ones. So I’ll include the problem of designing a good way to pronounce the “optional disyllables” in next issue’s Lo Nable.

But I disagree, Mark, that the four “natural diphthongs”, ai ao ei oi, should ever be pronounced disyllabically, much less always. To introduce a rule that they are always to be pronounced as two syllables would indeed simplify the rule-structure, but at the expense of inviting people to ignore the rule! People simply can’t say [ah-ee] fast without eventually saying [igh]—especially if their native language contains that diphthong (as most do)—any more than they can say [eigh-ee] swiftly and repeatedly without eventually compressing it to [ey]. To make it a “rule” to speak langorously and polysyllabically in the neighborhood of these potential diphthongs is to make a rule that won’t be obeyed.

8. There are at least two contraries of “arbitrariness”. One is “naturalness”, by which I mean giving the behaving organism the opportunity to do what it is inclined to do anyway. The other kind of non-arbitrariness is “systematicity”, i.e., the property that some arrays of elements have that makes it possible to deduce the function of an element from its position in the array. Apparently when you think of “reducing arbitrariness”, Mark, you tend to think of increasing systematicity, often at the expense of naturalness. But both kinds of non-arbitrariness are important in a designed language. In Loglan, you will find the systematic kind in its digit set, in its connectives, in its tense operators, and in many other places. You will find the natural kind in its attitude operators—especially in the ua ue ui uo uu set that I notice you don't want to change—but also in its personal pronouns, and in our unflinching observance of the Zipf rule that the most frequently used forms in any language must also be its shortest forms. The Zipf principle is linguistic naturalness par excellence. It is language bowing to what the organism wants to do, namely conserve its energy. It would be a mistake, I think, to destroy any amount of that kind of non-arbitariness in a language in order to install the systematic kind. To be sure, both kinds of non-arbitrariness make a language “easier to learn”. But your carefully worked-out proposals for reassigning the semantic values of the VV-form words seem to me to trade naturalness for systematicity at nearly every point. Your proposals would be unacceptable to the Academy on that ground alone... even if they didn’t impose a very considerable relearning burden on us Old Hands as well!

9. Reducing the number of primitive predicates by replacing some of them, like matma, with complexes, like ?fempenre or ?fempee, would be bad engineering on two counts. One, it would trade ethological naturalness (as defined above) for word-economy, and the first is much more conducive to ease of learning than the second. After all, every mammal is genetically equipped to perceive its mother as a “simple object”. Thus, in the mammalian world, "mothers" are not learned intersections of properties, like femaleness and parenticity [sic], but unique sources of care, food, warmth and safety that are—and need to be!— immediately perceived. And humans are mammals. The same reasoning applies to several hundred other predicates for things and properties that are probably also immediately perceived by humans: men, women, boys, girls, daughters, sons, laughter, love, stones, wind, knives, ropes, trees, days, nights, and snakes. Such words are primitive predicates in every human language, and therefore engage concepts which are arguably hard-wired in human heads. If they are, it would be both an ethological and a neurological error—the very height of arbitrariness—to treat these objects as analytically complex ones in the lexicon of any language meant to be used by human beings. Besides, your plan for reducing the number of primitives would replace a lot of short, commonly used words with longer ones, and that’s a crime against Zipf. Both moves are bad ones from an engineering point of view. They would inevitably make our language more difficult to learn and use.

10. The causes of the currently small size of the market for Loglan and Loglan-bearing products are only partly known, of course; but they would appear to have much more to do with the fact that we are currently sharing our natural market with our competitor-imitators, the Lojban group, than with any hypothesized difficulty we are having in learning the Loglan language. All my past experiences with apprentices and study groups —like my four 1977-79 apprentices and the 1974 Loglan Sogrun—seem to indicate that Loglan is in fact remarkably easy to learn...not perhaps when carried to the point of total mastery (which nobody has yet done). But when the goal is handling simple, everyday conversation, including the telling of jokes, Loglan does seem to enter one's head like a dream.

So let’s not fix it unless it needs fixing, Logli! As far as learnability is concerned, it’s probably just fine.

11. Your idea about having a dedicated “How to Say It” column in Lognet is a very good one. In fact, it has long been part of my dream for Loglan's GPA that one or the other of our two serial publications should have just such a "How to Say It" column, and I agree that it seems now better-suited for the (roughly) periodical Lognet than for La Logli. But what we’ve lacked until recently is a columnist capable of fielding such questions...that is, one who's not already too busy, as Jim, Kirk, RAM, Steve and I are, to take on anything else. In short, we simply don't have enough competent volunteers yet to handle all the dedicated columns we would like to do. But I've had my eye on a newer logli whose competence has been developing at an alarming (soi crano) rate and whose interests already seem to be pointing da toward "How to Say It" problems. You may have spotted da in the pages of Lognet. Anyway, I’m going to invite this logli to inaugurate just such a column in LN91/3.

12. Nemsia is normally pronounced /NEM-syah/ not /nem-SEE-ah/. But /ia/ is one of the “optional disyllables”. So until we arrive at a set of rules for pronouncing these optionals in all their contexts, you are perfectly free to pronounce them either way.

13. If I recall the context in which it occurs in L1, No mi totco tu is a response to a rather imperious demand by another person, namely No totco mi! (‘Don’t touch me!') If so, the response doesn’t need a tense operator; it can borrow the tense of the stimulus utterance...which, of course, being an imperative in this case, is the implied present ('Do it!' means 'Do it now!). Thus, na can be understood in the response: No mi na totcu tu. Hence the translation given in M1, ‘I’m not touching you’, which is a form of the English present tense, is reasonably accurate. If the speaker had wanted to say something more explicitly tensed, like ‘I don’t touch you’, in the sense of ‘I don’t ever touch you’, da could have said No mi suna totco tu = ‘It is not the case that I at-least-oneth (ever) touch you’. Or, in a word order that is odd for English, what amounts to the same claim: Mi rana no totco tu = ‘I all-times not touch you’ (‘I always don’t touch you’). But in the absence of such explicit adjustments to the tense vector, we assume, in Loglan, that the vector established by some previous utterance is carried over into the current one.—JCB

Copyright © 1991 by The Loglan Institute. All rights reserved.