(From Lognet 97/1)

Sau la Keugru (from the Caretaking Group = Academy)

In this issue the Keugru is making its longest—and some will say its most Loglandia-shaking—report ever. For we will be announcing the biggest addition to the language—the Loglan subjunctive—since 1982’s GMR (“Great Morphological Revolution”). There are also seven other, sometimes related matters on which we have recently made adoptions. We start with our adoption of a pretty new lexical device:

1. NI+UI Compounds: Nemomoui!

The Keugru have adopted a proposal to accept NI+UI compounds as part of the UI lexeme. This allows us to imitate the Spaniard’s graceful Mil gracias!, for example, for in Loglan we too can now say Nemosia! (A thousand thanks!). But we can also say Nemauu! (A hundred sorries!), or even Nemomoua! (A million wows!) should a particularly hearty variety of Eureka! be required (one appropriate, let’s say, to the discovery of room-temperature superconductivity). Such compounds are all single words, of course, and are typically pronounced with penultimate stress: /ne-MO-sia/, /ne-MA-uu/ and /ne-mo-MO-ua/.

This new sub-class of UI words makes a delicate modulation of emotional expression possible. Consider Pimaneuo. I am one-thousandth of a normal ‘Uo!’ (almost imperceptibly?) angry. Or, as this might be translated for the British stage, I am just the tiniest bit displeased.

NI UI phrases—such as Nema, ui = A hundred, I’m happy to say, as in answer to a question—are, of course, still sayable. But the possibility that these same syllables may now make up a compound will make it necessary for speakers of NI UI phrases either to pause between their words or to stress the last one if it is monosyllabic. For example, suppose someone is asked how our team of Loglandian athletes performed in the last Olympics: Ho logli pa gancu? How many logli won? The answer is still sadly none, and was once giveable as /NI-uu/ (with penultimate stress). But /NI-uu/ will now be heard as the rather sneering compound Niuu! Zero sorries! The meaning of None, I’m sorry to say can still be captured with either Ni, uu or Ni uu, however. The first has a definite pause, the second is without pauses but has final stress: /ni-UU/. (Morphological Note: Final stress tells the auditor that it’s not a word a’s listening to, but a phrase. This useful principle comes from JCB & RAM’s current Resolver work. Speech productions with a pause—in which any pattern of stress is permissible: /NI . uu/ /ni . UU/ /ni . uu/—are written with commas (Ni, uu), whereas pauseless productions with final stress (/ni-UU/) are written without commas: Ni uu.)

2. Declensional Suffixes

In SLK 92/3, the Keugru agreed to reserve a number of /-zVV/ forms for future use as declensional suffixes. The thought was that they might be used to make complex predicates declinable. (Complexes made from 3-letter affixes are not, of course, “naturally” declinable, as changing the final vowel of any such affix changes its meaning entirely.) It appeared, though, that in the five years during which this option has been available, no use had been made of it. So the possibility of dropping this reservation was considered by the Keugru. We mentioned that we were doing so to the e-connected logli (to the Logli List, to be specific).

At that point Bill Gober responded with a convincing argument for keeping at least some of these suffixes, and went on to propose two specific assignments. After some discussion between the Keugru and Bill, it was decided to adopt /-ziV/ as the “declension-carrier” on complex predicates when they are to have “ethnic” meanings—that is, when they are to predicate bits of languages, cultures, territories, or peoples—and to use /-zuV/ to put declinable endings on complex predicates when they refer to animals or animal-like beings.

The important semantical point here is that, without such suffixes, there is no way to apply the declensional vowel-changes to a complex predicate ending in a short affix. So having these suffixes available to us closes a semantic gap in the language.

Here are some of Bill’s examples:

The Loglan for the generic rabbit is lepsu. This word, being primitive and applying to an animal species, is of course easily inflected for sex and gender, lepso being available for the buck rabbit, lepsa for the doe, lepsi for the bunnies, and lepse for lapine or rabbit-like. That being done, however, one cannot apply the ethnic declination to such an “animaline” root. Lepsa still means female rabbit even if you (mistakenly) intend it to mean a bit of rabbit language. But in the book Watership Down, rabbits are given a culture, a mythology, and a language; in other words, they’re treated as an ethnic group. Given the new ethnic-declension suffixes, the translation of this book into Loglan is easy. An instance of the lapine language can be referred to with lepsyzia ([lep-suh-ZEE-ah]), of the culture of rabbits with lepsyzio, and of rabbit territories with lepsyzie.

Even in less whimsical contexts, as Bill remarks, “it may be useful or instructive to talk about peoples using descriptive names for them. For example, a speaker might want to focus on the common features of the languages of rice-eating peoples, using rismytcizia [rees-muh-chee-ZEE-ah]. Or on the common habits of rice-eating animals, with rismytcizua. On on anteaters with mantytcizuu… words that are often much clearer and more descriptive than any possible borrowing. For example, if we borrowed the Linnaean words for them we’d have mirmekofagidi for the generic sense of anteater and karnivori instead of miortcizuu for carnivora.”

While not endorsing the abandonment of borrowing from the vocabulary of science, soi crano, the Keugru has adopted Bill’s proposal, especially for the more whimsical, science-fictional uses he evidently has in mind for them. So the suffixes /-ziV/ and /-zuV/ can now be used to build complex predicates whose final vowels will convey the meanings reserved for the ethnic and animal declensions, respectively.

There are two easy mnemonics for English-speaking logli trying to learn these two new series of suffixes. For the animal series, just think of /zu/, which is the pronunciation of the E word zoo, and then put the endings /a e i o u/ on it. For the ethnic one, the /i/ in /-ziV/ can be associated with the ending used to convey the “people” sense of the ethnic declension, which is /i/. Thus the rice-eating people would be known collectively as lo rismytcizii [loh-rees-muh-chee-ZEE-ee]… like the old Roman names for Celtic tribes.

3. The Personal Variables Reconsidered

The system of Loglan “pronouns”—Free Variables (Loglan 1, 4.4) and Personal Variables (Loglan 1, 4.5)—has been recently rethought. The da de di do du system of Free Variables—variables referring to previously identified “3rd person” entities, in the manner of a pushdown stack—has faded from use and has now been largely replaced by the anaphoric use of letter-words in speech and either letter-words or “letterals” in text. These “Letter Variables”, as we may now call them, replace descriptions or names by repeating the initial letter of the principal element of the designation so replaced. Conventionally, lower-case letter variables are used to replace descriptions and upper-case ones to replace names. Thus mei, or simply ‘m’ in text, might replace la langa mrenu (the tall man) in later references to him, and Mai, or simply ‘M’ in text, could be used to replace la Betis Meisn in sutori references to Fum Meisn.

The Five Free Variables of the da-series are still used in two way: (1) explicitly, as when a long designation is linked by a ji or jio phrase to some da variable (Le mrenu ja pa godzi le cefmia gui ji da langa = The man who came to dinner, who is/will henceforth be known as X, is tall), and (2) implicitly, as when a “Demonstrative Variable”—one of the set {ti ta toi toa tio tao}—is replaced by a free variable when the speaker/writer wishes immediately to refer to its designatum again. (Ti bukcu. I da totnu. Ibou mi napa kalridle da. = This is a book. It’s thick. However I’ve completely read it.) Other ways of using the da-series are under consideration.

This reduction in the use of what were once the most generally used 3rd person variables also means that the three parallel series of 1st and 2nd person variables—mia, mie, … , mua, mue, … , tua, tue, …—are now almost never used. This observation, plus the recently increased emphasis on the distinction between “sets” and “multiples” when denoting plural entities (see JCB’s articles on this topic in LN 95/2:22ff and LN 69/1:7ff), has lead to a reform of the system of 1st and 2nd person variables.

The new forms that have been adopted both preserve and clarify the distinction between the “inclusive” and “exclusive” forms of we and you as well as allowing new distinctions to be made between their singular and plural senses. In addition, we now distinguish between the set and multiple senses of all plural forms of the personal (1st and 2nd person) variables.

In the next column is the complete table of the twelve personal variables. The two singulars, mi and tu, are old, familiar forms (although tu is now more precisely defined as singular) while the ten CVV forms are all plurals and new. These ten new forms replace the 15 old /miV muV tuV/ forms, a welcome return of 5 forms to the CVV pool. The old, ambiguous ?mu, which meant neither set nor multiple, has been similarly retired.

The most important novelty in this table is that the old functions of mu are now being served by two new words, mui and mua, one having the set, the other the other the multiple, sense of English word we. Thus mui means you and I jointly, i.e., we/you (speaker and auditor) in its “together” sense (Ea mui furvea levi hasfo = Let us buy this house) while mua means we/us (speaker and auditor) in its “independently” sense (Mua metro lio suto = We are each at least two meters tall). Notice also that tu has been slightly but significantly redefined. It now always refers to a single auditor. If a plural audience is intended, one now uses either tui or tua, words with the same vowels and logical structure as mui and mua. Thus tui treats the plural audience a set (Tui pa kambei leva tristaga, ue! = You lot carried that log (surprise)!) while tua treats it as a multiple (Tua metro lio sini, ue! = You’re each of you at most one meter tall (surprise)!). You will observe that these new forms are all extraordinarily succinct… at least compared to what it takes to make such designations in English.

There are some useful mnemonics for learning the new plurals. All 1st person plurals forms start with m, and all 2nd person plurals start with t. If one or the constituent designata of a plural form is tu then u is one of its vowels, and the converse is also true. If the last vowel in a plural form is e, then that form refers to a set (you may associate this e with the English e of set). If its last vowel is o, it refers to a multiple. The pair of pairs tue/tuo and mue/muo have similar patterns of vowels, constituents, and functions, as do the pair of pairs tui/tua and mui/mua. You may associate the final vowels of this last quartet with those of ti and ta. The ti-like tui and mui designate the more “closely linked” sets of individuals acting together (e.g., carrying logs) and the ta-like tua and mua designate the more “widely spaced” multiples of which something notable is true of each individual separately (e.g., being seven feet tall).

As mentioned, all the old /miV/, /muV/, and /tuV/ forms not reassigned here are now free to take on other assignments… except for mia, of course, which has been reserved for the Mia sense of the new Loglan subjunctive; see Sec. 7 below. /mu/ will also be given a new assignment; but the Keugru has not settled on one yet.

How should one use these new variables? They allow logli to make all the distinctions we often unconsciously intend with our personal pronouns when speaking English, but that we cannot easily make in that language. So they often go unspoken and unattended. Here’s an example of the distinctions we logli can now intend and make succinctly in Loglan:

During a telephone conversation one member of a logli couple says to a friend of theirs:

Muo napa ridle le cninu bukcu, inurau ea mue grujmi moi lepo takdui bei.

We (you, I, and my absent partner) have all (separately) read the new book, so let’s (all three of us collectively) get together to talk about it.

Everyone is aware of these distinctions… even in English, though we cannot easily make them in English. In Loglan, we can now make them easily. In fact we must make them if we are to covey our intentions. Let’s find out how difficult—or easy—that’s going to be for our still natural language (NL) shaped minds!

4. Set-Multiple Conversion via Descriptand

In the previous Sau La Keugru (that of Lognet 96/3), the Keugru gave a 2x2 table of descriptors of plural entities that showed the distinction between sets and multiples as one dimension, and the distinction between definite and indefinite description as the other. As logli studied this table the question arose—first from James Jennings—about “converting” between set- and multiple-descriptions. That is, if we had originally mentioned a set of, say, 10 persons, how can we then say something about the multiple comprised of the same 10 persons, and vice versa?

The Keugru thought about this and answered the question by adopting the convention that any simple change of descriptor, when used with the same predicate expression, is to be taken as designating the same collection but in the “mode” of the new descriptor. Thus all four transformations indicated by the arrows in the table below are now allowable:

JCB’s discussion of this new usage with the Keugru explains it well:

“After hearing Leu nete bukcu (The set of thirteen books I have in mind) spoken—perhaps by another speaker—one could simply repeat the quantifier and the “predexp” (predicate expression) but change the descriptor and say Le nete bukcu, and mean by that Each of the thirteen books I have in mind.

“When anybody asks what you do have in mind, tell them that you mean to designate as a multiple all and only the members of just that set that you or someone else had previously designated as a set with Leu nete bukcu.

“This is probably more workable than it looks. And if it is we could conventionalize the two le <—> leu transformations. Thus, on hearing ‘le NI <predexp>’ we could create a designation of the corresponding set by saying ‘leu NI <predexp>’; and vice versa. This might not work for the indefinites, however, since with them, there is nothing “in mind”!

“Still, why not! You don’t have to have the identity of two unknown robbers “in mind” to speak of the set of them. (Whoever they were, they were apparently tall enough collectively—standing one on another’s shoulders, perhaps—to reach that window.) We can also speak of their multiple. (They were probably both cigar-smokers, as there are two fresh cigar-butts in the ashtray.) So apparently we can conventionalize the moves ‘NI <predex>’ <—> ‘NIcu <predexp>’ as well.

“… Given Ve mrenu in the prequel, Vecu mrenu will indefinitely designate the set of those same nine men who had been referenced as a multiple earlier; and given Vecu mrenu in the prequel, Ve mrenu will indefinitely designate the multiple of those same nine men who had been referenced as a set earlier; and mutatis mutandis for the two definite transformations.”

That the two “NI”s are in parentheses in the Definite row of the table above means that the le <—> leu conversion will be applicable even if the descriptor does not include a numeric expression. Here is an example of such a usage:

Leu muzkao pa mutce gudbi pubkao vi le fomymia. I re le miatci ga laldo mrenu, ibuo le muzkao ga junti fumna.

The musical-group (that I told you about) played very well at the banquet. (And) Most of the diners were old men, but (all of) the musicians were young women.

5. Comma-less Names

Many learners of Loglan have found that the requirement for a comma after each name in text is something of a stumbling block. It was easy to forget to put them in, in writing, and in reading, the post-name commas often seemed to break up the flow of thought.

Upon being assured that LIP (the “Loglan Interactive Parser”)—in the next release, at least—can recognize names in written Loglan without depending on commas written after them, the Keugru has agreed that such commas should become optional in text. A logli writer may now put commas after names or not, as w chooses, but w should be consistent within any given document in whatever practice w adopts for d.

Logli should, of course, continue to insert either “long” or “short” pauses after words used as names in speech. (See SLK 96/1 for the reasons for these two kinds of pauses.) Recall that extremely short pauses, or “stops” (/./), which had long been used to separate vowel-initial words from their prequels (/la.ai-LIN/), have also been used since 1961 to separate elements in a serial name (/la-krist.DEN-li/ = La Krist Denli = The Christ Day), whereas longer pauses (/ . /) are required to separate names from ensuing predicates (/la-djan . SUC-mi/ = La Djan sucmi = John is a swimmer). At the moment LIP requires us to express long pauses in speech with commas in text (La Djan, sucmi); but in its next release, Takrultua (Grammarian) assures us, LIP will not require us to write them after textually identifiable names: La Krist Denli hapci ckemo.

This proposal comes from Alex Leith. To allow your eye to experience the new textual style, we shall use comma-less names in the Loglan text in the rest of this column.

[As the reader will have seen from previous articles in Lognet, there have been lengthy discussions about the representation of counterfactuals, subjectives, and conditionals in Loglan. In order to report the Keugru’s official action in this complicated area, it was felt that a newly-organized, complete laying-out of the problems and their solutions was required. Prof. Brown kindly undertook to produce such a summary document, and it is included here as the last three sections of our current report.—KS]

6. Sio, Dau, Biu: Three Probability Operators

In my article on “The Mia System of Subjunctives” (Lognet 96/2:3ff) I also proposed the adoption of three new probability operators, sio dau biu. These would serve as objective versions of the three grades of subjective conviction that are already indicated by the words ia io ii (certainly, probably, perhaps). Because the latter are “attitude indicators”, however, and so members of the UI lexeme, they have no effect on the logical structure of the claims to which they are attached. But these three new probability operators are members of the PA Lexeme, and so have all its privileges and restrictions. Among these are that PA words change the truth conditions of the sentences in which they appear. This is the most important reason then, for the Keugru’s adoption of these three new operators: they carry on the classical Loglan program of distinguishing between the “emotive” and the “cognitive” components of speech… a program that is almost as old as the language itself. The Keugru wishes to emphasize that we would probably have adopted these operators quite apart from their connection with the Mia proposal.

On the other hand, an important feature of the Mia System is the set of three “compound subjunctive operators” {siomia daumia biumia}, which are obtained by compounding {sio dau biu} with mia. We shall be reporting on those along with the rest of the Mia System in Sec. 8 below.

You may recall from my LN 96/2 article that the three new PA words are to be both derived from, and eliminated by, three especially-defined predicates sirto, dakli, and blicu. All three of these existed before the {sio dau biu} adoption but required both more precise and more uniform definitions to enable their uses as eliminators. Here they are as redefined:

sirto (3a) State or event V is certain to be or happen, or has a probability of 1.0 of being or happening, under conditions N given knowledge system S.

dakli (3a) State or event V is probable, or is likely to be or happen, or has a probability greater than 0.5 of being or happening, under conditions N given knowledge system S.

blicu (3a) State or event V is possible, or has a probability greater than 0 of being or happening, under conditions N given knowledge system S.

Sio from sitro, dau from dakli, and biu from blicu are all regularly derived CVV words. As each is a PA word, it may be used as an inflector, an adverb, or a preposition. Using sio as our example:

Da sio mamla.

X is certainly a mammal (under conditions obtaining at the time of speech).

Da mamla sio.

X is a mammal certainly (under these conditions).

Da mamla sto ta.

(That) X is a mammal is certain under those conditions.

All three sentences presuppose the existence of some unmentioned knowledge system x capable of guiding such probability calculations.

Note that the addition of sio dau biu to the language does not “disemploy” ia io ii. Among the following six sentences, the last three all have the same truth conditions—that is, make the same claim—while the claims of the first of these are not only different from one another but each of these first three claims is different from the common claim of the last three:

Da sio mamla.

X is certainly a mammal (given these conditions and some knowledge system x).

Da dau mamla.

X is probably a mammal (given these conditions and some knowledge system x).

Da biu mamla.

X is possibly a mammal (given these conditions and some knowledge system x).

Da ia mamla.

X, certainly, is a mammal.

Da io mamla.

X, probably, is a mammal.

Da ii mamla.

X, possibly, is a mammal.

The common claim of the last three sentences is, of course, made by the kernel sentence Da mamla; for the ia, io, and ii that adorn it are only expressions of the attitudes of the speaker toward s’s claim. But s’s attitudes make no differences in the truth or falsity of s’s claim. In contrast, sio, dau, and biu make very definite differences in s’s claims.

Any of the set {sio dau biu} may be eliminated from any sentence in which it occurs by using its “deriving predicate” to write an equivalent sentence of the following form:

Da sio mamla <—> Lepo da mamla ga sirto ti ba.

In the righthand sentence ti refers to the circumstances of speech while ba asserts the existence of some knowledge system x capable of guiding probability calculations. It will now be clear that such a sentence makes a very different, and much more elaborate, claim from that of Da mamla. So with { sio dau biu } we fill a semantic gap. Sentences using these operators will represent the objective or cognitive expression of epistemic relations that have previously been only emotively (when succinctly) expressed in Loglan.

7. Roi, Kau, Fui, Foi, Nui: The Five “Auxillary Verbs”

In my Lognet 96/2 article I also proposed five new “auxiliary verbs” for Loglan. These, too, have been adopted by the Keugru. They are to be used before the main predicates of sentences much as can, will, should, must, and may are used in English; it is in that sense that they are “auxiliary”.

The five auxiliary operators I proposed while discussing the Mia System were not so much part of that system as potential sources of confusion with it… unless, of course, they were clearly distinguished from it. Thus the original purpose of adding these five auxiliaries was to distinguish them from the subjunctive operators we anticipated adopting.

In the natural languages (NLs)—perhaps especially in English—auxiliary verbs often play dual roles, one in the subjunctive system of that language, the other in some other of its systems. For example, English could, which is sometimes the past tense of can, is often impossible to distinguish—even by context—from that other English could that is the “possibility” variant of the English subjunctive. (For example, He could swim the English channel. Is this the past tense? Or is it a subjunctive with some unmentioned condition?) Thus one must often guess at what the speaker of could “really means”. Unhappily, even the speaker of English is often unclear about what s “really means” by its ambiguous auxiliaries. It was to resolve such ambiguities that the Keugru has decided to adopt these five auxiliaries.

Having adopted them, however—and having used them experimentally in our own writing—several members of the Keugru have become aware of how much we’ve needed them for other reasons. Indeed, they seem to be “However did we get along without them!” sorts of words.

In many respects the uses of these new auxiliaries parallel the uses of { sio dau biu }. But in several other respects, the auxiliaries are quite differently used. This will be clear when we examine their eliminating formulas, for these tell us what we are “really saying” (even in English) when we use auxiliary verbs.

The five auxiliaries the Keugru have adopted are roi, kau, fui, foi, and nui (roi was *fuo in Lognet 96/2), and they correspond in a rough way to the ambiguous English series will, can, should, must, and may.

Let’s consider the new words one at a time, noting the meanings of their source predicates and the formulas by which they may be eliminated:

roi <— furmoi (4v) K will/intends/plans to do V for motive S under conditions N.

To remember roi, think or the “royal will”. It is used as follows:

X roi preda Y <—> X furmoi lepo X preda Y guo x ti, where Y stands for the set of sutori arguments of preda if it has any, x asserts the existence of some unknown motive S, and ti stands for the conditions N obtaining at the time of speech.

X will preda Y <—> X intends to preda Y for some motive x under these conditions.

Example:

La Meris roi godzi la Frans la Englynd.

Mary will/intends/plans to go to France from England.

Eliminative expansion:

La Meris furmoi lepo Mai godzi la Frans la Englynd guo ba ti.

Mary intends the act M’s going to France from England for some motive x, and this intention is “active” under these conditions, i.e., the conditions obtaining at the time of speech.

Thus the eliminative expansion of any inflectional or adverbial roi makes two hidden features of its claim explicit: (a) the existential claim (conveyed here by ba) about the actor’s motive S, and (b) the understanding (conveyed here by ti) that the actor’s intention is “active” under—has perhaps been precipitated by—conditions that are current at the time of speech.

But what if it isn’t? What if conditions not yet present are necessary for this intention to become active in Mary? Suppose it is now January and the speaker knows that Mary doesn’t intend to go to France until May? The conventions of all PA operators—and roi is a PA operator—tell us that in that case we can use roi prepositionally to designate just these not-yet-present conditions:

La Meris godzi la Frans la Englynd roi la Fermea.

Mary will/intends/plans to go to France from England in (the coming) May.

Notice how similar this usage is to the prepositional use of pa:

La Djan hasgoi pa la Ven.

John went home before nine.

With this understanding of roi we can now give the meanings and uses of the other four auxiliaries more briefly:

kau <— kanmo (3v) K can do V under conditions N.

X kau preda Y <—> X kanmo lepo X preda Y guo ti.

X can preda Y <—> X can preda Y under these conditions.

Example:

Ei la Djan kau sucmi lo lanbie?

Can John/Is John (now) able to swim to shore?

Eliminative expansion:

Ei la Djan kanmo lepo Dai sucmi lo lanbie guo ti?

Can John/Is John able to do the act of J’s swimming to shore, under these conditions?

No existentials are implied by the use of kau. (To remember kau, its only slightly irregular derivation from kanmo is likely to be sufficient. Kao, recall, has long been the “Actor” case-tag as derived kakto.)

fui <— funrui (4v) K should do V under conditions N according to moral authority S.

X fui preda Y <—> X funrui lepo X preda Y guo ti x.

X should/ought to preda Y <—> X should/ought to preda Y, under these conditions according to moral authority x.

Example:

Tu fui cluva letu snilii.

You should love your neighbor(s).

Eliminative expansion:

Tu funrui lepo tu cluva letu snilii guo ti ba.

You (a single auditor) should/ought to perform the act of you loving your neighbor(s), under the conditions obtaining now according to moral authority x.

Thus a hidden existential is always implied by the use of fui, namely that a moral authority for such injunctions exists. To remember fui, note that it is composed of the first two plus the last sound of funrui = should.

foi <— folfunroi (4v) K must do V under conditions N according to moral authority S.

X foi preda Y <—> X folfunroi lepo X preda Y guo ti x.

X must preda Y <—> X must preda Y, under these conditions according to moral authority x.

Example of a prepositional use of roi:

La Kein foi ta pa no mormao leKai brudi.

Cain, under those conditions, was obliged not to kill his brother.

Eliminative expansion of prepositional foi:

La Kein pa folfunroi lepo Kai no mormao leKai brudi guo ta ba.

Cain was obligated to perform the “act” of his not killing his brother, under “those” conditions (presumably those obtaining at that time) according to some authority x.

Like fui, there is a hidden existential behind every foi sentence. (To remember foi, note that it is the “strong” version of fui, which means changing the /u/ of fui to the /o/ of forli = strong.)

nui <— nurlei (4v) K is permitted by moral authority S to do V under conditions N.

X nui preda Y <—> X nurlei x lepo X preda Y guo ti.

X may preda Y <—> X is allowed/permitted by some authority x to preda Y under these conditions.

Example:

Tu nui godzi la Paris la Romas.

You may go to Paris from Rome.

Eliminative expansion:

Tu nurlei ba lepo tu godzi la Paris la Romas guo ti.

You are permitted by some moral authority x to perform the act of your going to Paris from Rome, under these conditions.

Again, an authority is implied; but this time S is in the 2nd place of the eliminating predicate (K nurlei S V N). Still our eliminating convention holds. We look for the “N” slot in the eliminating predicate. In the case of nurlei, this happens to be the last slot in its four-place structure. We fill that slot with ti in the eliminating expression. We do this, or course, if and only if nui is being used inflectionally or adverbially; for it is only in these cases that ti may be assumed.

Just as sio dau biu “objectify” the indicators ia io ii, so four of these new auxiliaries, roi fui foi nui, objectify four other attitudinals, namely ai oe oa oi.

It is often useful to use these auxiliary words prepositionally. For example, we often wish to specify the permission-giver in the case of nui. How do we do that succinctly? That is, without falling back on the rather untidy eliminative expansions themselves?

I have proposed to the Keugru that we try out the following usage. It will work for those four of the five auxiliaries that have 4-place eliminators—these are also the ones that have subjunctive indicators—and therefore have an “extra slot” in their place-structures after the “N”-slot is filled in.

Normally, when an auxiliary is used prepositionally, its operand is exactly a designation of the conditions N. But suppose the conditions are local conditions and so need not be explicitly mentioned? Even so, they may be designated by ti, and it will be useful to do so just in case we wish to fill in the extra slots as well. With the auxiliaries roi fui foi and nui we can do this in the following way:

Construct the operand of the prepositional auxiliary in the form ti ze <arg>. In this expression <arg> is a designation of the motive or moral authority involved… for example, the permission-giver if the auxiliary is nui. In all such cases ti is a perfunctory designator of current conditions, a mere place-holder. Of course if the conditions slot of the auxiliary also needs filling with something more informative than ti, we may use the form <arg₁> ze <arg₂> to build a compound operand. In this expression we may use <arg₁> to designate the conditions N and <arg₂> to designate the source S. In either case, we create something that is grammatically a single operand, as is required by any prepositional operator.

An example will show how understandable this novel usage actually is:

La Kein pa no mormao leKai brudi foi ta ze le hebro ze kristno gando.

Cain was obliged not to kill his brother under these conditions and jointly by the moral authority of the Judeo-Christian god.

Using ze to form designations of ordered sets—which is what is going on here (for the order of ta and le hebro ze kristno gandi clearly matters)—is thus a temporary expedient. If the trial use of ze in such positions suggests the value of an order-sensitive linking word, then the Keugru will probably adopt an allolex of ZE to build designations or this kind. (Of the five /zeV/ forms, LOD tells us that /zee/ and /zeu/ are still available for this allolex.)

The idea of using a ze-type link between designations of the elements of ordered sets is a proposal only. It has been on my personal “back burner” for some time. But this new evidence of its possible usefulness may well justify its adoption by the Keugru. Stay tuned.

8. Mia: the First of Two Subjunctive Operators

The Keugru has decided to adopt the Mia System of subjunctives as described in Lognet 96/2:3ff. But mia will probably be only one—though possibly the one more widely used—of what we expect will eventually be Loglan’s two subjunctive operators. (The second operator is still under consideration. It will probably have a sense fairly similar to those to those of the Fio/Foi operators proposed in James Jennings’ article “I Would If I Could” in Lognet 96/1:33ff.) In other words, the Keugru has decided to resolve the sometimes vigorous debate between these two schools of Loglandian thinking about the so-called “counter-factual conditional” by incorporating both their models into the language. This treats their theories as linguistic options rather than philosophic rivals. In short, we wish to let neither theory “carry the day”.

After nearly two-and-a-half years of discussing the intriguing problems—logical, philosophical, and linguistical—that are raised by the existence of counter-factual conditionals in human speech, the Keugru has come to the conclusion that the “subjunctive mood” (the linguistic device normally used in NLs to convey counterfactual claims) means quite different things to different speakers… even quite different things to the same speaker on different days! It is now apparent to us that the subjunctive, together with the numerous other kinds of conditionals and hypotheticals that adorn so many NLs, occupies an extremely muddy area of NL semantics. No single semantic theory explains all these devices; no single logic wipes them transformationally clean.

Even the two quite different interpretations of the counterfactual that we’ve been offered do not exhaust the possibilities. In the end, we’ve come to see that our task as the Loglan Keugru—the group that watches over the continuing development of this logical language—is not to decide the philosophical issue between these two opposing camps—that issue is a very large and serious one, but it is not our business to act as “philosophical referees”— but to make both ways of speaking legitimate in Loglan.

In fact, we’ve discovered, there are at least three ways of interpreting the NL subjunctive. We want to make all three of them available to logli with the sole proviso that whatever way is chosen by a speaker will be clearly understood by s’s auditors.

In short, we are offering the Loglandical community “options” again, but this time not only the option to be clear—to paraphrase Reed Riner’s felicitous phrase—but also to be clear about one’s philosophical position as well. Thus the Keugru has sought—and believes it has found—a broad and flexible subjunctive machinery that will permit logli to speak subjuctively when and if it suits them to do so, or not to speak subjuctively when and if it doesn’t suit them, and if subjectively, with meanings that, as Alex Leith has recently put it, are “blindingly clear.”

Our program for accomplishing this complex objective is as follows. We will provide at least three quite different ways of translating NL subjunctives into Loglan. Two of these ways will be described in today’s report. However, the third can only be adumbrated here. We hope it will be fully specified in our next report on this matter, a report to which we hope our Lodgru (Logic Group) will contribute decisively.

Here are the three ways:

First: We will provide the mia operator to convey the subjunctive sense of the English subjunctive conditional as it covers reports of planning, intending, imagining, scheming, joking, assessing, and other purely mental acts on the part of the speaker. Some examples of this widespread, Mia-type usage in English are given in my article on this “mentalistic” model of the subjunctive act: If you’d bring me that chair, I could reach this lightbulb. Even if you had taken my seat, I wouldn’t have minded. If a tornado were to come to my town, I’d seek shelter in my cellar. If he’d lived ’til next Thursday, he’d’ve been dead a month. All such untestable-in-practice claims—that is, they are all untestable by current scientific practice because they all refer to goings-on in the privacy of some speaker’s head, but they are certainly not essentially or forever untestable—can now be rendered with mia in Loglan. For example, the four imagination-dependent claims made in English above can be rendered with mia as follows:

Mia lepo tu kambei leva cersi mi guo, mi kau teutco levi litbui.

In the mental place where you carry that chair to me, I can reach this lightbulb.

Mia lepo tu papa tokna lemi nurski guo, mi no pa nurfau tei.

In the mental place where you had taken my seat, I was not troubled by it (your doing so).

Mia lepo lemi cmasitci ga nu tonteri guo, mi nuo karko vi lemi hafnilca.

In the mental place where my town is “tornadoed” (visited by a tornado), I shelter myself in my cellar.

Mia lepo da pa clivi pia la Neapri Fordei guo, da papa morto nia ne mensa.

In the mental place where he lived until Next Thursday, he had been dead for one month.

Thus, even contradictory arrangements can be usefully imagined in these “mental places” that we humans so readily manufacture. Even contradictoriness is evidently not prohibited by the neurological rules governing the construction of our mental worlds. [Those who have read my 1970 futurist novel The Troika Incident may remember my description in it of a device— there jocularly called the “Eager People Finder”— that would allow us to test such currently untestable claims about what goes on inside other people’s heads. Certainly science is moving very swiftly in that direction. Mia-claims are not likely to remain untestable for long!—JCB]

Second: Whatever Loglan operator is adopted for the objective—or, as it is sometimes called, the “Platonist”—sense of the NL subjunctive, it must convey a speaker-independent view of the world or worlds about which the speaker is making counterfactual claims. That is, whatever truths this second subjunctive operator helps us claim, they must be presumed independent of the claimant in roughly the the way that truths claimed about the observable world—about those five sparrows sitting on that telephone line, for example—are in some sense independent of the inner states of the observer. Note that the claims mia helps us make are not of this sort. They are in part, at least, about those inner states.

We have asked the Lodgru—a group of logli composed of our two active lodtua (logicians), Randall Holmes and Emerson Mitchell, and our one currently active fidsesmao (physicist), James Jennings—to propose a CVV-form operator that will mnemonically convey this alternative, objective meaning of the subjunctive to our next generation of logli. We have also asked them to propose the usages and elimination formulas of and for this “second subjunctive”. Their proposals have not yet been received. So until we have them let us call this second subjunctive operator simply xxx.

Many features of xxx are already knowable. We can expect xxx to have essentially the same grammar as mia, though obviously it will have a different elimination formula; it will also make the same sort of compounds with {sio dau biu}; and we trust that logli will use these xxx-type operators whenever they wish to make a challengeable claim about some currently unobserved—possibly even in principle unobservable—feature of some “possible world” about which they nevertheless believe themselves possessed of objective knowledge. Xxx-usages will therefore range from the serious realm of the scientific counterfactual (Stephen Hawkings-type claims about what we would discover in black holes, for example) to the very serious realms of spiritualism and superstition… indeed to the many other hidden corners of nature and supernature about which most humans solemnly believe they have certain knowledge.

Unequivocal examples of this xxx kind of NL usage are harder to supply than Mia-usages, but it can be supposed that at least some NL speakers who say things like If your dead mother were alive today, she would want you to you marry Eliza or If we were having this tea-party at the center of the Earth, our tea-cups would weigh exactly nothing have in mind a kind of objective truth about the listener’s once-living mother and the self-cancelling effects of gravity at the centers of massive spheres that are impossible for us to observe directly but are evidently to be taken by us as independent of the imaginings of the speaker.

In other words, for users of the xxx operator. these conditional events are not just taking place in some mental place that has been furnished by the imagination of some speaker, as Mia-events implicitly are. They are events believed to be happening in, and so to be true of, some sort of “logical space” or “real though unobservable reality” the nature of which may be difficult to specify but is nevertheless believed to be “knowable” by speakers who use their subjunctives in this Xxx way.

So when logli do wish to speak subjuctively in this Xxx way in Loglan—and the NL subjunctive is often used to claim knowledge of exactly this Platonic kind (If one knew the good, one could not choose the bad… a famous Socratic dictum)—they will not want to use the new L-word mia to voice their claims! For the Xxx-claims they wish to make are are a different logical and ontological order altogether from the Mia-ones, and a different sort of Loglan “subjunctive operator” will therefore be required to express them.

To suggest the kinds of claims these Xxx-claims are, here are the three English claims mentioned above as they might be rendered in Loglan once it has its xxx:

Xxx nepo letu morto matma na clivi guo, mei danza lepo tu mercea la Elaizas.

In a logically possible world in which your dead mother is now alive, she wants the event of your marrying Eliza (to happen).

Xxx nepo mui speni levi tcati hapvei vi le midpea je la Ter guo, lemua tcati kupta ga skatio lio ni.

In a logically possible world in which we are (jointly) experiencing this tea party at the center of Earth, (each of) our tea cups would weigh exactly nothing.

(I am indebted for this splendidly counterintuitive example of an xxx-belief to Kirk Stattley.)

Raba goi xxx rapo ba spedja lo gudbi guo, ba no kau tisra nepo durza lo zavlo.

For any x, in any logical possible word in which x recognizes “the good”, x cannot to do “the bad”.

In sum, the Keugru believes that Loglan must provide for clear and succinct talk under both these sets of assumptions about “the nature of absent reality” whatever may be the distribution of Mia-type and Xxx-type subjunctive usages in everyday talk… indeed, whatever may be their future distribution in what one trusts will be—in some measure aided by Loglan—an increasingly sane world.

Third: The Keugru also wishes to remind logli that they have a third and very powerful alternative to mia and xxx whenever they have an impulse to “speak subjuctively”. (Presumably such impulses will issue, from time to time, from their hypothetically still NL-shaped minds, soi crano.) Logli may render nearly any such impulse in the Loglan indicative mood, and so use neither xxx nor mia to express its conditionality. Thus logli may use some other prepositional forms, such as na or vi or even nacevi (when and where), to express the conditionality of nearly any “subjunctive” claim. For example, suppose we are driving in France and our logli companion tells us:

Nacevi nepo tocu tcaro nuo snigoi guo, le ritco ga lilkai lo godzi lilfurlei vi la Frans.

When and where an event of two vehicles jointly approaching one another occurs, the one on the right “has” (is legally granted) the right-of-way (the “go right”) in France.

This is a conditional, alright, but it is devoid of subjunctivity. All we have to do is wait until its condition is realized and see if it’s true

This bold linguistic strategy puts the conditional remark in the indicative mood. It avoids altogether the philosophical problem of what kind of “reality” is presupposed by the absence of some specified condition. When someone speaks indicatively of real-word events in this way, the interested listener simply waits for the world to change, or, alternatively, reads historical accounts of such events in the past or predictive models of their possible occurrences in the future, and from either direct observations or from such accounts and forecasts, decides whether the claim made by the speaker is true or not.

This strategy won’t work, of course, with the coming-to-life of dead mothers, nor with tea-parties held at the center of the Earth. Nor will it work with pleasant ladies claiming that they wouldn’t have minded having their seats taken, nor with storied Irishmen “living until they’d been dead a month”. But it will work with claims about things that are only temporary absent… like vehicles from intersections, or tornadoes from towns, or chairs under lightbulbs. In other words, much of the “subjunctive world” our NLs have shaped our minds to “see” may actually be illusory, and so may easily be turned into an indicative world just by speaking about it in a plain, old present, past, and future ways. This option will, we think, have strong attractions for many logli.

In any case, the normal, unmarked form of the Loglan sentence—the form that is free of both mia and xxx—that we can now take to be expressive of its “indicative mood”, provides logli with a clear alternative to the subjunctive way of speaking. It is one that probably ought to be used by srisu logli whenever conditions that are merely absent, and neither in principle nor in fact unobservable, are afoot.

On the other hand, the habitual mia-user might say, Why not refer to an imaginative rearrangement of reality with mia when one has already made it in one’s own mind? Isn’t it only fair to the auditor to do so? Isn’t it, in fact, the more honest thing to do?

Whether such practices are “honest” or not, or “fair” or not, are questions of personal linguistic style and not for language designers to decide. They are, moreover—and like all human speech—deeply personal manifestations of the mind of the speaker or writer. What we logli actually choose to do in these matters will depend entirely on the kinds of minds we have, and whether we chose to disclose or hide them when we speak or write. If we are possessed of highly imaginative minds that continually invent “mental places” in which observable realities are ceaselessly rearranged, then we will probably use mia a lot. If we are believers in large domains of truth that lie beyond our human powers of observation but nevertheless have momentous effects on our lives, then we will probably use xxx a lot. And if we are observation-oriented empiricists who don’t hold much truck with either imagined or unobservable-but-knowable worlds, then we will probably use the indicative conditionals a lot.

* * *

The example about right-of-way above was provided by Alex Leith (APL) when he and I were traveling together by car in France last summer. We were approaching a Y-junction that happened to be empty of vehicles except our own, and I, having lived in France some thirty years earlier, asked Alex if the French still observed that invariant right-of-way rule.

APL answered in English, and happened to use the subjunctive to do so. ‘If there were a car coming towards us over there,’ APL said, nodding toward the empty road on our right, ‘it would have the right-of-way.’ I was immediately struck by APL’s use of the subjunctive in this situation and began thinking, once again, about what the English subjunctive “really means”.

The most interesting point about the example of the French right-of-way rule is that, linguistically, it is absolutely neutral. It can go off in any of our three directions. Any partisan of the “possible worlds” scenario for the NL subjunctive will be perfectly free to use whatever replaces xxx in Loglan to report his knowledge of this French law. (After all, laws or rules of any kind are in a very clear sense “Platonic ideals”; they are residents of that same vast domain where the rules of human games and mathematics are stored.) Similarly, those partial to the “mental place” view of the subjunctive may use mia to express their knowledge of this same French law. (After all, it was into APL’s mind, and nowhere else, that the vision of a car approaching from the right flashed and so allowed him imaginatively to remind himself of how French law worked, and so to answer JCB’s question… and, if he’d been speaking Loglan, he could have used mia to do so.) And other speakers of Loglan—those so partial to the indicative that they make only sparing use of the subjunctive if they use it at all—will use neither of these subjunctive operators but some operator like na, or vi, or nacevi—as we have just done above—to express this “same” conditional relationship indicatively, that is, as a fact about the observable world.

But once we logli begin to deploy these three ways of expressing conditionals in Loglan, we will discover what will be plain to all who listen, namely that it is not the same relationship we’re talking about in these three ways but three very different kinds of relationship. For three different kinds of things are claimed to be afoot when we talk in Loglan about the world in these three different ways.

This may not be true in English, but it will certainly be true in Loglan.

* * *

We can now close this long report on the subjunctive with a summary of the components of the Mia System. (For more details, please see my article in LN 96/2:3ff. It has five main components:

1. Mia itself, being a PA word, may be used grammatically in any of three positions: the inflecting, the adverbial, and the prepositional positions. But of these, only the prepositional positions generates meaningful sentences. For think what ?Da mia preda claims. Expanded, it means Da preda mia ti = X predas in “this imaginary world”, presumably the world of speech. But worlds of speech are not imaginary… unless the speech is taking place in a work of fiction! So unless the speaker is someone like Alice in Wonderland, a fictional character who chooses perversely—and with rare self-consciousness, we might add—to speak of her own and her world’s imaginary character, ?Da mia preda remains undefined. It is either bad usage, or whimsical to a most extraordinary degree. In short, mia will (nearly) always be used as a preposition.

2. The operand (“object of the preposition”) of mia is normally a LEPO-clause, although it may be any single argument. All operands of mia may be taken to designate some “imaginary world” or one of its identifying features. Thus in (i) Mi hapci mia lepo mi bragai = I am happy in the imaginary place where I am king, lepo mi bragai designates an identifying feature of the imaginary world in which s is happy, the one that s literally has in mind.

Sentence (i) can, of course, also be translated in the customary English subjunctive form in which the relation between cause and event is expressed by a logical conditional, the if… then… connection: I would be happy if I were king.

One can also use a minimal operand and say (ii) Mi hapci mia ba meaning I am happy in some imaginary world x, leaving the imaginary cause of one’s imagined happiness unspecified. Such “mia ba” expressions may always be translated into English by using its “missing conditions” form, in this case, by the mysterious I would be happy. But notice that, for the reason given in (1) above, it is almost never correct to speak the Loglan subjunctive inflectionally or adverbially, and in that sense “incompletely”. For neither ?Mi mia hapci nor ?Mi hapci mia means I would be happy as we English-speakers might think it should. Instead ?Mi hapci mia means exactly this: I am happy in this imaginary place. This might be written by a fiction-writer for one of w’s characters, perhaps for some Alice in her Wonderland. But logli who want actually to say in Loglan what I would be happy says in English must give mia an operand, and the minimal operand is, of course, an existential of the ba series.

Mi hapci mia ba, Mi mia ba hapci, and Mia ba mi hapci are the three ways this may be done. We anticipate that the phrase mia ba will eventually become a single word, miaba, and be pronounced [mee-AH-bah].

3. The eliminator of the mia operator is also its source word, namely minsia, a complex predicate derived from smina sitfa = mental place. Minsia is a four-place predicate and means all of the following:

F is an imaginary/conceivable/thinkable/“mental” but not necessarily possible world/place in which event/state S occurs/obtains and causes (“would cause” if its world existed) event/state V, this world and its causal laws having been in at least part imagined/invented by thinker K.

The elimination of mia from a “normal” mia-sentence—i.e., one with a prepositional mia in it of which the operand is a LEPO-clause and in which something or someone X figures in both the causal clause and the sentence asserting its effect—may therefore be accomplished by the following transformation:

Da preda (a sentence asserting the effect V) mia lepo da preda (a mia-clause specifying the cause S)

may be expanded into a sentence without mia as follows:

Ba (F) minsia lepo da preda guo (S) lepo da prede guo (V) be (K).

Two arguments (F and K) appear in the mia-less expansion that do not appear—are only implicit—in the mia-sentence. Done in English, this same transformation may be written as follows:

X is a preda (V) in an imaginary world where X is a prede (S).

may be transformed into:

Something x (F) is an imaginary world in which the event “X is a prede” (S) causes the event “X is a preda” (V) to take place, the world x having been invented by some thinker y (K).

Here is an example of such a mia-eliminating move:

Mi hapci mia lepo mi bragai.

I am happy in the imaginary world where I am king.

Ba minsia lepo mi bragai guo lepo mi hapci guo be.

Something x is a mental world in which the event of my being king causes the event of my being happy, and that world has been invented by some thinker y (probably me!).

The actor (K) who invents the world (F) is seldom at issue. He or she can normally be supposed to be the speaker/writer. But suppose one is saying something subjunctive about Sherlock Holmes, e.g., Holmes would in that case…. Now the subjunctive world in which the imaginary action is taking place has been invented, not by oneself but by by Sir Arthur Conan Doyle. In this case one might at the outset make the world-maker’s identity explicit by mentioning Arqr Konen Doil in the fourth place of the minsia predicate.

4. Of strictly secondary importance to the Mia System, but adding a second semantic dimension to it, was my proposal that the “probability grades” of the Mia subjunctive be expressed by siomia, daumia and biumia. (These are best pronounced /SIO-mia/, /da-U-mi/ and /BIU-mia/. Mia alone, however, is probably best pronounced disyllabically, that is, as either /MI-a/ or with level stress as /mi-a/, seldom as the monosyllabic /myah/. Yet -mia figures most elegantly in its compounds as a monosyllable.) These three compounds are obtained by prefixing each of the three probability operators sio dau biu in turn to mia; see Sec. 6 above. Thus

Mi hapci siomia lepo mi bragai.

I would certainly be happy if I were king.

Mi hapci daumia lepo mi bragai.

I should probably be happy if I were king.

Mi hapci biumia lepo mi bragai.

I might possibly be happy if I were king.

These compounds allow speakers to estimate “effect-likelihood”, i.e., the probability of the imagined effect’s occurring given the prior occurrence of its imagined cause in the imaginary world. What this second dimension of the Mia model tells us is that even the imaginary effects of imagined causes are not always imagined to be certain. I might be happy, says Prince Charles wistfully, who really doesn’t know what kinghood has in store for him. A similarly circumspect Loglandian prince might say Mi hapci biumia ba.

5. Of tertiary importance in the Mia System’s is its third dimension: “world-likelihood”. World-likelihood is quite a different matter from effect-likelihood. For the latter, compound operations on the Mia operator itself can be performed. But the likelihood of the imaginary world’s occurring—what is the probability that Charles ever will be King?—can only be assessed in a separate sentence, one attached to the main subjunctive sentence by ice or icebuo. (In my original LN 96/2 article I recommended using ize for this connection; but I believe now that this was a mistake. The world-likelihood claim may safely be viewed as independent from the subjunctive claim.)

Using letter-variables to refer to the worlds whose probabilities are being assessed, and the three probability predicates sirto, dakli, and blicu to assess them, we can give these three examples:

Mi no hapci daumia lepo mi bragai, icebuo bei no blicu.

I should probably not be happy if I were king, however that (the event of my being king) is not possible.

Mi nu korji lemi morto matma lepo mercea la Elaizys siomia lepo mei livcea, icebuo cei no mutce dakli, soi crano.

I would certainly be ordered by my dead mother to marry Eliza if she (my mother) came to life, however that’s not very likely (smile).

[Livcea once meant adapt in the sense of the biological adaption of a species; it was made from the pre-1975 metaphor life-change. But livcea so obviously—and so more strongly now—suggests alive-become that I changed its LOD entry to mean come to life, and have added new words for the technical senses of genome, gene-pool, and adapt in its several biological senses. Is one allowed to do this? Yes. In fact one is obligated to do so whenever one sees an incompleteness or an ugliness in our dictionary. This is how we constantly improve our lexicon and the dictionary that tells us about it. It is one of the most powerful features of LOD that it allows its users to do this.—JCB]

La Sorpoi Tcariz mutce hapci biumia lepo Tai bragai, ice bei nurmue dakli.

(The) Prince (Six-Rank Power) Charles might possibly be extremely happy if he were king, and that’s moderately likely (to happen).

It seems awkward to us, as English-speakers, that the event-likelihood operator attaches to the preposition, not to the predicate of the sentence that asserts the event. But I expect we can get used to that. The way to think about this is to remember that the mia-clause is a modifier of the main predicate. It tells us, in his case, that the speaker believes that the kind of “extreme happiness” Charles will experience as king is, alas, only possible.

The rest of the Mia system seems quite natural.