(From Lognet 96/3. Used with the permission of The Loglan Institute, Inc.)
Letters policy: Unless otherwise stated, letters addressed to logli in general, to The Institute, JCB, or any editor of Lognet will be considered as offered for publication. But it would be good if the writer explicitly offers. We reserve the right to edit letters, mostly just to drop material that has to do with ordering books, etc. If you are writing us by paper mail and your letter is a long one, we’d be grateful if you’d enclose a soft copy on a diskette. We can read most word-processors and having your letters on disk saves us a lot of typing.
This first letter is from Roy Bigelow, a frequent contributor to this column:
Thank you for the [second] copy of LN 96/1 [Roy’s first one had been lost—JCB]. However, the gadfly can’t help asking some questions about the articles concerning the subjunctive.
First, I don’t understand the reason for the great amount of discussion. I believe that the only thing required for consistency with symbolic logic is a distinctive marker. I thought that the attitudinal eu fulfilled that role, and I note that LOD [the Loglan Online Dictionary] defines it to be the sign of the subjunctive. [Indeed LOD does, and until the Keugru changes or supplements eu, that is what eu is.]
Second, why conjure up imaginary worlds? Perhaps they are needed because Mr. [Mitchell] and Mr. Jennings, just like the article states, simply want a more direct translation of the English language words should and would. If so, would they be satisfied with the replacement of attitudinal eu with the attitudinals mia, siomia, daumia and biumia?
Third, am I in error? Is it not true that the English I would if I could now correctly translates into the Loglan Mi eu selhanduo, icanoi mi eu kanmo where the word eu may appear anywhere in [each of?] the two propositions?
Roy V. Bigelow
I think the biggest cause of the “great amount of discussion” we’ve had about the subjunctive—and it is has been great, having lasted more than two years, now—is the promise we have all of us seen in the “Kripke conditional”, a notational schema that allows us to sidestep, if not actually to defeat, the known formal inadequacy of conditional connectives—the if...then connectives so oddly called the “material conditional” in the jargon of logicians—for dealing with contrary-to-fact conditionals. That formal inadequacy is this: since any statement claiming that a contrary-to-fact condition is present is bound to be false, then any if...then claim relating that condition to any consequent whatever is bound to be true. For if p is false or q true, then p ⊃ q is true...thus trivializing the problem of how to express a contrary-to-fact conditional in a logically manipulable language. For example, if someone X who isn’t king claims that if X were king, then X would be happy, and if we interpret this claim as a material conditional, then we have emptied it of all sensible meaning. For so interpreted, such a conditional will always be true...even if spoken by someone who would hate being king!
To get out of this impasse—this trivialization of counterfactual conditionals by standard logic—Kripke logic turns the contrary-to-fact condition—kingship, in this case—into a prepositional feature of a declarative sentence, a sentence whose main clause will now assert the consequent...in this case, happiness. So the Kripke conditional comes out something like this: When and where X is king, X is/was/will be happy. The condition X is king is not satisfied in the here-and-now for any non-king, but it may have been or may eventually be satisfied. It thus makes a non-trivial claim. We can examine history or wait to look at the future. The additional twist that Kripke logic puts on this grammatical solution to the puzzle of counterfactuals—a solution that is grammatically available in all I-E languages, by the way—is to replace the indicative operator when-and-where with a special subjunctive one that says something like this: In any logically possible world in which X is king, X will be happy. All this is excellently explained by James Jennings in his paper in LN 96/1.
The reason it has taken us so long, in the Keugru, to adopt this apparently simple and loglandical solution to the problem of the counterfactual conditional, is that it is, in fact, neither simple nor obviously loglandical! It has taken us several years to make sense of the many uses of subjunctives in natural languages, and to find clear ways of separating these usages in Loglan. What, for example, does the natural language speaker mean when he uses the subjunctive mood? We’ve tentatively decided that he means one thing when he is an Irish comedian saying If Casey had lived ‘til next Thursday, he’d have been dead a month, and quite a different thing when he’s a driver looking at an empty road in France and saying If there’d been a car coming toward us on that road to the right, its driver would have had the right-of-way. To give just one small hint of our solution to these problems, we’ve decided that one of these sentences should take one sort of “subjunctive operator” in Loglan—probably Mia—and the other, another CVV-form operator not yet finally settled on.
In short, it is this very large can filled with these very energetically wriggling linguistic worms that has kept us busy trying to separate them into reasonably harmonious loglandical tribes.
I asked Alex Leith to comment on our somewhat irregular use of the word ‘subjunctive’ in these discussions of ours on the countrary-to-fact conditional, and this is what he wrote:-
In all the recent discussion of how to render the “subjunctive” in Loglan, referring to hypothetical claims like the classic “I would be happy if I were king”, I feel we may be using the wrong term.
English grammar uses terms borrowed mostly from Latin grammar—and they don’t always fit very well.
In Latin and the Romance languages it’s quite clear: the different tenses of the subjunctive mood are sets of inflections which are different from those of the indicative mood, and which must be used after certain conjunctions and verbal constructions. I’ve forgotten nearly all my school Latin, but do remember is that “ut” with the subjunctive means “in order that”.
In French the subjunctive is used in a number of verbal and adjectival constructions, of emotion, of wishing, of necessity, doubt, and probability. Although there is sometimes an element of counterfactuality in these, this isn’t what determines the use of the subjunctive. In the translation of our classic example “Je serais heureux si j’étais roi” the verb forms used are not subjunctive at all, but conditional “serais” and imperfect indicative “étais”.
The point is that “subjunctive” refers to how a language expresses certain concepts (by a different set of verb inflections), and not to the nature of those concepts. In Loglan we have no inflections of the verb, so we can’t have a real subjunctive. All the schemes proposed (mia, sio/dau/biu, and foi/fio) use operators to mark claims about states of affairs which are (temporarily or permanently) contrary to fact. What shall we call them?
There are two kinds of ‘if’ in Loglan. One is the logical ‘anoi’ which JCB has called ‘reverse implication’. I propose we call the other kind ‘hypotheticals’, which better describes the kind of claim that is being made, without referring to the mechanism of making such a claim.
Is this nit-picking? Perhaps...and if this use of ‘subjunctive’ were already firmly established, I could live with it. But if it’s confusing for English to borrow Latin grammatical terms, how much more so for Loglan, whose grammar is clearly defined in terms of its lexemes.
I think it might now be almost impossible to stop calling mia, for example, the “Loglan subjunctive”. Among the alternatives, the word conditional has been pre-empted by logicians to denote if...then... and ...if... linguistic connections, and Alex’s choice, hypothetical, is so much more clearly applied to the Let us suppose that... moves of mathematics and science. For these moves we already have the word eu, and I’d hate to stretch its usage over something so different as the imaginative counterfactual. Anyway, the wildly imaginative counterfactuals that are not only physically but logically impossible, such as If Casey had lived to next Thursday, he’d’ve been dead a month, are not really hypotheses in any of the usual senses of that well-defined word. They specify imaginary event-sequences created in individual speakers’ minds, and range from startling jokes to earnest plans. (If you’d bring me a chair, I could change this lightbulb; If there were a tornado, I’d seek shelter in my cellar; etc.) To me, after two years of spotting the tell-tale signs of them in English prose—signs that are especially rich in novels and biographies—and listening to them in English speech, this is now what the word subjunctive means. I shall take Alex’s remarks as an important cautionary note that the word also means something slightly different in Romance grammar...but something that is nevertheless not unrelated to the planning and intention-conveying functions—all imagination-using ones—of ordinary speech.
The next letter, a brief but fascinating note from Chris Hogan, was taken from Chris’s response to one of the questions in our June Questionnaire:
I am an APL-er and love the clean parallels between the maths notation of APL and the language structures of Loglan. I hope for a larger [Loglan] community.
I think one may be on the way, Chris. Here’s another interesting letter that accompanied a member’s response to that questionnaire:
I continue to be very interested in Loglan, its basis and hypothesis, and its ongoing development. Perhaps the following will explain why. Throughout my career [as a professor of music] I have read and speculated about how the following engender meanings in their various ways: 1) speech utterances (which precede graphic (re)presentations, I think); 2) visual and auditory signs and symbols (red traffic lights, beeper-sounds, ideographs, letters of alphabets); [and] 3) visual and auditory designs (maps, paintings, musical compositions). I hope you can continue the Loglan work for many more years.
To express his support of our ongoing work on this fascinating little language—and it is still not very big!—Gandias Kok is one of those who took the opportunity our Questionnaire offered to boost his membership status from Regular to Patron. I’m glad we gave everybody a chance to make those sorts of adjustment through our 1996 Questionnaire. Many took this opportunity to boost their support!
Talking about support, here’s some of the intellectual variety for Jeremy Dunn’s suggestion in LN 96/2, by one of our logicians:
I read the letter by Jeremy Dunn (LN 96/2:21) on antonyms and, basically, agree with him. This idea is a little like set theory, which I like. And since I have grown used to it, I have found the basic idea of set theory works well in English, too. It is very convenient to be able to speak of an x and a non-x. Antonyms are like this; hardly a day goes by that I am not searching for an English antonym, but we have no nifty way to make them. (Sometimes we make not [or non-] do the job.) My instinct tells me that this would be a good way to enhance any language, especially a logical one testing the Sapir-Whorf hypothesis.
There are some problems with Jeremy’s plan, but rather than discuss them here in a necessarily superficial fashion, I’d suggest that you, Jerome, work with Jeremy in shaping up a formal proposal you could then jointly make to the Loglan Keugru (the “Loglan Caretaking-Group”, or Academy). Then you can be sure it will be carefully considered.
Here’s a completely different topic that came up in the Keugru (Academy) recently and because it cuts pretty close to the ontological bone, we decided we had to have your input on it, Logl (that’s pronounced [LOH-gll]). As usual, Kirk Sattley is our Keugru spokesperson. (He happens also to be a spokesman, soi crano, but that’s just a another by-product of the sad fact that Loglan doesn’t attract many women, uu, so L has tended to remain a masculine enterprise. You women subscriber/receivers of LN and LL—and I know we have some—ought to change that. Please give us your perspectives on some of these ontological/metaphysical/philosophical/linguistical/sociological issues. Lognet’s Lo Lerci is open to everyone: djori and nordjori, mendi and femdi. Here’s an especially lively issue to give us your input on:
Hoi Logli, fie:
Reading some articles recently on Sumerian history, language, and religion, it occurred to me that, if one were to try to write such texts in Loglan, it would be extremely useful to extend the gender-inflexion scheme to the word gandi = (is a) god (of). This would amount to our saying that, linguistically at least, divine entities constitute a species.
Thus we would have:
gando = (male) god
ganda = goddess
gandi = godling
gande = god-like
gandu = deity (age and gender unknown or irrelevant)
Note that this discussion concerns only the common noun god. The proper name God remains, of course, La Gan. Godling sounds frivolous, but I recall seeing the word at least twice, once in a non-disrespectful story about the investiture of a young Tibetan boy as a high lama, and once in a story of the family life of the Olympian gods.
The kejgrudjo, invidually, are willing to include these terms in the Dictionary, but there was some [collective] feeling that this was a sensitive issue, and should not be legislated without asking the opinions of the logli.
The gender declension is [now] applied only to ... predicates[, primitive or borrowed,] that denote biological species, and I believe we should keep that rule. Even though men and women sometimes speak in jest of the other sex as if its members were of a different species, I argue that we should not accept mrenu (man) and *fumnu (woman) as species terms in the official dictionary. As Loglan becomes a living language, no doubt someone will make jokes using nonce-declensions of such words, but let’s not encourage it ahead of time.
There are a few other tempting quasi-species [predicates]: take for example kicmu, which already ends in -u, the [age-and-]gender-unstated form. Would there be much value in having primitives for specifically male (?kicmo) and female (?kicma) physicians? Again, I suggest not: in the rare event that we need to specify the sex of a doctor, the extra length of menkicmu and femkicmu [which would soon be shortened to menkiu and femkiu—JCB] isn’t inappropriate.
So what do we logli think? Are there problems with including ganda/-e/-i/-o/-u in the Dictionary? Are there any other metaphorical or literary species that deserve the dignity of a declinable Loglan [predicate]?
Looking forward to any comments,
This letter from Brian Kimerer was also sent us with his questionnaire.
Hello from a “silent partner”. I have noticed recently that my dues are overdue, and I thought that I would include with them a brief note to express some thoughts about your marvelous language as well as some of the reasons for the silence. Perhaps at times you wonder why some of your members are less active in Loglan than your core supporters. Well, here goes.
The primary reason for my lack of involvement is, of course, time. With a full-time job and two children in school there are no minutes left in a day once the chores are done. Enough said about that.
However, there are other reasons, one of which is the difficulty with which I am picking up the language. I have read L1 one-and-a-half times. As I read it, the details make sense. However my grasp of the language in a general sense is fugitive. Once the details are all understood, I still do not seem to know how to say ‘Good morning’ in Loglan. I believe that this is a side-effect of the language being as logical as it is.
Here is a description of the difficulty. I am a software engineer by trade, and have been programming computers since 1968. When I attempt to create a sentence in Loglan, I fall into a “programming” mental state. By that, I mean that my instincts make me attempt to get all of the details absolutely, positively correct in order to create a “bug free” statement. As you well know, a comma out of place in a computer program can be the difference between a working program and binary trash. This feels a lot like work... and it happens very slowly in my head. After nearly 30 years of programming, I must still think very carefully and slowly in order to make my programs work. It takes many passes through a computer program before I am confident that the bugs are gone. Even an intuitive feel for the language I am using does not speed up this process very much.
My observation is that the precise nature of Loglan causes me to work out complex, correct statements at a snail’s pace before expressing an idea (dictionary firmly in hand). In this situation, the value of Loglan as a conversational tool is diminished. This might not be a problem for a person who naturally thinks in a more orderly fashion (I knew a fellow once who spoke Fortran...true story!).
The situation is different with the natural languages, which are based more upon memory and less upon logic. If I utter a sentence in a natural language with a confusing ambiguity, it is not a big deal. The confusion can generally be cleared up in the ensuing conversation with examples and restatements of the proposition. The effort required to put forth an idea is more relaxed and less concerned with correctness. When conversing, I tend to take the path of least resistance.
All of this is not meant to disparage Loglan. I believe that it has an important place in the future. One major difference between programming languages and natural languages is the set of ideas which each is optimized to express. The natural languages are filled with words and constructs which are concerned with human experiences (weather, love, birth, death, food...). These are things which a computer can never experience. The programming languages are largely limited to concepts which exist in the context of computing (arithmetic, data storage,...). I have thought of Loglan as a sort of bridge between the two types of language. It is an unambiguous language (for the computers) which is capable of discussing human issues (for the humans). Its use in verbally directing intelligent software would seem a natural application.
Of course, the expressed purpose of this logical language is to force us to think in a more logical fashion. As I understand it, this is central to the Whorfian hypothesis. My personal prediction of the Whorfian effect is that it will turn out to be a filter. Basically, the individuals who do not think in an organized and detailed fashion will not use a logical language. People who do think in a very organized fashion will find Loglan natural and easy to use. I have seen this happen with programming languages. I know many people who cannot handle the tedium of making programs work. They try it for a while and then move on. The results do not seem worth the effort to them.
In any case, when viewed as a grand experiment, Loglan is an interesting effort. I am still interested in trying again to learn the language “off-line”. Because of a chaotic schedule, I cannot sign up for scheduled conferences or classes. I fear that I would simply miss them. However, I would like to see what happens with L3. Perhaps that could be the missing link in learning Loglan. I have included my response to your survey (sorry it is so late). Please sign me up to receive La Logli.
Brian S. Kimerer
I asked our tiftua (“offering workers”, volunteers) to comment on Brian’s thoughtful letter, and Alex Leith and Bill Gober did so:
Hoi Djim, fie:
Re the letter from Brian Kimerer, I’m not sure if I’ve mentioned this before on e-mail, or to you in person, when the discussion has been of “Is Loglan difficult?” But it had struck me that whereas in learning Italian or Inuit we try to say something, and never mind if it’s not quite right...and our Italian and Inuit friends clap and laugh and say ‘Bravo’.
However, with Loglan we seem to be trying from the outset to get it absolutely right. and if we don’t, then LIP or RAM or JJ or someone [will] tell us in no uncertain terms that we got it wrong, and worse, that we succeeded in saying something quite different from what we’d intended.
If Loglan is hard, it’s not because of the language (except in certain areas like quantification and connectives perhaps, balanced by the extreme simplicity of phonology and word forms) but because of Level of Aspiration.
In any event I’m sure L3 will help a lot.
Hoi Gandias Braon, fie,
[Re] Brian’s letter, I have to agree. A couple of years ago, I had a long chat with Steve Rice. Among other things (I can hardly believe the number of artificial languages he has materials for; e.g., both versions of Volapuek (I never knew there was more than one!)!) we talked about the current preferred method for writing Loglan: at the computer, using LIP [and LOD] to check the syntax [and the diction] of every utterance. [This is] the same way we write computer programs. It may be hard to turn such a language into a vernacular.
Hue Bil Gober
Naturally I hope Alex is right, and he certainly has a point: L, like computer languages, punishes faulty efforts rather severely. This is, I believe, what Bill is saying, too, namely that by having been punished for saying things we didn’t mean to say, we now take our L composition efforts very seriously. I believe we shouldn’t. I believe we should try to get out of this “perfectionist” mode and simply develop better and better speechways by speaking the language, but with people who happen to know it better than we do. That’s exactly what we do when we learn French by going to France, or Innuit by going to Baffin Island. But to do this we’ve got to actually create a speech community; for to learn any language we need the help of other speakers! At very least, we need “leading learners”...people who can sit back and listen to us, and murmur things like Io tu sanfurmoi li leu mrenu lu (Probably you intended to say ‘the set of men’) from time to time, and thus gradually help us shape up our speech into that of brana logli.
There is, I think, no substitute for a human role-model in learning a human language...whether it is a first or a sutori one doesn’t seem to matter. That’s the boot-strap we’ve got to hoist ourselves up by, Logl! The bootstrap provided by teachers and leading learners. Alex, Reed Riner, and I have some ideas about creating both them and access to them, ideas about which you’ll be hearing shortly. —Hue JCB