(From Lognet 91/1. Used with the permission of The Loglan Institute, Inc.)

Lo Lerci (Letters)

Letters policy: Unless otherwise stated, letters sent to The Institute, JCB, or any editor 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. Sometimes a given correspondent will have several letters in the hopper, so to speak, and we will combine them into one purely for the sake of clarity.  We’d be most grateful if you’d enclose a soft copy of your letter on a disk. We can handle most word-processors and having your letters on disk will save us a lot of typing.—Tisra

Dear Editor:

 On the cover of Lognet 90/2 the sign should say (if I’ve got it right) Mia takna la loglan. I hope that’s right; my books are still packed.

 Thank you,            James Jennings 

Mia takna la Loglan is almost right, but the predicate is wrong; it says 'We talked to Loglan'. You need to replace takna with lentaa, derived from lengu takna ('language-talk'), another sense of 'speak'.—JCB

 in the sense of 'speak a language'. Hence, Mia lentaa la Loglan. Why not mua ('I and you and others')? Because we cannot be sure the reader (tu) also speaks Loglan! (He may just be able to read it.) —JCB

Dear Editor:

....The English word 'computer' [see LN90/1:7] does not convey the meaning of what a computer is. Until this century, the word 'computer' meant either a person or a machine which calculated numbers; in other words it referred to a person constructing log tables or a mechanical calculating machine. Because when computers were conceived, their first applications involved numerical calculations, it seemed reasonable to call them “computers”. However, computers only deal with numbers. We have decided to interpret an electronic pattern  within the hardware of conventional computers as a binary number, which may pretend to be any kind of number (integers, real, complex, etc.), in any base, due to the mathematical properties of binary numbers. But those electronic signals in a computer may be interpreted as being things other than numbers. They could be letters in an alphabet, pictures, or more complex constructs. Computers are capable of simulating any real world object. Borrowing the word 'computer' or words which meant “numerical calculating” for Loglan would simply compound an initial mistake.

 I think the word which meant 'electronic brain' would be wonderful...for a word, that is, which means 'electronic brain'. Computers need not be electronic; mechanical computers are possible (made from gears and levers or from Tinkertoys) as well as completely metaphysical ones (Turing machines and similar automata which are mathematical models for computers).

In The Art of Computer Programming, Donald E. Knuth gives a definition for the word computer as “a data processor”. Not all countries have borrowed the word 'computer' from us. In the Netherlands, I believe that the word for computer science is something like 'informatiks' and I think something similar to this would be appropriate for Loglan. 

 Thank you.

 Sincerely,            Fred Schiff

Good thinking, I agree that ‘electronic brain’ is unnecessarily restrictive, especially with optical computing in the news. And there is a mechanical computer in the Computer Museum that plays Tic Tac Toe (Naughts and Crosses) and is made from Tinkertoys and duct tape! —Tisra

Dear Editor:

The recent flap between Rex May and JCB on methods of finding and creating predicate words has become what I feel is a nitpicking type of discussion. I do not have the horrible problems accepting the words as they are that Rex seems to. Recognizability aside, a foreign language is a foreign language and the particular words in that language will always seem odd and unnatural no matter what method one uses to create them. Rex might as well write to the government of France and complain about their word forms. Before I even started reading about this point of dispute I was impressed with how easy most of the words seemed to be to remember. JCB gets points from me on this one. Rex finds the 5-letter form too restrictive. It’s supposed to be! A logical language is not concerned with your aesthetic sense of how words are to be spelled out but only with their rules of construction. 

 Now before Rex gets too upset with me I will take a few jabs at JCB too. The argument between JCB and Rex on this issue is not that the predicates should be recognizable but the manner in which they are created to be recognizable. This is an important point. I find this issue of recognizability a little curious in some ways. The grammar of the language would miserably fail a recognizability standard and JCB didn’t seem to be too flustered about that. Why then this preoccupation with having the predicates so recognizable? To make it easier to learn the predicates, you say. Well the language would be even easier to learn if it was English to begin with. [For whom? For English-speaking learners, of course. But for a typical group of Earth's inhabitants? No.—JCB] In Loglan 1 JCB gives an estimate for the number of predicate words that are possible but misses the point that his recognizability process restricts the number of words that are available to be used. Given that there [are] now a large [number] of predicates in the Loglan dictionary that cover the vast majority of commonly used predicates I think that it is now appropriate to drop the recognizability process completely and simply pick unused letter combinations as needed. I don’t think that I am the only one to try word creating that has found that the optimum form has already been taken by some other predicate. By the time you get to the third best choice your word has become so unrecognizable that you might as well have chosen any permissible letter combination. [I feel that these discussions, such as between JCB and Rex, are actually much more than nitpicking. We are privileged to be present during the creation of a new language and are actually viewing the processes that, in nature, occur over centuries. Since few (none?) of the French who helped to create modern French are still alive this manner of discussion is not possible. Here we have a modern language created entirely within the lifetime of a still-living person and, so, are privileged to be present at this kind of discussion. I find it exciting and informative.

 Predicate recognizability is not the simple thing that Mr. Dunn thought it was. I may be forgiven for assuming that Mr. Dunn was only working within an English context in attempting to create predicates; imagine the same attempt with the same constraints that JCB imposed — eight languages with maximum recognizability in all eight! —Tisra]

 Rex’s choice of number names is silly in comparison to Loglan. In Loglan one can identify odd and even numbers by the letter they end with (with the exception of zero... rats!) why frustrate math teachers with a system no better then English? [Also, Loglan happens to embody a mathematical concept very nicely in the naming of the numbers. Most numbers are either odd or even; any given number may be positive or negative — except zero, which is unique. Therefore, the unique handling of its name is a constant reminder to us that what we are treating as any ordinary number, really is quite extra-ordinary and not all that much like the other numbers. —Tisra]

 You issued a call for comments on Loglan 1 and so here I am! My job is that of an electrical design CAD operator and thus can be considered a non-specialist in languages. I found the book fascinating, truly a unique and carefully thought out effort. There is definitely the need for Loglan 0. Loglan 1 would be too abstract for the average person to read without the help of a teacher to clarify the myriad questions that are sure to occur. The book is excellent for the individual who is well read and has some motivation to understand the ideas being presented...myself being a case in point.

 My only criticism is that I felt the time tenses were passed over too quickly and the words relating items positions in space also. [Time tenses too quick and the spatial positioning too far out??? (Sorry, but not very...)—Tisra][But he's right.—JCB]

 In reading the section on experimental uses of Loglan, I thought of a possibility that I haven’t seen mentioned. It is my hypothesis that a group of mentally retarded children who are taught Loglan from the beginning (as their natural language) will be able to think and speak more complex thoughts than retarded children who are taught English. My reasoning is analogous to the situation where a computer of lower performance can sometimes outperform much larger and faster computers if they have superior software in them — Loglan represents better software that occupies less space on the hard drive (the brain). What limits are there to the thought processes when the brain does not have to remember thousands of lingual rules? [This is an interesting thought-problem. Controls, similarity of handicapping, measurability of results, measurability of complexity of thoughts...I can’t begin to list all of the variables that would be involved in the scientific measurement of such an endeavour. The tragedy of mental retardation is usually unique to each person afflicted, the brain is such a complex mechanism that no two retardations are the same, thus invalidating any comparison. The basic concept, that of being better able to conceptualize, possibly will prove to be true in an "each in his own" manner, retarded or not. —Tisra]

 I think that Loglan should devote considerable push in the field of artificial intelligence and man/machine interface. Loglan is the ideal language for this purpose. Loglan could be the verbal equivalent of machine code. All computers should have a standard language that is common to all equipment in the world, it will not do to have some that can speak [only] English and some [only] Japanese, they should all have a common core language that can be used for communication between all equipment. This problem is ideally suited for the advancement of a universal artificial language. I feel that this is the only real arena to successfully advance any artificial language. Most people will not learn a language unless they see a practical advantage to doing so. They may be able to speak to a machine in English but how much nicer to learn Loglan and be able to speak to any device in the world! No machine in the near future could possibly be able to have translation programs between every combination of two languages, it would take up too much memory. How much easier for each machine to translate the natural language into Loglan. Perhaps a half and half approach could be a beginning wherein the individual could use Loglan grammar and his own predicate words and then later learn the rest. [Loglan grammar and English (French, German...) predicates?? I can’t say that I can even envision the concept. All these other languages persist in having verbs and nouns and adjectives and such instead of predicates! Attempting to plug nouns and verbs into Loglan grammatical structure brings to mind the round hole and the square peg. —Tisra][Still, this could be a practical application of our hard-won grammar.—JCB]

 It might also be interesting to advance the idea of Loglan as a communication language in projects involving the search for life on other planets. Why give aliens a hodgepodge like English to figure out? [SETI (Search for Extra-Terrestrial Intelligence) has been proceeding for several years, but the intentional broadcast of messages, as I understand it, was tried several years ago for a short time and has been discontinued in favor of listening for some time. It seems that it costs less to listen than it does to talk. Since we are listening, I guess you could say that we are listening in all languages and Loglan has as much chance as any other of being detected on the incoming message stream. —Tisra]

 I read a news item that unfortunately was about a year old. The item involved the channel tunnel project in Europe and how they were considering developing a language for the police on both ends. Is there still the possibility that The Loglan Institute could come to the rescue?

 You might consider the possibility of publishing future editions of Loglan 1 in a ring binder format so that new developments and corrections could be added when they come out. [This is what we plan for Loglan 6.—JCB]

 It seems that since Loglan is unambiguous it is possible to type a continual string of it without paragraph indentations. I find this a little disconcerting, I like to have occasional breaks in the visual field to keep from losing track of where I am on the page and ideas separated into chunks so that I can skim and select salient comments. A solid chunk of text is just a little too massive in appearance for my taste. [I agree that one solid chunk is too much. ‘Unambiguous’ says nothing about paragraph format! A paragraph, in any language, is still a device for grouping chunks of ideas together. The ideas should flow from one to the next logically and each paragraph should explore only one logical concept. An opening sentence, development of the concept and a closing sentence...a paper in miniature. Try reading the first and last sentences of each paragraph in a paper sometime to see just how the ideas (at least, in a well written paper) still flow. The only missing element should be the supporting development of each idea. —Tisra]  

 Jeremy Dunn

Dear Editor:

....[Here are] a couple of things I was wondering about:  [1] How would you translate 'when and if' as in 'I would go when and if you do'? [2] What in the world would we call the oceans? (Lots of variant names, depending on what language we started with? Loglan predicate names? Or something else?) Also, [3] it seems to me that we must have more than one name for places: what the residents call it, what the non-resident same language-speakers call it, etc. Sometimes the residents call it by various (similar) names. Or do we just go by what the residents call it? (This could be expensive to determine for each town, county, state, mountain...)

 Oh well, that’s about it for now. Kerju tu,

 Michael Demoulin

Re Michael’s question (1) about the Loglan for ‘when and if’: on the model of ifa (‘and later...'; see NB3:82-83) I’d say it was icanoicena and an eeshek, i.e. a member of the I Lexeme. I suppose you could reverse the elements and make a compound that was a member of the PA Lexeme, e.g., ?nacecanoi; but I’d be dubious about that one. It seems to me the connective should take grammatical priority over the PA-term and that these things in English are all grammatical variants of 'if'. Thus ‘when and if' and ‘if and when’ are semantically indistinguishable in my idiolect. But I suspect that, for a variety of reasons, in Loglan the logical element (icanoi, canoi, anoi, kanoi) should come first, yielding icanoicena, canoicena, anoicena and kanoicena as the four contextual varieties of Loglan ‘when and if'... any of which could also be translated ‘if and when'.

 (2) The several oceans: a naming problem with several interesting dimensions [and a lot of depth... —Tisra]. Rather than try to solve it offhand, I’m going to suggest that Michael take the problem up with the Purmao Diigru through their new Cerpeu, Stephen Rice, i.e., ask him or them for guidance. If policy issues get involved, the Diigru can bring them up with the Keugru.

 (3) The place-name problem. Ditto as for (2). I’d go to the Purmao Diigru for guidance. 

 (4) Finally, Michael scribbled Lo zvospa ja le rari zvolai on the outside of the envelope. Ok, I’ll take the challenge. I suspect that he meant to translate ‘Outer space (or just Space), the last frontier!' But I don’t think zvolai (zvoto landi) gets the sense of ‘frontier’ adequately.  Zvolai says to me an outland, an outside land, an exterior land, and this does not suggest to me what ‘frontier’ means. Instead, it is whatever is beyond the limits of a country, thus limbanlai or simply  limbae (from limji bande landi and limji bande, respectively, where bande is a new C-Prim meaning 'is beyond...on route...by interval...'; the etymology of this and other recently added words will be given in Steve Rice's column next time). And that, I now see, is the improvement on the original metaphor that Djori Demoulin probably intended: ‘the last beyond-frontiers’! Lo mutgudbi! Excellent!—JCB

[We may have met the only living American who doesn’t recognize, at a glance, the famous quote from Star Trek: 'Space, the Final Frontier!'.—Tisra]

Dear Editor:

 I must admit that I was somewhat surprised to find out that the membership of the Loglan Institute is as small as it is. I would like to increase your members by one and include myself in your elite group.

 A little bit of information about me: 

 My profession is computer software. I have been creating graphics oriented human-computer interfaces since 1981. However, the expression of some ideas using a point-and-click, icon based interface is very difficult. For example, try to design an icon for the function “Undo”. Hence the existence of a universal human/computer language is interesting to me for filling in some of those holes.

 I am enjoying L1. However, I must admit that it is a slow read. There are a lot of rules to get through. Learning any new language requires a lot of memory work and repetition. But the book seems to be very well written and professionally produced.

 Good luck in your membership campaign.


      Brian S. Kimerer

Welcome to the elite! Do you have the software to assist the learning process? I’d strongly recommend it.—Tisra

Dear Editor:

I am writing to you from Belgium. I am just getting set up here, so I do not yet know how to get access to word-processing or electronic mail. Do you have an e-mail address? [It's on the back cover. —Tisra]

 I am quite impressed with the new Loglan book. It seems unfortunate that a major revolution in vocabulary was necessary, but I think that the principles which guided the revolution were correct.

 I was itching to tell you 10 years ago that your statement that there was no way to establish unambiguity of Loglan was incorrect! As you now doubtless know, there is no way to establish the unambiguity of a context-free grammar in a uniform way, but it is possible to decide the unambiguity (or ambiguity) for grammars in restricted classes.

 Is it possible to obtain a complete description of the grammar as it now stands? [Yes; and Djori Holmes now has one. All that was necessary was to enter into a Trade Secrets agreement with us.; see LN90/1:16—JCB] As a mathematician, I prefer to have a complete description of a formal system I am looking at! One specific question I have is this: how does one punctuate the metaphor (po kukra sucmi) ditca, 'teacher of the event of fast swimming'?

 I conjecture ?po cui kukra sucmi cue ditca. [No, this won’t parse. —JCB] (or maybe ?po ge kukra sucmi cue ditca? [Yes, this is it. —JCB]) but I cannot tell from what you have given in Loglan 1. Also, I think that it would be useful to point out explicitly that (predicate je argument jue argument) forms can be used not only in descriptions but in predicate strings. You do mention this, but only “by the way” when discussing punctuation; it should probably be mentioned at the same time that you discuss inversion with go. [Good idea.—JCB] Continuing the example above, I am fairly certain that ?po kukra ci sucmi ditca says what I want to say [No; currently Da po kukra ci sucmi ditca parses as da ([{po kukra} ci sucmi] ditca), but obviously the grammar could be changed to make it parse as ?da ([po {kukra ci sucmi}] ditca), which would convey your intention. I suggest you study the matter, decide how it should parse, and convey the argument that persuaded you to us, the Keugru; and I expect we'll be persuaded too. This is the sort of grammatical change we make for good reasons all the time.—JCB], but a form with explicit parentheses would seem to be necessary for more complex expressions of this kind, and I suspect that cui, cue are already used in this way or could be used in this way (I’m not so certain after reading the appendix to L1).

 On the subject of Loglan usage for logic and mathematics; the designation system of Loglan (more exactly, the current stock of designators; the grammar is fine!) is inappropriate for logic and mathematics. Le preda means 'the object we are speaking of which appears to be preda'. In mathematics, it would be more appropriate to use a designator (provisionally ?laa) such that ?laa preda means 'the unique object [of which] preda holds (if there is such an object; otherwise a default value)'. When we talk about an object in mathematics, we actually know what it is (or, at least, we actually claim of it any predicate we use to locate it). ?Laa implements Russell’s definite description operator...more or less. ?Laa preda should refer to some default object, such as “the empty set” when there is no unique object of which preda holds. Another logical designator (provisionally ?lee) might be defined as follows: ?lee preda is an arbitrarily chosen object satisfying preda if there is such an object, otherwise the empty set. This implements the “Hilbert symbol”, which can in principle replace quantification: ?lee preda ga preda is equivalent to ba preda. Of course, the duplication of preda (which may be very long) makes this an impractical “linguistic move”. Also, ?Lee no preda ga preda <=> raba preda [is] again impractical due to duplication. ?Lee also implements the move “choose an object such that...”. The common feature of these two designators is that they make real claims about their referents: ?laa preda really does satisfy preda if there is one and only one object satisfying preda (and is fully specified otherwise as the empty set); ?lee preda really is an object of which preda holds if there is any such object and [is] otherwise equal to zero. 

 I disagree absolutely with your assertion that mathematical discourse can be carried on using identity sentences. This could not be carried very far beyond the level of your arithmetic/algebraic examples. The problem is that one cannot quantify into A bi B, because the contexts 'A' and 'B' are referentially opaque! Read Quine on this. [You’re right, of course. —JCB] One must provide a predicate (perhaps a modification or compound built on clika) for strict identity/equality and/or provide an element ?bii of the Lexeme bi with the appropriate sense. A bi B means 'The expression ‘A’ has the same sense as expression ‘B’';  one cannot quantify into A or B in the contexts 'A' or 'B' since there is no reference to A or B present! A bii B would simply mean 'A = B': A and B are the same. I think that both measures are desirable. [Don't both 'A = B' and Loglan 'A bi B' mean that the designata of 'A' and 'B' are the same?  How would the meaning of 'A bii B' differ from that?—JCB]

 For mathematical use, I (lightly) suggest ?torclika for strict equality, ?nerdzabi for mere existence; the numbers indicate the elimination of the last argument in each case (?torclika has 2 arguments, ?nerdzabi has 1 argument).

 It is possible to eliminate references to arguments in middle positions in Loglan:

Now A nu fu nu preda B <=> A preda ba B!  Nu ju nu interchanges 2nd and 4th arguments; fu ju fu interchanges 3rd and 4th arguments. [True but awkward. We found in 1977, when we studied these “compound converses” (see the early TL's), that they are essentially undecipherable on the fly. A preda ba C and C gi A preda are much more intelligible “incompletions” of A preda BC than A nu fu nu preda C is...besides being 4 phonemes shorter! —JCB]

 A further measure in this direction would be the addition of "reflexive operators" in the same lexeme as the conversion operators:

Note that ?nuo ?fuo ?juo reduce the number of arguments. Note that A ?fuo preda means A preda ba A;  this allows more eliminations in suitable contexts.

Also, A ?nufuonu preda BC <=> A preda BBC [because] A preda BBC <=> B nu preda ABC <=> B ?fuo nu preda AC <=> A ?nufuonu preda BC.

Lexical items such as nufunu, nujunu, ?nufuonu (and these are not such long words as not to be speakable) would be learned as separate words. Certainly no one will do these transformations in their heads! These should allow new methods of eliminating mention of argument places and different ways of getting argument places to match in combining forms. (I understand that all reflexive notions can already be expressed; nonetheless, one should gain combinatorial advantages from having these operators available! Also, one will [want to] perform transformations in written Loglan!

 On the phonetic level, I think that ?bii, ?nuo, ?fuo and ?juo are serious suggestions. ?Laa and ?lee are rather strange. Unless you like the idea that these designators require a stress, they should probably be chosen differently;  but I could make a case [for] these exact forms, as well . There is some analogy to la and le; actually ?laa is a way of forming designations of unique objects and ?lee is a way of introducing a somewhat arbitrary object to talk about. So ?laa is somewhat like naming and ?lee is somewhat like the usual designator le. Of course, the little words of the nufunu type stay as they are, since their form is dictated by the grammar. The language needs predicates for mathematical identity and membership to accommodate mathematical usage, in spite of the possible presence of ?bii and the presence of bie. Mathematicians would want to talk about identity and membership much as one talks about any relation (using a predicate) under many circumstances. On the other hand, the shorter ?bii and bie forms would also be quite useful.

M. Randall Holmes

Samto and djori (or nursei from nu setci) are all available for talking in plain-text about the relations more economically predicated by bi and bie. But let's get some discussion of these various proposals of yours here in LN. When they've salted down in your mind, you can put them formally before the Keugru. It's a good idea to remind us, by doing that, that a certain proposal has not yet been formally considered, and, as a consequence, has not been either rejected or incorporated into the language...even though it's obvious from our remarks that one or the other of us thinks highly of them, as I do of these! In short, remind us from time to time of our duties as kerju of the language as well as kinci in these discussions.—JCB

Dear Editor:

I am a graduate mathematics student at the University of Delaware. I’m very new to Loglan, but I do have some ideas.

I am glad to hear children have had fun with the language. Perhaps one avenue of propagating Loglan might be to encourage this. Elementary school gifted [children] programs would be the obvious place. [An extraordinarily fertile place, too, as my experience in Flagstaff in 1990 showed; see LN90/1:2.—JCB] To minimize drudgery and maximize fun, let the kids make up their own words within minimal rules of Loglan (namely, all words must be CVCCV or CCVCV form), but use the Loglan syntax (and little words). To get them started, assemble a children’s vocabulary list of Loglan words, preferably illustrated with cartoon pictures. Core words for the list might include words for various animals, colors, family relations, etc. (How about it, Rex May?) If it succeeds, we might wind up with dozens of Loglan dialects, all with the same syntax, differing only in some vocabulary. If the kids stay interested, they may even learn the mother tongue, and add a lot to it. Lognet Junior, anyone? [I think the idea of the core words being illustrated is a great idea. I think that the idea of making up an entire new vocabulary in each and every class is not a very good idea. A lot of thought and work has gone into the creation of the base-word vocabulary from the eight-language set, and the idea of re-inventing the wheel without the scientific base that Dr. Brown had in this area really gives me pause. —Tisra][Still, getting kids and pictures involved is a wonderful idea! Go to it! (Anybody thought of using HyperCard for this?)—JCB]

 Personally, I think the idea of X-rating Loglan (December 1989) is visually rather clumsy. It would be fine for interfacing with a computer, but I would vote for indicating the phonetic change with an accent mark, at least in written Loglan or more formally prepared documents. For interfacing with a computer, X would mark the spot just fine. [The isomorphism of the language (spoken and written being the same) would not be maintained by the idea of using accents in one form and Xs in another. Since keyboards seem to be the major input medium for the next foreseeable future, let’s not make their use any harder than necessary. Ask the Chinese, Japanese and Arabic speakers (writers) of the world how easy it is to use a computer or typewriter keyboard with all those accents floating around. —Tisra][In an upcoming LN paper (or is it going into LL?), Steve Rice will also argue for diacritics as a method of incorporating a larger kit of irregular sounds into Loglan. While I agree with the goal of these various proposals—to increase our ability to transcribe natural names—my own view is that a single diacritical mark, one that was universally applicable and available on every typewriter, would probably do the trick. An overstruck mark, like a stroke or a dash, would probably work best; and I have already begun to explore it, writing the name 'Chang', for example, as Tcan or Tcan.—JCB]   

 It might be worth distributing some of the teaching software as shareware. The concept is that shareware companies list the product in their catalog, and whoever is curious about it, orders it. Whoever likes it would be morally obligated to send a registration fee to the authors.

 The advantage for the Loglan Institute is that many more people will have the chance to examine Loglan without obligation and with very little cost to The Institute. If a thousand try it out, maybe 400 might like it, and 100 actually register. That doesn’t sound like much of a return, except that it’s 100 more registration fees than you would have gotten without distributing it as shareware, 400 more potential Loglanists, and 1000 find out what Loglan is. In any case, they have to go thru The Institute to get more materials (so increased book & membership sales should offset any lost revenues). [Sounds like our Loglan Businessperson in the making!—JCB]

 It would be handy if more Loglan articles (such as the Cortesi article in Dr. Dobbs Journal) were available as reprints. I have been having problems finding them. [Sounds like a good idea to me, what about it JCB? Perhaps we could add them to the ‘Offerings’. —Tisra][I don't know about putting them in our Offerings, but we could make xerox copies of these articles available...once we'd got permission from the publishers. I'll try to get them.—JCB] 

 Another suggestion is to distribute the dictionary on floppy disk incrementally as updates are made. [We've virtually decided to do this; see page 18, this issue.—JCB] This way those of us who are impatient can at least get a partially correct dictionary relatively soon, and any extra resulting revenue can go to the dictionary project.

 Also, a disk version of the final version of the dictionary might be very handy. Especially if it is computer readable. Then it could be tied into a spelling checker via LIP or a word processor, used as an input file for data base for Artificial Intelligence experiments, or whatever. The synthesis of a Loglan parser (based on LIP), with the dictionary and a logic programming language sounds like an interesting AI project. [Indeed it does; go to it!—JCB]

 A modifiable disk version of the dictionary would be very handy for anyone wishing to do a Loglan-German, Loglan-French, etc. dictionary.

 A programming toolbox of Loglan interpretation routines might be very handy for many experimenters. The procedures would parse a character string as a Loglan expression, and translate it into a data structure which would be computer usable (or vice versa). Versions in Pascal, ‘C’, and BASIC could ultimately be distributed. Source code for LIP could be used (if it’s portable), or modified. An "adventure" type computer game with Loglan as the interface language might be a very useful teaching aid, if there’re any hackers out there willing to take it on. [Durzo tio!]

 Regarding testing the Whorf hypothesis:

 Has anyone examined speakers of American Sign Language (ASL) for Whorfian effects? As I understand it, ASL has a unique syntax that is independent of English. Children who are taught to sign in English tend to develop ASL-like syntactical shortcuts— without exposure to ASL. Furthermore, sign languages that are not "transliterations" of a verbal language tend to share enough of a common syntax that even between vastly different cultures, speakers (signers?) need only a few days to become fluent in one another’s sign language. [The book Seeing Voices by Oliver Sacks includes an interesting recounting of American and Japanese signers being able to communicate on a basic level within a day or so. Not fluent by any means, but communicating anyway, and a lot easier than their hearing counterparts. —Tisra]

 Since ASL is a spatially oriented language, one would expect that native (from childhood) ASL ‘speakers’ might think about the world differently. I think that I have heard this is true, but I have not had much luck in finding out

 Along the same lines, there is a rumor that there is a disproportionately large number of Hungarian mathematicians, with the resultant speculation that there is something unique about the structure of Hungarian that fosters logical-mathematical thinking. Can anybody confirm, deny or comment on this?


        Dennis E. Sweitzer

Dear Editor:

 Enclosed are Rex May’s book review in Liberty magazine and a subsequent letter to the editor [by Prof. J. D. McCawley] in response to it, in case you didn’t have them already. I was disappointed in the book review; I thought May had a better grasp of the identity of Loglan. After his mediocre review and [McCawley's] scathing response to it, I am afraid that readers still do not really know what Loglan is. Worse, they may now erroneously think they know what Loglan is and are not that interested in it.

 My suggestion is that a concise, authoritative letter to the editor from the creator of the language might both clear up the misconceptions and engender the interest that the book review should have.

 What do you think?

 Kerju,         Kathy [Macedon] 

I wish I had the time to do this, Kathy, and I invite anyone else  who would like to write a  rebuttal to Prof. McCawley's misinforming letter to do so. . . Djori Macedon, for instance!—JCB