From C.D.Wright@SOLIPSYS.COMPULINK.CO.UK Thu May 04 00:07:42 2000 Return-Path: Received: (qmail 19378 invoked from network); 4 May 2000 07:07:42 -0000 Received: from unknown (10.1.10.142) by m4.onelist.org with QMQP; 4 May 2000 07:07:42 -0000 Received: from unknown (HELO nickel.cix.co.uk) (194.153.0.18) by mta3 with SMTP; 4 May 2000 07:07:41 -0000 Received: from solipsys.compulink.co.uk (solipsys.compulink.co.uk [194.153.10.165]) by nickel.cix.co.uk (8.9.3+Sun/8.9.1) with SMTP id IAA15553 for ; Thu, 4 May 2000 08:07:36 +0100 (BST) X-Envelope-From: C.D.Wright@solipsys.compulink.co.uk Message-Id: <200005040707.IAA15553@nickel.cix.co.uk> Comments: Authenticated sender is To: lojban@egroups.com Date: Thu, 4 May 2000 07:10:33 +0000 Subject: More publicity ... Priority: normal X-mailer: Pegasus Mail for Windows (v2.31) From: C.D.Wright@SOLIPSYS.COMPULINK.CO.UK coi rodo, There's a chap who writes a 900 word essay on linguistics twice a month. You can find it by putting "LangMin" into your favourite search engine - it's on bluemarble.net somewhere. I'll try to more more precise when I have more time. However, I managed to pique his interest about lojban and he's writing a column/essay on it. Here is his first draft, not to be made public, offered for comments about its accuracy, and any suggestions to improve it. He says: > Here, for better or worse, is a first attempt to put > the idea into simple words for my general public. I > always keep them to a maximum of 900 words, figuring > that people won't read much more, and this one still > counts out at 930 or so. Also, I suspect that my > transmission is going to lose the edits I carefully put > in, so you'll have to surmise where I have italic, > boldface etc. I haven't thought up an appropriately > snappy title yet. I'll be eager to hear your opinion! And below is the essay. Please email me directly or post any thoughts you have to improve the item. Feel free to go and look at the other essays to get the idea. Thanks. cdw. --- Look at these three sentences: 1 Harold is the father of Susan. 2 Harold accompanies Susan. 3 Harold is taller than Susan. In (1), what is said about him with respect to her is in the form of a noun, in (2) a verb, and in (3) an adjective. We think of these as three distinct ways in which we can talk about someone. All through history, whole libraries have been written about the metaphysical distinctions between SUBSTANCES, PROCESSES and ATTRIBUTES. But is there anything particularly nominal, verbal or adjectival in what is being said about Harold? Look how easily we can rephrase these, preserving the basic meaning but interchanging parts of speech: 1 Harold has fathered Susan. (noun becomes verb) 2 Harold is the accompanier of Susan (verb becomes noun) 3 Harold extends up above Susan (adjective becomes verb) Whenever we compare languages, we find one expressing something in verbal form (an action) that in another language is a noun (a thing), and other combinations. We might begin to suspect that using a particular part of speech just depends on what language we're talking (you'll find some specific examples of this in Miniature 34). If so, parts of speech may not be playing any essential role in describing what is said about Harold. Just like any other natural language, English looks like it's arbitrarily adding a metaphysical spin (substances, processes, attributes) to some fundamental things - genetics, activity, or size for example - claimed about Harold with respect to Susan. But English not only OVERspecifies, it also UNDERspecifies. Two examples: * When we negate ideas, we're often so vague about just what it is we're negating that we're really leaving to the listener the whole burden of choosing the intended interpretation. Does Daphne did not go to Memphis from Atlanta mean that she went to Memphis from some other city, that she went somewhere else from Atlanta, or that someone else went from Atlanta to Memphis? * Look at the difference between these two sentences: a. The demonstrators who were unruly were arrested. b. The demonstrators, who were unruly, were arrested. It's pretty important, isn't it, who we're saying was arrested here (certain ones in a., all of them in b.)? And yet to distinguish two quite different meanings we depend on nothing but those two little commas or our tone of voice. All languages make many distinctions that others don't, but all of them also leave unspecified some fundamental distinctions like the above. So shouldn't it be possible to design a language that states specifically and without any ambiguity anything we want to say, and do it without cultural metaphysics? Let's imagine we do this. * Saying something about someone or something is called predicating it. When we say something about Harold, we want exactly what we're predicating to stand out clearly, but we stop short of adding anything (parts of speech, for instance) that does not directly contribute to this. * When we negate something, we specify exactly what it is we're negating (Daphne, Memphis, Atlanta for instance) * When we say something about a group of people, we specify exactly whether what we're predicating (being arrested for instance) is true of all of them or only some. In the mid-1950s a fresh attempt was made on the centuries-old vision of a complete constructed logical language. This language followed the strict principles of symbolic logic, but this time the guiding idea was to clothe all those unfamiliar mathematical-looking symbols with easily pronounceable words. It was called Loglan, and you can find it presented in the April 1960 issue of Scientific American. For the last three decades or so, a small group of people has been extending and filling out the construction of a logical language. Its present-day incarnation is called Lojban. Loglan/Lojban has now become a highly flexible, culturally neutral, instrument of logical thought that allows the limitless expression that natural language does. Here are some of its main features. * It puts what is being talked about (Harold, Susan, Daphne, the demonstrators or anything else) in the form of a simple but flexible predicate relationship, following the principles of formal logic. * It uses none of the standard `parts of speech' since they are not part of logic, using instead a uniform set of predicate words. * Its grammar includes predicate words (what is being claimed), operators (for instance time and location markers), connectives (like conjunctions, showing how ideas are related), and indicators (including a wide variety of possible attitudes, and the speaker's assertion of whether what he says comes from direct observation, a second-hand report, or a surmise). * It does not use tenses (past, present, future) in the traditional sense. (Miniature 17 shows how little correlation there really is between tense forms of verbs and time concepts). Instead, there is a set of tense markers allowing fine distinctions and combining time (now/then) and space (here/there). * Each of these types of words has its own unique phonetic form (that is, combinations of consonants and vowels), so that in a sentence the grammatical function of every word is instantly recognizable. * Even punctuation has been made specific: all the usual punctuation marks appear as pronounceable words. * It is not a computer language, but since all the ambiguities of natural language have been designed out, it can be readily processed by computer. It therefore has potential to serve as an intermediate language in computer-aided translation between natural languages. The Lojban web site offers a complete explanation of its design, and both a summary and a detailed presentation of the grammar. TAKE A LOOK. [link to www.animal.helsinki.fi/lojban/] -- \\// ze'uku ko jmive gi'e snada