Wittgenstein argued (in the Tractatus) that language is composed of complex proposition which can be reduced to a series of elementary propositions (atoms).
It is said by some that he naively believed that the world could be built from these "atoms" but it is obvious that Wittgenstein had no great faith in language and maintained that "Philosophy is a battle against the bewitchment of our intelligence by means of language."
It is also obvious that there are many things that language cannot describe but if we were able to create a form of Wittgenstein’s “atoms” which did not have many of the pitfalls of conventional languages then maybe it would be possible to combine them and construct “molecules” – a step along the path.
It is my belief that a language like ClassyLang could go a long way toward making “atomic” definitions much more precise.
I'm sometimes amazed by my own naivety!
Here am I offering ideas that will help remove the enchantment from language on a site where most believe that the enchantment is more important than the message :O)
Of course there is even less chance of those already very favourably entrenched in the 'system' to want something which might alter the status quo.
Stipped of its power-words, emotive rhetoric, euphemisms and techno speak a lot of today's political and pseudo-scientific propaganda would be revealed to be the nonsense that it patently is and the Emperor would be displayed in all his naked glory and people might realise that the clothes don't make the man they just disguise him!
*********************************************************
Originally I simply wanted an improved thesaurus and inspired by the Dewar Decimal System I thought I would use a classification system that reflected what the word was about rather than what its title was:O)
What we generally think of as words are simply labels attached to a variation of agreed meanings that usually have little or no obvious connection other than by occasional, unreliable, grammatical rules.
There is added confusion in that the same label can have several diferent meanings and words like 'lead - the metal' and 'lead - to go in front' show that labels can even be pronounced differently without any indication from the label itself.
I wanted a system in which the word actually points directly to its meaning in a systematic and logical fashion. A lot like redefining Bar Code numbers to be related to the product they label rather than simply assigning them Ad Hoc.
The ambition grew into an attempt to define a new way of storing and classifying information - hence ClassyLang. Maybe I should have gone for Babylang :O)
I decided on a 32 bit word divided into 8 nibbles, which could be expressed as an 8 digit hexadecimal word, with the first 3 bits of the MSN (most significant nibble) classifying the word into Noun, Verb, Adjective, Adverb, Proposition, Conjunction and Literal and the other bit used as a flag (to denote the end of a sentence).
Depending on what type of word was denoted by the MSN different lists of properties would be set by the next nibble and so on down the remaining word...
i.e. for nouns properties like -
Abstract, Concrete, Absolute, Relative, Common, Pronoun etc.
and for verbs -
Transive, Intransive, Tense, Auxilary etc.
Started and Finished (completion) would be useful as well as perhaps 'Now' for triggering events.
Since all this would be transparent to the user the structure can be as inaccessable as machine code so long as it can be translated and interpreted with sufficient sophistication.
If it could be achieved then it should be possible to interpret the information in many different ways - into different languages, in a 'terse' or 'verbose' form, or even in a simple or technical form - simply by deciding what level of complexity the words should be allowed when output by the translator.
Originally Nouns were modified by Adjectives and Verbs by Adverbs and this was done by adding properties or adjusting their values.
Propositions modified relationships and Conjuctions allowed the joining of Classes or added new Members.
Later I began to try and modify the language to make it an ideal choice for AI...
since a language whose words actually had values which a computer could use to make decisions and comparisons is far superior to having to use Reserved words which must be defined and demand perfect grammar.
However, I finally realised that the job had become too big for any one man and would require a team of people from many different fields with a wide range of expertise.
My thoughts are that the early design attempts to be too ingenious and that it should be simplified to each nibble being used as a choice from a 16 member list. The members of the list (essentially the available properties of that type of word) would be dependent on the value of the preceeding nibble.
In general Nouns would own Adjectives and Verbs own Adverbs and the whole be owned by a sentence/concept.
There is much more to say about the benefits of such a system should it prove to be feasable and suitably implemented... I hope someone with the necessary vision and resources sees the value of such a language and continues its development!
Comments
FTSE100 | July 24, 2008 - 12:24
Cui bono? Suppose you finished your system, who would you give it to and how would they benefit from it? Ludwig needs to know.
mykle | July 24, 2008 - 16:52
I don't intend to try and finish 'my system' FTSE.
It would probably be best to start again from scratch with a small team of experts with previous experience of designing a computer language - like, say, Visual Basic.
My reason for posting this is simply in the hope that someone will see the potential of the ideas behind "ClassyLang" and develop it for the benefit of all those who seek knowledge.
The power of computers has not been used to store information in a logical way - as usual we have simply used modern technology to extend rather than truly revolutionise... although the Internet was a big step forward.
With luck someone will see the potential and realise that it can be done in steps with a small vocabulary and plenty of modifiers at first - then from the atoms, molecules...
FTSE100 | July 24, 2008 - 17:00
Sorry to be as persistent as a Paxman, but that doesn't really answer my question. Who is it for? Why will they be pleased to have it? Just because a thng can be done doesn't imply there's any point in doing it.
Once upon a time a salesman attended an inspirational lecture. 'To sell your product,' said the speaker, 'you must first discover your customer's problems, then point out how your product will solve them.' The salesman returned to the office and reported his understanding of the lecture thus: 'The geezer said we'll sell more of our products if we call them solutions.'
If you see the point of this anecdote, you'll see what question I'm asking. It's no good saying your system is a solution. What problem does it solve?
mykle | July 24, 2008 - 17:36
In the bible it says somewhere that God only gave us two tasks – to multiply and to classify...
If we are going to give names to everything we might as well make the naming rules consistent :O)
Seriously, I suppose all those people who can’t spell and hate the tyranny of the dictionary’s alphabetic sequencing would eventually benefit because they could enter simple words and combine them to get the more sophisticated word they were looking for.
People who try understanding technical or scientific literature but never get past the jargon.
All those long suffering people who use language translators only to find they produce total gobble-de-gook.
Eventually, I see a ClasstLang chip in every computer because everything stored in ClassyLang would be available to everyone online so long as they had a translator for their language – since it is a concept based language not a grammatical one.
Most important of all it is a stepping stone. People could have said “What’s the point of connecting a few computers together – saves a bit of time I suppose but isn’t it expensive and complicated?”
mykle | July 25, 2008 - 10:11
I’m very grateful for your interest FTSE!
Your questions inspired me to consider what ClassyLang might, ultimately, be a stepping stone to.
Since it is a conceptual language based on properties it would make an ideal front-end for a new type of operating system. Instead of just “Windows” which allows you to look into the digital domain it would allow “Doors” to actually get in there and affect things.
If Microsoft, or maybe Apple, were to get involved and design and add a Command library, which others could write proprietary drivers for, then you would be able to ‘talk’ to your fridge, TV or whatever in simple everyday language. Now, surely that is a ‘killer’ application :O)
Trying to do this with an ordinary language is virtually impossible because of the vagaries of traditional spoken languages but because every ‘word’ in ClassyLang has its own definition built in it gives it a precision that no other language could hope to match – unless it were based on similar principles!
mykle | July 25, 2008 - 13:08
I use Google not only because it seems to find things other search engines can’t but also because I love its fairly sophisticated “Did You Mean?” facility. To me this is a nod in the direction of a language like ClassyLang and I expect in the early days “Do You Mean?” would be very common...
Imagine being able to write an email in English and it being readable in any other major language.
Imagine being able to do a Google search through the whole Web because language was no longer a barrier to the information you seek.
Imagine chat sites with people from all over the world being able to write in their native language instead of being limited by their knowledge of the English language.
A truly World Wide Web!
mykle | July 28, 2008 - 05:44
To redefine my vision without going into specifics:
I propose a language which computers can use to store information - which can then be translated into any other language with the aid of a suitable translator.
The ‘words’ of this language would be defined by strict rules and would be an expression of the properties of the thing they described. I suggest a hexadecimal format of 8 or 16 ‘letters’ per word.
The ‘nouns’ and ‘verbs’ would be the fundamental building blocks which would be modified by ‘adjectives’ and ‘adverbs’ etc.
In other words I suggest a ‘Car, fast, big’ approach (rather than big, fast, car).
The language should be developed with Artificial Intelligence in mind as that would not only make in very useful in the future but also help to make it less depended on legacy languages.
Since the new design allows 16 different word types then to the original Noun, Verb, Adjective, Adverb, Proposition, Conjunction and Literal word types - I would add: Data, Variable, Value (absolute) and Instruction (since I feel computer design will move to having separate memory for data and code to aid stability and minimise virus problems) also, although it goes somewhat against the principle of ClassyLang I would add Reserved (word) to aid in interfacing with other computer languages.
I would also reserve at least 2 word types for future use.
I suggest that it be developed in stages starting with a simple vocabulary and plenty of modifiers...
Having spent some time in Thailand it is amazing how much you can say with a few nouns and verbs and plenty of “dee”’s – good and “markh”’s – extra/more : hence you can say dee, dee-dee, dee-dee markh up to dee-dee-markh-markh-markh before you get to dee- tee-sut which means The Best.
It would be up to the user and the translator to decide what level of sophistication is output in the target language where, for instance “excellent” could replace ‘very, very, good’ or whatever.
It strikes me that with the world’s biggest online community that China would find ClassyLang the perfect means of expanding the horizons of the Chinese people online. The fact that the interpreter could be designed to ‘censor’ chosen words and concepts might be a blow to the free speech ethic of the Internet but it would still allow a huge leap in the amount of information that would be available to the average Chinese surfer with little or no English!
mykle | July 30, 2008 - 11:40
I notice that Cuil (pronounced ‘cool’ I hear) the new Google challenger started today with a whimper rather than a bang... but then they had to over hype it to get noticed.
They should have called it Classy search as it is supposed to offer better classification of search related subjects.
I’m not sure all the pundits are right about what people want but Google has certainly been good at supplying it so far.
Perhaps Cuil are hoping Microsoft might make an offer – maybe they will but they’ll bide their time to see how the cookies tumble :O)
I’ll certainly give it a go when I can actually find a link to it :O)
Maybe they might consider adding a language like ClassyLang it certainly wouldn’t hurt their hype or their dented image :O)
I was wondering why Cuil appeared to perform so badly on its debut day – almost as if it were being sabotaged.
So far as I know it was designed by members of the team who originally helped Google to rise to its dominant position so they must know what they are doing!
Then it struck me that Cuil was hyped as giving you greater choice “(Cuil)...provides a more comprehensive guide to the web”.
Now isn’t that a good thing? Well, by being less limited to popular sites - would of course mean easier access (less searches) to find non mainstream thinking...
maybe the truth is out there but they don't want you to find it :O)
mykle | August 11, 2008 - 10:36
Well, I've done my bit, explained my dream:-
a NEW type of language for MEN and MACHINES.
So now I can leave it to fate and fortune to see who takes up the dream and turns it into a reality.
No doubt there has been some talk with everyone saying can't be done while their boffin boys in the brightly lit back room say "Less than a year."
It's my hope that Google wins the race because they are actually good at what they do!
Microsoft has stolen every good idea it has ever had and will try to do the same with this...
Google has the money and the power to prevent that and make Microsoft pay royalties for a change :O)
Heavens, we may even get an OS that works without a disk and looks where it is going so it doesn't crash when the drivers are poor :O)
mykle | August 12, 2008 - 08:24
In my search for a new language I decided to try and mimic human thinking.
So I began to analyse the human decision making process.
It occurred to me that the human brain is quite similar to a computer in that it tends to make binary decisions.
This, of course, leads to the serious problem of what to do with the “maybes, the don’t knows”.
I noticed that in many people the system worked like this...
“Do I want to go to the pub?” followed by a long list of reasons against.
There would be a single reason for going, which would be because they felt like it.
Now, in this process, there would be a huge number of don’t knows...
Can i afford it? Will I suffer in the morning? Will there be anybody there that I know? etc.
What always seemed to happen in the end was that all the “Don’t knows” would be counted as “Yes” and so to the pub it would be.
I called this the “optimistic system”, whereas when the “don’t knows” were counted as “No” it was pessimistic.
FTSE100 | August 12, 2008 - 09:55
Interesting idea, mykle. It would account, maybe, for illnesses like manic depression, or bipolar wossname if you want to be pc. During the manic phase all don't knows are 'yes', during the depressive phase all don't knows are 'no'.
The human brain is a weird thing. I've got one, so I know. How much of the time are we just kidding ourselves about why we choose to do this or that? As you know, you can hypnotise people, get them perform some entirely inappropriate act, then ask why they did it. They'll instantly come up with a reason, and truly believe that the reason they've invented was the cause of their action!
mykle | August 12, 2008 - 11:55
Thanks FTSE!
Yes, I vaugely remember a wierd phenomenon called Zombieism (don't know how it is spelt) which suggests that quite often that is the case that we invent resons for our actions rather than actions for our reasons.
Sorry, if that didn't come out well - I'm very tired and when I have had little sleep for a long time I tend to go a little 'crazy frog' and get lost in thoughts that I usually have enough sense to avoid :O)
I'm going to try and take a nap.
Thanks again.
mykle | August 13, 2008 - 16:32
Google Turns iPhone Into Universal Translation Device
http://www.tomshardware.co.uk/Google-iPhone-Translate,news-28939.html
Well, that was quick! Half way there already :O)
Mangone | May 17, 2009 - 08:48
While Google goggles will Alpha assimilate :O)
Wolfram says his Alpha creation is not so much a search engine as a "computational knowledge engine."
Like a search engine, but users can pose complex questions.
http://www.newsweek.com/id/197894
Mangone | December 22, 2009 - 07:59
http://blog.wolframalpha.com/2009/12/21/new-features-in-wolframalpha-yea...