My Linguistics Honours Thesis

Talking Through Every Possible Barrier

---

How To Communicate Without Words

2006年11月7日火曜日

§ 1.0 Introduction
My humble attempt at an honours thesis is to create a semantically coded language similar in purpose to LINCOS, but on a smaller semantic scale. This small scale coding could have applications later on in my M.A. when I try to create a full fledged language coding in the purpose of LINCOS but with improved information carrying capabilities. So this project as it stands is merely to built a galactic communication primer.

NOTE THIS BLOG WILL ONLY HAVE ONE VERY LONG ENTRY THAT WILL BE UPDATED AS TIME GOES ON.

§ 1.1 Situation Log
1. No Common Language - The language parsing factors between two completely unrelated intelligences could be a very broad bridge to try and cross, especially without direct contact. Completely different evolution (unless Panspermia is true)
1.1.1 Recently at a SETI conference anthropologists and ethnographers were called in to talk about what things are in common and what we can use to communicate with to an alien culture.
Something that I noted as salient was the fact that mathematics may not be primal. That is 1 + 1 = 2, may be a terrestrial prime, but on another planet it might be somethign like [approximately 1] plus [appriximately one] equals [approximately two]. Though in the theory of algebra we show that 0.999... actually equals 1.
That is that 0.999...=A, 10A=9.999..., 10A-A=9.999...- 0.999..., 9A=9, A=1
But even so even the way we look at numbers can be ethonocentric. So Below when I talk about using prime numbers please take the method with a grain of salt, and note than other methods such as consequtively listing numbers up to n over increasingly complex number systems up to i could work too.
1.2.1 Can only communicate by EMR - Messages are send in radio waves in an on/off pattern (zero or one, i.e. binary) [see 1.2.2], codinge multiple values into this (decimal) would increase the decoding complexity, perhaps by first requiring a decryptification (of binary) then a second interpretation to allow another system (Hex, Decimal etc.). If in the first section of the code, we could code in that this (decimal) would be the following paradigm of numbering. [
1.2.2 Again there is a problem, what method are we using for transmission. The entire problem of not finding intelligent messages blasting between interstellar space may be our inability to find the right frequency or we lack the technology to pick up the weak signals. The strongest signal sent was 3,000,000,000,000 Watts (3 Trillion) The strongest AM stations by comparison are powered at a mere 50,000 Watts, The Arecibo message was 60 Millions times stronger than the strongest conventional AM stations. But what use is all of this power if the recieving/searching end isn't looking at 2380MHz? All of our searches for ETI have been at 1420.45Mhz the freqency of the most abundant atom (Or in its higher harmonics). Other suggestions have been to look in the Water hole. That is the between the two main frequencies of water because alien biology may also use water as a solvent, but if ammonia or hydrocarbons are used (As Drake and Sagan discuss) the water hole search would be futile. Also the main source of finding interstellar messages is logically the civilisations that are long lived and distant. This communication would most likely not be on a search freqency but on a conventional dispatches/communique. And that senario only holds if a faster-than-light communication method has not been found. In which case we can only hope for our ET's putting their mirror Arecibo recievers on our transmission frequency. I include the picture below because a suggestion has popped up recently regarding the use of laser transmitters. We have a wide range to transmit in and receive in.

1.3. The person on the other side of contact will be a team of highly specialized and intelligent beings, but we have to initiate intelligible contact, by designing a creative system we can send messages in a method that will code the maximum information in the least data - Why wait when we can be ready, or try to initiate contact ourselves!
1.4. The message must be able to resist noise modification just due to the interference acting on interstellar messages.
1.5. The code should be compact for ease of transmission, and codification. Though since the distances are so great the message would not be responded to in less than 20n years anyway so it could code a very very long message.
1.5.1 Also how do we know how long to make the digits of our code. As I discuss the code is sent in a stream of on and off bursts, but how long should each one be? Should it relate to the wavelength of a common chemical such as neutral hydrogen? In any case the length of the digits are what our definition of time will be based off of these digits, as latter on [length]of[life]of[human]=[approximately][2,524,566,528][seconds], [1][second]=[4][digit bursts]
6. The previous attempts at creating a language for atronomical purposes ignores dexterity and value of semantic coding within the over logical code. For anything to be truly defined it has to have a place where the word being defined stops being defined by words that themselves require a definition. All words should end in a small system of primal words which all other words are intuitively based off of. This invoves the theory of dictionaries and the circularity of definition. Building a languae using the previously defined primes will help NSM researchers find new prime candidates and also make a conprehensive system for computation and codification of language.

§ 1.2 The Why
In 1998, D, Vakock wrote a paper concerning writing extraterrestrial messages, which was featured in Acta Astronautica 42 pp697-704. He gave several reasons why we should research this:

Concretely understanding the challenge of creating an adequate reply;

Helping decode messages from extraterrestrials;

Creating interstellar compositions as a new form of art;

Having a reply ready in case we receive a message;

§ 2.0 Principles of Development
Using the two values of a binary system would be severely limiting. One issue is how will the person on the other side even know that the beeps are intelligible?

§ 2.1 Engaging a Listener
There is a series of numbers that would be immediately recognisable. Try for yourself: 2,3,5,7,11,13,17,19... I know that If I was listening of an intelligent message I would identify this sequence immediately.

§ 2.2 The Message
Of course you can only send so much information by sending prime numbers. But what if we send a raster of on and off pixels through the numbers we can draw in code our own messages, instead of trying to code in messages after an initial set of primes on a single unending line of ones and zeros. i.e. 0,1,1,0,1,0,1,0,0...n.

Using as in the diagram above a referent line vertically shown here to align the data and set up exactly how large our raster will be. And If we make our raster the height and width of two prime numbers we can send our messages in discrete lengths, and our interlocutors will also have the clue of having a message in the exact length of two prime numbers multiplied.
But also to consider is what we could do if we had no restrictions on length of data or else each pixel was much faster than a second. We could infact add an z-axis to our message. This could be used to transmit 3d models of physiology or anything else that is very hard to picture in 2d or low resolution. For instance if a message is recieved one of the first things humans will think is what do our fellow sentients look like?
Something like this taking
This method of course will take a lot of space so it should be used sparingly in place of a 2d raster. This 3d image could take up to 960 million pixels. It would be equivalent to a giant black and white image 31000 pixels in x and y extent.

§ 3.0 The Content
Now that we have decided how to set up our message we can explore the different ways that such a message could be written.

§ 3.1 Communicating in Pictures
So having a raster we can actually draw what we want to convey. This is a simple way to work around defining many words to describe the situation. From this picture you can see our galactic location.
The data is send in a continuous stream and if they were to cut the data at 31 bits then repeat, to get a raster of 31x41 they would get the picture here.

If they cut it at 41 bits and repeat to get 41x31 they would get something very close to this.


Another consideration for debate is adding a single straight line down through the image which will allow for calibration due to noise and will allow a reference for where the lines of message connect vertically.

§ 3.2 Communicating in Code
Beside using drawing we can draw in regularised coding pictographs which is the means I will primarily use.

§ 3.3 Types of Coding
I want the readers to consider how the raster can be set up to give a message. The most inefficient method is to have every pixel carry a meaning, then these meanings can be combined to give a more complex meaning.

But if this was the method we would need to have a huge raster for every single word, however elegant the method the following is much better and the last best of all. In this second method every glyph is abstract ad has a unique meaning for the combination of pixels it has on/off.

This following method I believe is the best. We have the option of classifying our glyphs with the corner six pixels giveing readers a sense about what the word will mean before it is defined for them. They will know whether it is negating, what class of word it is, or even which particle it is being used with. But the real advantage is saving space, if we can add information into the six pixels here then we save 35 pixels for an additional character. I dont think we will need to use gender or case as we will have fixed word order and the grammar will be standardized but even with just six pixels we can have 400+ modifiers or combination of modifiers on a word, which should be sufficient but we can always update this later if it is too few, or in such a way that it will be able to resist noise.

§ 3.4 Established Codes

This image (from ThinkQuest linked below) show the first page of the Dutil-Dumas message sent, since it has been sent another message was sent using updated symbols, much like in the last of my style examples to account for astro-noise affecting the image, another interesting feature is that all the sections of code are boxed in, which will allow for compensation in the noise introduced. See more of Dr. Dutil's discussion on Astrobiology from his position in the Defence Research Establishment Valcartier.

This second image shows the definition of physical reality from mathematical ratios and graphical representations, which Dutil points out fervently, are the weakest point in the LINCOS based code.

§ 4.0 Semantic Realizations

to be presented to university of calgary staff and visting academics in April 2008

PREVIEW:

In Previous Attempts to create a universally decryptifiable language of transmission it has been a logical code, and in the classic book in the subject (LINCOS: A Language for Cosmic Discourse) this logical code switches to a discourse between actors to show more abstract concepts such as good and bad.

In my this thesis I will attempt to code the Anna Wierzbicka's Natural Semantic Metalanguage, the proposed prime (fundamental words) of all human languages, with logical code. With this system I will then build up a vocabulary using current NSM techniques, and also some Fregean Semantic decompositions.

For instance exactly at this moment I am looking to code PART OF and KIND OF as two different entities using examples in the previously defined words and, by virtue non-decomposable, NSM primes.

I think I will show kind of as PRIME# being a kind of NUMBER and then have part of shown by defining Pi π as a repeating fraction and then showing a part of the infinite fraction set as being part of π.