What would a language based on binary code be like? [Archive]

View Full Version : What would a language based on binary code be like?

druid91

2012-03-09, 06:59 PM

Ok, say for argument that someone was to make a language for conversing with other people, with the intent of using binary code as it's delivery?

How would it differ from english?

Ravens_cry

2012-03-09, 07:02 PM

Strictly speaking, you can use English with binary code. In fact, we're doing it right now, but it's being converted back into Roman text to be easier to read.
As for a spoken language, I imagine it would sound like Morse code, a series of clicks, thuds or other well carrying, distinctive sounds.

Mercenary Pen

2012-03-09, 07:10 PM

Unless converted back into something more intelligible, it'd be a massive pain in the neck to read and write- if only because you need lots of different binary numbers to convey complex information, and by the time you've got to eight or higher you have four digit numbers already...

A case in point can be seen in this youtube video (http://www.youtube.com/watch?v=0Kx9NVVOU88&feature=plcp&context=C4bc7fcfVDvjVQa1PpcFP9myyrWBQy0FxgVDXQIUTj DK4igA5WDrA%3D) from 2:30 onwards.

Worira

2012-03-09, 07:23 PM

I imagine the main difference would be that it would be in binary.

druid91

2012-03-09, 07:31 PM

I meant more along the lines of if someone took binary code, presumably with some way of quickly speaking/hearing it, and then built a language off of it. What would it be like?

Ravens_cry

2012-03-09, 07:33 PM

I meant more along the lines of if someone took binary code, presumably with some way of quickly speaking/hearing it, and then built a language off of it. What would it be like?
Long winded yet laconic.
It takes,
01001000 01100101 01101100 01101100 01101111 00100000 01110111 01101111 01110010 01101100 01100100 00101110
just to say (sans quotes) "Hello world."

wxdruid

2012-03-09, 08:58 PM

There's a Star Trek episode with this

11001001

The aliens speak in binary code

http://en.m.wikipedia.org/wiki/11001001

Grinner

2012-03-09, 09:27 PM

I imagine it would be conveyed in a series of beeps and whistles.

Siosilvar

2012-03-09, 10:26 PM

It would be horribly, horribly inefficient. You'd need to be exactingly precise with your timing or sit there clicking away for half an hour for a single sentence.

Imagine an tiny vocabulary of 4000 words for this language. Every single word would need to be 12 time-units long, because otherwise how can you tell the difference between "1 10101" and "1010101"? You could have a unique "space" word, but then you have to make sure that doesn't show up in any other word or possible string of two words back-to-back.

EDIT: Either that or your "space" word needs to be simply a pause longer than any other words in the language. If it's shorter, then what's the difference between "10000001" and "1 1"? There's simply too many restrictions to make it feasible.

Ravens_cry

2012-03-09, 10:34 PM

While I agree with most of that, one quibble. The zero doesn't have to be silence. It could be a squeak verses a squawk, for example.

Xuc Xac

2012-03-09, 10:38 PM

Talking in binary does not require talking like R2D2. It would only mean that the language would be limited to two syllables. How the language sounds would depend on what those two syllables were. A language made up entirely of "ko" and "ta" would sound very different than one made up of "foo" and "sha" or "mar" and "klar" for example.

Regardless of which two syllables it uses, it would sound very repetitive and the words would be very long because there are very few unique combinations that are only a few syllables long. It would be ridiculously inefficient in terms of both time and energy. There's a very good reason why human languages all use several methods of data compression such as using dozens of different phonemes, tones, pitch stress, vowel length, etc. When a rock is about to fall on you or a tiger is about to pounce on you, do you want your friends who see it to warn you with something like "move!","run!", or "look out!"? Or would you rather wait for them to finish something like "kokotakotakokotatatakotakotakokokokotatakotako!"?

celtois

2012-03-09, 11:10 PM

I think you are all thinking in a pattern that is far too restrictive.

If I were to develop a language based on binary code I would make various different combinations of 1's and 0's represent different sounds so
1 would be one sound
0 would be another
10 would be another
01 would be another
...
etc.

The advantage of this, is that verbal communication would be very quick, or at least as quick as any other language.

The trade off of this, is that writing would be even longer then what you were writing, so it would be a strictly verbal tradition, with writing being left to something like a priest caste.

factotum

2012-03-10, 01:32 AM

If I were to develop a language based on binary code I would make various different combinations of 1's and 0's represent different sounds so
1 would be one sound
0 would be another
10 would be another
01 would be another

But that's no longer a binary code--it's an analogue code masquerading as one. It's as if someone comes in and says, "How do I do binary arithmetic?" and you turn and say, "It's a lot easier if you use a few more numbers--I use 0-9". It might be true, but it's not answering the original question!

Eldan

2012-03-10, 07:43 AM

Adlan

2012-03-10, 12:39 PM

Ok, say for argument that someone was to make a language for conversing with other people, with the intent of using binary code as it's delivery?

How would it differ from english?

Does the Librarian from Discworld seem a good fit to you. He has says Ook*, or he is silent. But even then, he can say Ook, or Oook, people have little difficulty understand him, maybe because he can use lots of different inflections. So it's not really Binary.

Binary is Digital, the spoken word is Analog. So even a language of one word won't do. It's got to be a language of one sound. Silence or one sound, of uniform length, tone, volume ect. For this to be a language there would have to be a code to give it meaning.

*Yeah I know, Occasionally he says Eeeek!, but just run with it.

Though for fun, I like the Bit from Tron (http://www.youtube.com/watch?v=_fGujzulsas).

bluewind95

2012-03-10, 02:07 PM

I think that it'd be something that sounds akin to morse code (no, it doesn't need one sound. Two would do. The meaning would be one and none). Most of the meaning would be from tone and body language, and the vocabulary itself would be very simple.

Icewalker

2012-03-10, 02:44 PM

..- --, .. - .-- --- ..- .-.. -.. ... --- ..- -. -.. ... --- -- . - .... .. -. --. .-.. .. -.- . - .... .. ... -.. --- . ...

(Um, it would sound something like this does)
This allows for timing as well as binary characters though, which is of some note.

factotum

2012-03-10, 03:41 PM

Though for fun, I like the Bit from Tron (http://www.youtube.com/watch?v=_fGujzulsas).

Although the Bit isn't actually a bit at all, because he has *three* states--yes, no, and neutral (when he isn't saying anything). :smalltongue:

Tyndmyr

2012-03-10, 05:50 PM

While I agree with most of that, one quibble. The zero doesn't have to be silence. It could be a squeak verses a squawk, for example.

If that's the case, we add silence and use trinary instead of binary. You wouldn't base a natural language off binary unless you absolutely had to.

Siosilvar

2012-03-10, 07:22 PM

Re:
While I agree with most of that, one quibble. The zero doesn't have to be silence. It could be a squeak verses a squawk, for example.and re:
I think that it'd be something that sounds akin to morse code (no, it doesn't need one sound. Two would do. The meaning would be one and none). Most of the meaning would be from tone and body language, and the vocabulary itself would be very simple.

This has been said by a couple people already but I'll say it again:

If you have two sounds, then you have three possible things you could be doing: sound 1, sound 2, or no sound. That's not binary, that's ternary. Even two different tones lengths of the same sound has moved beyond binary*. Two different tones of the "same" sound is definitely past binary.

Binary is "on" or "off", just like a typical computer transistor (but not like an electrical current, which can be one direction or the other and of varying strength). There's no second sound.

*With the exception of twice the length of the sound meaning that sound twice, thrice meaning thrice, and so on. So "rarara" would just be a treble-length "raaaa".

Drumbum42

2012-03-11, 09:43 AM

Re:and re:

This has been said by a couple people already but I'll say it again:

If you have two sounds, then you have three possible things you could be doing: sound 1, sound 2, or no sound. That's not binary, that's ternary. Even two different tones lengths of the same sound has moved beyond binary*. Two different tones of the "same" sound is definitely past binary.

Binary is "on" or "off", just like a typical computer transistor (but not like an electrical current, which can be one direction or the other and of varying strength). There's no second sound.

*With the exception of twice the length of the sound meaning that sound twice, thrice meaning thrice, and so on. So "rarara" would just be a treble-length "raaaa".

Sorry, but that's not right. Binary is "on" or "off" from a logic standpoint, but that's not what is really happening. In actuality "on"=5v and "off"=1v, so if the input is 2v, it can be assumed that it was suppose to be 1v(off). This creates a built in error checker, to counter electromagnetic interference, and erroneous outputs. So, if that is used for a language there would be some flexibility for pronunciation. Say "zig"=1 and "plop"=0. So if someone has a lisp or something, "klop" or "slop" would be assumed to be 0.

The lack of sound is simply no verbal communication. It's still binary, just like transistors. The biggest difference is timing, and the fact it's not necessary. We can hear the end and beginning of each syllable, parsing the bits with out the need for a clock or timing to be used by both parties. (A computer must use a clock because it needs to know when a bit ends and the next begins)

erikun

2012-03-11, 10:07 AM

One other problem with a binary language would be conveying meaning. If we have the words "river" and "tree", then making a word "rivertree" would have some meaning even without explaining it. It helps that our rivertree word doesn't sound like other words, because both river and tree generally would not sound like other words.

On the other hand, binary only has two words: Yes and No. Yesno, or Noyes, or Nono, generally doesn't have much sensible meaning. Add to that that all of your words would involve the same sounds, so that "Yesnono-Noyes" sounds identical to "Yesno-Nonoyes" or even "Yes, Nonono, yes" and you'd have trouble conveying meaning in the same way.

Orzel

2012-03-11, 10:25 AM

I meant more along the lines of if someone took binary code, presumably with some way of quickly speaking/hearing it, and then built a language off of it. What would it be like?

Me and a few friends have 1 and 0 hand signals when doing things in place where we cant talk

01: Yes/Understood/Will comply
10: No/Did not hear/i aint doing that
100: Come
101: Stay
110: Follow
111: Go

the code for me is 11111000

Drumbum42

2012-03-11, 10:36 AM

Me and a few friends have 1 and 0 hand signals when doing things in place where we cant talk

01: Yes/Understood/Will comply
10: No/Did not hear/i aint doing that
100: Come
101: Stay
110: Follow
111: Go

the code for me is 11111000

That's actually a really cool idea. I should think about implementing this..... (I'm sure a few of my friends would enjoy this)

Trixie

2012-03-11, 04:00 PM

It would be horribly, horribly inefficient. You'd need to be exactingly precise with your timing or sit there clicking away for half an hour for a single sentence.

Inefficient? How so? :smallconfused:

Imagine a cricket chirping. When you only have 2 distinct sounds, you can do them much faster than humans speak their letters and still be well understood. In fact, short sentences and words would be far more efficient than ours, simply because it would need only rapid movement in two muscle states as opposed to dozens of moves needed to create our speech.

Imagine an tiny vocabulary of 4000 words for this language. Every single word would need to be 12 time-units long, because otherwise how can you tell the difference between "1 10101" and "1010101"? You could have a unique "space" word, but then you have to make sure that doesn't show up in any other word or possible string of two words back-to-back.

'Space' word is just making a pause. Morse isn't divided into any neat packages, and yet it works perfectly by context. You might as well ask what's the difference between 'there, ed' and 'the reed'.

But that's no longer a binary code--it's an analogue code masquerading as one. It's as if someone comes in and says, "How do I do binary arithmetic?" and you turn and say, "It's a lot easier if you use a few more numbers--I use 0-9". It might be true, but it's not answering the original question!

Not necessarily. Blurring sounds so that most common combinations are distinctive is simply feature of language that doesn't change anything. "They're" is still "They are" despite no one taking pains to pronounce it such. Language is still binary as always, using shortcuts doesn't change that.

To answer OP - that is not necessarily inefficient at all - mindset of race communication with only two sounds as letters would probably adapt to rapid information burst exchange, and seeing they wouldn't have any attachment to similar sounds like we do, the language would probably either remain static (no word evolution like 'gonna' instead of 'going to') through race's lifetime, or be well adapted to periodic massive changes.

Also, such race would probably be far better in using their speech than writing - one thing binary isn't good at is producing short sentences on paper. They would probably invent techniques similar to our data compression to make their script shorter, perhaps inventing decimal (or base 8, 12, 16, whatever) mathematics simply to use in their script, with 'real' mathematics perhaps being done in unshortened binary.

All in all, such topic would require a lot of assumptions before thinking. Are they hunters? Nomads? Gatherers? Herbivores? Advanced civilization? Do they have good eyesight? Etc, etc.

Gwyn chan 'r Gwyll

2012-03-11, 05:38 PM

So, a 'binary' language would only have 2 sounds, or to be truly proper, one sound and a pause, but that makes it impossible to actually work. As is, it WOULD take forever, as there's a limit to how fast you can keep going "tih" and "tuh", to take to simple examples. Try it. Don't voice it, just make the click and mouth shape, and go as fast as you can.

You're extremely limited by how fast you can physically talk.

Now, this system is basically DESIGNED so that any given word is going to have a ridiculous amount of syllables. More than you could imagine. How many words are there in the English Language? Let's say that we're only going to use 200,000 words, which is a VERY small number, considering that the OED has over 600,000 definitions.

It's been shown in a study that I can't find right now, that there's a relation in language between the speed of the language and how many syllables it takes for any given unit of meaning. Given that this binary language would actually be the most syllables POSSIBLE per given unit of meaning, it would have to be spoken UNBELIEVABLY fast.

I would say it would not be possible for humans.

Telonius

2012-03-11, 05:56 PM

Yeah, you'd need that many syllables to flesh out the whole language. But really, how many words does the average person use in a day? You'd only need to memorize a few thousand of the most commonly used; those would be the first couple thousand choices. 0 for "the," 1 for the indefinite article (wouldn't make sense for it to be divided into "a" and "an"), 00 for "I," etc. It would still be a royal pain in the butt, but not quite as awful as it could be unless you're using high-school vocab words in everyday speech.

Possibly use a smaller number of "category" words to start. 0000 means the code following is a color, then 0 for black, 1 for red, 00 for orange, 01 for yellow and so on. 0001 means number follows, and then just use the binary for whatever number you want. 0010 is noun, followed by indicator for person, place, thing, or idea. Subdivide as necessary.

Organized that way, you might only have a few hundred syllables at most.

Orzel

2012-03-11, 10:32 PM

Xuc Xac

2012-03-12, 07:51 AM

Possibly use a smaller number of "category" words to start. 0000 means the code following is a color, then 0 for black, 1 for red, 00 for orange, 01 for yellow and so on. 0001 means number follows, and then just use the binary for whatever number you want. 0010 is noun, followed by indicator for person, place, thing, or idea. Subdivide as necessary.

Organized that way, you might only have a few hundred syllables at most.

I don't think you realize just how huge that is. Many Asian languages already use category words but they are already one syllable. What you propose is that we can somehow save time and syllables by saying "oh-oh-oh-oh-oh-one" instead of "màu vàng" or "oh-oh-oh-oh-oh" instead of "màu đen".

No matter what shortcuts you try to take, you can't catch up to the efficiency of other languages that use the same shortcuts in addition to several other methods of data compression.

Telonius

2012-03-12, 08:42 AM

I don't think you realize just how huge that is. Many Asian languages already use category words but they are already one syllable. What you propose is that we can somehow save time and syllables by saying "oh-oh-oh-oh-oh-one" instead of "màu vàng" or "oh-oh-oh-oh-oh" instead of "màu đen".

No matter what shortcuts you try to take, you can't catch up to the efficiency of other languages that use the same shortcuts in addition to several other methods of data compression.

No, what I propose is that doing this could save time and syllables over the "one number for every word" idea. It's still going to be inefficient and horrible. It's the only way I can think of that would even be remotely workable or even possible for a typical human being.

Xuc Xac

2012-03-12, 09:04 AM

No, what I propose is that doing this could save time and syllables over the "one number for every word" idea. It's still going to be inefficient and horrible. It's the only way I can think of that would even be remotely workable or even possible for a typical human being.

But it's still one number for every word. The only difference is that all the words in the same category will have the same prefix so if you arrange things in "alphabetical order", related vocabulary will be in groups. That would make the Binary to English dictionary look organized, but it doesn't change anything in actual practice.

pendell

2012-03-13, 09:37 AM

This conversation is beginning to look suspiciously like Morse code (http://en.wikipedia.org/wiki/Morse_code), which is also binary. It was invented for telegraph communication before voice comms were invented. Morse tried for a long time to invent a translation machine before he realized that ordinary humans can be trained to understand Morse by listening and to "speak" it with a telegraph key. For years humans around the world "spoke" to each other in Morse, first with telegraph and then with radio, until voice communication became practical. Since then it's mostly forgotten, except by boy scouts.

Respectfully,

Brian P.

factotum

2012-03-13, 10:56 AM

Morse is another example of a code that's really trinary, though--you have three states; short dash, long dash, and silence (used to insert spaces between words).

Tyndmyr

2012-03-13, 01:39 PM

Morse is another example of a code that's really trinary, though--you have three states; short dash, long dash, and silence (used to insert spaces between words).

The difference between them is fairly minor, though.

That said, machine encodings such as ASCII are pretty purely binary, though, if it's important.

razark

2012-03-13, 01:42 PM

The difference between them is fairly minor, though.
The difference between decimal and octal is fairly minor, as well.

Doesn't mean you can substitute one for the other and get the same results, though.

pendell

2012-03-13, 01:53 PM

Even so, I would expect a human language based on binary to be very similar to morse code. I suspect that it would not be somewhat difficult to use for the same reason we aren't writing these posts in morse code -- the fewer symbols in an alphabet, the more complicated an encoding scheme you need to string them together in an intelligible way and the more pain it is both to encode and to decode. The more symbols in your alphabet, the more efficient communication is.

Respectfully,

Brian P.

Drumbum42

2012-03-13, 08:49 PM

The more symbols in your alphabet, the more efficient communication is.

I'd have to disagree with that. The English language has ~1 million words, and using that logic the most efficient language would be one with 1 million symbols. Think of that keyboard, it would take a 40ft wall just to display them all at the same time.

Also, the issue of pauses making the language base 3 instead of base 2 opens an interesting door. People don't just say words to convey ideas, it's HOW we say them that gives them connotation. Changes of facial expression and pitch of your voice, as well as pauses. Does this binary language need to convey ALL communication or just sentences?

Is this your apple? General question.
Is this your apple? Show's surprise that the apple is the object in question.
Is this your apple? Shows surprise of the apple's owner.

The words are the same, but the meaning is different. So does the binary language need to convey ALL information, or is it just words on a page and connotation can be derived by the way things are said.
All information: true binary
Just words: written is true binary, spoken is base 3,4,5,100 depending on how many ways you can say the same thing with different meanings.

pendell

2012-03-14, 10:17 AM

Well, I had originally written a much longer post which noted that ideographic languages like Chinese have many more characters but are unwieldy because there are too many to easily memorize. Which, again, maybe why romance languages are the ones used in web communication. There have been attempts to build ideographic computer languages like APL (http://en.wikipedia.org/wiki/APL_%28programming_language%29) but they've never achieved general acceptance.

So possibly efficiency can be described as a parabola: Too few symbols and it's difficult to encode it properly, too many symbols and it's hard to memorize them all. The most common languages, perhaps, are used because they have enough symbols to avoid the encoding problems, but not so many that they are difficult to memorize.

I would say that any communication medium has to have the ability to communicate subtext as well as content. The lack of the ability to show facial expressions is exactly why there's so much mis-interpretation on these forums, and that's why smileys exist!

Respectfully,

Brian P.

Razanir

2012-03-14, 03:33 PM

First off, I still don't understand this debate of binary v trinary. The way I read it, this is intended to be a language with only two different syllables. Even though in computer terms, that's technically trinary, common language accepts that as binary.

When a rock is about to fall on you or a tiger is about to pounce on you, do you want your friends who see it to warn you with something like "move!","run!", or "look out!"? Or would you rather wait for them to finish something like "kokotakotakokotatatakotakotakokokokotatakotako!"?

I think we're looking at this wrong. Natural language has a much larger phonemic inventory and can afford more short words. This language has 14, defining "short words" as "3 or fewer syllables" For sake of argument, I'll use "foo" and "bar" as the syllables.

Monosyllabic words:

foo- safe
bar- danger

Disyllabic words:

foofoo- yes
barbar- no
barfoo- German "doch" or French "si" (yes when no is expected)
foobar- no when yes is expected

I won't bother listing the trisyllabic ones, but they could be personal pronouns- I, You, He, One, We, Y'all, They, People

Finally, it would likely be like Chinese where some concepts are contextual. Things like tense, case, number, etc.