What language says the most with the least?

[Malice]

I just got back from my honeymoon in France and Spain where I was exposed to numerous languages from all over Europe. I noticed that on numerous occasions an idea expressed in few words in English took many more words and/or syllables to convey the same message in, for example, French.

While no doubt the reverse is also true, it got me wondering:

*What is the most "efficient" spoken language?*

Is there a spoken language that conveys more meaning in fewer phonemes or morphemes than any other?

Is there a language that conveys *less* meaning in a given string of sounds than any other (speech used by politicians excepted).

[SuaSponte]

Pig Latin.

Sua

[Una Persson]

I was of the opinion that among "common" languages, that is, not obscure ones, it would be Latin (not Pig Latin, shame on you Sua!). Every time I see a Latin phrase translated to English it seems that so many more English words and syllables are required to express the same thought.

[FloChi]

Well it sort of depends. I do know that patents people like writing stuff in German. They write one sentence in German whereas it would take at least a paragraph to describe the same thing (and all its permutations) in English. It is because they string words together (occasionally with verbs) and the order of the words denotes signifigance, etc.

But it still sounds like a mouthful when spoken.

Perhaps Finnish (maybe it is just because they don't talk?)

And Latin is rarely spoken anymore, so it wouldn't have mattered too much.

[Spritle]

I'm voting for either Turkish (is that the one where the verbs aren't conjugated, just modified with words like "tomorrow" or "Yesterday"?) or good Sign Language (either ASL or one of the international varieties). After years of study, I've always been impressed with how much is conveyed with so little movement.

Of course, this goes out of the boundaries set by OP.

handily,
Spritle

[Johanna]

Perhaps Chinese. Compare some translations of the Tao Te Ching, especially the Shambhala edition that includes the original Chinese text, and you'll see what a large amount of English verbiage is necessary to translate a few extremely elliptical Chinese characters.

[muppetsoup]

Chicken (http://www.plif.com/archive/wc072.gif)

[yabob]

Well it sort of depends. I do know that patents people like writing stuff in German. They write one sentence in German whereas it would take at least a paragraph to describe the same thing (and all its permutations) in English. It is because they string words together (occasionally with verbs) and the order of the words denotes signifigance, etc.
That's rather interesting - people that translate software always seem to comment that German takes up the most real estate on the screen. An aggravating factor is that because of the long words, word breaks tend to be awkwardly displayed.

Certainly, if you take product manuals which have been translated into multiple languages, the English presentation often seems to be the most compact, at least with reference to languages using the Roman alphabet. How much of that is a function of the original text having been done in English and having been translated into the others, I don't know. It does seem that the German translation tends to take more room than the French, English or Spanish.

[barbitu8]

Poetry.

[barstow]

Hawaiian has the fewest sounds.
The local joke is there are only 12 letters and 10 are vowels.

[yabob]

Perhaps Chinese. Compare some translations of the Tao Te Ching, especially the Shambhala edition that includes the original Chinese text, and you'll see what a large amount of English verbiage is necessary to translate a few extremely elliptical Chinese characters.

In terms of written language, I will certainly agree with this. You gain some visual compaction by having thousands of symbols. Any phonetic system is probably going to have to expand a bit visually over ideograms.

I might go for Chinese or one of the Southeast Asian languages for most compact spoken form as well, because those languages are tonal. Adding tonality should allow you to compact a bit more information per unit time into the spoken word, other factors being equal - just a WAG.

[Arken]

There was an invented language called 'Allnoun' which (big surprise) is all nouns. All other verbiage is expressed very simply with punctuation. It's extremely efficient. I don't have a URL for it, but I know there are a couple of websites out there.

[yabob]

Hawaiian has the fewest sounds. The local joke is there are only 12 letters and 10 are vowels.

18 phonemes, IIRC, and repeated vowel sounds. Which tends to make utterances run long - fewer combinations of phonemes for short words.

If we want to get strange:

How about Loglan? The low phoneme count reminded me of it. The radically wierd (but logical) grammatical architecture of Loglan, and the idea of reserving the "short word" space for operator words seems to make it very compact, in spite of a very minimal phoneme set.

[Johanna]

Spritle, Turkish verbs are in fact conjugated for person, number, tense, and mood. Turkish does this through agglutination, which means sticking the pieces that do the grammatical work onto the end of the word, without changing the base word.

You may be thinking of isolating languages, which do not change the morphology of words. Thai, Vietnamese, Malay, Chinese are examples of languages that work like that.

[Chas.E]

Let me suggest a different measure of "efficiency." Linguists sometimes measure a language's "redundancy." This is a technical measurement of how much unnecessary fluff is used in a language. Allow me to demonstrate. Here is a sentence with low redundancy (let's assume it is a real language).

f u cn rd ths, u ar prbly a lnguist.

And now the higher redundancy version:

If you can read this, you are probably a linguist.

The first version contains fewer redundant elements. In the second version, more redundant letters support the correct decoding of the word. If there are any elements in the words that can be eliminated without reducing comprehension, these elements are considered redundant. A "redundancy ratio" is a measure of the effiency of a language, on an abstract level.

Anyway, linguists have measured each language for the average redundancy, but alas, I don't remember where I saw the list of each language. I recall that English is sorta in the middle of the pack. Maybe some professional linguist can enlighten us with more redundancy info. Either that, or perhaps we could consult someone from the Department of Redundancy Department.

[jmullaney]

Originally posted by Anthracite
Every time I see a Latin phrase translated to English it seems that so many more English words and syllables are required to express the same thought.

Probably because the Latin phrases that have stuck around already porvay more with less. There's a good latin phrase for this logical flaw, but I can't recall what it is! :(

My bet is on Yiddish.

[Inky-]

I have to go with Chinese too because there are many different complicated inflections that can be used with each word.

As I understand it the inflections are catagorized as hot, cold, acending, decending and still (and I think there's one more) and each sentance takes additional meaning from any one or a combination of them.

[Turpentine]

I thought it was Newspeak.

[Arken]

Turpentine said:
I thought it was Newspeak.

Well you were doubleplus wrong then, weren't you? ;)

[wolfman]

I'd have to say computer languages are by far the most commpact in written form. Various groups of people have spoken representations for the symbols that are very efficient too, but there is no standardization.

ie. n>1? a=2: a=3
vs If n is greater than 1, then set 'a' to 2. If it is not, then set 'a' to 3.

[Maeglin]

Probably because the Latin phrases that have stuck around already porvay more with less.

While it is true that generally only pithy Latin sayings have stuck around, it is also true that Latin is an extremely concise and economical language. Ancient Greek is as well, though perhaps to a lesser degree due to handy little particles strewn about sentences to aid the reader in understanding the logical flow. Latin contains few such niceties.

MR

[bibliophage]

Contrary to the OP's experience, I think it takes more words and syllables to convey the same idea in French than in English. When I go to Canada, I often see notices in both languages side-by-side, and almost invariably the Fench version is longer.

Certainly Latin can convey an idea in fewer words than English, but Latin words tend to have more letters and more syllables than English ones.

I should add that redundancy in language isn't always a bad thing. Imagine if you're talking to someone in a noisy room, or that you're on a phone with a bad connection, and you can only hear half the words. In a language with high redundancy, you will be able to understand the idea anyway.

[malaka]

I'd wager that Manage-speak says the least while using the most words. :)

[ElvisL1ves]

Originally posted by Inky-
I have to go with Chinese too because there are many different complicated inflections that can be used with each word.

As I understand it the inflections are catagorized as hot, cold, acending, decending and still (and I think there's one more) and each sentance takes additional meaning from any one or a combination of them.

You may be thinking of Cantonese. Mandarin only has 4 intonations for vowel sounds: level, rising, falling-then-rising, and falling. Further, words are either one syllable long, or made of compounds of single-syllable words (each ideogram represents one syllable). There are no inflections (endings), no articles, not even real tenses (you convey past or future by specifying the time of the action). I don't think there are even any irregular constructions. There is a limited menu of initial and final sounds for each syllable-word, and not all possible combinations are even used.

The downside is that it is far more difficult to convey shades of meaning with such a limited vocabulary and highly-restricted grammar rules. Much meaning has to be inferred rather than stated. English, despite its often-nonsensical spellings and frequent irregularities, is far more useful even if nobody would call it compact.

[Sofa King]

I can back up FloChi on the German language's ability to create highly descriptive technical terms. I remember once learning the original German word for what we call "tanks." It was about forty letters long, and translated to something like:

"barbed-wire crushing, trench-crossing, heavy machine gun carrying, armored, tracked, self-propelled fighting vehicle."

It was a cumbersome word, but it conveyed a lot more information than "tank," which conveys no information about the vehicle at all. In fact "tank" is a ruse, designed to make one think that it is not a vehicle at all.

Then again, the Germans decided to can the original term and went instead for "panther."

[Derleth]

wolfman:
Is that pseudocode, or a specific language? It looks like a Basic dialect.

The link to Allnoun is: http://world.std.com/~tob/allnoun.htm
The maintainer no longer maintains the site, but the language still exists on the pages.

[Bobort]

I remember reading in a cryptography text somewhere that each letter of English text only contains about 1.3 bits of information (on average). Since English text is usually rendered in 7-bit characters, it should in principle be possible to devise a written form of english that is ~5 times more compact than what we have now. I'd be interested to see what this figure is for other languages.

[handy]

Sign language. Duh. One sign is worth a thousand words as you can give the emotional content, direction & a bunch of other cool things of a word as you say it.

[ianzin]

Perhaps we can usefully distinguish between related questions: e.g. which is the most efficient spoken language, the most efficient written language, and the most efficient notation?

In a sense, there is no way to compare like with like, since all languages involve a trade-off between rapidity and detail. It's easier to convey a given sentiment more quickly IF you omit some details which you trust the native speaker will simply infer from the context.

I used to make commercial video films with English commentary that subsequently needed to be translated into several different languages. French came out sorta okay, but German and Mandarin caused hours of extra editing to expand the duration of some scenes to accommodate the commentary.

What's also interesting is that of course some languages are simply spoken at a faster average speed than others. WIth the video scripts mentioned above, the Mandarin translation often LOOKED as short as the English script on the printed page, but it simply is not spoken as quickly. Native Italians are not trying to show off their incredible vocal dexterity - it's just a language which has evolved with rapid speech patterns. However, for natural conversational pace it is left standing in its tracks by something like Serbian. A perfectly natural chat between two native Serbian speakers can sound like two Uzis going off. They also seem to breathe through their ears, since a Serbian 'sentence' seems to go on more or less indefinitely.

I once worked with a Swiss colleague who had grown up speaking about 4 or 5 languages fluently, and could get by in a few others. She told me that for easy, efficient commmunication she always preferred Spanish, both for its relative grammatical simplicity and its vocal flow. She found English okay, but a little 'clunky' and lacking in flow, and of course riled at its weird, incomprehensible spelling (she was referring to UK English).

[bibliophage]

Originally posted by Bobort
I remember reading in a cryptography text somewhere that each letter of English text only contains about 1.3 bits of information (on average). Since English text is usually rendered in 7-bit characters, it should in principle be possible to devise a written form of english that is ~5 times more compact than what we have now. I'd be interested to see what this figure is for other languages. I do remember seeing an analysis something like this in a book I read. I can't really remember, but it may have have been in a book called Grammatical Man.

[HeyHomie]

Originally posted by Malice
Is there a language that conveys *less* meaning in a given string of sounds than any other?

Well, there's always Entish, wherein it might take a good 4 hours for a group of Ents to say "Good morning." But since ficitious languages are beyond the boundaries of the OP, my money is on any of the Aleutian languages. Not only are many words multisyllabic, but there are many, many words to describe any one concept, and the language is [insert some word that describes a tendency to make it up as you go along]. Cecil did a column on it a while back. I'll see if I can find it...

[HeyHomie]

Originally posted by rastahomie
Cecil did a column on it a while back. I'll see if I can find it...

Found it. (http://www.straightdope.com/classics/a1_297.html)

[Lightkeeper]

Originally posted by Malice
Is there a spoken language that conveys more meaning in fewer phonemes or morphemes than any other?

I could be wrong, but I'd vote for (are you ready for this?) the Arabic language. I've heard somewhere that it's referred to as a "high context language" due to this property.

Example 1: "If you ask for explanation, will they understand it?" can be said in 3 words.

Example 2: "Did you guys do it?" can be said in 1 word!

Example 3: The opening verse of the Qur'an is 4 words in Arabic. Saying the same in English would take 11 words.

Note: I'm referring to classical or "formal" Arabic, not the spoken/conversational dialects.

Of course, there must be some other high-context languages with similar properties.

[MrDeath]

Originally posted by jmullaney
Originally posted by Anthracite
Every time I see a Latin phrase translated to English it seems that so many more English words and syllables are required to express the same thought.

Probably because the Latin phrases that have stuck around already porvay more with less. There's a good latin phrase for this logical flaw, but I can't recall what it is! :(


Post hoc ergo propter hoc?

[schimmel]

<<Then again, the Germans decided to can the original term and went instead for "panther." >>

Panzer = armour: Panzerwagen = armoured vehicle

But "panther" is more interesting

[jmullaney]

Originally posted by MrDeath
Post hoc ergo propter hoc?

Yeah, I think that is it. After [the fact], therefore because of it.

[Johanna]

I'd vote for (are you ready for this?) the Arabic language. I've heard somewhere that it's referred to as a "high context language" due to this property.

Example 1: "If you ask for explanation, will they understand it?" can be said in 3 words.

Example 2: "Did you guys do it?" can be said in 1 word!

Example 3: The opening verse of the Qur'an is 4 words in Arabic. Saying the same in
English would take 11 words.

Note: I'm referring to classical or "formal" Arabic, not the spoken/conversational dialects.

Your example 3 is: Bism Allâh al-Rahmân al-Rahîm 'In the name of God the most gracious, the most merciful'.

A ratio of 4:11 = .36 --
I can top that: a ratio of .125 -- one word in the Qur'ân that takes eight words to translate in English:

fasayakfîkahumullâh 'But God will suffice thee as against them'. (verse 2:137)

So classical Arabic can verge on the polysynthetic structure of Eskimo or some American Indian languages: a whole sentence in one word.

[yabob]

Arabic sounds good to me, I guess. If I may, then, does anybody have any idea of how "high context" loglan or lojban is?

(background - loglan is an artificial language originally created to test the "Sapir Whorf" hypothesis, and as such, is an attempt to create an extremely logical language with no grammatical ambiguity and no difference between written and spoken forms. Its grammar is like nothing even remotely resembling any naturally evolving language (predicate logic based). What I find interesting is that it is a "speakable" thing which can be machine parsed, and could lead to practical natural language dialog.)

[dougie_monty]

In The Art of Plain Talk (1944), Dr. Rudolph Fleisch, who also wrote Why Johnny Can't Read, noted how efficent Chinese is with this comparison to English:
"...sign, meaning 'a mark.'
'(...)
"insignificance, meaning 'the making of no mark.'
"You have added no meaning but four empty syllables. Now you can be serious and philosophical and speak of the insignificance of man. A Chinese would say something about Man no mark.
Is that efficient or what?

[Lightkeeper]

Originally posted by yabob
...an attempt to create an extremely logical language with no grammatical ambiguity and no difference between written and spoken forms.[/B]

This seems very interesting. Do you have reference or a link to a page?

One thing I might add about Arabic, words are written the same way they are pronounced. If some word is written correctly/completely, and you can read properly, there's no way you can mispronounce it.

[sarteve]

'sign' language says the most with no words spoken!

[Astroboy14]

My fiancee and I have this licked! [Don Juan moment] We speak the language of love![/Don Juan]

[TQMshirt]

The arabic answer looks intriguing especially since I was thinking along the lines of hebrew.

"In the beginning" = one word

"I got dressed" = one word

"And to the dog"= one word

There is a whole lot of that. Mainly because there are a hole set of suffixes and prefixes that can be added to a word root to give it all sorts of meanings.

There are prefixes for
"to"
"in"
"and"
"the"
"like"
"from"
"of/that"

and suffixes for

possessive
plural



These prefixes and suffixes can be and are often used numerous times in one word. There are also grammatical methods to indicate the direction of action (to myself or to you).

So I can say "I said to him" in two words.
and "is in the hands of" in one word ("B'yday")

Oh yeah, and "did you guys do it" = one word ("Asitem?")

Hebrew also generally leaves out such conjunctions as "is" such that a literal translation of Hebrew would sound like "This not mine" etc....

I noticed this because when I translate Hebrew in my rabbinic duties it really stretches.

[yabob]

Referring to loglan / lojban:

This seems very interesting. Do you have reference or a link to a page?
Try this:

http://www.lojban.org/