“Speakers of some languages seem to rattle away at high speed like machine-guns, while other languages sound rather slow and plodding,” wrote linguist Peter Roach in 1998. A few months ago re­searchers systematically quantified Roach’s observation and offered a sur­prising explanation. Last year, in an issue of the journal Language, François Pel­legrino and his colleagues at the Univer­sity of Lyon in France published their analysis of the speech of 59 people read­ing the same 20 texts aloud in seven languages. They found Japanese and Spanish, often described as “fast lan­guages,” clocked the greatest number of syllables per second. The “slowest” language in the set was Mandarin, followed closely by German.

But the story does not end there. The researchers also calculated the infor­mation density for the syllables of each language by comparing them with an eighth language, Vietnamese, which served as an arbitrary reference. They found that an average Spanish syllable conveys only a small quantity of infor­mation, contributing just a fragment to the overall meaning of a sentence. In contrast, an individual Mandarin sylla­ble contains a much larger quantity of information, possibly because Mandarin syllables in­clude tones. The upshot is that Spanish and Mandarin actually convey information to listeners at about the same rate. The correlation between speech rate and information density held for five out of seven of the lan­guages studied, and the researchers conjectured that, despite the diversity of languages in the world, over time they all deliver a constant rate of information, possibly tuned to the human perceptual system.

The results of these studies could change the way we think about the di­versity of the world’s languages. In the 1950s linguist Noam Chomsky proposed the idea of universal grammar, which suggests that all languages, their ap­parent differences notwith­standing, possess a common set of abstract structures. This hypothesis galva­nized the field of linguistics, but truly com­mon structures proved tough to find. The current research suggests that languages can and do use a wide variety of structures, as long as they deliver information to listeners at a relatively constant rate. Thought of in this way, universal grammar is no longer an abstract notion but a linch­pin of human communication that ensures a steady flow of information from speaker to listener.

The article was published in print as "Fast Talkers."

16 Comments

1. 1. GG 11:00 AM 4/15/12

   If you bothered to learn Mandarin, you would have noticed the same thing that many others have said: in terms of grammar, Mandarin is a "baby language". There is not much grammatical correlation between words in a sentence. You don't have to conjugate any verb, you don't have to adjust for genders etc.
   
   The downside of the baby talk is that it does not encode a lot of information, and to make up for that you need to say more words for transferring the same amount of information. For example, you cannot say "went", because there is no past version of the "go" verb. To make up for this deficiency, you would need to say "go yesterday". And so on with everything else.
   
   The fact that a Mandarin syllables contain more information is upended by the lack of grammar.
   The result is that Mandarin speakers talk a lot, but they don't necessarily say a lot.
   
   German is very different. The complex grammar encodes a lot of information. German speakers can convey very complex facts and relationships in relatively short sentences. (they concatenate many words, which looks intimidating; but that's another issue)

   Reply | Report Abuse | Link to this

2. 2. conifer 12:29 PM 4/15/12

   I do not think we can believe the results based on an investigation of merely 59 people. That can not even get a proper average speed for one language in terms of the difference between areas using the same language.

   Reply | Report Abuse | Link to this

3. 3. promytius 08:14 PM 4/15/12

   59 people reading one set of words is a data point, not a study.

   Reply | Report Abuse | Link to this

4. 4. stargene 04:14 AM 4/16/12

   Just out of curiosity, what were the two studied
   languages which did not conform to the research
   result?

   Reply | Report Abuse | Link to this

5. 5. jgrosay 04:12 PM 4/16/12

   Brain is the best performing computer in the world, just its peripherals, way-in and way-out of information and actions are slower than Central Processing Unit. Thanks God, we have learned the way to produce prosthesis that improve our overall performance.

   Reply | Report Abuse | Link to this

6. 6. lionessofgd 04:14 PM 4/16/12

   This is very interesting - I wonder how fast Semitic languages are and if language speed in general has to do with the culture/lifestyle of a given linguistic population.
   
   As more of a visual person, I actually found this infographic on the Speed of Language to be pretty useful in understanding this concept. http://visual.ly/speed-language

   Reply | Report Abuse | Link to this

7. 7. bradster 08:32 AM 4/17/12

   Bull crap man, what about people who stutter?

   Reply | Report Abuse | Link to this

8. 8. JKendall 03:59 PM 4/17/12

   I browsed the original study[1] and found that the texts used for the other seven languages had all been freely translated from an original set of twenty British English source texts. Translation almost always involves loss and/or change of information (though it appears the source texts were chosen to minimize this effect). However, it wouldn't be unreasonable to consider the results biased in favor of British English given it never suffered from translation. Though it might require more work, they could achieve less bias by starting with an equal number of representative source texts from each language, translating them into the other languages to be compared before performing the main part of the study.
   
   1. Pellegrino et al. "A Cross-Language Perspective on Speech Information Rate." Language 87.3 (2011): 539-58.

   Reply | Report Abuse | Link to this

9. 9. CrassPip 06:18 PM 4/18/12

   The graphic for this article is wrong, isn't it?!
   The hare presumably is the fast talker, so he should have less dense information (the one big A), while the tortoise speaks slowly but with more meaning, ie the many small A's. The hare is Spanish or Japanese. The tortoise is Chinese.
   Of course the artist may have just read the misleading headline and been illustrating the so-called slow and fast languages. But in fact, the point of the paper seems to be the opposite- that when considered in totality, all languages encode about the same amount of information.

   Reply | Report Abuse | Link to this

10. 10. CrassPip in reply to GG 06:37 PM 4/18/12

    GG,
    Have you "bothered to learn Mandarin"? If so, I'm very surprised by your comment. First of all, Chinese's "baby" grammar results because most conjugation and declension in modern languages is redundant. If one says "yo soy americano" in Spanish (7 syllables), the "yo" is completely unnecessary since the verb already encodes the first person. The Mandarin "我是美國人" (5 syllables) does not need conjugation.
    When there would be ambiguity, Chinese usually relies on word order or context, and only occasionally requires clarifying words. Look at Classical Chinese. It is probably the most information-dense language ever known.
    Further, due in part to atonality most Spanish words are at least two syllables, whereas in Mandarin each phonetic syllable can be pronounced in one of five tones, so fewer phonemes are needed. Each syllable has meaning, and there are relatively very few words that have more than two syllables. The Spanish or Japanese speakers apparently make up for this lack of information density by speaking each syllable faster, making the total information conveyed similar. Still, when I hear native Chinese, it sounds quite fast to me! :-)

    Reply | Report Abuse | Link to this

11. 11. CrassPip in reply to CrassPip 06:40 PM 4/18/12

    * it killed my Chinese characters. It said, "wo shi mei guo ren" (pinyin syllables)

    Reply | Report Abuse | Link to this

12. 12. fhsu2007@gmail.com 01:11 AM 4/20/12

    Mandarin Chinese is an evolving language, and the current stage in its evolving way makes it sound like baby language. However, it is a very handy language, since it has no tense, and hence no burden of conjugation of verbs, no plurals, no genders, and most time the subject of topic is omitted -it becomes redundant- during the conversation. I believe Mandarin Chinese sounds slower than most other languages if it is spoken clearly and understandably, for it is limited to its basic four tones in pronunciation. Each tone can be shared with so many characters so if speaking too fast listeners may become confused in understanding the context. Another reason is that it is a monosyllabic language, and there is always a subtle pause between syllables and that slows down the speed.

    Reply | Report Abuse | Link to this

13. 13. priscianna 04:24 AM 4/22/12

    Not enough time to make many comments, so here are a few for your delectation.
    One aspect for evaluation appears to be missing -- precision.
    Chinese (as well as Hindi and Russian for example) does not have the distinction of A and THE. As others may have pointed out, there is no plural and no verb tenses.
    Some describe Chinese as 'telegram' speak.
    Long time no see = it has been a long time since I've seen you or since I saw you.
    As a writer of haiku (if keeping to 17 syllables), the longer words in Spanish make it more limiting (eg -ly is -mente). OTOH it doesn't need a subject personal pronoun.
    (Chinese of course can play more games with its pictographs and puns are easier in Japanese with so many similar sounds.)
    English takes great pains wrt time as in definite and indefinite past.
    We all know: Long time no see (it has been a long time since I've seen you).
    'Already' and 'yet' indicate a time continuum.
    From my observation in studying several languages, new words are born for new circumstances and obsolete ones die out (or are explained in the OED's word of the day :-) ).
    Observed by Horace (and better said):
    Men ever had, and ever will have leave,
    To coin new words well suited to the age,
    Words are like leaves, some wither every year,
    And every year a younger race succeeds.
    -- Horace, poet and satirist (65-8 BCE)
    Languages also have words that are 'more' apt than in others (which might take longer to express). In English one word is WRY. In Bahasa Indonesia one word is sudah.
    (British) English has borrowed oodles of words from its colonies making it all the richer. And this is on top of being a bastard language with the imposition of Norman French on Anglo Saxon so that English frequently has two words (each with a different connotation) whereas French and German just have one:
    labour/work; veracity/truth; flower/bloom/blossom; ~royal/kingly
    With international communication now there will undoubtedly be great words/expressions just too tempting not to use/incorporate.
    As Goethe said: "The strength of a language does not lie in rejecting what is foreign but in assimilating it."
    Every language has its strengths and weaknesses -- and its charm.
    And for me, its fascination.
    OTOH:
    Life is a foreign language; all men mispronounce it.
    -- Christopher Morley, writer (1890-1957)
    

    Reply | Report Abuse | Link to this

14. 14. fhsu2007@gmail.com in reply to CrassPip 11:15 PM 4/22/12

    How do you translate the question in English when you expect the answer is associated with nationality and is "Wo3 Shi4 Mei3 guo2 ren2?" It's not "Where are you from?" Or "what's your nationality?" Or "Are you American?" In Chinese we ask "Ni3 shi4 na3 guo2 ren2?" And I haven't figured out how to translate this sentence into English when I was teaching a class of college students Mandarin Chinese recently.

    Reply | Report Abuse | Link to this

15. 15. wwinde in reply to GG 10:54 PM 12/15/12

    Hi GG, I think you need to improve your Mandarin grammar. No one says "I go yesterday" for "I went". The correct translation should be "Wo qu le" for "I went". "Qu" is "go" and "le" is a particle to convert the meaning to past tense. For "I went yesterday", it should be "wo zuotian qu le". Mandarin contains high information density because it's concise instead of lack of grammar components (which is impossible). There are no conjugation, gender and plural form in Mandarin simply because they are almost useless and were abandoned when the language is evolving from ancient Chinese. For example, "He ate 3 tasty apples" in some languages can be very long. There are "gender" for apple (unnecessary and nonsense), different forms of "ate" for different subject (which is unnecessary because "he" is already mentioned in the sentence, unless you have no logic), plural form for "apple" (again, unnecessary because "3" already tells you that there are more than 1 apple), different forms for "tasty" according to the gender and the number of the subject (completely meaningless)...etc. In other words, Mandarin depends heavily on logics. Of course, there are also some useless grammatical components like these in Mandarin, but I guess the ratio is lower than the most languages in the world. I think another major reason why Mandarin can be shorter than many languages is having tones. Japanese lacks gender and plural form as well, but the language is much longer than Mandarin because the variety of sounds is smaller than Mandarin.

    Reply | Report Abuse | Link to this

16. 16. skyyrie in reply to GG 10:34 AM 4/12/13

    You feel a Germanic superiority towards Chinese speaker.IMHO,the Chinese don't need to attribute any grammatical gender to whatever a thing,e.g,Chinese don't need to think a chair is a female chair or a male chair,they think it is ridiculous to determine the sex an object,so ,the Chinese is more refined and elegant

    Reply | Report Abuse | Link to this