French linguists analyzed the speech of native speakers of 17 different languages and have found that they can combine the speed of information transmission. To do this, scientists calculated the speed of speech 170 media and information density of one syllable of each of their languages and found that they all share the speed of information transmission in bits by 39.15 per second. This figure is likely can claim the role of a universal parameter common to all natural languages, write the scientists in the Science Advances.
In the entire history of human existence, a host (according to various estimates, from four to seven thousands) of different languages. To track their occurrence is most often only possible to a common language ancestor or proto-language: for example, for Indo-European languages the accepted language is considered proto-Indo-European. The ancestor, you can understand why and how languages from the same language family are similar.
Of course, there must be something that unites all languages, regardless of their families — a kind of universality inherent in all of them. Language is primarily a means of communication, and thus tracks the language of universality is needed in the first place to look for in a human voice. Last year scientists, for example, was able to detectthat the deceleration of articulation in front of nouns is characteristic of speakers of natural languages from nine different families.
Linguists under the guidance of françois Pellegrino (Charles Pellegrino) of Lyon University has decided to calculate the information transfer rate for one unit of speech in 17 languages of the representatives of Austro-Asiatic, Indo-European, Sino-Tibetan, Turkic, Ural, and tai-kadai families, as well as Basque, Japanese and Korean. As a language unit for the analysis the researchers chose the syllable. Selected languages significantly differ in the number of new syllables: for example, in English syllables, about seven thousand, while the Japanese — only a few hundred. While Japanese and English allow their bearers equally effectively to convey information to each other. It is interesting, therefore, to see how to adjust the information density in each syllable, and how it affects the speed of information transmission.
For this, the researchers asked 170 native speakers of 17 languages that were used to read 15 short semantically related texts: texts had to be read several times that in the final version of the participants did not get off and read the text clearly. Recording read the lyrics divided into syllables: to be clear, were not considered canonical for a particular word syllables out of dialect variants.
Next, the researchers calculated two parameters: speech rate (number of syllables uttered per second) and information density (how much information in bits carries one syllable in the language). The last parameter calculated on the written corpus of each language: from reflects how many bits of information each carries a separate syllable (if simply how much of the semantic information carries single syllable of the language based on the same content of texts from various languages).
Despite the fact that the languages differed both in information density and speed of speech, scientists have noticed a pattern: the less “dense” languages differ in rapid speech, while languages with larger information density of speech is slower. For example, in the Vietnamese language with a density of 8 bits per syllable rate of speech was about 5 syllables per second, while in the Finnish information density was about 5.5 bits per syllable, and the speed of speech is slightly more than 7 syllables per second. Given such linear relationship, the authors calculated that on average in all studied languages speed of information transfer (information density multiplied by the speed of the speech) is about 39.5 per bit per second.