Loading player
Quantitative linguistics is a branch of general linguistics that studies the structure of languages through mathematical and statistical methods and concepts. It allows us to analyze and classify them from different aspects, but which is the “fastest” language is a question that does not have a univocal answer: it depends on the value taken into consideration to understand the speed and measure it. It can be the quantity of syllables on average pronounced by speakers in a certain unit of time, for example, or the quantity of information transmitted in the same unit of time.
The issue is complicated by the fact that, although it is a universal human trait, verbal language has few indisputably universal characteristics. Each language belonging to a different linguistic family differs from the others in ways and structures, and comparing them is an operation possible only within certain limits. In 2019 a group of researchers from the French language dynamics laboratory of the Université Lumière-Lyon-II published in the journal Science Advances a study that compared 17 languages from nine different Eurasian language families (Vietnamese, Basque, Catalan, German, English, French, Italian, Spanish, Serbian, Japanese, Korean, Mandarin, Cantonese, Thai, Turkish, Finnish and Hungarian) according to various parameters .
The analyzed sample is extremely diverse in terms of linguistic and typological characteristics, the research team wrote. From a phonological point of view, the 17 languages vary greatly in terms of the number of phonemes, i.e. the minimal individual sounds with distinctive features in a linguistic system: from 25 phonemes in Japanese and Spanish to over 40 in English and Thai. But they also vary in the number of distinct syllables (from a few hundred in Japanese to almost 7 thousand in English) and in tonal complexity in tonal languages (those in which tone variations – up to six, in the sample analyzed – can change the meaning of a word).
The study took into consideration the quantity of syllables pronounced per unit of time, counted according to the canonical pronunciation attested in dictionaries. For example, the English word probably (“probably”) was counted as a three-syllable word, although some pronounce it as a two-syllable word (“probably”). The researchers then collected recordings of 170 native adult speakers of the 17 languages, in which each speaker read a standardized series of semantically similar texts (for a total of approximately 240 thousand syllables, considering all the texts). Before being recorded, each participant had time to become familiar with the texts by reading them several times, in order to understand them and minimize reading errors.
In terms of number of syllables spoken per second, the fastest languages in the analyzed group were Japanese, Spanish and Basque, followed by Finnish, Italian, Serbian and Korean. The slowest were Thai, Vietnamese and Cantonese. The top three fastest languages all have only five vowels, while the three slowest have over 20 and are all tonal languages. As he explained to the site Atlas Obscura French phonetics professor François Pellegrino, one of the authors of the study, the quantity of syllables pronounced per second is only one of the possible ways of defining the speed of a language, not necessarily the most significant. Regardless of the articulatory aspects, there are subjective evaluations, for example, for which a certain unknown language spoken by a native speaker may sound more or less fast to the ears of the listener.
– Read also: The prospect of a world in which foreign languages are no longer studied
Then there is a variable that is very difficult both to define and to measure, but which according to Pellegrino also has to do with speed: the “information density”, that is, the quantity of information transmitted per second. A definition of density used in quantitative linguistics, which refers to notions developed by the American engineer and mathematician Claude Shannon, concerns precisely the speed with which a listener can reduce his uncertainty regarding the remaining part of the message as he receives it: speed variable depending on the language of the message. Measuring it requires a series of complex calculations of the number of possible syllables in a language, the relative frequency of each of those syllables, and the probability that a certain syllable will be followed by another.
From the study by the Université Lumière-Lyon-II, based on this type of calculations, an inverse correlation emerged between the quantity of syllables and the quantity of information that can be put together in a unit of time. Japanese, for example, has a very high number of syllables pronounced per second but also a very low degree of complexity of its syllables, that is, much less information encoded per syllable compared to other languages. It means that the syllables in Japanese flow much faster, but many more are needed to convey the same amount of information compared to a “slow” language, such as Vietnamese (one of the tonal languages).
In some ways it is possible to argue that English, for example, is a more efficient language than Japanese. Japanese syllables mostly contain a consonant followed by a vowel, such as ko, but English – which also has only five letters to represent vowels, like Japanese – has about twenty different vowel sounds. By using the “a” in different positions you can get the vowel sounds of cat, can, dog, calm and many others. English also has very complex syllables that include large consonant clusters, as in strength. In Vietnamese, the greater complexity of the syllables compared to Japanese is mainly due to tonal variations, whereby the melody or the height of a syllable can take on value.
Information density is a relevant parameter for speed because, in general, the greater the possible complexity of a syllable, the greater the amount of information it can carry. It makes sense to say that the speed of Japanese – about 12 syllables per second – is faster than that of English, but not that it is a faster language overall, since other languages can convey the same amount of information with much less syllables.
– Read also: Where invented languages come from
The question of information quantity and efficiency is also very thorny, because languages are complex, inconsistent and redundant phenomena. A direct translation of the English phrase I am in Spanish it would be “Yo soy”, for example, but “yo” is not necessary – Spanish allows, like Italian, the omission of the subject – and is usually omitted: it is one syllable less. To translate “I’m hungry” in English we would say I am hungrywhich however leaves out a piece of information: that relating to grammatical gender, for which a more complete translation should be something like I, a man, am hungrywhich has more syllables.
In Paama, a language spoken on the island of Vanuatu, possessive adjectives can include information about the relationship between the speaker and the object, which differs depending on the context. In an expression that can be translated as “my coconut”, for example, the word used for “mine” could mean “that I intend to eat”, or “that I have grown”, or “that I intend to keep in my family for non-food use”. ”. Even Paama could therefore be said to be a fast language, if a single word can have meanings that appear so different to us.
One of the limitations of the study conducted by Pellegrino and the other researchers – in addition to the limited definition of “information”, taken from Shannon’s theories – is the bias of the sample analyzed. Of the approximately 7 thousand existing languages, it did not include some very important ones such as Arabic and Swahili, spoken by billions of people. The results of the study, however, allowed us to notice how in general languages are equally efficient in transmitting information, each through a different compromise between complexity and speed.
People who spoke less complex languages pronounced more syllables per second, while those who spoke a more complex language seemed to have to make more effort to articulate the syllables, and therefore spoke fewer. However, information was transmitted at more or less the same rate in each case. «It’s like the wings of a bird. There are the large ones, which require a few beats per second, and the small ones, which instead have to beat much more, but the result in terms of flight is practically the same”, he told theEconomist cognitive scientist Christophe Coupé, one of the authors of the study.
