Hummer songs have a pattern similar to human language.

Humpsham songs have statistical patterns in the structure and are very similar to what can be seen in human languages. This does not mean that songs conveys complex meanings such as our sentences, but whales suggest that human beings can learn songs in a similar way as human beings begin to understand language.

Male humpback whales (Megaptera novaeangliae) Singing and acting is thought to be important for attracting colleagues. The song is constantly developing, and new elements appear and the population spreads until the old song is completely replaced with a new element.

“We think that everyone needs to do the same task, but it is similar to a standardized test that can show more tasks than others to make change and decoration,” he says. Jenny Allen At Griffith University of Gold Coast, Australia.

Allen and her colleagues were looking for a natural rescue pattern found in human language instead of trying to find meaning in the song. They analyzed eight years of whale songs recorded around the New Caledonia in the Pacific Ocean.

The researchers have started and started creating a spiritual code code that represents all the songs of all recording, including a total of 150 unique sounds. Allen said, “By default, it is a different sound group, so you can have AAB because you can do Grunt Grunt Squeak for one year. And in another year, you can make a moan. ”Allen says.

When all the songs were encoded, the linguist team had to find out how to analyze a lot of data. The breakthrough occurred when researchers decided to use the analysis technology that applied to the method of finding the word conversion probability.

“The speech is continuous and there is no temporary suspension between words, so infants must find a word boundary.” Inbal Arnon At Jerusalem Hebrew University. “To do this, we use low -level statistical information. Specifically, if it is part of the same word, sound is likely to occur together. Infants are likely to use this dip to find a word boundary along the sound of one sound. ”

For example, in the phrase “pretty flowers”, the child intuitively recognizes that the syllable “Pre” and “TTY” are more likely to go together than “TTY” and “Flow”. Arnon said, “If whale songs have a similar statistical structure, this signal will be useful for classifying it.

Using the speech version of the whale song, the team calculated the probability of transition between continuous sound elements, and the next sound element was surprisingly cut.

Arnon said, “This cut divides the song into a segmented subsequent. “We saw their distribution and surprisingly found that we followed the same distribution as found in all human languages.”

In this pattern, called Zipfian distribution, the prevalence of less common words falls in a predictable way. Another noticeable discovery is that the most common whale sound is the most common human horse, like the abbreviation of Zipf, tends to be short.

Nick Enfield Sydney University, which has not participated in this study, says it is a new way to analyze whale songs. “That means is an analysis War and peaceThe most frequent words will be twice as frequent than the following: The researchers identified similar patterns in whale songs, ”he says.

Team member Simon Kirby The University of Edinburgh said that this method did not think it was effective. “I will never forget the moment when the graph appears to know in human language,” he says. “This realized that we found deep in common between these two species, which were divided into tens of millions of years of evolution.”

But the researchers emphasize that this statistical pattern does not lead to the conclusion that whale songs convey the meaning we can understand. They suggest that the reason for the in common is that whale songs and human languages ​​are culturally learned.

Enfield said, “The physical distribution of words and sound in language is a fascinating feature, but there are millions of languages ​​about languages ​​that are completely different from whale songs.

In a separate study Publication this week, Mason Young Black At Stony Brook University in New York, other marine mammals have found that they can have structural similarities with human language in communication.

Menzerath’s law, which predicts that a sentence with more words should be composed of short words, was present in 11 out of 11 of the 16 copies studied. ZIPF’s abbreviations were found in two out of five, and the available data can be detected.

Youngblood said, “According to our study, the humpback whale song has developed more efficiently and easier to learn, and this feature suggests that this function can be found at the level of stationery and stationery within the song.

“Importantly, the evolution of these songs is biological and cultural. Some features such as the law of Menzerath can appear through the biological evolution of the vocal device, while other functions such as the ranking of ZIPF-frequency techniques are [the Zipfian distribution]The cultural transmission of individual songs may be needed, ”he says.