Speech is a continuous flow of sound, but humans can remarkably separate individual words with great accuracy and speed. To explore how this works, a group of linguists examined how long consonants last in various positions within words and phrases across multiple languages. They discovered that consonants at the beginning of words are typically about 13 milliseconds longer than those in other positions. The variety of languages showing this phenomenon suggests it may be a trait common to the species, acting as one of several important elements that enable us to discern the start of words in spoken language.
Identifying words in spoken language is a significant challenge, yet humans make it appear effortless, even when languages do not clearly indicate where one word ends and another begins. The sound cues that facilitate this skill remain largely unclear and are often overlooked in most of the world’s languages. For the first time, linguists have identified an acoustic pattern that may serve as a recognizable marker across different languages: the consistent lengthening of consonants at the start of words.
Researchers from several institutions including the Max Planck Institute for Evolutionary Anthropology, CNRS Laboratoire Structure et Dynamique des Langues (SeDyL), Humboldt Universität zu Berlin, and the Leibniz-Centre General Linguistics (ZAS) utilized the unique DoReCo corpus for their study. This corpus incorporates extensive linguistic and cultural diversity, featuring samples from 51 populations across all inhabited continents, and it offers precise timing information for over a million speech sounds. “The comprehensive nature of the DoReCo corpus is vital in revealing species-wide patterns in human speech, particularly given the vast diversity of languages,” explains Frank Seifart, the senior author of the study, who is associated with CNRS in Paris and HU Berlin, as well as being a co-editor of DoReCo.
Could Word-Initial Consonant Lengthening Be a Universal Trait?
“Initially, we anticipated finding evidence that contradicted the idea of word-initial lengthening as a universal linguistic feature. We were quite surprised by our findings,” states Frederic Blum, the first author of the paper and a doctoral student at the Max Planck Institute for Evolutionary Anthropology, who spearheaded the research. “Our results indicate that this phenomenon is indeed prevalent in most languages around the globe.” Significant evidence of lengthening was found in 43 out of the 51 languages studied, while the remaining eight languages produced inconclusive results.
The researchers concluded that lengthening may be one of several elements that assist listeners in recognizing word boundaries, thereby segmenting speech into individual words. Other factors, such as articulatory strengthening, which has not been thoroughly examined across different languages, also play a role. Moreover, some languages in this study exhibited a shortening effect after pauses at the beginning of utterances, supporting the conclusion that no additional cues for identifying word boundaries are necessary when pauses are present.
This research enhances our comprehension of the acoustic processes that are common across all spoken languages. By emphasizing non-WEIRD (Western, European, Industrialized, Rich, and Democratic) languages, the researchers aim to expand our understanding of cognitive processes related to speech that are not limited to specific populations.
