Research indicates that individuals from Glasgow, Belfast, Dublin, and the north-east of England are more skilled at identifying when someone is mimicking their accent compared to those from London and Essex.
Among the tested groups, people from Belfast were the most adept at spotting fake accents, while those from London, Essex, and Bristol had the lowest success rates.
A study published today in Evolutionary Human Sciences showed that participants from Scotland, the north-east of England, Ireland, and Northern Ireland accurately recognized whether brief recordings of their accents were genuine or not between approximately 65% and 85% of the time. In contrast, individuals from Essex, London, and Bristol demonstrated success rates ranging from just over 50% to 75%, which is only slightly above random guessing.
This extensive research, which gathered responses from 12,000 individuals, revealed that participants across all regions could distinguish fake accents better than chance, achieving accuracy rates just above 60%. As expected, individuals who were native speakers in the accents being tested performed better than non-native listeners, some of whom failed to do better than random guessing, highlighting regional differences.
“We observed a significant discrepancy in the ability to detect faux accents among these regions,” explained Dr. Jonathan R Goodman, the lead author from the Leverhulme Centre for Human Evolutionary Studies at Cambridge, along with Cambridge Public Health.
“We believe that the capability to identify fake accents correlates with the cultural cohesion of an area, specifically how closely aligned its residents are in terms of cultural values.”
The researchers argue that the accents characteristic of Belfast, Glasgow, Dublin, and the north-east of England have developed culturally over centuries, during which conflicts arose, particularly with the cultural group from southeast England, primarily London.
This, they propose, has likely driven individuals from certain areas in Ireland and the north of the UK to prioritize their accents as key markers of their social identity.
The findings suggest that the high social cohesion in places like Belfast, Dublin, Glasgow, and the north-east may have fostered a stronger apprehension of cultural dilution from outsiders, thus enhancing their ability to recognize and respond to imitation of their accents.
Conversely, people from London and Essex struggled to identify fake accents possibly because these regions have less defined ‘cultural group boundaries,’ leading to exposure to a wider variety of accents, which may desensitize them to detecting imitation.
The study notes that many people speaking with an Essex accent have been newcomers from London over the last 25 years, while the accents of individuals in Belfast, Glasgow, and Dublin have been shaped over centuries amidst cultural strife.
Interestingly, one might have anticipated that residents of Bristol would have performed better at recognizing their accent. However, Goodman points out that “the city has become increasingly culturally diverse.” The researchers are interested in collecting more data specifically from Bristol.
An enhanced skill
Previous studies have indicated that to delineate their identity for cultural reasons, people’s accents tend to become more pronounced. Additionally, the capacity to identify and deter ‘free riders’ is believed to have played a crucial role in the evolution of large, complex societies.
According to Dr. Goodman, events like cultural, political, or even violent disputes often compel individuals to amplify their accents as they strive to preserve social unity through shared cultural traits. Even relatively minor disturbances, like the influx of tourists during vacation seasons, might have this influence.
“I am keen on exploring how trust develops in society. One of the first assessments individuals make about others, which informs their decision on whether to trust them, is based on speech. Understanding how humans learn to extend trust to someone who might not belong to their group has been fundamentally significant throughout our evolutionary past and remains essential today.”
Overall, while the study found that participants could detect fake accents at rates better than chance, it raises the question: Why did so many fail 40-50% of the time?
The authors clarify that participants were presented with only 2-3 second audio clips, making the accuracy rates of 70-85% achieved by some quite impressive. The researchers suspect that if participants had access to longer recordings or engaged in face-to-face interactions, the accuracy could increase but would continue to vary by region.
How the tests were implemented
The research team designed a series of sentences aimed at revealing phonetic differences among seven targeted accents: north-east England, Belfast, Dublin, Bristol, Glasgow, Essex, and Received Pronunciation (RP), which is recognized as standard British English. These accents were chosen to guarantee a diverse range of phonetic elements within the sentences.
Examples of test sentences included: ‘Hold up those two cooked tea bags’; ‘She kicked the goose hard with her foot’; ‘He thought a bath would make him happy’; ‘Jenny told him to face up to his weight’; and ‘Kit strutted across the room’.
The team initially engaged around 50 participants who naturally spoke these accents, asking them to record themselves reading the sentences in their authentic voice. Subsequently, the same individuals were asked to mimic sentences in other accents they naturally did not speak, chosen at random. Males mimicked male speakers, and females mimicked female speakers. The researchers selected the recordings that closely replicated the respective accents based on the reproduction of crucial phonetic characteristics.
Participants then listened to recordings made by others with identical accents, from both genders. Consequently, speakers of the Belfast accent evaluated recordings made by native Belfast speakers, alongside recordings imitating the Belfast accent by non-native speakers.
Participants were tasked with determining the authenticity of the recordings. Each participant assessed whether the speaker was a mimic for 12 recordings (six imitative and six genuine), presented in a random order, resulting in 618 responses collected.
In a subsequent phase, the researchers recruited over 900 participants from across the UK and Ireland, regardless of their native accents. This expanded the comparative control group and increased the sample size of native speakers. In this second phase, 11,672 responses were gathered.
“The UK provides a unique context for investigation,” Dr. Goodman noted. “The rich linguistic diversity and deep cultural heritage encompass numerous groups that have inhabited the same regions for extensive periods. Over time, very specific nuances in language, dialect, and accent have emerged, constituting a compelling aspect of language evolution.”
