Why do I have an Accent and Should I be Ashamed of It?

By: Jingshu Helen Yao, Master of Museum Studies student / Summer Intern at the CLM

I have an accent and I’ve never liked it. I sometimes say “sank” instead of “thank” and “worm” instead of “warm”. If you say “had” and “head” without any context I won’t be able to tell the difference. Pronunciations are the coordination of our brain and muscles. Sometimes even if my brain knows what to do, my tongue and my facial muscles are just not used to moving that way. It is valid to feel pressure when speaking with an accent, since it is the first thing others notice when you start to talk; accents can be connected with negative impressions like “broken English”, or even “uneducated”. My language insecurity is one of the reasons that led me to linguistics in the hope that I could “fix” my English, but the study of language science took me on a very different journey.

Pitch, Spectrograms and Formants

Before diving into the details of speech sounds and their features, let’s get the terminologies out of the way. As the physicists suggest, sounds are waves. The way in which our mouths create such waves is similar to how musical instruments do. Our teeth, the roof of our mouth, and our tongues, are like keys or strings that determine the speech sound we produce. Their position and placement will change the sound produced by our vocal fold vibration, which becomes the dynamic speech we eventually hear.

Researchers in different academic fields have tried to visualize sounds in order to better study them. Sounds can be characterized by features such as pitch (frequency) and loudness (intensity/amplitude). A sound with a constant pitch (simple periodic sound) can be represented by a simple waveform as follows.

Sin curve showing periodic oscillation every 17 seconds.
Simple waveform (Image from ResearchGate)

However, human speech is very complex. It consists of waves at different frequency levels that add on to each other and can be difficult to visualize. Spectrograms are employed by acoustic phoneticians to analyze human speech.  A spectrum is a display that shows the intensity of each level of frequency and allows us to study the features more closely. Formants are dark lines that are formed in the areas with a high intensity, which can be seen most clearly in vowel sounds. Formants are represented with a red line in the following image.

Spectrogram of 1.475873 seconds of speech showing area of highest intensity in the centre of the speech
Insert formants. (Image from EdHUK)

For the purpose of this article, we will only be looking at the first and second formants of vowels. They indicate the height and frontness of our tongue position when we produce the sound.

Phonetic Differences Between Languages

Every language has a phonetic inventory, which consists of all sounds (phonemes) that are possible in the language. For example, the “th” sound in English is rather rare in other languages and makes it difficult for a second language speaker to master. Not being able to produce the exact phonemes is the top reason for having an accent.

While I am fully aware that I sound different from a standard Canadian English speaker, I was curious to find out what my vowels look like. Thus, I did a small test using Paart, a program developed by phonetic scientists from the University of Amsterdam, Paul Boersma and David Weenink, to create spectrograms from audio recordings.

I was relatively young when I learned English and lived in an English-speaking environment for several years. Therefore, I can sometimes distinguish and produce the sounds that are not in my first language: a notable pair is “a” as in “bad” and “e” as in “bed”. When I pay a lot of attention to my speech and consciously remind myself to differentiate the phonemes, I can distinguish the pronunciation of “a” and “e” in careful speech. However, in a natural conversation setting where I am simply trying to get my meaning across, my pronunciation of the two sounds tends to be very similar.

With Paart, I recorded three audios: in two of them, I carefully pronounced “bed” and “bad” respectively, and in the third audio, I put both words in one sentence “My bed is bad so it hurts my back.” Then I analyzed all the audios and noted down the values of the first and second formant of each vowel.

            Helen’s careful speech:

                        Bed: F1: 530 F2: 2041

                        Bad: F1: 751 F2: 1890

            Helen’s natural conversation:

                        Bed: F1: 589 F2: 1833

                        Bad: F1: 586 F2: 1756

            Standard Canadian English:

                        Bed: F1: 600 F2: 2930

                        Bad: F1: 860 F2: 1550

In addition, I noted the standard Canadian English value (according to research data from the University of Manitoba) as a reference. A higher number in F1 corresponds with a lower tongue position, while a higher number in F2 corresponds with the tongue further forward. The data show that when Canadian English speakers produce the ‘e’ sound in ‘bed’, the tongue is high in the mouth and forward; for the ‘a’ in bad, the tongue is pulled back and lower.

It is easy to tell from the data that both of my productions, in careful speech and in natural conversation, deviate from the standard Canadian pronunciations.  This suggests that I will always sound a little different no matter how hard I try. However, in careful speech, there is indeed a distinct difference between my ‘e’ and ‘a’ sounds, whereas in my natural speech, ‘e’ and ‘a’ are almost identical, especially with respect to the value of the first formant.

Accents are not avoidable when it comes to producing sounds that are not originally in our first language. Language learning is closely related to one’s age because the ability to produce new sounds decreases as one grows older. Having an accent doesn’t imply deficiency or unskillfulness but simply the fact that one’s muscles are not used to moving in a certain way. Speakers can be trained to change their pronunciation in careful speech, but accents are unlikely to be removed completely.

I started learning linguistics because I wanted to improve my English. However, instead of “fixing” my accent, I learned to consider it from a different point of view. In my studies I learned about the wide variety of sounds that are used in the languages around the world. Each individual language normally has an inventory of between 20-37 different sounds (UCLA database). Instead of struggling over the sounds that I am not capable of producing, I have learned to appreciate how diverse human languages are and how unique each language can be.

Click this link to the International Phonetic Alphabet to see and hear the different sounds in languages around the world.

International Phonetic Alphabet