On Accents, Founders and the Rest of the World
I am a startup founder, and I have an accent. This per se would have been enough for me to follow the debate around Paul Graham’s latest post ‘Founders’ Accents’ with much interest, but there’s more to it.
As a matter of fact, the problem that PG describes — the negative correlation between heavy accents and business success — is one of the reasons why I decided to launch my first startup, MonoLibre, which was a finalist at TNW Conference Latin America’s Startup Rally in São Paulo last week, and whose private beta we are about to release.
As a native French speaker, working on MonoLibre was very much a case of scratching my own itch, and the same goes for my Spanish-born co-founder and CTO, Guillem Pérez: we both speak English, but our fairly strong accents can make it difficult for us to be understood. This inspired us to create MonoLibre, whose goal is to help users “reduce” their foreign accents.
Let me clear this out first: like PG, I am only referring to accents that are strong enough to hinder intelligibility, in the same way that a speech impediment would. A lot of this controversy could have been avoided if people had made the effort to read his original quote in full:
“You can sound like you’re from Russia. It’s just fine, as long as everyone can understand you,” he told the New York Times a few weeks ago.
When accents and diversity don’t mix
It was somewhat disheartening to see some voices equate PG’s fact-based observation with allegedly rampant racism and discrimination. It feels unfair to the US, a country of immigrants that has been quite accepting of accents throughout history. In my home country, France, I doubt that so many people with a strong accent would get hired as professors and TAs, despite the fact that French pronunciation is hard for foreigners to master.
Even Mark Suster, whom I always enjoy reading, suggested that PG’s view ignored the importance of diversity. I see it slightly differently, and I would argue that the more diversity there is in the room, the more founders’ accents are likely to get in the way.
As a former tech blogger for TNW, I have attended my fair share of Demo Days, listening to pitch after pitch for hours — the very same setting that sparked PG’s comment — and I perfectly remember feeling my non-native English speaking brain unintentionally tune out when someone with a strong accent took over the stage.
If I noticed that issue, it is precisely because I made the conscious effort to try as hard as I could to understand that person. However, I am not sure every investor in the room did the same. I also know that fellow non-US startup founders reported how difficult it was for them to understand pitches delivered with an accent they weren’t familiar with — and that often included regional accents from native English speakers.
In that sense, Eran Galperin made an interesting comment when he stated that “an accent can be an advantage” when it pushes self-conscious founders to simplify and clarify their speech — maybe even more than a native would. Still, this is more the exception than the rule, and I am tempted to say those who manage to do this have already reached the level I hope MonoLibre’s users will aim for.
While too many players in the accent training space promise that users will achieve “accent-free” speech, my co-founders and I are being careful not to do so. On one hand, we don’t want to commit to something we may not be able to deliver to all users; on the other hand, accents are part of our identities, and there is no point in making them disappear as long as speakers are understandable.
Accent training can work
I was surprised to see how many commenters believed that adults can’t even hope to improve their pronunciation in a foreign language. Research shows that even grown-ups never stop learning, and that also applies to language skills. As the New Scientist reported earlier this year, “a study by Yang Zhang at the University of Minnesota in Minneapolis that focused on the acquisition of foreign accents in adults suggests we may simply be suffering from poor tuition.”
There are many reasons why people who study and practice a language for years still struggle with pronunciation. In my experience, lackluster teaching methods are usually to blame, in addition to the fact that our speaking skills are rarely tested in exams. This reinforces our own bias as learners: we typically spend way more time on grammar and vocabulary acquisition than on phonetics, and neglect to make fluent communication our main goal.
I am not saying that improving one’s accent is easy, and the training it requires is different from mere immersion or from traditional English classes, which means that people who haven’t done it yet aren’t necessarily “clueless”. In short, it requires us to work simultaneously on two core skills: listening and speech. To pronounce correctly, we first need to be able to hear the difference between words that sound very similar to us due to the influence of our mother tongue (for instance, “sink” and “think” for French speakers).
A few years ago while living in Brazil, I decided to do some speech therapy sessions to improve my pronunciation in Brazilian Portuguese, mostly out of curiosity (I have always been interested in language learning). I don’t regret this decision, as it helped me reach a level of fluency I haven’t attained in the other languages I speak. The process was also fairly quick, mostly because I had been lucky enough to get a quick grasp of hearing the language’s main sounds. Still, it took quite a few appointments — and a fair amount of money — for me to master uttering them.
While accent coaches all over the world are helping a certain number of people become better speakers, I am sure that technology can help us do the same in a more scalable, affordable and convenient way. For instance, MonoLibre is using a custom speech recognition algorithm that our CTO has developed to automate the feedback users will receive on their pronunciation.
Well beyond founders
Interestingly, the issue PG pointed out is anything but a first-world problem. While Demo Days are the brainchild of Silicon Valley’s tech elite, the need for foreigners to be understood is as great in any selling and/or competitive situation.
In other words, what is true of a Y Combinator founder is just as relevant to an unemployed Spanish computer engineering graduate doing phone interviews for his dream job abroad; or to a young professional in Central America, India, the Philippines or any emerging country who is hoping to get hired by a call center.
It’s with all these use cases in mind that I realized that MonoLibre would have to be very different from our competitors. For instance, mobile apps enabling bite-size learning are one of our top priorities. More importantly, we knew it was important for our solution to be as cheap as possible, and came up with a business model that lets us offer our service for free to end users.
How will we do that? By monetizing the big data on accented speech that our users will generate while practicing. This information is highly relevant to companies involved in speech recognition, whose current products struggle to understand non-native speakers (yes, I am talking about you, Siri!).
As voice interaction becomes a growing part of the way we interact with our devices, it is crucial for speech-related solutions to be able to understand anyone.
No matter how deeply I am convinced that accent training can work, I am also fully aware that many foreign-born people don’t have the time, resources or motivation to go through this process. This doesn’t mean that they should be left out of this (r)evolution, and we are hopeful that our approach will help tech companies become more inclusive.
This aspect is perhaps what I love the most about our entire project, because it takes it to a whole new level. Not only will it have a positive impact on actively engaged users, who may get access to better opportunities in life by improving their pronunciation, but it will also make the daily life of non-users easier, even if they never get rid of their accent.
Call me a utopian, but my ambition is to create a world with fewer misunderstandings, where people will be able to communicate easily and be taken seriously regardless of their birth place. As a founder with an accent, I know I will have to work hard to beat the odds and make this vision a reality.
