Linguistic Injustice in the Workplace and Generative AI

Approximately 80% of all scientific and research communications are published in English, even though only about 5.4% of the global population are native English speakers. English is also recognized as the dominant international language of business. This disparity creates significant challenges for non-native English speakers in professional and academic environments.

Non-native English speakers often spend up to 30% more time on the same tasks due to language barriers. Despite these challenges, they typically produce output comparable in quality and quantity to their native-speaking counterparts. However, their efforts often go unnoticed. Native English speakers might not fully understand or appreciate the additional hurdles non-native speakers face to overcome these barriers. It can take 7-10 years of immersion and practice for a non-native English speaker to achieve fluency comparable to a native speaker, depending on individual learning conditions and the complexity of the language.

Let me provide some personal examples. In some languages, words like goal, objective, target, purpose, and aim are often translated into a single term. When newcomers arrive in English-speaking countries, they often struggle to choose the right word for a particular context, leading to hesitation or miscommunication.

Pronunciation also poses a challenge. Non-native speakers may have learned incorrect pronunciations for certain words, including people's names. When they discover the correct pronunciation in a new environment, they may find themselves juggling two competing pronunciations in their minds. Over time, selecting the correct one becomes a cognitive task in itself. It takes practice and exposure for the "correct" version to dominate, which can take months or even years.

Sometimes, these issues lead to amusing misunderstandings. Here’s a personal anecdote: a long time ago, I misheard and mispronounced the term tacit knowledge as tactic knowledge. Since I was already familiar with the word tactic, I internalized the incorrect term as meaning implicit knowledge. For years, I used tactic knowledge in my academic writing, and no one—not my co-authors, reviewers, editors, or copy editors—ever corrected it. Even readers likely didn’t notice.

It wasn’t until my brother questioned me after failing to find the term tactic knowledge online that I reconsidered. I initially dismissed his observation, confidently relying on my academic background. Later, when I used the term again in a new piece of writing, I asked ChatGPT to edit my text for grammar and typos. It corrected tactic to tacit. Fortunately, I caught the change, researched the term, and realized my long-standing mistake. This example illustrates how a simple mispronunciation can perpetuate for years and how tools like AI can help rectify such errors.

Generative AI's Role in Addressing Linguistic Injustice

It is often said that generative AI will help non-native speakers in several ways:

• Improved Communication: AI tools like ChatGPT or Grammarly can assist with grammar, sentence structure, and vocabulary, enabling non-native speakers to produce clear and professional communications.
• Pronunciation Training: AI-driven language apps can help users learn and practice correct pronunciations through real-time feedback.
• Cultural Nuance Guidance: Some advanced AI tools can offer suggestions based on cultural and contextual nuances, which are often challenging for non-native speakers to grasp.
• Accessibility and Speed: By automating routine corrections, generative AI can save non-native speakers time, enabling them to focus on the substance of their work rather than its linguistic accuracy.

However, generative AI cannot fully eliminate linguistic injustice in the workplace. For example:

• Bias in Hiring Practices: Studies have shown that up to 25% of job applications are rejected solely based on the applicant's non-English-sounding name, regardless of qualifications.
• Fluency as a Barrier: Fluency in spoken English continues to matter disproportionately in hiring decisions, even for roles where verbal communication is unnecessary.
• Cultural and Social Perceptions: AI cannot fully address biases and stereotypes about accents, which often affect how non-native speakers are perceived in the workplace.

In the short term, linguistic injustice remains a pervasive issue in workplaces and academia. While generative AI offers significant promise in leveling the playing field, it is not a complete solution. Addressing linguistic injustice requires systemic changes in hiring practices, workplace cultures, and educational systems to value and support linguistic diversity.

Imagine an extreme future where AI tools or avatars can simultaneously translate people’s words from one language into another in real time, with cultural nuances and context perfectly preserved. In such a scenario, language requirements in most job applications could become obsolete, and linguistic injustice might no longer be a barrier. This could usher in a truly inclusive workplace where talent is evaluated based solely on skills, creativity, and qualifications, rather than language proficiency.

What do you think about linguistic injustice? Have you encountered or observed examples of this in your own experiences?

?