Harness the Lionbridge Lainguage Cloud to support your end-to-end localization and content lifecycle

Lainguage Cloud™ Platform
Translation Community
Workflow Technology
Smairt Content™
Smairt MT™
Smairt Data™
Language Quality

Our people are our pride, helping companies resonate with their customers for 20+ years. 

About Us
Key Facts
Trust Center



Pedestrians cross a busy street.

Reducing Implicit Bias

Addressing inherent prejudicial beliefs in translation, documentation and code base

The coronavirus that causes COVID-19 is not the only pandemic facing the globe today. Discrimination, particularly racism, continues to weigh down society as a whole. In the last few years, growing protest movements and social justice organizations have pushed for reforms to fight both explicit prejudice and its more subtle forms.

One insidious byproduct of kyriarchy is implicit bias: the unconscious devaluing of groups of people and their attributes. The very words we use can contribute to this pattern, but if addressed, carefully chosen language can also help fight it.

Lionbridge has always worked to remove explicit bias. And as an extension of those efforts, we have also increased the attention to implicit bias. Here's what some of Lionbridge's leaders had to say about increasing fairness and building a more inclusive society with language.

The following Q&A includes answers from:

  • Jaime Punishill, Chief Marketing Officer
  • Mark Aiello, General Manager, Life Sciences
  • Lisa Yip, Solution Architect
  • Rafa Moral, R&D Director, Linguistic Engineering
  • Brian Randall, Vice President, NA Strategic Accounts
  • Claire Goodswen, Senior Director, NA Solutions

How have you seen the conversation about implicit bias change recently?

JP: Over the past few years, the mainstream seems to have woken up to the idea that it’s not enough to not be racist, you have to be proactively anti-racist. People in positions of power and influence need to do more than talk about change: Here’s an opportunity for them to proactively make an effort to demonstrably change.

MA: Our clients and prospective clients often ask us about diversity efforts and corporate social responsibility. Discussions about implicit bias are a natural extension of the diversity questions. From the perspective of the life sciences industry, it makes a lot of sense given the push for greater patient engagement. Pharma is constantly looking to diversify in every sense—how studies are designed, how study participants are engaged with and how supply chains are managed—and addressing implicit bias is one way to do this.

LY: I have seen a lot more questions about diversity and inclusion in our RFPs. Those questions have increased since mid last year. 

CG: It has come more to the forefront in the past year and even the past five months. Before that, it was more on the fringes. Now it’s something people see as a necessity. Even if they’re not addressing it yet, it’s on the agenda. Because of things like the #MeToo and BLM movements, people are reexamining their own bias and looking at how this is ingrained in society. 

RM: We’ve definitely seen an increase in requests for this sort of service, especially after the 2020 increase in protests against police brutality following the murders of George Floyd and Breonna Taylor, among others. 

"We all have the opportunity now to make changes that better reflect our values." -Brian Randall

What are the first steps to address implicit bias in language?

CG: We need to be open minded and not make the assumption that we’re immune. No one is immune to it...We also have work to do because it’s ingrained in our own industry. 

BR: Companies interested in these projects have two separate issues at hand: correcting any past decisions that demonstrate bias, and updating workflows for the future. The latter is an opportunity to create an inclusive language style guide.

RM: Identifying problematic words or phrases is part one. This is where our semantic annotation services and customized transcreation services come into play; we can use our existing terms lists in combination with any lists or tools a customer already has. 


How does Lionbridge’s experience lend itself to projects for reducing implicit bias?

LY: We already partner with diverse suppliers and vendors through our community management team. It is critical for certain projects and clients that LSPs understand the client’s scope, and that is something we excel in. Not everybody has the skillset required for these projects; you need to be very culturally aware and have a lot of conversations. Often people want these changes made quickly, which is tricky to do without losing quality. We have experts who can help complete projects quickly and skillfully. 

MA: We live inside the world of linguistic management, so everything we do can expand into this space. We can really develop a plan from scratch because we work so closely with our partners. For example, when our pharma clients request us to work with members of our community who are diversity providers, we’re able to create a resourcing plan in order to meet their needs of linguistic quality, as well as vendor selection.

RM: For some time now we’ve had Linguistic Toolbox Offensive Language rules as part of our workflow to make sure inappropriate terms don’t appear in translations. Expanding from more traditionally unwanted words like profanities and explicitly racist, sexist and homophobic terms to subtly coded terms is simply a wider application of our skillset in this area.

What are some barriers in finding and reducing implicit bias? 

CG: Once you have found where the bias is, I think it’s education. You need to build awareness for content creators—having things like glossaries and quality checks are key.

It’s a matter of hearing without becoming defensive. People often see this conversation as a judgment on them personally when really it’s a product of our history. The path forward was historically based on the path of least resistance, and so for years this wasn’t considered—this is not a judgment on what you’re doing right now, but on how we can change for the better moving forward.

RM: Many of the terms that are offensive can be inoffensive in certain contexts. That’s why we’re working on disambiguation capabilities in our machine learning models. Combining gazetteers and machine learning in a hybrid approach is one of the best ways to identify offensive text. Gazetteers increase recall and machine learning can disambiguate to reduce false positives.

BR: In languages with grammatical genders—for example, Romance languages—sometimes no non-gendered version exists. Spanish-speaking communities have adopted the -x ending in some cases (think Latinx), but that spelling often gets caught by filters for being ungrammatical as well. So even if the bias is identified, it can be challenging to find alternatives.

From an interpersonal perspective, it’s important to invite people into the conversation instead of excluding them because they were doing something wrong. That’s how we can make the most progress.

An abstract design similar to a blurred night photo of a bridge.

How is Lionbridge moving the needle on implicit bias?

RM: We’re continually in a research and improvement phase for reducing implicit bias. Right now we’re developing ways to analyze bias and put a number or rating to it.

We’re also compiling lists of terms that are biased but not profanity or slurs—this ranges from something as common as the word “landlord” or “spokesman” to more obviously offensive terms like “spinster” or “old maid.” There’s a lot of subtle gender references in words like these. 

MA: There are a ton of words we use in industry every day that have a history based in racism. “White label” is a good example, as is “blacklist.” And beyond that, they aren’t actually informative labels. We have all these terrific translators and translation tools on our teams; we can use those same resources to “translate” into more modern, descriptive and inclusive language.

CG: I think having the conversation with customers is important. Companies are struggling with the enormity of how to address this. We can open that door and tell them yes, there’s a lot of road to go, but we can support you along the journey. We have the tools and solutions to help. When it’s offered, people really grab onto it and really want to talk about it—it’s kind of like a life raft. 

Why is it important for corporations to recognize and reduce implicit bias?

CG:  The world is very diverse and that diversity adds value to society and enriches our potential for connection around the globe. It’s important that people focus on inclusivity and acknowledge that not everybody is the same. From a corporate perspective, we need to make sure that our messaging connects with as many people as possible. And for a company to be successful in the next decade recognizing the importance of these changes really opens up who feels welcome in your audience. 

LY:  Who you are is more than what you look like on the outside. This effort is a good move and honestly, has probably taken too long to start. 

JP: This is the right thing to do. It’s part of being a good corporate citizen. Particularly in the tech community—there’s a sense of responsibility that there’s more that they can be doing. If a company pours lots of money into Diversity and Inclusion efforts while recruiting and then those new software engineers go into a codebase with these antiquated terms, then you’re exposing them to these implicitly racist or sexist phrases and you haven’t really made an inclusive workplace yet.

How you set the foundation is really important because that dictates the process of making everything relevant and culturally appropriate locally. It’s putting your money where your mouth is, and at some point it will become hard to explain why you haven’t done this work when other companies have.

BR: At some point you have to get really “real” about how these words can impact your teams and your customers alike. It’s a matter of making your employees feel they belong and are valued and doing the same for your customers. We all have the opportunity now to make changes that better reflect our values. To paraphrase Dr. Maya Angelou, you did what you knew how to do, and when you know better, you do better. 

A woman looks at her phone while riding the bus.

At Lionbridge, we are committed to learning more and doing better every day. It is part of our mission to break barriers and build bridges globally. Our work eliminating words and phrases with problematic histories can help your team do the same.

linkedin sharing button
  • #technology
  • #blog_posts
  • #translation_localization

April M. Crehan
April M. Crehan