Two Memorable & Hilarious Translation Fails (And What They Reveal About the Strengths and Flaws of Both Humans and Machines)
- Ka Yee Meck
- 12 minutes ago
- 3 min read
Over the years, I’ve seen my fair share of translation hiccups.
But two, in particular, stand out not just for how hilariously wrong they were, but for what they reveal about the nature of human versus machine translation.
Both happened while I was working at NET-A-PORTER, the world’s leading online luxury fashion retailer, and they still make me smile to this day.
Let’s dive in.
Mistake 1: Cotton or… Noodles?
Source language: English
Target language: Chinese (Simplified)
One line in the product details simply read:
“100% cotton”
But what the third-party supplier submitted was:
“100% 面”
(That’s “noodles,” in case you’re wondering.)
The whole department ROARED with laughter when my Chinese colleague and I shared this gem during a team presentation. It was so ridiculous – and yet, so easy to understand how it happened (if you happen to speak Mandarin!!)
For those who don’t speak Chinese, here’s a quick explainer:
In Mandarin, the word for “cotton” is 棉 (mián). The word for “noodles” is 面 (miàn). When typing Simplified Chinese using a pinyin-based input system, you type out the pronunciation (e.g. “mian”) and then manually select the correct character from a list (as show in the screenshot below)
Here’s the catch:
As you can see in the screenshot, option 1 is 面 (noodles)
Option 2 is 棉 (cotton)
If you’re typing quickly, not paying close attention, or relying too heavily on predictive typing, it’s incredibly easy to pick the wrong one – and voilà, you’ve got a... noodle t-shirt.
Classy.
What does this tell us?
Oddly enough, this is one instance where a machine translation system or even a basic CAT (Computer-Assisted Translation) tool might actually have saved the day. Since MT typically works with entire phrases and larger context, it’s unlikely it would confuse “100% cotton” with “100% noodles.”
In this case, we saw human fallibility in action — the classic typo with serious LOL potential.
Mistake 2: I Got My Degree from St. Martin's... CANAL?!!!
Source language: French
Target language: English
This one came from our French colleagues and caused nearly as much laughter as the noodles incident.
The context: A product write-up for a designer stocked by NET-A-PORTER.
The intended message: “Designer XXX studied at St. Martin’s” – a clear reference to Central Saint Martins, the iconic London-based fashion and design school.
What we got:
“Designer XXX studied at St Martins Canal.”
…Which, oddly, suggests a scenic waterway in Paris – Canal Saint-Martin – rather than one of the world’s top fashion institutions.
Unlike the noodle example, this wasn’t a typo. This was a context fail. The translator (a human) had simply never come across “St. Martins” in a fashion context and made a best guess based on surface familiarity – a classic case of “false friends” between cultural references.
What does this tell us?
This is exactly the kind of nuance that machine translation still struggles with. Without proper training data or context, even the most sophisticated AI could easily make the same error.
The irony? While the cotton/noodle mistake was something a machine could’ve caught, this one shows why humans are still essential. But only when we’re well-informed and culturally attuned.
So… Machine or Human?
These two stories might seem like nothing more than comic footnotes, but they highlight a deeper truth:
Both machine and human translation come with strengths and weaknesses.
Machines can reduce careless errors – like the noodle mix-up – especially when context is simple and terminology is repetitive.
Humans can grasp context, nuance, and culture – but we’re not immune to fatigue, oversight, or unfamiliarity.
That's why I argue that the best approach is to combine both: using the efficiency of machine tools with the insight and experience of a qualified human translator.
Real Expertise
If you’re looking for reliable, culturally-aware translation that won’t accidentally serve up a noodle dress or a misplaced design school, I’d love to help.
I’m a fully certified, experienced translator with expertise in fashion, marketing, and luxury branding.
Comments