Can you read Gerard Nolst Trenité's The Chaos without tearing your hair out? This 1922 poem contains over 800 examples of irregular spelling in the English language, including gems such as:

"Finally: which rhymes with 'enough,'
Though, through, plough, cough, hough, or tough?"

If your brain isn't hardwired from childhood for such weirdness, the fact that 'beard' and 'heard' fail to rhyme might always cause a niggling annoyance. Trenité was a second-language speaker from the Netherlands (of course), explaining why he was able to pinpoint so many quirks that fly over the heads of native speakers. It raises an interesting question: can the 'weirdness' of different languages be quantified? And what does it mean for people in our line of work - translators, proof-readers and copywriters potentially working in languages they learnt in adulthood? Weird and hard-to-explain features may go some way towards revealing why a translator doesn't need to be an outstanding user of their source language, but rather a well-informed recipient of its implications. A gut feeling sometimes makes the difference between a good and bad text.

Through Gaston Dorren's excellent book Babel: Around the World in 20 Languages, I recently became aware of the World Atlas of Language Structures (WALS), a database comparing the structural properties of thousands of languages. Dorren uses WALS to investigate why it's tough to learn German, analysing which of its features are least common internationally. Perhaps unsurprisingly, the most unique languages tend not to be widely spoken. Mixtec, for example, is the world's only language not to differentiate between questions and statements in any way, including pitch, intonation or word order - and is spoken by just a few thousand individuals in Mexico.

This doesn't stop unusual languages from popping up in unexpected places: Nenets of northern Russia clinches second place in the weirdness rankings but has still managed to contribute the word 'parka' (a large windproof jacket) to English. At the other end of the spectrum, some languages that are often perceived as weird may not deserve their reputation. Basque is a complete isolate, unrelated to any other living language, but its core linguistic properties are commonplace. Its subject-object-verb word order is the world's most popular, shared by nearly 40% of languages. Like 70% of them, it also opts not to classify personal pronouns by gender.

When considering the world's biggest languages, it's German that comes out on top for weirdness. English, however, makes it to a respectable third place thanks to an array of rare features. Inverting the word order of sentences to form questions takes place in just 1.4% of languages, with many others simply adding a question particle at the end of a sentence. English's question particles - 'right?' and 'innit?' - are still reserved for only the most colloquial situations. Additionally, only 7.5% of languages include a /th/ sound, which non-native speakers will often accidentally substitute for the letter D - so of course English stubbornly uses it in its most common word, 'the'. In terms of weirdness, it's the subject-verb-object word order that English fall down on, with 35.5% of languages sharing this property.

And the least 'weird' of the big players? It's Hindi-Urdu, indicating that languages with many common features are not inherently easy to learn. Weirdness is interesting, but when it comes to acquiring that gut instinct needed for the perfect translation, contexts such as exposure, previous experience and quality of teaching will probably play a bigger role. After all, if you grew up with weirdness - whether it's German, English or Mixtec - it can be an uphill battle just to understand what the rest of the world considers the norm.

- Josh