They say money can’t buy happiness. But at Mrs Wordsmith, we fully believe that a rich vocabulary can. Well-chosen words can be the route to love, to success, and to pretty much anything that is worth asking for. Words transform lives. Words matter.

That’s why our mission is so gloriously simple: improve children’s outcomes through vocabulary learning. And we’re fulfilling this mission by combining cutting-edge data science with extraordinary creativity.



With approaches to literacy teaching having remained remarkably unchanged for more than twenty years, we knew that it would take something truly special if we were going to make a real breakthrough in the world of vocabulary learning.

Our plan of attack really came down to two key questions. Firstly, how would we decide which words children need to learn? And secondly, what would be the most effective way to teach them?

Finding the words that matter

Historically, the system by which vocabulary is taught in schools has been wildly unscientific and largely random. With no better option available to them, it’s often a case of individual teachers making in-the-moment judgments about which words their students might find useful. We were convinced that there must be a more precise solution, so we started digging deep into the data and came up with some surprising insights.

For instance, we found that people only tend to use the same small selection of words over and over again. 98% of the time, we use the same 5,000 simple, everyday words. Beyond these initial 5,000, though, things begin to get interesting.

The next 20,000 most frequently used words are common enough that you might regularly come across them in books and newspapers, but not so common that children can simply ‘figure them out’ by seeing them in context once or twice. These are often the words that bring meaning to life, and are the key to understanding a variety of texts that can make a real difference to a child’s academic achievement.

Indeed, one of our key aims from the outset has been to make sure that our words are genuinely useful – that they empower children to take ownership of their learning by helping them understand more of what they read.

In our relentless pursuit of learning efficiency, though, 20,000 words seemed unmanageably clunky. We needed to get the wordbase down to a more definable challenge.

Somehow, if we could meticulously dissect and analyze an endless supply of novels, newspapers and academic writing, we could trim the English language down to its most beneficial 10,000 words. The trouble was, doing so would require a level of complex linguistic analysis that no lexicographer – let alone teacher – has the time or resources to achieve.

It was a man-made problem, but there wasn’t really a man-sized solution. What we really needed was a machine.

Imagine our delight, then, when we found out that some of the world’s leading experts in computational linguistics and language processing were right on our doorstep at the University of Cambridge. We hastily arranged a meeting, and what they showed us was revelatory.

Not only did they have a system that could process unimaginably huge quantities of data, it was a system that became more perceptive and efficient the more data it processed. Machine learning could facilitate an unparalleled level of insight into the English language –  we couldn’t wait to unleash it in education.

And so, led by our new Chief Scientific Advisor Ted Briscoe and Machine Learning Lead Tamara Polajnar, we packed our machine with texts of all kinds and began analyzing, selecting, categorizing, and curating the largest language on Earth into the 10,000 words that make a significant difference to children’s academic success.

Showing them the words

10,000, though, is an awful lot of words to learn, even if you spread them over ten years from the ages of 7 to 17. With that in mind, we knew we had to come up with a motivating and time-efficient way to present them.

There’s a wealth of research that proves knowledge is easier to retain when it’s introduced visually, and also that humour is surprisingly effective for activating long-term memory. The answer was simple: we would illustrate each word with a striking, hilarious cartoon that reflected its meaning.

At first, we experimented with a variety of different artists and illustrators, desperately searching for a house style that we felt truly matched the beauty of our vision. But it wasn’t working. Somehow, the meaning didn’t seem to jump off the page, nor were the illustrations able to capture children’s imaginations in the same way that, say, the best children’s movies could.

What was it that these blockbuster animations were doing that we weren’t? We pored over the works of Disney, Pixar, and Dreamworks, deconstructing the visual style of each and analyzing what it was that made each character so memorable.

We became totally obsessed with the Madagascar movies in particular, watching them endlessly in envy, wishing we could create something equally funny and charming. Then it struck us – if you can’t beat them, hire them.

And so, using IMDb as our guide, we began our quest to find the people behind Madagascar’s characters. We cold-called whichever obscure crewmember Google would offer us, from sound editors to lighting assistants, and they all told us the same thing: if you want funny, you need Craig Kellman.

Craig was Madagascar’s Lead Character Designer and a seasoned Hollywood pro. Reaching out to him felt a little bit overambitious – after all, who in the right mind would want to swap the glamour of the movies for an eccentric little edtech company?

As luck would have it, Craig was not in the right mind. As a father himself, Craig was passionate about taking the same creativity and magic that keeps hundreds of millions of kids glued to screens, and applying it to education. If he could help us do that, we knew we would be accomplishing something seriously powerful.

Craig was on board, and his primary task was a momentous one: to design the characters that would appear throughout our Mrs Wordsmith story. Immediately, we could see that Craig shared our vision – the characters were charming, relatable and hilarious, and drawn in a timeless style that we knew children would love.

With the Mrs Wordsmith identity becoming more fully formed daily, Craig assembled a team of Hollywood’s funniest artists to begin work illustrating each word of our 10,000 Word Journey.

Using linguistic metadata provided by our machine learning team, the artists were tasked with bringing the words to life in a way that was simultaneously hilarious and educational, unequivocally demonstrating their meaning. This is no small task for even the sharpest creative and, indeed, each illustration goes through a rigorous vetting process.

When done to our satisfaction, though, the results are magical. Not only are the illustrations beautiful, but our qualitative testing shows that they help children grasp the meaning of each extremely quickly. Rather than requiring the 6-12 exposures typically required when learning a word from a dictionary on when inferring it from reading, our images express meaning so clearly and memorably that they dramatically accelerate vocabulary acquisition.

Over time, we aim to prove that Mrs Wordsmith really does take the “rote” out of learning – and that a brilliant, hilarious picture can be truly unforgettable.

So many more words to come

Mrs Wordsmith’s own word journey, meanwhile, is only just beginning. We’ve already got so much in the pipeline, from the creation of our Early Years and Growth Mindset Journeys to the expansion of our Narrative Journey with spelling, punctuation and grammar (SPAG) worksheets.

Ultimately, we dream of creating an entire joined-up curriculum universe, allowing teachers and parents to maximise children’s learning across the home/school divide.

And above all else, we want to make vocabulary rich. Filthy rich.


Facebook Comments