THE GIFT OF THE MAGI: MAKING AUTHENTIC TEXTS ACCESSIBLE

Making Authentic Input More Accessible: Elaborated Input

Cheers to 2025 and the thrill of new experiments!

All experiments begin with good ideas.

I’ve never been a huge fan of adapted versions. Authentic language? Gone. Nuance? Bye-bye. That said, I can totally relate to the struggle of scraping through unfamiliar words. Case in point: still vivid memories of my university professor reading Beowulf in the original. ‘Feel the beauty,’ he said. Hah, when the only word you can recognize is Beowulf, beauty is the last thing on your mind.

One idea I’ve been experimenting with over the past few years is elaborated input (a huge thank you to Geoff Jordan and his blog!). Originally introduced by Michael Long, this approach focuses on exposing learners to fully authentic texts without simplifying or ‘dumbing them down’. Instead, it makes texts more accessible by adding just enough explanation to scaffold understanding while preserving their natural flow and original beauty – much like we do in everyday communication to enhance comprehension without losing authenticity.

However,

All good ideas are often deceptively simple.

Manually elaborating a text is a laborious process, with interest and engagement peaking in the first paragraph, dwindling by the fifth, giving way to escapist thoughts, and ending with checking out available simplified aka Titanic-in-30-seconds versions, ultimately defying the IKEA effect. – Author’s observations.

LLMs have been a big help in this process – check out this post from two years ago, where I shared prompts for input modifications – Generate, Simplify, Elaborate and Modify. Now, I’ve fully automated the process with the Elaborated Input Maker, which delivers quite consistent results.

My idea was to apply this approach to video captions to support listening comprehension. I shared the workflow for creating multimodal input, using captions with elaborations at a session on AI for multimodal storytelling this year, so I won’t get into the technical details here (but please let me know if you’d like me to!).

Also, I played around with visually adjusting the elaborated input. I used a slightly deeper, less vibrant shade for the elaborations compared to the authentic text. This subtle tweak made the elaborations less attention-grabbing while still keeping them easy to read against a black background.

Here’s one of the experimental videos, featuring O. Henry’s timeless masterpiece, The Gift of the Magi, in its full, unaltered text. It’s part of the Gift of the Magi Reading Maze and is designed for students to watch on their own.

Text: The Gift of the Magi by O. Henry (public domain)

Voice: Synthetic

The Gift of the Magi Reading Maze (html) – online version

The Gift of the Magi Reading Maze (pdf)

We’ve planned a session on Personalisation 2.0 for the end of January, where I’ll be showing how to personalise input and output using AI, including transforming stories into reading mazes, among other things. Stay tuned for the invitation and more detailed information!

And with that, here’s to new experiments!