An experiment in learning Spanish - writing a daily log, with a twist

Posted Jan 13, 2023

By Chris Lovejoy

4 min read

UPDATE (Feb 2023): I decided to stop at day 16. It was a helpful exercise, but writing itself wasn’t the rate limiting step. I realised I needed to learn more words, and I came across an awesome tool called LingQ for doing so - so that’s what I’m focussed on right now.

UPDATE (Sept 2023): I decided to re-start from day 16. I’ve spent a bunch of time reading and listening, and feel that synthesising this learning through writing is the next step. (Hopefully there’s a noticeable improvement from the first 16 days!) I’ve also added functionality to my local Obsidian vault which automatically generates feedback on my Spanish writing with GPT.

For the next 30 days I’m going to write 250+ words in Spanish every day and share them on this site.

I’m not reinventing the wheel here. The whole “do X for 30/60/90 days” isn’t new.

However, I’m going to take a systematic approach, which I haven’t (yet) come across elsewhere.

(context: I’m learning Spanish, am around B1 / lower intermediate level).

Topic and grammar looping

I’m going to identify 6 topics and keep looping through them. Given 30 days, this means I’ll re-cover each topic 5 times. I’m hoping to beyond initial superficial thoughts on a topic - and force myself to reach a deeper level of expression in the new language.

As well as looping through topics, I’m going to loop through:

grammatical structures
styles
use of translate to find new words

For example, I might focus on using the future conditional on one day and past imperfect on the next. Or incorporating a specific learning point (e.g. use estar vs ser in Spanish). The cadence for this will be different (ie. more than 6 approaches I cycle through), so it’ll be deliberately out-of-sync with topics.

In the final six days, I’ll focus on bringing together all that I’ve written on each topic into a final, original essay in Spanish.

What else I hope to learn

I’m also going to experiment on other axes, with a view to answer the following questions:

Does the nature of my writing vary when I write directly into Spanish vs when I write first in English then translate that into Spanish
How can AI best be utilised to support the language learning process? (E.g. Can GPT give me good feedback on my writing? Can it give me prompts to write about?)
How best to utilise translate? Do I remember more or less if use translate when I can’t think of words?
What is the best type of feedback? Can I categorise my errors?

What I’ll write about

The 6 topics will be:

Sleep
AI for language learning
Writing
Work
Progress
Principle of learning

The logs

Why do it this way?

1. Writing is the best way to learn grammar

I’ve consumed a lot of Spanish and understand well, but am slow to produce output (both spoken or written). I believe you can learn faster with written, because you can pause and think more - and therefore learn the grammar. When talking, you can often skip on grammar and still convey the desired meaning - and improve less here as a result.

I loved the Total Language Transfer course and fully believe in the principle:

Think slowly, learn quickly

(or words to that effect.)

Even if nobody reads this articles, the fact that somebody may just check out this article and see that I haven’t followed through can be a powerful motivational force.

Initial inspiration for this came from stumbling across Jamal’s notes.

3. I’ll learn words that are useful for me

I’m a believer in learning the words that you’re most likely to use. No point learning the names of dinosaurs in spanish if I’m never going to have a conversation about it.

Given that I’m writing about areas I’m interested in, there’s a pretty decent chance I’ll have future conversations about these topics too.

Actualizaciones

En todos los textos, estoy escribiendo una versión editada. Para esto, estoy usando mi herramienta CheckMyWriting.

Spanish Language-learning

This post is licensed under CC BY 4.0 by the author.