The settings

In the beginning of the year, I have decided to meet a few good friends in Italy this September. Having had no prior experience with the Italian language except for hearing conversations of others, I decided to start to learn Italian autodidactically from scratch approximately one month prior to my trip to Italy. Simply as an experiment about what is feasible in such a short amount of time.

My stategy

Since I have had good experiences with Duolingo for freshing up some rusty language skills as well as starting a completely new language, I decided to use the Duolingo app on my phone during the time of the experiment. Since I was working full time professionally, I limited the time spent on Duolingo to 5 minutes per day.

Different from how I used Duolingo earlier, I decided to accompany the Duolingo activity with daily review activities on Anki. I constrained myself to 5 minutes of Duolingo time because I wanted to put extra emphasis on Anki: Whatever I learned through Duolingo, I planned to type in Anki to repeat the Duolingo content even further. I knew that the time spent on Anki grows roughly linearly with time in the beginning of using Anki for a new subject so that utilising only 5 minutes per day for Duolingo plus the time spent on Anki to repeat the Duolingo content seemed realistic to me. I configured Anki to introduce 3 new cards to me per day.

I wanted to focus on Anki because I’ve had phenomenal experiences with Anki regarding other subjects. I use it daily for reviewing flash cards that deal with Physics, computer science and language knowledge. My experiences show that whatever I type into Anki burns into my mind, justifying spending more time on Anki than on Duolingo for this experiment.

The leading questions

The purpose for conducting this experiment is manyfold. The following questions were of interest to me, in increasing generalising order:

1. How much of Italian is it possible to learn in one month?
2. How well do Duolingo and Anki play together? What are the timescales one is making progress for each of them?
3. How well does the strategy work out for a language? I conjecture that the answer to that question generalises to other languages as well.
4. How realistic is such a strategy during working full time?

The results

I conducted the experiment successfully without any days left out. Below, the Duolingo activity overview after half of the experiment can be seen.

The Duolingo activity overview taken from the weekly emails one gets when using Duolingo.

The following figure - taken from the Anki statistics feature - shows my Anki statistics after the whole experiment and reveals how many new words I learned, how much time I spent and how many flash cards I have reviewed, amongst others.

My Anki statistics of learning Italian for one month. Each letter corresponds to one evaluation and the letters (i) and (ii) correspond to the monthly and yearly statistics, respectively. The latter is shown since I learned Italian for a few days more than one month.

The most apparent results are the following ones. I …

• … spent 1 hour of reviewing Anki flashcards with an average of 1.4 minutes per day.
• … reviewed 708 flashcards with an average of 14.4 flashcards per day.
• … spent an average of 5.9s per cards.
• … added 162 flashcards altogether. After the whole month, 26 flashcards remained unseen.

The latter aspect reveals that Anki operates on a slower timescale than Duolingo does since I used solely Duolingo content for my Anki studies. Thus, when using Duolingo as an input for new content, the additional learning process with Anki is delayed. However, whatever is entered into Anki really burns into my mind. For me, it felt that if I progressed with Duolingo as fast as possible, the progress speed with the content from Duolingo would be slightly too fast for me. With that knowledge, I can adjust any future experiments with that fact in mind.

Another interesting fact was that my very small experience of hearing Italian conversations really helped in studying Italian. It seemed to me like I knew the basic phonetics and potential ways to pronouce the words that I learned. This finding emphasises the advice that many experienced language learners give: Expose yourself to the spoken language as much as possible while learning the respective language.

Along the same lines, learning a language in conjuntion with visiting the corresponding country once in a while makes sense. Not only does it boost one’s motivation but it also familarises with the sound of the language. Last but not least, the native speakers you meet make up a large portion of the fun of learning a new language!

Future directions

The most apparent extension to this experiment is to do it for a longer time, e.g. for one summer or one year. Since I would like to refresh another language first and want to try to use Anki for other types of knowledge at this point in time, I am not able to continue this experiment with Italian for now. However, I will use it with the afore mentioned other language.

Conclusion

Let me present to you what my conclusions are:

1. One does learn quite a bit in such a short amount of time, even with limited time spent. However, since I was able to check my progress right in Italy, I have to admit that there is a lot more to be learned before a language is usable after an experiment like this. Anyway, it gave some good intro to that language and I’d do it again for either the same or other languages.
2. The experiment showed to me that Anki alone is not the best. By that, my experiment confirmed Anki’s recommendation of not to use Anki for stand-alone knowledge but to use it in conjuntion with other types of classes or tutorials.
3. With the former, it is clear that also applying the gained knowledge is advisable. For instance with an additional language course, a tandem buddy, by writing texts or traveling.

No writing about Italy without a pretty picture!

The cozy beach of Napels with the volcano Mount Vesuvius in the background.