Learn Chinese Like an Engineer (Part 2)
You can only improve what you can measure. You'll need a weighing scale to lose weight. You'll need an equivalent to dramatically accelerate language learning, too.
As I’m writing this from a neighbourhood caffé in Brussels, where the staff speaks one of the most complex languages on Earth, I realise that we’re almost in September and this language-learning series might accidentally be very timely for many who are still studying or for lifetime learners who synchronised their life cycles with those of their children who’ll soon return to school again.
The last issue ended with a mention of an agreed-upon yardstick used to measure the complexities of languages.
FSI, or The Foreign Service Institute, classifies Mandarin Chinese as a group IV language class that requires around 88 weeks, or 2200 hours of practice, to reach a decent proficiency level.1 This approximates how long it might take to understand others and express yourself, but not becoming fluent or mastering it.
2200 hours…
Is this number set in stone? Are we really at the physical limit? We used to think we’ve reached human speed limits for standard running distances or that flying was exclusively a bird’s prerogative. All of the above, we now know, were instances of a fixed mindset, a very understandable cognitive bias.2
Most Chinese people in my entourage tell me that vocabulary was taught through rote memorisation. It has always been like this and probably always will be—their words.
But what if there were engineering optimisation techniques that allowed us to shave off precious hours of our lives to achieve the same result faster? I believe we can borrow some engineering paradigms to accomplish precisely that. It’s one of these instances where solutions to problems in one domain can be ported to solve similar issues in a different, non-related one.
You Can Only Improve What You Can Measure
A friend and colleague have been nagging me about observability or measuring application performance for a while. Although we’re nowhere near where we’d like to be as a progressive tech company, we both understand that one can only improve what one can measure. We need eyes to see where we’re headed.
When an engineer is tasked with measuring performance, the first step towards this goal is instrumentation. Instrumentation is equipping yourself with relevant measuring devices. If you want to lose weight, you need a weighing scale. If you need to improve your application’s performance, trace where it wastes most of its time.
When the application's code has been instrumented with the measurement devices, you can finally see what was previously wholly opaque. One of the most helpful observability artefacts is the performance flame chart.
A graph like the abovementioned immediately reveals bottlenecks. The longer the line, the lower the performance. Ideally, you want your entire timeline from left to right to be as short as possible. To get there, however, you’ll need to identify and deal with the most extended stripes, one by one. Once done, rinse and repeat until you can no longer optimise.
…
I’m discussing observability, instrumentation, and flame charts because language acquisition is a typical application request processing. The most significant difference is that the length of this flame chart is measured not in milliseconds but in years.
As is the case every time you click a website button, learning a language is also composed of layered, overlapping activities such as grammar, pronunciation, listening, speaking, writing, and vocabulary to get you the result you’re after. Depending on your abilities and inclinations, you’ll spend more in some flame chart stripes than in others. You aim to shorten most of them, which is how an engineer would challenge the commonly agreed upon FSI 2200 hours “limit”.
Divide and Conquer
You learn a foreign language like you eat an elephant: one bite at a time. This divide and conquer paradigm is one of the core life patterns that applies to almost everything and never fails.3
This means that you’ll need to dissect a language into its building blocks to master them separately, then try to assemble them into a coherent whole, disassemble them again, and reassemble them into a bigger, better version once more. Repeat until done.
One of such building blocks you’ll need to focus on is vocabulary.
The first part of this series introduced concentric circles of a language vocabulary that, similar to onion layers, progressively builts subsequent layers on top of more fundamental ones. Although this approach is characteristic of all languages, it’s even more applicable to Mandarin Chinese because of its Lego-like philosophy.
As an engineer, I love tinkering with this erector set, so I’ve developed my approach to learning the vocabulary. I believe it can help you shave off up to 70% of that vocabulary acquisition stripe from the language learning flame chart. It’s based on the associative nature of the human brain, and I’d like to share it with you in the next issue. Stay tuned.
https://blog.rosettastone.com/the-complete-list-of-language-difficulty-rankings/
https://www.mindsethealth.com/matter/growth-vs-fixed-mindset
https://www.linkedin.com/pulse/divide-conquer-help-your-startup-navigate-unknownpart-alimoekhamedov-jsepc/