Three months ago, I decided to apply for Hungarian citizenship. I qualified because some of my ancestors were Hungarian. The passport would come in handy for living and working in the EU, as well as in Switzerland, where I reside. Problem #1: as part of the citizenship application process I would need to go through an interview and prove higher-intermediate knowledge of the language (upper B1 or B2 in the so-called European Framework. Problem #2: I spoke zero Hungarian three months ago. Problem #3: studying the language would be very low on the list of my everyday priorities.
Still, in late May, I walked into the consulate in Bern, had a nice thirty-minute chat with the Consul in more or less intermediate Hungarian, and passed the interview. I had spent exactly sixty hours actively studying the language to get to that level, divided into 43 hours of individual study and 17 one-hour Skype lessons.
Since most sources that I’ve consulted claim that reaching the B1/B2 level requires 500+ hours of study (examples here and here), I thought my experience might be useful to other language learners looking for some tips on how to speed up their study, be it of Hungarian or of any other language.
To get this out of the way — I already speak seven languages fluently and another two or three at an intermediate level. This seems like a lot, but comes mostly as result of circumstances — I’ve lived and worked all over the world, and travelled for years on end. I’ve learned from girlfriends, colleagues, roommates, neighbors and fellow travellers, from books, (rarely) through formal classes, as well as just on my own, for my own pleasure. The setup of my life is such that I speak about five languages daily at work and at home. You decide whether, on the basis of this, you write me off as someone who’s just unusually talented and lucky or read more closely for the lessons I might have learned from all these experiences. My language history notwithstanding, the Hungarian experiment was the first time I approached a language methodically and with a clear purpose in mind; it was a very exciting challenge.
Estimating the difficulty
There isn’t much point in rating the difficulty of a language you have to learn, but I did have a quick look at the five compartments that are usually the source of complications — writing system (Latin alphabet — no problem), vocabulary (practically “alien” — complicated), gender (no gender — no problem), conjugations (moderate complexity) and cases (moderate complexity). Overall, with about 2.5/5 on the complexity scale, Hungarian would be a language of an average difficulty, most of it coming from the vocabulary, conjugations and cases.
I had one focus: to learn the language _fast_, i.e. to reach the intermediate level in the minimum amount of time. I defined the intermediate level as being able to sustain a 30-minute conversation about my life, work, household, Hungary and similar topics, with more or less decent grammar and without struggling for words too often.
If you’ve ever taken a comprehensive language exam then you’ll know that there are several components to language proficiency — for example reading, writing, speaking, listening, and grammar. Since I was going to have to pass an interview, I focused heavily on speaking and listening. This only works to a point, however — writing and grammar exercises help solidify the knowledge, and reading expands the vocabulary, so learning cannot be too unbalanced. I decided to spend one hour speaking for every three hours spent on individual study (reading, writing, grammar).
I started by reading the Wikipedia article on Hungarian. This is where I got an overview of how its engine works, and mapped out the easy aspects to leverage (gender, pronunciation) as well as the tough ones to look out for (definite/indefinite conjugation, cases).
This website had a nice, encouraging overview of Hungarian for beginners. Some of the analysis is unrealistic but it’s a confidence-booster.
There is a website called Hungarian Reference which has an exhaustive guide to Hungarian grammar. I used it sparingly, to try and grasp the concepts and group them together, so as to avoid having to learn by rote. For example, skimming through the section on imperatives made me understand that there are basically no rules and that I’d need to learn many conjugations by heart at first. On the other hand, skimming the section on comparatives and superlatives showed me how easy their construction was, and I learned that on the spot.
As a beginner (and only as a beginner) you need a book, any book, to follow for a while, and I chose this one. It is outdated and a bit dry, but it gave me a first intro to the grammar concepts and my first 1000 words or so (the first 10 lessons). It came with audio recordings, which were handy for starting to understand the spoken word and self-correcting the accent.
I found my Skype teachers on italki. I needed to go through four or five to find the two that I liked working with.
I used Anki as my spaced repetition flashcard software.
Over the years, I’ve seen dozens of products which claimed to revolutionize the way we learn languages, and I’ve never been impressed by any of them. This especially goes for “scientific” methods that want you to learn a language supposedly the way a child does. How language is acquired remains in many ways a mystery, despite much research over the last few decades— it is probably a propensity inherent to human beings which developed relatively swiftly 100,000 of hundred years ago, and it gets “activated” in children of a certain age (those who grow up in isolation, for example, cannot learn a language in adulthood). Therefore, my method has little revolutionary in it. Instead, it simply aims at optimizing the time spent learning for maximum impact.
It studied daily. Even if only for ten minutes, to go through a few flashcards, I never skipped a day. I did about three book lessons per week. On proper working days, it would take me about 30–45 minutes to go through a lesson in the book. I’d add the vocabulary to Anki, glance at the grammar, do the exercises until I’m happy I understood most of the material, and prepare questions for the teacher. I would read the lesson out loud at the end, and follow that with listening to the recording. I would speak to the teachers about twice a week for an hour, on average. I often Googled my questions, especially once I was past the beginner stage.
Repetition, daily limit and importance of sleep
While everyday learning is crucial, there is a limit to how much active learning I can fit in a day. If I study for less than thirty minutes the brain hasn’t fired up yet and I’m slow. If I learn for more than three hours then I stop retaining as quickly and I start getting tired. The sweet spot is around the two-hour mark, which I aimed at hitting over the weekends, with up to an hour of learning per weekday.
Sleep was important. Sleep resets my ability to retain. Sleep somehow “packs” things, if I’ve learned them properly, in their proper place, and speeds up their recall. I’ve come to appreciate this and expect it to happen through my experimentation with not only language learning but also the practicing acrobatics (parkour, skydiving), playing instruments, etc. Conversely, irregular and insufficient sleep debilitates my learning capacity, as well as my willingness to learn.
Thinking in Hungarian
The greatest opportunity to speed up the learning is structuring the time when you’re not actively learning. I used this “free time,” when walking around, or resting, or riding the motorcycle, to think to myself in Hungarian. I would give myself a topic and come up with sentences, or imagine a conversation. As I would come up with words that I didn’t know I would take a mental note of them, or write them down. Eventually I began thinking in Hungarian, even the everyday thoughts that normally brew in the background of my mind. For this to happen I needed enough vocabulary to get the conversation going, as well as the habit of everyday practice, in particular going through the flashcards in the morning, which would “tune” the brain to the frequency of a language, so to speak.
Thinking in a foreign language might seem far fetched, but in fact comes naturally — if you’d never spoken anything but English and then went to live in Rome and immersed yourself in the language, after a couple of weeks you’d be thinking in Italian. I just make a slight conscious push to start the process sooner and keep it going for longer. Thinking to yourself improves vocabulary, accent (since you “hear” yourself think) and recall speed, and all it takes is a little willpower to adopt.
Spaced repetition flashcards
Spaced repetition is the next most powerful tool, but only when used properly. The flashcard concept is simple — you write a word in one language on one side and a corresponding word in the other language on the other. This way you can test yourself at regular intervals. There are many apps that help you determine those intervals based on a bit of science and how successfully you recall a word. I use Anki, because it works without hassle. There could be better apps out there, I haven’t explored much. This is how I use flashcards:
- I make them myself. You can find flashcard “packs” for many language pairs, which seems a convenient option, and sometimes is, but such collections often contain words that are not useful or are badly translated. In addition, making the cards myself gives the brain some valuable time to focus on the word and remember it.
- The flashcard shows the language I know, and asks me the word in the language I’m learning. This is definitely harder than the other way around, but this is the way recall works when I actually need to use the word.
- I use whichever base language the term comes the most naturally to me. Most of my flashcards use Serbian, English or French as base languages but others use Italian, Indonesian, Russian and others, depending on how well the language expresses the concept and how “wired” the concept is into my mind in that language. This helps recall, since it maps the way I think about a term onto the target language.
- I create a visual association, or sometimes a mini story, for every single word, especially in an alien-sounding language like Hungarian. So if I’m learning the word “jelenleg,” meaning “currently,” I’ll break it up in the following way. “Jelen” means deer in Serbian. I append the English word “leg” to it. So “jelen+leg” makes me think of a deer that is _currently_ being treated at an animal shelter since it hurt its leg. It sounds complicated, but I’ve discovered, notably through my experience of learning Chinese, that my brain recalls a word/character for which I’ve created an association nearly five times as successfully as a word that I tried to learn by just “looking” at it for long enough. I sacrifice about 10 seconds of my time every time I create a new flashcard in order to form the association, but this brings enormous dividends later on in terms of better retention and fewer repetitions.
- If possible, I try to learn a word in its context, for example in a short phrase. This is what many people recommend, but I’m careful with the technique for two reasons: firstly, if the phrase contains many words I don’t know it will be very tough to learn; secondly, learning words through phrases might make me more likely to learn by heart and to reuse those exact phrases, leading to somewhat formulaic speech. Still, this technique has its benefits in some cases.
- I learn many more words per day than the app recommends. The default setting is about 20–30 new words and 20–30 old words (reviews), and I find this way too low. I don’t mind going through up to 100 new words on a given day and up to 200 reviews. Even this kind of workload doesn’t take more than 25–30 minutes per day and keeps my recall rate at 80%, which I am happy with. Again, I have no doubt in my mind that the “secret” to not dropping the recall rate as I crank up the number of words per day is creating an association for every single word.
- Finally, for my interview preparation I could obviously choose which words to include in flashcards, and it would have been tempting to stuff the deck with the words that, in my estimate, would be more likely to come up. For example, the word “passport” is certainly more important than “giraffe.” And yet, a long enough conversation can move in mysterious ways, so, in my experience, it makes more sense to include words organically as you learn them through conversation than to try and anticipate too much. That said, I had about twenty interview-specific words on the list, relating to different forms I would need to fill out.
In summary, spaced repetition flashcards are what you make of them — they can be a fantastic tool if used correctly, and can help you build vocabulary with amazing speed.
“Teaching” the teachers
My Skype lesson would normally go like this. Taking my teacher Rose as an example here.
We decide on a topic in advance and I prepare a list of questions. We spend a couple of minutes at the start just exchanging pleasantries and making the conversation enjoyable. Then she quizzes me on the agreed topics. She also does her fair share of speaking for my listening practice. When I make mistakes or use a non-Hungarian word she corrects me or translates the word for me, doing so by directly typing into Skype instead of interrupting me. We rarely stop for this; I take in the corrections on the fly as I see them appear. When either of us feels that more explanation is needed (a new grammar concept for example) then she starts explaining and I stop her when I feel like I “got it.” Then I give her another example to show that I’ve understood, and we move on. When I feel like I need to slow down, or I already have say 100+ new words in Skype, I maybe ask a grammar / culture / vocabulary question and she speaks for a bit. After the call I have spoken for 30–45 minutes and have many new words for flashcards. I never do grammar or reading exercises together with a teacher.
This way of optimizing time works for me, but it takes a bit of coordination even with the best of teachers to get it going. Typical problems when choosing a teacher to be aware of include:
- Too set in their ways. One lesson takes one hour, hell or high water, and you go through a book, lesson by lesson. If the lesson requires you to learn 5 different types of fish soup and how to write a telegram (true story) then so be it. All students are the same, and classes are usually about reading texts and doing grammar exercises, the same pace for everyone. If you finish the lesson quickly they will simply say goodbye and move on to the next student.
- Taking too much time when explaining. This isn’t bad per se, but I often encounter teachers that dominate the conversation by over-explaining, taking half the lesson to explain the instrumental case or what a prime minister is. Better teachers are those that give small hints of info and wait for the student to ask for more; the best ones are those that throw in hints of the kind “did you notice a connection between this and that” or “see how you can remember this easily,” without stopping the flow of the lesson.
- Not versed in grammar. A teacher-student pair who can talk grammar-speak to each other can advance very quickly. The teacher should know what adverbs, gerundives and locatives are, how to justify the definitive conjugation and when subjunctive is necessary. If you have a teacher who doesn’t know about grammar then you must be an experienced learner who can ask tracing questions to infer grammar rules and regularities.
If you notice any of the three problems above then you need to talk to your teacher to try and remedy them; if that doesn’t work then it’s time to move on.
In all cases, the teacher must know the target language perfectly well; this is obvious. It is less obvious that, especially if you are a beginner, your teacher must know your language perfectly as well, since you’ll be using it to ask questions about and all sorts of subtleties, such as tricky words and elusive grammar topics.
Grouping concepts vs. acquiring “the feel”
Learning progresses along two lines — conscious learning through associations and concepts, and acquiring the “feel” for when things just “sound” right. Concepts are crucial, since that is how you take a disorganized mess that is a foreign language and separate it into understandable chunks that form a coherent system and make sense. But the “ear” is a powerful ally as well. For example, over time you group the different case triads (for example “towards,” “at,” and “away from”) together, and learn a few classes of objects that are used with each, but thereafter it’s your “ear” that takes over to “know” that you say “Budapestről” instead of “Budapestből,” regardless of what seemed logical in the beginning. The more you can guide the ear conceptually the better, but sometimes it is best left to itself — it acquires the feel faster than your brain grasps a very complicated concept. For example, I prefer to leave the “autopilot” on when conjugating in the past tense, keeping track of the mistakes I instinctively make, rather than preprocessing based on a myriad of rules.
What to focus on first
Apart form learning the most common verbs, I learned the indefinite present tense conjugations of the ten most common verbs, and went through about a dozen connectors (“and,” “but,” “therefore,” etc.) so that I could build sentences. The whole edifice of the language can be constructed gradually from there.
Things not to focus on
I intentionally deprioritized a few iteam:
- All rules without much explanatory power (i.e. those with loads of exceptions) and which can’t be explained in a simple sentence should be ignored until they keep relating to something that comes up in conversation. Speaking practice should take precedence over grammar study, so that you in a way almost “anticipate” rules over time, inferring them from conversations.
- Learning words that haven’t shown up organically in conversation should be avoided.
- Writing, too, should take a back seat to speaking (but not be altogether ignored) if you’re preparing for an interview, like I was.
- Reading texts that are too difficult (books, and some newspapers) might be tempting, and give an occasional thrill when you, say, manage to understand a full passage, but is a waste of time since the enjoyment is low if you need to look up five words in each sentence, and you will not have a feel for the exact meaning of the word anyway.
I believe that following the above guidelines can take a dedicated learner to an intermediate level in most languages in 50–100 hours, or 1–2 months of study. The path to fluency from there is straight, but often long. The better you get the more effort it takes to get even better, especially once you can already speak well, which is the final goal for most people. I think of this process as of an inverted pyramid which depicts the effort required to make progress. In the beginning you easily double your knowledge every few days. At the advanced level it can be tedious to learn the remaining exceptions and remember the fifth synonym for a word, in particular when you no longer really practically need to.
A few remarks on what normally holds people back when learning a new language, as well as the meaning and importance of speaking languages.
The Swiss education system is great. Yet here in Geneva (where French is spoken) a typical kid will go through a thousand hours of learning German (another official language of the country) by the time they leave the system, and few of them will have achieved fluency; indeed, many will not be able to hold any kind of conversation. Some of them come to regret this come job-hunting time. I’ve seen this pattern all over the world. Tuition is unimaginative, repetitive, grammar-heavy and immersion-light. Under such conditions, and in the absence of an obvious payoff, learning is a miserable chore. Yet it is exactly language learning that should be tailored so as to arouse interest through immersion and gamification, since tangible benefits come later than when learning most other skills. The opportunity for systemic progress here is substantial in many countries.
The immense challenge
When they hear (and believe) that it takes an English speaker 2,000 hours of study to learn Arabic even the most confident learner starts with the wrong mindset, expecting to perish inevitably along the way like a character out of a Kafka novel. A much better strategy is to see the learning as a sequence of rewarding steps which proceed in a rapid sequence. If you jump-start the language the way I did Hungarian then you may be a few weeks away from your first meaningful conversation, a couple of months away from a prolonged phone call, and you could be fluent (in most everyday topics) in six months, all with part-time study and while living your life for other priorities. Those who churn out estimates of the 2,000-hour kind probably assume classroom-style group lessons (which move very slowly) traditional learning methods (many of which are obsolete) and average ability to retain information (some of this is innate but most of it can be trained).
I wrote it above and I’ll repeat it here: to learn a new language it is useful to already speak one and truly understand its engine. It is not absolutely necessary (as proven by children everywhere) but it is very useful. Understanding the engine of the language means understanding why and how it uses tenses, declensions, gender, prefixes and word construction to convey meaning. In other words, it means knowing grammar. Until you can speak the meta-language of grammar sufficiently well the most useful language to learn is actually your own.
The fear of making mistakes is the biggest obstacle to learning that I’ve ever seen. Maybe this is a feature of our time in which we’re all supposed to project an image of success at all times, or maybe it was always like this, I don’t know. If you want to learn a language this is a very disadvantageous trait to have, and yet we all have it to a degree, I included. What helps is role-playing. When speaking a foreign language pretend you are someone else, so that mistakes don’t have to be shameful revelations of what a “stupid human being” you are deep inside. Exaggerate the accent, wear a silly hat, do whatever is necessary to detach your fragile inner core from the character you are playing.
Why be a polyglot
People around me are quick to ascribe my relative prowess to innate talent, and there might be some of that as well, but far more relevant is my methodology, as well as my genuine interest in languages. This interest comes from an image of myself which I retain from childhood — a Corto Maltese-like figure who travels around, engages with local people easily and hits the ground running wherever he is, from Juarez to Shinjuku to Bangui. I often fall comically short of living up to this image, but I also pull it off from time to time. Silly or not, this is where my drive comes from.
There is no inherent value to being a polyglot. Languages are merely a tool. You do it to read Baudelaire in the original, speak to your in-laws, understand an oratorio or have a beer with people at the local bar. Languages should not be one’s proudest achievement by themselves. Some of the most inspiring people I’ve met have been polyglots, but so have some of the saddest ones. The utility of language study can be easily questioned. Maybe we’ll all just be speaking English down the road, or artificial intelligence will step in to translate seamlessly. Or maybe we’ll need ever more polyglots to bring the world together, as it tends to neo-isolationism. It’s hard to tell. Such lofty concerns don’t create polyglots. No, they are created through their irrational drives and fortunate circumstances, as is everyone else who ends up making a difference, I suppose.