Is German really hard to learn—myth or truth? Here’s what Steve thinks:
Is German really hard to learn—myth or truth? Here’s what Steve thinks:
I moved to Austria in 2023. I was there for almost a year, but I went there with zero German, thinking, ‘I’ll just learn it once I’m there’… well, terrible idea. First of all, I made the mistake of speaking English whenever I felt overwhelmed or when people didn’t understand me. Secondly, I didn’t study for a single minute; as I mentioned, I thought it was just going to pop into my mind. So, I already felt frustrated. However, learning German is really important to me, since my girlfriend is Austrian and I want to return to her country. I would say I have a sense of the language, since I’ve picked some things up with her and I have been using LingQ for German for the last two months. I know basic stuff and basic vocabulary (1,700 words on LingQ). I do, however, know some rules which help me. As for speaking, I barely do it with my girlfriend due to my lack of vocabulary. I believe I should start with that later on, though now she helps me by mentioning things in German, like speaking randomly. I understand more than I speak, obviously. But now I really need to hustle and break the barrier. I believe that listening (Comprehensible Input) is really important, but in my humble opinion, you can´t just learn properly a language only through that. I think we do have to, at a certain point, do some things which we don´t like in order to progress. Yes, we are understood and being able to communicate is our main goal, but from my perspective I feel annoyed when I say “der” and not “dem”, and so forth. Writing and speaking (a few short sentences) at the early stage can be really beneficial for it. And I do recommend it even more when you are learning a similar language. That´s how I have stopped mixing languages like Portuguese and Spanish; and that´s also how I am getting better at German. It´s only my perspective
Mr Kaufmann presents a table of lexical similarity to prove that German shares a lot of vocabulary with English. Unfortunately he has not realised that the values he presents are calculated for a small list of the most common words, such as father and Vater. Lexical similarity is typically calculated for lists containing 100 words. The reason why linguists do this is to examine the genetic history of languages. If the most common 100 words are very similar, then that suggests that the languages share a common ancestry. The table tells us absolutely nothing of value as language learners, since we need to learn thousands of words. In fact if we look at lists containing tens of thousands of words we see that French and English have a lexical similarity of roughly 60%, whereas for German and English, it is roughly half that value.
He confidently tells us that German is easy to learn. I disagree. The grammar is not that hard, as it is pretty regular. What makes German hard is the vocabulary as English and German share far fewer words than English and French or English and Spanish. Perhaps he learns words easily, hence for him German is easy. I struggle to learn words, and hence for me German is hard, Grammar? That’s not so hard, no harder than French IMO.
I do wish he would not assume that his experience applies to all of us.
I think it does. It’s just a question of degree. And perhaps even more importantly, our tolerance of feeling stupid. If I need x time to learn something, you might need x0.8, a bright person will use x0.5 and a dimvit x1,5. The bright one could still however be the most frustrated: “why doesn’t it stick on the first go?!”, “why cant I understand Rilke after a full year of study?!”, and so on…
I have started German myself rather recently. Definitely feeling dumb, especially when I venture into the declination tables, but I couldn’t care less. Mann, Sebald and Kantorowich is in my sights, and I’ll get there…
You are simply missrepresenting facts here. You just can’t compare looking how much of english come from which source, but instead you have to look what are common between languages. It’s true that there is some 25% words of germanic origin in english, but if you look at the total shared vocabulary that’s about 60% because german and english share also huge number of words that have latin/romance origin. So just about the same as with french. That’s also my experience while learning both french and german. French maybe shares spelling more than german, but what is gained from that is lost in the learning of the spoken language. Personally I find it a lot easier to decipher those small differences in spelling and thus have direct visual link to learn the spoken as well, instead of trying to learn those spoken differences.
AI can gather this kinda information easily and if you like also show those most common shifts even if they are quite obvious if you try to find them.
There is just one little random part of Tarzan book that I’m reading analyzed by AI. Only small part of the shared words are unhelpful. Same is also true between english and french; not all cognates are helpful.
To me german has one advantage and that is that it makes words using existing words with prefixes/suffixes/combining two words so that you already have little information of the word and/or have something familiar that you already recognize instead of having to learn completely different words, which is often the case with romance languages where words have had more time to diverge from their common origin.
Biggest hindrance in german to me was lack of material in lingq in beginner 2 level, which is why I went almost directly to intermediate 1/2 and just battled through couple books. Some languages have better material, but all languages I have learned have common that getting through around 4000-8000 known words level gives a little headache. Even norwegien, that is highly similar to swedish which I leaned before it. That’s why I have found it important to just push on and not to get stuck in details. Everything get’s clearer the more information you collect.
I find German quite different from English despite the amount of vocabulary they share. So, regardless, do you think German is a language where you can just say “Oh, CI is the key” and magically learn it? Or perhaps SK has a different perception of what knowing or speaking a language entails. In my view, while it’s great to be able to communicate, can you really claim to speak the language if you make a mistake in every single sentence?
For the record, here’s Steve Kaufmann’s account of learning German:
My ability in German is considered good by native speakers and I have no difficulty in conversation, or in understanding radio and newspapers. Although I had exposure to German in my younger days at home, while working on a German ship and while working on construction in Vienna, I really could not carry on a conversation. In November 1986, I decided to spend the month learning German.
I went to many secondhand bookstores in Vancouver looking for reading books in the German language which had vocabulary lists. I must have bought about ten or fifteen of such books as well as various cassette tape products in German.
The readers were full of the scribblings of the people who had used them before. For one month I read and listened to tapes in German, with the help of vocabulary lists. Of course I achieved a dramatic improvement in my German. I was subsequently able to improve on this during visits to Germany. But it was hard work.
In German, the nouns have three genders. Word endings change depending on case. This is very hard to master in speech. Only repetitive listening has enabled me to cope with this difficulty, not explanations and lists. I just speak German the way I hear it and hope for the best. My German is natural and fluent, but I am resigned to the fact that I am not perfect in German grammar.
I never had to pass a test in German, but I have done business in that language. I have also traveled to the wonderful medieval towns of Germany, sat down in restaurants and engaged the locals in lengthy conversations in German. I know that I frequently get my genders and cases wrong. It has not held me back. On the other hand, I know that to improve my grammar I just have to listen and read more, and occasionally refer to grammar texts to reinforce what I am experiencing in the real language. The isolated study of cases and genders, without a lot of language exposure, will not enable me to improve my grammar in real situations.
–Steve Kaufmann, “The Linguist”
I don’t think Steve Kaufmann is a language savant, but he is a very intelligent, highly motivated person with a particular passion for languages. He also seems to be an extrovert which is helpful on the speaking side.
By the time he got serious about German in 1986, he had already been exposed to German via his parents, plus some vagabonding around Europe as a young man. He had also locked in French well enough to study at a top-level French university, then learned Mandarin and Japanese well enough to function professionally.
So, yes, I would agree that Steve’s experience of learning German likely doesn’t extrapolate well to the rest of us beyond the broad strokes of staying inside the language.
Quite the opposite.
For reference I believe this is the source for his table:
Mr Kaufmann stated that the lexical similarity between German and English is about 60%. He did so to support a claim that they share a high percentage of words. In fact he has misrepresented the lexical similarity calculation, as it is usually based on a small list of the most common words in the two languages. English and German share a common ancestor, or closely related ancestors. The most common words, such as mother and father, son and daughter, change the least, That is why the lexical similarity of the top 100 words is so high.
Note that although the table in the link does not give the word list size, the article links to Swadish lists, and a standard form has 100 words.
If instead you consider a far higher number of words, you will find that the lexical similarity is much closer to 30%. (I can’t find a source, but you can search if you wish, for lexical similarity calculations based on tens of thousands of words.)
That is incorrect.
You are not a native English speaker, you are a native Finnish speaker, so your experience is relevant to L1 Finnish speakers, not L1 English speakers.
I agree that there is a lack of good material below B2. Most of it repeats the same limited subset of vocabulary. That said, organisations such as Deutsche Welle might have good paid material.
My own experience is that I could learn a lot of French by listening to podcasts, even words I did not know were often intelligible due to shared word origins. With German that simply is not the case,
You are making a fundamental mistake which is to accept the output of an AI engine as fact (1). Nonsense in, nonsense out, You need to go to original sources, and then check their references,
Firstly, the above states that 60% of English vocabulary shares a common origin with German. I do not believe that. If it was true, I would be learning German vocabulary fast, Secondly, we are actually interested in the percentage of words in modern German (including words of French origin) that have a recognisable cognate in modern English, and that value is according to estimates that I seen of the order of 30%.
Yes glue words in English tend to be Germanic and the (small) core vocabulary is quite close. But no, I find it much harder to understand unknown German words compared to French. And even though glue words such as ‘in’ appear to be shared, they change their meanings and/or usage significantly. The prepositions bei, an, in, über, auf, and aus for example do not map to English prepositions. Where we might use in, Germans might use auf, and so on.
(1) Russia is taking advantage of the rise in AI. Each day they publish online hundreds if not thousands of articles. These contain propaganda, telling us that the West is corrupt, and President Putin is a good kind man who has the welfare of his people at heart. China does likewise, AI is trained by scooping up online data, and hence AI engines are absorbing this propaganda, and then feeding it to users. A French media outlet did an investigation, and Grok for example would tell users that Taiwan is part of China, and the Chinese territorial waters extend almost to the costs of Vietnam, Thailand and other nearby countries. And of course Grok gave a false account of the treatment of the Uighurs.
Exactly.
As you said, he has passion for languages, and the determination to work hard, and perhaps that is his secret.
With all due respect, that is simply not true. He makes mistakes constantly. In a language where word order is relatively flexible but declension is fundamental-such as German-one can be easily misunderstood. It is not as simple as saying (wrong) “he are”; an incorrect declension can alter the meaning of the entire sentence. He has fallen into fossilization, and the same has happened with his Portuguese. Relying solely on CI and assuming it mirrors Spanish often results in speaking Portunhol (a common pitfall), making a lot of mistakes and mixing concepts that are different in each language.
It is, indeed, a compelling idea to believe that CI is all that is required, and while it is undoubtedly a crucial part of the process, it is only one piece of the puzzle.
Even native speakers, who indeed study grammar at school, fall into those mistakes. This is just nonsense.
Here’s the thing, on paper I am very bright as I have a degree and PhD in physics. When given a maths problem, I can often ‘see’ the answer intuitively. I can’t explain how, but I subconsciously solve the problem. Whereas many other people can’t. And yet when I was learning to play hockey, I was a slow learner. Others would pick up drills by watching, and skills developed naturally. I struggled to pick up drills, and I never did pick up skills that others acquired easily. I am a good ice skater, but I learn slowly, and cannot pick up moves ‘naturally’ the way others do. In short, at sport I am a moron.
It’s the same with languages. I have to plod my way along. However, I have a very analytic mind, and I have been told by countless people that my accents are very good including Scottish English, Northern Irish English, and Yorkshire, and several native speakers said that my pronunciation of the Welsh language is very good. Some Germans were surprised at how well I pronounced a phrase they gave me, when I knew no German. I don’t find grammar so hard, probably because it involves learning a small number of patterns. I cannot learn it ‘naturally’ i.e. I cannot infer from context.
If you’ve ever taught maths, you’ll know that people learn differently. For some it is intuitive, for some they need it to be spelt out step by step.
You can find on YouTube an interview with Dr. Atsushi Mizumoto, whose specialism is vocabulary acquisition. In the video he states that his research shows that different people learn vocabulary in different ways. So there is no one size fits all learning method.
Can we learn to learn? To some extent, yes, in my opinion. I think a person can improve by finding out what suits them.