Thanks to all who participated. Here are the (most important) word pairs that caused problems:
Scale: 1 – Sound the same (mit/mit). 2 – Very Similar (drücken/drucken). 3 – Similar (Pfüte, Pföte). 4 – Different (Pfeet/Pfit). 5 – Completely different (Meat/Mat)
Some analysis: The long vowels in biete/bete [i]/[e] and bote/bute [o]/[u] are extremely close in German, and I wasn’t surprised to see them at the top of the list.
Then comes a whole mess of difficulty between ströck [œ]/strück [ʏ]/struck [ɪ]/strick [ʊ]. Every combination of [œ], [ʏ], [ɪ] and [ʊ] was difficult, except for struck [ɪ]/strick [ʊ], which exists in English already (put/pit). I had trouble saying them a few times, which I’m sure didn’t help, but it’s a pretty challenging little group!
Aside from the major problem between the German [u] and [o] (bute/bote), there were some minor difficulties with every combination of büte [y], böte [ø], bute [u] and bote [o] except Büte/Bote [y]/[o]. Why that last combo seems easier than the rest of them may just be experimental error (For the statisticians out there: average 3.8, standard deviation 0.8)
The other minor difficulty seemed to be strock [ɔ] in contrast with strack [a] and struck [ʊ].
I’ve been preparing a list of tricky sounds in German, and this may be the last piece of the puzzle. I didn’t bother to test short vowels (Offen [ɔ]) vs long vowels (Ofen [o:]), because I already knew they were hard, and the quiz would have taken forever. Examples of these pairs are:
Offen/Ofen [ɔ]/[oː] Ass/Aas [a]/[aː] Busse/Buße [ʊ]/[uː] (Easier: similar to put/poot) Bett/Beet [ɛ]/[eː] Widder/wieder [ɪ]/[iː] (Easier: similar to pit/peet) Hölle/Höhle [œ]/[øː] fülle/fühle [ʏ]/[yː] Bären/Beeren [ɛː]/[eː]
The only other vowel problem in German seems to be the two schwas (which is quite difficult!):
schöne/schöner [ə]/[ɐ]
I could also imagine some tricky consonant pairs between German and English:
German/English
dich/dish Buch/book Bach/balk rot/wrote pfan/fun Tier/tear (drop)
As well as some vowel/diphthong pairs between German and English:
German [o] (Not) vs American [oʊ] (know) German [e] (See) vs English [eɪ] (Say) German [ɔʏ] (neu) vs English [ɔɪ] (boy) German [ɐ] (schöner) or [ə] (schöne) vs American [ɚ] or British [ə] (river) and perhaps German [u:] (Ute) vs some dialects of American [ʊu] (loon)
So….what?
As I mentioned in the previous post, I’m planning on making some effective pronunciation tools for as many languages as possible. There just isn’t very much for pronunciation out there, aside from Youtube tutorials and a few pronunciation guides online, and very few tests with immediate feedback, which seem to be the most effective method studied so far (there are a few apps that will test you on your ability to hear Chinese tones, but I’m not aware of anything else like that):
Tester: Is this “schöne” or “schöner?” (pause) ”Schöner” Student: Answer: “Schöne” Tester: Incorrect. It was “Schöner.”
In the linked study above, three 20-minute sessions of this sort of testing were able to (permanently) teach Japanese speakers to hear the L-R distinction in English. Anki provides immediate feedback, so there’s no reason why we can’t have some decent tools based upon this research, if we can find out where the problem sounds are and make good recordings. So stay tuned. I don’t know how long they’ll take, but I’ll get them done.
Some final notes on the quiz (and a link to the excel file!):
This was not a robust scientific study, although I do think I got good, useful data from it. I’m not a native speaker, so I would have to do it again with a native speaker recording to get valid data. Also, I’ve learned that I’m not supposed to use the 1-5 Likert scale for this sort of test. Rather, I should say “Struck/Strück…..Struck” or “Struck/Strück…..Strück”, and you’re supposed to say which one it was. I kind of prefer the Likert scale because there are times when I can hear a difference, but it’s so slight that I know it would cause me problems in the future, and I certainly would have trouble saying it, but I definitely understand the criticism. I’m also not supposed to say why I’m doing the test (“This is to figure out the hardest sounds in German”, etc). I’ll keep some of this in mind for the next test (probably French).
It might seem like a small thing, but I must say I am impressed to find someone on the Internet who (correctly!) cites research in support of a claim about language acquisition. Very interesting article, too.
(1) I do not find vowels too difficult because a long, long time ago I got interested and learned about the IPA, and by “learning” I mean the logic behind their symbols (tongue up-down and back-forward, and lips stretched-rounded)
That’s pretty much all there is to it, and the rest is simply seeing your target word described in IPA and reading it aloud a couple of times (and later, perhaps, although ideally at the first encounter - hearing it said by a native speaker)
BUT ANOTHER REASON FOR MY COMMENTING HERE is this:
(2) I read about your method, and it seems my own experience overlaps with yours by 70 percent, probably.
I am experimenting with learning foreign languages by reading frequent word entries in a monolingual dictionary for foreign learners (most European languages at least are well provided with those).
I do not try to remember them actively (with a flash card program or some such gimmick - those I would consider inferior as the format does not allow for multiple senses and tons of examples easily).
I just read the dict like one could read a book - an exciting book, which I do not need to break from (all explanations are within).
WHY? - to break from the “frequency cage” THIS IS THE MOST IMPORTANT PRINCIPLE here.
Most common words cover largish parts of texts (e.g. 2000 could come to 80% for non-specific texts ) - therefore if one tries to acquire new vocabulary from reading, as is often suggested, one has to fight against the “frequency conspiracy” - words beyond the first 2000 will come up after thousands of frequent words, and only in one of their senses.
By reading a dictionary you beat the Frequency Conspiracy against you - you are presented with all meanings, in correct contexts and collocations, in examples drawn from natural corpora.
You save enormous amounts of time by not trying to get vocabulary from reading.
THEREFORE - reading a dict becomes one’s major method of language acquisition.
As dictionaries do not come with soundtracks, I read them in my own internal voice, at this stage.
This remains almost my sole practice until I acquire PASSIVELY 2000-3000-4000 words.
THERE IS ANOTHER MAJOR REASON for this: there is no known human intellectual ability, perhaps, in which passive skills would not cover areas several times as extensive as those in which one’s active skills are manifested. Language is no exception.
It is utmost stupidity - in all language courses - to demand that new vocabulary and grammar would be remembered and used, all of it, immediately upon learning. NO! - get yourself presented with 2, 5 times more of correct language first.
It coincidentally avoids being crushed by excruciating effort, as ‘just reading’ a self-explanatory text, which a monolingual dictionary is, is pleasure, not pain
Then I switch to (a) correcting how the lang sounds, and (b) to activating my knowledge.
(a) comes as a Harry Potter book in the target language (there are few if any without translations) accompanied by an actor reading the full text (Yes, best actors have been hired and produced audio versions of those).
With say 3000 to 4000 thousand words passively in my head already (took me about 1 month of reading Robert dictionaries for my French about 2 hrs per day), it’s a pleasure, not a toil.
The froggy actor (at 3/4 the full speed, modern digital recorders can slow it down without actually distorting the tone of the voice or making it unnatural) is buzzing in my ear, I stop the recording at each and every sentence making sure it registers in my head, and so it goes on. All properly said, with 100% native intonation and ways to convey emotions etc - as the actor performs with gusto on the audiobook version I have.
This sentence-by-sentence reading/listening is different from just watching some Potter-based movies with rewritten dialogues.
And the next stage is ACTIVATING this all.
First - by rereading already studied chapters, then by retelling them to myself in my head, then by starting to write all kinds of opinions, micro-essays, or doing scenes suggested by regular textbooks - only now, when those have long fell “below” my level of knowledge.
THEN and only then it would be listening to radio/TV (i.e. recognition of unknown language) - it is useless to try to recognize speech when you have no idea what those words and/or idioms mean
GRAMMAR is something that I go on very lightly, it’s actually not as important as many, many examples of words in correctly constructed sentences picked from language corpora, i.e. 100% natural native speech — which are supplied by the monolingual learner’s dictionaries
———
P.P.S. If you are interested, I could recommend you a couple of dictionaries for learning Russian for reading.
One is based on “the Longman Activator”, i.e. contains about 600 thematic, notional categories written as dictionary entries, good for “just reading”
The other one is mimics the famous Collins COBUILD definitions style and covers about 7000 most frequent Russian words (which are marked in bands, so one could pick “the first 2000″, or the next band to, say, 3000 or 5000 and so on).
Both can be found on the NET
P.S. If you are interested, I could recommend you a couple of Russian learner’s dictionaries for dictionary reading.
One is based on “the Longman Activator”, i.e. contains about 600 thematic, notional categories written as dictionary entries
good for “just reading”. This one is good for collecting several ways of expressing notions together.The other one mimics the famous Collins COBUILD definitions style and covers about 7000 most frequent Russian words (which are marked in bands, so one could pick “the first 2000″, or the next band to, say, 3000 or 5000 and so on).
Both can be found on the Net.
——————
You never mention on your site how familiar you are with, let’s call them, “applied linguistics” marketable products generated by the extensive EFL industry.
Since 1986 most learner’s dictionaries of English became corpus-based (gradually, but for all major publishers now). This is a new era for linguistics, when facts about a given language can be based on the realities of its actual usage, rather than on what dictionary compilers think, even if their opinions are informed by card-collecting activities.
But the pioneer in the field, Collins COBUILD project, went farther than that. They redesigned definitions to become complete English sentences, which are written very precisely to demostrate who or what, for example (a person, a thing, a subclass of those, e.g. “politicians”) does what in what phrase to what class of people or things - in other words, their full-sentence definitions are not only “readable” in the fullest sense, but show correct patterns and collocations, the environment in which the word lives.
This is further illustrated with 100% natural corpus-drawn examples of usage.
In the beginning of 2000 a group of Russian linguists inspired by the then well-known COBUILD dictionary line, decided to do a similar description of the core of the Russian language. They collected a corpus of actual Russian usage (mostly documents published on the Internet by that time), and formulated their explanations of the 7000 core Russian words (each having a number of meanings and usages) in the style of COBUILD entries.
You can really read it, like a novel.
Here is an example entry for you:
(1) Конец какого-либо периода времени - это его последний момент, час, день и т.п. “Конец учебного года”
(2) Конец какого-либо объекта - это его последняя граниза, точка в пространстве
(3) Концом какого-либо дела, действия является его завершение
(4) Если вы в каком-либо деле идете до победного конца, значит, вы стремитесь к его успешному завершению
……….
……….
(21) Вы говорите нет конца (и края) чему-либо, если хотите подчеркнуть многочисленность или продолжительность этого. “Вашим жалобам нет конца”
…….
(23) Вы используете выражение из конца в конец, когда хотите подчеркнуть, что движение происходит от одной крайней точки территоррии до другой
Hi Alex !
Those Russian examples are really lovely. Just to confirm, you’re talking about the Толковый словарь Дмитриева?
http://dic.academic.ru/dic.nsf/dmitriev/1897/конец
Exactly.
I did not realize it was on-line in full. Great!
(I have a scanned electronic text, rather a set of pictures of its pages packed as a pdf file).
It is almost ideal for foreign learners: has all forms of the word up front, indicates stress, and of course those COBUILD-style definitions are so helpful.
The printed version also marks the frequency band for each entry.
——-
It’s great it’s on-line. Means one can browse, then copy and paste entries into his (Anki) or whatever other tools or even text files one finds it easier to keep the information in.
——
The other one is like a thesaurus with notes. So if you’d like to “compact” various words into notional groups you could use that book (Шушков “Толково-понятийный словарь русского языка).
That one is based on the idea of Longman “Language Activator”
.. only it looks like this on-line version contains just 3200 most frequent words.
(80 groups, 40 words in each, it looks like)
The full version in my PDF (or the one in a bookstore in Moscow) has 7000 of them.
But even so, this should be a great resource: it’s a true corpus-based learner’s monolingual dictionary of Russian with COBUILD-like full sentence definitions that reflect collocations and usage patterns
гранизаграниЦаSorry for all other typos and errors. Ignore them.
Hi Gabriel!
I suck at the tones thing - no musical ears for me. I have been going over the IPA chart, and I understand the concepts, but it is not something that I can easily replicate with my own mouth apparatus. I was wondering what tips do you have for someone to improve their hearing and pronunciation skills. I know you say to enter a language pronunciation first, but I am awful at it… I’ll never learn a language this way because I’ll stay stuck at this level… I accept it’s just something that will slowly get better with time (it took me a while until I was able to hear different tones when I lived in China, just my ear getting used to it) I can usually mimic a native speaker right away with the sounds correctly, but when I try to recall the words laters, I am totally off mark.
Just wondering as a musician if you know of specific exercises to loosen up the tongue and fine tuning one’s ears?
Thank you!
M
If you’re working on Chinese, you’re in luck. You’re in the one language that actually HAS some good pronunciation tools. You’re looking for programs that quiz you on the sounds and give you immediate feedback. http://www.archchinese.com/mandarin_chinese_tone_drill.html
or
http://www.bbc.co.uk/languages/chinese/games/tones.shtml
are both decent examples of what I’m talking about. Once you can reliably hear them, you should start imitating the sounds you hear. That *should* do the trick
Gabriel!
I am learning Russian. I have a real hard time with vowels in general with foreign languages - I am a spanish speaker, so for me there is just one sound for a/e/i/o/u each. I can’t distinguish the difference of vowels btwn father/about or beach/beech/bitch and it is where my American accent falls apart.
Any tips?
Yeah! You need to test yourself with immediate feedback. Get recordings off of Forvo or Rhinospike with words that are different by only one sound (beach/bitch, for example) and quiz yourself on them with Anki (“is this recording beach, or bitch?” If you do enough tests, your brain will rewire itself relatively quickly (the study I cited above taught Japanese speakers to hear Rock/Lock in 3 20-minute sessions)
Great post! You really went into great depth here. As someone who spoke German as a first language it’s easy to pick up on the nuances found in the vowel sounds of the German language. Yet, I see so many English speakers having a hard time distinguishing between these sounds.
Your post was really insightful, thank you.
Gabriel
This site is the bomb and I am currently using many of your techniques to brush up my long neglected German. I’m really enjoying the process and can see myself extending it to other languages in the future. I am particularly grateful for your recommendation to learn the IPA which I’m going through at the moment.
The breadth and quality of the information on the site is excellent. I did notice a small mistake in this article that could be corrected. In places you mention: struck [ɪ]/strick [ʊ]. Shouldn’t this be struck [ʊ]/strick [ɪ]?
Once again thanks, and apologies for being a pedant!