It would be interesting to know 1. how he challenges the idea (what is meant by that specifically) and 2. whether he provides any clues or proof to his claim.
I would argue that even our perception of the effectiveness of a strategy impacts said effectiveness. You can see this on a bigger scope with people thinking that talent is important. If they assume they won’t understand it anyways, they most likely won’t (or at least not very fast). The same will apply to learning strategies, I assume. If you consider comprehensive reading to be superior to spaced repetition, the first might indeed end up working better for you (at least to some degree).
Spreading out learning over time is a bit of a vague formulation. The original idea of spaced repetition is that the time that passes before reviewing an item should be choosen so that the review happens shortly before you are about to forget the item. Using variable spacing intervals like ANKI does is just one system to apply this theory, and I myself mistakenly described this as spaced repetion in on of my earlier posts. However, the term spaced only refers to having long time intervals between the repetion, but they can also be static. Studies even suggest that while spaced repetion is an effective learning strategy, there is no proofable difference in efficiency between static and variable repetition intervals.
Applying interleaving in the sense as in the quote might not be doable if the studying sessions are relatively short. However, interleaving could also happen on a grander scale. So that instead of studying the same way every day, you change it in a somewhat irregular fashion.
A similar approach is applied by the International School in the city I am living. Instead of having a five day schedule (so a schedule for monday, tuesday, …) they use a seven day schedule. This leads to a shifting so that the lessons you have on monday, tuesday and so one differ from week to week.
They also exchange the teachers every two years. The idea is that this avoids the students to become too focused on routines and improve their mental flexibility. I don’t know whether their is any research that back ups this approach or that investigate possible downsides.
As you and @michael.mcd were referring to a book, which was referencing research, I was pointing out that I am sceptical that this re-reading not being as effective as retrieval practice might not necessarily be referring to language learning. How many of the studies were testing on language learning? I imagine it’s incredibly hard to measure the total impact in language learning when re-reading. In 5 minutes, you could read 500+ words. This is slowly drilling in all those separate words, collocations, phrases, grammar, etc. (so 1,000+ variables), whereas 5 minutes of flashcards might be drilling, what, 10 words or maybe 50 words if you are using sentence cards (so idk 100 variables?).
It depends on your goals. Sometimes there is nothing wrong with a “limited degree of testing.” As an example, consider someone who reads Italian literature. They will encounter alternate, “archaic” spellings of words. Surely it depends on your goals on if it’s important to “highlight gaps in knowledge” on if you can actually produce or correctly identify it from a misspelling. I looked up an alternate spelling which I may have seen only a handful of times just now as an example (leggiero) and I spelt it correctly first go. For me, even if this was a knowledge gap and I couldn’t accurately produce this alternate spelling found in literature, personally, I don’t consider it important, based on my goals. My goal in this case is being able to solely recognise the word in context. Were the studies against re-reading as a strategy taking into account that sometimes the goal is merely recognition in context, not recognition out of context or production?
Are you intentionally using Stephen Krashen’s definition of “acquired” vs “learnt” here?
In my case, I think it’s partially because the sheer volume of vocabulary is huge, but also I do opt for more ‘holistic’ techniques. Reading while listening is really practising many aspects at the same time - vocabulary, reading ability, listening comprehension, listening speed, understanding different accents, understand grammar, etc. Even though the pop sci book you two are quoting only refers to re-reading, not re-reading while listening, it would be so hard for the researchers to measure every single micro improvement on every single variable in language learning, hence why I’m sceptical that their advice was even applying to language learning.
This is clearly a faulty premise. It’s just about as accurate as saying that all have the same skin colour. Similarly there are host of invisible differences. I mean, that’s the whole basis of evolution. This applies to comprehensible input aswell. It doesn’t work the same to everyone, whether it’s because of innate or assumed differences, but it’s likely that it works to big portion of people relatively similarly. If it takes someone 1000 hours to reach same level that someone reached in 500 hours, it still doesn’t mean it doesn’t work if you don’t feel difference after 100 hours.
The ideas presented in the book are based on numerous research studies carried out in the field.
They don’t adopt the Stephen Krashen and Noam Chomsky approach which can be summarised as “It’s true because I say it’s true, I’m a very distinguished scientist at a renowned world class university with more papers than you can wave a stick at, so I’m not going to waste my valuable time arguing with a little worm like you, so just go home”.
Guys, thanks for your responses to my post. It generated a little more heat than I expected.
Some very well-informed observations from both @LeifGoodwin and @nfera. And some good advice from @Pr0metheus and many more. Thanks.
A little worrying that some folks dismiss references from a popular scientific book on learning research because it’s popular.
FYI: the quote about rereading is in the context of students rereading their course notes and is not referring to reading or reading with some (20%?) novel words à la LingQ or graded readers, etc.
PS My observation was simply that the stats here were misleading for me, and I was wondering what others here thought.
You do come across more words by reading, but the science shows that in order to learn a word, you need to come across it many times over a long period of time, and for uncommon words reading may not provide sufficient encounters with the word for it to be moved into long term memory. So in this case quantity does not best quality.
It’s all very well coming across obscure words, but most likely you won’t move them into long term memory through reading alone.
FWIW I adopt a mixed approach. I spend one hour doing free listening which consolidates known words and grammar, and might get me some new words and grammar, but more likely I will forget them. Then I do an hour assisted listening with a transcript, occasionally stopping to consider new words and phrases. If I spend enough time playing with a word, using Google to find pictures say, I might learn it there and then. Usually it has to go into Anki so I can train my Gruyere based brain to retain it. I make no claims that this is the best approach, but it seems to work for me.
No, but I have triggered you. I’m sure you twitch at the mere mention of Saint Stephen’s name.
Are you siding with Stephen Krashen?
Ultimately language learning is just learning, albeit with a huge quantity of information to ingest.
You might be interested in the Declarative Procedural Theory of language. This holds that language does not have a unique cerebral apparatus, rather language evolved by subverting existing areas of the brain, and in particular declarative and procedural memory. This is consistent with how evolution works. Thus language learning proceeds in the same way as other forms of learning. The theory does make testable predictions which are consistent with observations. I’m afraid you’ll have to do your own research for more details.
I din’t think it is fair to refer to “Make it Stick” as pop-sci. The two authors are recognised experts in cognitive science who summarise the research in non academic language. Pop sci is largely junk.
One point I will make is that there are techniques to make reading more effective. Research has shown that memory formation is more effective when associated with emotional stimuli. Thus we recall certain events from our childhood, whereas memories of day to day events have long since gone. I remember the day when airliners hit the twin towers, and the day when men first orbited the moon in 1968. I can sometimes remember words if I build up an emotional and contextual wrapper i.e. add pictures, emotions etc. I don’t see this discussed anywhere, though I’m sure these are well known methods. And I’m sure that good teachers use this phenomenon to improve learning outcomes. Or to put it simply, the students have fun.
I would counter that. Books generally use much more wide vocabulary and are probably the only way to encounter words that aren’t commonly used in spoken language. Writers also tend to have a certain style and frequent some words more than their fair share. If you don’t encounter words by reading, I doubt they are that important. Besides going for C2.
You do need to read, and listen, in order to encounter new words. However, research suggests that to get those words into long term memory, you really need to have additional tools such as Anki. Of course if you want woodworking vocabulary, then listening to a shedload of carpentry videos might do the trick. But just waiting for words to pop up is not so good. Don’t forget that a native acquires vocabulary ‘naturally’ but it takes many, many years and 100% immersion.
You do make a good point that many words are not really needed, so the student needs to decide on their aims. In my case I want a near native vocabulary, but only that of an average native, not some French literature graduate.
I would still say you get enough repetition if they are words that are used enough. In the meantime you get repetition of words that are more common, which might in the end be the one that does the trick to learn those less used words. Many times it’s the context what makes the words stick, not repetitions
I am a literature guy, but I just gave up reading Wolfe’s “The Electric Kool-Aid Test” in French because Wolfe accesses a huge reading vocabulary, lots of cultural references and slangy expressions plus experimental writing techniques.
Furthermore, the translator was clueless about the 60s hip scene in California.
I am ambitious in my French reading. When I started I didn’t mind wading into a text with over 50% unknown words. But “Acid Test” just wasn’t worth it in terms of ROI (Return on Investment).
Oh yes. I found it all productive. I entered my study of French in a highly passionate state, which I still can’t explain, but it was all exciting and beautiful.
Two years later French is my old lady. Not so passionate, but settled in our routines and happy.
Thank you for taking the time to respond and share your experiences.
For context, I’m in a spot now with Chinese where I know most of the very common and “connector” words instinctively, and when I know the words in a sentence I can almost always get the meaning.
That means the next phase is going to be mainly just grinding vocabulary for quite a few months.
Native content is more interesting and higher quality, but because my vocabulary is small it’s frequently very high % unknown words. (This is complicated by LingQ’s issues with splitting words in Chinese, but that’s another topic.)
I’m motivated enough to slog through this kind of content, but I’m unsure whether it would be more productive to just keep doing intermediate content geared towards learners. Your experiences are valuable input for me, and I thank you again for sharing them.
You missed the point. You are still practising even if you aren’t repeating some words frequently enough to make them stick. With flashcards and unknown words you are spending quite a lot of time just to remember those words. Before you learn the words, you aren’t gaining anything. Using those to increase frequency that you meet them might be useful, but if you don’t meet those/can use them in real live they aren’t going to stick in the long run very effectively.
I used to listen to content where I only understood the gist, not the details. It helped me automatise what I already knew, and wasn’t as boring as learner content. Was it efficient? Was it a good idea? Do more experienced lesrners do this? I have no idea.
With flashcards you aren’t picking up anything else except the words you are trying to learn and those only after you remember them. With reading you are using the whole time to practice. It might feel like it goes slowly, but your brain is learning as fast as it can. There are now shortcuts. Those that you have learned with flashcards you only know in isolation. You still need go through the process of learning them in context to pinpoint the meaning. For example if you try to translate english word X to your target language, you aren’t choosing the first option or randomly choose one of the translations. You choose what sounds right in the context you are going to use it. That is quite impossible to learn using flashcards. Better would be to create content where your target words are repeated, if you feel like you need to learn specific words.
I don’t need to read about. Comprehensible input also is spaced repetition and I know it works. I just don’t see flashcards as an effective use of time for reasons that I mentioned above. I would argua that flash cards are what gives the illusion of learning, to go back to the original topic.
I think learning expert Justin Sung makes some good arguments against (wider) Anki usage in this video:
The main points are:
Anki is inefficient and gives a low level of memorization
It’s not how we learn vocabulary in our native languages
Not all words are equal (there are words that we pick up instantly, words that require seeing a few times, and words that are hard to get to stick, where Anki is only useful for the second group)
I can see its usefulness for learning Hanzi, Kanji, minimal pairs, etc., but it’s both inefficient and boring for learning vocabulary.
I would say that sums up my ideas, even if sum up isn’t what you would call a 30 min video. I like that he briefly touched on etymology of words. Personally I have found that, on top of learning several languages, helpful for seeing those parts of the unknown words that might be already familiar. Although I don’t read about etymology just to learn languages, but I also find it interesting.
