When it becomes so fun to read

At what word count it becomes so fun to read that you can’t stop?

Hi LearnTillDeath,

You mean some kind of “flow state”? I’d say it depends on

  1. the distance of the L2 to be acquired from your L1 and the L2s you already know well
  2. the level of difficulty of the literature
  3. the reading mode, i.e. assisted or unassisted reading.

Ad 2)
If you have to read difficult, e.g. scientific literature, it will always be hard (= no fun) as long as you haven’t mastered the specific knowledge domain (i.e. the theories, methods, etc.). This is true for native speakers of a language and even more so for non-native speakers.
The same is true for other text genres such as poems, which are often notoriously difficult.

And even if you choose popular fiction: if it isn’t contemporary (e.g. literature from the 17th, 18th or 19th century), you should also expect many comprehension problems due to outdated vocabulary / language structures, which isn’t “always fun” either.

Ad 1) + 3)
Case 1: assisted reading

  • If the L2 isn’t too distant and you practice assisted reading using audio reader software like LingQ, I’d say the threshold is ca. 2-3 million words read, depending on the text genre (Note: contemporary popular non-fiction texts are usually easier to digest than their fictional counterparts!).
  • If the L2 is more distant (let’s say an English / German / French, etc. native speaker tries to learn an Asian language), the threshold is probably higher: maybe 3-4 million words read (if I’m not mistaken here).

Case 2: unassisted reading, i.e. without LingQ and Co:

  • For not too distant L2s >= 4-5 million words read
  • For distant L2s >= 5-6 million words read.

However, you should take all those numbers with a grain of salt. They are only “rough estimates” and in concrete individual cases there may be more or less large deviations.

Hope that helps to get a feel


Hey Peter you cited a bunch of studies on a thread that I can’t seem to find and I just kind of encountered the things discussed in the article.

Language learning is like any other skill like free throws in basketball and your brain and body adjust to make you better at the skill. But in Language learning there is like you describe an educated native speaker who will struggle with something written 300 years ago or poetry for example, considering this how does language learning fall into this basketball paradigm? How does flashcarding vs reading/listening either each level fit into this free throw model? After reading your posts I would very much appreciate your opinion on such a thing. Obviously language learning is just exposure but if one were to conceptialize it with models, levels, and domains/content areas how would this look? You are very good at simplifying the complex as seen with this post above. Any help would be much appreciated! Thank you!

And how about the number of known words we need to get into that flow of fun reading (of ‘‘normal’’, everyday texts)?

If you want to read books above your level and still enjoy the plot in flow state then you can start with bilingual books. This way you can simply take out the problem of facing unknown words and dictionary lookups. Check out dtv website or Amazon for such books.

They offer bilingual books of famous English novels in foreign languages.

Also, watch a video of Olly Richards on 8 ways to increase your reading input.


Hi Hagowingchun,

“Language learning is like any other skill like free throws in basketball and your brain and body adjust to make you better at the skill.”
The problem is that these are high-level generalizations such as the following: “All processes of skill acquisition are practical. And if someone doesn’t practice, skill development can’t take place.”

However, the devil is in the details. That is, at more detailed levels, skill A can be very different from skill B. Or rather, there can be some similarities, but also many differences. Thus, it really depends on which aspects are compared at which level of abstraction or detail.

For example:

  • All cats are carnivorous mammals.
  • Some cats are predators that can kill humans, while their domesticated equivalents can be a member of a human family.
  • A domestic cat can be a house or farm cat, but there are countless cat breeds with (very) different features (see: Cat - Wikipedia).

In short, the higher the level of generalization/abstraction, the easier it is to compare cats because all the low-level complexity is abstracted away.
However, once you focus on the details, all the low-level complexity will “haunt” you to show you how different cats really are!
The same is true for comparing skills and the corresponding skill acquisition processes!

Therefore, the general lesson to be learned here is: There’s a connection between reducing complexity by generalization / abstraction and information loss.

“Obviously language learning is just exposure but if one were to conceptialize it with models, levels, and domains/content areas how would this look?”
The answer depends on the

  • selected level of generalization / abstraction (see above).
  • aspects you are interested in
  • models that you find useful in one context, but maybe not in another (see the “General Model Theory” by Herbert Stachowiak: Conceptual model - Wikipedia).

So if you’re interested in the performance of experts in different fields, e.g., pro basketball, polyglotterie and top-level mastery of chess, you could try to identify similar patterns of behavior or mind-body connections that can be generalized and applied to even more distant fields.

See for some concrete examples that (might) lead to general theories of skill acquisition / learning:

“will struggle with something written 300 years ago or poetry for example, considering this how does language learning fall into this basketball paradigm?”
In this case, there are “a lot” of differences so that many comparisons / analogies will fail.
However, on a more abstract level you could ask, for instance:

  • Are there similar mechanisms reg. focused attention that are valid both for deciphering (old, poetic, etc.) texts and the body movements of your opponents in a basketball game (in the sense of “reading / anticipating” the game in a kind of “flow” state)? And if so, how could you train them?
  • Or in a similar vein: Could general pattern recognition mechanisms be at work here?

In many cases such analogies/comparisons will lead to nothing. But there may be some cases where some “creative” analogies may also lead to new intellectual discoveries :slight_smile:

I’m not sure if those answer attempts are useful to you, since your questions touch on difficult topics (about which tens of thousands of scientific pages have already been written):

  • How do we construct categories / prototypes in a non-essentialist way?
  • How do we deal and interact with complex phenomena that are completely different from complicated phenomena (Note: The latter can be reduced to some simple or basic principles, but the former can’t).
  • How important are analogical processes for human thinking?
  • How do we build and use (mental) models, frames, scripts, and schemas?

If you are interested in such (more general) questions, Doug Hofstadter (“Gödel, Escher, Bach”; “Metamagicum”, etc.) and other greats of cognitive science/psychology are good starting points for further investigation.

Hope that helps,


1 Like

It depends on the “distance” of the L2 to be learned, the difficulty of the text (“Harry Potter” is easier to understand than “The Lord of the Rings”, for instance) and if you want to read contemporary popular non-fiction or fiction texts:

Some “rough estimates” for not too distant L2s are:

  • ca. 40-50k for contemporary popular non-fiction texts.
  • ca. 50-70k for contemporary popular fiction texts.

However, it’s better to focus on the overall number of words read and listened to. If you do that, your other stats on LingQ will skyrocket (including the count for known words).


@PeterBormann: Thank you for sharing these “rough estimates”.
I had been under the assumption that I should be able to read books at advanced 1 level, however I was struggling with some of the french books I have tried (even though some books worked just fine). So it is useful to get the confirmation that I very likely simply need to spend more time with the target language.