Does the Input Hypothesis & Comprehensible Input work for learning Chinese? Can Mandarin pronunciation and tones be acquired through comprehensible input alone or is in necessary to learn rules, use rote memorisation and repetition drills to become proficient? Can we acquire Mandarin to a high level without deliberately studying grammar rules? In my latest podcast I invited two Mandarin learning experts Professor Karen Chung & Daniel Nalesnik of Hack Chinese to take a critical look at Stephen Krashen’s ideas: Does Stephen Krashen’s Input Hypothesis & Comprehensible Input Work for Learning Mandarin Chinese? – I'm Learning Mandarin
6 Likes
As always, thanks to Mischa for providing such stimulating content.
After reading some of Krashen’s papers, I have to say, I found his hypotheses quite underwhelming and that is both, from an intellectual perspective - even I, non-linguist, can’t help but notice that many key terms lack proper definitions (comprehensible input, i+1…); and also from a language learner’s perspective. Of course there is a good chance I’m missing something. Searching around for other, more qualified opinions, I stumbled upon this reddit thread:
https://www.reddit.com/r/LearnJapanese/comments/gtwzpl/immersion_isnt_quite_all_you_need_heres_why/
Of especial interest to me is this paper, which is linked in above thread: http://jehdnet.com/journals/jehd/Vol_4_No_4_December_2015/16.pdf (1)
It provides an evaluation of Krashen’s thought from an academic point of view, I found it quite elucidating.
I don’t have much time today, but I’d like to pick on one particular idea of Krashen’s. That is the concept of comprehensible input which is a cornerstone of his Input Hypothesis (2) and one that is probably cited most often in language forums, conversely it is a concept I struggle with. According to Krashen (3) input can be made comprehensible by simplifying it, examples he cites are caretaker speech, “motherse” or “baby talk”. As an adult language learner, I have to concede that I am not at all interested in this kind of input. To the contrary, I actually signed up for LingQ to study political commentary, history and similarly intellectual topics, topics that revolve around abstract concepts. The material is expressly intended for native speakers, no concessions are made to accommodate my inadequate level of Chinese.
It can rightly be said that this content was way above my level, and did not at all conform to the i+1 hypothesis; but I claim, that I used LingQ to help me make sense of it, that is, I availed myself of dictionaries and translations to make it “comprehensible”. I use comprehensible here in the dictionary sense of “able to be understood, intelligible”(4).
Krashen’s other examples for comprehensible input like “foreigner talk” or “teacher talk” are equally unattractive to me. Actually, I feel that I stayed with content intended for learners for too long and avoided contact with the “real” language.
Don’t get me wrong I’m not against Krashen, he comes off as likable and his messages are relatable - does anybody really like grammar drills, language classrooms, embarrassing yourself while speaking a foreign tongue etc.?
But one could play devil’s advocate, and say he over-emphasizes the comfort aspect, that he encourages learners to avoid practice and feedback (see his recent conversation with Steve). I think Mischa and his guests have a point in stressing the importance of deliberate practice and feedback, however uncomfortable that might be.
(1) Liu, D. (2015) A Critical Review of Krashen’s Input Hypothesis: Three Major Arguments
(2) Krashen, S. (1985). The Input Hypothesis: Issues and Implications
(3) Krashen, S. (1982) Principles and Practice in Second Language Acquisition (p.22)
(4) New Oxford American Dictionary
6 Likes
I plan to listen to the podcast when I have time, but real quick… I’ve only listened to about 300 hours of Chinese, so I still have a long way to go, but tones don’t seem any more of a problem than gender does in French.
I think there might be an interplay going on. Yes comprehensible input is key, but conscious knowledge of what you’re listening to might also be helpful. If I had no knowledge of tones at all, it might have taken quite some time to even start to notice what’s going on there. From day one I focused on hearing the tones with very comprehensible input, but I had an idea of what I was listening for. Listening to tone pairs I think can also be helpful, which is another form of input, hearing the same tone pattern but in different contexts. I think it’s the same with grammar, just giving me little tid bits of grammar here and there helps make the content I’m consuming more comprehensible.
So that’s my theory, a little bit of conscious knowledge about the language goes a long way toward making the input more comprehensible, and thus more effective.
To achieve good pronunciation you do have to do a certain amount of listen and then mimic, how else would you develop the proper muscle movements of your vocal tract without doing some practice? (your podcast about letting yourself “hear” your echoic memory repeat the word/sentence first before you say it was very interesting, I’ve been trying that with my current babble/mimic practice).
3 Likes
In reading Thai, knowing the tone rules specifically is absolutely necessary.
For speaking, if you at least know the tones and how to pronounce them, you could get by with comprehensible input alone, as many learners do.
Good points, bamboozled (as always 
)!
I think we can remove the main distinction “input / output” in
this context altogether because the technical data transfer model
isn’t helpful for explaining human communication processes
(socio-emergent communication models, for instance, are a much
more “advanced” alternative that is available nowadays).
Then we can focus on the more “interesting” questions when it comes
to processes of practical skills acquisition (beyond language learning).
For example:
- How important is “incidental” learning (in SLA: by reading and listening)?
- And how important can this kind of “incidental” learning be if “focused attention” is the bottleneck?
- And how important is “intentional” learning in this context?
- And what kind of “intentional” learning in which domains (in SLA: pronunciation, grammar, etc.) is necessary to have more success? Possible Keywords: spaced repetition, active (i.e., minimally discomfortable) recall, deliberate practice with feedback loops, etc.
- How important is the role of “frustration / discomfort” tolerance in learning processes?
- How did the idea come up that the ideal learning process should be “effortless” / “comfortable”?
etc. pp.
“I think Mischa and his guests have a point in stressing the importance of deliberate practice and feedback,”
Absolutely. There is enough research to know that this style of intentional learning works very well in all kinds of processes of practical skills acquisition.
And it’s probably not about contrasting these two learning styles in the sense of: “(purely) incidental” vs. “(purely) intentional” learning.
At least from a practical point of view, it’s more about the blending of these learning styles (= a both / and - relationship).
In SLA, it boils down to the question: when should we start with “intentional” learning? And this leads, for example, to the old controversy:
- Is it better to start speaking earlier or later?
- And how late is “later”? Or, vice versa: how “early” is earlier?
3 Likes
Thank you for your comment.
“…tones don’t seem any more of a problem than gender does in French.”
I think in one sense tones pose the same kind of problem as French genders. In effect our brains tend to ignore an aspect of the language which doesn’t exist in our L1 so we may need to turn to deliberate memorisation to get it right.
However, in the same sense, tones present a tougher challenge. Firstly, getting gender wrong will rarely cause you to be misunderstood. In this sense getting to a reasonably high level of accuracy (say 80%+) is not strictly necessary if the goal is just to be understood. Not so with Mandarin tones.
Assuming the goal in both cases is to reach 80% accuracy this can be achieved much more easily with genders. In French there are two genders so learners start at around 50% from day one. In Mandarin there are 20 tone pairs and the chances of guessing one right by chance are around 5%.
Finally, in French only nouns are gendered so you may only need to remember the most common 100 or so nouns to be right 80%+ of the time when you speak. In Chinese not only does every single word have tones that need to be memorised, every single syllable of every word must be pronounced with the correct tone.
5 Likes
BTW, I think it’s a misunderstanding to reduce “intentional (explicit)” learning
to just learning explicit grammar and other “rules”.
Instead, it should rather be used to learn collocations with sentences (and other context info) coming from authentic native speaker communications - and then this kind of intentional learning based on deliberate practice using SRS can be both very effective and efficient!
And that’s the reason why Will Hart’s approach for becoming fluent in Mandarin, for instance, worked.
2 Likes
I should have added that I welcome corrections, I really don’t know anything about linguistics and just read some old papers by Krashen, supplemented by an academic paper that happens to be rather critical of Krashen.
Most of Krashen’s works are freely available online by the way: http://www.sdkrashen.com/
Regarding the psychological component, there again we see a contrast. On the one hand we have Krashen who emphasizes how important it is for students to learn in low anxiety and pressure-free situations and that they not be forced to output before they are ready. On the other hand we have people like Benny Lewis who seem to seek high-stakes situations, and also Mischa’s guest who underline the importance of such situations.
Indeed, language learners probably best avoid the extremes and maybe take some inspiration from the likes of Aristotle (the Golden Mean), Confucius (中庸 - Doctrine of the Mean), and others: Golden mean (philosophy) - Wikipedia
3 Likes
Thanks everyone for your thoughtful comments. I was slightly apprehensive about posting this podcast because in my experience the moment you dare raise any questions about Krashen, CI or the Input Hypothesis - let alone criticise any elements of these ideas - a number of his more extreme followers become very angry.
There seems to be a community of learners who treat Krashenism as a religion and the man himself as a god. They sincerely believe that Krashen’s hypothesese are laws of physics and get extremely upset and angry if you raise questions about them. This rant which someone left at the bottom of my blog is a perfect example: Does Stephen Krashen’s Input Hypothesis & Comprehensible Input Work for Learning Mandarin Chinese? – I'm Learning Mandarin
If any of you would like to politely to reply to the comment linked to above on my blog I invite you to do so and would be very interested in reading your thoughts.
2 Likes
“I really don’t know anything about linguistics and just read some old papers by Krashen”
No problem, Florian.
I think this forum is more about “brainstorming”, i.e. coming up with or spreading / amplifying interesting ideas regarding SLA than anything else. It can’t be about language (learning) -related “science” because most of us aren’t linguists, SLA researchers, cognitive psychologists, etc.
“best avoid the extremes”
I agree. For language learners, it’s often about finding the right mix that works for them.
In other words, there’s “no panacea” in this context, and the mix can change for various reasons (e.g., time / budget constraints, possible deadlines, language level, etc.).
So if someone claims that there’s only “one” way to learn a language,
- they are either clueless/inexperienced
- or they are saying that for marketing purposes
In some cases they may even be both at the same time
3 Likes
“On the one hand we have Krashen who emphasizes how important it is for students to learn in low anxiety and pressure-free situations and that they not be forced to output before they are ready. On the other hand we have people like Benny Lewis who seem to seek high-stakes situations, and also Mischa’s guest who underline the importance of such situations.”
Yes. The “psychological make-up” is definitely an important aspect in this context.
I’ve witnessed this in my own family:
My sister and older nephew tend to collapse in high-pressure situations. I, on the other hand, thrive in such situations (exams, job interviews, etc.). So what they perceive as “stressful” is often just “exciting” for me.
I’ve observed similar tendencies regarding the handling and framing of stress in many of my students.
5 Likes
“There seems to be a community of learners who treat Krashenism as a religion and the man himself as a god”
Funny, I’ve been learning L2s for decades, teaching languages for more than 10k hours (both in companies and the private sector), and have been highly active in this forum for ca. 3 years, I’ve never met any “fanatical Krashenists”
And if there were, I wouldn’t care one iota.
It’s been a common practice, esp. in the West, to criticize anything that tries to avoid criticisms (common sense “truths”, religious, political, etc. “beliefs”, ideological structures, whatever).
That is, the whole public discourse in the West has thrived on criticism for about 300 years from the early Enlightenment thinkers through countless critics of ideology and latency to today’s deconstructionists. Ultimately, it boils down to the fact that there is only this critical dynamic (in communication processes) left.
And that was already nothing new for the brilliant Karl Marx in the 19th century:
“All that is solid melts into air, all that is holy is profaned.”
On an international political level, only authoritarian regimes like the CCP don’t seem to understand how “anachronistic” they are
PS -
I just scanned Katie’s comment on your blog, she seems to be “passionate”, but I wouldn’t use the word “fanatical” here
“For three years I actually lost a lot Chinese I had previously learned because the teachers were so against anything but grammar explanation and never provided any input outside of the dialogues from the one textbook.” (Katie from your blog)
This is definitely one of the main aspects why Krashen’s CI approach has become popular in the language learning / teaching community since the 1970s.
And although I’m “pro” blending incidental and intentional learning styles in SLA, I’m against the traditional
- a-communicative
- grammar (rule)-
- translation-
- textbook-
oriented style of language learning / teaching based on an obsession with perfection/accuracy, which still seems to be prevalent in Asian countries.
When it comes to intentional learning using deliberate practice, we can do better than that nowadays. Possible “ingredients” are then:
- being highly communicative
- grammar light, esp. at the beginning
- using authentic language material instead of artificial examples in textbooks / graded readers
- being process- and not result-oriented
- being error-friendly
- relying on habits and not on empty talk about motivation, etc.
- independent and not simply instructor-led learning
- emphasizing one’s own special interests
- freedom of choice instead of pre-chewed compulsory topics
etc. pp.
4 Likes
Interesting that you mention Marx. This may seem like a strange analogy but I think Krashen occupies a similar space in language education to that which Marx occupies in modern political/ economics discourse. Both are serious thinkers in their respective fields whose work has influenced mainstream thinking. Both have legions of religious accolytes who hang on their every word and get very angry when others question or disagree.
Katie’s comment reminds me of arguing with religious Marxists when I was more into political debates. They enter into the debate already at the end of their patience. They believe that they alone hold the truth and have nothing but scorn for people who they assume are not as well read in the holy texts.
Instead of presenting arguments and exchanging ideas calmly and coherently they foam at the mouth, repeatedly accusing their opponents of total ignorance without explaining what exactly the problem is. They accuse others of lacking evidence for their claims, then proceed to make extreme categorical statements providing no evidence to support them.
2 Likes
Thanks for your kind and inisghtful comment and for the links.
“with religious Marxists”
Do they still exist? I’ve been in hundreds and hundreds of philosophical, sociological and political discussions in my life (because I specialized in the humanities and social sciences first before switching to computer science later). but I’d say since the late 1980s “marxism” is largely irrelevant in public discourse (at least here on the continent).
In the last 30-40 years post-modernist / post-structuralist discourses inspired by the writings of Foucault, Derrida, Lyotard, Deleuze / Guattari, etc. or equivalent systems theoretical approaches (Luhmann,Dirk Baecker, etc.) have often dominated the intellectual scene.
Compared to these “distinction-based” approaches, (vulgar) marxists have absolutely nothing to offer - and nobody takes them serious any more.
Nowadays, on the other hand, the intellectual scene seems to be quite “diffuse” while technological developments (the Internet in general, big data, artificial intelligence, etc.) tend to dominate everything more and more.
In short, modern technology monitors us (esp. smartphones are the “perfect” surveillance tool - and the best thing is: people carry it around “completely voluntarily” all the time. Not even the Gestapo or NKWD / KGB could dream of something like that!), the algorithms know more about us than we do, etc. -
compared to that, distinction-based approaches à la Derrida, Luhmann, etc. seem to be rather “helpless” because advanced technology simply creates “facts” the scope of which we often don’t understand…
Be that as it may, Marxen’s “Das Kapital” has still some interesting insights to offer when it comes to economic processes in the modern era, but apart from that he’ s mainly interesting for understanding the 19th century. That’s all - and that’s probably not enough for becoming a Marxist zealot nowadays - maybe the non-theory of the “flat Earth” is more attractive to them, who knows?
Krashen never had a similar influence in the language (learning / teaching) community. And while Marx was brilliant, but clearly “authoritarian”, Krashen is rather brilliant, but highly self-deprecating, which is a characteristic of most intellectuals / scientists nowadays.
Moreover, “authoritarian” attitudes have completely lost their charme since the student revolution of the late 1960s in the West. And people who still cling to them are often regarded as “uneducated and naive fools” (e.g., populists like Trump, Bolsonaro, etc.).
But the dangerous thing is when such fools come to political power so that they
can do a lot of harm!
Krashen doesn’t belong in this category. I’m pretty sure of that!
And besides, SLA is not that important overall
3 Likes
I don’t want to dwell on this analogy for too long because it is not reallt relevant to the central discussion of this thread.
However, I’d like to clarify wasn’t comparing Marx and Krashen as individuals. I was comparing their overall influence in their field and the behaviour of their most extreme supporters. There seem to me to be similarities there.
I am happy for you that you’ve managed to avoid dogmatic Marxists and Krashenites in the language and politics forums you visit. But to answer your question yes they do exist and there are many of them.
1 Like
“There seem to me to be similarities there.”
Maybe, maybe not.
Probably, we need more empirical research on the SLA community here
Anyway, I wonder if as “humans” we face a more serious problem related to the Internet: sometimes people tell me that we don’t need experts any more because we have “all the knowledge we need at our fingertips”.
The problem is: we already have to know a lot (i.e. theories, methods, etc.) to really be able to understand the nuances of the information that is presented to us on the Internet. That’s usually impossible for laymen.
Laymen tend to read a Wikipedia article here, a blog post there, and may often be happy when their biases are confirmed. The funny paradox is then: on the one hand, they claim that “experts are irrelevant” nowadays, while on the other hand, they overestimate their own instant expert perspective.
In a certain way, that’s the Dunning-Kruger bias on digital steroids
If that’s the new social norm, then we can probably say goodbye to “public discourse” because it’s one thing not to be able to convince others (which is usually the case), but it’s a completely different thing if we aren’t even able to “agree to disagree”.
However, if “orange is the new black”, maybe “instant expert confirmation bubbles” are the new spheres of public non-discourse
1 Like
You (michilini) will likely jump on your “but chinese is different” shtick here but I think jahufford is right to some degree. A similar thing happens in Russian as you are both describing. In the Russian situation, over a period of time you build up the ability to comprehend spoken or written Russian even without formal grammar training. I believe it is some form of compression similar to color blindness. A color blind person would be able to get most of the information from a picture but might not be able to tell that a certain part is red or green. They would know from experience that it is red OR green but they would not know which it was.
In reverse, if this color blind person painted with two tubes of paint, red or green and a variety of other colors, then a non-color blind person would see immediately that the color blind person had painted in red or green whereas the color blind person would not know which of the two colors he had used.
I believe this is a similar situation in Russian with incorrect cases as it would be in Chinese with incorrect tones. In other words a person who doesn’t really know cases hears something spoken she could say that it could be this or it could be the other. Likewise with the tones they could say it could be horse, mother or whatever.
Conversely they would be screwed.
TLDR: A person with poor knowledge of tones/cases can have a “it must be one or the other or the last” type of understanding of what they are hearing and try to guess from context. Whereas a native speaker hearing a person speaking a correct sounding (but wrong) case or tone will misunderstand what the non-native is trying to say.
i.e. it’s not necessary to know the tones or cases in order to be able to understand but it is necessary to be able to reproduce the tones or cases correctly in order to be understood.
5 Likes
I think while you’re right, you’re also focused on the final end of the language learning journey - “crossing the finish line”.
My personal experience is that Krashen type methods (my own method is a version of these - but input only) will get you to an intermediate level of comprehension rapidly and efficiently.
Where it gets interesting is how do you “cross the finish line” and get to an advanced level which includes accurate (though not necessarily “native”) output.
In this area, my own method with no modifications won’t do it. I believe output with feedback is key to “cross the finish line”.
I think lingQ will also not do it in and of itself. If you look at what Steve K does: he doesn’t only use lingQ (which includes listening and reading) but he also starts to talk to native speakers.
1 Like
“it’s not necessary to know the tones or cases in order to be able to understand but it is necessary to be able to reproduce the tones or cases correctly in order to be understood.”
Makes sense to me (but I’ve never learned a tonal language like Mandarin, Cantonese, Vietnamese, etc.) because “guessing from context” definitely helps.
However, I’m not sure how many nuances of understanding are lost this way.
But “understanding the nuances” can be very important, depending on the situation
1 Like