Does speaking actually improve the language skill that much?

[Mactuary]

Dear fellow forumers,

Does speaking actually improve the language skill that much? And I mean speaking in the strict sense of output.

I've been thinking about an issue regarding this and I'm sure there have been studies done (I would appreciate it if you could send me the link to some.).

Let me explain first my situation and I'm sure there are lots of people who are in my boat.

On top of being rather introverted I've discovered (a long time ago) that my people skills are lower than my Arabic skills; and I've only listened to a handful of Eastern Arabic tapes, if you know what I mean. I don't generally enjoy speaking to people even in my native language, let alone with local Chinese people considering the added burden of a lack of language skills.

So here's my argument for the against:

1. Speaking only draws upon existing knowledge, so it will not increase it no matter how much I do it.

2. Speaking may have the negative effect of reinforcing mistakes, since mistakes are inevitable when learning any language.

Here's the rebuttal:

1. While speaking, strictly speaking (haha), doesn't help, it's the interlocution that help develop language skills.

2. On the other hand speaking helps to uncover unresolved grammar, colocation problems etc.

Here's the counter rebuttal:

1 & 2. Through my experience of speaking with natives, extremely few people are diligent enough to point out mistakes, or maybe they think it's impolite to continually correct my mistakes. I can generally make myself understood, but as to correct grammar or elegant sentences...So this nullifies the advantages of speaking. I almost always have to ask them again to confirm "Is it correct to say...?". This is ever so tedious.

3. After receiving the correction, it is strictly the listening that helps my language skills.

So it seems like I'm missing something here, since everyone I meet tells me to speak more to native speakers or whatever language one is learning. But considering the above, particularly for such a personality, is there not a better way?

Also people tell me to not be afraid to make mistakes, this would only work under the assumption that the listener will correct these mistakes immediately, no? Otherwise it'll be just like my argument number 2.

I wonder if it's possible to just listen to podcasts, news, movies and tv series in the target language? This way I would get correct information on how to say things in certain situations, the obvious drawback is that I can't control the situation. But considering the alternative of quiet and awkward situations with my "new friends"...

When I do actually go out with the locals I usually prefer if it weren't a one on one situation. In this case I would usually end up just listening to the locals talk and not say much myself.

As for online chatting, I have found that it is more likely for my mistakes to be corrected. I guess this may be due to the reduced real-timeliness inherent in online chatting.

What do you think?

Do you still think I should still force myself to be more outgoing and hang out with locals whenever the situations presents itself?

I always tend to write super long posts, so thanks to those who read from start to finish, particularly if you respond commensurately.

[li3wei1]

Does watching videos of Tiger Woods make you a better golfer?

Several people on this forum recommend listening to tapes of native speakers, and repeating what they say, recording that, and doing it over and over again until what you're saying is exactly (sound, stress, intonation, timing) what they're saying. So you're speaking, but you're not in a social situation, and in a sense you are being corrected (by yourself). The actual physical movement of your lips and tongue and lungs is important, just as the movement of your body when learning how to play golf. Do this enough times and native-sounding phrases will roll off your tongue with ease, and the more of these you can produce the better you will sound.

In real-life conversations, even if the other person isn't correcting you, you will hear them repeat words and phrases that you've just said or are trying to say, and doing it correctly. Also, if they don't understand you, you'll have to try again until you say something good enough for them to understand. Or, you'll hit a word you don't know how to say, ask them how to say it, they tell you, you use it in a sentence. Word learned, at least for a little while.

[renzhe]

Controversial topic.

Speaking is certainly important, you need mental practice, you need to produce words at will, you need to apply grammar rules in real time, and you need lots of muscle training to perfect pronunciation.

Personally, though, I think that input is far more important, especially in early stages. You are exposed to correct grammar, and only correct grammar. You're exposed to correct pronunciation and only correct pronunciation. You need to hear before you can produce and it often takes people months before they can even HEAR things correctly. By all means, you should repeat, get feedback, record yourself, try to compose sentences in your head, talk to yourself, etc, but it's not always essential to go to a bar and start up a conversation with a person.

IMHO, I agree with people who argue that correct, comprehensible input should fill the majority of your time. The faster you get to the point where you can understand most language, the faster you'll be able to say most things, since you must be able to understand something before you can say it yourself. You will start speaking fluently a bit later, but you will have a better accent and better grammar, at least in my experience. People who start speaking with poor grammar and pronunciation immediately usually end up speaking with poor grammar and pronunciation forever. These are not absolutes, of course, since it all depends on many factors and the person.

Personally, I've tried different things with different languages, and ended up shifting from the "speak all the time from the start" approach I used with German towards "understand first, speak later" school, which seemed to work really well for me with Portuguese. The first one got me speaking faster, the second one got me speaking with a better accent.

Does watching videos of Tiger Woods make you a better golfer?

Yes, absolutely, without any doubt.

Assuming, of course, that you understand the basics of golf already.

https://en.wikipedia...i/Mirror_neuron

EDIT: I'd just like to point out that I'm NOT advocating "not speaking". Speaking is important, active production is important, and they must be a part of a complete programme. I'm just commenting on the relative importance and required amount of input/output especially in the early stages of learning.

[Silent]

I guess the answer to your question ain't a simple yes or no. Everyone is different and what works great for one might not work at all for someone else. Personally I'm quite introvert too, but I see real benefits to speaking (as in a dialog). I think the most important factors are that you use the language actively and thus bump into problems and that feedback is immediate.

Active use can also be done by writing or giving a speach, but for those your will receive feedback with a huge delay making the feedback far less effective. You may, as suggested, tape yourself and compare it to native audio. I think the feedback delay would still be quite big compared to talking to someone but it also has as the disadvantage that you have to find the errors yourself. A native will listen quite differently and notice other things then yourself. E.g. many beginners have a hard time to hear tones correctly, so it's unlikely that selfcorrection will result in good tones.

Another huge benefit for me is that social interaction provides more anchors to remember things you learn. You're far more likely to remember what you learn in a funny chat over a beer or on a long dialog to get a simple message across then what you learn by reading a textbook or watching a soap.

[陳德聰]

EDIT: I realized after I finished this post that it's really long. If you want my short opinion without the facts, scroll down.

====

The first freely available "academic" discussion of this topic that came up on google is here.

Unfortunately, I say "academic" because it says 論文 at the top and not because it is actually particularly academic. The article is almost wholly anecdotal, but has references to some more credible academic research by Swain (1985, 1995, 1996) and Swain & Lapkin (1995).

To paraphrase:

The author supports a theory contrasting Krashen's Comprehensible Input called Comprehensible Output, which would supposedly fill the gap in Krashen's original theory that one can learn language solely from input. In Comprehensible Output, learners receive added benefit from speaking in three key ways: 1) noticing the gap between their intention for communication an their actual ability; 2) hypothesis forming and testing with new vocabulary or structures; and 3) engaging in meta-linguistic discussion. The author also notes that it is possible to engage in some of these behaviours through internal verbalization (essentially producing language in your head instead of out loud), though it does not necessarily lead to the same level of achievement.

1) Gap noticing: when you speak you take what you intend to convey, e.g. "Would you mind if I opened the window?" and try to turn it into comprehensible output using whatever knowledge you already possess. If there is a gap in your knowledge, you are more likely to notice it if you find yourself unable to remember how to say "窗户", unable to remember whether you need a "我" or not, etc.

2) Hypothesis forming & testing: when you have a feeling you might know how to say something even though you've never heard it before, and you take your knowledge and try to apply it to the phrase in conversation. Perhaps you've never heard anyone use the word "丰富", but you've read it, and you're pretty sure you are allowed to say "这本书的内容很丰富" even though you've never heard or read that sentence before. If you say it to a Chinese person and they understand --regardless of whether or not it is grammatical, since a Chinese person will probably never straight up just correct you anyway-- what you were trying to say and the conversation progresses without problem, then you have successfully tested your hypothesis*.

*I personally do not buy the author's argument here, because I imagine that even if you said something like "请把盐瓶通给我", you would still get the salt and then what, you would have solidified in your internal lexicon that 通 can be used here? I imagine the benefit of this tactic is largely dependent on the actual context of discourse.

3) Meta-linguistic discussion: when you talk about language. Much like what we do here, talking about the target language has been shown to benefit growth in the target language.

====

A quick search of my university's database brings up a pretty comprehensive review of speech models and past studies done on Swain's original Output Hypothesis: "Why Output?: Theoretical Claims and Empirical Evidence in SLA" by Sun Hee Kwon (2008). Besides being quite recent, this paper discusses 3 models of cognitive processing relevant to language and several past Input vs. Output studies very succinctly (go Sun Hee Kwon!). It's definitely worth reading the whole thing, but I'll paraphrase for those who aren't partial to 25 page papers*.

*even though paraphrasing a paraphrase is bound to leave out important info.

It's my vacation so I am going to write these summaries in super informal speech--

Levelt's Speech Production Model:

Levelt describes speech production as a feedback loop with several processes/components. The loop starts with a conceptualizer which generates the pre-verbal message, runs it through the formulator which takes lemmas and forms from the lexicon (in the form of syntax and phonology) and spits out the phonetic plan. The articulator takes the plan and creates the overt speech as a phonetic string, which is then --if you're not deaf-- picked up by the audition process, which throws the phonetic string (what you hear) into the speech comprehension system which takes that phonetic plan from before and compares it to the phonetic string, also dipping into the lexicon for those lovely lemmas and forms to make sure what came out of your mouth is what you wanted to come out of your mouth. The speech comprehension system parses the speech, and then your situation knowledge/discourse model takes information from the situation and informs your conceptualizer if there are any changes it needs to make.

McLaughlin's Information Processing (IP) Model:

McLaughlin says there are two types of processing, and a restructuring function that helps things go from one type to the other. The IP way is to have either controlled processing or automatic processing. Controlled would be when you're a beginner and you are very carefully taking your set of grammar rules and putting sentences together like building blocks in your mind, and automatic would be when you're just spitting out sentences and saying what you want to say. Restructuring happens at some point when your brain takes controlled processes that have happened a lot and turns them into automatic processes. "Practise makes perfect".

Anderson's Skilly Learning Theory:

Anderson came up with Adaptive Control of Thought (ACT) to work out how humans generally learn new skills, not just language. It basically is the same as McLaughlin's IP model, except it calls the two stages declarative knowledge and procedural knowledge, and focuses a lot more on how exactly your brain does the restructuring.

Previous Studies have concluded:

The results are contradictory.

Some studies said that output-based instruction (where people focus on speaking rather than reading/passively listening) sucks and input-based instruction rules. Only problem is that the main study that claimed this has some rather big methodology problems. A replication of the study with one added measurement (written narrative test) found that there is no measurable difference between the two approaches. The authors of the two studies are not very pleased with each other.

Some other studies have found that output is better than input, but again the methodology... Come on guys, get it together.

Some studies found that output is better for output-based tasks, and input is better for input-based tasks. That is, output-based instruction led to better performance on production tests, and input-based instruction led to better performance on comprehension tests.

How can all of these studies be so contradictory? The problem lies in the methodology and testing.

====

You might have noticed I lost steam somewhere near the end there. All of the references are listed at the bottom of the articles, so y'all are free to do the reference checking yourselves, but I'd like to offer my opinion based on all of that reading:

Empirical evidence shows that both input and output affect performance in second language acquisition, so the whole argument over which is better is silly. Focusing on only one over the other is likely a poor way of achieving optimal results. Thus, finding the proper balance between the two is important to me. Now, the actual quantifiable "balance" may be a bit elusive, but I imagine trying to do as much of both as possible is probably a sure-fire way to make sure you are getting enough input and output to achieve optimum language learning.

So in short, OP, yes. Speaking actually improves your language skill that (see above) much.

[jkhsu]

I like renzhe's approach post #3 and I follow a similar learning strategy. Speaking early is OK as long as you continue to learn from proper written (reading) / audio (TV, radio) input with correct grammar and pronunciation. However, there are some (on a 3 month program) who believe you can just "talk your way to learn" from the start. You might be able to verbally communicate earlier doing this but you'll inadvertently pick up / create mistakes which may affect your learning in the long term.

[li3wei1]

You can study all you like the relative efficiency of various methods, but at the end of the day, if the human being employing them doesn't like a particular method, or isn't as comfortable with it, he/she just won't do it as much. So if the OP isn't into talking in social situations, he should do something else, and if I'm into reading manga so much that I can do it for six hours a day, then manga it is. A balance between reading, writing, listening and speaking would be ideal, but for some people, and in certain situations, it just isn't going to happen.

[jbradfor]

Speaking only draws upon existing knowledge, so it will not increase it no matter how much I do it.

I disagree with this.

Mostly, "knowing" has different levels. At first, you "know" a word by being able to recall it with difficultly, and pausing in your speech. Then you "know" it by being able to use it, but it still takes mental effort. Then you "know" it by being able to use the word without any conscious effort. The more you speak, the more you "know".

Also, even if you don't "know" a word, when you try to "speak around it", the other person might use that word in reply, and, now you "know" that word. Or, even if they don't, you might remember that gap (especially if it were particularly embarrassing -- I find the more I'm embarrassed by an error, the more I remember it and not make that error again....), and actively learn it later.

[陳德聰]

@li3wei1

I was under the impression the OP was asking whether or not speaking (output) actually has any relation to language development. I don't advocate for any one particular method of language learning, but it is probably important to know that when you don't speak you are sacrificing a part of your development.

EDIT: What's more, if you're going to refute something you should probably read the whole text so that you know what you're refuting. This is not me trying to appeal to verbosity, but the OP asks about role of the output in contrast to role of the input. The output plays a role, and this does not mean going out and finding people to talk to, but it means recognizing that there is something important about those interactions that would elicit further language development.

I didn't get into them all in my post, but if you read any of the bazillion studies done stemming from Swain's original hypothesis you can see the actual importance of making mistakes in discourse situations. Any method that can simulate the kind of situation where a learner would be forced to modify their output would be alright in my book, regardless of whether it involves reading manga or talking to strangers.

[imron]

Several people on this forum recommend listening to tapes of native speakers, and repeating what they say, recording that, and doing it over and over again until what you're saying is exactly

One of those people would be me (although I wouldn't recommend tapes so much as mp3 recordings).

Not only that, but doing so seems to assuage all of your concerns.

1) You will come across many new words doing this, so it doesn't just draw on existing knowledge, plus as jbradford mentioned, you will strengthen your familiarity and comfort with words you already know

2) You're copying native speakers using material that doesn't contain mistakes (at least not if you've chosen the material well, but even if not, they'll be 'mistakes' acceptable by native speakers too), so no need to keep asking people if this is correct. You can also go back over the recording yourself, listening for mistakes.

3) It helps with both interlocution, colocation and helps resolve unfamiliar grammar and sentence patterns.

4) You can quite happily lock yourself away in your room with nothing but your computer, and practice your introverted heart out

[querido]

I hope you listen to imron.

But also... A year or so ago I "forced myself" to make friends on QQ and I found a live tutor too. It has been a productive year. It has been one of the best years of my life. (So, my answer to your question is yes; for me it has been essential.)

[li3wei1]

Sorry, 陈德聪, I wasn't refuting anything, and I did drift a bit off-topic. I was following a chain of thought that started up from another thread (something about the value of looking up words in dictionaries vs. electronic pop-up style apps, I think), about the possibility of researching the relative efficiencies of different learning techniques. Such research may find, for instance, that using Chinese definitions on your flashcards is x% more effective (how you'd measure it I don't know, because it's not just the target vocab, it's all the vocab in the definitions as well), but if the student, outside the laboratory, gets a headache after 10 minutes of Chinese-definition flashcards but can do 30 minutes of English-definition ones, then the efficiency is wasted. I suppose I'm saying that other factors have to be considered, such as fun, practicality, affordability, etc. Listening to a podcast may be less valuable than, say, writing an essay, but you can listen to a podcast while driving, or washing the dishes. Some things you can do when you're tired or in a distracting environment, some need a quiet room and a cup of coffee.

This is not a refutation of anything that's gone before, and it may not be an answer to the OP's question, just a warning that no advice about what's right or best should be taken dogmatically.

[Mactuary]

I kind of already half expected the responses before I made that post, so that's good in a way. I will continue on my quest to find suitable speaking partners or dare I say "friends".

Thanks for all your contributions.

[imron]

You may find that suitable speaking partners and friends have different sets of requirements that may or may not overlap. You'll also find that speaking practice in isolation will provide flow on benefits when speaking to real people :-)

[wakdjunkaga]

Hi, just wanted to add my two cents here as a non native speaker of English who has spent about 10 months speaking the language for many hours every day while getting mostly Chinese input. My experience is my English speaking skills haven't improved at all, I still make quite a lot of silly mistakes, especially if I'm tired (I can almost always tell when I made a mistake and how to correct it, but I can't keep up with the speed involved in real conversations where you are actually thinking about the content and not the language).

So I would say yeah, speaking helps, but it will only get you so far. (although now that I think about maybe it also has to do with learning English not really being a major goal for me right now)

I would also add that, to me, one of the most important things speaking does for you is getting you to that state of internal verbalization that 陳德聰 pointed at: having friends to whom you speak in your target language will usually get you rehearsing conversations in that language in your head. Imo this is akin to, for those who play an instrument, involuntarily rehearsing a song even when you don't have the instrument with you; when you are in that state where you just can't help doing that, even if you can't play for a week, when you go back to the instrument the song will sound much better than before. I don't think it's really that this rehearsing away from the instrument that you catch yourself doing is particularly effective, to me it's maybe just a reflection that you have achieved a state where your brain, mostly on an unconscious level, is really working to get you to play that song. Same goes for speaking a language.

All these are just personal impressions and are to be taken with a big fat pinch of salt

[giraffe]

I'm an introvert too and not too likely to strike up a conversation in my native tongue with a random stranger or even the guy I see every day in the elevator. But I must say that I get a most delicious pleasure from speaking Chinese and being understood. It's just like I've learned that I possess this extraordinary ability, however imperfect, to talk to the birds or to creatures from another planet and, to our mutual surprise and delight, we understand each other. Climbing out of my shell once in a while is totally worth the effort.

[Olle Linge]

Speaking only draws upon existing knowledge, so it will not increase it no matter how much I do it.

I disagree with this.

Mostly, "knowing" has different levels. At first, you "know" a word by being able to recall it with difficultly, and pausing in your speech. Then you "know" it by being able to use it, but it still takes mental effort. Then you "know" it by being able to use the word without any conscious effort. The more you speak, the more you "know".

I think this is spot on. If you think you can learn without speaking very much, listen to any Chinese student who has learnt English this way and you know what it will sound like. Of course, you're not likely to learn many new words simply from speaking yourself and so vocabulary learning is not part of the goals of speaking practice.

If you don't speak much, you will never become fluent, not because you don't know the words, but because you can't recall them fast enough. Fluency is in this sense a skill that can and should be practised be practised (here is a fun word game I use with students which targets fluency). I also agree with others who have said that speaking is essential for noticing weaknesses (or for allowing others to find weaknesses) in our own language skill.

That said, I don't spend much time actively practising speaking now, but looking back on the past few years, I can say that the years that contained huge amounts of verbal communication with natives were also the years that I felt that I developed the most overall. This might be because I have a fairly large passive vocabulary, which can be turned into more a more active one simply by trying to use it, but I don't think this is unieque for me.

[laurenth]

Confusedlaowai recently published an interesting blog post briefly discussing the different "strands of language learning", including input- and output-based. He also mentions Swain (see post #5 by 陳德聰).

It's here