Examples of Input Only

General discussion about learning languages
User avatar
reineke
Black Belt - 3rd Dan
Posts: 3434
Joined: Wed Jan 06, 2016 7:34 pm
Languages: Fox (C4)
Language Log: https://forum.language-learners.org/vie ... =15&t=6979
x 6082
Contact:

Re: Examples of Input Only

Postby reineke » Thu Feb 28, 2019 4:32 pm

Literature review

"3.4 Nagy, Herman and Anderson (1985) Learning words from context

This classic study provides clear and convincing evidence that learners acquired new word knowledge as the result of a reading event.

This study is meticulous in its design. Many studies in the read-and-test tradition establish incidentally acquired gains in one of two ways, by comparing posttest scores on a word knowledge test either to pretest scores or to the performance of a control group. This study does both. The researchers entered prior knowledge of the targets (scores on the checklist pretest) into the regression analysis so that the unique contribution of the reading exposure to the targets became clear. Including in the analysis the performance of control groups who had not read one of the texts made it possible to show unequivocally that exposure to text was the sole source of word knowledge gains.

This study is often cited for its finding that learners acquire new word knowledge incrementally. The use of measures sensitive to partially correct knowledge of words made it possible to show that word growth occurs in stages. For instance, at the most basic level, a learner might recognize that a gendarme is a noun. Realizing that a gendarme is a person represents a somewhat more advanced stage, while being able to pick policeman out of a list of definitions that includes spy, bellboy, waiter and bodyguard represents a much fuller knowledge of gendarme. Making this important point required the labor-intensive procedure of writing multiple-choice questions, one at each of three levels of difficulty, for each of the 30 targets. Posttesting occurred immediately after participants read the experimental texts.

Neuman and Koskinen (1992) Captioned television as comprehensible input: Effects of incidental word learning from context for language minority students

Captioned TV watchers — the participants who had access to pictorial, sound and written input — consistently outperformed participants in the other conditions on all measures.

...it is troubling that the students who simply read the science texts, all 2700 words of them, learned little more than controls who read something else and much less than the participants who watched TV. Perhaps the reading-only participants realized that students in other classes had access to a more appealing TV version of these materials and refused to take interest in the project.

The one factor that did predict learning outcomes was the extent to which target words were pictured clearly in the televised materials. Thus the beneficial effects of picture support observed with L1 learners in the study by Elley (1989...

4.2 How do studies measure word knowledge?

Most of the reviewed studies of incidental vocabulary learning used multiple-choice tests to assess learners' knowledge of target words after the reading treatment has been completed. Even when other kinds of tests were used (e.g. Jenkins et al., 1984; Neuman & Koskinen, 1992), researchers tended to focus their discussion on the multiple-choice results. Thus the main type of vocabulary knowledge explored was the type multiple-choice tests measure: the ability to recognize (rather than produce) definitions.

Typically, the stem of a multiple-choice question presented a decontextualized word that occurred in the reading treatment and the testee was required to select one of four or five meaning options. In some L2 studies (e.g. Dupuy and Krashen 1993; Rott, 1999), learners matched an L2 word to an L1 translation equivalent. Others required testees to match an L2 target to a synonym or definition in the L2 (e.g. Brown, 1993; Ferris, 1988).

Tests of recognition knowledge seem an appropriate way to measure learning acquired through reading, which is itself at least partly a decoding process, but there are limitations to how much recognition knowledge such tests can assess. Note this example of a question from the test Nagy et al. (1987, p. 267) used:

headlamp

a) a tower with a bright light to warn and guide ships
b) a small electric light powered by batteries
c) a light on the front of a train, car or truck
d) a set of electric lights used to control traffic
e) don't know

It is clear that answering the question correctly requires a very complete understanding of headlamp. Knowing that it means some kind of light would not be helpful, nor would simply knowing that a headlamp is a noun. In an earlier 1985 experiment, Nagy and his colleagues created multiple-choice tests at three levels of difficulty for the very purpose of tapping such various levels of knowledge. But the 1987 study by Nagy et al. and many of the reviewed studies used a multiple-choice measure that required full definitional knowledge as in the example above. Such tests are probably not very sensitive to small changes in word knowledge, exactly the kind of changes that might be expected with encountering words once or twice in reading. Multiple-choice tests may be even less sensitive if learners must understand definitions written in the L2 in order to answer correctly...

...of 0.25 words per minute rate established by Nagy et al. (1985) for incidental acquisition through reading..."

Lextutor.ca


0 x

User avatar
reineke
Black Belt - 3rd Dan
Posts: 3434
Joined: Wed Jan 06, 2016 7:34 pm
Languages: Fox (C4)
Language Log: https://forum.language-learners.org/vie ... =15&t=6979
x 6082
Contact:

Re: Examples of Input Only

Postby reineke » Sat Mar 02, 2019 2:04 pm

Extensive reading & extensive listening

We have tons of data from ER but almost nothing for EL. This seems totally weird and unbalanced to me... Everything we know about ER's benefits (for vocab learning, for overall motivation, for language consolidation, to build fluency etc etc etc) need to be confirmed (or rejected) for EL. All the canonical work done in ER needs to be mirrored with EL. E.g.

What's the uptake of vocabulary from EL?
What known/unknown word ratio facilitates this?
How much listening is needed?
How does one's EL ability develop?
What prevents/ encourages comprehension in fluent listening?
How do students segment the sound stream in EL?"

Rob Waring

This researcher advocates a very high comprehensibility percentage for extensive listening activities.
2 x

User avatar
reineke
Black Belt - 3rd Dan
Posts: 3434
Joined: Wed Jan 06, 2016 7:34 pm
Languages: Fox (C4)
Language Log: https://forum.language-learners.org/vie ... =15&t=6979
x 6082
Contact:

Re: Examples of Input Only

Postby reineke » Sun Mar 03, 2019 7:36 pm



A lone learner may try to fit a whole side of beef into that pot.

Krashen uses the metaphor of the human digestive system. While it's true that the system is the same everywhere, not everyone has a good stomach and not everyone is a good cook.Unfortunately it is also true that that too many cooks spoil the pot.

Chinese drama with its talking heads, eunuchs and characters that spontaneously vomit blood is a chunky ingredient that may not be wholesome and digestible to everyone. Live TV is closer to having the whole village contribute to the pot and the two soldiers could represent...you and the common sense/tutor/personal guide. Your average course is a....box of grits. I am having fun with this.

We instinctively use metaphors when we talk about language learning because it's such a complex subject:

"Regardless of how comprehensible a message may be (or how comprehensible a teacher thinks it is), only a portion of input is generally converted to intake during the early stages of language acquisition. Intake is the portion of the input from which the learner is able to derive meaning, and at best, that may be a vague or ambiguous representation of what was actually communicated.

Thus far, the intake has strictly been focused on deriving meaning, progressing from vague representation to robust and more complete understanding. Comprehension is where language acquisition begins, but what happens after comprehension is achieved is so complex that there is not a scientist on the planet who can figure out exactly what happens during the process."

"Research has revealed that learners have the same built-in mechanisms for acquiring language and that language generally develops in predictable stages. These stages, although predictable, are not neat and tidy, and they sometimes resemble a commute from the suburbs to downtown."

"What makes analyzing language acquisition even more challenging is the evidence that suggests that the process is completely unconscious."

Bill VanPatten
Adding Madness to Method

https://forum.language-learners.org/vie ... 62#p134625

I hope this is helpful.
2 x

User avatar
smallwhite
Black Belt - 2nd Dan
Posts: 2016
Joined: Mon Jul 06, 2015 6:55 am
Location: Hong Kong
Languages: Native: Cantonese;
Good: English, French, Spanish, Italian;
Mediocre: Mandarin, German, Swedish, Dutch;
Some: Greek, Turkish.
.
x 3587
Contact:

Re: Examples of Input Only

Postby smallwhite » Mon Mar 04, 2019 1:30 pm

reineke wrote:
In light of all this, why is Patrick being painted as a near failure?

Re: The limits of comprehensible input?
smallwhite wrote:Patrickwilken is probably the most immersed member here - lives in Gemany, [here she lists Patrick's private business] consistent multiple hours of media for multiple years. English native learning Cat II language. If Patrickwilken is still not doing that great then it seems to me relying on comprehensive input to learn anything but sister languages is unrealistic and infeasible.


... and Patrick's success is being whittled down to failure.


I haven't been following this thread but Reineke's saying I painted Patrick as failure and near failure is just horrible. My description of Patrick "not doing that great" - far from "failure" - was a quick summary immediately following Patrick's own post in which he described himself as -

- "I still find reading books a bit of a slog..."
- "My sentence structure is not standard German..."
- "I think to a large extent I just need to do a ton more Comprehensible Input to..."

so still "a ton" of input away from his goal. Patrick's post here.

Reineke has become the reason I will read LLorg regularly no matter how busy I am.
0 x
Dialang or it didn't happen.

The real LLorg

User avatar
zenmonkey
Black Belt - 2nd Dan
Posts: 2016
Joined: Sun Jul 26, 2015 7:21 pm
Location: Germany and France
Languages: Spanish, English, French trilingual - actively studying German (B2/C1), Hebrew, Tibetan, Setswana.
Some knowledge of Italian, Portuguese, Ladino, Yiddish ...
Want to tackle Tzotzil, Nahuatl
Language Log: viewtopic.php?f=15&t=859
x 4970
Contact:

Re: Examples of Input Only

Postby zenmonkey » Mon Mar 04, 2019 1:50 pm

smallwhite wrote:
reineke wrote:
In light of all this, why is Patrick being painted as a near failure?

Re: The limits of comprehensible input?
smallwhite wrote:Patrickwilken is probably the most immersed member here - lives in Gemany, [here she lists Patrick's private business] consistent multiple hours of media for multiple years. English native learning Cat II language. If Patrickwilken is still not doing that great then it seems to me relying on comprehensive input to learn anything but sister languages is unrealistic and infeasible.


... and Patrick's success is being whittled down to failure.


I haven't been following this thread but Reineke's saying I painted Patrick as failure and near failure is just horrible. My description of Patrick "not doing that great" - far from "failure" - was a quick summary immediately following Patrick's own post in which he described himself as -

- "I still find reading books a bit of a slog..."
- "My sentence structure is not standard German..."
- "I think to a large extent I just need to do a ton more Comprehensible Input to..."

so still "a ton" of input away from his goal. Patrick's post here.

Reineke has become the reason I will read LLorg regularly no matter how busy I am.


But was he really "not doing that great"?

Patrick notes:
patrickwilken wrote:I think people are misreading what I said a bit. I am (admittedly self assessed) at C1 for reading and listening in German. So CI worked really well for me. The last time I looked a couple of years ago I had a knowledge of at least 10000 words in German, but that just doesn't equate to knowing all the words on a page. There are still lots of gaps as soon as you start reading serious newspapers or books, though I can still read and understand these things without a dictionary.


So despite the issues he states he has, he also states a C1 proficiency, a huge vocabulary knowledge base and not issue in understanding of films or newspapers. All in 5 years.

If you take him for his word for achieving C1 in 5 years, that's doing pretty damn good.
(Still, personally I'm not going to "buy" a CI input only approach, than you very much.)

It's funny how subjective success is - with "not doing that great", I'd be really happy with. I think part of the issue is the words he uses from post to post to either complain about his weaknesses vs communicating his success - we only see part of the picture.

Also, and this is my concern with CI methods - if I focus much less on output then I'm going to build less output.
1 x
Tagged posts: Language Method Resource
Please feel free to correct me in any language, critique my posts, challenge my thoughts.
I am inconsistency incarnate.
Go study! Publisher of Syriac, Aramaic, Hebrew alphabet apps at http://alphabetsnow.zyntx.com

User avatar
reineke
Black Belt - 3rd Dan
Posts: 3434
Joined: Wed Jan 06, 2016 7:34 pm
Languages: Fox (C4)
Language Log: https://forum.language-learners.org/vie ... =15&t=6979
x 6082
Contact:

Re: Examples of Input Only

Postby reineke » Mon Mar 04, 2019 6:05 pm

jeff_lindqvist wrote:
garyb wrote:Maybe being too invested in one learning philosophy is a language learning weakness? I remember a poster on the old forum (whose name I forget) who was determined to stick to the input method: avoiding speaking German whenever possible despite living in Berlin with a German wife, and steadfastly refusing to study grammar even though he admitted his grammar was a major weak point. He was very frank about his results, and eventually admitted what most of us know: input gets you quite far but most learners need other work on top of it in order to become a competent speaker in a reasonable amount of time. He's just one data point of course, but probably the best-documented example we've had of someone persisting with very input-focused learning over several years.


I think that was patrickwilken and his mega thread German: massive input in Berlin. More thoughts about it here: Krashen & some failures for massive input.


Monox D. I-Fly wrote:
garyb wrote:Maybe being too invested in one learning philosophy is a language learning weakness? I remember a poster on the old forum (whose name I forget) who was determined to stick to the input method: avoiding speaking German whenever possible despite living in Berlin with a German wife, and steadfastly refusing to study grammar even though he admitted....


OCD: Any learners' worst enemy


Fair play is sorely lacking here. The thread is entitled "Your language learning weaknesses".
1 x

User avatar
Tristano
Blue Belt
Posts: 605
Joined: Mon Jul 20, 2015 7:11 am
Location: The Netherlands
Languages: Native: Italian
Speaks: English, Dutch, French, Spanish
Understands but not yet speaks: Romanian
Studies: German
Can't wait to put his hands on: Scandinavian languages, Slavic languages, Turkish, Arabic and other stuff
Language Log: viewtopic.php?f=15&t=5141
x 912

Re: Examples of Input Only

Postby Tristano » Mon Mar 04, 2019 9:02 pm

golyplot wrote:You can add me to the list of people who study almost exclusively through input. Of course, that's partly a matter of expediency, because it is a lot easier to watch TV than to do anything which requires actual effort.


I'm an example, even though you are probably looking for someone more accomplished than me.
I don't use the TV method. I rather use translation and srs (memrise, clozemaster) to kickstart a language and move on reading extensively.

At that point my production sucks but gets very soon better, if explicitly trained. The languages I can't speak are also the ones I forgot the fastest, even if i can read in that language. For me an example is Romanian: I can still understand a good deal but can't recall anything that is not dragostea dintei or mamaliga. (Ok, yes, buna ziua)

I found that training production has a positive effect in enhancing recall and listening comprehension.
4 x

garyb
Brown Belt
Posts: 1168
Joined: Mon Jul 20, 2015 12:35 pm
Location: Scotland
Languages: Native: English
Advanced: Italian, French
Intermediate: Spanish
Beginner: Greek
Language Log: viewtopic.php?f=15&t=1855
x 3444
Contact:

Re: Examples of Input Only

Postby garyb » Tue Mar 05, 2019 9:23 am

reineke wrote:
jeff_lindqvist wrote:
garyb wrote:Maybe being too invested in one learning philosophy is a language learning weakness? I remember a poster on the old forum (whose name I forget) who was determined to stick to the input method: avoiding speaking German whenever possible despite living in Berlin with a German wife, and steadfastly refusing to study grammar even though he admitted his grammar was a major weak point. He was very frank about his results, and eventually admitted what most of us know: input gets you quite far but most learners need other work on top of it in order to become a competent speaker in a reasonable amount of time. He's just one data point of course, but probably the best-documented example we've had of someone persisting with very input-focused learning over several years.


I think that was patrickwilken and his mega thread German: massive input in Berlin. More thoughts about it here: Krashen & some failures for massive input.


Monox D. I-Fly wrote:
garyb wrote:Maybe being too invested in one learning philosophy is a language learning weakness? I remember a poster on the old forum (whose name I forget) who was determined to stick to the input method: avoiding speaking German whenever possible despite living in Berlin with a German wife, and steadfastly refusing to study grammar even though he admitted....


OCD: Any learners' worst enemy


Fair play is sorely lacking here. The thread is entitled "Your language learning weaknesses".


I think it's unfair to quote that without mentioning Patrick's reply and my apology for my over-simplistic interpretation of his method which (as is worth mentioning especially for the purposes of this thread) was input-heavy but certainly not input-only; he never shied away from speaking when he did have the chance, he started with classes, and he admits that reaching C2 likely requires lots of output and feedback.


In general, I just struggle to find any real examples of adults learning from input-only. Plenty people claim to have learnt a language, particularly English, from "just watching TV and films" but when you ask for more details it almost always turns out that there has been some amount of classes, study, conversation, and/or travel involved. These will usually be downplayed with statements like "I didn't start to see real progress until I started watching TV" and even criticisms of their past teachers, and in fairness the number of hours of input are usually orders of magnitude higher than those of the other activities, but with those other variables it makes it impossible to judge the efficacy of just input.
1 x

User avatar
zenmonkey
Black Belt - 2nd Dan
Posts: 2016
Joined: Sun Jul 26, 2015 7:21 pm
Location: Germany and France
Languages: Spanish, English, French trilingual - actively studying German (B2/C1), Hebrew, Tibetan, Setswana.
Some knowledge of Italian, Portuguese, Ladino, Yiddish ...
Want to tackle Tzotzil, Nahuatl
Language Log: viewtopic.php?f=15&t=859
x 4970
Contact:

Re: Examples of Input Only

Postby zenmonkey » Tue Mar 05, 2019 12:09 pm

Tristano wrote:
golyplot wrote:You can add me to the list of people who study almost exclusively through input. Of course, that's partly a matter of expediency, because it is a lot easier to watch TV than to do anything which requires actual effort.


I'm an example, even though you are probably looking for someone more accomplished than me.

....

I found that training production has a positive effect in enhancing recall and listening comprehension.


if you are training production how is this Input Only. As for SRS, are you saying you never pronounce the words you study?
How is SRS input only? You somehow need to output the answer to show you got it right.
1 x
Tagged posts: Language Method Resource
Please feel free to correct me in any language, critique my posts, challenge my thoughts.
I am inconsistency incarnate.
Go study! Publisher of Syriac, Aramaic, Hebrew alphabet apps at http://alphabetsnow.zyntx.com

User avatar
Tristano
Blue Belt
Posts: 605
Joined: Mon Jul 20, 2015 7:11 am
Location: The Netherlands
Languages: Native: Italian
Speaks: English, Dutch, French, Spanish
Understands but not yet speaks: Romanian
Studies: German
Can't wait to put his hands on: Scandinavian languages, Slavic languages, Turkish, Arabic and other stuff
Language Log: viewtopic.php?f=15&t=5141
x 912

Re: Examples of Input Only

Postby Tristano » Tue Mar 05, 2019 12:31 pm

I'm a lazy person so I don't use srs courses where I have to type the answer.
I further study with a long phase of input only, after which I still need to train production if I want to get good at it. With input only I could reach a a2 level at production and b2/c1 at passive skills.

Since I won't need to speak many other languages, I am mostly going to use input only, but if I also desire to speak the language then I don't see the point of not training production at all.
0 x


Return to “General Language Discussion”

Who is online

Users browsing this forum: No registered users and 0 guests