The ALG-shaped hole in research on second-language acquisition

Every academic discipline has its research gaps and holes—those areas that haven’t been adequately investigated and the questions that haven’t been answered, or even asked.

The field of second language acquisition is no exception.

What’s fascinating, and at times maddening, about looking at the Automatic Language Growth, or ALG approach, alongside existing research on second-language acquisition is how so much of what ALG asserts and touches upon has not really been examined.

These gaps coincide so much with ALG theory and practice that I would say it’s as if the research on second language acquisition has an ALG-shaped hole.

In this post I will focus on what I think are the most prominent parts of this research hole: scientific control in research on the issue of age and second language acquisition, and the so-called “silent period”.

Scientific control

A fundamental part of the scientific method is to control for variables in order to be clear about what is causing the effects we observe.

Existing research has looked at language learners and found that the older one begins learning a language, the lower one’s attainment is likely to be.

The general conclusion is that going from childhood to adulthood, our brains lose the ability to pick up languages and effortlessly attain high levels of ability in them.

This conclusion however ignores a major confounding variable—that is, a factor that changes with age: adults typically experience and do very different things from children when learning a language.

These differences are so great that we can’t simply assume that age and inherent changes in the brain with maturity are to blame.

Adults learning second languages often lack the same environment of comprehensible input that children get to pick up language from, and they tend to use abilities that they have gained with maturity, such as the ability to practice and consciously think about language.

ALG theory posits that the use of these abilities interferes with adults’ acquisition, and asserts that if we give adults an environment that allows them to pick up language as children do and have them approach it the same way, they can become as native-like as if they had learned the language as a young child.

Surprisingly, there has been little research that tries to control for the differences between typical adult and child language learning approaches and experiences to see how much how we learn, as opposed to when, affects the results.

Of course, it may be difficult or even impossible to control for all these differences.

But given the importance of controlling for variables in scientific research, we should expect as least some attempt to control for some of these variables, over some period of time.

Yet, even this sort of limited research is quite rare.

Silent period research

One key area to examine in such research would be whether a long “silent period” of first listening to a new language before speaking it much leads to better pronunciation and other abilities.

ALG suggests that the adult ability and propensity to try to speak a new language without this foundation of listening experience are largely to blame for the pronunciation problems and “broken” grammar we commonly observe—not a loss of childhood ability as commonly believed.

Dr. James Marvin Brown, the American linguist and originator of ALG, put it this way in a treatise on the approach:

THE MISTAKE – Children can do something that adults cannot.

THE UNASKED QUESTION – What would happen if an adult were to just listen for a year without speaking?

OUR ANSWER – Both adults and children can do it right, but only adults can do it wrong.

Imagine a 4 year-old child and an adult reacting to somebody talking to them in a foreign language. The child most often just listens, while the adult usually tries [to] talk back. Now suppose that ‘not trying to speak’ was the child’s secret. It could be. After all, doesn’t it make sense that listening to things that are always right would tend to build the language right, while saying things that are always wrong would tend to build it wrong? It makes you wonder what would happen if adults were to do the same thing children do, (that is, just listen for a year or two without trying to say anything). It would be worth finding out. But it seems that this experiment was never tried. Not until recently, that is.

Brown goes on to describe the AUA Thai Program where the ALG approach has primarily been implemented:

In 1984, the AUA language center in Bangkok started doing precisely this in its Thai classes. The students just listened for as much as a year without speaking at all. We found that adults get almost the same results that children do. If adults understand natural talk, in real situations, without trying to say anything, for a whole year, then fluent speaking with clear pronunciation will come by itself. A lesser period of not speaking will produce proportionately less-perfect results.

Despite a few studies that suggest listening before speaking results in better pronunciation, this research is very limited, and it appears nothing has been carried out at the scale of the AUA Thai Program, which has had many students listening for hundreds of hours with little or no speaking.

However, while AUA claims good results, with some adult students reportedly even going on to achieve native-like levels of ability, it has not been the subject of formal scientific research.

A research proposal

A rigorous “silent period” study would have to have subjects not only avoid speaking from early on but also avoid things like reading, which involves speaking mentally.

Ultimately, what Brown appears to have discovered is that real issue is not speaking or not speaking but what mental processes are involved (hence my use of quotation marks when referring to the “silent period”).

Adults like children can make mistakes when first speaking, but if they have internalized the language first through enough listening experience their speaking will eventually correct itself.

In contrast, if an adult doesn’t speak out loud but is practicing the language their head and comparing it to their own language, they may still create the problems we commonly see in second language learners.

Considering that most AUA students do not follow the ALG approach and try to speak, enforcing not only a “silent period” but also trying to keep subjects from even thinking about the target language in certain ways might be very difficult to implement in a study.

AUA suggests that the most successful adult students of its Thai program are those who didn’t try, and in some cases came there having no interest at all in learning Thai.

Psychological research often involves keeping subjects unaware of the true purpose of an experiment until after it is over and they are debriefed.

Perhaps ultimately, to more fully control for the differences between adults and children, experiments could be created where adult subjects were exposed to a new language in a situation where they are not aware that they are there to learn a language.

But of course, this could be very complex and expensive.

However, even more limited research on the silent period could show a strong enough effect to suggest that the differences in results between adult and child language learners comes more from how they typically learn languages, not when.

This could spur further research in this area, and hopefully, lead to better ways and opportunities for people to learn languages.

This “silent period” research is a major part of applying scientific control on the issue of age and second language acquisition, which I think is the biggest part of the “ALG-shaped hole” in language learning research.

In my next post I will look at other areas that are also important to ALG theory and methods but haven’t really been examined in second language acquisition research.


2 thoughts on “The ALG-shaped hole in research on second-language acquisition”

  1. I’m sorry, but I can’t personally see this gap you’re talking about. If you check Google Scholar for the words
    silent period second language
    you get about 876,000 results. The silent period is not a particularly heavily researched part of SLA at the moment because it has gone out of fashion – previous research hasn’t really given anything that useful.

    One of the reasons for this is one that you allude to in your own post – most people aren’t completely “silent” internally when learning a language. You suggest that this inner non-silence is an impediment to learning, whereas others say that it’s a part of the learning process.

    Others still say that it doesn’t matter which is the case – this internal noise/silence is completely beyond the control of the teacher. If half your students are truly silent and half aren’t, and the course works for one or the other, then you end up with a course that it guaranteed to fail half your students, and pragmatically that’s not a great situation to be in.


    1. Certainly the concept of the “silent period” has been discussed a lot in SLA literature, much of it in the context of childhood SLA, hence the number of results you get (although putting “silent period” in quotes for the search returns far less).

      My interest is specifically in the question of whether a “silent period” of focusing on input before output can result in more native-like production. I don’t know if this aspect was ever truly in fashion as I can find only limited discussion of it the literature and very limited research, which suggests a positive effect, but not nearly enough to be conclusive one way or another.

      Some academics have noted this research gap, as I noted in my follow up to this post quoting the introduction to Input Matters in SLA, which the editors dedicate to Harris Winitz, who has done some research on this question:

      “Harris Winitz prompted us in our own research and teaching to take another look at input, particularly in its relation to the development of a second phonology. One of his intriguing ideas…has gone relatively unexamined, apart from Winitz et al. (1995). This is the claim that production from the start of L2 exposure influences the route of phonological development, ultimately resulting in an end state that is more distant from the target than would be the case were production delayed. Among the studies that the readers of this book might be roused to undertake, we hope that at at least one of you will take up the challenge to investigate whether an initial silent period does, in fact, ultimately lead to more native-like L2 phonology.”

      Regarding your comments that “most people aren’t completely ‘silent’ internally when learning a language,” and that I “suggest that this inner non-silence is an impediment to learning, whereas others say that it’s a part of the learning process”: I would say there could be truth to both these things. While it might depend on what specifically the learner is doing internally, I would describe “non-silence” more as a possible double-edged sword.

      For example, suppose an adult learner is deliberately practicing constructing sentences in their head after being exposed to some of a new language. This practice might help them remember and use the language more fluently in the short run but I would suggest it could also inhibit their ability to approach native-like levels of fluency with accuracy by ingraining pronunciation or other errors.

      My contention here is this: it appears likely that the increasing tendency with maturity for L2 production and analysis in various forms to precede internalizing L2 input is the main cause of difficulties and problems we commonly see in adult L2 learners—not any inherent loss of ability to acquire languages. I think a major gap in SLA research is the lack of attempts to control for this difference between early childhood and adult SLA.

      As it is today, practically all deliberate language teaching and learning that happens from childhood on involves at least some elements of study and practice from a beginning stage. As I have explored in other posts on this blog, there are generally not really any practical options for adults who want to pick up languages without involving these things, even though there could be.

      If more research was done on things like the effects of a “silent period” on pronunciation and other aspects of “input before output” that showed positive outcomes, this might do a lot to inform language teaching and spur the development of more opportunities for language learners of all ages.

      I suspect there are many situations today where learners would be much better off focusing on gaining understanding of the L2 through listening to high-quality comprehensible input initially rather than being expected to speak, read, analyze, and so on from the start.

      I am absolutely with you in being concerned about the pragmatic question of how whatever might work in theory actually plays out in the classroom, or any other language teaching or learning situation. We should be mindful of individual differences or the possibility thereof, such as you describe.

      On that note, I hope you would see the value of developing abundant opportunities for comprehensible input that don’t involve other study or practice, similar to the AUA Thai Program, which uses the ALG approach.

      Besides comprehensible input being widely considered critically important to language learning, I think at least, such opportunities to pick up languages this way might benefit a sizable minority of adults much more than what is currently available.

      For example, one prominent academic suggested based on my description of the AUA program that it might help those who don’t do well in traditional language programs, echoing reports from AUA of students who have had difficulty with other language study taking to it particularly well.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: