Non-invasive nerve stimulation boosts learning of foreign language sounds: study | Hacker News Reader

Non-invasive nerve stimulation boosts learning of foreign language sounds: study

medicalxpress.com

132 pointslelf60 comments

This is near-quackery. The vagus nerve is composed of ~100k fibers of which about 80% are afferent fibers from organs. Each of the fibers conveys information in a fairly well ordered manner as do most neurons. Imagine if you will a highway with 100k lanes entering an absolutely massive city. Then you periodically turn on and off an earthquake machine which causes some cars on the roads to go faster, some to crash, and some to slow down. Then you observe the behavior of the city's economic output to see what happens. In one study one measure of the economy increases and you publish a paper about it.

There is both a scale disconnect (the signal being introduced by the electrodes vs the thousands of isolated communication channels which operate independently) and a causal disconnect (you have no idea what happens between the disruption and the observed behavior). A lot of TMS studies suffer from these methodological flaws as well.

Rather than saying things like this "opens the door to improving cognitive performance across a wide range of domains" they should be saying "we have no idea why this works at all, and you should absolutely not try this at home."

A simpler and cheaper brain-stimulating intervention to improve language learning has been known in the literature since 1972. No fancy electronics necessary.

“The results of this study are interpreted as confirming the hypothesis that the ingestion of small amounts of alcohol, under certain circumstances, does lead to increased ability to authentically pronounce a second language.”[1]

Note that the improvement only happens after consuming exactly 1.5oz of cognac. Anything more than that and performance drops precipitously.

[1]https://deepblue.lib.umich.edu/bitstream/handle/2027.42/3404...

Ah, they found Ballmer's peak before its time!

For those wondering: https://xkcd.com/323/

Anecdotally, phenibut has the same effect without being so sensitive to getting the right dosage.

It should also be mentioned that phenibut can create very serious dependency issues if taken regularly.

> Note that the improvement only happens after consuming exactly 1.5oz of cognac. Anything more than that and performance drops precipitously.

Anecdotally, this appears to be true for billiards as well.

Does it have to be cognac, or can I use some other brandy?

Per the paper, the treatment was actually a punch that was half liquor, half other things. The liquor was a mix of cognac and light and dark rums.

And what about whisky?

Vagus nerve stimulation (noninvasive or surgical implants) is used to treat many conditions like pain, epilepsy and depression. Vagus nerve stimulation is also used to tread bad cluster headaches.

ps. Nerve stimulation is more targeted and precise than psychiatric medications. Meds just go everywhere.

These two statements - this one and the parent post's - are entirely compatible.

It's amazing how often treatments are introduced, and even become quite popular, despite being supported by only scant evidence. Sometimes this leads to good outcomes, sometimes, not so much.

One could make a strong argument that a more liberal approach, even outright grasping at straws, makes sense if you're dealing with a serious condition like chronic pain, epilepsy, or depression, and more proven treatments have already been tried without success.

It's maybe not so clear that it's wise to use an intervention that's also supposed to be effective for major ailments like pain, epilepsy, and depression, just because a preliminary-looking study with low statistical power says it might help you burn through your Anki deck 13% faster. Side effects do happen.

Desperation is a big driver. Medical doctors quickly give up on anything tricky. When your in horrific agony, and you can see a doctor for 5 minutes every three months. Well people find other sources of information.

Facebook and reddit groups are a wealth of finding out what options are available and how they might affect you. There is a lot of “quack” info in the groups. But also a lot of great tips.

Topomax is a common example. Major side effect is it makes people stupid. Not everyone, but a large percentage. Neurologists often fail to mention this.

What does "noninvasive or surgical implants" mean?

Anything without peer reviewed multiple intergenerational studies would not trust worth. All these new biohack tech I would only consider them to be an experiment. It is find that you want to volunteer for the, just treat them as an experiment and accept the risk involved.

Parent is not arguing that Nerve stimulation is not used.

The argument is (as with many of these therapies) that it is rare to have robust enough methodologies when testing this and that the positive effects claimed are often null/negligible or worse non-reproducible.

Arguing peripheral nerve stimulation is effective because it is used as a therapy, gives it as much evidence-based robustness as homeopathy.

> The argument is (as with many of these therapies) that it is rare to have robust enough methodologies when testing this and that the positive effects claimed are often null/negligible or worse non-reproducible.

parent did not argue that. The parent criticized the method and did not challenge the result.

> Imagine if you will a highway with 100k lanes entering an absolutely massive city.

Name me any medical drug that doesn't work by statistics.

Your post reminded me of "Can a biologist fix a radio?", [1].

[1] https://www.cell.com/cancer-cell/pdf/S1535-6108(02)00133-2.p...

Does it do any lasting damage? Then why not try it? Lots of things 'stimulate nerves' including playing the piano or typing. If typing while learning Mandarin, improved your Mandarin, would we be admonishing the experimenter?

Interesting. There's a whole class of these non-invasive field methods to affect human brains - https://en.wikipedia.org/wiki/Transcranial_magnetic_stimulat... is another similar idea.

I have this cool book that I've recommended before on this site called Neurochronometrics of Mind or something close to that which discusses a lot of the data around these methods. It's super dense but I was interested in trying to build a device 10 years ago and was discouraged by most. Seems like the data points to it potentially yielding a plethora of positive outcomes if done properly, but probably not by a non-clinician. Really interesting stuff. Imagine watching your own brainwaves with a 4096 channel OpenBCI kit (openbci.org iirc) and then engaging in a biofeedback experiment where depending on how you consciously react to seeing your various brainwaves in real-time, the hypothetical combo device also has a series of rTMS tori that each selectively expose you to varying gauss levels (give me some rope here I'm not a mad scientist) that the various parts of your brain get exposed to. Not sure how applicable or useful it might be but it's interesting to think about. A colleague of mine worked on visualization of brain waves for technological applications and told me it was a pretty wild experience to see your brain waves in real-time while also watching them in parallel, so I imagine creating a tight feedback loop with an inbound magnetic field in response to your thoughts might have some weird and possibly pretty trippy effects.

Disclaimer: don't do this, this is probably how you get spontaneously generated warp singularities appearing in your lab

Personal brainwave biofeedback sessions seemed like a tantalizing idea too, when as a teen in the 2010s I was struck by it while researching neurology. I think it was a video of a long-time meditator who demonstrated total control of his brainwaves that really got my attention. Anyway, I happened to be in the Netherlands for a summer and ordered a mini OpenEEG PCB from it's designer from neighboring Germany. That's when I found out I had no idea how electronic circuits worked nor how to assemble SMD boards. I relegated it to a pipe dream after that, but have always been on the lookout for an easy to use device that could provide the same.

My wife is having a TMS treatment right now.

Feom what I can tell something has changed, and she seems happier.

ECT is one of the most effective treatments we have for depression, and TMS certainly promises to allow a procedure that's similarly effective but less invasive/without anesthesia - reducing costs significantly.

You can become manic from ECT, and I'd assume you can with TMS as well, but I'm sure your wife's psych knows all about the risks.

Hope you wife is doing well. How have the long-term effects been for her, and on the relationship, if you don't mind my asking? Any noticeable shifts in mood and behavior?

It looks to me like listening practice, really. This is something that people usualy do when they want to learn a new language - we are learning languages not by sitting in class but through immersion. I'm wondering how the results would look like if they'd introduce third group - people who actively watch/listen to Chinese TV shows or podcasts.

I’d bet this is useful for a range of memorization tasks other than language learning. I imagine this works because it strengthens the signal to noise ratio by pairing auditory stimulus with vagus nerve stimulation. It’s kind of like how reading something out loud can help you remember it, only in this case it’s like you are reading aloud and listening to someone else read at the same time.

I wonder if it could be used to improve reading comprehension. Say you are reading a story, maybe each character has their own waveform, and as you read or listen the stim pulse changes based on who is talking.

Or lets say you are reading a math textbook. You could label different variables with different stim pulses, which might make it easier to follow complex logic or minimize the need to re-read different sections.

Imagine if scientists could one day figure out how to expand the linguistic critical period into adulthood, we could all be polygots.

I'm a polyglot and half my languages were learned as an adult. I really think you just need to immerse yourself for a few months.

It feels so so strange to me that the science is apparently still out on that.

I too find it strange. The most obvious explanation to me is that adults simply don't have the time, energy, or money to devote to consistent practice with a fluent speaker where ones effort is being tested in some way like children. And no Duolingo doesn't count as.

> consistent practice with a fluent speaker where ones effort is being tested in some way like children

I agree with the total time and deliberate practice aspect but in my opinion you are underestimating one extra factor for adult learning.

As an adult if you practice a new language with natives (outside the context of a formal lessons) the majority of people will be a lot more accomodating of mistakes and mispronunciation simply because of the social pressure of wanting to 'save face' and not trying to demotivate the person who is learning.

There are maybe a handful of close friends who will be brutally honest enough to correct you, in the same way they would correct an infant. For the most part as an adult it is more common that you end up practicing with an 'excessively tolerant' feedback loop from other language natives.

I tried being “brutally honest” with an Indian coworker about her English. She started to studiously avoid me, and I realized I’d made a huge misstep following the “golden rule” so to speak. If I learn a new language I want people to constantly correct me.

I have another coworker who is Eastern European and she has explicitly told me “point out EVERY mistake I make. Be brutal, I want to learn perfect English.” So I do. It seems like she really is getting less of an accent from it. It’s pretty cool to see.

People are nice and don’t point out mistakes because some people don’t want this. It’s taboo on both sides, at least in the corporate US world. So they never learn to speak fluently just enough to get by.

Use app.

While I agree in principle, there are things that are much harder for adults to learn -- like perfect pitch.

Perhaps children are much more immersed into the activity than adults.

I recently stopped all cartoons for my 2.5 year old daughter, at the recommendation of her educator and multiple therapists because she only ate when watching cartoons -- but if she ate normally, I wouldn't do so -- she picked up english(not her primary language) words and idioms and little songs because she enjoyed the cartoons so much.

Would I pick japanse words and phrases if I watched cartoons as easily as her? Perhaps, if I was as interested and enthralled.

You can't learn perfect pitch as an adult, lookup the Rick Beato video on this.

That doesn't change the fact that during the critical period people are muuuuch better at acquiring language than after. Moreover, while people are fairly equal in their ability to acquire a language during the critical period, there is _much_ more variance in ability, all else being equal, after the critical period. The science isn't still out on that.

Variance in language-learning ability after the so-called “critical period” can be to large extent explained by increase in its optionality.

As a child you really have no option but to learn your first language. The entire environment around you screams “learn the language” every hour of every day. People around you are incessantly encouraging you to learn it.

When you’re an adult, it’s very optional for obvious reasons. In fact, certain social dynamics might actively discourage one from learning a new language.

(Bear in mind the difference between first- and higher-order volition. The former does not readily lend itself to conscious control and is probably impossible to measure—merely spending large amounts of time on learning a language is not a reliable indicator.)

Maybe they are just much better at not being so easily discouraged.

And as for the study:

By the end of the training, those participants were 13% better on average.. There's a general feeling that people can't learn the sound patterns of a new language in adulthood, but our work historically has shown that's not true for everyone," Chandrasekaran said. "In this study, we are seeing that tVNS reduces those individual differences more than any other intervention I've seen."

"In general, people tend to get discouraged by how hard language learning can be, but if you could give someone 13% to 15% better results after their first session, maybe they'd be more likely to want to continue."

Sounds like placebo to me. ROMANES EUNT DOMUS.

I think it's hard to bring this study to bear on the critical period hypothesis in toto since its scope is narrow in both time and grammatical breadth (just sound production, and no other parts of comprehension or production), so we should maybe leave that aside.

You're right that the critical period hypothesis is still not settled, but let's go back to your proposal, that anyone who's immersed and motivated enough will be able to acquire nativelike proficiency. While we're in the realm of anecdotes (since you make reference to your own experience), I witnessed several peers of mine in a months-long language immersion program fall behind others in the program by quite a margin, with (IMO) little hope of nativelike proficiency even with months more of intensive classes, despite apparent motivation (sometimes much more than the better performers) and 24/7 immersion in the target language.

Going beyond anecdotes now, let's consider results from the literature against the assumption that if only adults could not be so easily discouraged they would be able to attain more nativelike proficiency. In that case, we'd expect at least a few places in time and space to have second language learners who do a phenomenal job of acquiring the target language, yet as far as I know that's never been found in the literature. Birdsong[0] reports that the range of attested rates of nativelike proficiency in L2 learners is between 5% and 25%. Moreover, there's a wealth of literature that documents in detail the specific ways in which language acquisition mechanisms falter beginning around age 6[1], which is in line with the suggestion that nativelike proficiency becomes progressively harder with age.

Neither of these results pushes your suggestion beyond plausibility, of course, but it does seem to me like the preponderance of evidence is on one side. If you know of any studies with findings that support the idea that motivation can be a decisive factor in acquiring nativelike proficiency in a second language I'd be really interested to see them.

[0]: https://web.stanford.edu/~hakuta/www/research/publications/(...

[1]: https://www.jstor.org/stable/pdf/44488300.pdf?casa_token=YS2...

> In that case, we'd expect at least a few places in time and space to have second language learners who do a phenomenal job of acquiring the target language, yet as far as I know that's never been found in the literature.

Respectfully, your literature is gobbledygook. :)

If the preponderance of evidence is on one side perhaps the academic researches need to get out more and collect better evidence.

I suggest as a start going to touristy places in developing countries (SEA, Peru, 'etc) where the guides are economically incentivised to speak half a dozen languages. I've encountered this sort of person over and over. They simply had to learn, as adults, and so they did.

Bet you'd learn French too if you worked for tips and your kid was hungry tout de suite!

And as for literature, try cracking a few books open from previous centuries that involved extensive travel for the author.

I was intrigued in your experience as a polygot and was buying your arguments until I read your suggestion to listen to guides then that completely reversed it because almost all the many many guides I've had are definitely not fluent in their non-native language in the same way children gain fluency. Most have broken language in the same way older immigrants have broken language despite the fact that are forced to navigate in their adopted language full-time.

Have to start your search somewhere. Watch out for an older distinguished-looking gentleman giving proper tours, not the fellows that accost you at the exit of the bus/train station with a few simple sentences.

If I was a researcher with no travel budget I'd try cross referencing a list of medium-small touristy towns (like Cusco) with decades that had upticks in tourism and a very bad local economy. Narrow it down to above-average income college grads and start asking them if they are or know of a polyglot. Can probably get pretty far with nothing but Facebook.

Just a guess, maybe I'm wrong about this pattern. Let me know what you find out. ;)

Edit: I was also intrigued by your comment until I read "in the same way older immigrants have broken language" - this is indeed the common but rather narrow experience. I don't think I'll be able to convince you of anything. Best assume you are corresponding with the Loch ness monster.

My work requires me to travel with guides, I have no idea if they are all above average income 'gentlemens' as you say (before you edited it out) so I can't really comment on that. I'm an immigrant myself along with my community so there's that. If you consider that a 'narrow' view not sure what more to say. There's a difference between fluency and native-like fluency that kids obtain. I would love you to post recordings of you speaking and ask the public to see if they can tell if you're a native speaker or not because I genuinely am curious.

The stimulation is imperceptible

> That doesn't change the fact that during the critical period people are muuuuch better at acquiring language than after.

I don't know about direct comparison of abilities of an adult and a child. The trouble is: child does nothing but learning, not just language, but he/she has a lot of daily practice. At the same time adult learning a language could spend a few hours per week learning, forgetting about the language at all other times. This difference itself might explain differences in outcome, there is no need to invent one more reason like "critical period" for the explanation.

To learn language child needs ~3 years. Give me 3 years of learning language for 5 hours per day, I'll learn it no worse than a child or even better.

I think people are different. I spent two years in Tanzania, and managed to pick up maybe only ~1000 words of Swahili while my girlfriend became completely fluent.

Maybe I wasn't trying hard enough, but what I can say is that simply immersing yourself for a couple months doesn't cut it.

Funnily enough, a friend of mine spent 2 years in Japan with his wife and he never learned anything while she ended up fluent (just like my gf).

Maybe men are worse at learning new languages? Or maybe they care less?

Oh interesting, what's your native language and did you only know one language before adulthood? There are studies that have shown that the likelihood of being a polygot is influenced by having a multilingual childhood.

Not OP but I moved to Germany as a single language speaker (well, some French from school that I was never much good at) in my mid 20s. Language classes for a few months and immersion made me pretty fluent over a year or two. Adults can learn languages just fine with motivation and immersion, I think.

My children who were born here later on already speak better German than I do. Of course, I work in English and they go to German school, so it’s not exactly like-for-like. But their English accent is better than my German one and they’ve a LOT less experience there than I do in German.

I work for a foreign language instruction company. Our entire business is based on being able to teach adults languages. And they aren't easy languages, our largest program is Chinese. We've been at it since the 80s.

It is definitely possible to learn a language as an adult.

The main barriers adults face are a lack of time to practice, a lack of motivation and already being fluent in a language to "run away" into.

Learning a language is tough, most native speakers have simply forgotten how much of a struggle it was in their younger years.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3848041/

"Valproate reopens critical-period learning of absolute pitch"

Maybe some possibilities for language, too.

tVNS devices are available to buy at https://vagus.net

It seems similar devices are available on Amazon for considerably less - any idea why the significant hike in price?

Devices on amazon are repurposed tENS devices which control voltage rather than current. tVNS is an emerging area of research and to repeat studies you need to understand the "dosage" which means the current delivered, because the same voltage can result in different currents depending on resistance. The accuracy of the biphasic waveform is also a factor.

I checked your bio and this is obviously an area you know a great deal about, and one I'm interested in.

Do you happen to have any high-level resources on the differences, pros & cons of TENS/tVNS/tDCS/PEMF and why one would be chosen over the other?

I'm slowly starting to piece things together, but frankly I'm a bit overwhelmed by it all

Thanks. They're all experimental and YMMV. What's exciting with tVNS cf tDCS is that you can use HRV as a measure of efficacy.

Every time I read a neuroscience paper I remember that one study where they had neuroscientists try and fail to reverse engineer some classic arcade games with their methods.

I use a tdcs device 20 minutes a day while doing dual n-back practice, and would recommend this regimen to others who are interested in boosting working memory.

Tdcs devices are really flexible and different electrode placements have different effects. Check out https://totaltdcs.com/ for more info.

https://tdcsbrain.com/product/apex-type-a-4ma-std

Looking at this product, it seems failsafe. You just put the sponges on your scalp and adjust the voltage?

I've never used that particular device, but generally you soak the sponges in a saline solution, place the electrodes on the desired locations, and choose your desired voltage. Anything under 4ma is considered safe, though I personally never go past 2ma.

I use the improved attention montage: https://totaltdcs.com/electrode-placement-montage-list/impro...

That day when no wants to put any effort into anything will be a fun day in the history of mankind.

Perception of phonetic features which are not markedly meaningful in your native language is a very hard thing. It's not just a matter of putting in effort of not.

Reply on news.ycombinator.com