Bird Brains Explain How Humans Learn to Talk

eaglebtc writes "A team of neuroscientists at MIT have made tremendous progress in understanding how birds learn to sing: a part of the brain called the basal ganglia is primarily responsible for controlling the learning of movement and the production of speech. This circuitry is also present in humans, and it is the same way that a baby's random babbling eventually becomes the proficient speech of adults. It is hoped that this research can provide further insights into Parkinson's Disease, an inherited genetic condition that causes rapid breakdown of motor control and speech production. The full research study is available as a downloadable PDF."

Comments

Mirror

[alienfluid]

Wow, a 5.0 MB link from the main page of Slashdot! Here's a mirror of the PDF documents if the original site goes down.

parkinsons isn't very inherited.

[Vellmont]

Parkinson's Disease, an inherited genetic condition

While there's some genetic risk factors, it's not know what causes parkinsons disease. According to wikipedia having a parent with Parkinsons increases your lifetime risk of getting it from 2% to 6%.

Re:parkinsons isn't very inherited.

[Anonymous Coward]

Yup, some doctors think that Parkinson's is caused by insect poison, others think it is a virus.

Re:parkinsons isn't very inherited.

[kris_lang]

There are also other known causes for Parkinson's or parkinson's like syndrome:

badly brewed methamphetamines will have a high percentage of MPTP which will cause severe damage to the substantia nigra leading to Parkinson's like symptoms even in young people. It was in fact a bad batch of MPTP causing these parkinsonian tremors and symptoms in a group of 20 year old drug abusers that led to some scientific studies and discoveries and the creation of a primate model of parkinsons by injecting MPTP into primates.

Vasular: a stroke or microinfarct to the basal ganglia or substantia nigra can cause parkinson's like symptoms.

Toxic: I read about the insect poison thing too, but don't remember any key details or whether it has any MPTP like properties.

Parkinson's has also been treated in the past with Fetal Cell tranplants directly into the substantia nigra and into the globus pallidus.

It has also been treated by implanting electronic stimulating electrodes into the thalamus, more specifically into the VIM nucleus, in an attempt to disrupt the rhythmic tremors of Parkinson's. Very cool stuff.

Re:2 years later...

[cuerty]

Here is the link to the article (In spanish) no a too much tecnical review: http://old.clarin.com/diario/2003/01/09/s-02601.ht m and google translated for those who doesn't speak spanish: http://www.google.com/translate?u=http%3A%2F%2Fold .clarin.com%2Fdiario%2F2003%2F01%2F09%2Fs-02601.ht m&langpair=es%7Cen&hl=en&ie=UTF8.

Re:Chomsky

[grepMeister]

Chomsky is not in any kind of debate with Pinker as far as I know, and as far as I can tell from the article the parent linked to. The two have very similar viewpoints on the matter.

Chomsky put forth the following for how children learn language: it is a simple statistical fact that there are entire kinds of sentences that you understand and produce correctly that you are likely have to never heard before -- like English "Is the woman who is walking her dog Tom's neighbour?" where you are asking a question but there are two mini sentences, one about the woman walking her dog, and one about her being Tom's neighbor.

So you couldn't possibly be learning how to speak English (and not speak non-English) just by observing how often things occur -- because you wind up understanding that perfectly even though it just doesn't come up.

Both the claim that this really is a 'simple statistical fact' and the claim that you couldn't learn they are okay just by observing the statistical pattern have been HOTLY debated. But that is Chomsky's claim.

Something is missing, the argument goes, and that missing chunk must be mechanisms in the brain specifically dedicated to language. This too has been hotly debated, but it is both Chomsky's and Pinker's position (less so Pinker).

TFA bears on this issue in a tangential way. It's known that there's a circuit in bird brains that is required for them to learn their songs properly ('anterior forebrain pathway'). This research sheds some light on why it is required. The way I understand it (IANA neuroscientist), it is a keep-on-trucking circuit that says, 'okay, do it again. and again. and again.' Something like this mechanism is also found in humans, in the basal ganglia, so now we have maybe learned something about us.

However, many people on Chomsky's side are very suspicious of any 'generic' learning mechanisms like trial-and-error. For example, the article mentions babies' repeated babbling as a mechanism of trial-and-error to get the sounds of a language right very early on. But a potential alternate 'language-is-totally-innate' -- Chomsky/Pinker -- explanation might be that babbling does not feed back on itself, and a baby doesn't learn anything from it; rather, the baby's 'language faculty' has not matured (which happens with minimal help from what the baby hears) to the point where it can do anything else. I made this up - this is not a theory about babbling that can be found in the literature -- but it is the kind of alternative to 'general intelligence' that is often proposed.

The parent-linked article is about a book which vehemently denies Chomsky/Pinker's point of view. (FWIW I've read some of the author's other works and I think he's just a troll.)

Parkinson's Not Rapid

[Ranger]

Parkinson's Disease, an inherited genetic condition that causes rapid breakdown of motor control and speech production.

It is not an inherited genetic condition. There may be genetic factors. Nor does it cause rapid breakdown. The disease is a slow breakdown over many years. And a person can have a normal lifespan. It is treatable. My grandfather had Parkinson's. He lived to be 90. He had a shuffle walk and didn't have serious tremors like other sufferers.