Among other things, I work on the use of language processing to find treatments for spinal cord injuries, so every once in a while, I go to a conference on computational neuroscience. Computational neuroscience is a very broad field, and I might hear talks on subjects as diverse as optic nerve injuries in rats, the mechanics of how to measure brain activity in a mouse, and some of the truly horrifying things that you can do to a rodent in the name of science. A couple months ago, at a computational neuroscience conference, I heard a very nice talk by Jesse Goldberg of Cornell University on the topic of dopaminergic error signals in songbirds.
Why is it important to study dopaminergic error signals in songbirds? Songbirds present an excellent opportunity for studying how humans learn language. Not all birds sing. Ornithologists distinguish between bird calls and bird songs. All birds have calls, but only songbirds sing. Song acquisition in songbirds has some similarities to language acquisition in humans. We tend to think of animal behaviors as instinctive, and many of them are, but even instinctive activities may require some amount of social context and experience to learn correctly. Language is very much like this. Humans have an instinct to acquire language, but the specifics of the language that they learn have nothing to do with genetics—it’s completely a matter of which language(s) they’re exposed to in childhood. If I had been adopted by a Malayali family, I would be speaking Malayalam today, not English. Similarly, songbirds have to be exposed to adults singing the song of their species, or they will not learn it.
The similarities between acquisition of song by songbirds and acquisition of language by humans do not end there. Human infants babble. So do songbirds. If human infants don’t have feedback, they stop. (Human babies, whether hearing or deaf, babble vocally. They also babble manually. Hearing babies who aren’t also exposed to sign language stop babbling manually, but continue babbling vocally. Deaf babies stop babbling vocally, but continue babbling manually. Expose a hearing baby to both spoken language and signed language, and they will continue babbling in both modalities. I don’t know whether or not songbirds that aren’t exposed to song from other members of their species stop babbling.)
If the songbird is going to transition from babbling to the song of its species, it will have to be able to tell whether or not it is singing well. How can a bird tell if its song is any good or not? It looks like it happens through dopaminergic signaling. Dopamine is an organic chemical that is best known for being associated with rewards and pleasure in humans.
There are things that you can do to a baby songbird that you cannot do to a baby human. You can raise them in isolation from other members of their species. (Once upon a time you could deafen them, but I hope that no institutional review board would allow that today.) If a songbird is going to learn to sing, it can’t be exposed to the song of its species at just any old time—it must be exposed in its childhood. Raise a songbird to adulthood in isolation and then put it in a community of its species as an adult, and it’s too late—the bird will not learn its song. There is what we call a “critical period” for this kind of instinctive learning behavior—the learning has to take place by a certain point, or it can’t happen. The same is, unfortunately, true of language acquisition in humans. It is why I will never speak French natively, no matter how hard I study. It is why when deaf children are raised in isolation from other deaf people, they end up in the horrific situation of not having any native language, spoken or signed. (They will typically invent a signed language within their family, but can’t use it to communicate with anyone else, and can’t later learn another signed language natively.) It is the reason for the horrific story of Genie, which I won’t depress you with if you haven’t already heard it. The bottom line is that there are situations that can lead to human beings who have no language, and this is a horrible, tragic thing. If we can learn things from songbirds that might help us to help these people—and there are far too many of them in countries where there are not adequate services for deaf people—then I, personally, am happy to have some of my tax dollars (as we say in the US) go to research on how songbirds learn to sing.
In Jesse Goldberg’s talk about his research on dopaminergic signaling and the role that it plays in the acquisition of song by songbirds, he sang the song that the baby birds were trying to learn, making for one of the more memorable scientific talks I’ve ever attended. He sang beautifully! Since the French Wikipedia article on songbirds is short and not very interesting, here’s some vocabulary from the French Wikipedia article on dopamine:
La dopamine (DA) est un neurotransmetteur appartenant aux catécholamines, issu de l’acide aminé tyrosine. Dans le système nerveux central, elle active les récepteurs dopaminergiques postsynaptiques. Elle est principalement produite dans la substance noire et dans l’aire tegmentale ventrale2, situées dans le mésencéphale (partie supérieure du tronc cérébral). Bien que la dopamine, avec la noradrénaline et la sérotonine, soient très minoritaires dans le cerveau, puisqu’ensemble, elles concernent moins de 1 % des neurones, elles jouent un rôle modulateur final essentiel des sorties motrices et psychiques.
- être issu de: to stem from, to derive from; to be produced by.
- bien que + subjonctif: Although. We’ve seen this construction before, but it bears repeating, as I forget about the subjunctive every time, I’m sure.
- minoritaire: in (the) minority
- la sortie: I think that in this context, it’s an efferent nerve, but I’ll have to check—it’s not in WordReference.com. It looks that way from another token of it that I found on the French Wikipedia page on the insect nervous system.
Zipf's Law 