[NSF GRFP 2010]

On an unrelated note, I find it irritating that clinical people are applying and GETTING the NSF - they're taking their previous research and spinning it as social psych. I know a couple clinical students on fellowship...and if I know a couple, then it's a sure bet that there are more out there...grrr.

I totally understand this. The website explicitly says that clinical isn't the place for this, and anyone in clinical should know that NIH is more important. But people will do what they want.

I'm also somewhat hoping that since funding for the NSF GRFP went up this year, and there should be more awards, I won't have to worry as much, but we'll see. It's good practice to write this sort of thing anyway.

[NSF GRFP 2010]

I applied! I have alright GREs (1350, 760 math, 5.5 analytical), a strong GPA (3.88 at time of application), really strong recommendations, and a pretty strong statement and proposal. I spent forever talking about the broader impacts and scientific merit in all my sections, and may have written two or three drafts before submission. I don't know that I've got a stronger application compared to anyone else, especially since I'm still in undergrad, but I do think I wrote some pretty coherent material.

We'll see how it goes. I'm pretty nervous about the whole ordeal. Coming into this straight out of undergrad seems to always give me this inferiority complex.

[What's your end game?]

My dream job is actually to move to Europe and work at MPI (Leipzig, probably, but I'm not terribly picky) forever. Great place to research, and that's largely all you have to do. Otherwise, professorship somewhere.

[BU, UChicago, Carnegie Mellon, U Toronto, and Georgetown]

One of the girls in my lab (applying for cognitive neuroscience -- mostly reading, dyslexia, neuroimaging) heard from CMU a couple weeks ago. Visiting weekend is next weekend.

[What's your specialization?]

I voted for computational psycholinguistics. Computational linguistics isn't "just" engineering; I'm interested in language acquisition, and one way to test a model of language learning is to write a computer program that does exactly what the model does, give it real data, and see if it learns the linguistic structures that we know humans must be able to learn.

As I see it, this is the only real way to test claims towards the ``poverty of the stimulus:'' claims that the speech kids hear doesn't contain very much information about linguistic structure. Real speech might contain a lot of cues that are not useful in isolation but can be useful when integrated by powerful (and psychologically-evident) statistical algorithms. And the only way to see how these things hash out is to actually expose these algorithms to the stuff kids hear and see how well they do.

You would probably love some of the work done by my thesis advisor, and some of the people he's worked with (Bannard, Tomasello). I'm into computational psycholinguistics, but as applied to adult language processing and production, so I'm glad to see someone else here. I voted for psycholinguistics itself since computational work isn't necessarily a subfield, just a methodology.

[What are you reading?]

God Bless You, Mr. Rosewater - Vonnegut

It's really strange to distract yourself from reality by reading satire.