As requested, this space is now dedicated to
dissemination of updates.
Latest update(12 Nov) :
I created a darcs
repository on Jones to hold my python-based OT work. More will follow.
I update my journal sporadically, but it's usually about programming.
About
My name is Nathan Sanders, one Christian and hacker, sometime styled ZackMan, the translator. I spend most of my time as a computational linguist in the Ph D program at Indiana University. My web site (which is what the title says, in Esperanto) has some information on my current projects, some miscellaneous hacking, and some old writing. I also maintain a blog, which I recently switched to standard software, and have some pictures.
Dialect Distance
Primarily simple phonological distance. (Older code is available here, as is some newer code.) I am in the process of exploring syntactic distance next; I have been chopping up the ICE-GB into regions and using a test of statistical significance that Nerbonne and Wiersma used previously. If you are at ACL 07 this year, you can see my poster there.
Optimality Theory
Specifically, computability and learnability of optimality theory, and computational implementations thereof.
- I have contributed to OTableau, which is a graphical tool that will produce LaTeX-formatted tableaux. I hope to extend it to graphically show various constraint-demotion algorithms, and perhaps some other OT learning algorithms. If you use Latex to write phonology papers, I highly recommend OTableau.
- I have a small web-based demo of constraint-demotion learning algorithms at the IU linguistics site. Although it is supposed to integrate with the OT constraint wiki on Jones (the IU comp ling server), I have not had time to make a nice web interface to the new code. However, if you have darcs, you can pull the code from here.
- I created a programmatic interface for creating LaTeX tableaux: see ot.zip. This is mostly superceded by OTableau, which has had more bugs shaken out of it. The phonological representation is a little ad-hoc in places because I hate parsing. In the available version, the winner-finding algorithm loops forever on ties (which are not strictly supposed to happen in classic OT).
- Now that there is a wiki on Jones, I hope that some Real Linguists will provide some more definite ideas on what is important to represent phonologically so that I can figure out the best representation. So far this hasn't happened.
Utility Code
Several years ago, after reading Paul Graham's On Lisp, I started a library of utility code. My experience with solving research problems in a functional, agile method is condensed here. The code brings a common environment to Scheme, Common Lisp and Python that is based on functions from On Lisp, Common Lisp, Python and Haskell. Download the zip file here.
Hacking
A few small things I've written that I like. Typically these are implementations of things we talk about in class, or expositions on some particular style. I have a problem with understanding formal things unless I've tried to implement them formally. My brain is not very formal. Check out what Sussman has to say about this.
Avocational Interests
Every Ph D student has a list of weird interests. Here are mine. Of course I don't actually have time for them.
- Playing video games, preferably old Japanese ones.
- Aikido. OK, I lied, I do have time for these first two, three hours a week.
- I don't know if I should count cooking or Chinese food as a hobby.
- Translating video games (at a fan level)
- I really like comic strips and their history. My favourite is Pogo, followed by Krazy Kat and Calvin and Hobbes. Also, I recently discovered Dinosaur Comics.
- I like programming languages, but this is sort of cheating since this isn't exactly avocational. I would like to find a way to relate natural languages and machine languages at some level besides purely semantics. So far the closest thing we have it "Perl has adjectives", which is not inspiring to me.
- Also, I prefer to spell things using the Queen's English. My hypothesis is that I read more English literature as a child than American and got used to the style. I can corroborate this with compositions from my entire school career. Since I was home-schooled, I don't believe that I ever got corrected for it.
Contact Nathan at
- ncsander@indiana.edu