<p>^Well, I’ve had a mini-love affair with the place for a while, and had considered applying EDI before summer circumstances changed my mind. But yes–LAC + ling + cogsci (almost a self-designed minor) + comparative lit + interpretation theory. Just generally winning at the complex game of offering weird fields of study.</p>
<p>I agree with many posts here. The one missing is Johns Hopkins. While they don’t have a linguistics department, their cognitive science department has linguistics as one of the concentrations. The department has real live red blooded linguistic scholars, something that you won’t find more than two of at any LAC, except for Swarthmore and Pomona. </p>
<p>It’s seems pretty easy to double major in Computer Science and Cognitive Science with Linguistics as one of the concentrations within Cognitive Science. It sounds like a serious option for you. It’s worth at least checking out.</p>
<p>^JHU does have an extensive cogsci program with sufficient linguistics. However, I haven’t heard many good things about its undergraduate focus; quite the opposite, in fact. (But the same goes for Stanford.) Any thoughts on similarities between LACs and JHU?</p>
<p>Keilexandra, my S is a linguistics nut, too, and is applying to some of the same schools. The tough thing is that he, too, would prefer an LAC, which makes the linguistics issue a bit tough, so, assuming he has some options to choose from come April… he will have to decide what’s most important to him. It seems like most schools have at least some courses scattered around the disciplines. He does not have as specific a focus as you, though. He just loves words, grammar, derivations, etc. He follows several linguistics blogs; do you? I also think Swarthmore would be a great place for him, but getting in there is extremely tough!! </p>
<p>I feel that Yale made a poor decision today. I have noticed your posts on this forum, and have been extremely impressed with your insights and how articulate you are. I will be curious to see where you end up. If by chance you and my S end up at the same place, I think you guys would be good friends!</p>
<p>Heh, I’m not quite a linguistics nut yet, I think; I’ve tried following Language Log, but the technical stuff goes whoosh. I do find the structure of language fascinating, in a similar way that I find literary theory fascinating (on a larger scale, of course).</p>
<p>Keil, I’d actually second the recommendation on JHU. When I was planning out my visit to JHU, I sort of just tossed it onto my list of schools to see, thinking I might as well see it while I was on the east coast. I was <em>completely</em> blown away; I really, really like Hopkins a lot. It seemed very friendly, the students who I talked to were very nice, and they had a lot of good things to say about the undergraduate experience at JHU. I had been wary of what I had heard in terms of a competitive atmosphere…for some reason, Hopkins seemed sort of cold and unwelcoming in my mind. I was completely wrong, though. They talked about research opps for undergrads a lot, which has more to do with physical sciences/engineering, but I would think that whole ‘philosophy’ would still apply. </p>
<p>I’m not totally sure how it compares to LACs in terms of more objective criteria like TAs, etc., but JHU was definitely my top “school I didn’t think I’d like, but then it turned out I loved it”…that kind of thing. I think it’s at least worth a second look.</p>
<p>^Good to hear an opposing anecdote. I have friends attending, so maybe I’ll try to swing by for a visit. Heard their need-based FA isn’t so good, though.</p>
<p>I know I’ve made this recommendation before, but I just looked at the Symblic Systems courses at Stanford and it seems that you could basically replicate those courses at Haverford (and probably Grinnell and Carleton as well).</p>
<p>^Definitely not replicable at Grinnell, where I sat in on a/the sole linguistics “theory” class and talked to the professor afterwards. Grinnell’s ling concentration is mostly interdisciplinary, with one CS professor who has an undergrad degree in linguistics and has offered a computational linguistics course alternate years. Carleton has 3 linguists on staff, so it’s probably replicable; I don’t know whether I could get a “Symbolic Systems” major approved when there’s already a “Cognitive Science” minor. But it’s more of a possibility.</p>
<p>At Haverford, wouldn’t I have to take upper-level linguistics at Swat or Penn? I really hate commuting; that’s what the small liberal arts college is for, so I feel like I’m attending class and living on the same campus.</p>
<p>
</p>
<p>As similar as night and day, as…well, you get the picture.</p>
<p>have you considered Bard? They have a CogSci concentration, a fabulous Literature Department, seminar style classes, tutorials, flexibility in designing majors, etc. The linguistics courses are scattered around and can be found in these departments: Literature, Classics, Psychology, Philosophy and maybe some other ones that I didn’t see (I didn’t even look at the Hard Sciences). In the Psychology Dept., they offer classes on Language Technologies and Psycholinguistics. </p>
<p>If you like Oberlin, Grinnell and Swarthmore, I would think Bard would appeal to you, too.</p>
<p>
</p>
<p>Chomskian “linguistics” has been exposed as a treadmill for some time. Only politics (literally!) has kept it in business. If that weren’t enough, the methods in use today for natural language processing and speech recognition are, in effect, a huge vindication of Skinner over Chomsky.</p>
<p>^ The dominant approaches to much of natural language processing and speech recognition today do depart from the Chomskyan, rationalist approach to language modeling that dominated in the 1960s-80s. I don’t know that these approaches are a huge vindication of behaviorism, or even particularly address related philosophical issues at all. They seem to be, if anything, a vindication of the usefulness of certain statistical modeling techniques and the brute processing power of modern computers. </p>
<p>Johns Hopkins University has embraced this statistical NLP approach. One major contributor has been Professor Frederick Jelinek, Director of JHU’s Center for Language and Speech Processing.
[Frederick</a> Jelinek - Wikipedia, the free encyclopedia](<a href=“http://en.wikipedia.org/wiki/Frederick_Jelinek]Frederick”>Frederick Jelinek - Wikipedia)
[Frederick</a> Jelinek’s home page](<a href=“http://www.clsp.jhu.edu/~jelinek/]Frederick”>Jelinek Fellowship - Center for Language and Speech Processing)
I know someone with a Ph.D. in Mathematics who once tried to take one of Dr. Jelinek’s NLP courses, but had to drop it because it was too demanding. That’s not to say that his presence at JHU does not lend light to the work of others, people who might be more accessible. David Yarowsky is another prominent JHU faculty member specializing in NLP (machine translation, etc.)
[Yarowsky</a> algorithm - Wikipedia, the free encyclopedia](<a href=“http://en.wikipedia.org/wiki/Yarowsky_algorithm]Yarowsky”>Yarowsky algorithm - Wikipedia)</p>
<p>The one word that comes to mind in distinguishing JHU from most of the other schools on Keilexandra’s list is research. If the Washington Monthly rankings are to be believed, Hopkins spends approximately $1.5 billion per year on research, the highest figure of any American university. This translates to abundant research opportunities for students, including some (I would think) in NLP.</p>
<p>A good introduction to statistical natural language processing has been written by Christopher Manning & Hinrich Schütze. To get a little of the flavor of the book (and of the field) you can browse a couple of chapters here:
[Foundations</a> of Statistical Natural Language Processing](<a href=“http://nlp.stanford.edu/fsnlp/promo/]Foundations”>Foundations of Statistical Natural Language Processing)</p>
<p>Creative Writing is another respected field at JHU. You can take a Linguistics minor there. Hopkins is much more than a world-famous hospital and medical school, but that reputation probably draws many pre-med students. If you are attracted to the rural environment of a place like Middlebury Vermont, Baltimore is likely to be a negative (though the campus itself is beautiful).</p>
<p>Bard looks interesting; I’ll do some research.</p>
<p>No comment on the Chomsky vs. X vs. Y debates, which I assume are ongoing because academics don’t agree very easily. Regardless, his theory was integral to the history and development of modern linguistics.</p>
<p>I’m still looking into JHU; the cogsci/linguistics looks very promising, to be weighed against negatives of urban environment and simply being a graduate-focused research university. But it’s worth mentioning here that their Creative Writing department isn’t what it could be. A close friend, majoring in environmental engineering but having attended multiple selective summer CW workshops, is pretty unsatisfied with her workshop experiences at Hopkins. She has declared a minor but still finds it nearly impossible to register for classes, as they (at least the intro/intermediate courses) are first-come first-serve based on registration time rather than portfolio review.</p>
<p>^K wrote “At Haverford, wouldn’t I have to take upper-level linguistics at Swat or Penn? I really hate commuting”</p>
<p>It depends how many courses you’d want to take. According to Stanford’s Symbolic Systems, you’d take two Linguistics courses (one in the Language and Mind category out of 5 options and one in the Linguistics theory category out of three options). It appears that two courses (one from each category) at Haverford/Bryn Mawr would not be a problem. </p>
<p>See [Haverford</a> College: Course Catalog](<a href=“http://www.haverford.edu/catalog/linguistics.php#Courses]Haverford”>http://www.haverford.edu/catalog/linguistics.php#Courses)</p>
<p>^Ah, but I’d probably take the Natural Languages track of Symsys, which includes far more than two ling courses: [Symbolic</a> Systems Program](<a href=“http://symsys.stanford.edu/ssp_static?page=concentrations/natural-language.html]Symbolic”>http://symsys.stanford.edu/ssp_static?page=concentrations/natural-language.html)</p>
<p>
</p>
<p>That the models are statistical and often black-box in nature (neural nets, genetic algorithms, etc) is behaviorism. No top down description, “just do it” through training. The data sets are larger than Chomsky imagined, because the relevance of economical top-down rule sets is smaller than Chomsky imagined. Chomsky’s hostility to statistics, which has set back the field by decades, is apparently a personal issue: the need to not concede a battle that only he invented in the first place.</p>
<p>It is certainly possible that less behaviorist approaches will rise in the future. But they will not develop out of anything Chomsky has done. They will come from functioning areas of science not driven by politics or a personality cult. CS, math, statistics, physics, biology, or the old-fashioned empirical study of numerous human languages.</p>
<p>At any rate, I’d settle for “a huge discrediting of Chomsky’s self-asserted competence to judge the scientific issues back in the 1950-60’s” as a separate conclusion from any positive assessment of behaviorism itself. 50 years of politics and intellectual cancer while hundreds of real languages went extinct. Thank you, Noam!</p>
<p>
</p>
<p>“Chomsky is awesome” read like an endorsement. There aren’t ongoing debates on these questions, just a silent withering of the Chomsky stuff’s share of funding. As his flock of grad students and grandstudents from the 1960’s boom years retires, the decline will accelerate. Know thine field.</p>
<p>
</p>
<p>The same could be said of flat earth, geocentrism, phlogiston and other discarded theories. Theory revision doesn’t usually run its course until the old school is dead.</p>
<p>Keilexandra, I think you have an outstanding list developing.</p>
<p>I’d drop the ones with questionmarks. If a school lacks scholars in an area that you are interested in, it’s pretty hard to study it. They all lack something that you care about and you have Rochester that lacks nothing. It is very LAC like. </p>
<p>I still think only Swarthmore and Pomona are the only “perfect” LACs for everything you’ve said. Both Carleton and Macalester each have one full professor and one non-tenured assistant processor, and Carleton has an additional contract scholar, who has very little chance of still being there by the time you’re even a sophomore. They will both probably be fine, but I have less confidence that they can sustain linguistics in the face of budget issues than Swarthmore, which seems to have a serious commitment to linguistics, and Pomona, which because of the consortium, has enough students to fill the linguistics classes. If you have time, you might just add Pitzer to increase the probability that you end up on the Claremont campus. Given that you’ve already completed a hard application in Yale, it probably takes less time to write one more common app than it does to continue to research the schools that you already know don’t meet your needs. </p>
<p>I’m also not sure why Chicago’s not there. I just assume that you’ve thought about it and eliminated it.</p>
<p>If you drop the ones with question marks, and maybe add Pitzer you have a great manageable list and you can mow through the applications in the next few weeks. Good luck. </p>
<p>Reaches
Yale, Stanford
Pomona, Swarthmore</p>
<p>Matches
Carleton, Macalester, Scripps, Pitzer?</p>
<p>Safeties
Rochester (cogsci is psych/neuro-heavy, but strong CS and ling)
UDel (state flagship; B.S. cogsci program is NOT a fit for me, but I could maybe design a compling major?)</p>
<p>I forgot to add, don’t neglect that access you would have at a Claremont school to Harvey Mudd’s computer science department.</p>
<p>Siserune - I still claim to know very, very little about ling; that’s what I’m going to college for (hopefully). But your vehemence is instinctively suspicious; even knowing nothing, I am disinclined to take you at face value because theory and social science are not so black-and-white.</p>
<p>ClassicRockerDad - Thanks for the analysis. Assistant prof is tenure-track though, isn’t it? I agree that Carleton’s contract scholar is to be discounted, but it takes a lot for a good LAC to start firing tenure-track profs. Unless I’m misreading the usual conventions of academic titles.</p>
<p>Pitzer fell off my list because it doesn’t allow majoring in CS… I’m not sure why, since Scripps also has zero CS courses but is happy to let a student major at Pomona or HMC or a combination of both (self-designed).</p>
<p>I have thought about Chicago endlessly, it seems; it has amazing positives (individually very strong academics for my interests, LAC-like university) and also significant negatives (urban location, Core making it very difficult to double-major and still explore electives, would not be able to major in CS unless I took a full year of physics or chem).</p>
<p>I should note that the applications for all of my “old” schools are complete, except for one Stanford essay. Expenses add up, of course. But work level isn’t too important, as I’m reasonably adept at writing college essays to flexible prompts.</p>
<p>Time to re(-re-re)consider Pitzer, Chicago, Oberlin/Grinnell/Middlebury. I think JHU is too “university” for me, and Bard–in addition to not meeting full need–seems to lean toward Vassar on the Vassar-Oberlin spectrum (Vassar being decidedly artsy/hipster vs. environmental/activist, though both schools are obviously ultraliberal). FYI for future linguistics people, apparently W&M has SIX linguists? I can’t apply, though, because their OOS financial aid is atrocious.</p>