<p>Hello everyone… well, as the title says, I want to do a double major in linguistics and computer science when I am in college. I realize that the two aren’t very obviously related, though, and I’m not really sure what the best school to go to would be for both. I’m thinking of Carnegie-Mellon… but I’m not sure. Would that be a good choice? Are there better schools out there for what I want to do?</p>
1 Like
<p>Well the obvious overlap for linguistics and computer science is computational linguistics, and MIT is the clear leader in that field. So if you have the stats, MIT is the best school out there for what you want to do. Stanford and a number of the UC’s are also strong for linguistics. A linguistics program that emphasizes computational linguistics will probably be your best bet.</p>
1 Like
<p>MIT, Stanford, UC Berkeley, Penn, Harvard, and Carnegie Mellon are all strong in both computer science and linguistics. You should check with each university, however, to see if it’s even possible to double major in these two fields. At Penn, for example, linguistics is in the College of Arts & Sciences while computer science is in the School of Engineering & Applied Science (SEAS). I don’t know whether they’ll let you be enrolled in two schools at the same time.</p>
<p>Also, keep in mind that MOST undergraduates change majors at least once, so choosing a college based on your intended major entails some risks. That’s why it’s often wiser to choose the strongest school overall.</p>
1 Like
<p>Artificial intelligence is a classic are for the combo of linguistics and computer science. That said, my impression from the outside is that AI seems to be moving away from a linguistic approach and towards a biological approach.</p>
<p>^ It’s true - a combination of CS and linguistics predated AI and essentially started the field of AI. As biological methods become better understood, they’re more relevant, but neuroscience is still rudimentary here. Linguistic methods have become more important within the past few years: since around 1990, there’s been a focus on statistical methods, and researchers in AI (specifically those working in language-related problems) have increasingly realized that a hybrid between linguistic/symbolic and statistical methods is necessary. As always in AI, there’s a trade-off between accuracy (where linguistics comes in) and efficiency (where statistics comes in). The engineering problem then is to find a happy medium, and since natural language processing has led to a ‘bottleneck’ of sorts with only statistics as a basis, more linguistic methods are necessary to engineer systems that can do higher-level work (with efficiency as an important criterion, of course).</p>
<p>MIT used to be the top school - many of the renowned profs in the field were once at MIT. Unfortunately, now it has only one professor working in the field, as their other big-time prof left for Columbia this year. Its strength in general AI, however, allows it to remain one of the few strong schools in the field. Stanford is solidly at #1 for computational linguistics now.</p>
<p>OP, I sent you a PM. If anyone else who has an interest in this, send me a PM.</p>
1 Like
<p>
</p>
<p>You must have missed Watson on Jeopardy.</p>
<p>Sure MIT is great, but for a “safty” I think this topic is a PHD program in CMU, although its never too early to get it onset in the UG…</p>