![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
trocheeThe title here comes from this discussion of a peculiar snowclone
Today I had a sense of visiting The Time Machine's future world, 800,000 years in the future. I went back and forth between two lands.
I woke up in a dark room in the messy, grungy lab, having spent the night bashing Perl scripts and gigabytes of data into shape to spit out a small table of statistics. I read my email, made some coffee and staggered off to class.
In my first class, we are studying semantics: an abstract philosophy of language, where "but it's not a very elegant solution, and it's difficult to hold it in mind, so we should reject it" is a valid line of argument. If a formalism doesn't fit on a blackboard, we wave our hands and say "I don't want to talk about these messy details, but you can see that it will work." Sometimes, this debate looks like an "angels-on-the-head-of-a-pin", and though I enjoy it, I can't help but think of the happily-oblivious Eloi.
Now it's back to Morlock land. In the next, it's math and Bayesian learning -- teaching machines to think, primitively but powerfully. Formalisms matter, numbers crunch, all the Greek is meaningful but occasionally obscure. These are the people for whom the scrolling transparency was designed -- the formulae just keep going and don't stop. Then it's lunch with a crew of engineering grad students, amusingly and banteringly practical people.
After lunch, back to the lab -- all still as Morlocked as it was when I left it. I spend an hour and a half explaining the grungy, unpleasant decisions in the software I've written; I have the distinct sense of being the 21st century analog of a Bunsen-burner-scarred, glass-beaker wielding industrial chemist, complete with the soot burns. A lab meeting underscores the hyper-practical nature of this group -- it's all about NFS [network filesystem problems]; the mood is briefly lightened by one of the PIs calling them NSF problems [National Science Foundation, one of our funders!].
I dash off from there to the linguistics colloquium, and I'm suddenly back in Eloi land -- more semantics, arguing about whether "Bob briefed Alice before every meeting" happens at one time or many times.
It seems to come down to something I've posted about before: a sizable number of the linguists here seem willing to ask and explore questions that only have to do with things that you can sit down and "just think about". The engineers, and certain linguists, actually want to know how widespread is this generalization or can I use this generalization to do something useful.
But this latter attitude is really quite Morlock-ish. I may sleep in the lab on occasion, I may have to learn the contemporary equivalents of the Morlock monkeywrench. It's so odd to see these two worlds existing so closely together, yet with so little communication between them. It's grim to think of this as a Morlock/Eloi distinction, but I wonder if it's accurate:
I hope we engineers won't start kidnapping linguists to eat them any time soon. Or maybe they have, and they're just fattening me up.
Today I had a sense of visiting The Time Machine's future world, 800,000 years in the future. I went back and forth between two lands.
I woke up in a dark room in the messy, grungy lab, having spent the night bashing Perl scripts and gigabytes of data into shape to spit out a small table of statistics. I read my email, made some coffee and staggered off to class.
In my first class, we are studying semantics: an abstract philosophy of language, where "but it's not a very elegant solution, and it's difficult to hold it in mind, so we should reject it" is a valid line of argument. If a formalism doesn't fit on a blackboard, we wave our hands and say "I don't want to talk about these messy details, but you can see that it will work." Sometimes, this debate looks like an "angels-on-the-head-of-a-pin", and though I enjoy it, I can't help but think of the happily-oblivious Eloi.
Now it's back to Morlock land. In the next, it's math and Bayesian learning -- teaching machines to think, primitively but powerfully. Formalisms matter, numbers crunch, all the Greek is meaningful but occasionally obscure. These are the people for whom the scrolling transparency was designed -- the formulae just keep going and don't stop. Then it's lunch with a crew of engineering grad students, amusingly and banteringly practical people.
After lunch, back to the lab -- all still as Morlocked as it was when I left it. I spend an hour and a half explaining the grungy, unpleasant decisions in the software I've written; I have the distinct sense of being the 21st century analog of a Bunsen-burner-scarred, glass-beaker wielding industrial chemist, complete with the soot burns. A lab meeting underscores the hyper-practical nature of this group -- it's all about NFS [network filesystem problems]; the mood is briefly lightened by one of the PIs calling them NSF problems [National Science Foundation, one of our funders!].
I dash off from there to the linguistics colloquium, and I'm suddenly back in Eloi land -- more semantics, arguing about whether "Bob briefed Alice before every meeting" happens at one time or many times.
It seems to come down to something I've posted about before: a sizable number of the linguists here seem willing to ask and explore questions that only have to do with things that you can sit down and "just think about". The engineers, and certain linguists, actually want to know how widespread is this generalization or can I use this generalization to do something useful.
But this latter attitude is really quite Morlock-ish. I may sleep in the lab on occasion, I may have to learn the contemporary equivalents of the Morlock monkeywrench. It's so odd to see these two worlds existing so closely together, yet with so little communication between them. It's grim to think of this as a Morlock/Eloi distinction, but I wonder if it's accurate:
The Eloi ... had decayed to a mere beautiful futility. They still possessed the earth on sufferance: since the Morlocks, subterranean for innumerable generations, had come at last to find the daylit surface intolerable. And the Morlocks made their garments, I inferred, and maintained them in their habitual needs, perhaps through the survival of an old habit of service. ... But, clearly, the old order was already in part reversed. The Nemesis of the delicate ones was creeping on apace.Problem is, the engineering solutions work, ugly though they are, and they have funding for that reason.
I hope we engineers won't start kidnapping linguists to eat them any time soon. Or maybe they have, and they're just fattening me up.
![[identity profile]](https://www.dreamwidth.org/img/silk/identity/openid.png){kind=small}
no subject
Date: 2004-02-01 05:46 am (UTC)I'm a dual major, Math and Linguistics (UNC-CH), and i'm a unix admin in the physics department. I spend about two-thirds of my day in rigorous, anal, detail-oriented systems, and then step into happy handwaving with my linguistics classes. It's depressing.
For example, i'm taking the introduction to syntax course. Our first assignment was to prove A as a seperate lexical category from D, N, and V given a little dataset. I turned in a couple pages of logic, most people turned in a single sheet of paragraphs. He handed mine back, ungraded, with the comment "Please re-do this, writing your respons in clear and prsuasive prose." Most frustrating.
The only justification i can see for the feeble-minded attitude linguists favor is that there aren't enough Really Smart people to handle all that formalism. Also, we Morlocks still don't have tools quite up to the task of abstracting out to the hand-waving level.
That said, i'm keen on starting the Anal Linguist Underground. It's the ALU, not an ALU...