Sherry Turkle and my Discontent
A couple of weeks ago I read Sherry Turkle's Simulation and it's Discontents. I read it because simulation is most of what I do for work, because I think a lot about how simulations succeed and fail, and should probably think harder. The first half consists of a pair of ethnogrophies of MIT professors, first in the 1980's, and the second around 2000--2005.
I was particularly intrigued by the explanation one physicist gave of the importance of experiment. The intro physics classes I've taken have covered subject matter that we've understood for at least a century. In that context, it's very hard to maintain a critical stance towards accepted physical models ('laws'). So it's great to hear a professor talking about the difference between using science and doing science. The student who asserted that physics is superior (to engineering) because it deals with the universal, without the encumberment of humanity, was also fabulous---I know physicists think this, but they so rarely come out and say it.
Turkle's description of the attitudes towards CAD in the 80's helped me to understand the rhetoric and departmental politics that I witnessed during my undergrad. The second ethnogrophy overlapped my time at MIT, and closely matched what I experienced.
As with several other ethnographies that I have picked up recently, I found Turkle's heavy filtering of evidence to support her thesis frustrating. (Unlike the other two, I actually finished this one.) It's obvious that she believes we have embraced new technology too readily, with too little regard for what is lost. (I gather her latest book is even more explicit in this regard.) Strangely, she seems to believe that the reader has not heard any of these criticisms of simulations. Every professor I had at MIT, and every experienced engineer I've worked with since, has expressed strong reservations about the use of computer models. Every one has encouraged me to draw by hand, or use back of the envelope calculations, either instead of or to supplement simulation.
I am disappointed that Simulation did not record any advice from practitioners on deciding whether simulation is useful, and how to combine it with other scientific methods. These are decisions I make constantly, and I'd much rather learn from others' mistakes than from my own. More on that in my next post. This entry was originally posted at http://bergey.dreamwidth.org/1392.html. Please comment there using OpenID.
I was particularly intrigued by the explanation one physicist gave of the importance of experiment. The intro physics classes I've taken have covered subject matter that we've understood for at least a century. In that context, it's very hard to maintain a critical stance towards accepted physical models ('laws'). So it's great to hear a professor talking about the difference between using science and doing science. The student who asserted that physics is superior (to engineering) because it deals with the universal, without the encumberment of humanity, was also fabulous---I know physicists think this, but they so rarely come out and say it.
Turkle's description of the attitudes towards CAD in the 80's helped me to understand the rhetoric and departmental politics that I witnessed during my undergrad. The second ethnogrophy overlapped my time at MIT, and closely matched what I experienced.
As with several other ethnographies that I have picked up recently, I found Turkle's heavy filtering of evidence to support her thesis frustrating. (Unlike the other two, I actually finished this one.) It's obvious that she believes we have embraced new technology too readily, with too little regard for what is lost. (I gather her latest book is even more explicit in this regard.) Strangely, she seems to believe that the reader has not heard any of these criticisms of simulations. Every professor I had at MIT, and every experienced engineer I've worked with since, has expressed strong reservations about the use of computer models. Every one has encouraged me to draw by hand, or use back of the envelope calculations, either instead of or to supplement simulation.
I am disappointed that Simulation did not record any advice from practitioners on deciding whether simulation is useful, and how to combine it with other scientific methods. These are decisions I make constantly, and I'd much rather learn from others' mistakes than from my own. More on that in my next post. This entry was originally posted at http://bergey.dreamwidth.org/1392.html. Please comment there using OpenID.