Architecures of Knowledge

I am occasionally asked whether or not it's a good idea to go to grad school. As a matter of objective facts, including salary and career prospects, it's hard to conclude that it is, or could be. If you go to graduate school in the social sciences or the humanities, you will be earning a very small amount of money for a (potentially) very long period of time, in order to become one of the 200+ who will be applying for whatever jobs there might be; four or five years of bouncing around temporary gigs and assembling what used to be a tenure-worthy cv will get your foot in the door. The sciences are better, to the extent that the funding is better and the programs (usually) are shorter. But the successful track to running a lab in, say, the biological sciences, requires at least one and more typically two post-docs, each of three to five years, much of which time will be devoted to writing grants and writing progress reports. Also there are fewer jobs than there were. Also the federal funding which provides for most basic science research is in jeopardy. Watching my friends in the sciences collectively realize this has been disciplinary schadenfreude only matched by watching them write their dissertations, which they initially cannot believe would take more than a couple weeks to pull together.

If I am asked, instead, whether I regret going to grad school, the answer has to be 'no': I met a number of interesting people, including my wife, got the chance to spend time reading, thinking, and writing about much of the stuff most interesting to me, and have an endless supply of anecdotes from my various experiences with teaching and administration.

The actual skill I gained that I think most worthwhile is in the architecture of knowledge. Research in the social sciences and the humanities is a matter of knowing how to phrase a question, and how to look for the answer. This gives me the ability to do astonishing things, like search for the universe of books and articles on a topic and identify which ones might be helpful; the ability to confirm that suspicion on approximately five minutes of reading, sometimes less; and find and digest what's useful in those results in a very limited amount of time. As it turns out, this is a useful skill in all parts of life: how do I know if this cookbook will be any good before I order it? How do I determine which of these potential pieces of furniture is the best value? How do I determine the best way to get from A to B in Durham's maze of streets? Figuring out where you are in a body of knowledge, analyzing what you find, synthesizing it into useable form: all of these make life a lot easier.

Explaining this to, say, undergraduates is quite difficult. On average, they're content to google their way to the answers to questions, accept wikipedia as a neutral and authoritative source, do the minimum amount of research for the assignment. I used to intentionally assign students to find more sources for a paper than were required for the final product, because you could see people prepared to ride or die with the first three, or five, or however many things they found. Discriminations of quality and usefulness were generally lost on them.

Which brings me to this commercial for google:



I think it's supposed to indicate to us how amazing google is at recognizing and answering our questions. A closer look at the result:

Google's great wisdom points us to an answer given by, italics required, mobile.omgfacts.com. Now, maybe omgfacts is a really great, reputable website; maybe their answer is correct. None of that should give any confidence at all that google will return a decent search result. If you're leaving the results of your search for an answer to google, or Amazon, or Facebook, or the local library search engine, you will get the results that someone else thinks best, and that someone is designing a search algorithm with goals that are unknown but not necessarily your own. If you don't know how to tell when an answer is wrong, how will you ever know that it's wrong?