What does it take to be truly great lawyer? Or an outstanding psychologist? Or historian, or computer scientist, or sociologist? Is it an encyclopedic memory of all the details in one's field, which could be spit back at will? Knowledge of facts may help you get a job, but would it make you truly great? Most people would say no, as even the most mediocre scholar could look up the details of his trade in seconds. And as Albert Einstein has said, "Imagination is more important than knowledge."

A historian, for example could know the complete history of every officer above the rank of major that fought in the civil War; but aside from impressing people at cocktail parties, this historian could easily have an entirely unremarkable career as a Civil War scholar unless he could apply this knowledge to come up with an interesting evaluation of the events he is studying.

If Harvard truly believes that it is molding great minds who will contribute to world scholarship--and no one can deny that the Harvard mystique suggests great things for all its graduates--then a system of education must be developed in which the focus is not the collection of knowledge, but the ability to interpret it. This aim would best be served by revamping current testing procedures to include more open-book testing.

Open-book testing should go a long way towards fostering truly great minds. Students would be forced to go beyond the mere pretense of knowledge, in the form of spit-back ability, and develop to a point where their understanding is genuine.

Some critics of this system would claim that open-book testing would lead to a situation in which students don't need to study at all. This would hardly be the case, as understanding of material would be impossible without a certain degree of familiarity. Others would claim that a certain body a of knowledge is necessary before qualified conclusion. before qualified conclusions could be drawn could be drawn on any topic.

"For some classes is seems appropriate," says Jay Phelan, head Teaching Fellow for Science B-29. But for science B-29, he insists "a lot of the course is under standing the diversity of the biology of animals...If you need to rely on your books, you probably don't understand the material as well as you should,."

It appears that this is the reasoning behind the testing policies of most Harvard course, as open-book tests are indeed rare. Clearly it is the case that a vital goal of many course is the building of a body of knowledge making analyses possible, but open-book testing would not undermine this objective.

It is unreasonable to assume that a student given access to a textbook during a test on the Shavante Indians would proceed to read about their rites of passage for the first time during the hour allotted to him to take the exam. The student would still need to have a working knowledge of the topic, in addition to knowing where to find useful details. Coupled with finely-tuned analytical skills, these are the only tools that a great anthropologist would need.

Some doubters would all still claim that this system would be impractical. In theory this plan might work, they say, but put into effect it would lead to nothing less than academic anarchy. Well, those people should take a look at the testing policy of Computer Science 50, in which students this year progressed to material for the next level course--despite being subjected to the perils of open-book testing.

CS50 students this fall were encouraged to bring to tests any of the course textbooks that they felt would be useful. Even so, the course gave rise to students who could write a variety of programs quite different from any of those they had seen in class. How was this possible?

By allowing the use of textbooks, Assistant Professor Margo I. Seltzer rendered a test of the minute details of C programming pointless. The only thing that could be tested remained the students' ability to apply the knowledge that they had acquired. They still had to have a very complete working knowledge of the language, and they could use the textbook to recall fleeting little bits of information that really shouldn't have been necessary to memorize.

"What we try to do is to simulate a real world environment as closely as you could in an exam." says Seltzer. She went on to explain that programmers wouldn't be asked to come up with algorithms in the real world without algorithms in the real world without the use of textbooks or similar resources, so it would not make sense to force students to do so on a test. The open-book policy will be continued in CS50 next year.

Let's face it: even if you believe that, mass memorization before tests rarely gives rise to long-term retaining of information. In classes like CS50, the fundamentals are thoroughly understood, while the details that would be forgotten anyway are looked, up, as they should be.

Who know, maybe open-book testing would even lead to a better recollection of facts. As I learned in psychology, information that is well-understood is much more easily encoded into long-term memory.

Ironically, in this psychology class much like my others, tests measure the ability of students to cram as many isolated facts as possible into their heads, only to forget them as soon as they are written down, psychologist is supposed to be fostered in this environment, I still don't understand.