ARCHITECTURE PROFESSOR EXPLORES
INFLUENCE OF SPACE ON RELIGIOUS RITUALS
When he is not teaching Division of International Programs Abroad students at the Florence Center, Professor Alexander Fernandez ’94, G’97 is exploring an ancient world.
For the past two years, Fernandez has conducted an in-depth study of Carthusian monasteries in Italy that grew out of a studio project he worked on as a graduate student. His research tracks the development of the Carthusians, a Catholic contemplative religious order, and how their use of architecture responds to their religious ideals by providing a balance between the individual and the communal. “I’m interested in the relationship between ritual and architecture and how one informs the other,” Fernandez says. “The research draws attention to the monk’s daily rituals and customs in relationship to his cell and his communal domains.”
To better understand this relationship between routine and architecture, Fernandez conducted a translation of the order’s statutes, originally written in French. He also spent a year analyzing two structures, San Lorenzo in Val Di Ema in Galluzzo, and the Certosa di Maria e San Giovanni Battista in Calci. “A textual reading of the statutes and an assessment of the building condition revealed a close relationship among building, space, and function,” Fernandez says. “This suggests architecture was the primary framework for the rituals and customs of the Carthusian order.”
The two well-preserved Italian monasteries, located about 20 miles from each other, provided a unique opportunity to examine this relationship. Both were designed within a 20-year span in the late 1300s, and are of similar scale and design.
With periodic assistance from visiting students, Fernandez conducted interviews with monks, pored over archival data, visited other Carthusian monasteries in Europe, and compiled sketches of the spaces. Translation of the statutes, known as the Consuetudines Anticua, was supported by a grant from the University’s Office of the Vice President for Research and Computing.
Fernandez’s goal is to bring a new perspective to what is already known about the structures. “Publications on the development of most Carthusian monasteries in Europe are extensive,” he says. “The purpose of this study is not to duplicate this printed information, but to use it as support for the various hypotheses and particular readings into the subject of ritual in architecture.”
Since he approached the research from a visual rather than textual perspective, Fernandez placed great emphasis on graphic documentation. “Small spaces and details were drawn with a high degree of precision to represent materials with as much accuracy as possible,” he says.
STUDENTS FIND COMPUTERS DON'T
MAKE THE GRADE LIKE PROFESSORS DO
When students in Professor Robert Van Gulick’s Minds and Machines philosophy class were challenged to write bad essays for a good grade to test a computer grading system, they rose to the occasion. In a psychology essay on childhood attachment, William Campi ’01 wrote: “The parent is the child’s shepherd, and just as the shepherd’s final intent is the slaughter of his brood, parents too intend to consume. The guardian must instill the infant with confidence and a sense of security by securely attaching the child to a radiator, although a ceiling fan might serve as well.” Such silliness earned a grade of 9 points out of a possible 10 from an electronic rater.
The Bad Essay Contest was part of a hands-on research project for students in the course, which addresses philosophical issues concerning artificial intelligence. The students examined the question of whether machines can understand, learn, and think rationally. According to Van Gulick, the idea for the exercise developed from a class discussion about a New York Timesarticle on the Education Testing Service’s use of a computer, called an e-rater, to help grade the essay section of the Graduate Management Exam. The e-rater—which replaces one of two human reviewers grading each essay—is reportedly capable of assessing not only form and grammar, but also content. If there’s a large discrepancy between the e-rater’s grade and the reviewer’s grade, the machine grade is discounted.
Interest in the subject grew when, at the next class meeting, a student presented a Chronicle of Higher Educationarticle that questioned the e-rater’s usefulness. The article also pointed to web sites where students could write sample essays to be graded by the e-rater, allowing them to judge its effectiveness for themselves. “I then designed a contest that rewarded the student who could earn the highest score for the worst essay,” Van Gulick says. Once the students’ essays were scored by the e-rater, Van Gulick presented those earning the highest scores to the class for a reverse rating system in which the worst essays received the highest scores.
While the students didn’t rule out the future value of the e-rater, they did express what Van Gulick calls a “healthy skepticism” about the appropriate use of machines in roles traditionally held by humans. “Students came away from this project with what I would want them to come away with,” Van Gulick says. “They determined machines aren’t yet ready to grade their essays. There is still a lot of work to be done before electronic grading systems can be widely used.”
Ironically, although in one sense the Bad Essay Contest reflected poorly on the intelligence of machines, the experiment itself could not have been executed without using technology. Students relied on the Internet to gather the original articles that sparked the idea for the contest, to receive instantaneous results from the e-rater, and to collect related information that contributed to their understanding of the exercise. “Machines not only helped make this active learning experience possible,” Van Gulick says, “they also helped make the lesson an entertaining and interactive one.”