THOUSANDS OF ALUMNI converged on Cambridge this week to stay for a few days in the undergraduate houses they remember so fondly.

The Class of '92 will return in 25 years to do the same. Dunster and Adams and Eliot and the other residential houses may still stand, but they will be dorms like those at any other college--buildings in which students live-and nothing more.

The administration so far has refused to listen to undergraduates who with to retain the character of Harvard's houses. Only alumni can convince the Harvard administration to stop the agent which is destroying the houses' character--non-ordered choice.

NON-ORDERED CHOICE first infected the houses in the spring of 1990, after two years of bitter struggle over the apparatus by which first-year students would choose their houses. It is a halfbaked compromise between complete randomization and the old ordered-choice lottery instituted in the 1970s.

Under the ordered-choice regime, first-year students in the College received lottery numbers and then selected, in order of preference, three houses they wished to live in.

Those students whose lottery numbers were too high to allow them entry into their first-choice house received their second (or third) choice, or were randomized into houses which had yet to be filled.

Ordered choice certainly was an improvement over older ways of choosing who lived in which house. Those in the Class of '42 no doubt remember that house masters often used to choose their student population through personal interviews.

This non-system allowed masters to choose students on the strength of such arbitrary attributes as a student's social standing and family background. And it fostered an exclusionary mystique in many of the houses.

Engendering a distinct character in each house, however, was an important goal of the house system's creators. The houses were modelled after the Cambridge and Oxford colleges.

Harvard houses, to be true to the spirit of the house system, should be the intellectual, social and cultural centers of College life. If they become mere microcosms of the College community as a whole, as some administrators seem to think they should, there is a slight chance that they will remain so. What they won't be, however, are distinct communities.

IT WAS BECAUSE the houses succeeded in remaining distinct communities under ordered choice that College administrators proposed the complete randomization of house assignments.

In a mad rush to diversify the houses for egalitarian purposes, Dean of the College L. Fred Jewett '57 and Associate Dean for the House System Thomas A. Dingman '67 decided that every Harvard house must be different only in name and architecture.

In the spring of 1990, Jewett, with the approval of then-President Derek C. Bok, instituted a compromise plan proposed originally by a student group, the Committee Against Randomization, which also presented Jewett a petition signed by over 1100 members of the Class of '93 criticizing his randomization proposal. Non-ordered choice began its infection.

Non-ordered choice is supposed to be a trial program which attempts to diversify house populations without destroying the unique character of houses. Its trial period is over this year, and next year complete randomization may ensue. If it does, Harvard's houses will become nothing more than bland dormitories, the student population will become fragmented and the cultural life of the College will suffer.

Proponents of diversity disparage house "stereotypes," saying that the previously distinct character of houses such as Adams causes students to narrow their outlooks.