Objections to the Anchor/Cluster/Williams House Proposal

This document explains my primary objections to the cluster housing proposal set forth by the CUL. These objections range from reservations about particular details to several fundamental issues with any division of the campus into clusters. I have shared this outline with other students and found that it covers many of their core objections to cluster housing as well. It is my hope that the CUL will take these points into account before rushing into any proposal. These are points that I feel have not been adequately addressed by the CUL—in fact, I believe that most of them can not be answered within the framework of the proposed cluster system. If the proposal goes forward on the 28^th without regard to these points, it is my hope that the CUL will remain open and responsive to these points during the one-year gap between proposal and implementation.

1) Divisions. This objection is a fundamental one: the Williams campus simply should not be divided into clusters. The campus does not support equitable clusters, and the student body is too small. The campus is geographically tiny anyways.

a) Percieved boundaries. The Record article on Middlebury shows that despite Midd administrators’ insistence that ‘there are no walls or boundaries’ between clusters, students perceive boundaries—and that is what counts most, because students will act on their perceptions. Sure, nothing stops a student from making friends with someone in another cluster, but they are certainly prevented from living with them. They are certainly prevented from choosing their dorm based on house location—a major factor in current students’ decisions—or character.

b) House identities stunted by geography. Currently, Williams students can choose from houses with their own identities. Clusters are too small for houses to develop such identities, with only a few options available to each class year—because cluster draws are by seniority, dorms will still be picked in order of house desirability. In some clusters, there are perhaps three houses available for a class year; while in some, there is only one. Take the Berkshire Quad cluster: Prospect will house sophomores and some juniors, Currier will house juniors and some seniors, Fitch will house seniors. The divisions created by cluster boundaries do not leave students any sort of elbow room or margin for error.

2) Dorm equitability. Inter-cluster dorm equitability is an issue that the CUL recognizes as vital to the survival of housing clusters, however it has not succeeded in drawing any equitable cluster boundaries. Given the limitations imposed by the extreme variation in dorms on this campus and their geographically spread locations, this is a nearly impossible if not completely impossible task. With the constraint of geographic unity, equitable clusters are impossible. The CUL has failed to accommodate students’ opinions that some dorms are inherently more desirable than others, based both on their physical characteristics (ie, Chadbourne, Perry, Dodd) and locations (ie, West for sciences, Greylock for theater majors).

a) Disadvantaged clusters as drawn. The Berkshire Quad dorms are not inherently physically desirable (except perhaps to art majors). Currier is the only planned anchor house that is not currently picked early in the rising senior room draw—there is a reason for this. The Berkshire Quad cluster is severely disadvantaged. Compare to the Wood cluster, containing four incredibly nice row houses along with junior/sophomore housing comparable to all the Berkshire Quad houses. (Note that current desirability for Berkshire Quad dorms is based largely on the established social character of those houses. They really are not the nicest houses on campus.)

b) Geographic spread of desirable and undesirable dorms. Several of the most desirable and most undesirable dorms share geographically exclusive areas of campus. The Berkshire Quad is a concentration of less appealing housing, while the high-quality row houses are on the other end of campus. Also consider the major concentration of desirable sophomore-level housing that is Mission Park. The layout of the Williams campus and the spread in quality of these dorms are not conducive to cluster planning. In fact, they make equitable clusters realistically impossible. Even if the CUL was to advocate renovating a large percentage of the undesirable dorms, it would have to consider several key issues first:

i) Relative chronology. It is imperative to cluster success that all clusters be equal; thus any and all renovations would have to occur before implementation of a cluster housing system. Otherwise, some clusters will end up inherently undesirable compared to others for an indefinite transition time.

ii) Concentrations of undesirables. Both Currier and Fitch would have to undergo significant renovations at once, putting the Berkshire Quad cluster at a severe disadvantage while construction was in progress. Social events in these clusters would be displaced, and quality of living would decrease.

iii) Middlebury. The CUL claims incomplete renovations as a major factor in Middlebury’s Commons’ failure in students’ eyes. We do not want to emulate another school’s failures along with its successes (or, especially, without its successes).

3) Entry affiliations. The manner in which entries are affiliated with clusters is a point of contention with many students. Students objected to the forced diversity of clusters composed of entries, and they objected to the extension of the first-year community to encompass the entire campus. It is my firm belief that entries ought not to be affiliated in any way with any cluster. This affiliation—even ‘social affiliation’—may well be perceived as a limitation by students—and it is their perceptions that they will act on, not the perceptions the CUL tells them they should have (as in insisting, ‘there are no walls between clusters’).

a) No history of first-year affiliation. Quite simply, entries have never been and should never be associated with upperclass housing. There is no basis for this. This is perhaps the single most radical aspect of the anchor house proposal.

b) Separate freshman community. Incoming freshmen are not ready for, and wouldn’t want, a sudden introduction into the entire campus community. Entries provide a transition to college life as a separate community for first-year students. First-year students bond extensively with other first-year students through entry and inter-entry activities, going out of their way to meet other students (particularly those doing the same) during First Days, and sharing the common new experience of a first year at college. The promised freshman experience is a big part of the reason why many students choose Williams. This should not be interfered with. Sophomores and older students necessarily share a ‘more experienced’ view of college life, and are thus somewhat distant from first-year students. The JA program is an excellent way to foster inter-class interaction and allow dedicated upperclassmen to mentor and provide advice to freshmen. This responsibility should not be pressed onto the shoulders of all upperclassmen by expecting cluster members to provide for their ‘socially affiliated’ first-year students in any sort of capacity. Not everyone wants to be a JA. Perhaps a sophomore-freshman one-on-one affiliation would help both the first-to-second-year transition as well as the sophomore experience—but do not expect an entire cluster to take a vested interest in entries that are listed somewhere as ‘social affiliates.’

c) JAs. Junior Advisors must already throw off (or, at least, set aside) remaining entry identity from their first year and replace it with entry identity towards the entry they are now responsible for—and an integral part of. The same will necessarily be true of JAs’ clusters and their new entry’s ‘social affiliation.’ Thus JAs will be a group of students bred to have very little cluster loyalty, or to have very changeable loyalty. This is not desirable—adapting to a new entry is one thing, but it is entirely another to expect a student to spend their first year in an assigned entry, their second year in a randomly chosen cluster, abandon that cluster after a single year of membership for a new entry and a likely new ‘social affiliation,’ and then either re-insert into the cluster they have less loyalty to than their new entry or use the opt-out—giving these students a different set of loyalties for each year they spend at Williams. This proposal takes the rock-solid core of one of Williams’ greatest pillars and demands that it be extremely flexible and changeable in its identity.

d) Vestige of ‘version 1’ proposal. The ‘version 1’ proposal included entries in clusters for two purposes: First, to establish a ‘four-year community;’ second, to ensure cluster diversity and prevent ‘special-interest’ housing. Neither of these purposes is useful any longer. The means towards the diversity goal has been shifted to random assignment of rising sophomore picks. A ‘four-year community’ has never existed at Williams, entries have always been separate entities, and the CUL has already moved from cluster affiliation to ‘social affiliation,’ a much more vague and diffuse link between clusters and entries. This link is being maintained as a remnant of the ‘version 1’ proposal, and ought to be done away with entirely.

4) Sophomore ‘Mission experience.’ The CUL has sorely underestimated the improvement the Mission Park renovations have had on sophomore life. They have also underestimated the effect on this improvement of splitting Mission Park up amongst clusters in the ways proposed. Living together as a class in Mission Park—or hanging out there with other class members if a student doesn’t actually live there—is a huge part of the Williams sophomore year experience. This is a boon rather than a curse, despite the appearance of decreasing inter-class interaction. The cost of increased class-year interaction enforced by sophomore-junior residential mixing is too high in terms of sacrificed sophomore year enjoyment and class-year identity. The CUL further fails to recognize that Mission Park functions much more as one giant house than four smaller-but-connected ones. This arrangement is not conducive to splitting between clusters, and provides a huge obstacle to cluster equitability. The classes of 2006 and 2007 ought to be surveyed about the new Mission experience before the CUL throws its wrench into the works.

a) Already-improved sophomore experience. Certainly renovating a building does not address issues like the lack of sophomore advising—but then, neither does picking and choosing the houses sophomores will be able to select in the future. The Mission renovations have led to increased sophomore sociability in much better, brighter, and more open common spaces; increased intra-Mission socializing in both large formal events and suite-to-suite gatherings; inter-suite mixing since Mission suites are both large and in close proximity (if not actual physical contact) with one another; the best dining experience on campus; increased sophomore class identity. Previous class years living in Mission have complained that it is a minor social taboo to walk through another Mission suite—this is not true any more. Sophomore year in Mission Park has become something to look forward to.

b) Concentration. Currently, Mission Park is the place to live for sophomores. Rising sophomores who, by chance and chance alone, cannot select Mission Park end up in doubles scattered among the ‘desirable senior housing.’ Sometimes these unfortunate sophomores bond with their houses, but sometimes they don’t—and often because the senior HC’s focus activities on the about-to-graduate seniors. These sophomores tend to go to Mission to hang out with the people that they can and will bond with. Mission provides them with an incentive to cross campus to be with their friends, as opposed to a system where sophomores have been distributed evenly in five different places—that would result in even more sophomore insularity, not less, as sophomores hang out with the sophomores near them and not with those across campus. Current sophomores across campus are as much a part of the Mission experience as those living in Mission Park itself.

c) Removal of Mission experience as an option for 3/5 of the College. An unfortunate feature of the ‘version 1’ proposal was that it removed any and all chance to experience Mission Park from two out of six clusters. The current plan not only dilutes the Mission experience (lowering incentive to cross campus and hang out in Mission), but prevents a full three out of five clusters from ever experiencing it. By arbitrary random assignment, 60% of Williams College rising sophomores will be told that they can never have the advantages listed in item (a) experienced by two (now three) years of Williams sophomores. Students should not be limited in this way—especially when the limitation takes away something incredibly good. In this instance, the CUL is hardly acting under its mandate to improve and build on what already exists. (This is probably because they still insist that ‘improvement’ is directly marked by ‘class-year interaction.’ That assumption is not true, and the measures of class-year interaction should stretch beyond classes living in close proximity—there already is substantial class-year interaction in academics and intramurals, not to mention just plain hanging out with other classes.)

5) Suites and ‘serendipity.’ A great feature of free agency is that is haphazardly mixes students within dorms—but still allows students to choose their priorities when choosing a living space. Students form groups for only four, a number less than the average number of rooms in a suite, so suites contain students from more than one pick group. Suites do not simply encompass self-contained groups of a bunch of friends with no desire to leave their floor—there is an element of ‘newness’ to every suite formed. Many students go into the room draw looking at the lists of people who have already made their selection and thinking, ‘I had a class with that person…I met that person a few times, they seemed nice…That person was on my IM Frisbee team…’ They select a room based on where they see people they could get to know, not just where they see people they are already friends with. Over the course of a year, these ‘serendipitous’—yet still selected—relationships can be solidified. For many members of the class of 2006, living in Mission with suitemates who perhaps weren’t already friends, but who students felt they wanted to get to know, helped form some amazing small communities. These ‘serendipitous’ friendships have, in fact, carried over to this year and the junior living arrangements in the Greylock Quad. These suites are extremely well bonded, have great senses of pride in their suite or floor, and they interact with each other—from suite to suite. These are genuine communities. There is absolutely no reason to suggest that communities formed inside clusters would be any more ‘genuine’ than these communities. In fact, many students believe quite the opposite. By limiting pick selection to within a fifth of the school, and by splitting Mission and Greylock in any of the ways proposed, anchor housing hampers the possibility of these communities getting off the ground and succeeding.

6) Opt-out. The opt-out options, in their various guises, were the last remnant of free agency under the ‘version 1’ proposal. Yet they were intentionally made less desirable than within-cluster picks, as if to discourage or reprimand students wishing to take advantage of an ability to choose their living companions. The opt-out was originally envisioned as a means for uncomfortable students to ‘escape’ a cluster, but the very design of the option meant that students had nothing better to escape to. Under the latest proposal revision, the option remains as a way for small groups of friends to remain together, but these groups are unnecessarily limited and their formation is still discouraged. If four, five, or six students find that they want to live together enough to really want to use this option, then why prevent them? As the proposal stands now, the only way for those students to have what they really want is to apply for a co-op; yet there is already a high demand for those spaces and some students see losing their meal plan as a disadvantage. Why is the option to choose a larger group of good cross-cluster friends open to some students (those entering the co-op lottery) but not to others? If five good friends lack enough cluster identity to use the opt-out individually, why prevent them from using the option together and forming genuine bonds among themselves?

a) Undesirability. The limitation to three students opting out makes the option an undesirable one to use. The chances of a student finding two companions with whom they want to spend the rest of their Williams lives is smaller than the chance of them finding four, five, or six people with whom they get along well but would be interested in getting to know better and spending more time with—deepening their genuine bonds. Restricting the size of opt-out groups has the appearance of a punishment against students opting out. If the CUL feels that cluster loyalty will result in few students opting out, then there is no reason to restrict their group size. If the CUL feels that a larger opt-out size would dramatically increase the number of students wishing to opt out, then there is a systematic problem with the anchor system. Additionally, the implication that students will ‘apply’ to opt out could be interpreted as a means to discourage any students from using the option. By requiring students to give a reason for wanting to opt out, there is the possibility that a dean somewhere will decide that applications stating, ‘I hate my clustermates,’ ‘my cluster dorms suck,’ ‘the people in cluster X are so much cooler than mine, please put me in cluster X,’ or anything of the sort could decide that the reason isn’t good enough and reject the application. An application process itself might be too involved to allow students who want to use the option to feel it is accessible.

b) Inflexibility to character change. What should happen if a student finds that they can’t stand living with their 5 companions from freshman year after all? What if he lives on Mills first floor and decides the five students in Dennett 1 sharing a common room with him are much cooler? What if the ‘cooler’ five live in another cluster? What if a student uses the opt-out alone to escape from an undesirable cluster and ends up in an equally undesirable cluster? What if a student who picked in alone as a rising sophomore finds a larger (again, say five) group of students that year, or later, that they would like to live with the next year? In this last case especially, a lone student who finally comes out of their niche is not given the opportunity to do so. These situations happen commonly among students, and free agency provides them with a means to solve their problems and seek out a social setting in which they will flourish should their character or desires change from year to year.

c) Vestige of ‘version 1’ proposal. Why is there a need to discourage students from opting to reaffiliate with 4, 5, or 6 of their closest friends if that is what they really want to do? Why is the opt-out size half the size (or less) of an intra-cluster pick when the opt-out pool of students is five times as large? There is no need for this limit except that it is a leftover from the old entries-lead-to-clusters proposal, when it was envisioned as a means to live with one or two ‘soul mate(s).’ Now that entries are no longer cluster members, and there are only three years to shape true cluster identity, an opt-out means that a student lacks cluster loyalty much more than opt-outs in the ‘version 1’ proposal. Students without cluster loyalty anyways should not be discouraged from or punished for finding those with whom they would identify.

7) Economics. The current system is defined by two forces: chance and free-market economics (supply and demand). In the long run, chance will affect everyone equally, thus any steady state reached in the Williams free-agent system is independent of chance. Free-market economics has long proven itself to be the most efficient way to allocate resources to a group when compared to engineered or controlled systems. In fact, the Williams housing system currently is in a steady-state, equilibrium condition. The rigorous definition of equilibrium is a position to which the system will be restored if it is displaced slightly from that position. We saw two displacements over the past two years: the Mission renovation and the Baxter demolition. Coming up on the horizon are more displacements: the new student center completion, Prospect renovations, library renovations (to a lesser extent). The past two years have demonstrated that the Williams housing situation has not changed significantly—despite the Mission renovation and Baxter demolition, sophomores still live mostly in Mission, juniors mostly in Greylock or the Berkshire Quad, and seniors in row houses or the Berkshire Quad. The anchor housing proposals seek to deliberately twist this equilibrium and forcibly shape a new one (most noticeably in Mission Park); one that does not match with students’ needs or demands as expressed by the current equilibrium. Clusters should fit the free agency equilibrium, not try to force it to adjust to them.

8) Manner of proposal. This point encompasses a set of serious objections to the CUL’s manner of data-gathering and presentation rather than the proposal itself. Despite CUL chair Will Dudley’s insistence that he’s not a believer in ‘spirit from the top’ in the case of HCs or party planning, the anchor house proposal itself looks like an imposition of ‘spirit from the top.’ The CUL asserts that establishing anchor housing would result in some sort of spontaneous wellspring of cluster pride, cluster spirit, and everything good that anchor housing could possibly bring. However, this carries more of an appearance of a top-down executive decision on students’ residential—and social—structure without consultation of those most affected.

a) Timescale. The proposal has been in the works for four years. However, it has not occupied the campus as an issue for any longer than three weeks. Current seniors vaguely remember the idea of anchor housing failing in 2002, and only now does the issue appear again. The ‘version 1’ proposal burst onto the scene in mid-January, students were (1) informed that the proposal would be finalized and presented by late February and (2) given the impression that the administration supports the proposal and would essentially rubber-stamp whatever crossed the President’s desk. The mostly negative immediate student reaction induced the CUL to make alterations to some major details of the proposal, but they stick to their initial self-imposed deadline despite changing the proposal within the space of a day or two. This willingness to alter the plan is at odds with the insistence that four years of work lie behind it and that it should be trusted for its four years of continuous effort.

b) Student information and input. Student input was not visibly sought before the proposal appeared in the campus consciousness in mid January. Since then, CUL strategy has focused more on selling the proposal to students and convincing them that they would not lose anything rather than seeking additional student input and advice on both the details and the fundamentals of the plan. Now that the implementation has been delayed, the CUL ought to carry out student opinion-gathering with as much transparency as it possibly can. Until now, CUL operation and data-gathering has been an extremely opaque process. The CUL also needs to remain open to fundamental objections to the plan rather than accepting anchor housing as an undisputable premise and recognizing only student input on details.

i) Non-representative student committee members. As a corollary to insufficient student input gathered, the student members of the CUL are not representative of the student body as a whole. They were appointed from a self-selecting group: only those students who see a problem to be solved applied. Only those students who believe that a change of housing system will spontaneously result in improved social and academic life are advocating this proposal. Furthermore, comments from some student CUL members suggest that they see very little suite identity across campus. These are not majority opinions among the student body.

c) Faulty data. The CUL often cites a 2002 campus survey, the 13% statistic, comparisons to other schools, and ‘conversations’ as the primary sources of data in both developing and supporting the anchor house proposal. However, none of these carries substantial weight—especially after the CUL revised the proposal several times within a span of two weeks, suggesting that their prior attempts to gather data were incomplete.

i) 2002 survey. This survey was conducted nearly a full ‘institutional lifetime’ ago—three years. Thus, current seniors vaguely recall a survey, but other students have not been systematically polled. This survey also predates (1) the Baxter demolition, (2) move to four-person pick groups, (3) Prospect renovation, and, more importantly, (4) Mission Park renovation. The College is dealing with an entirely new group of students now, a group with new experiences and priorities; a group that has not been surveyed.

ii) 13%. This piece of ‘data’ cannot be evaluated without more background. We got a glimpse of that with the knowledge that the percent of ‘very satisfied’ students was 25% in 1999—but this works against the CUL, suggesting that the 2002 room draw changes, the only changes to occur within that timespan, are responsible. Regardless, we need to see this ‘very satisfied’ figure over more than two years—and we need to see the distribution (in other words, while ‘very satisfied’ students declined, did even more ‘very dissatisfied’ students become ‘satisfied?’) and comparable figures for other school with and without cluster-like systems.

iii) Comparisons to other schools. The comparison to Yale’s residential college system is flawed, as it is designed for a larger university, and designed specifically for the particular ‘small-college’ dorm setup, which Williams lacks. Yale’s colleges also include such dubious features as the ability to pick the same college as a parent who attended Yale. Middlebury implemented a clusterlike system, but students there apparently view their Commons as social engineering and silly. They see no cluster pride. That Middlebury Commons failed because Midd lacks the dorm structure originally planned is an unacceptable rebuttal because (1) their system has been in place and failing for a number of years now and (2) other schools with worse dorms have clusters that the CUL claims work well (Bowdoin). There also need to be comparisons with schools that have free agent systems in place.

iv) Conversations. Why is the CUL now neglecting the students who are ‘conversationally’ disputing the merit of anchor housing in favor of an approach defending anchor housing to all students? Why did the opinion-gathering stop? And why, if student opinion is important enough to merit direct conversations, do we not have a campus-wide survey or poll beyond the planned ‘renovation priority’ survey?