MIT Faculty Newsletter Vol. XVII No. 1 September/October …web.mit.edu/fnl/vol/171/fnl171.pdf ·...

this issue features an editorial on “Affirming Freedom of Expression at MIT” (page 3),the “Preliminary Position of the Faculty Policy Committee on Faculty Governance” (page 5), our Teach Talk feature (page 8), commentary on undergraduate education, with three articlesbeginning on page 10, a report on how faculty medical benefits change upon retirement (page 16), a piece on the Engineering Systems Division (page 18), our MIT Poetry feature (page 20), an article on academic computing (page 22) and Web accessibility (page 24), and anOpenCourseWare Update (page 25).

MITFacultyNewsletter

Vol. XVII No. 1September/October 2004

WE HAVE PR E PAR E D TH I S ARTI CLE in response to arequest for comments on the management of the MITEndowment. With the increased importance of the Endowmentin the financing of activities at the Institute, this is a wonderfulopportunity to more completely describe the management ofthese assets to the faculty, a core constituency of the Institute. Abetter understanding of how these assets are managed and howdistributions from the Endowment are determined is essential tothe setting of realistic expectations of the potential distributionsfrom these funds in the future, and to further an understandingof the underlying obligations of the Institute to these funds,which are the result of gifts from many generations of alumniand friends of MIT.

In setting out these comments, we have drawn from recentpresentations to the Academic Council and various depart-ments. We are available for more direct discussions with otherdepartments and laboratories to further elaborate on the com-ments presented here, and look forward to the observations ofthe faculty on the management of these assets and the distribu-tion policies described.

WE AR E EXCITE D AN D HOPE FU L over the arrival of a newpresident at MIT. Your experience at other universities and yourknowledge of the biomedical sciences bring new resources intothe Institute’s leadership. A new hand at the helm providesopportunities to make strategic and policy adjustments that maybe needed. We hope you will view the faculty as your colleagues,allies, and advisors.

Opportunities for extensive discussions with faculty will behard to find, but necessary. Our faculty represents an extraordi-nary body of experience, knowledge, and engagement in therelationship of the Institute to both teaching and research mis-sions, to the internal community of students and staff, and to thelarger external community of institutions, corporations, govern-ments, and nations.

Continuing globalization of the economy and the interna-tional sharing of knowledge means our role as a national resourcehas to be adjusted towards our also being an internationalresource. Given the increasing polarization of the U.S. economyinto haves and have-nots, MIT will have to work even harder toavoid serving and training only the children of the haves. We arelooking forward to working with you to achieve this goal.[Editor’s note: President-elect Hockfield has agreed to addressthe faculty in the January/February MIT Faculty Newsletter.]

continued on page 26

http://web.mit.edu/fnl

MassachusettsInstitute ofTechnology

Allan S. Bufferd

The Management of theMIT Endowment

Welcome Aboard President-elect Hockfield!

Vol. XVII No. 1 September/October 2004

2

The MIT FacultyNewsletterEditorial Board

Alice AmsdenUrban Studies and Planning

*Jeanne S. BambergerMusic and Theater Arts

*John BelcherPhysics

*Nazli ChoucriPolitical Science

Erik DemaineElectrical Engineering & Computer Science

Olivier de WeckAeronautics & Astronautics/Engineering Systems

Ernst G. FrankelOcean Engineering

Stephen C. GravesManagement Science and Engineering Systems

*Jean E. JacksonAnthropology

Gordon KaufmanManagement Science and Statistics

Daniel S. KempChemistry

Samuel J. KeyserLinguistics & Philosophy

*Jonathan KingBiology

Stephen J. LippardChemistry

David H. MarksCivil and Environmental Engineering

*Fred MoavenzadehCivil and Environmental Engineering

Ronald PrinnEarth, Atmospheric and Planetary Sciences

David ThorburnLiterature

George VergheseElectrical Engineering and Computer Science

Rosalind H. WilliamsSTS and Writing

Kathryn A. WillmoreVice President and Secretary of the Corporation

David LewisManaging Editor

*Editorial Sub-Committee for this issue

AddressMIT Faculty NewsletterBldg. 11-268Cambridge, MA 02139

Websitehttp://web.mit.edu/fnl

Telephone 617-253-7370Fax 617-253-0458Email [email protected]

Subscriptions$15/year on-campus$20/year off-campus

Photo credits: Page 1, MIT Museum; Page 17, Calvin Campbell

01 The Management of the MIT EndowmentAllan S. Bufferd

01 Welcome Aboard President-elect Hockfield!

Editorial 03 Affirming Freedom of Expression at MIT

04 Teaching this fall? You should know . . .

From The 05 Preliminary Position of the Faculty Policy Faculty Chair Committee on Faculty Governance

Rafael L. Bras

Teach Talk 08 Developing Musical Structures: A Reflective Practicuum Jeanne S. Bamberger

10 Work of the Committee on theUndergraduate Program, 2003 – 2004J. Mark Schuster

12 Task Force on the UndergraduateEducational CommonsRobert J. Silbey

14 Some Reflections on Aspects of theUndergraduate Education PolicyJames B. Orlin

16 Benefit Changes for Faculty Upon RetirementShawn Foley

17 Short Takes

18 Establishing Leadership in the EmergingField of Engineering SystemsDaniel Hastings, Daniel Roos

MIT Poetry 20 Concerto for Erhu and SubwayTunney Lee

22 Spaces, Software, and Services – Supporting Educational Innovation and Sustainability with TechnologyVijay Kumar

24 Web Accessibility:What Faculty Should KnowKathleen Cahill, Edward Barrett

OpenCourseWare 25 What Was it Like Working with OCW?Update Mary P. Rowe

contents

MIT Faculty NewsletterSeptember/October 2004

3

EditorialAffirming Freedom of Expression at MIT

TH E EVE NTS OF 9/11 and the envi-ronment generated in response to 9/11, inparticular the passage of the Patriot Actand the expansion of Homeland Securityapparatus, have generated new dangers tothe maintenance of free inquiry, expres-sion, and speech that is the lifeblood ofgreat universities. In particular, the issueof freedom of expression has emergeddramatically as a result of events at MIT’sgraduation last spring, and we feelstrongly that the right to freedom ofexpression at MIT needs to be affirmedquickly and decisively.

At the June 2004 MIT Commencement,the MIT campus police prevented fourmembers of the MIT Social Justice collab-orative – three undergraduates and onealumnus – from peacefully leafletingmarchers at graduation. They were activelyprevented from handing out leaflets, andone was arrested. We believe that these actsconstitute a serious violation of theConstitutional rights of the students andalumnus involved. Their leaflets spoke toissues of campus, scientific, and nationalrelevance, and civilian vs. military priori-ties in the budget of the National Institutesof Health. The leafleters accosted no one,nor did they interfere with the progress ofthe graduation ceremonies. The subse-quent release of the arrested leafleter, withcharges dropped, is not an exoneration ofthe police actions. Arrests are chilling, andarresting and then releasing and droppingcharges is a classic mechanism of suppres-sion of free speech when no law has beenviolated.

Freedom of expression, particularlypolitical speech, is a cornerstone ofmodern democracy. The right of free

expression is of particular importance inthe university, in part in support of thebroader democratic goal, in part becauseof the necessity of the fullest freedom of

expression for optimal progress in theacademic enterprise. For these reasons,the MIT faculty needs to insist that:

1) The suppression of freedom ofexpression at Commencement 2004was unacceptable and needs to becondemned by the MIT administra-tion, not excused.

2) The protection of the rights ofexpression of everyone – students,staff, and faculty – is a fundamentaltask of campus police, and that

3) University rules and regulationsprotecting freedom of expression bestrengthened in the coming period.

Among the reforms that we feeldeserve serious consideration:

• Establishment of a Faculty/Staff/Student Review Board, modeled afterCivilian Review Boards, to overseepolice actions on the campus.

• The expansion of the MIT CampusPolice mission to explicitly includethe protection of the rights offreedom of expression and assemblyof students, staff, and faculty.

To give some historical context to theviewpoint we express above, we note thatdespite its role as a leading partner in mil-itary research and national defense policy,

MIT has a long history of sustainingdissent – from the nuclear disarmamentmovement led by physicists VickiWeisskopf, Herman Feshbach, Phillip

Morrison, Aron Bernstein, and HenryKendall, to the Scientists Strike for Peaceduring the Vietnam War, led by DavidBaltimore, Ethan Signer, and others, to theMiddle East critiques by Noam Chomskyand associates, to the student critiquesand actions against the Gulf War and thecurrent war in Iraq.

There have always been attempts tolimit expression on such controversialissues, not only with respect to nationalpolicy debates, but also with respect toMIT policies. This faculty newsletter wasfounded because faculty members real-ized they had no independent means ofpublicly addressing each other or collec-tively criticizing the administration. TheNewsletter’s survival in its early yearsdepended critically on an active struggleby faculty supporters.

Thus we are particularly sensitive tothe issue of freedom of expression atMIT. Leafleting of those attending thegraduation march has been a frequentactivity during the past four decades.Why do we suddenly witness such consti-tutionally protected speech being sup-pressed? The most likely explanation isthe changing political climate and the

continued on next page

Leafleting of those attending the graduation march hasbeen a frequent activity during the past four decades. Whydo we suddenly witness such constitutionally protectedspeech being suppressed?

MIT Faculty NewsletterVol. XVII No. 1

4

current federal administration’s promo-tion of the Patriot Act and related initia-tives, the chilling effect of which is toencourage self-censorship and inhibitmany forms of political expression andassembly. We see this in the corralling ofdemonstrators at the Democratic andRepublican National Conventions, and inthe treatment of four young people onour campus.

Those of us old enough to have expe-rienced the McCarthy period or the CivilRights struggle are familiar with the useof a supposed internal enemy to justifysuppression. In this coming period, MITand other colleges and universities havean added responsibility to defendfreedom of expression and assembly asgranted in the Bill of Rights, and toensure that the truth is not a major casu-alty of the war on terrorism.

We need to take the first steps in ourown backyard, and guarantee that sup-

pression of dissent does not becomeaccepted campus policy. There is somedanger that the actions needed on thisissue will fall between the old and newadministration, with neither takingresponsibility. We feel that the facultymust ensure that these deep issues aredealt with seriously and effectively. Ourfailure to do so will abrogate our responsi-bility to maintain the atmosphere of freeinquiry, expression, and speech that is thelifeblood of great universities.

Editorial Sub-Committee

Affirming Freedom of Expression at MITcontinued from preceding page

Teaching this fall? You should know …the faculty regulates examinations and assignments for all subjects.

Check the Web at http://web.mit.edu/faculty/termregs.Questions: Contact Faculty Chair Rafael Bras at x3-2117 or [email protected].

First and Third Week of the TermBy the end of the first week of classes, you must provide a clear and complete description of:

• required work, including the number and kinds of assignments;• an approximate schedule of tests and due dates for major projects;• whether or not there will be a final examination; and• grading criteria.

By the end of the third week, you must provide a precise schedule of tests and major assignments.

Tests Outside Scheduled Class Times:• may begin no earlier than 7:30 P.M., when held in the evening;• may not be held on Monday evenings;• may not exceed two hours in length; and• must be scheduled through the Schedules Office.

No required classes, examinations, exercises, or assignments of any kind may be scheduled after the last regularly scheduled class in a subject, except for final examinations scheduled through the Schedules Office.

No Testing During the Last Week of ClassesTests after Friday, December 3 must be scheduled in the Finals Period.


5

Rafael L. BrasFrom The Faculty ChairPreliminary Position of the Faculty PolicyCommittee on Faculty Governance

L A S T S P R I N G T H E Faculty PolicyCommittee (FPC) started to codify someconcrete ideas to improve faculty gover-nance. Nobody considers the system asbroken, but the Committee feels thatimprovements are possible, and indeednecessary, to facilitate two-way communi-cations between the faculty and ouradministrative leaders; to strengthen theinput of the faculty as a whole on impor-tant policy and strategic decisions; torender the committee structure morestreamlined and efficient; to make facultymeetings more relevant and betterattended; and to provide the officers withthe tools to represent the faculty appropri-ately. FPC’s discussions are guided by aseries of principles, to be respected as ideasfor changes emerge. These principles are:

1. Preserve the concept of a unifiedfaculty;

2. Minimize bureaucracy;3. Maintain and enhance the close

working relationship with colleaguesin the administration;

4. Keep MIT’s governance above pettypolitics.

The following concepts have remainedas possible propositions for experimenta-tion or implementation. Faculty gover-nance is discussed in terms of threeelements: the officers of the faculty, thefaculty meetings, and the committeestructure.

The Faculty Policy Committee wantsyour input and suggestions. The FPC isalso requesting the comments ofPresident Vest and President-electHockfield, as well as other colleagues inadministration. Once all comments arereceived, we hope to formulate a proposal

that will be discussed with AcademicCouncil and then in a faculty meeting.Hopefully this process will lead toimprovements in governance to be imple-mented in the near future.

Officers of the FacultyThe Officers of the Faculty should work asa team. Each should have well definedroles, yet it should be recognized that theirmost important activity is to be stewardsof the policies, regulations, environment,and processes that make the faculty theultimate body directing the educationaland research enterprise of the Institute.The officers, and particularly the Chair ofthe Faculty because of her/his participa-tion in key committees, have the responsi-bility to represent the interests of thefaculty.

Office of the Chair of the FacultyThe Officers of the Faculty should have apermanent and visible locale well knownto the MIT Community, particularly thefaculty. This area should:

1. Have meeting space,2. House a full-time administrator and

assistant,3. Be the repository of files and

archives, and4. House the Faculty Newsletter activi-

ties. It is understood that the FacultyNewsletter is an independent outletfor faculty communications and isoperated by an editorial board onwhich the Chair of the Faculty is justanother member.

The “administrator” should be a full-time individual responding to the Officersand with the Chair as direct supervisor.

The role of the administrator shouldinclude:

1. Assisting the Chair, Associate Chair,and the Secretary in their duties,

2. Arranging logistical set-up of facultymeetings,

3. Writing the first draft of minutes offaculty meetings to be completed bythe Secretary,

4. Maintaining a list of outstandingcurrent issues,

5. Dealing with all day-to-day opera-tions of the office,

6. Staffing the Faculty Policy Committee,7. Staffing some of the other standing

committees. Some committees, likethe Committee on UndergraduatePrograms (CUP), Committee onCurricula (CoC), and Committee onAcademic Performance, (CAP) arenaturally best served by staff fromcorresponding administrative offi-cers, as they are now.

8. Serving as a communication link toall standing committees and theirstaffs,

9. Promoting and facilitate communi-cation with the Office,

10. Serving as the spokesperson of theOffice in ways to be defined by theChair.

The Office of the Chair of the Facultyshould have an annual budget for opera-tions, with the Chair as the responsiblesupervisor of its use. It is important to notethat FPC is not suggesting that the Chair ofthe Faculty be physically at the Office of theChair of the Faculty. In fact, the opinion isthat the Chair should physically remain attheir normal office location.



6

Duties of the OfficersChair of the FacultyThe term of the Chair of the Facultyshould be three years; it is currently twoyears. Any changes in term will not applyto the current Officers of the Faculty.Duties should include:

• Setting the agenda of faculty meetings • Chairing the faculty meetings in the

absence of the President• Chairing the Faculty Policy Committee• Working with the Nominations

Committee in appointing Chairs ofstanding committees

• Working with the President and otherMIT officers on appointment of adhoc committees or task forces

• Convening the Chairs of standingcommittees twice a year

• Sitting on Academic Council (includ-ing the Deans and promotions sub-groups)

• Participating ex officio on key com-mittees (those dealing with issuesthat could result in major policyimplications)

• Addressing faculty grievances• Leading the faculty at graduation• Attending meetings of the MIT

Corporation• Appointing faculty committees on

presidential searches, in collaborationwith the other officers of the faculty

• Hosting receptions and events• Recognizing faculty and their service• Introducing the Killian lecturer• Managing the election of Institute

professors• Promoting communication via fre-

quent formal and informal meetingswith key individuals at MIT

• Overseeing the adherence to termregulations

• Overseeing the regular review of theRules and Regulations of the Facultyand helping Academic Councilmaintain the currency of policies andprocedures

• Communicating and controlling com-munication with the faculty-at-large.

Associate ChairThe Associate Chair should serve the sameterm as the Chair. Duties should include:

• Setting the agenda of the facultymeetings

• Assisting the Chair and representingthe Chair in her/his absence

• Serving as a member of the FacultyPolicy Committee

• Serving as an ex officio member andliaison to the Committee onUndergraduate Programs

• Addressing faculty grievances• Working closely with the Chair in all

her/his duties, as needed.

Chair-ElectThe Chair-Elect should be a full-fledgedmember of the Faculty Officers and par-ticipate in their work during the year-in-transition. The Chair-Elect shouldbecome a guest of the Faculty PolicyCommittee.

SecretaryThe Secretary should serve the same termas the other officers. Roles should include:

• Setting the agenda for faculty meetings• Sending out calls to faculty meetings• Overseeing and monitoring written

presentations of motions and othermaterial for presentation to thefaculty

• Writing and distributing minutes ofthe faculty meetings

• Addressing faculty grievances• Coordinating on-going records of

standing committees for the Officers• Overseeing archiving of all records• Coordinating memorial tributes to

faculty members• Authorizing the Registrar’s list of

degree candidates• Serving as a member of the Faculty

Policy Committee• Serving as an ex officio member of the

Committee on Graduate StudentPolicies.

Faculty MeetingThe Faculty Policy Committee sees thefaculty meeting as an instrument topromote debate, influence policy, and

provide the faculty with ownership ofmajor decisions affecting the educationaland/or research enterprise of the Institute.Historically, faculty meetings are poorlyattended, except on rare occasions. Thereare several (negative) reasons for the poorattendance. First, there are issues of formand structure that discourage attendanceand do not promote a sense of ownershipand open discussion by the faculty. Second,there is a feeling that all decisions arealready made and the meeting is a “rubberstamping” exercise. Third, there is a sensethat the issues brought to the meeting areat best unimportant. A fourth (and posi-tive) explanation for the lack of attendanceand participation, is that the faculty truststhe committee structure and the adminis-tration to make the right decisions.

FPC has compiled the following list ofpossible changes. Several of these can beimplemented as experiments withoutchanges to the Rules and Regulations of theFaculty. For example, under present regu-lations, it is possible for the President tocede the chairing of the meeting to theChair of the Faculty. As stated earlier,many of the suggested changes are onlystructural.

1. Some faculty meetings to be led bythe Chair of the Faculty. Both theChair and the President should con-tinue to be present at all meetings.

2. Request that each departmentappoint three representatives toattend faculty meeting, for staggeredtwo-year cycles. These individuals,generally senior faculty, will have theresponsibility of reporting to theirunits. Note that the meeting willcontinue to be open to all facultymembers and all faculty in atten-dance will have a vote.

3. Change the time of meetings not toconflict with classes. A possibility isto alternate times.

4. Have separate business and infor-mational meetings. Develop a setof agenda items for generaldebate.

5. Introduce evening dinner meetings,maybe at the beginning of eachterm, in order to define issues.

Preliminary Position of the FPCBras, from preceding page


7

6. Publish position statements onissues that come up to the faculty,(faculty committees to present suc-cinct arguments for and against).

7. Standing committees should reportto the faculty at least annually.

8. Have the ability to have at least somefaculty meetings closed to thepublic. Dinner meetings could servethat role.

9. Develop a Web-based voting proce-dure for some of the most routine,but necessary, votes.

Committee StructureThe committees of the faculty are the keyto good governance.

Overall the committee structure workswell. Improvement is needed in:

1. Communication of the FacultyPolicy Committee with othercommittees,

2. Communication between committees,3. Overlap with other faculty commit-

tees, and4. Charge of committees.In thinking of the committee struc-

ture, it is important to keep in mind that,in our system, the faculty meeting is acommittee of the whole.

Definition of CommitteesThe standing committees are: AcademicPerformance, Curricula, Discipline,Faculty Administration, Graduate SchoolPrograms, Library, Nominations, OutsideProfessional Activities, Student Life,Undergraduate Admissions and FinancialAid, and Undergraduate Programs. Inaddition, there is a CommunicationsRequirement Subcommittee of theCommittee on Undergraduate Programs,Edgerton Faculty Achievement SelectionCommittee, and the Killian AwardSelection Committee.

Most committees work well and havefull agendas. Nevertheless, it is necessaryto review the charge of all committees,their relevance and composition relativeto their tasks. At least two committees –Outside Professional Activities andFaculty Administration – are currentlydefined in ways that limit their potential.

Some ThoughtsAs a way to improve communications, allstanding committees should have a repre-sentative from the Faculty PolicyCommittee. All committees should reportto the Faculty Policy Committee on anannual basis, via short written or oralreports, as necessary. In some cases,reports to the faculty at large may be nec-essary (as is already the case with theCommittee on Discipline). Committeechairs should meet with the Faculty PolicyCommittee and the Faculty Officers twicea year. All meetings of committees shouldbe recorded in minutes and they shouldall be archived. The minutes should beshared with the Faculty Officers.

The mandate of the Committee onAdministration should be clarified andthe membership redefined; the sameapplies to the Committee on OutsideProfessional Activities. An outcome maybe that the existing functions of thesecommittees could be folded into otherexisting or new standing committees.

The Committee on Graduate SchoolPrograms should parallel the CUP/CoC,possibly splitting into two functions,policy and curricula, and include afaculty member as chair, with the Deanof Graduate Students as an ex officiomember, as is the case with undergradu-ate committees.

The outgoing Chair of the Facultyshould name the membership of theNominations Committee. Nominationsto committees and to the Chair of theFaculty should be requested from thefaculty. The Chair should appoint theAssociate Chair of the Faculty and theSecretary.

The membership of the Faculty PolicyCommittee should increase to includemore faculty members. The Secretary ofthe Faculty should be a voting member.

A clear process should be in place forperiodic review of standing committees,and a process to add or eliminate com-mittees. Monitoring the “health” of com-mittees could be a role for theChair-Elect during her/his year-in-train-ing, so a review would occur at leastevery three years.

All ad hoc committees of the President,Provost, or Chancellor should be dis-cussed with the Faculty Policy Committeeto make sure that the role of standingcommittees is considered and member-ship discussed.

New standing committees may be nec-essary on:

1. International engagements,2. Graduate program policy,3. Graduate programs curricula and

degrees, and4. Faculty regulations and policies and

procedures.

Concluding RemarksMIT operates better than most other aca-demic institutions. Its system of gover-nance allows for a speedydecision-making process; it aims to beconclusive; it is not caught in politicalintrigues; and most importantly, hasavoided the “them versus us” syndromebetween the administration and thefaculty. After all, the academic administra-tion is composed of faculty. All of theabove are characteristics that we must pre-serve. Nevertheless, the system mustevolve and adjust to the times. If thefaculty is to retain the responsibility of theacademic well being of the Institute, thenit must become more involved and thesystem of governance should encourageand facilitate that involvement (see“Improving our System of FacultyGovernance,” MIT Faculty Newsletter,Vol. XVI No. 4, February/March 2004).

The suggestions made in this docu-ment are made in the spirit of the aboveparagraph, and evolve from discussions inthe Faculty Policy Committee. MIT isnever afraid of change and always ready toimprove. FPC is eager to experiment inthe search for improvement in gover-nance. Please share your opinion with us.You can write to [email protected] or participate in discussion throughhttp://web.mit.edu/faculty/.

Rafael L. Bras is a Professor, Civil andEnvironmental Engineering and Earth,Atmospheric and Planetary Sciences; FacultyChair ([email protected]).

Jeanne S. BambergerTeach TalkDeveloping Musical Structures: A Reflective Practicuum

[These comments are excerpted from thepaper,“The Development of Intuitive MusicalUnderstanding: A Natural Experiment.”Psychology of Music: January, 2003.http://www.sempre.org.uk/journal.html]

OVE R TH E YEAR S I have taught themusic fundamentals course subscribing tothe usual rule-based music theory prac-tice. But recently, out of a sense of dissatis-faction, I redesigned the course, mostsimply to make it work better – moreappropriate to students’ largely untutoredbut none-the-less well-developed musicalintuitions, and more responsive to stu-dents’ active and self-motivating ways oflearning.

As I saw it, there was a major problemwith previous approaches to the funda-mentals course: we have been asking stu-dents to begin with what we believedwere the simplest kinds of elements, butwe were actually confusing smallest ele-ments with simplest elements. Wefocused on these small, isolated, decon-textualized pitch and duration elementspartly because they are the easiest todefine, and thus also the easiest to assesswith respect to whether students havelearned them or not. More important,the symbols that represent these ele-ments are the tools of the trade for sea-soned musicians – they are what wedepend on for communicating with oneanother, for saying what we heard and fortelling others what they should hear andplay. But in doing so, we are not distin-guishing between, on one hand, our ownmost familiar units of description, thenotes shown in a score and our analytic

categories, and on the other, our intu-itive, contextual units of perception –those which we all, in fact, attend to inlistening and making sense of the musicall around us.

From everything I have learned so far,these “units of perception” are highlyaggregated, contextually and functionallymeaningful entities such as motives andphrases, their boundaries marking thelandmarks, the goals of motion, as wefollow the continuously unfolding per-formance of a composition. We don’tlisten to “notes” anymore than we listen toletters or even phonemes in following theunfolding of ideas in a lecture or a play.

It is not surprising, then, that students,often those who are best at improvisingand playing by ear (as well as those whoare best at improvising when making andfixing mechanical gadgets), are sometimesbaffled and discouraged when we askthem to start out by listening for, lookingat, and identifying the smallest, isolatedobjects. For in stressing isolated, decon-textualized objects to which our units ofdescription refer – to measure and nameobjects in spite of where they happen andtheir changing structural function – weare asking students to put aside their mostintimate ways of knowing.

The new course, called “DevelopingMusical Structures” (21M.113) perhapssurprisingly shows certain similaritieswith the innovations implemented byTEAL in physics (Belcher, MIT FacultyNewsletter, Vol. XVI No. 2, October/November 2003) and the comments ofWarren Seering in mechanical engineer-ing (Seering, MIT Faculty Newsletter, Vol.

XVI No. 1, September 2003). For example,instead of starting with exercises drawnfrom canonical music theory, studentsbegin by actually making music throughcomposition projects aided by the com-puter music environment, Impromptu.Design PrinciplesTwo very basic principles have guided thedesign of the course and Impromptu. First,computers should be used only to dothings we can’t do better in some otherway. Second (borrowed from HalAbelson), an educational computer envi-ronment is valuable to the degree it causesits developers to re-think the structure ofthe relevant domain. Thus instead ofsaying, “Here is this computer with allthese neat possibilities, what can I do withit?” I said, “Here are some things thatbeginning music students can do already,how can I use this intuitive know-how tohelp them learn to do what they can’t yetdo in a more musically relevant, intuitive,and accessible way?

Impromptu evolved in answer to thesequestions coupled with related issues ofrepresentation. Music notations representmusic at the “note” level and I wanted togive beginning students more aggregatedand perceptually meaningful elements. But“notes” are necessary to make them. So, Iwas drawn to the potential of the computeras an interactive medium because I couldcreate programmable, clickable icons thatwould immediately play just such alreadyaggregated melodic motives. Theseplayable icons would function for begin-ning students in their initial composingprojects as both units of perception andunits of work. We called them tuneblocks.


8


9

The Working EnvironmentIn the screen shot below, the icons on theright, when clicked, play meaningfulstructural entities (motives); in thisexample, they include the motives withwhich to reconstruct the melody, “Ode toJoy.” To build the melody, students dragtuneblocks into the Playroom and arrangethem in order so that they play the wholemelody.

Blocks 3-1-3-2, the opening twophrases, are shown in the Playroom.Notice that as students build up a melody,they are actually involved in “constructive

analysis”– i.e., they are reconstructing thelarger structure of the melody as embod-ied by the sequence of icons/motives. Thegraphics window at the bottom of thescreen shows a more fine-grained repre-sentation of the sounding blocks – “pitchcontour” graphics.

The most interesting work developswhen students are given what they called“strange” blocks borrowed from unfa-miliar pre-tonal or post-tonal styles.Students are asked to make a melody

“that you like and that makes sense” bylistening, arranging, and rearrangingblocks in the Playroom window and alsomodifying the “contents” (pitch andrhythm) by “opening up” the blocksusing the edit window. It turns out thatalmost everybody can do that. However,in any one class of 10 or so, given thesame materials, no two students come inwith the same tune. [To listen to theblocks and to hear some studentmelodies, please see the online version ofthis article on the Faculty NewsletterWebsite, http://web.mit.edu/fnl.]

Most important, students are asked toreflect on their process of composition asan integral part of the process, itself. Asthey work, students keep a log comment-ing on their decisions, and how thisinforms their emergent “model of a sensi-ble tune.” Students’ papers, together withthe performance of their compositions,become the center of our class discus-sions. Of course, students are often sur-prised, even confused, that the focus inclass discussions is on their puzzlements

and insights rather than on collectingnotes drawn from the instructor’s knowl-edge and information. Instead, as instruc-tor, I am interrogating, probing,questioning – in order, collaborativelywith the students, to make sense of andbuild on their sense-making.

The text, Developing Musical Intuitions(Bamberger, 2000) and recorded exam-ples on an accompanying CD, illustratehow composers have used and extendedsome of the structural principles that areemergent in the students’ own work. Inaddition to the conventions of notationand other vocabulary, the basics of musicfundamentals are couched in terms ofgeneralizable principles, thus informingencounters students have had in compos-ing, listening critically to one another’swork and to the recorded examples.

One of the gratifying results of theclass is that instead of my devising ques-tions to test what the students havelearned, it is their continuing investiga-tions into their own and one another’smusical understanding that becomes thegenerative base for developing newknowledge. Searching for answers to ques-tions that they have put to themselves, stu-dents begin to build a developing theoryof musical coherence. At the same timethey are developing hearings and appreci-ations of music that go beyond what theyknow how to do already, to knowingabout and knowing why. And in thatprocess, they are also learning to hear andto notice aspects of music that previouslypassed them by, thus helping to broadentheir musical taste and their listening pref-erences. Rather than giving up their intu-itions, they learn in the service of betterunderstanding them.

Impromptu Window

Units of Work: Multiple Representations

Jeanne S. Bamberger is a Professor of Musicand Theater Arts ([email protected]).


10

J. Mark SchusterWork of the Committee on theUndergraduate Program, 2003 – 2004

T H E C O M M I T T E E O N T H E

Undergraduate Program (CUP) overseesthe undergraduate academic program,particularly the freshman year and inter-departmental programs. CUP pays partic-ular attention to long-term initiatives andpolicies. It is responsible for encouragingexperimental innovation in undergradu-ate educational policy and has the author-ity to approve and supervise limitededucational experiments. This authorityextends to granting exceptions to theRules and Regulations of the Faculty inorder to facilitate such experiments.

CUP works closely with the Committeeon Curricula (CoC) and the Faculty PolicyCommittee (FPC), and together these threecommittees make up the core skeleton ofthe faculty governance structure. The workof these committees is not often visible tothe full faculty, but the decisions that theymake impact us all. That is why I am takingthis opportunity to give the faculty anupdate on the work undertaken by CUPduring the 2003-2004 academic year.

Last year’s CUP agenda focused on fourmain topics: reviewing two CUP-licensedexperiments that were reaching the end oftheir trial period; considering proposalsfrom two new undergraduate programsthat have requested temporary devices torestrict enrollment; monitoring the imple-mentation of various changes in the under-graduate program; and working with theCommittee on Student Life (CSL) to artic-ulate new guidelines for the advising andmentoring of upper-class students.

Review of CUP-Licensed ExperimentsWhen CUP authorizes an educationalexperiment, it does so for a limited lengthof time and the experiment is subjected toperiodic reviews by the Committee as toits promise and success. In 2003-2004, twoeducational initiatives came to the end oftheir experimental periods: the Special

Freshman Program in Media Arts andSciences and III-C, the experimentalundergraduate major in Archaeology andMaterials.

The Special Freshman Program inMedia Arts and SciencesThe Special Freshman Program in MediaArts and Sciences (MAS) was designed tooffer freshmen a hands-on experienceinside the Media Lab, which does not offeran undergraduate major. This programhad been run, since it was established in1998, as an experimental alternative fresh-man program endorsed by CUP. Duringthe 2003-2004 academic year, CUP took aclose look at the track record of theprogram to determine whether the com-mittee would endorse permanent status.

CUP discussed the MAS program overthe course of several meetings, duringwhich reservations were expressed con-cerning the coherence of the studentexperience in the program and the lack ofavailable data about and from partici-pants. Members of CUP had some linger-ing concerns over the quality of theexperience offered by the program, andthere was a feeling that the program hadnot done enough to ensure its separateidentity as an alternative freshmanprogram. MAS was asked to address theseconcerns through a clearer articulation ofthe goals and objectives of the program,development of a more coherent fall aca-demic program for participants, and anoutline of how its success will be moni-tored. MAS has since presented a plan toCUP, which includes improved use ofFreshman Advisory Seminars, greater vis-ibility for MAS.110 (the core fall subject,which is now approved as a communica-tion-intensive HASS subject [CI-H] and aHASS elective), and clearer articulation ofthe role of UROP placements in thisprogram.

As a result, CUP has endorsed perma-nent status for the program, but it alsobelieves that the MAS Freshman Programshould be reviewed in the context of allthe alternative freshman programs. Thus,CUP has recommended to the Task Forceon the Undergraduate EducationalCommons that it articulate criteria andguidelines for alternative freshman pro-grams as part of its deliberations. Oncesuch criteria are articulated, CUP wouldthen review the MAS Freshman Programand all other alternative freshman pro-grams accordingly.

S. B. Degree in Archaeology andMaterialsOn rare occasions, CUP has taken theresponsibility for authorizing experimentalmajors; the undergraduate programs inComparative Media Studies and inArchaeology and Materials are the mostrecent examples. In 2003-2004, CUPundertook a review of the experimentalundergraduate major in Archaeology andMaterials (III-C) – the first to be informedby the Guidelines for the Approval of NewUndergraduate Programs endorsed by thefaculty in spring 2003.

CUP determined that, over the dura-tion of this educational experiment, III-Chad evolved into an innovative andvibrant undergraduate program withcommitted faculty and excellent students.III-C has all the hallmarks of a successfulprogram of the sort that CUP would liketo encourage. Accordingly, at the May2004 faculty meeting, CUP introduced amotion to make the S. B. in Archaeologyand Materials a permanent major. Thefinal vote on this motion was taken andpassed at the September 15, 2004 facultymeeting.

Because the III-C program has raisedseveral issues that have been under discus-sion for some time, some further explana-


11

tion of CUP’s endorsement is in order.One of the original criteria articulated byCUP for evaluating the success of III-Cwas the number of students enrolled inthe major. The committee no longer con-siders this criterion to be determinative.By endorsing III-C, it is not the intent ofCUP to divert attention from the size ofproposed programs, but to assert that thenumber of majors is not the only indica-tor of relevance. Overall enrollment in thesubjects associated with this program hasbeen substantial and continuing, and theHASS concentration has been markedlysuccessful.

In the committee’s deliberations, othercriteria emerged as relevant in assessingthe quality of the program, and they arethe basis for CUP’s recommendation thatthe program be made permanent. In par-ticular, the III-C program has made acommitment to teaching students aboutthe origins of the disciplines that it incor-porates. This merging of the social sci-ences, humanities, science, andengineering within a single curriculum issomething that the CUP has long favoredand encouraged.

Temporary Restrictions on EnrollmentsDuring the 2003-2004 academic year,CUP received proposals to limit tem-porarily enrollments in two new academicprograms: the minor in Management anda major in Biological Engineering.

Minor in ManagementThe Sloan School has been developing aproposal for an undergraduate minor inManagement. The proposal will bebrought to the Committee on Curriculaearly in fall 2004. The expectation is thatwhen it is launched in fall 2005, thisminor will become an attractive optionfor undergraduates across the Institute,but there is a great deal of uncertainty asto what the actual demand will be.Consequently, Sloan has been workingwith CUP to develop a rationing mecha-nism that could be used to ease the transi-tion into the new minor. Thus, while theminor in Management is being proposedas a permanent program, Sloan has asked

CUP to authorize a rationing mechanismon a transitional (experimental) basis, andthe committee has endorsed the plan touse a lottery integrated into the existingSloan bidding process to allocate space inthe minor. The first such lottery will takeplace in spring 2005 and may continuefor up to four years.

S. B. Degree in Biological EngineeringThe Biological Engineering Division isdeveloping a proposal for an undergradu-ate major in Biological Engineering. It,too, is concerned about the potentialvolume of students who might declareBiological Engineering as their major.CUP has had several discussions with thefaculty who are developing this program,and it has become clear that the mostimportant binding constraint is labora-tory space for required subjects. CUP hasadvised the Biological EngineeringDivision as to what forms of allocationmight be acceptable, and a final proposalis pending.

Changes in the Undergraduate ProgramCUP is responsible for overseeing changesin the undergraduate program, particu-larly in the General Institute Require-ments (GIRs). Recently, CUP has beencalled on to monitor the roll-out of thenew Communication Requirement, theexperiment with the sophomoreExploratory Subject option, and changesto spring-term grading for freshmen.

Responsibility for day-to-day oversightof the Communication Requirement fallsto the CUP’s Subcommittee on theCommunication Requirement (SOCR).Over the past three years, SOCR has beenparticularly involved with the approval ofCI-M subjects, those communication-intensive subjects that are offered as partof every major. There is considerable vari-ation as to how departments implementthis component of the requirement, andSOCR is working to make sure that thisvariation is sensible and appropriate tothe disciplines.

2003-04 was the second year sincespring-term freshman grading changedfrom Pass/No Record to A, B, C/No

Record. CUP has been monitoring theimpact of these changes in detail, but it istoo early to tell how they have beenreceived either by the faculty or students.

Linked to the changes in freshmangrading has been an experiment allowingsophomores to designate one subject eachsemester as “Exploratory.” While therewere some start-up issues linked to theroll-out of the Exploratory option, 289sophomores chose to designate one oftheir subjects as Exploratory in the fall(ultimately, 64 of these students opted toconvert the subject to Listener status), and320 designated an Exploratory subject inthe spring (with some 34 conversions toListener status to date).

CUP will continue to monitor all ofthese changes in the coming year, particu-larly as more data become available as totheir success (or failure).

Advising and MentoringOver the past several years, CUP has had anumber of conversations concerning thequality of undergraduate advising andmentoring. CUP members are concernedabout the relatively low number of facultyinvolved in freshman advising in general,and in Freshman Advisor Seminars inparticular. We are also concerned aboutthe quality of advising of upper-class stu-dents once they declare their majors.These concerns are shared by theCommittee on Student Life, and the twogroups have been developing a series ofrecommendations that we hope to presentto the faculty this fall. We will also becoordinating this work with the TaskForce on the Undergraduate EducationalCommons.

It is my hope that these commentshave made the work of CUP a bit moretransparent to our faculty colleagues. Aschair of CUP, I urge any faculty memberwho has an idea or a concern relating tothe work of our committee to contact medirectly at: [email protected].

J. Mark Schuster is a Professor of UrbanCultural Policy; Chair, Committee on theUndergraduate Program ([email protected]).


12

Robert J. SilbeyTask Force on the UndergraduateEducational Commons

IT I S N ECE SSARY, from time to time,for a great university like MIT to take stockof its undergraduate educational pro-grams from a fundamental perspective. Asa matter of course, these programs evolveslowly over time as faculty introduce newideas and new teaching techniques. Sincethe last thorough examination of theundergraduate curriculum, the MITundergraduate student body has changeddramatically, becoming more diverseacross a wide range of dimensions.

For these reasons, this is an appropri-ate time for us to reevaluate undergradu-ate education at MIT, and to ask if ourstudents – when they graduate – areappropriately educated and have acquiredthe skills and attitudes necessary to makepositive contributions to their field and tosociety. The report of the 1998 Task Forceon Student Life & Learning discussedcommunity life at MIT and made recom-mendations for improving our environ-ment – some of which have been actedupon. Building on that foundation, thenewly formed Task Force on theUndergraduate Educational Commonswill affirm and update the goals of an MITundergraduate education and proposeimprovements to the core educationalexperience that are tailored to the studentswe teach and the world in which they live.

During the spring term, the Task Forceon the Undergraduate EducationalCommons met bi-weekly to educate itselfabout the current state of the core educa-tional program. The General InstituteRequirements (GIRs) are designed tobroaden our students’ academic horizons,improve their problem solving and ana-lytical skills, and provide a solid founda-tion upon which future learning can be

built. The Task Force dedicated a numberof meetings to fully understanding whatthe GIRs encompass, the history that hasled to their current form, how successfulthey are perceived to be, trends in enroll-ments and other data, and the issues andchallenges faced by those who deliver theGIRs. As each requirement was reviewed,familiar themes emerged: pressure andpace; a desire to add to the curricularrequirements coupled with a reluctance totake anything out; and issues with theretention and application of materiallearned in the first year. There was alsodiscussion of a perceived lack of excite-ment and engagement among students inthe first year and a need to articulate thepurpose and goals of the core curriculumin a more effective manner.

Individual members of the Task Forcegathered input from various stakeholdersin the educational commons, includingthe faculty who teach the core subjects,the MacVicar Fellows, the facultyUndergraduate Officers in each depart-ment, the Engineering Council onUndergraduate Education, and the DUEVisiting Committee. These groups raisedconcerns over the allocation of timewithin the curriculum and the trade-offsnecessary to add new components to theMIT educational experience. In addition,many members of these groups expresseda specific need to revisit the purpose,goals, and implementation of all theundergraduate requirements.

The student members of the TaskForce reached out to the larger commu-nity through an open forum and twosmaller student roundtable discussions.They also gathered student opinionsthrough the UA Website and in conversa-

tions with the Student Senate. Much ofthe student sentiment focused on theneed to simplify the HASS requirementand broaden the School’s subject offer-ings. In addition, a group of studentsstrongly advocated the development of a“diversity requirement.” There was unani-mous praise for undergraduate researchopportunities, but students expressed adesire for interaction with facultymembers in a wider variety of settings.

Dean of Admissions Marilee Jonesattended an early Task Force meeting toreport on the profile of the current gener-ation of students, and how this profile haschanged in the past 10 years. According toMarilee, our students have broader inter-ests than MIT students of the past. Theyhave been encouraged throughout theirlives to engage in a wide range of activitiesand feel significant pressure to succeed inall of them. These students are accustomedto having little unstructured time and havehad minimal experience with failure. All ofthese factors have significant implicationsfor how we teach our students and conse-quently what and how well they learn.

During an intensive work week heldshortly after commencement, the TaskForce heard from instructors of a few ofthe innovative, hands-on subjects that aretaught throughout the Institute, such as12.000 (Solving Complex Problems);2.000 (How and Why Machines Work);and 6.002X (an experimental version ofCircuits and Electronics). The group con-sidered whether these classes could serveas models for additions to the educationalcommons that would increase enthusiasmand conceptual learning among students.Dr. Lori Breslow of MIT’s Teaching andLearning Lab joined the group for a dis-


13

cussion of recent pedagogical researchunderway at MIT and elsewhere. Thegroup reviewed the success of active learn-ing approaches and debated the methodsand feasibility of incorporating this type ofteaching into a wider range of subjects.

The Task Force spent a full morningtalking with Associate Dean of EngineeringDick Yue, chair of the School ofEngineering Council on UndergraduateEducation (ECUE). In addition to hearingabout the results of a number of surveys ofSoE students and faculty (including a studyof engineering student workload patterns),Professor Yue shared ECUE’s thoughts onpotential links between engineering educa-tion and the core educational program.

The Task Force also reviewed prelimi-nary findings from this year’s SeniorSurvey and requested additional analysisfrom the Institutional Research staff of theProvost’s Office. As in the past, the dataindicated that students at MIT placegreater importance on developing analyt-ical, quantitative, and problem solvingabilities than on understanding andappreciating the humanities, arts, andsocial sciences. The Task Force hopes to beable to track the priorities of students overtime and compare MIT results to those ofother institutions, shedding light onwhether it is reasonable for the MIT edu-cational commons to encourage greaterbalance among these areas.

The remainder of the work week wasdedicated to reviewing the findings ofprior committees regarding the goals ofan MIT education and the principles thatguide the teaching of our students. At theend of the week, the group broadlydefined four focus areas on which to con-centrate. Members divided into smallgroups and will report their progress tothe full Task Force this month.

The Task Force was charged to engageactively with the entire MIT communitythroughout its deliberations, and as thegroup moves in the upcoming monthsfrom learning mode to generating a draftset of educational goals and ideas, weintend to live up to this commitment.Members of the Task Force will begin anactive outreach to departments, faculty,

students, staff, and alumni to share ourwork-in-progress, as well as to solicit feed-back. In addition, our student memberswill establish a student advisory group toensure that we receive regular input fromthe wider student community. While theTask Force has made progress, there is stillmuch work to be done. Now that thegroup has developed a solid understand-ing of the current state of MIT’s educa-tional program and the forces that areimpacting the MIT experience, we canfocus on what aspects of the curriculumneed to be addressed and how we can best

achieve educational reform. As we formu-late a vision of the MIT of the future, wewill look to you to provide your perspec-tive. Contact your colleagues from themembership list to share your ideas forenhancements to the undergraduate edu-cational commons.For more information,seethe Task Force Website: http://web.mit.edu/committees/edcommons.

• Robert J. Silbey, Task Force ChairChemistry Dean, School of Science

• Rafael L. Bras, Guest Member Civil and Environmental Engineering Chair, Faculty Policy Committee

• John G. Brisson II Mechanical Engineering

• Margaret S. Enders, Executive Officer Associate Dean, Office of the Dean forUndergraduate Education

• Steven D. EppingerSloan School of Management

• Elizabeth L. GreenwoodMathematics, Class of 2005

• Thomas J. GreytakPhysics

• W. Eric GrimsonElectrical Engineering and Computer Science

• Paula HammondChemical Engineering

• Diana HendersonLiterature

• Kip V. Hodges, Associate Chair Earth, Atmospheric and Planetary Sciences

• Chris KaiserBiology

• Thomas KochanSloan School of Management

• Tomas Lozano-Perez Electrical Engineering and Computer Science

• John Maeda Media Arts and Sciences

• Anne McLeodStaff Associate, Office of the Dean forUndergraduate Education

• Albert R. Meyer, Ex Officio Electrical Engineering and ComputerScience Chair, Committee on Curricula

• Haynes R. Miller Mathematics

• David A. Mindell Science, Technology, and Society ProgramEngineering Systems Division

• Heidi Nepf Civil and Environmental Engineering

• Dava J. Newman, Associate ChairAeronautics and Astronautics

• Robert P. RedwinePhysics Dean for Undergraduate Education

• Jessica B. RheeChemical Engineering, Class of 2006

• J. Mark Schuster, Ex Officio Urban Studies and Planning Chair, Committee on the UndergraduateProgram

• Charles Stewart III, Associate ChairPolitical Science Associate Dean of the School ofHumanities, Arts, and Social Sciences

• Christopher A. SuarezElectrical Engineering and ComputerScience, Class of 2006

• J. Kim VandiverOcean EngineeringDean for Undergraduate Research

• John R. VelascoPolitical Science, Class of 2005

Robert J. Silbey is a Professor of Chemistry;Dean of the School of Science; Chair, TaskForce on the Undergraduate EducationalCommons ([email protected]).

Task Force Membership


14

James B. OrlinSome Reflections on Aspects of theUndergraduate Education Policy

R ECE NTLY, I HAVE B E E N giving morethan my usual amount of thought tosome aspects of undergraduate policy, inpart because of issues that have arisen inteaching undergraduates, and in partbecause of my new role as a member ofthe Committee on Academic Performance(CAP). My goal in this note is to raiseissues for discussion, and not to proposenew policies, per se.

Possible Limits on UnitsThis past semester, I became aware ofthree different students in my class whowere taking over 100 units each. Each ofthem asked to take the final exam a daylater than the scheduled time, not becauseof conflicts, but because of the stress theywere feeling. I was surprised to learn thatthere is no upper limit on the number ofunits taken by undergraduates after theirfirst year. A limit on units would havehelped in these situations because itwould have lowered the stress level, andbecause it would have made it much lesslikely that these students would haveasked for special privileges.

I was relieved to learn that taking suchan excessive number of units is not wide-spread. I obtained data from theRegistrar’s Office for this article, andlearned that only six students out of morethan 3000 students (2nd year and higher)completed the semester with 100 or moreunits, and fewer than 3% of studentscompleted the semester with 75 or moreunits.

The table summarizes the number ofunits taken by students.

Even if the problem of students com-pleting the semester with very largesubject loads is not common, it still maybe worthwhile to consider limits on units,and to see whether it would be good edu-cational policy. There are clearly pluses topermitting students to take as many unitsas they want. It shows that MIT valuespersonal autonomy of students, andmakes it easier for students to doublemajor or to graduate in three years, orpossibly both. It also avoids the need formechanisms for limiting loads. But thisliberty also comes with costs. It leads tostudents spreading their focus, and notgiving the necessary attention to individ-ual subjects. It increases the stress level. Itnegatively affects subjects that have groupprojects. Given the efforts and thoughtthat went into limiting the number ofunits of first-year students, it is time tobroaden that discussion to consider stu-dents after their first year.

Drop and Add DatesRelating to the issue of overload is theissue of when drop and add dates occurin the semester. It seems to me (perhapsbecause I am naïve), that the primaryadvantage of having drop dates so late inthe semester (and later than the dates forcomparable universities) is so that stu-dents have even more time to assess whattheir final grade in the class will be, andthus make a more informed decision onhow to improve their GPA via selectivedropping. Undoubtedly most undergrad-uates value this option; however, it seemsto me to hinder education at MIT ratherthan aid it. It results in students deliber-ately taking overloads, and spreadingtheir efforts too thinly. And, it encouragesthem to overly focus on the grade ratherthan on the education. And for instruc-tors, it means that the class size is notdependable. It is also incompatible withsix-week subjects, which is a time period

Summary of Information on Units Taken During Spring 2004

2nd year 3rd year 4th year all studentsAvg. Number of Units

Before Add Date 64 64 58 62After Add Date 57 57 49 54After Drop Date 54 53 45 50

Maximum Number of UnitsBefore Add Date 150 207 267 267After Add Date 126 186 195 195After Drop Date 111 138 99 138

% taking 75 units or moreBefore Add Date 20.5% 21.2% 18.3% 20.0%After Add Date 5.1% 7.6% 3.9% 5.6%After Drop Date 3.1% 3.5% 2.0% 2.9%


15

that is becoming more common at theSloan School.

Perhaps the greatest problem createdby late dropping is the enormous waste ofintellectual efforts and resources.Students waste enormous time in takingsubjects for half a semester, and facultyand TAs waste enormous resources inteaching and grading these students.Given the scarce resources available, MITshould rethink when is an appropriatetime for the drop date.

What is Acceptable AcademicPerformance?When discussing academic performanceof undergraduate students, we flag stu-dents who are taking fewer than 36 units,or who have a term GPA that is below 3.0.As I recall, this was the same criteria usedwhen I arrived at MIT some 25 years ago,despite substantial changes in our under-graduate population and despite the pos-sibility that there has been grade inflation.This may be a good time to review what isrequired for performance to be acceptableat MIT.

There are several issues to considerwith respect to what constitutes accept-able academic performance. The firstissue is what performance merits awarning. Given the grade distribution atMIT, I propose that anyone with a GPAunder 3.0 merits a warning, as do someother students taking too light a load. Thesecond issue concerns the circumstancesunder which a student is required to with-draw from MIT. Here I suggest that MITshould consider being much stricter, and

not permit students to continue at MITwith warnings in many different semes-

ters. Personally, I view a cumulative GPAof less than 3.0 after the sophomore yearas not meeting what should be MIT stan-dards, and except in unusual circum-stances, such students should be asked towithdraw. While I do not expect everyoneto agree with me, I do believe it would bebeneficial to discuss, as a community,what constitutes acceptable performance.

Low Achieving StudentsAll undergraduate students at MIT weregreat achievers in high school, and arrivehere with great academic potential. Butfor a variety of reasons, not all studentshave academic success after they arrive.Moreover, for a number of underachiev-ing students, MIT is not only a source ofconstant stress and disappointment, butit can do serious damage to motivationand sense of self. This situation is mademore complex because many students atMIT view a transfer as an admission offailure.

The MIT community needs toacknowledge the simple and obvious fact

that some students will do better, behappier, and be more successful by trans-

ferring to another university. We need tochallenge the widely held (and incorrect)view that a transfer out of MIT is anadmission of failure. We at MIT are doinglow achieving students (and some otherstudents as well) no favor if we blindly letthem continue at MIT without present-ing academic counseling that includesalternatives.

I suggest that advice to students includethoughtful information about transfers toanother university and advice concerningfinancial assistance. Furthermore, MITshould consider having one or two advis-ing deans who specialize in transfers. Andfor those students interested in transfer-ring, we should do our best to make surethat they can transfer to a university wherethey will be both happy and successful. Inso doing, we would be serving these stu-dents quite well.

There are several issues to consider with respect to whatconstitutes acceptable academic performance. The first issueis what performance merits a warning. Given the gradedistribution at MIT, I propose that anyone with a GPA under3.0 merits a warning, as do some other students taking toolight a load.

James B. Orlin is a Professor of OperationsResearch ([email protected]).


16

Shawn FoleyBenefit Changes for Faculty Upon Retirement

YOU WI LL B E AN M IT R ETI R E E if,when you leave the Institute, you are age55 or older, and have completed 10 yearsof MIT Retirement Plan eligible serviceafter age 45. The following is a summaryof your Massachusetts Institute ofTechnology Health and Welfare RetireeBenefits.

Cost of Health Coverage at RetirementAs we announced to the MIT communityon June 16, 2003, MIT implemented anew method of determining MIT’s andretiree’s share of the cost of retiree healthcoverage beginning January 1, 2004. Youare not affected by the change if you wereborn before July 1, 1930 and were retire-ment plan eligible, or were born beforeJuly 1, 1940 and completed 10 years ofretirement plan eligible service beforeJuly 1, 1995.

Health Insurance at RetirementAs an MIT Retiree, you, your spouse/partner covered by an MIT Plan at thetime of your retirement, and eligibledependents qualify to elect to participatein one of MIT’s health insurance plans.Your coverage options are dependentupon your age at retirement.

Health Insurance If You Retire Before Age 65If you retire before age 65 and are notenrolled in Medicare Part A (HospitalCoverage) and Part B (Medical Coverage),you will continue to be covered by one ofthe health plans available to active MITemployees.

If your eligible spouse/partner isunder age 65 at the time of your retire-ment, your spouse/partner will remain inan active MIT health plan. Retirees andspouses/partners without dependent

children will change from family coverageto two individual plans.

If you have dependent children onyour family plan, the plan coverage willcontinue as family coverage until youand/or your spouse turns age 65 or yourunmarried, dependent child is eligible forcoverage through an employer or othergroup health plan or the end of the monththey turn age 25, whichever occurs first.

Health Insurance If You Retire On orAfter Age 65If you retire on or after age 65, MedicareParts A and B will become your primaryinsurance.You may add to that coverage bychoosing one of the MIT Group Medicaresupplement plans. MIT currently offersseveral Medicare supplement plans:

1. Indemnity Supplement Plan – MITGroup Medex Plan

2. HMO Supplement Plans – ManagedBlue for Seniors & Tufts MedicareComplement

3. Medicare HMO Plans – FirstSeniority & Secure Horizons

Note: In most cases, under the MITGroup Medex Plan, you may continue touse the MIT Medical Department for yourservices.

If Your Spouse/Partner is Age 65 orOlder When You RetireIf your spouse/partner is age 65 or olderwhen you retire and not actively workingor receiving coverage through anotheremployer or group health plan, and yourspouse/partner has Medicare Parts A andB coverage, your spouse will be eligible forcoverage through one of the MIT GroupMedicare Supplement Plans. Yourspouse/partner is not obligated to have thesame supplemental coverage as you andmay elect any one of the plans available.

Cost of Health Insurance Effective1/1/2004 – Under Age 65Any cost will be deducted from your MITpension check or you will be billedmonthly. MIT’s share of the cost of cover-age will be based on your years of retire-ment plan eligible service with theInstitute after age 45. The percentage willbe based on the full monthly cost of thePlan.

• MIT’s share will equal 50% forretirees with 10 years of such serviceafter age 45;

• MIT’s share will increase two per-centage points for each additionalyear of such service beyond the initial10 years of required service to amaximum of 70% for retirees with 20or more years of such service after age45; and

• Partial years of such service will beprorated at .167% per month.

Over Age 65Any cost for you and your eligiblespouse/partner, will be deducted fromyour pension check or you will be billedmonthly.

Cost sharing for retiree health plans:• MIT’s share will equal a percent-

age of the cost of MIT GroupMEDEX Plan.

• This percentage will be based onyour years of retirement plan eligi-ble service after age 45, will be pro-rated for partial years, and isidentical to the percentages thatapply to retirees under age 65, asexplained above.

• You will pay the remainder of the costof your coverage.

• The dollar amount MIT pays willchange as the cost of MIT GroupMEDEX Plan changes.

Years ofService

After Age45

MITShare

RetireeShare

20 70% 30%

19 68% 32%

18 66% 34%

17 64% 36%

16 62% 38%

15 60% 40%

14 58% 42%

13 56% 44%

12 54% 46%

11 52% 48%

10 50% 50%


17

The following chart illustrates theInstitute’s and retiree’s cost sharing byyears of service.

For More InformationPlease contact the Benefits Office by e-mail at [email protected] orby telephone at (617) 452-3694 from9:00 a.m. to 5:00 p.m. The office islocated in E19-215. The office at LincolnLaboratory is located at LL-A-128; thenumber is (781) 981-7055. Details of theplan changes can be found athttp://web.mit.edu/hr/benefits/retire_prep_health.html.

Shawn Foley is Assistant Manager,Department of Human Resources([email protected]).

Newsletter Redesign Wins AwardJ U D G E S F R O M T H E U C DA AnnualDesign Competition have chosen theredesigned MIT Faculty Newsletter toappear in the 2004 Design Show.Appearance in the Design Show repre-sents a Gold Award of design excellence.

The Newsletter redesign was done byTim Moore and Jan Moscowitz of thedesign firm Moore Moscowitz, with inputfrom representatives of the NewsletterEditorial Board and the President’s Office.

Additional entries from MIT to appearin the Design Show include theMIT_1990_2004.zip brochure, and theMIT FREE2B brochure. The exhibition canbe viewed at the UCDA AnnualConference, October 2-5, in Vail, Colorado.

Additional Design Certificates may beordered for MIT departments. For moreinformation, please contact MooreMoscowitz directly at (617-731-7783;http://www.mooremoscowitz.com).

Nominations Open for 2004/2005MIT Excellence AwardsNOMINATIONS FOR THE M IT ExcellenceAwards are being accepted online athttp://web.mit.edu/hr/rewards/ex_nominate/.These awards honor the exceptionalachievements of staff across the Institutein the categories of leadership, clientservice, collaboration and innovation,community service, and work-life balance.

The nomination deadline isWednesday, October 13, 2004 for the2004/2005 awards. The Excellence Awardsceremony will take place on March 2,2005 in Kresge Auditorium at 12 noon.For more information and the nomina-tion form, please visit http://web.mit.edu/hr/rewards/excellence/ or contact KandeCulver, program administrator, [email protected].

Times Square Offers Concerto forErhu and SubwayTH I S I SSU E’S M IT POETRY offering,“Concerto for Erhu and Subway” by Prof.Tunney Lee (page 20), was submittedsome months before a July 6, 2004New York Times article “Asian Music,Accompanied By the A Train.” The articledescribed the New York subways as a per-formance space in which the ancientmusical arts of China, including music forthe instrument described by Prof. Lee, iskept alive by master musicians.

“Before the attacks of Sept. 11, 2001,many of the musicians said, an eight-hourday of performing at a subway stationfetched an average of $70,” according tothe article. “Since then their income hasdropped by roughly a third because of theeconomic doldrums and, they speculate,increased suspicion of foreigners.”

Where Is This?

Short Takes

Can you identify this location on theMIT campus?


18

Daniel HastingsDaniel Roos

Establishing Leadership in the EmergingField of Engineering Systems

I N D ECE M B E R 1998, TH E School ofEngineering established a second newdivision, the Engineering SystemsDivision (ESD), which focuses on theengineering of complex systems. ESD’screation responds to the rapidly evolvingfield of engineering where there is a needfor the development of new approaches,frameworks, and theories to better under-stand engineering systems behavior anddesign. It also responds to a need withinthe School of Engineering for the devel-opment and support of educational pro-grams on complex systems and designsynthesis that will prepare students forleadership positions.

To quote from a recent letter fromDean Magnanti:

“The engineering profession today facesa number of unprecedented challenges,many reflecting the changed context inwhich engineers practice. It is no longerenough to design a product or a systemwithout accounting for the world in which itwill operate. Today, many large-scale,extraordinarily complicated systems call outfor a systems-driven engineering approach.Just consider a few of these critical systemschallenges:

• redesigning transportation systemssuch as airline, rail, and urbanhighway systems that have increasinglyreached their capacity and createdenormous delays;

• using information technologies tocreate products that are more timely,less expensive, and increasingly respon-sive to consumer needs;

• reconciling the inevitable growth inworld-wide energy demand withpotential environmental costs;

• creating product development systemsthat address the full spectrum of con-ceiving, designing, and developing anew product; and

• developing manufacturing systemsthat are more attuned to the humanimpacts they generate, from wageattenuation and job losses to disloca-tions linked to globalization.

At MIT, our role is to help meet theseand other societal needs, through leadershipgrounded in technical excellence and inno-vation. Indeed, we feel an obligation to focusour attention on addressing these challeng-ing issues. We believe that the convergingforces of increased system complexity andthe social impact of technology – combinedwith a need for increased leadership byengineers – create opportunities for newdirections in engineering education andpractice. The most successful engineers mustpossess superb professional skills as engi-neers, including a keen understanding ofsocial, regulatory, environmental, cultural,and other forces. In short . . . we needEngineering Systems.”

To fully appreciate complex engineer-ing systems requires an integrative holisticview that bridges traditional engineeringapproaches with insights from manage-ment and social science. Therefore, ESD isan integrative effort that cuts across theSchool of Engineering departments, theSloan School of Management, and theSchool of Humanities and Social Sciences.The Division has over 40 faculty membersincluding two Institute Professors (JoelMoses and Sheila Widnall). All ESDfaculty have either a joint or dual appoint-ment with another academic unit. [Thedual faculty appointment was introduced

when the two new Divisions were formedin the School of Engineering. Dual facultyshare their time equally between twounits; the division and a department.]These shared appointments enable ESDfaculty to work with their engineeringdepartments on system related initiatives.Overall, the Division provides an institu-tional framework and intellectual homefor engineering systems faculty to developand support system oriented educationaland research programs, facilitate theadmission of students to various interdis-ciplinary academic programs, andprovide governance on key issues such asfaculty hires, promotion, and tenure.

ESD brings together several systems-oriented educational professional pro-grams and research centers that weredeveloped at MIT over the past severaldecades. Five Master’s-level interdiscipli-nary professional practice educationalprograms at the Institute are serving over300 students today. These programsinclude Leaders for Manufacturing(LFM), System Design and Management(SDM), Technology and Policy Program(TPP), ESD SM, and Master ofEngineering in Logistics (MLOG). TheESD research centers are the Center forInnovation in Product Development(CIPD), Center for Technology, Policyand Industrial Development (CTPID),and Center for Transportation & Logistics(CTL). Like the ESD academic programs,these centers are interdisciplinary, involv-ing faculty from engineering, manage-ment, and the social sciences

ESD builds upon these pre-existingeducational programs and researchcenters. The mission of the Division is to


19

create a new field of study and to broadenEngineering education and practice. Toaccomplish this mission ESD haslaunched several new educational andresearch initiatives described below.

ESD PhDESD received approval from the MITfaculty in 2003 to offer a Doctoralprogram. The mission of the program isto undertake fundamental in-depthresearch oriented around theory, policy,and practice associated with engineeringsystems. All doctoral students take a corecomposed of courses in system theory,quantitative methods, and socio/technicalcontexts. The ESD PhD acquired theinterdisciplinary Technology, Policy andManagement (TMP) PhD. It currently hassome 40 doctoral students including stu-dents from the legacy TMP program.

The Engineering Systems SymposiumOn March 29-31, 2004, the EngineeringSystems Symposium brought over 360leading academics, industry, and govern-ment representatives, and students toMIT to learn about the emerging field ofEngineering Systems and to consider waysto work together. In the opening session,Dr. Vest noted, “This is a remarkable,perhaps historic, event of great import toengineering education and to ourInstitution. If we are to continue to be agreat Engineering school in the future andhelp address complex problems like anti-terrorism, the Columbia Shuttle tragedy,globalization and sustainability in waysthat benefit humankind, we will need tobe great in Engineering Systems.”In addi-tion to Dr. Vest, speakers included MITSchool of Engineering Dean ThomasMagnanti; Institute Professor SheilaWidnall; William Wulf, president of theNational Academy of Engineering, Dr.Joseph Bordogna, deputy director of theNational Science Foundation, and TravisEngen, president of Alcan. Several of thepresentations are available on MIT Worldat http://mitworld.mit.edu/series/57/.

A key feature of the Symposium wasthe release of the Engineering SystemsMonograph by ESD faculty and staff. In

addition to a paper by Dan Roos on thehistory leading to ESD’s creation and apaper by Daniel Hastings on ESD’s futureand the creation of Engineering Systemsleaders, there are six papers on the foun-dations of Engineering Systems. Aframing paper on foundational issues byJoel Moses is followed by five papers onvarious aspects of the field. Dan Whitneywas principal author of a paper onsystems architecture, Richard de Neufvilleplayed a similar role in a paper on uncer-tainty, Tom Allen wrote on enterprisesystems, David Marks on sustainability,and Nancy Leveson on systems safety. TheMonograph papers can be found athttp://esd.mit.edu/symposium/monograph/.The remaining papers presented at theSymposium can be viewed athttp://esd.mit.edu/symposium/agenda_day3.htm.

At the Symposium, Dan Roosannounced that over 20 of the top engi-neering schools in the U.S. and Europehave agreed to work collaboratively todefine and evolve the field of EngineeringSystems by sharing educational materialsand information on job opportunities forPhDs in Engineering Systems, and holdinginter-university student colloquia.

New ResearchESD’s TPP program, along with theCenter for International Studies (CIS), theDepartment of Political Science, and theScience, Technology, and Society (STS)program was awarded $2.9 million fromthe National Science Foundation’s presti-gious Integrative Graduate Education andResearch Traineeship (IGERT) Programfor a multidisciplinary program on assess-ing effects of emerging technologies.

The Program on EmergingTechnologies (PoET) is led by four prin-cipal investigators: ESD/TPP’s DanielHastings and Dava Newman; KennethOye of the Department of PoliticalScience, ESD, and CIS, and Merritt RoeSmith of STS. A workshop entitled“Emerging Technologies: RecognizingUncertainty and Assessing Implications”(also the fourth annual TPP sympo-sium) was held on April 12, 2004. More

information is available athttp://poet.mit.edu/igert.htm.

ESD’s Center for Transportation &Logistics signed a multi-year, multi-million dollar agreement with the govern-ment of Aragón, Spain, to help create aninternational education and researchprogram in logistics and supply chainmanagement. The MIT-ZaragozaInternational Logistics Program is part ofa large-scale initiative to develop theAragón region of Spain, around its capitalcity of Zaragoza, into a significant logisticscenter in Europe. The MIT-ZaragozaInternational Logistics Program is theCenter’s flagship effort. In addition to con-ducting cutting-edge research, CTL willwork with the Zaragoza Logistics Center tooffer graduate and executive education inlogistics to students from around theworld. The offerings will include a Master’sdegree modeled on MIT’s Master ofEngineering in Logistics (MLOG), aDoctoral degree, and a set of executiveeducation courses leading to certificates invarious logistics-related disciplines.

The FutureESD is working concertedly to build uponthis foundation and to strengthen its lead-ership position. With our interdisciplinaryfaculty, new PhD program and researchprograms, we have laid the groundwork forcontinuing to define and develop the newfield of Engineering Systems. However,there is much work to be done. We havemade an excellent start on defining theintellectual foundations of engineeringsystems and in the future years we willdeliver on that promise. ESD is taking aleadership role in engaging the extendedcommunity, including students, faculty,alums, partner companies, and staff withinESD, and reaching out into the world ofacademia, government, and industry atlarge. We are building a lifelong learningcommunity that encourages active and sus-tained partnership from all of our con-stituencies over the short and long term.

Daniel Hastings is Director, EngineeringSystems Division ([email protected]). Daniel Roos is Founding Director, EngineeringSystems Division ([email protected]).


20

on the uptown platform at Times Squareblack brown pale tall short fat skinny people swaddled in wool leather fur fleece jacketshoods hats balaclavas gloves mittens filling all the bits of left-over space

on the downtown platform a train screeches to a haltgraffiti and scratchiti is forbidden violators subject to fine and imprisonment indecipherable language gurgles from the loudspeakersis it baluchistani chechen cham chamorro chimbu, chukchi or maybe esperantodelayb mmnnnot stoppming at ppenmstaysnnnn bewarnm pinkpomkets

creating a sense of crisis nature and location unspecifiedjack-hammers join in adding an emphatic stop and go beat

the Chinese man age indeterminatejet black hair brush cut wearing a navy blue Fila sweater brown corduroy pants down jacket on a crate elevated from the grit ground into grime

hesits reed straight aloofon his folding stoolhis left hand holds the erhu by its long slender sandalwoodneck

MIT Poetry

Concerto for Erhu and Subway

Erhu. A two-stringed, vertical fiddle introduced into China

from Mongolia in the Song dynasty, 960-1279.


21

curved elegantly

at the top

punctuated bypaired pegs

sitting on his left thigh the ebony base black and polishedcovered with python skinblack and white pattern bold strokes of calligraphy

fingers flit like cicadas over the strings

his right hand sweeps the bow its two strings of horse hair remembering the captive barbarians mourning for their homeland

the uptown train thunders in brakes squealing adding togarble jack-hammer cell-phones chatter

eyes closed he persists with fingers and bow swaying to an inner voice

the plaintive wail sad mysterious almost human rises dips

rises dipsweaving through the din

a temporary structuring of the cacophony

Tunney Lee is a Professor Emeritus in UrbanStudies and Planning([email protected]).

–Tunney Lee


22

Vijay KumarSpaces, Software, and Services –Supporting Educational Innovation and Sustainability with Technology

ACAD E M IC COM PUTI NG, PART OF

MIT’s Information Services & Technology(IS&T) organization, is committed to sup-porting faculty in creating innovative usesof technology for education. Tied to thiscommitment is Academic Computing’sobjective to enable and assert MIT’s globalleadership in educational technology, byleveraging its expertise and experiences.

Looking ahead to the coming aca-demic year, we will see major growth andtransitions in the spaces, software, andservices which are key to supporting edu-cational technology, with a particularfocus on sustainability. Following aresummaries of five of AcademicComputing’s current inivtiatives charac-terizing this growth and transition.

Spreading the Word – CollectivelyOne of Academic Computing’s goals is tofind better ways to communicate tofaculty what educational technology serv-ices are at MIT, and to report on projectsand activities. The result is two new Webresources. The Teaching with Technology(http://web.mit.edu/teachtech/) Websiteis designed to make it easier for faculty tofind the educational technology servicesprovided by numerous organizationsthroughout MIT. A new interactive,online newsletter, the Ed Tech Times(http://edtech.mit.edu/times/), replacesthe old paper Insider. Ed Tech Times pub-lishes the latest information on educa-tional technology projects, services,events, and trends.

Other outreach for the coming yearincludes the popular Crosstalk(http://web.mit.edu/acs/Crosstalk) seminarseries, and the second Ed Tech Fair.

Crosstalk gives MIT faculty and the largerMIT community a forum for informationand intellectual exchange about educa-tional technologies. Following the successof the first MIT Ed Tech Fair, held duringIAP 2003, another Ed Tech Fair is being

planned for November 2. This will be anopportunity for MIT faculty, students,and others working on educational tech-nology projects to demonstrate their workto all of the MIT community.

From Computing Clusters to FlexibleLearning SpacesTrends in student use of laptop comput-ers, commercial advances in software andhardware, and changing pedagogicalpractices challenge MIT to provide spacesmore aligned with contemporary learningand technology needs. Surveys of MITstudents and faculty have substantiatedthese needs on the campus. Redesigns of afew traditional Athena clusters as well asother computing spaces were launchedthis summer. These are pilot efforts toexplore different approaches for support-ing student computing, and to provideinformal and flexible learning spaces.Parts of four public clusters (W20,56-129, E51-075, and Hayden Library)were renovated to accommodate studentsneeding group collaboration space.

Featuring comfortable soft seating, largewall-mounted flat-panel displays, andWeb-enabled white boards, these spacespromote cooperation and teamwork as anessential part of an MIT undergraduate’seducation.

Academic Computing will be hostingtours of these clusters and other newlearning spaces on campus in September.To join a tour or for more information,send e-mail to [email protected]. Therenovated clusters and other MIT newlearning spaces will be featured in aSeptember workshop on learning-spacedesign, co-sponsored by the NationalLearning Infrastructure Initiative (NLII)and MIT (see http://www.educause.edu/nlii/keythemes/spacedesign.asp).

Undergraduate High-PerformanceComputing (HPC) ArrivesEmerging interdisciplinary computa-tional requirements are driving faculty tolook for high-performance and high-throughput computing resources for theirteaching and research. In response,Academic Computing has started a pair ofprojects to shape HPC services for teach-ing. The first is a Website to support acommunity of practice for those inter-ested in high-performance parallel com-puting (http://stellar.mit.edu/S/project/

Emerging interdisciplinary computational requirements aredriving faculty to look for high-performance and high-throughput computing resources for their teaching andresearch. In response, Academic Computing has started apair of projects to shape HPC services for teaching.


23

computationallyinten/) (HPPC). TheHPPC Website serves as a focal point forinformation exchange and referencematerials related to clustering hardwareand software technology, as well as aforum to guide decisions regarding hard-ware and software procurements.

The second project leverages MIT’scomputing infrastructure in two pilot,undergraduate HPC teaching clusters.One cluster has Intel hardware runningRed Hat Linux (the ROCKS Beowulf dis-tribution), and the other cluster hasApple PowerPC hardware running MacOS X (using the Sun-grid engine). TheAthena student computing space in 4-035was repurposed to house the LinuxBeowulf cluster; the Mac OS X cluster islocated in Building W91, and is accessibleremotely. The implementation of thesetwo HPC pilots in classes this year willprovide information about what isneeded to effectively use and supportthese systems.

Prof. Dave Darmofal, of Aeronauticsand Astronautics, intends to use theBeowulf cluster for his 16.100 class thisfall. In the spring, Profs. Gerd Ceder andNicola Marzari, both in Materials Scienceand Engineering, will be trying it out withtheir students.

The Apple HPC cluster uses OKIauthentication and authorization, andfocuses on specific applications such asgridMathematica, a parallel implementa-tion of Mathematica. Prof. Jim Elliot,from EAPS, and Prof. Raul Radovitsky,from Aeronautics and Astronautics, areamong those considering use of thisresource for their classes in the spring.

Faculty who are interested in learningmore about the HPC clusters, or usingthem for their class work, should contactPhil Long, [email protected].

Open Source – The Future of OnlineLearning ToolsMIT’s collaboration with the University ofMichigan, Indiana University, andStanford University on the Sakai(http://www.sakaiproject.org) projectprovides direction for the on-going workand sustainability of Stellar.

Supported by the Mellon Foundation,Sakai will develop and share open-sourcesoftware for learning tools based onservice specifications from OKI. Whilethere will be limited pilots of Sakai tools infall 2004, the production implementationof a Sakai-based course managementsystem (CMS) is planned for fall 2005.The Sakai CMS will include new features,plus tools and applications developed incollaboration with other universities andcommercial vendors. The framework willultimately support new tools and servicesdeveloped by MIT faculty and staff, suchas an online gradebook that promises tomake the grading process easier for MITfaculty.

Shop Online for EducationalTechnology ProjectsTo help faculty “shop” for technologywhich might support particular pedagog-ical goals, Academic Computing is creat-ing an online catalogue of educationaltechnology projects at MIT.

We are beginning to inventory suchprojects in close collaboration withMITCET and the Teaching and LearningLab. The inventory seeks to identify anddescribe both large and small-scale uses ofeducational technology. The collecteddata will take advantage of the CarnegieFoundation’s Snapshot tool to displaydetails about each project’s pedagogicalgoals, technical information, and assess-ment outcomes.

The new initiatives described above areby no means the only activities keepingIS&T Academic Computing busy in the

coming months. We continue to maintainand update third-party software for theAthena computing environment, offerspatial data and GIS services in conjunc-tion with the MIT Libraries, providefaculty consulting services through ournewly named Educational TechnologyConsultants (formerly know as theFaculty Liaisons), and offer training tofaculty in the use of the educational tech-nologies available to them.

The period ahead will be characterizedby an assessment of academic computingpriorities as Academic Computing, collab-orating with other groups engaged in sup-porting educational technology, shapesnew services and responds to the sustain-ability needs of new initiatives. Faculty andstudent input is paramount to our effortsin identifying and building a rich andenduring ecosystem for educational tech-nology. We welcome and invite your feed-back. If you would like to find out moreabout any of the services or projectsoffered by Academic Computing, or tosend us your thoughts, feel free to contactthe Educational Technology Consultantsat x3-0115, [email protected].

Quick Survey on PDA Support for Faculty

Information Services and Technology (IS&T) is exploring faculty interest to extend our serv-ices for and use of handheld devices – PDAs (Personal Digital Assistants) for access to MITservices such as e-mail, TechTime, and the Web. At this time, IS&T is building support for e-mail access via devices such as Blackberries and Treos, and we already offer support forPDAs running PalmOS and PocketPC (please see http://web/ist/topics/hardware/pda).

We would be grateful for a few minutes of your time to provide feedback to guide our effortsgoing forward. If you would like to weigh in on the usefulness of such support and to sharewith us your interest in devices beyond those already offered, please fill out the feedbackform at http://web.mit.edu/ist/survey/pda/.

Vijay Kumar is Director of AcademicComputing, Information Services andTechnology ([email protected]).


24

Kathleen CahillEdward Barrett

Web Accessibility: What Faculty Should Know

A S I N C R E A S I N G N U M B E R S O F

course materials migrate from paper tothe Web, the issue of equal access for allbecomes more than just a trite saying. Forpeople with disabilities, especially thosewith visual, hearing, or motor disabilities,the World Wide Web presents an oppor-tunity to find and read materials they maynever have had access to previously. Howto author a Web page so that all users canretrieve the same information is a vitaland important part of Website design. The2000 U.S. Census estimates that approxi-mately 15-20% of residents surveyed reporteda disability.

Take an example of a blind computeruser. Many blind and visually impairedusers have screen readers installed on theircomputers, which read text out loud. Thescreen readers can only parse text, whichmakes it important for Web pages to havetext equivalents (also called ALT text) forgraphics, pictures, and other non-textualinformation. Some people with disabili-ties use assistive technology (such as thescreen reader described above) to helpthem access a conventional computer.Other examples of assistive technologiesinclude screen magnification software,voice recognition software, head pointingdevices, eyegaze devices, or refreshableBraille devices. Some of these are availableto try out in the ATIC Lab (seehttp://web.mit.edu/atic/www/).

The World Wide Web Consortium(W3C), based here at MIT, has developedWeb Content Accessibility Guidelines(http://www.w3.org/TR/2004/WD-WCAG20-20040311/) to assist Web developers. MITuses the Web Content AccessibilityGuidelines in its own MIT AccessibilityPolicy and Guidelines (http://web.mit.edu/atic/www/sw/) which have been in placesince 1999. However, many Webmasters,or those who do Web page updates, useWeb editing software such asDreamweaver or Home Page, and do not

know how to code in HTML for accessi-bility. Luckily, Dreamweaver, which is anMIT-supported product, has a built-inaccessibility checker that allows a Webpage author to make sure a Web page is asaccessible as possible.

There are federal laws that apply toaccessibility of programs and servicesoffered by entities receiving federal funds.MIT is one such entity. Section 504 of theRehabilitation Act of 1973 requires thateducational programs be made accessibleto people with disabilities. That includesaccessibility of information and materialspresented, be it over the World Wide Webor on paper.

For Web accessibility, the issue oftenbecomes one of consciousness raising andeducation. Most Web developers wouldnot knowingly design an inaccessible pageand once given the information, arewilling to make the necessary changes.And the changes are not difficult ones tomake. Many Website changes involveadding ALT text, making links moredescriptive, making sure the Web page canbe navigated with the keyboard, and usingcolors and fonts that are easy to read.

One of the biggest challenges involved inWeb accessibility is making non-textualinformation accessible, be it online video,audio,simulations,or graphs and charts (seehttp://web.mit.edu/atic/www/sw/develop-web.html to find out more about Web pageaccessibility and Adobe™ PDF accessibility).

Some of the MIT ATIC lab staff serveon the Information Systems andTechnology Usability Team, which assistsdevelopers of Websites and applications inevaluating ease of use. ATIC lab staff havemade presentations on Web accessibilityfor various departments and groupsincluding 21W.785, Communicating inCyberspace, (instructor, Dr. Ed Barrett).In this class, students work in small col-laborative groups proposing and imple-menting a variety of Websites.

Early in the semester, ATIC Lab WebAccessibility experts visit the class whenstudents are beginning the design process.A Web Accessibility expert demonstrateshow a blind person interacts with severalmainstream Websites through a screenreader. What students hear is a cacophonyof sounds emanating from his laptop ashis screen reader attempts to voice infor-mation from sites that have not beendesigned in accordance with Web accessi-bility guidelines.

The Web Accessibility experts thendemonstrate several sites designed inaccordance with Web accessibility guide-lines. Compliance with these guidelinesnot only allows visually impaired users toretrieve information more easily, but alsostrengthens basic information architec-ture within the site, making it more coher-ent and organically related to interfacedesign elements sighted users perceive ontheir screens.

So, in summary, what should a facultymember do to make their course Websitesand information more accessible?

• Perform a 5-minute quick check foraccessibility,located at http://web.mit.edu/atic/www/sw/developweb.html#check.

• If you are composing a page inDreamweaver, check the accessibilityof your page in File > Check Page >Check Accessibility.

• If you use PDF files on your site,please take a look at information onmaking PDF files accessible:http://web.mit.edu/atic/www/sw/developweb.html#pdf.

If you need further help, please contactthe ATIC Lab at 253-7808 [email protected]. We would be happy toreview your site and offer feedback.

Kathleen Cahill is Team Leader, AdaptiveComputing (ATIC) ([email protected]). EdwardBarrett is a Senior Lecturer, Program in Writingand Humanistic Studies ([email protected]).


25

Mary P. RoweOpenCourseWare UpdateWhat Was it Like Working with OCW?

I N 1985, I COU LD NOT find a courseanywhere that linked negotiation theorywith conflicts within the workplace. So Iset out to apply the theories of my Sloancolleague, Bob McKersie, to my work as anorganizational ombudsman. Most of thetheory and practice of modern-day nego-tiation draws on the 1960s work ofRichard Walton and Robert McKersie.Their theories have been widely applied –to world peace, the martial arts andmodern dance, corporate strategy,courtship, and real estate transactions –but my course appears to have been thefirst on “Negotiation and ConflictManagement.”

In the spirit of McKersie’s work,“Course 15.667: Negotiation and ConflictManagement” was designed to delivertheory to practitioners and to developtheory from practice. And, followingMcKersie, my course deals with intangi-bles (how do people feel about the negoti-ation, about the conflict) as well astangibles (what will they get). The courseis realistic (and messy) since my cases aredrawn from real life. (There are no“optimum solutions,” just ranges of betterand worse solutions depending on thenegotiators and the setting). In the spiritof MIT, I have taken an integrated systemsapproach to organizational conflict man-agement. Over the years, I have con-tributed a bit, here and there, both tonegotiations theory and conflict manage-ment theory.

Fortunately, this subject matter turnedout to be very popular. However, like otherlong-term faculty, I now get calls and e-mails from colleagues all over the world.Having my course published on theOpenCourseWare Website lets me just refermany teachers, practitioners, students,alums, and perplexed negotiators to OCW.

So, what was it like, working withOCW?

I figured the process of publishingwould be awful. I mentioned that thecourse is “messy,” and so, I thought, weremy files. I have revised the course ~20percent each year, so my computers werefull of stuff – cases, brief negotiation“recipes” I had developed, outlines for lec-tures, an Introductory Test, instructionsfor a “Negotiations Journal,” and advi-sories for negotiators. I thought gloomilythat it would take a year to organize thesecourse materials into a coherent Website.My spirits sank lower when I got a callfrom OCW asking, “Could you deliver15.667 in a week?”

I thought this was hopeless. Maybe Iwould get it done in 2010?

But a long-suffering OCW departmentliaison – a recent MIT alum with goodnegotiations skills who could not be dis-couraged – came and downloaded every-thing, and then organized it overnightinto the distinctive OCW template.Presto! The course looked much betterthan ever before.

I edited, over a weekend on e-mail,with the patient OCW wizard who keptindefatigably encouraging me. It got done.(Of course it is never done. But I now havean exceptionally useful framework for myteaching notes, and for adding new mate-rial. And some of the burden on our won-derfully able office assistant has been lifted– I am no longer asking her quite as oftenhow to find something.)

In the last several months since 15.667was published on OCW, I have welcomednew colleagues who visit or call fromaround the world, having found thecourse on OCW. (These are people Iwanted to meet – OCW takes care ofrandom e-mail). A few cloned courses are

springing up around the world. Studentsuse fragments from the site. Some MITalumni tell me they are happy to see15.667 course materials available for themto teach, and use, at work. OCW has alsosaved many hours that would have beenspent in correspondence with conflictmanagement practitioners and theirsenior managers.

I was concerned about intellectualproperty. To my relief, this concern disap-peared with all the others. OCW took careof the permissions process for third-partymaterials I had borrowed for the course,just as Graphic Arts had done for me inthe past.

In addition, I am also happy about thefate of my own intellectual property. I amsuddenly getting more credit, includingacknowledgement for a number of myideas from the past. So one happyoutcome was that I found one can actuallyget more recognition for one’s work viaOCW. Moreover, people are much morelikely to find something I have written onOCW than on my office Website. (Ofcourse this new recognition is sometimesexpressed in quixotic terms – “you meanyou came up with that concept?”)

So – overall – what was it like workingwith OCW? I have great respect for OCW.For a member of the faculty gettingstarted or an instructor who is reallyworking to pull together a life’s work, thisseems to me a very unusual gift from MIT.

If you are interested in participating inthe OCW, please contact Jon Paul Potts,OCW communications manager, [email protected] or 617-452-3621.

Mary P. Rowe is an Adjunct Professor ofNegotiation and Conflict Management, SloanSchool; Ombudsperson ([email protected]).


26

Organization and Conceptual OverviewThe investment policies governing theendowment are established by a standingcommittee of the MIT Corporation, theInvestment Committee, in accordancewith Section 16 of the Bylaws of theCorporation of the Massachusetts Instituteof Technology. The Investment Committeeis comprised of nine members of theCorporation, two of whom, the Chairmanof the Corporation and the Treasurer ofthe Corporation, serve as ex officiomembers. This Committee meets three orfour times annually to review investmentpolicies and monitor investment activitiesand results for the Endowment and otherassets of the Corporation. The implemen-tation of the policies approved by thisCommittee is carried out by the Office ofthe Treasurer through external investmentmanagement organizations. The Corpor-ation recently approved the formation ofthe MIT Investment ManagementCompany, effective July 1, 2004. Themajor impact of this change will be to addadditional members to the Board oversee-ing the management of the Institute’sassets.

Endowment assets are those assetsresulting primarily from gifts to theInstitute where the donors have imposedthe condition that only the income pro-vided by the investment of the gifts mightbe expended for the educational andresearch purposes of the Institute. Theseassets are known as True Endowmentassets. In addition, the ExecutiveCommittee of the MIT Corporation maydesignate otherwise expendable funds tobe managed as Endowment funds. Theselatter funds are known as Quasi-Endowment (or Funds Functioning asEndowment). Either of these two classes ofendowment assets may be restricted as topurpose of use, such as strictly for scholar-ships or fellowships, for professorships, orfor research; or they may be unrestrictedas to purpose of use, in which case theyare known as Unrestricted Endowmentand are available for any general purpose

of the Institute. Since the major focus ofthis discussion is on the management ofthese assets, rather than the use of thesefunds, we will not focus further on thesedistinctions, other than to note that it isnecessary for the Institute to carefullybalance its budgetary needs with thenature of the resources available to it fromthe endowed assets.

Another aspect of this discussion is tonote that the primary investment pool forthe Endowment is known at MIT as PoolA. Although Pool A is neither the com-plete Endowment nor is it comprised onlyof Endowment assets, it is for this discus-sion a good proxy to discuss the manage-ment of the Endowment and the annualdistributions that support operations. Adepartment’s funds that are invested inPool A will receive annual distributionsbased on the number of units held in thePool. The units are valued monthly andnew gifts or other additions to the Pool arecredited with new Pool A units based onthe current month’s unit market value. Inmany respects this is similar to the cre-ation of shares in an open-end mutualfund, about which many may have directpersonal experience.

The Pool A assets are managed tomaximize total investment return relativeto appropriate risk, which in this caserefers to the volatility of returns. Thechallenge is in developing investment and

distribution policies which over timesupport the operations of the Instituteand protect the underlying purchasing

power of the endowed assets. Investmentincome earned (in the form of interest,dividends, and rents) and a portion ofinvestment gains are distributed in amanner that, over the long term, allowsfor the retention and reinvestment of anamount at least equal to the inflation rate,thereby satisfying the legal obligation toendowed funds.

Financial Strength of the InstituteBefore more completely describing themanagement of these assets, we empha-size that the Institute is a financiallystronger institution today than it was20, or even 10, years ago. A fewnumbers describing the general invest-ments of the Institute, the bulk ofwhich is the Endowment, will illustratethis point. In presenting this informa-tion, and other data to follow, we notethat the financial records for Fiscal 2004will not be finalized until October 2004.In addition, some information on peerand benchmark measures are not yetavailable for Fiscal 2004. We are there-fore presenting both 2003 and 2004information where available, not tosuggest that the short-term perform-ance should be the basis of presenta-tion, but rather to provide bothcomparable information and the mostcurrent information possible.

A tripling of the ratio of total invest-ments to operating expenditures over thepast 20 years is only one of many measures

Management of the MIT EndowmentBufferd, from page 1

FiscalYear

Total Investments $ Million

Total OperatingExpenditures

$ Million

Ratio ofInvestments/Expenditures

1983 767.3 588.7 1.303

1993 2,126.1 1,133.9 1.875

2003 6,174.1 1,686.6 3.661

2004 7,251.9 (est) 1,844.1 (est) 3.932 (est)

Table IRatio of Total Investments to Total Operating Expenses for Selected Fiscal Years


27

that demonstrate a financially strongerinstitution than in prior years. This hasoccurred as a result of two enormouslysuccessful capital campaigns, theCampaign for the Future, which raised$710 million in 1987 to 1992, and theCampaign for MIT, now closing in on $2 billion in gifts and pledges over theperiod 1999 to 2005. The generally favor-able investment climate for much of thelast 20 years is another important factor, aswell as an appropriate balance betweencurrent and future needs through theapplication of a moderate distribution rateand the reinvestment of a portion of theinvestment return for future generations –the intergenerational equity issue. This isnot to suggest that this institution does notface continuing challenges on both anabsolute and relative basis. It is only toemphasize that we face this future muchstronger than we did a few years ago.

Endowment Investments, Past andPresentFigure 1 presents the investments of PoolA as of June 30, 2004. Pool A assets were$6,141.8 million, an increase of $748.1million from the $5,393.7 million of June30, 2003. This change in market value,13.9%, is not the investment return for theyear. It is the change in asset value which iscomprised of income, market valueappreciation (or depreciation), new giftsadded to the investment pool, lessamounts distributed for spending.Investment return is that measure ofchange in value due solely to investmentresults and does not take account of newgifts or of the amount distributed forspending. For the fiscal year just closed,the investment return was 18.1%. Sincethe amount distributed for spendingalways includes a portion of gains, thechange in asset value will always lag theinvestment return, unless gifts receivedare unusually large. In reviewing Figure 1and other comments on the asset classescomprising Pool A, reference may bemade to the Glossary (back page) for abetter understanding of the asset classes inwhich MIT invests.

The overall policy allocation of Pool Ahas changed over the past 10 years, aperiod of time during which economicoutlook went from euphoria to malaise.The Pool A strategic asset allocation isreviewed annually by the InvestmentCommittee and changed only modestly asa result of periodic studies which incorpo-rate a broad range of inputs on expectedreturns, volatilities and correlationsamong the asset classes considered forinvestment. These portfolio optimizationstudies, including Monte Carlo simula-tions of longer term expected values forthe endowment, are inherently sensitive tothe quality of the inputs and tend toreflect recent performance results andoutlook. The range of inputs from con-sultants, investment managers, and otherinstitutional investors helps to provide ageneral guideline to an “optimized” port-folio, optimized in the sense of expectedreturn and volatility (risk). The optimizedportfolio further incorporates boundaryconditions and other practical constraintsthat reflect a sense of the execution skillsof staff and the viewpoints of the staff andthe Investment Committee as to wherethe best risk-adjusted investment return

will be realized. The objective of this effortis the construction of a portfolio withassets showing low correlation with eachother, to minimize overall portfoliovolatility while maximizing expectedportfolio return over time. The mostuncertain aspect of this approach is theshort-term volatility.

Table II shows the changes in the assetallocation over the past one, three, fiveand 10 years back to Fiscal 1994. Someimportant changes in outlook arereflected in these allocations.

While there are few unidirectionalchanges during the past 10 years, therenow is a substantially decreased exposureto the domestic equity markets and fixedincome markets, and a significantlyincreasing exposure to marketable alter-natives (see Glossary), private equity(both venture capital and private capital,see Glossary), and real estate during thisperiod. As described above, these changesreflected policy guidelines and outlooks asreviewed periodically with the InvestmentCommittee of the Corporation.

In practice, we are guided by a policyportfolio with allocation weights centered

Figure 1Pool A Asset Allocation on June 30, 2004 (See Glossary)


MarketableAlternatives

18.4%

Cambridge Real Estate

7.9%

R. E. Pools2.2%

Cash2.3%

Fixed Income6.8%

Real Assets2.6%

U. S. Equity26.4%

International Equity15.0%

Private Capital7.7%

Int’l Private Capital3.6%

Venture Capital7.0%


28

on a single value for each asset class, butwe allow for variation within an accept-able band around that value. The Pool Apolicy portfolio and the acceptable bandsfor fiscal 2004 are shown in Table III.

The very slight differences between actualallocations and the policy range for U.S.Equity primarily reflects implementationissues with regard to allocating funds toexternal managers and to some degree theresult of market action. Significant out-of-range allocations are rebalanced periodically.

PerformanceHow does the MIT asset allocation comparewith that of other colleges and universities?This information is collected periodically byCambridge Associates, a consulting organi-zation with whom MIT has worked formore than 25 years. Figure 2 compares the MIT Pool A Allocation to that of theCambridge Associates Mean Allocation at

the end of Fiscal 2003. Peer information forFiscal 2004 is not yet available.

In general, our underweight in GlobalEquity (Domestic and InternationalEquities) and Global Bonds (FixedIncome) is offset by an increased alloca-tion to Non-Marketable Alternatives(Private Equity) and Real Estate. Such dif-ferences are less pronounced in compari-son to the largest endowments in thesurvey group, as they have a greater frac-tion of their assets in alternative invest-ments and a commensurate reduction inthe marketable securities areas.

Figure 3 presents a comparison of theinvestment results for each Pool A assetclass to its respective benchmark for thefive years ended December 31, 2003, thelast date at which all benchmark informa-tion was available for the broad range ofalternative assets.

All the benchmarks against which thePool A performance is measured arewidely accepted standard benchmarks,with the exception of that forInternational Equities, which was customdesigned by MIT with the assistance ofProfessor Stewart Myers of the SloanSchool for the purpose of deemphasizinglarger capitalization markets in the inter-national sector and affecting externalmanager behavior. Since adoption in

Management of the MIT EndowmentBufferd, from preceding page

Asset Class 2004 2003 2001 1999 1994Fixed Income & Cash 9.1% 11.5% 15.1% 15.5% 17.6%Real Assets 2.6% 1.0% 0.0% 0.0% 0.0%U.S. Equity 26.4% 26.6% 38.5% 47.2% 49.3%International Equity 15.0% 7.5% 7.7% 10.8% 20.3%Private Equity 18.4% 18.3% 19.0% 15.0% 8.0%Marketable Alternatives 18.4% 20.3% 8.7% 4.6% 0.0%Real Estate 10.1% 14.8% 11.0% 6.9% 4.8%

Table IIPool A Asset Allocation for Past One, Three, Five and Ten Years

Asset Class Target Allocation Acceptable Range

Fixed Income 10% 5-15%

Real Assets 5% 0-10%

U.S. Equity 21% 16-26%

International Equity 14% 9-19%

Private Equity 20% 15-25%

Marketable Alternatives 20% 15-25%

Real Estate 10% 5-15%

Table IIIFiscal 2004 Pool A Target Allocation and Acceptable Ranges

34.1

9.3

18.120.3

14.8

3.4

0.0

46.6

19.2

8.5

18.6

3.02.2 1.9

0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

40.0

45.0

50.0

Global Equity Global Bonds Non-MarketableAlternatives

MarketableAlternatives

Real Estate Cash &Equivalents

Other

Pe

rc

en

t o

f T

ota

l A

ss

ets

MIT Pool A Cambridge Associates Mean

Figure 2Comparative Asset Allocation of MIT Pool A to Cambridge Associates Mean

as of June 30, 2003


29

1998, the MIT Custom InternationalBenchmark has exceeded the index morecommonly used in the investment man-agement field, and our managers haveoutperformed the Custom Benchmark. Itis to be especially noted that for each ofthe asset classes the Pool A managers haveoutperformed the respective benchmarksfor the five-year period, a few by quite sig-nificant margins.

Another way to review the longer termhistory of the investment results is tocompare the aggregate results to a proxyfor the portfolio. In this case we haveselected a composite of 85% in thedomestic equity market, measured by theS&P 500 Index, and 15% in the domesticbond market, measured by the LehmanAggregate Bond Index, as well as themedian return of all funds in the TrustUniverse Composite Service (“TUCS”),

the broadest comparison of institutionalasset pools, over the period 1994 through2004. We chose the 85/15 proxy as Pool Ahas been comprised of 85% equity orequity-like assets. Figure 4 illustrates thesecomparisons.

Recent comments about endowmentperformance over this period recognize thatthe Institute participated significantly in thestrong markets of the late 1990’s throughthe early part of Fiscal 2001, substantiallyoutperforming a broad universe of compa-rable funds and in particular a portfolioproxy without any alternative assets partici-pation. However, the fall in relative andabsolute performance for the past threeyears, as illustrated in Figure 4, has raisedconcerns about the investment policy andalso the distribution policy which has beenfollowed. We also have had inquiries aboutdistribution models that may be used byother major private universities.

While there have been adjustmentsover the past few years in the allocation ofassets, the general tenor of our policy hasbeen constant over the past three years.The recovery noted in this past yearemphasizes the underlying rationale ofthe Pool A policy portfolio.

Let us examine the history of the dis-tributions from Pool A with thesethoughts in mind.

2.7

20.0

6.9

10.5

14.9

30.1

-0.1

7.1

3.3

9.58.6

12.1

-5.0

0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

US Core Equity US Small-Cap InternationalEquity

Total RE MarketableAlternatives

Private Equity*

An

nu

aliz

ed R

etu

rn

Figure 3Asset Benchmarks and Results for the Five Years Ended December 31, 2003

MIT Benchmark

$0

$100

$200

$300

$400

$500

$600

FY1994 FY1995 FY1996 FY1997 FY1998 FY1999 FY2000 FY2001 FY2002 FY2003 FY2004

Figure 4Comparison of Investment Returns from Fiscal 1994 through Fiscal 2004

for Pool A, the TUCS All Master Trust Median, and an 85%/S&P/15% Lehman Aggregate Bond Indices Proxy


Asset Class BenchmarkU.S. Core Equity Russell 1000Small Cap Equity Russell 2000International Equity MIT CI20% EM/MSCI EAFETotal Real Estate NCREIF (1 qtr. lag)Marketable Alternatives T-Bills Plus 5%Private Equity* Cambridge Associates (1 qtr. lag)

*The Private Equity returns for MIT and the Benchmark are calculated by linked quarterly IRR’s in this graph. The annualized five-year point-to-point IRR returns would be 98% for MIT and 31% for the Benchmark.

Endowment returns 85% S&P - 15% Lehman Agg TUCS Median


30

Distribution PolicyOver the 28-year period from 1970through 1998, the annual distributionfrom the endowment grew 6% annually,with no decrease from year to year. Thepolicy guidelines over this period varied asthe respective Investment and ExecutiveCommittees looked to formulations asguidance in balancing the need forcurrent resources against sustaining thelonger term purchasing power of theEndowment. In addition, from 1973 to1998 there was only one annual increasegreater than 10% (10.4% in 1989) in thedistribution rate. During this period spe-cific formulas were used as guides, ratherthan being followed explicitly. The most

recent formulation sets the distribution ratein the range of 4.75% to 5.5% of the lagging36-month average unit market value.

This formulation and the sharplyincreased unit market values due to thestrong investment performance in the late1990’s resulted in the distribution increas-ing at an average annual compoundedgrowth rate of more than 19% from 1998to 2003, as illustrated in Figure 5.

This rate of growth in the distributionrate coincided with the very robustmarkets and a call for increased resourcesto fund initiatives outlined by PresidentVest’s 1998 report, entitled The Path to theFuture, and described in more detail by

Provost Brown in his article titled“Financing MIT” in the December/January 2004 issue of the MIT FacultyNewsletter (Vol. XVI No. 3). As the base-line distribution indicated by the formula(the long-term distribution rate appliedto the three-year average market value)

did not meet demands for resources out-lined in that report, the balance was sup-ported by an additional allocation by theExecutive Committee of $500 million ofEndowment de-capitalization (technicallya draw on unrestricted quasi-endowment).That program, authorized for the period2000-2010, is being implemented on thatoriginal schedule.

Comparison to Other DistributionPoliciesThere are many other distributionmodels in use at peer endowments whichserve either as firm formulae or as indica-tive guidelines. These include a cap on the

annual rate of change in the distributionrate, a model followed by at least onemajor endowment during the “bubble”period. Another model defines the distri-bution as a weighted average of prioryear’s spending and a percentage of thisyear’s market value, a so-called 70/30formula. The formula generally uses the“spot” market value, as the prior year’sspending itself incorporates prior yearmarket values (through a recursive appli-cation of the formula). This approach iseffectively a modified exponentialweighting of market values, adjusted forinflation.

Figure 6 presents a comparison ofwhat the MIT distribution would havebeen with the 70/30 formula applied tothe Institute’s actual investment resultsfrom Fiscal 1973 through Fiscal 2005. Thespot unit value used here is of theDecember preceding the fiscal year forwhich a budget is being prepared, as theInstitute’s budget is finalized in the wintermonths preceding a fiscal year. While theresulting distribution rate would havebeen smoother from 1998 through 2005than we experienced under our currentformula, the distribution would have beencumulatively almost $26 per unit less thanwhat was historically distributed, half ofwhich was post-1998. To the extent opera-tions and the new initiatives were tofollow their original time plan, the drawon quasi-endowment assets would havehad to increase by that amount. With anominal seven million units in theEndowment over the 1998-2005 period,the difference of $13 per unit distributionwould have resulted in an additional $85

Management of the MIT EndowmentBufferd, from preceding page

$-

$10

$20

$30

$40

$50

$60

FY70 FY75 FY80 FY85 FY90 FY95 FY00 FY05 FY10

$/unit

Figure 5Pool A Distribution per Unit from 1970 to 2005

Even with the subsequent decrease in the distribution rateover the past two years, the distribution rate from 1998through 2005 will have grown at an annually compounded rateof over 10%.

Historical unit payout 6% average growth rate ('70-'98) 8% modified target growth rate


31

million of draw on unrestricted quasi-endowment funds.

Even with the subsequent decrease inthe distribution rate over the past twoyears, the distribution rate from 1998through 2005 will have grown at an annu-ally compounded rate of over 10%. Whilethe purposes for which the increased rateof change in the distribution was effectedwere appropriate and directed to impor-tant initiatives, there is the question ofwhat can we reasonably expect for thefuture rate of growth in the distribution?

This longer-term outlook is suggestedby Figure 5, above. The average annualgrowth rate in the distribution rate was6.15% for the 1970 through 1998 period.This period included years of economicstrength, of economic weakness, of highinflation, and of low inflation – essentiallya full spectrum of economic results. Basedon this history, it seems difficult to expectthat distribution rates should grow onbalance much more than 6% annually. Ona rough, intuitive level, the 6% growth in

distribution is the net of investment returnof the Endowment, less the annual distribu-tion. In those terms, the average invest-ment return of the Endowment was 11%over that period. This is only a rule-of-thumb, as the process of calculating thedistribution incorporates a 36-monthaverage of market values, and the distri-bution rate as a percent of market valuevaried over that period. But as such, itgives a ballpark figure for us to considerthe impact of the current market environ-ment and a 5-6% effective distributionrate (incorporating the impact of the useof quasi-endowed funds raises the effectivedistribution rate). To continue the com-parison, Figure 5 shows two projections,one at 6% and one at 8%. It is to be notedthat had the 1998 distribution rate beenincreased at a 6% rate from 1998 forward,the fiscal 2005 distribution would havebeen $26.40 per unit, and the projection at8% would have yielded a $30.10 per unitdistribution, in comparison to the $36.00per unit currently being distributed.

The challenge for the future is toachieve an appropriate coupling of thedistribution rate to a longer term view ofwhat can be expected from investmentresults, and by doing so, balance thefunding of our current needs with ourobligation to ensure the financial, andultimately the operational, flexibility ofthe Institute.

We welcome the opportunity to discussthese issues further with interested membersof the community and thank the editors forproviding an opportunity for a broaderdiscussion of the investment program andits implications for support from theEndowment.

$(10)

$-

$10

$20

$30

$40

$50

1973 1977 1981 1985 1989 1993 1997 2001 2005Fiscal Year

Difference (MIT - X-Y)

X-Y model (70%-30%)

MIT Distribution

Cumulative total: $25.82 per unit

Figure 6Comparison of the Actual MIT Pool A Distributions Per Unit to Implicit 70%/30% Model

MIT vs. X-Y Payment PolicyFiscal Years 1973 to 2005

Allan S. Bufferd is Treasurer ([email protected]).

Editor’s Note: An exception was made tothe normal length restrictions of Newsletterarticles for the above report on the MITEndowment, due to its unusal content andinterest to the MIT community.


TH E ASS ETS OF TH E Institute’s majorinvestment pools are divided into eight sub-portfolios. This Glossary has been preparedto assist in understanding these eight port-folios and the assets and strategies inwhich the Endowment participates. [Seearticle, page 1.]

Fixed Income and CashThe Fixed Income and Cash portfolioemploys a conservative investment strategydiversified among U.S. Treasuries, U.S.Agencies, high-quality corporate debt, andmortgage-backed and other asset-backedsecurities, with a modest allocation to high-yield and non-U.S. bonds. This type of strat-egy is typically described as “core fixedincome.” The portfolio benchmark is theLehman Aggregate Bond Index.

Real Assets (TIPS) The Real Assets portfolio is designed tofurther diversify the assets and, more impor-tantly, to provide inflation protection. Theportfolio includes Treasury Inflation-Protected Securities (“TIPS”) other inflationhedges, commodity futures and naturalresource-related equities. The currentbenchmark is the Lehman U.S. TIPS Index.

U.S. EquityThe major portion of the U.S. Equity portfo-lio is focused on large-cap domestic equi-ties but includes the capability to invest upto 20% of the sector in large-cap interna-tional equities. Approximately 5% of thesector is currently invested in internationalequities.

The large-cap portfolio and an Energysector fund comprise about 75% of theU.S. Equity portfolio.

The balance of the U.S. Equity portfoliois the Small-Cap sector. We follow variousapproaches from broad diversificationwithin the sector, to size – and industry –specific strategies. The applicable bench-marks are the Russell 2000 Index and theAMEX Biotech Index.

International EquityThe International Equity portfolio has threedistinct sectors. The first is a large-cap allo-cation that is the largest commitment in thisportfolio and is benchmarked against aderivative of the MSCI EAFE Index, one thatwas developed and customized by MIT withthe assistance of Professor Stewart Myersof the Sloan School almost six years ago.The MIT Custom International Index consis-tently has outperformed the standard MSCIEAFE Index and serves as a higher stan-dard for our managers.

The second sector of the InternationalEquity portfolio is dedicated to emergingmarkets, debt and equity. The benchmark isthe MSCI Emerging Markets (Gross) Index.

The third sector of the InternationalEquity portfolio is comprised of internationalsmall-cap equities. One of the mandates isa long/short* strategy that was developedby and for MIT. The benchmark is theCitigroup EMI Ex US Index.

* In a long strategy, one purchases asecurity expecting to sell at a future time ata higher price. In a short strategy, one sellsa security not owned (“short”) hoping topurchase it in the future at a lower price.

Private CapitalThe Private Capital portfolio is comprisedpredominantly of investments in private, andusually non-technology, companies, domes-tic and international. The Institute was one ofthe early institutional investors in this strat-egy, executed through various partnershipswith selected managers. Originally termed“leveraged buyouts,” it is more correctlycharacterized as growth financing in additionto buyouts. The portfolio is highly diversifiedamong domestic and international man-agers, with very little overlap in specific hold-ings. There are a few direct positions as wellas some public holdings in this portfolio. Thisportfolio and the Venture Capital portfoliocomprise the area known as private equity.The Cambridge Associates Private EquityIndex is the benchmark.

Venture CapitalThe Venture Capital portfolio is comprisedpredominantly of investments in private,usually technology-based, domestic compa-nies. The Institute was one of the early institu-tional investors in this strategy, executedthrough various partnerships with selectedmanagers. This portfolio has more overlap inspecific holdings than experienced in thePrivate Capital portfolio. There are a fewdirect positions and some public companiesin this portfolio. This portfolio and the PrivateCapital portfolio comprise the area known asprivate equity. The Cambridge AssociatesVenture Capital Index is the benchmark.

Marketable AlternativesThe Marketable Alternatives portfolio is com-prised of diverse investment strategies inhedge funds and other marketable alternativeinvestments. The Marketable Alternativesportfolio includes areas such as: event andmerger arbitrage, distressed debt and credit-oriented funds; special situations; and long-short equity management. Approximatelyone-third of the portfolio’s exposure acrossthe strategies is outside of the United States.Most of the underlying investments are in mar-ketable securities, although some of theinvestments are private positions or subject tolimited liquidity due to the organizational struc-ture of the manager. The benchmark is the 91-Day Treasury Bill Rate Plus 500 Basis Points(i.e., a 5% return over the risk-free rate).

Real EstateThe Real Estate portfolio is 75% in Cambridgethrough land holdings and direct investmentsin office and laboratory space, more or lesscontiguous with the MIT campus. While thisreal estate has a strategic purpose, purchasecriteria include a sufficient return for theEndowment to commit capital. The two mostsignificant holdings are University Park at MIT,and the Technology Square development. Thebalance of the Real Estate portfolio is investedin diverse property types, domestic and inter-national. The NCREIF Property Index is thebenchmark.

The Management of the MIT EndowmentGlossary

Date post:	22-Jul-2018
Category:	Documents
Upload:	vothuy
View:	218 times
Download:	0 times

MIT Faculty Newsletter Vol. XVII No. 1 September/October …web.mit.edu/fnl/vol/171/fnl171.pdf ·...

Documents