A Conversation with George G. Roussas
aa r X i v : . [ s t a t . O T ] A p r Statistical Science (cid:13)
Institute of Mathematical Statistics, 2010
A Conversation with George G. Roussas
Debasis Bhattacharya and Francisco J. Samaniego
Abstract.
George G. Roussas was born in the city of Marmara in cen-tral Greece, on June 29, 1933. He received a B.A. with high honors inMathematics from the University of Athens in 1956, and a Ph.D. inStatistics from the University of California, Berkeley, in 1964. In 1964–1966, he served as Assistant Professor of Mathematics at the CaliforniaState University, San Jose, and he was a faculty member of the Depart-ment of Statistics at the University of Wisconsin, Madison, in 1966–1976, starting as an Assistant Professor in 1966, becoming a Professorin 1972. He was a Professor of Applied Mathematics and Director ofthe Laboratory of Applied Mathematics at the University of Patras,Greece, in 1972–1984. He was elected Dean of the School of Physicaland Mathematical Sciences at the University of Patras in 1978, andChancellor of the university in 1981. He served for about three years asVice President-Academic Affairs of the then new University of Crete,Greece, in 1981–1985. In 1984, he was a Visiting Professor in the Inter-college Division of Statistics at the University of California, Davis, andhe was appointed Professor, Associate Dean and Chair of the Grad-uate Group in Statistics in the same university in 1985; he served inthe two administrative capacities in 1985–1999. He is an elected mem-ber of the International Statistical Institute since 1974, a Fellow of theRoyal Statistical Society since 1975, a Fellow of the Institute of Math-ematical Statistics since 1983, and a Fellow of the American StatisticalAssociation since 1986. He served as a member of the Council of theHellenic Mathematical Society, and as President of the Balkan Unionof Mathematicians. He is a Distinguished Professor of Statistics at theUniversity of California, Davis, since 2003, the Chair of the AdvisoryBoard of the “Demokritos Society of America” (a Think Tank) since2007, a Fellow of the American Association for the Advancement ofScience since 2008, and a Corresponding Member of the Academy ofAthens in the field of Mathematical Statistics, elected by the member-ship in the plenary session of April 17, 2008.
Key words and phrases:
Personal and professional life, milestones,Marmara, Thessaloniki, Athens, Berkeley, Madison, Patras, Davis.
Debasis Bhattacharya is Professor of Statistics,Visva-Bharati University, West Bengal, India, and afrequent visitor in the Department of Statistics,University of California, Davis e-mail:debasis [email protected]. Francisco J. Samaniego isDistinguished Professor, Department of Statistics,University of California, Davis, California 95616, USAe-mail: [email protected].
This conversation took place in George Roussas’office at the University of California, Davis, on the15th of May, 2009.
This is an electronic reprint of the original articlepublished by the Institute of Mathematical Statistics in
Statistical Science , 2010, Vol. 25, No. 4, 566–587. Thisreprint differs from the original in pagination andtypographic detail. D. BHATTACHARYA AND F. J. SAMANIEGO
Fig. 1.
George Roussas (in the middle of front row) in the5th or 6th grade in Thessaloniki, 1944–1945.
EARLY YEARS AND FAMILY BACKGROUND
Debasis and Frank : George, it’s a pleasure to havethis opportunity to chat with you about your lifeand career. We’re coming at this conversation fromdifferent angles, one of us as a regular research col-laborator over the last ten years and the other asa long time departmental colleague. Our commonground is that we are both long-time friends andadmirers.Let’s start at the beginning. Tell us a bit aboutyour early days.
George : Let me say, first, that I’m greatly honoredthat you asked me to have this conversation, and Ihave been very much looking forward to it.I was born in the city of Marmara, broadly ina family of educators. Marmara is a small commu-nity (of maximum population of about 1350) on theGreek mainland. It is widely thought to be withinthe location of Achilles’ ancient kingdom. I attendedthe elementary school in Marmara and in Thessa-loniki, where part of my family was. My high schooleducation was also divided, started in Thessalonikiand completed in Athens. The schooling was highlystructured, as was typical in Greece, and very rigor-ous. The environment in Marmara was idyllic, andI still have very fond memories of it.
Debasis and Frank : What can you tell us aboutyour family background?My paternal grandparents had three sons (my fa-ther Gregory and my uncles Hercules and Constan-tine). These two uncles obtained university degrees,but my father was business oriented. In the early1920’s he left Marmara and went to Thessaloniki,where he entered in the dairy business. He had con-siderable success in this, holding a prominent place in the distribution business in Thessaloniki for al-most two decades. Indeed, he was a self-made mil-lionaire! My parents (Gregory and Maria) had fourchildren, daughters Aggeliki, Demetra, and Stella,and myself. Our family moved to Thessaloniki, andlater to Athens, “in stages,” as the locations wheremy father’s varied business interests were centeredchanged over the years. Of course, a civil conflict inthe country had its influence on this as well.My father was a product of the classical Europeanliberalism, which in the 1920’s and early 1930’s waswell represented in Greece by a remarkable states-man, Eleftherios Venizelos. My father was an activemember of the liberal party and a staunch supporterof Venizelos. He retained this position throughoutthe late 1930’s, when Greece was run by a politicianturned dictator, and during World War II, and later,when Greece was savaged by a civil war. His politicalliberalism and outspokenness cost him dearly. Dur-ing the German occupation of the country, he wassent to prison (released as a political prisoner at theend of the war, upon the liberation of the country).During part of the civil war, he was exiled to a re-mote deserted island by the governing party. Thefact that my father saw some glitter of light in therepressive soviet system, such as the availability ofabundant opportunities for well-qualified studentsto pursue their educational goals, did not sit wellwith the party in power at the time. From a financialviewpoint, he believed strongly in currency ratherthan in property. Consequently, his hard-earned “for-tune” became worthless with the nullification of theGreek currency during World War II, and a million-aire became virtually penniless! It was in this kindof environment in which I grew up. This environ-ment shaped my determination to distance myselffrom any business per se and to pursue education toits highest level possible.
BECOMING A MATHEMATICIAN WHILESEARCHING FOR ITHACA
Debasis and Frank : What were your main inter-ests when you entered college? Did you have strongfeelings about what you wanted to specialize in?
George : In high school, I developed a strong affin-ity to the humanities and social sciences, with margi-nal interest in mathematics and physical sciences.Soon, I discovered that any weaknesses in mathe-matical and physical sciences would deprive me ofmany options in later years. So, I decided to inten-sify my efforts, and graduated with a strong record
CONVERSATION WITH GEORGE G. ROUSSAS in all my subjects. This standing put me on a solidposition to compete for a position in the air forceacademy; it was my youthful dream to become anair force officer. But that dream would never come tofruition. In addition to succeeding in a competitiveexamination, I would also have to have the writtenconsent of both of my parents. I thought I could talkmy father into it, but my mother was adamantlyopposed to the idea. Instead, I was advised by myuncle Hercules (the dean of the classicists, as he wasoften referred to) to take the entrance examinationin the department of mathematics at the Univer-sity of Athens. Reluctantly, I took his advice, butI never checked the results of the entrance exami-nation. My fixation was still with the air force, andat this time, I targeted the aeronautical engineeringschool of the air force. However, there seemed to bea problem here. Namely, those competing for a posi-tion were more than 300, and the positions availablewere 6–8! In view of these imposing odds, my parentsdid not attempt to dissuade me from preparing forsuch a competition, and actually taking the exami-nation. Why should they? It was, clearly, a hopelesseffort! For about a year, I exhaustively studied mathand science. When the examination time arrived, Iwas an enthusiastic and determined participant. Inmilitary schools, the exams were taken serially, andonly the successful participants in one subject wereallowed to continue with the next subject. In thismanner, I reached the last examination in chemistry,which was taken by less than a couple dozen people.I later learned that I had earned the top overall scoreon the examination.And it was here when the drama began. Succeed-ing in the examinations was extremely tough, butthat was only part of the admissions routine. Thecandidates for all military schools, and, in partic-ular, for such an elite institution as the air forceaeronautical engineering school, also had to be cer-tified on their political beliefs, on the basis of severaldegrees of family connections. It was here where thesorry political past of my father entered the picture.As became known later, the disqualifying certificatearrived at the examination committee’s headquar-ters right after the examination papers in chemistrywere corrected. It was the duty of the committee toflunk me, no explanations provided. The chairmanof the committee, an air force colonel-engineer, tookit upon himself not to post the results. My uncleHercules, who was highly regarded and had manyinfluential acquaintances, tried vigorously to obtain an exception for me, unfortunately, to no avail. Hewas given to understand that there would be direpolitical consequences if the effort to gain my admis-sion to the elite school of aeronautical engineering ofthe air force succeeded. So, I was officially certifiedto be a . . . communist (!), and I was abruptly deniedthe realization of my dream. Debasis and Frank : Disappointing and demoraliz-ing! What happened next?
George : For more than a year, my odyssey insearch of my Ithaca went on, without much sat-isfaction. After quite some time, disappointed andshaken, I decided to visit the department of math-ematics of the University of Athens, just to inquireabout the previous year’s entrance examination. Iwas told that I was successful, but since I did not en-roll, I lost the right of enrollment. Fortunately, therewere a couple of openings in the following year’sclass, and one was allocated to me. Apparently, mymanifest destiny was to become a mathematicianrather than an air force officer-engineer! I graduatedfrom the University of Athens in four years with highhonors. During the last two years, I also served asa teaching assistant to professor D. A. Kappos, whowas a student of Constantine Carath´eodory and heldthe chair of mathematical analysis. I was fortunateto take many of my courses from him. While mystudies in mathematics were quite broad, I had notyet been introduced to probability or statistics.
CHOOSING STATISTICS—THE UNIVERSITYOF CALIFORNIA, BERKELEY YEARS
Debasis and Frank : That’s a fascinating story, Geor-ge. It’s interesting how sometimes bad things seemto happen for a reason. Certainly your ultimate ca-reer path is a good example. It’s intriguing that youdecided to pursue graduate work in a field that youhad yet to be formally exposed to. How did thatcome about?
George : It was my determination to pursue gradu-ate work abroad. The decision to go for statistics—despite the lack of any relevant background—wasdue to a liking I took in probability (by attendingan occasional seminar, and also by studying on myown), but primarily it was due to the advice of Pro-fessor Kappos. His own expertise was in measuretheory and probability in abstract structures. Theabsence of statistics from the curriculum was an ad-ditional reason. Since studying abroad was well be-yond my family’s financial means, another financial
D. BHATTACHARYA AND F. J. SAMANIEGO
Fig. 2.
A (nonabelian!) group of the mathematics graduating class at the University of Athens, fall 1956. (George Roussas,in the middle, kneeling). source would have to be located. Fortunately, theGreek government did provide some relevant fellow-ships, based on a series of written examinations andservice in the armed forces. So, I was inducted intothe army, where I served for two years as a private(an unusually low rank for a young man with mybackground), largely because I was still plagued bymy unfortunate experience with the air force. Butmy low rank army service did me some good, as Ispent almost the entire period close to home, and Ihad the possibility to pursue my study of the En-glish language. Soon after my discharge from thearmy, I participated in an examination for the selec-tion of fellows to study applied mathematics (whichincluded probability and statistics) abroad. Profes-sor Kappos insisted that, once the decision to studystatistics was made, the place to go would be theUniversity of California in Berkeley.
Debasis and Frank : You began your graduate stud-ies in the states in 1960. What were your first im-pressions of Berkeley?
George : I traveled from Athens to Berkeley byway of London and New York. My first impressionof New York was awful, but California and, in par-ticular, Berkeley, was another matter. The climateis about the same as that of southern Greece, thecity of Berkeley is charming, and the university cam-pus is of exceptional beauty. Nothing, of course, is needed to be said about the academic standing ofthe entire university, and of the department of statis-tics, in particular. David Blackwell was the chair ofthe department, and Lucien Le Cam was the grad-uate advisor.
Debasis and Frank : Tell us a bit about your grad-uate studies and, in particular, about the facultymembers who had a strong impact on you.
George : The first graduate level course in proba-bility and statistics I took from Thomas Ferguson,during the summer session. Early on, I took courses
Fig. 3.
George Roussas when serving his military service inthe Greek Army, 1957–1959.
CONVERSATION WITH GEORGE G. ROUSSAS Fig. 4.
George Roussas in the Yosemite National Park inhis first summer in the UC Berkeley, 1960. in measure theory, functional analysis and topologyfrom the department of mathematics. The hypothe-ses testing and point estimation courses I took fromE. L. Lehmann, and decision theory form Le Cam. Itook courses in measure-theoretic probability, sec-ond order processes and sufficiency form EdwardBarankin. I took a one-year course in probabilityfrom Lo`eve, and a course in Markov chains fromDavid Freedman. From Blackwell, I took a courseon coding theory and another one on programming.Also, I took a course in Markov chains from K. L.Chung during a summer, and another course in em-pirical Bayes methods from Herbert Robbins whenhe was a visitor in Berkeley. It was almost a crimi-nal omission that I did not take the ANOVA coursefrom Henry Scheff´e, and at least one of the coursestaught by Jerzy Neyman. I did, however, study theScheff´e book thoroughly and, later, taught out of it.The faculty of the department of statistics in theUC Berkeley was an almost . . . suffocating constella-tion of stars! I had immense respect and admira-tion for each and every faculty of the department.Neyman—founder of the statistical laboratory andof the department of statistics—was an imposingfigure in the department. He was very kind to me, and more than once mentioned to me his experienceduring a brief visit in Greece as an international ob-server. Barankin, in addition to being an outstand-ing mathematical statistician and probabilist, wasalso well versed in philosophy and in the classics. Itwas not unusual for him and me to talk about Plato,Aristotle and Sophocles. I learned asymptotic the-ory primarily from Le Cam. His seminal work oncontiguity of sequences of probability measure andits statistical implications were the key for my en-trance into the field of large sample theory. As iswell known in the statistical community, Le Camwas deeply knowledgeable in a broad area of mathe-matical sciences, and exceedingly helpful to all thosewho sought his advise. Le Cam’s vast knowledge of-ten created communication problems between him-self and a student. However, those who persistedwould manage eventually to chip away bits of hiswisdom.David Blackwell was my great discovery at UCBerkeley. It is not a secret that UC Berkeley wasthe repository of great scientists. So, in this con-text, it would not come as a surprise that Black-well belongs in that exclusive club. What is ratherrare, however, is for a great scientist to be endowedwith exceptional human qualities. That is, indeed,the case, which puts Blackwell in a class of his own.He is endowed with a refined, friendly and appeal-ing personality, and he treats people in ways thatbuild their self confidence and inspires relationshipsbased on mutual respect. He’s been a wonderful rolemodel for me and many others.As one would expect, the time in the UC Berke-ley was academically challenging, but overall pleas-ant, and certainly extremely constructive; it pro-vided unmatched academic training.
THE UNIVERSITY OF WISCONSIN,MADISON, EXPERIENCE
Debasis and Frank : After taking a temporary po-sition in 1964 (while considering a possible returnto Greece), you joined the Statistics faculty at theUniversity of Wisconsin, Madison. George Box wasthen in the early stages of organizing the statisticsdepartment there. What were the highlights of yourtime in Madison?
George : In the fall of 1965, I was invited to inter-view at UW Madison. At the end of my interview,Irwin Guttman, then the acting chair of Statisticsthere, made me an unofficial offer, and I accepted it
D. BHATTACHARYA AND F. J. SAMANIEGO
Fig. 6.
George Roussas’ family in Marmara, in the summer of 1966. From left to right: George, sister Demetra, mother,sister Stella, father, sister Aggeliki (kneeling), and nephew John.
Fig. 5.
George Roussas upon his graduation from the UCBerkeley, 1964. on the spot. I had already fallen in love with Madi-son, both because of its physical beauty and becauseof the superb academic climate there. I did not allowmyself the time for the usual bargaining to improveupon the rather low academic salaries offered by theUW at the time! So, I joined the department of statistics of the UWMadison, in the fall of 1966, as an assistant profes-sor. At the same time, another four assistant pro-fessors were hired: Asit Basu, Richard A. Johnson,Gouri Bhattacharyya and James Bondar. Existingfaculty members, in addition to Box and Guttman,were Norman Draper, John Gurland, Bernard Har-ris, William Hunter, Jerome Klotz, George Tiao,Donald Watts and Sam Wu, as I recall.George Box had created in the department an aca-demically demanding and rigorous climate, but atthe same time, comfortable, nonoppressive, and en-joyable. Those who did their work well were rec-ognized and rewarded. I was promoted to associateprofessor (with tenure) in 1968 and to full profes-sor in 1972. As all other faculty members, I usedto teach two courses per semester, one graduate-level course and one undergraduate course. The un-dergraduate course was alternated between an up-per division probability and mathematical statisticscourse, and a pre-calculus statistics course. The lat-ter choice was often made, because that was wherethe interesting students were! We recruited somevery good ones into the statistics major!
Frank : I understand that at least one of them wasrecruited into marriage! (Laughs.)
George : You are right about that! It was in one ofmy pre-calculus statistics classes that I met Mary
CONVERSATION WITH GEORGE G. ROUSSAS Fig. 7.
A faculty meeting in the Department of Statistics of the University of Wisconsin, Madison, sometime between 1968and 1970. From left to right (part of the faculty only): Jerome Klotz, Grace Wahba, George Roussas, John Gurland and JohnVan Ryzin.
Louise Stewart, who was destined to become mywife. She was a Ph.D. candidate in food manage-ment with a minor in statistics. This was in the fallof 1969. During the spring semester of 1970, I was onsabbatical leave, which I spent in the famous mathe-matics institute of the University of Aarhus in Den-mark as a guest of Barndorff-Nielsen. It was alsothere that I wrote the draft of my book on contigu-ity and where I met the great K. It¯o and attendedhis ergodic theory seminar.Sometime early in the fall of 1970, after my returnto Madison, I contacted Mary, and she respondedpositively. We started dating regularly, and were en-gaged in the summer of 1971. I took Mary to Greeceto meet my parents, sisters and close relatives in thesummer of 1971, and upon our return to Madison,we had our civil wedding ceremony on September11, 1971.
THE PENDING ISSUE OF RETURNING TOGREECE—TRAVELING BETWEEN MADISONAND PATRAS
Debasis and Frank : Things went quite wonderfullyfor you in Madison, both personally and profession-ally. We know, however, that you were faced witha difficult choice in the early 1970s—whether to re-main at Madison or return to your native Greece. What were the main issues you were dealing with atthat time?
George : Madison was great for us in so manyways. However, there was a recurring issue that cau-sed quite a bit of discomfort. I began to receiverepeated notifications from the Greek governmentabout my contractual “obligation” to return to Greeceand my need to discharge this obligation. At the
Fig. 8.
Mary Roussas in George Box’s class on Time SeriesAnalysis as a graduate student at the UW Madison when shewas still Mary Stewart, 1970.
D. BHATTACHARYA AND F. J. SAMANIEGO time, Greece was under military rule, and that mademy return there problematic on many counts. Froma purely practical viewpoint, I was highly contentwith my life and career, and I had no desire to leaveMadison. Further, Mary and I were already planningto start a family. Philosophically, I was strongly op-posed to serving under a military regime. Finally,a return to Greece seemed unsafe to me, as I hadbeen active in opposing the military regime. On theother hand, I had no wish of being deprived of myGreek citizenship, as was being threatened. Inciden-tally, I became an American citizen on May 28, 1971,and ever since, I have been grateful to the Americanpeople for the privilege of citizenship bestowed uponme.
Debasis and Frank : So, how did you resolve thisvexing conflict?
George : I decided to respond to demands madewith a proposal that seemed like it had virtually nochance of being accepted. I indicated that I couldconsider a return to Greece only if I was offered anacademic position there comparable to the one I washolding in the States. Since there were quite a lim-ited number of professorships in Greek universitiesat that time, and occupying a full professorship was
Fig. 9.
George and Mary Roussas newly married in Madi-son, Wisconsin, 1971. only for the well connected, the possibility seemed,to say the least, remote. So, on this account, I feltfairly safe. Unfortunately (for me), the governmentcame up with an open chair in Applied Mathemat-ics (which included probability, statistics, numeri-cal analysis and a few other subject matters) in thenew and promising technologically oriented Univer-sity of Patras (UP) (situated about 150 miles west ofAthens), and insisted that I submit a candidacy. Stillfeeling safe for the reasons I cited above, I submittedan application, and lo and behold, I was elected (toa full professorship). However, the Minister of Edu-cation refused to ratify the election and ordered forthe chair to be declared open again. At this time,electors (all full professors of the School of Phys-ical and Mathematical Sciences) pleaded with menot to object to the resubmission of an applicationon my behalf. The process of election commencedanew, and I was elected again! This time, the Min-istry of Education kept the outcome of the electionin its drawers—neither rejecting nor ratifying it—for a few months, until a new Minister, a civilian,came in to replace the previous person, who was amilitary officer. This fellow was a chemical engineerwith a Ph.D. degree from McGill University, whohad worked for the Shell Oil Company in the Statesover many years. As soon as he was informed aboutthe long pending ratification of my election, he ap-proved the election immediately. That was the rightthing for him to do, but it did not serve my purposeswell. I was strongly urged to go to Athens to takethe oath of the office, and perhaps be given leave ofabsence for a limited period of time. The compro-mise reached was to take the oath of office in theConsulate General of Greece in Chicago, and reportfor duty in early 1972.Under these circumstances, Mary selflessly aban-doned her studies temporarily (she had already takenher Master’s degree, and was well on her way to-ward fulfilling all requirements for the Ph.D. de-gree), took a crash course in the Greek language, andstarted preparing herself for the forthcoming adven-ture. The colleagues in the department attemptedto dissuade me from going to Greece, and insistedthat I retain my appointment at Madison while tak-ing a leave of absence of indeterminate duration. Ihave always appreciated this gracious gesture.
Debasis and Frank : So, this is when your triumphant,if somewhat reluctant, return to Greece began.
George : In a manner of speaking, yes. In Febru-ary 1972, Mary and I departed for Patras. Now, the
CONVERSATION WITH GEORGE G. ROUSSAS city of Patras and the university campus are builtin a beautiful location, on slopes overlooking a bay,with the western part of Greece opposite it. How-ever, being acclimated in a new community (and,for Mary, a foreign community at that) did poseconsiderable problems. In the university itself, I wasreceived well by some, and not favorably by others.In dealing with the authorities, both in Patras andAthens, my strong point was that I had a safe escaperoute and, therefore, I had no problems in behavingin my natural way. I started teaching immediatelycourses in probability and statistics, organizing theLaboratory of Applied Mathematics, recruiting TA’sand personnel for the lab, mentoring students withan interest in probability and statistics, attendinghours-long and stormy faculty meetings, etc. Marymade a valiant effort to adjust to the local condi-tions, and tried hard to improve her Greek vocabu-lary. I am happy to say, though, that she was treatedexceptionally nicely by all involved.In the fall of 1972, we returned to Madison, andin the winter back to Patras. In the fall of 1973,we returned again to Madison. Mary also gave birthto our first son, Gregory, that October 18. Unfortu-nately, soon thereafter, I had a rather serious oper-ation (removal of slip discs) in the University Hos-pital, and I could offer little to the department atthat time. By the end of the year, we returned toPatras, where I completed my recuperation. Debasis and Frank : Traveling between the two pla-ces must have been quite cumbersome!
George : Yes, the moving back-and-forth betweenMadison and Patras eventually necessitated for usto essentially retain two households. It was finan-cially challenging and physically tiring. At the be-ginning, it was Mary and me (and Mary’s three catsfrom her student days!), and now it was Mary, meand Gregory (in addition to the three cats!). It wasclear that a decision was due soon as to where wewere to affix our affiliation. At that time, everythingpointed toward Madison, as my academic experi-ence in Patras had been a disappointment to me.In addition to all my painful efforts to organize andstaff a new unit, I repeatedly encountered what Iconsidered to be harassment from the Minister ofEducation (who seemed intent on imposing politi-cal considerations into university affairs). In time,we were able to establish a tentative truce, allowingme to proceed with academic matters in ways thatAmerican academics consider natural and perhapssometimes take for granted. It was about this time that a decision about re-turning to Madison permanently was due, when allof a sudden the military regime collapsed (in July1974), and a civilian government took over. The peo-ple, by and large, were elated with the change. Amilitary regime is not a normal and natural regimefor free people, in particular, for the country wherethe concept of democracy was invented and firstpracticed. Nevertheless, it appears that every regimehas its excesses. Even the new civilian regime waseventually credited with its share of excesses, in par-ticular, in the academic world. In any case, despitesome shortcomings, the new civilian regime was re-sponsible for establishing a semblance of normalcyin Greece and for opening up some new horizons.This promising outlook led me to decide to remainin Greece and to resign my appointment at the Uni-versity of Wisconsin, Madison. I did this in the fallof 1976, thus culminating four and a half years of ajoint appointment between the two institutions. Italso, by no means painlessly, terminated a ten-yearassociation with the great university and beautifulcity in which I had met the love of my life, flour-ished personally and professionally, and spent themost enjoyable years of my academic career.
THE EXPERIENCE IN GREEK UNIVERSITIES
Debasis and Frank : The next chapter in your pro-fessional life was spent as a faculty member and ad-ministrator within the Greek university system. Tellus about those years.
George : Yes, I was now fully identified with theUniversity of Patras. I felt that it was incumbentupon me to do all I could for the benefit of the in-stitution, while also looking after my own scientificsurvival. I had worked hard seeking out the bestavailable candidates (of Greek descent, as required)whenever a faculty position became available. I ex-panded this effort to the entire spectra of biolog-ical, natural and physical sciences. These kinds ofactivities were not universally appreciated, but Iwas nonetheless narrowly elected as the Dean of theSchool of Physical and Mathematical Sciences (bythe full professors of the school). Around this sametime, Mary became pregnant with our second son,John. He was born in Bloomington, Indiana, on Au-gust 10, 1977, while I was spending the summer asa research professor at Indiana University.
Frank : Is that when you were offered a starringrole in the bicycling classic film “Breaking Away”? D. BHATTACHARYA AND F. J. SAMANIEGO
George : No, that came later! (Laughs.) This time,I just went to visit and work with Madan Puri, anold friend of mine since our UC Berkeley days, andmy former student from Greece, Michael Akritas,now an outstanding senior statistician, as we allknow.
Debasis and Frank : Following this leave, you re-turned to Patras to take on the deanship with re-newed energy?
George : Exactly! I did not feel bound by academictraditions that didn’t seem to work. My main guideswere common sense and my experience with US uni-versities. One of my early “accomplishments” was toreform the manner in which faculty meetings wererun. Regular school-wide faculty meetings were al-ways held in a large room with the faculty seated,according to academic seniority, around a huge ovaltable. Meetings were seen as both business and so-cial affairs. The agenda was typically unrealisticallylong, so that meetings dragged on and on for hours,often without any truly useful work being done. Iintroduced a new system in which the agenda wasprioritized, placing first the items on which facultyinput was essential. I made a strong effort to ex-clude items which were politically driven, and to setaside strictly administrative issues which could behandled without taking up the faculty’s time. Whileexerting control of the agenda, I resolved to be firmbut also fair and impartial.
Debasis and Frank : By the way, reconnecting withyour family during this period must have been aspecial pleasure.
George : Most certainly so! My parents especiallyenjoyed seeing our children on a regular basis.
Debasis and Frank : How was your approach to thedeanship taken by the faculty in Patras?
George : While my methods were considered newand different, most of my colleagues were pleasedwith my performance and some of them urged meto stand for election for the office of the chancellorof the university. In those days, the chancellor waselected by the totality of the full professors of theuniversity; this was the old continental Europeansystem. I was elected by a comfortable margin. Iserved as a chancellor-elect for a year, and took overthe chancellorship the year after.
Debasis and Frank : An important political changebegan in Greece in 1982 when A. Papandreou formeda new government. What did you know about himat the time?
Fig. 10.
George Roussas’ parents in the late 1970’s.
George : Papandreou came from a political family(his father was the leader of a political party, andhad served both as a minister and prime ministerin the past). He left Greece right after high school,studied economics in Harvard, and served on thefaculty of several US universities, most notably UCBerkeley, where he was also the chair of the depart-ment of economics for some time. It was there whereI came to know him. He was a noted economist,clearly a political leader with highly respected cre-dentials within the European Union (EU). The newgovernment was welcomed by a substantial major-ity of people as a turning point in Greek politics, andjustifiably so. A modern and knowledgeable economistat the helm of the government would surely put thevast amounts of resources flowing form the EU togood use, developing and modernizing the Greekeconomy. It was generally expected that a man ofhis background would also revitalize Greek educa-tion at all levels, helping to recruit a substantialnumber of Greek academics from home and abroad,thereby infusing Greek universities with new bloodand highly qualified scientists.
Debasis and Frank : Around this same time, youyourself made a change within the Greek Universitysystem. Tell us about that.
George : Actually, this was not a change of mybase, it was the undertaking of temporary additionalduties. Undersecretary of Education George Lianisoffered me the position of the vice president for aca-demic affairs of the then new University of Crete.
CONVERSATION WITH GEORGE G. ROUSSAS Fig. 11.
George Roussas delivering a speech at the University of Patras at his inauguration as the Chancellor of the university,1982.
Fig. 12.
At the reception right after George Roussas’ speech. From left to right: Mary Roussas, George Roussas’ mother andGeorge Roussas. (His father had passed away). D. BHATTACHARYA AND F. J. SAMANIEGO
The appointment would allow me to remain in Pa-tras as chancellor, but it would require weekly meet-ings, in either Athens or Crete. My primary new re-sponsibility was to oversee and chair the elections offaculty members in the Departments of Mathemat-ics, Physics, Chemistry, Biology and Computer Sci-ence. I felt I could accommodate these duties in myportfolio without disturbing my family, now grownto include Mary and three sons, with the addition ofGeorge-Alexander on December 12, 1980. I launchedinto my new responsibilities with gusto, and beforelong, the University of Crete was staffed by scien-tists of considerable international reputation. Therewere some challenges to be faced, but I’m saving thedetails for a mystery novel I will compose in the nearfuture! (Laughs.)Something about the Greek system of adminis-tration might interest you, as things are quite dif-ferent in the States. Neither the deanship nor thechancellorship I held provided any additional pay-ment or stipend beyond the professorial salary. As achancellor, I had at my disposal a car and chauffeurfor university business around the clock, and thatwas all. For my service in the University of Crete, Iwas simply paid travel expenses and a nominal perdiem. In Greece, academic administration is viewedas within the normal scope of a professor’s activities.
Debasis and Frank : What seemed like a very promis-ing beginning at Patras and at Crete was, unfortu-nately, destined to run into insurmountable obsta-cles. What were the root causes of the change in theacademic climate in Greece?
George : Perhaps not unexpectedly, politics trumpsmost other forces in our society. While the condi-tions were ideal to usher Greece into a new era ofachievement and prosperity, it soon became clearthat this was not going to happen. It became ap-parent that Papandreou had no intention of beingthe architect of such a feat. What he did insteadwas to continue preaching and practicing his pre-election populism, even after he was in power. Hesquandered the national wealth and the resourcesprovided by the EU for partisan purposes and otherunworthy causes; consequential productive investingwas nowhere to be seen.The education system of the country, and, in par-ticular, the higher education, was in dire need of re-organization. Papandreou’s “reorganization” essen-tially abolished the administrative structure in theprimary and secondary schools, and virtually dis-mantled the universities. The existing small num-ber of university professors was marginalized with the flooding of the ranks by new appointees, andthe universities were turned into unlikely arenas ofcompetition of the political parties. By political col-laboration of teachers and students, the resultingmajority was then in a position to elect the uni-versity authorities at all levels (departmental chairs,deans, vice chancellors and chancellors). Needless tosay, the result was predictable chaos and a dramaticlowering of academic standards. It has been morethan a quarter of a century since these measureswere put into effect, and the results are everywhereto be seen. Furthermore, there is no hope for deliv-erance form this evil anytime soon; the genie is outof the bottle, and it is not easy (or even possible) toconfine it in there again!
THE TURNING POINT—RETURNING TOTHE STATES
Debasis and Frank : That is indeed a tragedy. It’sclear that you harbor both sadness and anger aboutit; sadness for your native land and anger about theway things were changed for the worse. As you com-pleted your term as chancellor of the University ofPatras, you could see the handwriting on the wall.That’s about the time that you took a sabbaticalleave at the University of California, Davis, is it not?
George : That’s exactly what happened. Althoughour next move was not yet clear, Mary and I de-cided that a year of sabbatical leave, spent outsidethe country, would be a welcome and much neededchange, and help us work out a plan for the future.That future could have been a suitable position inthe EU. Nevertheless, we decided to spend the yearin the States, and that is how I found myself atthe UC Davis in the capacity of visiting professor,starting in the summer of 1984. P. K. Bhattacharya’swork on nonparametric statistics was one of the rea-sons that I was drawn to UC Davis. Statistics at UCDavis at that time was organized as an IntercollegeDivision of Statistics headed by an associate dean.Professor Robert Shumway was the acting associatedean, and he was prompt and most accommodatingin his response to my request about visiting the UCDavis for the year.As you well remember, Frank, the UC Davis statis-tics unit was formed in 1979 in the usual manner,that is to say, by grouping together statisticians af-filiated with other departments, such as mathemat-ics, epidemiology, etc. It was a solid group of fairsize, and its first associate dean was Julius Blum, a
CONVERSATION WITH GEORGE G. ROUSSAS noted probabilist. Other members of the unit in theearly 1980’s were P. K. Bhattacharya, Alan Fenech,Wesley Johnson, Y. P. (Ed) Mack, Norman Matloff,Frank Samaniego, Robert Shumway, Jessica Utts,Alvin Wiggins and Neil Willits. Jane-Ling Wangcame aboard the same year with me in 1984. Theidea behind this mode of organization of the unit,that is, as an Intercollege division of statistics ratherthan a department of statistics, was to gather to-gether all statistical activities under one roof, ratherthan having them spread over the campus. In theUC Davis there is also a rather novel idea at work,that of a Graduate Group, which brings togetherfaculty with common research interests serving invarious units on campus. Actually, it is the grad-uate group which controls the graduate curriculumand supervises graduate degrees. So, there also was agraduate group in statistics, and the associate deanof the intercollege division of statistics was, ex offi-cio, the chair of the graduate group. Blum passedaway unexpectedly in his third year as associatedean of the unit, and Professors Bhattacharya andShumway reluctantly served in succession in an act-ing capacity, while an active search was launchedfor a permanent appointee as associate dean. As Irecall, Frank, at that time, you were serving in acampuswide administrative capacity as the Assis-tant Vice Chancellor for Academic Affairs. THE UNIVERSITY OF CALIFORNIA, DAVISYEARS
Frank : It seems that the stars were aligned thatyear, as Davis was searching for a new head of itsStatistics unit and you were seriously looking fora new position and new challenges. I clearly recallthat you were the unanimous choice of the Statisticsfaculty in our search in 1984–1985. You joined theIntercollege Division of Statistics in July, 1985, asProfessor, Associate Dean and Chair of the broadlybased Graduate Group in Statistics. You served asthe head of our unit for 14 years without takingeven one quarter of sabbatical leave. Your serviceto the unit was both visionary and very effective.From your perspective, what were the highlights ofthis period?
George : Thank you, Frank, for your generous de-scription of those years. Upon shouldering the lead-ership responsibilities in 1985, the faculty, in con-junction with the university administration, designeda strategic plan to expand the unit up to the pointof achieving a critical mass, and turn it from a solidunit to a unit of national and international standing.We proceeded with the implementation of the planby hiring a number of bright new faculty members,including Prabir Burman, Chris Drake, Hans-GeorgMueller, Wolfgang Polonik and Chih-Ling Tsai, andby making efforts to attract established superior level
Fig. 13.
The Roussas boys. From right to left: Gregory, John and George-Alexander, in Greece, 1983. D. BHATTACHARYA AND F. J. SAMANIEGO statisticians, such as Rudolph Beran from the UCBerkeley and Peter Hall from Australia. These lat-ter two recruitments became realities soon after mystepping down as Associate Dean. At the same time,we laid the foundation for a program in biostatis-tics, which subsequently developed into a programof national repute. A decisive role in founding anddeveloping the biostatistics program was played byHans Mueller, who was by training a statistician, abiostatistician and a medical doctor. Within a fewyears, it became apparent, and certifiably so, thatour objectives and goals were well on their way ofbeing realized.I’m sure you recall that, in an evaluation studyof 300 statistical research institutions around theworld—carried out by the National Sciences andEngineering Council of Canada (NSERC) for theperiod 1986–1990—statistics in the UC Davis wasranked 14th (top 4.7%) worldwide, and 11th withinthe United States (top 3%). This ranking was reaf-firmed and even improved in a follow-up study, car-ried out by Christian Genest and Mireille Guay (
TheCanadian Journal of Statistics , Vol. 30, No. 2, 2002,pages 392–442). In this study, the authors employed several criteria of evaluating the same as above in-stitutions. On the basis of one of these criteria—essentially, published research papers per capita inthe “top 25” research journals in the field—statisticsat the UC Davis was ranked 4th (top 2%) among 202institutions studied. And these hard facts are be-yond and above the general reputation of UC Davisstatistics faculty as excellent researchers, teachersand contributors to the profession. This really isan achievement for which our faculty as a wholedeserves credit, and I am extremely proud of thehard-working yet congenial group that constitutethe statistics faculty at Davis. Naturally, I am proudas well of the role I had the opportunity to play inhelping to shape this group.
Frank : George, would you like to mention at thispoint some events and turning points we faced as astatistics unit?
George : I surely would. Heading the statistics unitat Davis was not without its challenges, especiallyduring periods in which the California economy wasweak. In the early 1990’s, for example, during one ofthe more severe financial crises of the State (with in-evitable repercussions for UC), the dean in charge ofday-to-day oversight of the Intercollege Division of
Fig. 14.A.
In front of Kerr Hall at UC-Davis, in 1986. From left to right: Keh-Shin Lii, UC-Riverside; Y. P. (Ed) Mack,UC-Davis, Murray Rosenblatt, UC-San Diego; Peter Hall, Australian National University; and George Roussas.
CONVERSATION WITH GEORGE G. ROUSSAS Fig. 14.B.
In the island of Spetses, Greece, during a NATO Advanced Study Institute in 1990. From left to right: Y. P.Mack, UC-Davis; Paul Deheuvels, L.S.T.A., Universite’ Paris VI, France; and George Roussas.
Fig. 14.
George Roussas giving a seminar after his appoint-ment as Professor, Associate Dean and Chair of the GraduateGroup in Statistics at the UC Davis, 1985.
Statistics recommended the merger of statistics andmathematics as a cost saving device. Not only hadthe dean (a fine humanist, but largely unschooledquantitatively) forgotten that that was where wehad started—separating from mathematics in orderto realize the breadth and potential that statisticsrarely can achieve within a mathematics department—but he had failed to reflect on both the unit’s statureand its many applied contributions (including con-sultation across the campus through our StatisticalLaboratory, collaborations with applied scientists oncampus and a broad spectrum of courses taught asa service to students in other majors). The facultywent into overdrive to come up with ideas and strongarguments against the proposed merger. Also, fora period of about three months, I lobbied heavilya number of higher-level administrators and otherinfluential people who were supportive of our con-tinued independence. As a result of our collectiveefforts, I submitted a detailed and impassioned let-ter in defense of our status as a free-standing unit.In the end, the administration conceded that themerger of statistics and mathematics would be a se-rious strategic mistake. We held our status as anIntercollege unit for 21 years. While we cherishedour considerable independence as a mini-college oncampus, as well as the access it gave us to variousforms of support from all other schools and colleges, D. BHATTACHARYA AND F. J. SAMANIEGO we also realized that we were not big enough to with-stand ill-conceived attacks. It was this kind of rea-soning that led us to seek and achieve the (lesser butsafer and, let’s face it, more traditional) status of adepartment. This went into effect in 2000, and Jane-Ling Wang was the first chair of the department ofstatistics. She was succeeded by Rudolph Beran, andthen by the current chair Wolfgang Polonik.
Debasis and Frank : Your friends and colleaguesclearly appreciated your accomplishments and yourgenerous service to several institutions and to thestatistics profession generally. They threw quite a“party” in your honor!
George : All this was something of a surprise tome. My old friend and former collaborator MadanPuri of Indiana University organized a volume of re-search papers featuring work in the general areas inwhich my own research was focused. This project re-sulted in the Festschrift “Asymptotics in Statisticsand Probability: Papers in Honor of George Gre-gory Roussas,” VSP International Science Publish-ers, 2000. It was edited by Professor Puri, and con-sists of 25 papers by 48 authors from 13 countries,with a preface co-authored by Madan L. Puri andmy good friend and well-known statistician YannisYatracos.Subsequently, Hans Mueller and the departmentof statistics conceived of the idea of organizing atwo-day workshop at UC Davis at which the Fest-schrift would be officially presented to me. The con-ference took place at UC Davis on May 19–20, 2001,with the participation of a select group of researchers,including a member of the French Academy of Sci-ences, a chancellor of a German university, the holderof a name-chair in the London School of Economicsand three chairs of statistics departments.
Frank : At the time, I made note of the fact thatthe conference was scheduled right in the middle ofthe period that I was on sabbatical leave in Ireland.I chose not to take offense. (Laughs.) But seriously,George, I would have loved to have been on handto give my hearty congratulations for a roundly suc-cessful and meaningful career. And, of course, thebeat goes on.
George : It was indeed unfortunate, Frank, thatyou could not be with us on that occasion, but ap-parently there were many constraints the organizershad to abide by. I always regretted your absencefrom that festive event, in contrast to your everpresent continuous support and counsel throughoutmy UC Davis years.
Debasis : Let me add, George and Frank, that I wasfortunate enough to be present, and the proceedingswere thoroughly enjoyable!
Debasis and Frank : How have you structured yourprofessional life since leaving the administrative postsyou held up to 1999?
George : Well, I continue to be an active memberof the department of statistics, and a member ofthe graduate program in statistics and the graduategroup in biostatistics, both housed in the depart-ment of statistics. I have concentrated both on avariety of research problems in the areas in whichI’ve always been interested, and have enjoyed myteaching assignments, but I have also had the lux-ury of time to work on pet projects. I have writ-ten three books [“
An Introduction to Probability andStatistical Inference ” (2003), “
An Introduction toMeasure-Theoretic Probability ” (2005) and “
Intro-duction to Probability ” (2007), all published by Aca-demic Press]. I have already revised the
Measure-Theoretic book, and I am in the process of revis-ing another two books. Also, I am in the process ofworking collaboratively on a new book (tentative ti-tle “
Probability and Statistics for Non-majors ”). So,I’ve kept quite busy, both with professional projectssuch as these and with family, to whom we haveadded my daughter-in-law Casie, wife to my sonJohn, and my delightful granddaughter Sophia Agge-liki, as well as my daughter-in-law Laura, wife of myson Gregory.
SOME MEMORIES FROM ROUSSAS’PROFESSIONAL CAREER
Debasis and Frank : What are your fondest mem-ories over a career spanning almost 50 years?
George : In retrospect, I feel that there are manyreasons that I should be grateful for my long pro-fessional life. I am certainly most grateful to all ofmy professors in the UC Berkeley for the truly solidtraining imparted in me; the full value of it did notbecome evident until later in my academic career. Ioften reminisce about my ten productive and pleas-ant years at the University of Wisconsin, Madison.Madison was, after all, where I met my wife, MaryLouise, and where our first son, Gregory, was born.I do not regret the ten to twelve years that I in-vested in seeking to contribute to higher educationin Greece, although the net result was almost negli-gible. I feel that I gave it my best try, but, realisti-cally, there are many other factors which influencedthe final outcome.
CONVERSATION WITH GEORGE G. ROUSSAS Fig. 15.
Faculty of the Department of Statistics at the UC Davis, Fall 2000. Top line: Alan Fenech, Wesley O. Johnson,Y. P. (Ed) Mack, Prabir Burman, Hans-Georg Mueller, Wolfgang Polonik, Rahman Azari and Juanjuan Fan. Second line:Richard Levin, F. J. Samaniego, J.-L. Wang, George Roussas, Christiana Drake and Jessica Utts.
I am certainly most grateful to UC Davis for theway I was received, and the opportunity I was givento do here what I was not allowed to do in Greece.Genuine thanks are also due to my professionalcolleagues who honored me with my election as aMember of the ISI (1974), admitted me as a Fellowin the RSS (1975), and elected me as a Fellow ofthe IMS (1983) and the ASA (1986). Special thanksare also due to the scientific community at largefor electing me a Fellow of the AAAS (2008). Andlast but not least, I am grateful to a select group ofGreek scholars—the membership of the Academy ofAthens—for electing me a Corresponding Memberof the Academy of Athens in the field of Mathemat-ical Statistics (April 17, 2008).
Debasis and Frank : And on a personal level?
George : I feel exceptionally fortunate that I havespent my adult life surrounded by a wonderful, sup-portive and endlessly interesting family. For my sta-mina and persistence, I must thank Mary, especially,both for her support and encouragement over theyears but also for her sage advice. I feel singularlyfortunate to have three healthy, intelligent, beautifulsons—Gregory, born in Madison in 1973, now a com-puter scientist, John, born in Bloomington in 1977,a practicing attorney, and George-Alexander, bornin Patras in 1980, a UC Davis graduate in politi-cal science, aspiring to the legal profession. Also, weare delighted with the relatively new arrival (March26, 2008) of our first grandchild, Sophia Aggeliki, daughter of John and Casie, also a practicing attor-ney and our newest daughter-in-law Laura. Manywholeheartedly felt thanks are due to my sisters fortheir immense moral support and consequential ma-terial support when that was most needed.One thing we regret is that we did not have enoughtime to enjoy the house that we built in Patras in1980. Its setting is truly idyllic: It lies on an acreof land full of trees, (including an olive tree grove)at the foot of a wooded hill with a mountain in thebackground, and faces the Patras bay.On a personal level, it has also been painful that,by expatriating myself for most of my adult life, Iwas deprived of the opportunity to spend any signif-icant amount of time with my parents, sisters andother members of the immediate family. In retro-spect, I also believe that, by devoting unduly muchtime to my professional duties, I deprived my ownfamily and myself of the opportunity of spendingmore time together. But I truly believe that eachphase of our lives is a nonrecurrent event, and mustbe appreciated, as it comes, to the greatest extentpossible.
Debasis and Frank : Since you’ve served in a widevariety of administrative capacities during your acade-mic career, perhaps you’d like to share your thoughtsabout what it takes to do this type of work well.
George : I am pleased to do so. I believe that beinga good, efficient and inspiring administrator requiresan inborn talent. Beyond this, one has got to be D. BHATTACHARYA AND F. J. SAMANIEGO
Fig. 16.
One aspect of the Roussas’ house close to the university campus in Patras. honest, just and straightforward, and by word anddeed, convince one’s co-workers about that. One’scommitment to these values must of course be real,but it is also important that they be clearly per-ceived by those with whom you deal. Furthermore,without in any remote way being dictatorial, onehas got to convey the clear message that there isonly one leader at a time. One should stand and defend well established principles, and not bend ac-cording to the prevailing winds. Vision is important,but it is also essential to have the ability to explainone’s vision in ways that gain the needed supportand engagement from others. The ability to listenis extremely important. It is essential that differentsides of a controversial issue be weighed. I have neverfound it difficult to take a position that may be un-
Fig. 17.
At the baptism of the Roussas’ first grandchild, Sophia Aggeliki, in Athens, September 2008. The parents John andCasie Roussas with the baby, and George and Mary Roussas.
CONVERSATION WITH GEORGE G. ROUSSAS Fig. 18.
George and Mary Roussas with Sophia Aggeliki, right after her baptism. popular, but I never wished to do so without beingconvinced that I had the relevant facts in hand.I recall when once working at my desk in the chan-cellor’s office in the UP, I heard outside my door arather heated discussion. Inquiring about it, I wastold that it was a committee of cleaning ladies whowanted to see me and present to me a perennialunsolved issue of theirs, but the receptionist wouldnot allow them to do so. During my entire tenureon the university campus, the rumor spread widelyabout this professor from the States interacting withpeople at all levels. However, for the receptionist itwas inconceivable that a cleaning lady would askto see the chancellor. On this particular occasion,I invited the committee into my office, listened totheir concerns and was able to resolve them to theirsatisfaction that very day.There were several incidents with highly politi-cized student and TA groups, which could have de-veloped to the point of explosion, but fortunately,they were decisively contained to the point of dissi-pation. It is probably best that I not elaborate onthem further.
BRIEF DESCRIPTION OF MAIN RESEARCHINTERESTS
Debasis and Frank : We haven’t spent much timeon the areas of Statistics and Probability that youhave concentrated on during your career. This con- versation would be quite incomplete without yourgiving us a brief tour.
George : Thanks for asking! As you know—and es-pecially, you, Debasis—my early work is based onLe Cam’s concept of contiguity and Local Asymp-totic Normality (LAN). Roughly speaking, LAN al-lows for a more or less arbitrary parametric familyof probability measures to be replaced (asymptoti-cally) in the neighborhood of each parameter pointby an exponential family of probability measures.Contiguity ensures the establishment of asymptoticnormality under moving parameter points, when suchnormality under a fixed parameter point is alreadyknown. This theory has important statistical impli-cations. Roughly speaking, whatever can be done forexponential families can also be done, in the limit,for the given family of probability measures. Thoseresults may then be transferred to the original fam-ily, for which they are going to hold at the asymp-totic level. Such results were developed, originally,for discrete time-parameter Markov processes.Nonparametric estimation in special cases of Mar-kov chains has been around for a long time. However,nonparametric estimation in a general setting of dis-crete time-parameter Markov processes was largelyan open area for investigation in the late 1960’s. Itwas exciting to be in on the ground floor in thisproblem area. I published a series of papers, begin-ning in 1969, which established some foundational D. BHATTACHARYA AND F. J. SAMANIEGO results and opened the door to further research inthe area.In Markov processes, the future depends on thepast and present only through the present. One wayof incorporating the entire past, when that mattersis by introducing various modes of dependence con-ditions, referred to as mixing. The basic idea in mix-ing processes is that the past and the future are ap-proximately independent, if they are sufficiently farapart. In a way, it is the natural evolution beyondMarkovian dependence. I was introduced into thisarea by my colleague Y. P. Mack (a student of Mur-ray Rosenblatt) in 1984, the first year of my visit inthe UC Davis. From a probabilistic viewpoint, therewas a huge amount of work already done, mostly bythe Russian probability school (Davydov, Gorodet-ski, Ibrahimov, Kolmogorov, Lifshits, Rosanov, Vol-konskii and others), and also by probabilists in theStates (first and foremost Rosenblatt, then Bradley,Kesten, Peligrad, Philipp and others), as well as byother researchers (e.g., F¨oldes, Iosifescu, O’Brien,Withers, Yokoyama and Yoshihara). However, therewasn’t a body of work on statistical inference onsuch processes. These problems intrigued me, and Igot some nice results. These were published in a se-ries of papers, starting in 1987. Ever since, there hasbeen an explosion of papers in this area, includingcontributions by Doukhan, Louhichi, Masry, Shao,Tran, Yu and many others.The next area of my interest has been that of as-sociated processes. The concept of associated ran-dom variables was introduced by Esary, Proschanand Walkup in a seminal paper, and it was exten-sively used in the book by Barlow and Proschan ina reliability framework. The concept of negative as-sociation was introduced by Joag-Dev. Associationwas also introduced and used in the context of math-ematical physics by Fortuin, Kasteleyn and Ginibre.Although I had a peripheral interest in this areadue to my overall interest in modes of dependence,my interest was accentuated significantly after anextended visit to UC Davis by Frank Proschan. Again,there did not seem to have been a systematic ap-proach to statistical inferences in such processes,and this fact stimulated my interest in such a kindof work. As a result, there has been a stream ofpapers between 1997 and 2001 by me, my studentsand other collaborators in which a variety of suchproblems have been posed and solved. More impor-tantly perhaps, this seems to have instigated the formation of a “school” in this area with much ac-tivity in China, South Korea, France and Portugal.Some of the noted contributors in association, ei-ther in probabilistic developments or statistical in-ference, have been Birkel, Bulinski, Cai, Doukhan,Ioannides, Louhichi, Oliveira, Prakasa Rao, Shashkin,Taylor, Yoshihara and others. Special mention is de-served for a seminal paper on this subject by C. M.Newman.In the last ten years or so, I revisited, with Deba-sis, the area of contiguity and LAN, and extendedprevious work to the so-called Locally Asymptoti-cally Mixed Normal (LAMN) families of probabilitymeasures, so coined by Jeganathan in 1982. In thislatter framework, we produced a number of paperson distribution theory with applications to statisti-cal inference.Finally, my current interests include conditioning,sampling from continuous time-parameter stochas-tic processes, and the theory and applications of cop-ulas.
Debasis : George, I’ve truly enjoyed the opportu-nity to work with you. We’ve worked on a wide va-riety of topics, including, of course, contiguity. Your1972 book on contiguity has become a classic! I knowit’s been translated into Russian and perhaps otherlanguages. Have you given any thought to writinga new edition of the book that would include themany new results that we and others have obtainedin the area?
George : The contiguity book, which was publishedby Cambridge University Press in 1972, was writtenin an attempt to obtain a deeper understanding ofthe concept of contiguity and its statistical applica-tions, and also to help disseminate this very impor-tant concept. Le Cam’s original paper in 1960 is notparticularly easy to read. Of course, he employedcontiguity in his all-inclusive 1986 book (“
Asymp-totic Methods in Statistical Decision Theory in Statis-tics ,” Springer-Verlag). A much more accessible dis-cussion of contiguity and its repercussions are pre-sented in the 2000 monograph (“
Asymptotics in Statis-tics: Some Basic Concepts ,” 2nd edition, Springer)by Le Cam and Yang. I was therefore somewhat sur-prised that the Cambridge University Press put out(in 2008) a reprint of my book in a paperback form;apparently, there still seems to be some continuinginterest in that work.And now, in order to answer directly your ques-tion, Debasis: I don’t really have any plans to do
CONVERSATION WITH GEORGE G. ROUSSAS what you suggested. However, should you take theinitiative, I might be persuaded to join in! (Laughs.)Incidentally, some time in the recent past, I hadthought of organizing some material on associatedprocesses and their statistical applications. This ten-tative plan is now aborted with the recent publica-tion of an excellent monograph on the subject mat-ter (“ Limit Theorems for Associated Random Fieldsand Related Systems ,” World Scientific, 2007) byBulinski and Shashkin.
MUSICAL INTERESTS
Debasis and Frank : We know that you have agreat appreciation for classical music. How did thatlifelong interest originate, and which composers areamong your favorites?
George : I developed a strong liking for classicalmusic early on, during my high school days. I’mnot sure what drew me to it, other than its sheerbeauty. No one in my immediate environment wasparticularly musically oriented. At the same time,I have also always liked good folk music, as well assome Greek popular music as exemplified by the twonoted composers Hadjidakis and Theodorakis. Also,I enjoy selected pieces of popular American musicand light jazz. However, my passion is classical mu-sic. In general, I am fond of the Germanic (German–Austrian) composers. I enjoy everything composedby Beethoven, and, in particular, his third, fifth,sixth and ninth symphonies, and his fifth (the em-peror’s) piano concerto. I love many of Mozart’scompositions with special preference for some of hissymphonies, piano concertos 20, 21 and 22, and hisoperas The Magic Flute, The Marriage of Figaro,Idomeneo and Don Giovanni. Above all, I adore hisrequiem. I very much like a number of symphoniesby Brahms and by Haydn. Also, I enjoy many ofMendelssohn’s compositions and, in particular, theScottish and the Italian symphonies. I much enjoythe eternal Messiah by Handel, and on a lighter side,his water music and royal fireworks. Somewhat sur-prisingly, I never developed a true liking of Wagner’scompositions, although I have immense appreciationfor them.I also greatly admire many Russian composers.Among them, Tchaikovski ranks first followed byothers, such as Stravinski, Rimsky-Korsakov, Mus-sorgsky, Prokofiev, Rachmaninov, Sostakovich andBorodin.I much enjoy many compositions of the Italiancomposer Corelli, some of Vivaldi’s compositions, and the arias of operas by Rossini and Verdi. Itwould be an omission to leave out my liking of Filan-dia, and of symphonies number 2 and 4 by Sybelius,of some compositions by Chopin, Liszt’s Hungarianrhapsody number 2, and also of a couple of piecesby Bizet. And of course, everybody enjoys Ravel’sbolero!
Debasis and Frank : This is clearly more than a pass-ing interest. It seems that it ranks right up there withJack Kiefer’s appreciation for mushrooms! (Laughs.)
George : Well, you did not ask me about my foodpreferences! What a coincidence! I will never passup—if I can help it—a Saturday brunch of a mush-room omelet! I do have a preference for certain kindsof mushrooms, but in the end, any nonpoisonousmushrooms will do!
GENERAL OUTLOOK—CLOSING REMARKS
Debasis and Frank : On another topic close to yourheart, what are the main tenets of your politicaloutlook?
George : At least in the recent past and currently,the usual terms employed to characterize politicalideology are those of being “liberal” or “conserva-tive.” However, these terms are quite tentative, andhave had different connotations in different periodsof time. They are interpreted differently by differentpeople. I like to think of myself as not really fittingthe modern interpretations of either camp. My basicbeliefs are that one should be interested in preserv-ing the accumulated wisdom of our collective societyand in respecting the greatest intellectual achieve-ments of the human species over the millennia. Ifthis is this “conservatism,” so be it! At the sametime, it seems to me essential that one keeps anopen mind and a positive disposition toward newideas. That is “liberalism” in my book. However,before a new idea of any importance is adopted, itmust be deeply contemplated and should be vigor-ously debated and tested. The novelty of an idea inno way guarantees its worthiness and its usefulnessto society. The unquestioning adoption of “progres-sive” ideas, just because they are novel, may be trulydeleterious for the well being of a society; that isill-conceived license for perhaps emotional but cer-tainly not rational behavior, having nothing to dowith liberalism. At the same time, I believe that itis a mistake to resist change and adhere to the sta-tus quo, simply because it’s what we know and arecomfortable with; that is simply reactionary. When D. BHATTACHARYA AND F. J. SAMANIEGO
I find myself needing to take a position on a politi-cal or social question, I try to combine what I knowor can learn about the alternatives under considera-tion and form my opinion based on both experienceand newly found information. In short, I believe thatwe should respect existing social structures, but weshould not do so unquestionably. I have never beena “party-line” type of citizen, and I tend to vote forthose candidates and propositions that seem to meto stand the best chance of solving real problemsand of generally enhancing the quality of our livesand of the times we live in.
Debasis and Frank : This topic seems to naturallysegue into your general philosophy of life. How wouldyou summarize that?
George : I would say that the most important thingis to live a “worthy” life, based on ethical princi-ples, respecting valued traditions, yet trying to leavethings better than you found them. Trying to makesome meaningful contributions to society is impor-tant, as is the avoidance of excesses that distractone from one’s more noble goals and aspirations. Ibelieve that a commitment to “excellence,” in a gen-eralized sense, is also very important. This applies to one’s chosen profession, that is, to the way onedoes one’s work, and also to one’s personal affairs.In both of these areas, integrity and respect for oth-ers, and for society’s needs, should always be primeconsiderations. Mary and I have striven to raise oursons to have a sincere appreciation for these sameprinciples.
Debasis and Frank : What advice would you giveto young people just beginning their careers as aca-demicians?
George : Aim high and then work hard, with en-ergy, imagination and persistence, to achieve yourgoals. Strive to live a worthy life. Determine whatyour main strengths are, and use them to try tomake a difference, both in your professional activi-ties and in your personal life. Take pride in your bestachievements! At the same time, accept responsibil-ity for whatever failures you encounter and, mostimportantly, learn from them.
Debasis and Frank : George, this conversation hasbeen a distinct pleasure. You’ve had an extraordi-nary career, with consistently strong contributionsthrough your research,your teaching, the highly re-spected books and monographs you have written
Fig. 19.
Debasis Bhattacharya and Frank Samaniego interviewing George Roussas (in the middle) in his office in the De-partment of Statistics at the University of California, Davis, on May 15, 2009.
CONVERSATION WITH GEORGE G. ROUSSAS and your many achievements in administrative ca-pacities. It’s been most interesting to hear aboutyour personal trajectory. You didn’t set out withthis trajectory in mind, but we feel very fortunatethat it led you to Davis. We have all benefited fromthe leadership and collegiality that has character-ized your 25 years here. Thanks for taking the timeto talk with us about your life and career. George : The pleasure has been all mine. I’d liketo express my deep appreciation to both of you,Frank and Debasis, for this precious opportunity. Years from now, perhaps my children’s children willread this and be surprised that “old pappou” hada pretty interesting life, and one that was blessedin many ways.