Showing posts with label Daniel McFadden. Show all posts
Showing posts with label Daniel McFadden. Show all posts

Sunday, 23 May 2010

FIAT LUX?




The person you are looking at is Professor Thomas Marschak, an eminent authority in Economics of Information, son of Jacob Marschak, an even more eminent authority in the same field. More importantly, he was also my teacher and mentor way back in 1976/77 and is still conveying bonhomie, as well as knowledge, to ever younger student generations glad to be his pupils.
Has time stood still? Is it possible, after long and successful service, to become a Professor and Teacher Eternal? 

Before answering this and other weighing questions, permit me, as a prelude, to explain about my own dealings with UCB, that is, about my reasons for coming to Berkeley back then and my experiences, once arrived. In order to make this more palatable to you, let me intersperse it with some pictures of flowers I took at a rainy day on Campus.

I started academic studies (in Vienna) already in 1962, but it took me ten more years to get my major in economics and be accepted as a graduate student in Stockholm. My first thesis adviser, the markedly creative economist Bengt Christer Ysander, encouraged me to write a thesis in the field of information (My main adviser was Professor Lars Werin, who aptly guided me through the thesis' final stages). 

In those days, it was deemed important to improve the assessment of investments in public information systems, in particular the large scale EDP (electronic data processing) systems being introduced overall in Swedish administration during that time.


If you are an economist you realise, of course, that any assessment method for a given class of investments must start with developing an identification model for that investment class. For you normal people, this means that you have to know, and be able to describe, the things you are assessing. But how do you identify the changes in information flows generated by buying a huge mainframe computer, letting systems analysts run amok for a couple of years and, as end result, driving office staff to the brink of exhaustion by forcing it to enter data, on a video screen, according to procedures defying established logic?

Whilst I was pondering these strategic questions, and not going anywhere with my thesis, Roland Andersson, a good old friend, came to rescue. He was just back from Berkeley and looked the part, with a long blond moustache and hair cascading down his shoulders. He advised me to go to UCB and said: “Whatever problem you may have with your thesis, there will be someone there to make sense out of it for you.” More to the point, he also lent me his UCB General Catalogue, in which I saw an interesting interdisciplinary graduate course with the impressive title “Economics of Decision, Information and Organisation”. This course was stretched out over the full academic year and appeared to solve all my problems.


So off we went, my wife Alice and myself, to this temple of knowledge. Upon my arrival, I was well received by two advisers assigned to me, George Åkerlöf and Daniel McFadden, both still young then but, I am glad to say, Nobel Laureates by now. They had bad news for me: the one year course had been reduced to a short seminar, to be given during the third quarter in spring 1977 and taught by Thomas Marschak.

The good news was however, as they put it, that this would enable me to garner the necessary prerequisites for the seminar, which consisted of courses in mathematics and statistics at intermediate level and which I had been sadly unaware of. With those digested, I would also be able to take a third quarter course on Game Theory, taught by John Harsanyi (another Nobel Laureate; where do all these geniuses at Berkeley come from?).

So a major part of my stay was not to be spent on studies directly relevant for me, as I saw it then. But how I was mistaken! These preparatory courses were among the most interesting academic exercises I had ever experienced. The course I remember most fondly was about “elementary” classical mathematical analysis. It was given by Tosio Kato, an amiable, somewhat absent-minded, scholar.

He distinguished himself by rarely being able to finish the day’s lecture topic within the allotted hour. He would put the proposition of the day on the blackboard, hesitating about its precise phrasing, and keep changing details for about 15 minutes, mumbling to himself, with his back to us faithful students. Once satisfied with the topic of the day, he would then proceed by outlining the proof of the proposition, but soon lose himself in all kinds of diverting arguments; and so it would continue until the clock tolled and us confused students were released from his guardianship.
Professor Tosio Kato in 1988 (UCB Website)
All this led to some frustration among the younger students in his class, but I had already some 12 years of study behind me, so I did not mind that much. In fact, a bit slow in oral understanding that I am, I always had been neglecting lectures in my early student years, concentrating instead on reading the written course material. So this is what I did also in this case, plowing the thick course book “Elementary Classical Analysis” by Marsden.

Unfortunately, every two pages or so, I would get stuck in the mathematical arguments. This forced me to permanently occupy the poor professor’s premises during, and often also after, his office hours. Fortunately, I was the only one daring to do so. The often heated discussions that resulted usually ended with realising that my lack of understanding had its cause in printing errors or errors of presentation, and with Tosio congratulating me for pointing out these small issues in the book.

When it was time for the finals, Tosio had thought to make things easy for us students, by starting with two very simple warming-up questions, that actually belonged to the basic course, which I had never taken (not being prerequisite for the seminar I actually was interested in). Naturally, I was unable to answer those, but did my best to scribble in some formulas nonetheless. The last two questions were far more interesting and to the point, since they consisted of propositions we had to prove, but which had hitherto not been treated in class. Fortunately, both had been given as exercises in Marsden’s book and solved by me at home. I guess that only a few of the students were able to provide the right answers.


The week after, I visited Tosio for the last time, to discuss some final issues with Marsden’s book and to ask him, how I should repeat the course, which I was sure I had failed. To my surprise, he had given me an A grade and motivated it by saying that it was more important for the students to have independent thinking than slavishly repeated knowledge learnt by heart. He also conveyed to me Marsden’s gratitude for helping him “proof-read” his first edition with my comments, which Tosio had forwarded to his colleague forthwith. I think this story shows a bit the highly intellectual, but at the same time free-wheeling, atmosphere prevalent in Berkeley in those days. Let me just corroborate this by a citation, mentioning just the first of Tosio’s many achievements as scientist,

“Kato achieved early mathematical fame with his proof, published in 1951, showing the self-adjointness of stationary Schrödinger operators for physically realistic singular potentials. This result crowned a program, initiated by John von Neumann, of providing a consistent mathematical foundation for nonrelativistic quantum mechanics.” (UCB website)

So, there you have it; both Professor Kato’s and my credentials firmly established! Now the time is ripe to tell the story of the reunion with Professor Thomas Marschak. We have to thank Lars for this, for he asked me, when he heard I would be staying in Berkeley, whether Thomas was still in office. I took this as a challenge and, already in the first week of my stay here, I trudged up to the newly built Haas School of Business, an impressive, if just a bit too bulky, building complex high up on campus.



You may recall that this was the mellow week of students having a good time on Memorial Glade. In the sunny weather of early April, also Haas seethed with activities, from free dancing to collecting funds for the graduation ceremonies to be held in May. As a former student of the good professor, I of course put some money in the collection box, asking the two youngsters you see below, whether they knew Professor Marschak. As answer I got an enthusiastic “Yes”; they knew him indeed, and appreciated him, as their teacher in decision and information theory. Somewhat taken aback by this amazing staying power of an octogenarian I rushed up to Thomas’ office, but had no luck in finding him there that day.
Fellow (young) students of Professor Marschak
Since I don’t have a phone here in Berkeley, I e-mailed him my greetings, asking for a meeting to pay my respects. Alas, no answer was forthcoming during the following weeks and I almost forgot about this initiative, having plenty of other blog postings to prepare and publish. But the week before last, bad conscience caught up with me, urging me to call Thomas on a pay phone, only to get a peep from his answering machine. Stating my business, I indicated that I would follow up the call with a renewed e-mail. This time I got a very polite answer, inviting me over to his office on Monday last.


I can tell you that the good professor still looked a spritely middle age, not at all like the wizened wizard you might have expected. His welcome was cordial and he was eager to recall the seminar I had the privilege of attending 33 years ago. His eyes lit up when I told him that he had asked us students, at every second meeting, to present our own topics of interest, leaving only half of the seminar to his prepared lectures. This made him recall the course and - at least I hope - my humble participation.

He added that it would be impossible to have such an educational set-up for present-day students, not even graduate students; the young generation was far to impatient and focussed on efficient transfer of knowledge to accept such a scheme. I thought that the students may have a point there if they had the ambition to finish graduate studies within four standard years, concurrent with extra-curricular activities such as the one shown below.


I then asked him how it was possible to stay in office at the venerable age of 80. Apparently, a relatively recent Federal Law had made it possible for tenured professors to stay on active duty after their normal pension age, if they so wished. There was a cut-off date however, the Law applying only to teachers born after – if I remember it right - 1924. This affected his colleague, John Harsanyi, who had to leave upon reaching pension age, so Nobel Laureate he was. He deplored that Swedish professors did not have the same freedom to stay on, whilst asking me to convey his regards to Professor Lars Werin, whom he had met shortly back in 1982.

From this we switched our discussion to two friends of mine who also had participated in the seminar, or in the more lengthy one-year course. The first, Pavel Pelikan, he remembered very well. He had met him already back in the sixties, whilst being a visiting scholar in Yugoslavia and participating in a conference in Prague from there. Subsequently, Pavel had received a grant to visit Berkeley for a year. Thomas was intrigued to hear that I knew Pavel from his time at the Swedish Institute for Industrial Research. He also recalled Eva (see my posting on “El Escorial en Alta California”) who not only had taken the course, but later also had stayed shortly as scholar at UCB.
Fellow (more mature) students of Professor Marschak
Finally we had a short discussion about his research activities. To my surprise, he was as active in research as in education. He was very occupied with a theme that appears to me as current in Sweden as in the States, the role of improved information, through computerization, in increasing productivity. He was not sure in his mind, whether computerized information was a complement or a substitute to traditional production factors and thrived to investigate this, treating information not as an on/off choice, rather, dealing with it in the form of increasing degrees of refinement.


After this interesting discussion, which took more than an hour, we bid each other farewell, somewhat sadly, for would we ever meet again? Thomas will go off to a well-earned holiday in France after the finals, whereas I will stay in Berkeley two more weeks for new experiences and blog postings. When I left Haas, trying to think about the topic for the next posting, the impressive staircase I had to descend welcomed me with a light drizzle. This brought back some melancholic thoughts and I decided to slowly amble home along the nature paths crossing campus, whilst paying tribute to my numerous memories from a time long gone.