Category: Blog

Every major in college education has a standard set of courses that a student needs to take in order to get a degree. For example, a physics major needs to take classical mechanics, electrodynamics, statistical mechanics, and quantum mechanics; another example, a chemistry major needs to take inorganic, organic, physical, and analytical chemistry. Traditionally, these courses have been deemed a requirement for future jobs. Successful graduates will be able to call themselves physicists or chemists, or be identified as physicists or chemists by others. They serve as convenient labels in the job market and in their future careers.

But the reality of today’s workplace suggests that those labels seem almost meaningless. Perhaps the most revealing manifestation of this reality is in scientific and technological research. The research today is conducted at an industrial scale, with nations and corporations pouring resources into universities and research centers in the hope that the outcome of the research (which is the return of their investment) will boost the national and/or institutional competitiveness at a global scale. It has evolved and become quite different from the research, perhaps only a century ago, when researchers were those individuals who had the luxury and curiosity to explore how things work and why.

One important feature of the research at the industrial scale is the increasing degree of specialization: because a lot more has been made known in the world of science and technology, one person cannot know it all so one has to specialize in a particular area of research. This is where the preexisting labels, however convenient in the job market, get in the way. On the one hand, an individual might feel unwilling to proceed with a line of research because he/she is not specialized in whatever discipline the research is deemed belong to. On the other hand, one’s superiors might feel an individual unqualified for a particular area of research just because that individual is not trained as such or such. Both situations exemplify the failure to take into account the dynamic nature of an individual’s intellectual capacity, and perhaps, a failure to recognize that existing disciplines were themselves just labels of specializations a few decades ago,  and they are not labels of individuals’ mental characteristics. 

Now with the advent of online education, an individual can tailor one’s knowledge profile and skill set seemingly at will, so how should one choose? There could be a variety of opinions on this question: job market reality, future earnings, societal needs, and of course, one’s curiosity, individual gifts and talents, etc. I will venture to give my own, given the context set in this post, in the realm of research. I believe that a researcher should learn what is necessary in order to create something new.

Once getting into research, one quickly realizes that understanding at the textbook level, particularly undergraduate textbooks, are not enough. (Their idealized version of the development of concepts could even be misleading.) Moreover, the selection of topics to be included in and excluded from a particular textbook is often a personal and subjective judgment of the authors. In research, however, it is often said that nature doesn’t ask questions addressed to a specific discipline, and the answer is certainly not included in a specific textbook. A researcher ultimately bears the responsibility of making sure what he/she creates is something new, not just a slightly modified copy of something already existing. That responsibility necessitates familiarity with the existing literature in a given field. Given the interdisciplinary reality of today’s research, one’s existing education, however comprehensive, often has gaps that prevent a researcher from appreciating all the existing literature on a given subject. A responsible researcher should try to fill those gaps, and to fill those gaps is perhaps the best use of online courses for a researcher.

Online courses have been around for a while. MIT Open Course Ware (OCW), established in 2001, is one of the earliest examples. In OCW, the videos of the lectures are recorded and posted online, as well as some of the course materials such as lecture notes and homework assignments. However, the structure of the courses is still based on traditional classroom learning. The online learners get partial learning experiences of those who actually took part in the classroom. Although online, these courses are not designed for online learners. 

Recently, I took a few online courses in Cousera and edX, two major online education platforms. On Coursera, the Machine learning course by Andrew Ng at Stanford University has over 2.7 million people enrolled and over 120 thousand ratings (course completion). On edX, the Statistics and R by Rafael Irizarry at Harvard has over 200 thousand enrollment (no data on the number of people completed). Both courses are free. But if you want a certificate, the machine learning course costs 71 euros, and the Statistics and R costs 99 US dollars. 

Neither of the courses has any midterm or final exams. It makes you think whether these exams are helpful for education in the first place. Homework assignment already provides the test and feedback one needs to assess if one understands what is taught in the lectures. Arguably, midterms and final exams serve only to increase the stress level of students, which could be counterproductive for learning. 

Moreover, moving a course online redefines the idea of peers. People can learn the course anytime and anywhere. As a result, one’s peer group could be of any age. In this sense, education has finally moved from the industrial age to the information age. To paraphrase Sir Ken Robinson, an iconic feature of manufacturing in the industrial age is the batch number or the date of production, and in the realm of education, it is the year (age) of the class. The problem with education at the industrial scale is that people differ, and hence the age of a student is not always a good indicator of his/her learning ability/speed and much less of his/her learning style.  Online courses give people more control over their learning and hence more responsibility in his/her own education. It is liberating, but it can be terrifying as well. Given so many online courses, what should one take? 

Javert evokes conflicting emotions. He is an agent of the law and, by extension, the state. His chief traits, as Hugo described, are his respect for authority and his hatred for all crimes, which make him a great detective. To illustrate his effectiveness as a detective, Hugo described the arrests following the ambush of Jean Valjean, where the most notorious criminals in Paris at that time all readily capitulated without daring to put up a fight. The main fault of Javert, as Hugo pointed out, is that he pushed these good traits to the absolute extreme that he became a fanatic. In his words:

“This man was a compound of two sentiments, simple and good in themselves, but he made them almost evil by his exaggeration of them: respect for authority and hatred of rebellion; and in his eyes theft, murder, all crimes were merely forms of rebellion. … He had nothing but disdain, aversion, and disgust for all who had once overstepped the bounds of the law. He was absolute, admitting no exceptions. On the one hand he would say, “A public official cannot be deceived; a magistrate is never wrong!” And on the other, “They are irremediably lost; no good can come of them.” He fully shared the opinion of those extremists who attribute to human laws an indescribable power of making, or, if you will, of determining, demons, and who place a Styx at the bottom of society. He was stoical, serious, austere: a dreamer of stern dreams; humble and haughty, like all fanatics.”

He pursued Jean Valjean out of a sense of duty as much as out of the absolute conviction of his beliefs. The extreme nature of his beliefs is attributed to his birth in prison, which is humiliating to him, and his origin in the gypsy race, for which he felt an “irrepressible” hatred. One can say that he was biased while representing justice. As a result, he was blinded to the good deeds of Jean Valjean as M. Madeleine, the major of Montreuil-sur-mer. Upon Jean Valjean’s confession in the courtroom, Javert felt his conviction rewarded. His bliss was “hideous” and “pitiful”, as Hugo superbly described below.

“Javert, though hideous, was not base.
Probity, sincerity, candor, conviction, the idea of duty, are things that, when in error, can turn hideous, but—even though hideous—remain great; their majesty, peculiar to the human conscience, persists in horror. They are virtues with a single vice—error. The pitiless, sincere joy of a fanatic in an act of atrocity preserves some mournful radiance that inspires veneration. Without suspecting it, Javert, in his dreadful happiness, was pitiful, like every ignorant man in triumph. Nothing could be more poignant and terrible than this face, which revealed what might be called all the evil of good.”

At last, when saved by Jean Valjean in the barricade, he could no longer ignore the fact that this “criminal” has been “kind” to him and has done good deeds and is an exception to what he believes criminals are. In not arresting Jean Valjean, he was compelled to admit that he has submitted himself to something higher by disobeying his duty as prescribed in the codes of the laws. This admission broke his belief and him, and Hugo considered his suicide inevitable:

“What was happening in Javert was the Fampoux of a rectilinear conscience, the derailment of a soul, the crushing of a probity irresistibly hurled in a straight line and breaking up against God.”

I think Hugo created Javert as a metaphor for the French criminal laws at his time. The severity of the penalty is reflected by the extremity of Javert’s beliefs. Through the suicide of Javert, he suggested that the laws, at the least the severe penal codes, needed change. His reasoning is that the severe punishment, especially for petty crimes such as stealing a loaf of bread, had become a form of repression of the strong against the weak, of the state against the poor, in so much that the justice system has become unjust and that the wrongdoers had been wronged.

If there were not too much weight on one side of the scales— on the side of the expiation. If the excess of the penalty were not the eradication of the crime; and if the result were not a reversal of the situation, replacing the wrong of the delinquent with the wrong of the repression, to make a victim of the guilty, and a creditor of the debtor, and actually to put the right on the side of the one who had violated it.

It reminded me of what Confucious said about laws and punishment: people obey but have no shame. This is the state of the mind Hugo attributed to Jean Valjean before he met the bishop. On this issue of crime and punishment, Hugo and Confucious would agree with each other.

As a note of levity, does Javert have to suicide? An interesting comparison can be made between Javert and Jason Bourne. Both had been effective agents of the state, and both suffered the “traumatic” experience of committing an act of kindness. Javert killed himself while Jason Bourne transformed himself. I guess it all depends on what the author or director wanted to say.

Another interesting comparison: Javert and the Joker. In a sense, Javert is the exact opposite of the joker. Javert, a detective, the strict upholder of the rules of the codes; Joker, an anarchist, who followed absolutely no rules at all. On the spectrum of order and chaos, they are on the polar opposite, (with Batman probably somewhere in the middle). Imagine Javert vs the Joker. OK. Time to stop.

The movie/musical is great, so I read the book. The book is long. I read it on Kindle, so it is a bit hard to tell how long the book actually is. It is only when I finally finished it a few weeks ago that I realized it is the only fiction I have read this year. It is said that reading this long book is like taking a long solo trip: you are not quite the same when you finished it. I feel this is indeed the case.

There are so many occasions when Hugo’s words had me bang my fists on the table or jump out of bed and walk around in order to calm down. The movie now feels only like a sketch: it gets the major points across, but with necessary omissions and adaptations so that it is still a movie, not a TV series. In addition, the movie gives the impression that the story revolves around Jean Valjean and Javert. I believe there are good reasons for doing so in the movie (think about Gladiator vs. Wolverine). But the book is obviously much richer than that. Hugo summarizes, at the beginning of the fifth volume of the book, as follows:

 The book the reader has now before his eyes—from one end to the other, in its whole and in its details, whatever the omissions, the exceptions, or the faults—is the march from evil to good, from injustice to justice, from the false to the true, from night to day, from appetite to conscience, from rottenness to life, from brutality to duty, from Hell to Heaven, from nothingness to God. Starting point: matter; goal: the soul. Hydra at the beginning, angel at the end.

For Jean Valjean, the summary is pretty clear. But for Javert, Fantine, or Thenardier, it is far from obvious. I have been organizing my reading notes for the past a few weeks, thinking about what the author wanted to tell given the storylines of each of these characters. I think I will start with Javert.

From 1 July 2011 to its official publication last week, it is exactly 7 years.

What I have learned over the course of the 7 years.

First, the difference between having an idea and realizing an idea.

The conceptual competence is in the quality of the idea, but the technical competence is in the quality of the details. Having an idea is an important first step, but it is only a first step. To go from the initial idea to the final product (in academia, this means paper), requires a lot of technical mastery, and more often than not, the idea would have evolved into something quite different from what it was at the beginning. The original goal of FLIPS is to be used as anti-icing surface. That goal is not even in the paper. To quote the words of Steve Jobs:

It’s the disease of thinking that a really great idea is 90% of the work. And if you just tell all these other people “here’s this great idea,” then of course they can go off and make it happen. And the problem with that is that there’s just a tremendous amount of craftsmanship in between a great idea and a great product. And as you evolve that great idea, it changes and grows. It never comes out like it starts because you learn a lot more as you get into the subtleties of it.

Second, the interdependent nature of research is a source of humility.

One only needs to look at the long list of coauthors on the paper to realize how much collaboration it took to make it happen. The experience of working with other people is an important learning process, not just in terms of technical knowledge, but also in terms of emotional growth. Being technically independent is an important first step, but being emotionally capable to deal with the demand of the interdependent teamwork is by no means a small feat. Frustration is common, and agreement is rare. Having the humility to realize that the whole is bigger than the sum of the parts and that each person has something to contribute, is the key to avoid intra-team competition and the costly ego battles.

Third, the desire to create quality work is a more powerful motivation than the desire to be the first.

To be first is a good motivation, whether it means to be first chronologically in publishing an idea, or to be first in terms of impact in a research field. But the satisfaction in knowing a piece of work well-done is a more powerful motivation. It takes time to do quality work, and it often runs contrary to the practical wisdom on the law of diminishing returns. At the end of the day, however, the satisfaction associated with quality work lasts because it is based not on the paradigm of competition (which the desire to be the first is), but on the paradigm of contribution. In the long run, quality matters more, always.

PS. Video credit: Harvard SEAS and Wyss Institute. See Wyss official press release.

German is a more complex language than English. I have been trying to learn enough German in the hope to write this post in German or a post in German, but alas, that day may come in the still distant future.

The image above shows the variation of articles and adjectives endings (and some noun endings) in German, which is one of the main chief culprits for the complexity of German. All the nouns are classified into the masculine, feminine, or neutral categories. They all have different definite (der, die, das, die) and indefinite articles (ein, eine, ein, keine). Moreover, depending on the case of a noun in a sentence (nominative, accusative, dative, or genitive), the forms of definite and indefinite articles change. Finally, the adjectives and some nouns following the adjectives will change their endings too.

This is just to give you a taste of the complexity of German grammar, as compared with English. After a while, one gets used to these grammar rules. Native German speakers, like native English speakers, don’t think about the rules while they speak. If it doesn’t sound right, then something is wrong. Just like English, or for that matter, any language.

Learning German is an interesting experience for me; it reminds me of the days and the drills of learning English. For all of its complexity, going from English to German is a small step, as compared with going from Chinese to English. The backbone of its grammar is still similar to English, just with an added layer of variations and many exceptions to rules. There are multiple occasions when I just have to remind myself that just stop questioning and just start trying to memorize.

Every once in a while, I question myself, what’s the point of learning German? Especially given the fact that the most spoken sentence for me is “Koennen Sie Englisch sprechen? ” (Do you speak English?)

One Saturday, as I was preparing to leave my apartment for shopping, two Germans knocked on my door, and when the door was opened, they started immediately speaking mandarin to me. As we chatted more, I learned that they are missionaries from a local church, and they work mostly with the Chinese people in the church. I was stunned by the level of fluency in which they spoke. There was immediate goodwill towards them, something that would not happen if they spoke to me in English. Even though I didn’t end up going to the church, the goodwill has remained.

Reflecting on this particular experience, I think trying to speak in a foreign language to its native speakers indicates an interest in the culture it represents and the people in it. It shows an effort to understand the culture and the people, and it is a genuine sign of respect.

I read this book because of Sir Ken Robinson’s TED talks. In a sense, the book is a big compilation of his presentations on the subjects of human intelligence and creativity.  He defines the Element as “the meeting point between natural aptitude and personal passion”. To find one’s Element, he suggests, one also needs to have the right attitude and actively seek opportunities.

One of the key messages is that we should think of human intelligence as something much more than academic abilities. The list includes mathematical, linguistic, musical, spatial, kinesthetic, interpersonal, and intra-personal (knowledge and understanding of oneself). The three features of human intelligence are that they are diverse, dynamic, and distinct, which indicates that each person should seek to find his or her own unique Element.

Creativity, he defines, is applied imagination. It is “the process of having original ideas that have value”. It takes a considerable amount of time to develop an idea, so in this sense, it is a process. It also requires a medium, in the sense that creativity can manifest itself in all kinds of human activities such as literature, maths, sciences, music, and dance. The last and perhaps the most important feature is that one has to develop skills associated with using that particular medium, being writing, calculating, experimenting, singing, or dancing. The corollary is that the teaching of creativity should go hand-in-hand with the development of skills in that medium.

Just like his presentations, the book is filled with the characteristic British self-deprecating humour. My favourite one is the comparison of earth’s size with other planets and stellar objects in the universe. One version of this joke can be seen here.

There is one thing, however, that I think I have some disagreement. It is about his assessment of the MBTI (Myer-Briggs type indicator). He disapproves it on the ground that it tends to put people into boxes and thus limiting the possibilities. I have read the original book Gifts Differing by Briggs and Myers, and I think the idea behind MBTI is never to put people into boxes and to limit people’s potential. It is more about understanding how people’s talents may differ (thus the title Gifts Differing), which is precisely what Sir Ken teaches too. However, I do agree that the subsequent widespread use of MBTI, particularly by the human resource department to sort people into different categories automatically, ignores the dynamic nature of types and tends to have a deterministic view of types.

It has been a while since the last time I published, but finally here it is. It is open access, so everyone can download it.

The article is written from the perspective of science, the development of ideas and realization of them. However, there is an equally important perspective, the one from technology. It is about what the techniques are that made the study possible, and more importantly, at the design phase of the experiments, what available technology to use to realize the scientific ideas. The first question is a technical one and is a boring one if you are not part of the research community. The second one is more generic, and the answers are video and 3D technology.

The first thread of technology development is video. Images are great, but videos are even better. Youtube is only 12 years old, and Vimeo 13 years old, even though they seem to have been there since the beginning of time. Their popularity means a large pool of professional tools to acquire, store, analyze, and compose videos. It sounded trivial, but recording and analyzing videos are the enabling methods to study spatiotemporal patterns. To put it differently, we chose to study spatiotemporal patterns because the technology for studying them is ready.

The second thread is 3D. 3D modelling software used to be the tools of only professional architects, animation designers, and product engineers, and requires significant training to use them. Now with 3D printers becoming more and more readily available, modelling software becomes cheaper and easier to use, so people start to make their own things. For researchers like me, it means using parametric design capability of these software to create models with precise specifications.

In a book on Bell labs, the author said something like this: in technology, being early is not necessarily different from being wrong.  I think it is true. There are certain study that can only be done once the technologies for it are ready.