On Online Classes

2020 05 26, Tuesday

A friend and I were discussing this:

After Weeks Of Online Classes At IIT, Here's The Truth - NDTV https://www.ndtv.com/opinion/online-classes-sound-cute-on-paper-heres-the-reality-2237464

WhatsApp turned out to be an unsuitable medium for this, so I wrote my comments out separately.

My father used to say that no comparison is fair.  This can be extended to ... .

I believe that any new scheme in publishing, education, learning, ​etc.​ does not replace the earlier schemes.  It only adds to it.

Old proverb:

The old order changeth, yielding place to the new.

New version:

The old order endures; it looks a bit different, since the new one has a place in it.

Example: In the old old days, people had to travel to a Gurukul, stay as a member of a Guru’s (extended) family, and study and learn.  Almost a 1-to-1 ratio.  OK, not exactly, but maybe 1 to N, when N was a single digit (usually).  [Disadvantage: the topper had to marry the daughter of the Guru!]

Then came the technique of writing (however crude it was).  And people thought that one can learn from the writings of the Guru and his shishyas.  That was possible, but the Guru was not replaced, only his students increased, and they had to meet and discuss their ideas and difficulties with the Guru.

Then came the printed book (Johannes Gutenberg).  The book did not replace the Guru, nor writing, nor discussions and interactions with a Guru.

Then came photography.  It added to the books, e​tc.​.  But people still did go out to look at stuff, samples, trees, hills, rivers, waterfalls, birds, dogs, ... .

Then came cinematography.  Another addition, note: NOT a replacement.  People watch the movie, and read the book, often do both.  Ditto for TV.  All that it did was to bring the cinema into homes.  Nothing of the old was replaced.  The order did change, but by addition, not by replacement.

Remember ​The Ascent of Man​, by Jacob Bronowski (BBC, 1973)?  He created the BBC TV Series, and then wrote the book (a must-read for any mature human).  His comparison of television and book is here (screenshots from Google Books):

I have not seen a neater comparison between television and the book.  But note: they both add to each other, the book (which came earlier) is not replaced by TV.  

After TV, we had on-demand TV, videotapes and discs, and then the internet.  Each item added to our ‘order’, it did not replace it.

We cannot hear the views of Dr Bronowski on the current order!  He died in 1974.

When I created the MOOC: ME209x Thermodynamics, I realised that it does not and will not replace any course, any series of lectures, or any book.  It may be considered as another book in a library.  Some ‘students’ will browse through it, some will taste it, some will swallow it, and a few will chew it and digest it (after Francis Bacon:​ Of Studies​).  We have done that with so many books (of all kinds).

The current situation: we are trying to replace real classes with online ones.  The idea of replacement is not correct.  Online lectures, video-recorded lectures, ​etc.​ will only add to our repertoire, it will not replace it or even change it significantly.  Ditto for evaluation, examination, ... .

Arrangement of Books in a Library

A few days ago, there was an item the BBC website:

Coronavirus: Library books rearranged in size order by cleaner

The title says it all, but one has to look at the photo to appreciate the result.

During the lockdown, a team of cleaners went through the shelves and re-arranged all books in the order of their heights!

This may be shocking for regular users of the library.
But that set me thinking:

When books are arranged in topic-order (typical for a scientific/technical library, which uses a coding, e.g. the UDC), the result is an arrangement with random heights.   However, this is very convenient for people (students/teachers/readers) to find books on a particular subject, field, or topic.

However, this means that the vertical spacing between shelves is dictated by the tallest book, and hence lots of empty space is also 'stored'.  This may be considered a waste of space - a scarce resource in many places.

Suppose the whole library arranges books by height.  That means, jumbled up w.r.t. subject, etc..
But on a given shelf, there will be books of almost the same height.  That means, we can adjust the empty space above them to the minimum, and thus store more books!  The library will now need less overall space.  Or, in a given space, we can store more books.

There is another advantage of this.  Serendipity and happiness will increase.

Consider the scene: I visit a library, I go through the indices (necessary now), and find that the book on Thermodynamics that I want to access is on Shelf 23B4.  I go there, and find the book I need.  However, there is a good probability that surrounding that I find a novel that I wanted to read (say Jules Verne's Around the World in Eighty Days).  I am happier, and my life is richer.

As a result of this, engineers may look at fiction, they may also read some sociology!  You can extend this and imagine yourself in such a library.  Communication between scientists, engineers, sociologists, economists, and laypeople will improve.

Does any small library want to try this out?  I will help out.  And, the reduction in space will be nicely quantifiable!

Entropy

A few days ago, a good friend of mine (since by IITB-student days) wrote:

This morning me, my brother and niece (who are both chemical engineers) were discussing thermodynamics and in particular, entropy. 

My brother and niece were convinced that entropy is a purely theoretical, abstract, and dry concept, and need not be taught. 

I narrated our old story of ejector design in detail and how you  helped with a revised analysis covering entropy (which was not considered in the original design), and how it worked perfectly well after that.

Since you have taught entropy in your teaching of thermodynamics many times, tell me, how many of your students really understand entropy? 

He wanted me to comment.  So, here goes:

Among engineers, I believe Mechanical Engineers ("We", henceforth) live happily with entropy.  We need it to analyse and design compressors, turbines, pumps, ejectors, mixers, ... .
We develop a feel for it.

Imagine our student (and for many, professional) life without an entropy axis (and it is always the x-axis, the major axis) on  T-s,  p-s,  h-s,  and so many other diagrams!

Even the good-old-professor  B. B. Parulekar  (of Khandu Patil fame) was happy to work with entropy, even though he kept away from the mathematics involved in its derivation.

And without entropy, we will find it difficult to define Gibbs Function, and hence chemical potential.  How do we study chemical equilibrium without it?

Maybe 10% of students who study thermodynamics understand entropy fully.

To paraphrase  Francis Bacon:
Some students just taste entropy, others just swallow it, and some few chew and digest it.

But fractions are confusing.  When I started teaching (as a formal teacher at IIT Bombay), my first lectures on thermodynamics were to a group of some  42  UG students.  But there was a core group of about  20  to  25  students (and they typically sat in a  5x5  or  4x6  formation, front, centre of the classroom, even their relative positions were almost fixed).  These were the alert ones, trying to understand everything, and ready to jump at the teacher if anything went off-track or was confusing.  I soon realised that I was teaching, essentially, these students.  [The class size went from 42 to 60 to 90 to ..., but the number in the core group remained the same, more or less.]

And I think these were the students who understood entropy.  And some of them continued to flirt with it, some had it as their girlfriend (half or otherwise), some are married to it and are living happily! 

While discussing with Prof Achuthan on not-any-particular-topic (and this was common), we realised that there are perhaps three types of manipulations that we need to do.

From the simple to the not-so-simple, these are:

Numerical manipulation (NM):
Arithmetic++.  All students are good at this.  And with the advent of calculators, our effective ability at NM has improved significantly.  Traditionally, active members of the trading class were excellent at NM.  We continue to do, actually need to do, NM throughout our life.

Symbolic manipulation (SM):
Algebra++.  A large majority of students are good at this.  This begins in high school with algebra, then trigonometry and calculus, then calculus++.  A large number of science and engineering graduates are good at SM.  And, the advent of software for symbolic manipulation has improved our effective ability in SM.  We keep on doing some sort of SM later in life.  Scientists and engineers do it continuously throughout their professional life.

Concept manipulation (CM):
This is the ability to handle different concepts, and then fields, together.  We have to do it, as students, to absorb high-level concepts in mathematics, science, and engineering.  Later, only perhaps researchers and professors do it!

Remember our school/college/maybe even university exams?

Questions were usually chapter-wise.  One chapter - one question (if at all).  In geometry, each question was (a) some theorem to prove, and (b) a rider in which that theorem was the main idea needed.  This requires a low level of CM, if at all.

But if you are confronted with a problem which requires material from two or more chapters from the textbook of a given subject, then you have to do a bit of conceptual manipulation, and this increases the degree of difficulty.  Real-life situations need ideas and concepts from different fields.  For example, detailed modelling of an electrical machine needs that we use electromagnetism, fluid mechanics, thermodynamics, heat transfer, as well as solid mechanics (to determine thermal stresses, and structural deformation).  This needs a reasonable amount of concept manipulation.

Even in a given subject, we need to combine and extend concepts (terms, definitions) to create newer concepts.

In thermodynamics, the major concepts are energy, temperature, and entropy.  Of these, energy straddles many (almost all) branches of physics, and is not very difficult to grasp.  Temperature is a basic thermodynamic idea, and requires some concept manipulation, using the zeroth law of thermodynamics. 

The idea of entropy involves a significant amount of concept manipulation.  That is because it is (almost) the final extract of a derivation beginning with the second law of thermodynamics.

Just consider (this will be familiar to those who are still friends with entropy):  We start with
the second law of thermodynamics - the Kelvin-Planck statement.  Then we take a very scenic tour through heat engines, a definition of efficiency, thermal reservoirs, 2T-heat-engines, a hierarchy of temperature levels, a definition of 'higher' and 'lower' temperatures, (possibly a detour to refrigerators/heat-pumps, and their performance parameters,) the idea of reversibility and reversible processes, the Carnot theorem, the Carnot engine, thermodynamic temperature scales, thermodynamic Kelvin scale, (the Carnot cycle, and proof of equivalence of thermodynamic-Kelvin and ideal-gas-Kelvin scales,) the Clausius inequality and its proof; and then using the equality part of Clausius inequality we arrive at the definition of entropy.  This requires a significant amount of algebraic as well as conceptual manipulation.

Another fact is that the ideas of energy ("the ability to do work") and temperature ("degree of hotness") are introduced in school, with very simplistic methods.  One of my (self-assigned) tasks in IIT Bombay is to make our students get rid of (unlearn) many such ideas from their school days.

Unfortunately, for entropy, there is no 'short-and-sweet' definition (however incorrect).  Hence, the belief that entropy is a dry, purely theoretical, and abstract concept! 

We should also understand that many of the terms we use are short-forms (terminology).  In thermodynamics, we use the word 'system' as a short form for 'a region of space, with well-defined boundaries, in which we are interested'.  'Property (of a system)' is a short form for 'relevant measurable characteristic (of a system)'.  Imagine our frustration if we do not use such terminology during our study of thermodynamics (or for that matter, any science).

Unfortunately, for 'entropy', there is no 'expanded form'.  It is an extraction of all that the second law of thermodynamics stands for.

That is what makes an appreciation of entropy difficult.  I mentioned that maybe 10% of my students understand it.  In professional life, even a smaller number need to use it.  This small-usage fraction is true not just for entropy, it is true for a large number of topics and concepts, sometimes whole subjects!  Any reader of this blog with an engineering and/or science background will appreciate this.

So, perhaps, the reasoning of the brother-and-niece was this:  we did not use entropy in our professional life (and maybe we did not understand it as students), so it need not be taught.

Imagine implementing the same logic on X (replace X by your favourite concept) .  If one takes this to its logical end, we will have to close down, maybe, our whole education system!