Mandate oriented classroom design

"Our engineers were building a bridge across a river, when suddenly there was a flash flood and everything got washed away. What did they do? They protested, saying it was an out-of-syllabus question" -- An old joke among engineering college teachers.

Some problems with our current modes of teaching are well known. Education is too exam and metrics oriented; there is a tendency to study for exams, rather than study to learn something; coursework reduces to a competitive activity, while learning is typically a collaborative activity; and so on..

To be sure, our current model of homogenized, standardized, calibration and metrics-based system, has its merits. From a country, which was reeling under rampant illiteracy and drought, a century ago, our education system has created one of the world's largest talented pool of engineers and other professionals.

A standardized, process-oriented system is somewhat like a factory -- it can mass produce, acceptable quality outputs at acceptable costs.

A process-based system also goes through several phases of evolution. Anyone familiar with capability maturity models can recognize the following phases of process maturity: Ad hoc, repeatable, measured and optimized.

Early systems of pedagogic processes were ad hoc in nature -- typically dependent on coherence between the teacher and the pupil. But as education needs increased, ad hoc processes became no longer feasible. We needed to scale, and for this we needed processes that gave satisfactory results and were repeatable and replicable in several places. Once, a repeatable process is in place, it is not long before we start calibrating and measuring different elements of the process. And once we have metrics, we can optimize on different elements and produce the best possible output for a given input.

However, process maturity does not stop at optimality -- and this is often overlooked. Unbridled optimization invariably eventually leads to overfitting. Overfitting is a state where the system becomes hypersensitive to metrics and starts reacting to, and trying to optimize on every measurable element.

Somewhat similar concerns must have prompted Andy Grove to say, "Every successful organization contains the seeds of its own destruction."

Success breeds overfitting. And this is not just with education. Consider media overfitting to TRP ratings, and you see what I mean. Overfitting gives specific narrowly defined metrics, much more importance than the underlying spirit of the activity.

The solution to overfitting, is to shift the focus from processes to principles. We need to shift the focus from asking how to conduct a course, to what is the course meant to achieve.

*~*~*~*~*~*

This kind of background reasoning led me to experiment with a new kind of learning model, after having a teaching experience of over 8 years, in the conventional model of lectures, tutorials, assignments and exams. 

This model is what I call: Mandate-oriented learning. I have been implementing it for the past three years and it has evolved over time with greater clarity. 

The underlying assumptions behind the model are the following: 

• This model is applicable for a higher-education setting, where the teacher can assume a reasonable level of maturity and background proficiency from the students 
• This model assumes that the students are genuinely interested in learning the subject, and have taken the course out of their own free will, with a desire to learn, and not out of peer-pressure or any other extraneous motivations

The course itself is organized in the form of mandates. A mandate is broader than a learning objective. A mandate simply specifies a set of topics that need to be learned by the class. There could be several specific learning objectives within a mandate. A mandate typically requires familiarity with a set of topics and proficiency in some of them.

The class as a whole (including the teacher) works towards fulfilling a given mandate. The teacher contributes to the mandate by providing lectures, facilitating discussions, etc. Students contribute to the mandate in several ways -- by offering talks themselves, writing essays, leading and contributing to forum discussions, contributing to the classroom glossary, etc.

Each mandate runs for about 4-5 weeks. At the end of this time, the class reflects back on the mandate by summarizing what has been learned. The teacher then does a random sampling from the class to assess how the class as a whole, fares with respect to the mandate. If the class as a whole is found to be satisfactory with respect to the mandate, the current mandate is closed and the course moves to the next mandate.

And that is the overall idea. There is no individual-level quiz or exam at the end of a mandate. The teacher does ask questions -- but anyone could be asked. So the class as a whole needs to be prepared for the end-of-mandate session.

Grading is based on the evidence shown by the students (on an online platform like Moodle) towards their contribution to the mandate. The more evidence that students show that they are helping the class meet the mandate, the better it is for their individual grades.

At the end of the course, typically we have a (summer/winter) break for a month. The last mandate in the course is a reflective mandate, and is open through the break. Students can use this time to reflect back on the course and try and implement something based on what they have learned through the months, and contribute it back to the rest of the class.

The design objectives behind the above model are as follows:

• View learning as a collaborative activity, rather than as a competitive sport. The entire class is assessed against a mandate and students have to help one another to ensure that the entire class makes it to the next mandate 
• Encourage reflection and assimilation. Usually, in conventional coursework, stress builds up to paranoid levels by the time of exams and right after exams, students tend to unwind completely and forget everything that they learned. The spillover mandate helps in reflecting back on their course after the semester is over, so that the learning lingers on
• Allow the student to discover and develop their unique mode of learning. Not all students are equally good at answering well-defined questions, and not all of them are equally good at implementing something. Learning a subject involves approaching the subject from our own unique vantage point.

The model has met considerable success so far, but I have thus far refrained from quantifying its "goodness" against conventional models, to prevent overfitting. Thus far, the goodness of the model is apparent by the (formal) feedback received by students and the (informal) observation that students who have gone through this model in the past seem to be applying their knowledge well, in their professional lives.

This post here is written to introduce the background thinking behind the mandate model and to solicit comments and suggestions on how we can improve it.

Does language "shape" or "distort" thought?

The question whether language affects thought has been of immense interest since several decades. There are several variants to what is meant by language "affecting" thought and let me give my perspective on this. (The terms Language and Thought are used in capitalized form when referring to them as objects.)

The connection between Language and Thought, has attracted the attention of philosophers since several centuries. There are two basic forms of hypotheses regarding Language and Thought [1].

The "mould" hypothesis thinks of Language as mouldable clay that can be cast into some "form" by means of Thought. What this means is that, while Thought is necessary to form a cognitive structure, without Language (the clay) there is no Thought. The only way we can see Thought is by its footprints in the clay.

The second hypothesis thinks of Language as a "cloak" or "dress" for Thought. Here, Thought is supposed to exist on its own and is packaged by Language whenever it needs to be communicated. In the cloak hypothesis, the stuff that Thought is made of, is universally innate in all of us and we just need to cloak it in Language in order to communicate.

One of the early forms of the mould hypothesis is the Sapir-Whorf hypothesis. A strong form of this hypothesis is called "linguistic determinism" which says that language determines what people think, essentially that thought has no place without language. A weaker form of this hypothesis is called "linguistic relativity" which basically asserts that language influences our thought and what we think is relative to what language we speak/think in.

Sapir-Whorf hypothesis is mostly equated with linguistic relativity and linguistic determinism has failed to gain traction because of lack of enough supporting evidence. It however caught the fancy of science fiction writers like George Orwell, who introduced the notion of "Thought Police" in the book Nineteen Eighty-Four. The Thought Police of Orwell's book held a powerful grip on the vocabulary used by the population. Terms representing personal freedom and liberties were completely removed from records and spoken language, so that no one really thinks them.

Such intense forms of linguistic influence theories were rejected by the scientific community as there is ample evidence to suggest that Thought does exist without Language. For instance, we often get into a state where we do not have the right words to express what we are thinking -- which basically shows that we are thinking without words.

Linguistic relativity though is an issue that cannot be easily settled. Linguistic relativity says that the language in which we speak and think influences the way we think, although the stuff of Thought itself is independent of Language. Let me return to linguistic relativity in a while, after we briefly visit the cloak hypothesis.

The cloak hypothesis, also called the universality theories contend that the stuff of Thought is something that is innate and universally applicable across all humans irrespective of the language they speak. In the cloak hypothesis, any notion from one language may be translated to any other notion in some other language, as Language is merely a packaging layer over Thought.

The cloak hypothesis is especially popular among philosophers of science. It is easy to see the universal nature of scientific theories -- the laws of physics for example, will be the same, no matter which language it is expressed in.

The cloak hypothesis also fits in well with Analytic Philosophy, that is widely seen as the underpinnings of philosophy of science. Prior to the 20th century, Western philosophical thought was dominated by the works of philosophers like Kant and Hegel, and was based on the notion of Absolute Idealism. The underlying idea is a kind of universal monoism, somewhat similar to the core tenets of Advaita school of thought in Indian philosophy of the 8th century CE. It states that what we perceive as different objects in the material world and concepts in the mental world, are basically part of a unified whole containing everything. The only element that exists really (is "simpliciter") is the unified whole, so speaking about any concept in isolation is meaningless.

Analytic philosophy on the other hand departs from this position and claims instead that concepts do exist on their own (are "simpliciter") and there may be an infinite number of such concepts that exist on their own. It is just that the mind cannot readily "see" these concepts and their characteristics.

For instance, consider the concept of a prime number. Prime numbers have been shown to have several characteristics -- some proven, some unproven and several (potentially infinitely more) unknown. But then do prime numbers really exist or are they simply a fabrication of our minds? When there were no humans, and dinosaurs roamed the face of the earth, were there prime numbers?

Analytic philosophy argues that prime numbers do really exist and existed during the time of the dinosaurs too. It is just that the dinosaurial mind was not evolved enough to discover them. Even the human mind has not discovered their complete properties and every time we find something new about prime numbers, it makes news among mathematicians.

To discover properties of prime numbers, our minds followed some kind of "light". Our minds could conceptualize the notion of quantities, the notion of addition, multiplication, division and so on and slowly as we went around following these trails, we discovered the existence of prime numbers.

The conceptual world is explained by what is called the "Plato's Cave" analogy. Imagine we are trapped in a cave and the conceptual world outside is casting some light and shadows inside the cave (our minds). Based on what we see inside our cave, we theorize and discover properties of the world outside.

The notion of Language does not feature anywhere in this model. It is only when we need to communicate with one another what we are seeing in this conceptual world, we need Language.

The cloak hypothesis is also supported by Chomsky's theory of Universal Grammar and Pinker's theory of the Language Instinct. Both of these theories argue that humans have an innate capability for learning linguistic constructs. And this innate ability comes from our innate ability for abstract conceptual modeling.

But we are still left with the question of linguistic relativity..

Linguistic relativity states that while language does not determine how we think, it does influence how we think.

Among the most recent results in this is a recent paper by a cognitive scientist from Stanford named Lera Boroditsky, on how languages shape thought. Boroditsky details a number of very interesting experiments conducted in different parts of the world, showing how language used by the community influenced their thinking.

In some languages like in most Indian languages, the gender of a person is an integral part of any sentence involving that person. Hence, in these languages, we cannot say something like, "I went to a movie with a friend," without specifying the gender of the friend. Which in turn directly or indirectly influences our mind to think about the gender of the friend in addition to other pertinent thoughts about the sentence.

The paper also mentions about an Aboriginal community in Australia, where sentences in their language have cardinal directions as an integral element. Rather than using relative terms like "left" and "right" the language uses absolute cardinal directions like north, south, east and west. So rather than saying, "The one sitting to the left of me is my cousin," one would say "The one sitting north from me is my cousin.." This means that our mind needs to think of and about cardinal directions in just about every situation.

So the language in which we express a thought influences the kind of concepts that are brought into the working memory of our brains. Even though the central concepts representing the semantics of an uttered sentence would be the same across languages, the quirks of each language causes it to bring in other peripheral concepts like gender or cardinal directions, into our working memory, in addition to the central concepts.

So how does linguistic relativism reconcile with the cloak hypothesis and Analytic Philosophy?

To connect these two elements, we need to bring in another notion -- that of a mental model.

A mental model is an abstract, incomplete, but consistent view of the world around us. We always think within the framework of a mental model and our brains may store several mental models. The different mental models often correspond to the different "hats" we wear in our lives. How we view some situation as a parent, may be different from how we would view it as a doctor or as a leader.

Mental models are built from our experiences and our interpretations of our observations. Every time we parse something, it is done within the framework of a mental model, and the semantics extracted goes to augment the mental model. Hence, when we say a sentence like "I went to a movie with my friend, yesterday" in any Indian language, the mental model at the recipient's end, would make connections between concepts representing me, my friend and the movie, and also on the gender of my friend. Which means that whenever the recipient's mind loads knowledge learnt about me into memory, the gender of whoever I had been with, is also loaded as part of the model.

In this way, yes, Language does influence the way we think. But I would argue that it is more accurate to say, Language "distorts" our thinking, rather than Language "shapes" our thinking. Thinking shapes our language, but linguistic quirks distorts our thinking. Let me explain.

The cloak hypothesis is based on the contention that concepts are independent of language. Analytic philosophy goes one step further to say that concepts are what are "real," while language is basically a means of communication.

For instance, the concept of hunger affects all animals even though they don't understand the word "hunger."

Our stuff of Thought is essentially to discover properties of concepts as comprehensively as possible.

For instance, we all generally understand the concept of "Ownership" which establishes a relationship between an "owner" and an "owned" object and giving certain rights for the owner on the owned. Let us say that there is a highway that is privately owned by a company. And because the highway is the property of this company, let us say that the company introduces the following restrictions on other users of the highway:

1. People using the highway need to pay a toll for every use 
2. All blue-coloured vehicles may not drive at more than 20 kmph on this highway

We can see that, while the first rule is considered "valid" the second rule would raise eyebrows or even protests. But then, why not? It is a private highway and they get to set the rules right? Despite the knowledge that it is a private highway, we would still argue that the second rule is somehow unfair and perhaps invalid.

So basically there are some characteristic properties of what constitutes "Ownership" which just exists "out there" for us to discover. We "generally" understand what is Ownership, but not comprehensively.

To help us maximize the chances of discovering the characteristics of "Ownership" (or any other concept) it is important that we approach this question with a mental model that is "least biased". The model needs to have just the right set of ingredients with little or no extraneous associations, in order to help us understand a concept. If we bring in elements concerning the gender of the owner, or the cardinal orientation of the highway, it only dilutes our attention on the main issue of Ownership.

From this perspective, I would argue that if anything, linguistic quirks mostly distort Thought by diluting or placing hurdles in our understanding of concepts, rather than "shaping" Thought. Language does not "shape" Thought in the same way as Thought "shapes" Language -- by giving it its characteristic properties.

The primary building blocks of our linguistic constructs are directly shaped by conceptual modeling abilities, while the way language influences thought is by bringing in extraneous concepts into our working memory, which are only peripheral to the issue of concern.

References:

1. Bruner, J. S., J. S. Goodnow & G. A. Austin ([1956] 1962): A Study of Thinking. New York: Wiley

Understanding Saturation and Stagnation

Adversity is one of the primary factors that molds the culture of a population. Just about any aspect of a society's culture -- be it the way they dress, the food they eat, the social protocols they follow -- have some roots in past adversity that the society has had to face.

Some kinds of adversity are of an instantaneous nature -- the cause and effect of the adversity are clearly visible. For instance, adversity due to cold weather or hot deserts or volcanic action or earthquakes, are clearly visible. Cause and effect are both apparent from the adversity.

However, there are some classes of adversity that are of a more insidious nature. This article talks about two such kinds.

The first is the kind of adversity that arises due to saturation of resources in a population. Resources at first, appear to be plenty and not much thought is given to what happens when the resources deplete. Most of the thought would have gone into utilizing the resources.

Saturation is a process that builds up slowly and has no one root cause. There is no one place that saturation starts -- it is everywhere. From an individual's perspective, saturation starts as a mild irritation. This irritation persists and slowly grows till an extent that it is no longer possible to ignore it or live with it. And this process typically takes such a long time, perhaps even over generations, that the physical cause of saturation is lost. Most of our response in turn, would be directed against the irritation than the source of the saturation itself.

The dangers posed by saturation is best illustrated by the parable of the boiled frog. Whether literally true or not, the idea is basically the same -- we do not equip ourselves sufficiently to deal with slow moving dangers and when the extent of the danger is apparent, it is too late to do anything.

Be it fuel crisis, depleting forest cover, increasing pollution, depleting water table, terror threats or climate change, slow moving dangers are everywhere. Because their threat is not immediately apparent, they tend to get trivialized or rationalized away by citing more immediate, pressing crises. Unfortunately, there will always be immediate, pressing reasons at any time to rationalize away slow moving dangers. And neglecting the slow moving threats can undermine all that we do to keep ourselves safe in the immediate and present state.

The second kind of insidious danger is that of stagnation. Stagnation is a state where our mental models of the world around us, have ceased to evolve. Stagnation ironically, usually sets in when we experience long periods of relative stability in our society. When a society does not face large crises for an extended period of time, our worldview does not have much motivation to change. And when the society does indeed face a real crisis, its response to the crisis is often woefully inadequate or incompatible with what is required to address the problem.

As an example, consider the earthquake that struck Latur in 1993. It was a magnitude 6.4 quake, which resulted in a loss of more than 9000 lives. However, a much stronger earthquake of magnitude 6.8 struck Seattle in 2001, resulting in just 1 casualty.

But, as the saying goes: earthquakes don't kill people, buildings falling on people, kill them..

Seattle had bigger and more multi-storeyed buildings than Latur and in the face of such a strong quake, it should have resulted in a much higher casualty. But then, building technology in Latur had not evolved to face up to the threat of earthquakes, while technology in Seattle had evolved. This stagnation in building technology and practices is perhaps because Latur was not as prone to earthquakes as is Seattle. But the stagnation resulted in a greater tragedy when the crisis did occur.

The same is true of a lot of practices related to safety. I remember once when we went on a sailboat over the sea, we asked the guide whether we will be given life jackets to wear. To which, he laughed and gave this all too familiar reply -- that he has been sailing from the past 22 years and nothing has happened, so just trust him.

A lot of safety practices look unnecessary and even amusing; and most of the time, they are redundant and do not matter. If we stagnate ourselves and neglect the importance of safety, in the rare cases when they do matter, they make a difference between life and death.

Stagnation, like saturation is not immediately apparent. It is hard to tell when our mental models have stagnated and are not keeping up with the time to face potential challenges.

However, there is a telltale sign of a stagnated population. This is when a population looks for "social" solutions to just about every problem, almost impervious to the more fundamental physical reality underneath.

As an example, I was reading a news article recently where the parents of a sailor who was drowned in the choppy seas in a storm last October when their ship ran aground, were asking for the "guilty to be punished."  It is almost as if the storm had nothing to do with his death -- it has to have someone guilty to be the cause of his death.

If our worldview provides only social explanations and looks only for social solutions to all problems, it probably means that we have not been jolted enough to think deeper about the physical reality and its underlying challenges. And that should be an ominous warning sign..

Understanding joint ownership of conserved and non-conserved entities

Here is a sharp distinction in definitions that I've found in my professional world. This pertains to the concept of "joint ownership" of something.

I have collaborated with several companies and usually when it comes to this issue of joint ownership of the fruits of collaboration, things get messy.

By one definition of joint ownership, which I have predominantly seen coming from MNC companies is that, if two parties jointly own some entity, then neither of them can use the entity without the permission of the other. And for this reason the companies are keen to buy off full ownership of the entity from collaborators.

On the other hand, I have seen this alternate definition of joint ownership, predominantly coming from local companies -- especially startups. This says that, if two parties jointly own some entity, they are both empowered to use it in whatever way they want and does not need the other's permission for anything.

This disparity in definitions was intriguing and I tried to understand the source of this disparity. And here is what I've realized.

The former definition of joint ownership (that were in use primarily in the MNCs that I've interacted with) was developed by lawyers in the brick-and-mortar era where the object of ownership was predominantly a material entity. For instance, companies jointly owned an airplane or a warehouse or an estate, and such things. The property of material entities is that their usage is a conserved operation -- or a "zero sum" game. That is, if I jointly own a car with my friend, my usage of the car will hamper his usage of the car at the same time. So, we need to co-ordinate our activities and inform one another of our plans to use the jointly owned entity. Not doing so, is clearly crude and impolite.

But the second definition of joint ownership has been conceived in a completely different environment. All of the local startups that I have interacted with, are software companies. The thing with software is that it is not a material entity -- it is an information entity. And the property of information is that it is a non-conserved entity. If I give a piece of material to someone, I won't have the material with me anymore; but if I give a piece of information to someone, both of us will have the information. Same is the case with software. If two of us jointly develop a piece of software, we can both own it completely. Its usage by one of the parties will not hamper its usage by the other parties. So, there is no point asking for permission from the other parties, to use something that is wholly owned by you.

Established legal practices are seldom changed because each change brings with it enormous unknown implications which need to be understood. So it is quite rational for an established MNC to just use existing practices without having to go through reinventing legal modalities for information entities. But for a startup that is primarily working in the information space, the most natural thing to do is to adopt the newer definition of joint ownership.

There is another angle why companies insist on explicit permission. They are worried that one of the collaborators may use the entity in a way that adversely affects the businesses of the other collaborators. But this can be addressed by a non-compete usage agreement for software entities. It is not necessary to pose a hurdle in the form of explicit permission for every usage.

This is one clear case of change in rules between the erstwhile brick-and-mortar businesses and the information era businesses. Who knows how many more such paradigms are being fundamentally altered?

The two pillars of emotions

These are some thoughts that I've been entertaining after a particularly insightful conversation yesterday with an old friend and fellow philosopher, who I met after more than a decade. Not sure how appropriate it would be to name the friend, so let me focus on the ideas instead.

Over the last few years I've been particularly interested in cognitive models -- models explaining how our brain thinks and makes sense of the world. One thing that I am quite convinced of now is the notion of "emotional disposition" -- basically how emotions affect our worldview. Think of our emotional disposition as an underlying framework that in turn determines what is the mental model we adopt for explaining observations and reasoning based on assumptions.

For instance, suppose we see a stranger approaching us in an unusual manner. What intentions we assume about the stranger and how we respond, now depends on our emotional disposition. The very same observations may give vastly different explanations depending on our emotional disposition.

Our emotional disposition is partly nature (who we are) and partly a function of what we have experienced. For instance, we could be suspicious by nature, or we could be suspicious because of what has happened to us in the recent past.

The question that is now pertinent is to ask, how many emotions are there? Or how many dispositional states are there?

There does appear to be a vast variety of emotions and researchers have long since tried to map out this space. One of the well known results in this area is Plutchik's "Wheel of Emotions" (Image source: Wikipedia)

Essentially it posits that there are "basic" emotions and there are "compound" emotions that are built from the basic emotions. Plutchik's theory identifies eight basic emotions.

But the theory I heard yesterday is far more elegant in my opinion. It basically posits just two basic emotions: a positive emotion called "love" and a negative emotion called "fear."

The positive emotion is like an engine that propels us outward, while the negative emotion is like the brakes, which propels us inward.

All other emotions are rooted in these two basic emotions. For instance, empathy, care, trust, desire, curiosity are all rooted predominantly in "love" while, distrust, suspicion, jealousy, lust, greed are all rooted in "fear."

To lead a smooth life, we need both emotions -- just like we need both the engine and the brakes to use our car. A smooth ride however, will use the engine far more than brakes, while a novice driver may use brakes far too often, making the ride bumpy, inefficient and damaging to the car.

To make our drive smooth, our driving style needs to be developed well. We need to have foresight, understand intentions of others in the road, and have a mental picture of the outsides while driving. Similarly, to lead a smooth life, we should be driven far more by our positive emotions than our negative emotions. And to do this, we should be insightful and thoughtful and strategic in our actions. Knee-jerk reactions and haste lead us to a drive where we are forced to use the brakes (negative emotions) very much.

It is not just individuals, but entire societies can be driven by positive or negative emotions. Management theories are rife with debates stemming from "Theory X" and "Theory Y" where the former adopts a mental model that is suspicious of people by default, while the latter adopts a mental model that is trustful of people, by default.

In order for Theory Y to work in a society, people in the society have to think strategically and be insightful and thoughtful in their actions. Theory X is the model of choice, when the population is not insightful. Such models often lead to self-fulfilling prophecies. Because the population cannot be trusted, we treat it with suspicion, and because we treat it with suspicion, the population does not act in a trustworthy manner and strives us to treat it with suspicion even more.

Much of our recent crises have blown up so much simply because our society seems to be run by fear, than by love.

The myth of a global language or Why language preservation is necessary

As the world becomes more and more connected, there are increasing concerns about local cultures, especially languages, dying out in a matter of a generation. This concern is wide spread, including in the erstwhile English-speaking world, who see the globalized version of English as a distortion or corruption of the basic structure of the language.

Several arguments are put forth to urge people to invest in preserving their language. One of the most incredulous arguments I've seen is once when a speaker mentioned theatrically that, "..if you neglect your mother tongue, it is no less than neglecting your mother!"

Dramatics apart, the issue of preserving one's language or culture never fails to be emotionally charged, and the usual model is that of a global evil force swooping in, an erasing all local cultures and ideas. This "evil force" theory is not relegated to India or other developing countries, where perhaps it is just more prevalent. I've seen people in countries of Western Europe, US, Japan and other such arguably developed countries, give the same argument.

Finally, a more intellectual sounding argument for preserving languages says that, "language shapes our thinking." Without enough access to a rich language, the argument goes, we cannot have access to certain thoughts at all.

One of the most famous examples of this theory is the book "1984," by George Orwell. This story describes a futuristic (at the time of writing) society set in 1984, that is run by a totalitarian government which has reduced all its population to essentially zombies. And one of the most insidious thing they do, is to change the English language and erase the term "freedom" and all its related terms from public memory. No one is taught this word, and this word is purged from books and all other forms of recorded communication. Thus, with no word to describe freedom, no one knows the concept of freedom at all.

While this sounds horrendously innovative, it is hard to believe that people will not know the concept of freedom, just because they do not have a word to describe it. We often see children and animals communicating complicated concepts, even when they do not have the words to describe them. So, the absence of linguistic constructs does not impede our ability to think. It may impede our ability to articulate and communicate, but thinking itself is more fundamental. In fact, if a population does not have a word for freedom, they'd soon enough coin a word, if the concept of freedom is important to them. It would be more accurate to say that "thought shapes language" rather than saying "language shapes thought." Sophisticated linguistic constructs would not have been invented in the first place, if the corresponding sophisticated thoughts that they represent, hadn't appeared first.

Here is an example of a monkey expressing rather sophisticated thoughts about displeasure and expressing a desire for fairness..

Language is essentially a tool to help us explore the space of concepts -- the concepts are already there -- language does not create concepts.

Think of the world of concepts and meaning as a vast ocean that already exists (and you are the earth, hosting the ocean). Languages are like different kinds of vehicles meant to navigate this ocean. Each vehicle (language) has its own structure, but all of them are meant to navigate and explore this ocean. So, we may have sailboats, huge steamers, rowboats, submarines, and so on. The structure of each language provides it some kind of edge to explore some part of this conceptual space. Theoretically it may explore all possible parts of this ocean, but its structure makes it optimized for certain specific kinds of explorations.

For instance, some languages may be very well suited to express emotions, while some other languages may have concepts of social relationships an integral part of their structure. Some languages like mathematics (yes, math is a language) are very well suited to express ideas with precision, while some other languages are well suited for speaking about physical reality.

In fact, the more "elegant" a language is, the more specific its suitability. A language structure may be so well optimized to describe some part of the conceptual space, it looks very elegant when applied to it. But trying to apply the same constructs in some other setting may make the language appear so unnatural. Think of the several instances of poetic beauty lost in translation, between languages and you know what I mean.

I would now make a pretty bold claim, that there can be no language that is optimally suited to describe all kinds of meanings with as much clarity as required. And this is regardless of the nature of the language. Be it a lexical language like English, where letters are strung together to form words representing concepts, or a ideographic language like Chinese, where concepts are represented as words directly, or a phonetic language like most Indian languages where letters are used to build sounds that represent concepts; languages are simply ill equipped to represent all possible meanings in the conceptual space.

(For the mathematically inclined, let me offer a quick proof. A discrete representational structure can only create a countably infinite set of representations. But I'll argue that the space of concepts is uncountable. For instance, any subset of a set of concepts described by a language can be a concept in itself. And all such subsets cannot be assigned a name in a countably infinite world, as the powerset of a countably infinite set is uncountable.)

There have been so many efforts to unify all the world's languages under a common, global representational system (Esperanto is one that comes to mind.) All of them have fizzled away, and this is not because of apathy or cultural pride. This was bound to happen, as there cannot be a globally optimized representational system for meaning.

Historically, as the world was large and communication and travel were relatively infrequent, each language developed more or less independently of one another. Every language of the world developed to explore and explain every aspect of the conceptual world. However, each part of the world was faced with a different kind of challenge -- some places had to deal with cold weather, some places had to deal with volcanic activity, some places had to contend with wild animals, and so on. This meant that the need for communication of meaning was irregularly distributed across the conceptual space and languages evolved to optimize themselves for that part of the conceptual space that the population had to largely contend with.

Hence, even though the language itself was meant to be a generic description structure, it became more optimized for communicating certain kinds of meaning, while not very suitable for describing certain other kinds of meaning. For instance, the sublime nature of several writings in Kannada about life in the Malnad region, is very difficult to convey, using English, even though appropriate words seem to be available in English. A similar lament is made by Bengalis who say that the Gitanjali, translated to English is nowhere as beautiful as the Bengali original.

Similarly, trying to explain concepts in advanced mathematics or quantum mechanics, in Kannada has the same unnatural experience. Several people have tried valiantly to express complex physical concepts in Kannada, but the fact remains that on a public knowledge portal (like the Kannada Wikipedia) the number of articles on physics is far lesser than the number of articles on literature, arts and social commentary.

So a single language replacing all other languages is as unnatural a prospect as a single kind of boat replacing and making obsolete, all other kinds of boats. Sure, the distribution of language users may be highly skewed with a long tail and such, but I believe, there will always be several languages in this world.

Gap filling versus path breaking research

Update: Added a few examples on request by one of the readers.. 

After having read and evaluated a lot of research theses over the years, I've come to see two clear classes of (genuine) research. The first is what may be called "gap filling" research and the second is what might be called "path breaking" research.

By definition, every research thesis extends our known body of knowledge. The difference lies in the approach towards how this body of knowledge is sought to be extended.

Gap filling research starts by a comprehensive understanding of the state of the art and looks at the various results we have known thus far. Then, it looks for knowledge gaps in the body of knowledge and proceeds to fill them.

Path breaking research typically starts with a pertinent vexing problem and aims to understand the problem. In doing so, it surveys literature through the lens of the problem at hand. Often times, the search ends in a near positive match -- someone, somewhere would have addressed a very similar problem. The research then starts by questioning the assumptions behind which existing solutions are proposed and sees if any of them are unrealistic and can be relaxed. Usually relaxing an assumption opens a can of worms and the researcher stares at lots of new problems. Path breaking research, then aims to create a consistent alternate model, with alternate assumptions and proposes it as the thesis.

Gap filling research helps in nailing down a research area and helps pave the way for the entry of commercial interests into the research area. Path breaking research on the other hand, opens up new, alternate ways of thinking and seeing our world through a whole new lens. It opens up new research questions and primarily appeals to other philosophers and academics.

Here, the term "path breaking" is not used in the sense of "great". Sure, all great research have been path breaking, but not all path breaking research have had "great" impact. For every path breaking research with "great" impact, several more genuine research outputs of the path breaking kind would have died a silent death, full of possibilities.

The term "path breaking" here simply means breaking the underlying mental model within which a particular problem was addressed hitherto. Gap filling research on the other hand, works within the existing mental model.

In an overall sense, gap filling researchers have greater acceptance by the "community" since they have demonstrated comprehensive knowledge that the community cares about. They however, have much lesser freedom in formulating their research problem because they are already constrained to think within the paradigm adopted by the community.

Path breaking researchers on the other hand, primarily see research as a licence for free thinking and an escape from community norms and practices. Path breaking researchers typically have a much tougher time with community acceptance and they end up treading on a number of raw nerves in their process of questioning assumptions.

Even though a large fraction of path breaking research never sees the light of the day, it is that occasional path breaking work that creates a disruptive change in the way we understand our problems.

Here are some examples of the two kinds of research:

Arguably, present day Data Mining was born with the idea of mining frequent itemsets and association rules from a database table. Till then, the predominant notion was that a database engine required a query in order to provide any answers. Essentially, we cannot obtain knowledge from data unless we already know what we are looking for.

In the early 1990s a series of papers proposed the idea of "mining" a database table without the context of a query.  Essentially, you could ask a database, "tell me something interesting from the data" without specifying anything further about what you are looking for. The key idea behind these papers were the concept of an "Apriori" property, where a series of interesting patters could be discovered very efficiently.

This is clearly an example of path breaking research. Following this, the Apriori algorithm has been stretched, applied, modified and enhanced in a variety of ways by a variety of researchers. Even today we still find papers on Apriori rule mining being presented in conferences and journals. Each of these are essentially gap filling research. While the Apriori paper changed the paradigm with which we view databases, the rest of the papers worked within the new paradigm introduced by Apriori. Gap filling research are also important because, they are what enabled commercial data mining products to emerge. A lot of people became rich by gap filling research over Apriori, but not all of these wealth trickled down to the initial researchers who introduced Apriori (not that they needed it..).

There are several such examples of path breaking and gap filling research. Einstein's relativity theory for example, is path breaking. It questioned some of the assumptions of Newtonian physics, like the absolute nature of time, and created an alternative, consistent mental model of physical reality. Relativity theory itself could not be directly commercially exploited; but it paved the way for numerous gap filling research that gainfully applied relativistic thinking to a variety of different areas.

Not all path breaking research necessarily leads to great paradigm shifts. But lack of impact, does not mean lack of merit. A good example are research efforts in non-Eucledian geometry. These geometric models question some of the axioms underlying Eucledian geometry and formulate alternative, consistent geometric models. They have opened up a whole new way of thinking about geometry and resulted in several gap filling research endeavours; however, their overall impact is not as big as say, relativity theory.