Tuesday, May 23, 2006

Sapir-Whorf hypothesis: My thoughts and comments

I encountered the Sapir-Whorf hypothesis when a friend forwarded an article about it some time ago. In short, it asserts that "thought is constrained by language"- a very interesting and 'thought' provoking idea.

Researchers and practitioners of systems theory, especially the kinds of John Warfield, George Chandy (of OPMS fame) and yours truly, believe that language is more than mere words and sentences- somewhat like that cliche "the whole is bigger than the sum of the parts". The various parts of speech in a sentence, what the specific words mean to the audience, how the sentence was presented, by whom and the context in which the sentence was used are all equally important. You just CANNOT look at any portion of communication in isolation. Ok. How does all this relate to the Sapir-Whorf hypothesis? It relates because, further to my discussion couple of sentences ago, language is not just a group of words. If it were so, you could mug up a French dictionary and talk like a native Parisian. We all know that we can't. Still, how does this relate to the hypothesis I started this article with? Elsewhere (I can't recall where), I read some people's conclusion of the hypothesis that "we can't understand ideas that we do not have words for". I disagree. And quite strongly so. Let me try a pseudo-mathematical/ logical proof:

1. Lets assume that I am Einstein and this is early 20th century

2. Lets assume that we really can't understand what we don't have words for.

3. Can I come up with the theory of relativity? Think about it. I don't have a word for the theory yet.

4. Well, I do come up with the theory AND THEN invent word or words to describe the theory.

5. Because I can do (and Einstein, like so many others before and after him, has done this) the above, assumption 2 MUST be wrong.

Anyways, language is more than just vocabulary, which is just a cog in the communication machinery. And how this cog works 'with' the rest of the machinery is important to understand anything. Lets return to the original Sapir-Whorf hypothesis. Can language really limit our thought? Let us see how these two things are related. Think of language and thought as a dyad (put simply, dyad is any two unit combination towards something meaningful- refer to works of Charles Sanders Peirce). We can depict the dyad as in Fig 1 below:

Fig 1: The Language<==>Thought Dyad

How are they related? Language contributes to building mental models and mental models contribute to linguistic models depicted as sentences in a language. Hence, limitation of one can (and will) seriously inhibit the other. No rocket science in there.

Ah...mental models. What are they? They are, put very simply, the way our brains structure information to make "sense" of it. It makes models of everything it perceives, including sentences. Its model of a sentence includes lot more than mere array of words- context, prior experiences, notions, assumptions and a myriad other components come to play. Roughly, it is akin to how a computer operating system converts a binary executable to low level constructs the processor understands. Thus, the network of mental models is (loosely) an operating system of the human brain. We even have a word for it- mind. If brain is a biological entity consisting of neurons and other physical components, the mind could be thought of as a meta-model of various mental models, and their interactions.

Lets return to our dyad of Language<=>Thought from Fig 1, and our assertion that one's limitation affects the other adversely. How does the mind get out of this negative loop? It uses another faculty that probably only humans possess- Imagination. Along with Language and Thoughts, Imagination forms a "Triad" (sets of three ideas that are applied collectively and integratively) shown in Fig 2 below. A Golden triad, in fact.

Fig 2: The golden Triad

Together, they give rise to many marvels of human intelligence like Creativity, Problem Solving, Decision Making, Learning etc. The power of such Triadic constructs has long been established since Peirce.

Ok. Once again, how does all this related to the Sapir-Whorf hypothesis? The above discussion was aimed primarily to put forth my idea that

1. Language is lot more than just words

2. There is a specific relationship between language and thought

3. Mental models are a crucial mechanism of human intelligence and thought process

4. A meta-model of mental models serves as one way to think of 'mind'

From which, I derive that the limiting "language" the hypothesis talks about is words and mental models, and not just linguistics. Even using a common language, some people understand things others have difficulty in grasping- only because of the way the two groups prepare their mental models.

In conclusion,

1. If we could change our conventional communication style to use prose+pictures effectively, we will come quite close to depicting, communicating and working with mental models effectively

2. Such a style will help communication processes rise a tad (yes, just a tad but that makes hell of a difference in the real world) above pure linguistic dependencies in communication

3. While Sapir-Whorf hypothesis is worth its weight in gold, people who have mastered the art and science of systems thinking coupled with prose+picture style of communication are at a significant advantage

One of the best and most effective of methods of working with mental models (in isolation as well as in groups) are Warfield's graphical modeling techniques (like Interpretive Structural Modeling, Field Representation etc.) and George Chandy's extension of these models into meta models in his One Page Management System.


  1. I reached your blog from google. Maybe becos i don't know systems theory, but i had to read the article three times to understand it. But, when i understood it makes lots of sense. I like your approach of prose+pictures, but your links to Warfield and OPMS are not informative enough. If possible, please write about these in a future article and also point to Pierce's work on dyads and triads that you referred to. I don't think i fully understand it.

  2. George Chandy, creator of OPMS, mailed me the following response to this blog. I am posting it here on his behalf:

    I think we need to be careful about the phrase "prose + pictures". We are, often enough, doing that in every illustrated book, in every presentation, etc - but to enable effedtive understanding, the "pictures" have to elucidate "structure of the systems" being discussed..

    What Warfield has developed is a practical way to enable people to understand the relationships between the factors within the systems under discussion in prose - this is "structural graphics". So I use the term "prose + structural graphics" to denote the minor extension to our conventional mode of communication that can help us add significant power to both thought and discussion.

    Check out www.jnwarfield.com for some basic information about Warfield's books, etc - describing the seminal advances he has made to "General Systems Science" and particularly to "Systems Design". You can find out more about a practical aid to problem solving and decision-making based on Warfield's contributions - the "One Page Management System" - OPMS - by joining the group called "Towards Democracy" at YahooGroups - I am happy to send invitations to join to anyone who expresses interest. Participants are enabled to understand a sizable bit about systems, the need for understanding 'structure' and they have access to a prototype software that realises the problem solving tool, the OPMS. .

    What the OPMS specifically does is to enable anyone, at any level, to understand quite advanced systems theory with relative ease and to apply the sophisticated science of this to issues of his/her interest.

  3. nice, cozy place you got here :)..

  4. So you are suggesting we go around drawing pictures?