The core linguistic reasons for TAD

Don't like reading?
Here is a keynote speech that was delivered by Sabu Francis at a conference. It is about an hour long; but it would give a nice intro on the why of TAD . Even the what is also explained – as also the social context of architectural practice (especially in developing countries) https://youtu.be/b-7AkmA1xqM


You want a shorter version that gives a broad overview? Read this


Architecture is one of the last few specialized subjects (or maybe the only one?) where the indirect communication (aka “written” or “archived” representations ) of the subject is fraught with a serious linguistic flaw.

Let me explain: architects traditionally make drawings to explain a design.

But drawings only explain what is built – architects leave dark marks on the paper ( these marks are called by a french term:Poché) to represent what is built. Drawings represent spaces in buildings rather poorly. Spaces, need to be interpreted by the observer. And interpretations often go wrong.

So the serious linguistic challenge in architecture is that it has a figure-ground issue.

Those who want to understand any built-environment in depth need to know both the built-matter (e..g. walls) as well as the what is NOT the built-matter aka spaces… But spaces emerge in the built-matter because the architect left behind built-matter…. and hmmm, the built-matter was left there because the architect thought of the spaces. So both the figure and the ground are hopelessly intertwined.

Just like the famous figure-ground illusion.

Of course, when architects talk about “creating” space they mean modulating some part of the universal space in such a manner that some opportunities to function are thwarted and some are promoted.

(And as everyone knows; space in this universe is something that can neither be created nor destroyed. It just is there. And we humans happen to live in a tiny sliver of inhabitable space on this blue planet)

For e.g. When Trump visited Ahmedabad in India, in 2020 February; the local government in that city built a wall along the road which would have been taken by Trump. The reason? They wanted to thwart the opportunity for Trump to see the ugly, higgedly-piggedly slums behind. And they promoted the opportunity to uphold the image of a neat and clean city.

That was just a simple, dramatic example. But if you look at any built environment, you would see this interplay.

So when any built-environment needs to be discussed in any serious manner, the representation system should not ignore this figure-ground problem and instead, it should be able to deposit all the alphabets that are needed in the subject. That means both spatial representation and representation of the built-matter would be needed.

BTW, there are other subjects that do have this unique representational challenge. Music, for example, puts out both musical sounds (notes, drum-beats, etc) as well as silences between those sounds. When music need to be communicated indirectly (i.e. via a written medium) that written script MUST necessarily talk both about the sounds as well as the silences. And the musicians do have musical notation system to do this rather well.

(Why this parallel was given here; is to explain that it is not just architecture which has a figure-ground issue)

Imagine the linguistic situation in another way: Imagine asking a writer to write some novel without using a few of the alphabets. Say, he is told (as an example) do not use e, k, m, z … now possibly, such a novel could be written but it actually cripples the representation; and can leave a lot to the imagination (or interpretation) of the reader.

On similar lines, when the final means of communication of architecture reside in drawings – which accurately only talks of the built matter; and the spaces merely interpreted… the world would be receiving a poor set of indirect communication.

No wonder, architects are famous for being isolated in their own silos of knowledge. With nobody really sure of what really went into these buildings during the design phase. To risk an analogy: It is akin to a chef coming to you with a well made dish, but refuses to explain the process behind making that dish

TAD solves this problem due to a mathematical insight by Sabu Francis. His theories won him the 1991 special award by JIIA (Journal of Indian Institute of Architects)

Using the taxonomy he discovered, TAD actually can talk about ALL aspects of the built environment; be it the built matter or the spaces. This is the first design software that does this quite well – and now this way of looking at architecture has started gaining traction.

This way of looking at architecture also makes the data very tractable by automated systems peering into what was left behind by architects using TAD. Thus one can envisage writing AI programs that can get meaning from the data without requiring architects to manually interpret and provide the feedback.

In fact TAD has a built-in AI scripting language called ARDELA (ARchitectural DEsign LAnguage) that can be used for such AI routines. TAD itself is written in a version of Prolog called Visual Prolog; and even ARDELA is a variant of Prolog. As you may know, Prolog is one of the oldest AI languages which has been used for lots of critical projects.

Why was this linguistic problem not discussed earlier?
Few architects did discuss this, but probably not with the perspective of use for the Internet era. Rob Krier, for example, did talk about three types of Spaces, similar to the ones that Sabu talks about.

Possibly there is a vested interest in NOT talking about the need to leave behind cogent, useful information for the future. The history of architecture is deeply intertwined with the history of power play, politics, economics and so on. People are more bothered about the final buildings out there. To mix-up a metaphor: The proof is in the eating. So they just want the final dish. But then if the process is not known or is mired in controversial interpretations that one has to make – then there would be people in the know who would snicker in the background, thinking “Let them keep squabbling… they will not come to know our nefarious reasons hehehehe”

But there could be a simpler explanation too for the status-quo.

The act of making technical drawings was a powerful discovery

The mathematics of descriptive geometry and projections were extremely useful in laying out the final built-forms. When a mason holds a trowel in his hand; he really does not know what would be the space that wall is enclosing. His job is to ensure that the wall is made. That's it. So he goes about analyzing the drawing and detecting where the walls are to be laid out. Spaces emerged as a natural consequence and that seemed good enough for the dish to be tasted.

Drawings got a spurt in its life when Brunelleschi made his famous dome at Florence. What people forgot is that architectural linguistics has got more complexity (possibly here it differs from music) the kind of representation that is needed during the design process is not really the same as what is needed during construction.

And what Brunelleschi was impressing us was the use of drawings for construction! Somewhere along the way, we forgot drawings were for construction and instead, we thought they were for the whole architectural design process from inception onward…


Press F1 inside the application to read context-sensitive help directly in the application itself
Last modified: le 2023/04/22 20:59