Part I: To sketch or not to sketch? That is the question

As mentioned in the last post, this post was to deal with role that gestures and movement of the body play in spatialization of design ideas as sketches or images in the mind’s eye. We are deviating from that sequence. It was felt that an overview of the series would be more relevant before we move on to the complex role that gestures and movements play in design thinking.

Let us look back at several findings presented in the last seven uploads starting from ‘Why do designers sketch?’ We plan to take a bird’s eye view of the findings and reflect on them in this post. Now that we know the role sketching and mental imagery play in design problem solving, we can now reflect on the series.

Designing as resolving spatial issues

The designer’s decision making seem to be similar to most often-cited example; being in a room and working out alternative furniture arrangements to get the most effective layout at the end. Whether you use sketching or imagery most design problems are typically resolved through spatial decisions and spatial manipulations of elements (in this case furniture pieces). For this part of the discussion, let us treat designing as resolving spatial issues to achieve pre-determined functional goals through design actions. So, whether you are creating a space (layout problem) or a smaller object, what designer’s handle/create are the spatial elements and their locations in space. In part II, we will revisit these issues through the lens of design thinking.

There are two ways to handle such an assignment. In both approaches, actions are in this ‘real’ space, which becomes a context for situated cognitive activity. Whether you use sketching or decide to work it out in the mind’s eye, the process involves perception and monitoring of representations that are being continuously created and altered by reactive design actions. Most reactions typically result in yet another new representation.

Let us focus on each of the two ways to handle spatial issues.

Sketching as a medium for conceptualizing of objects

Though this series does not directly report research on sketching, the author had worked on the role of sketching prior to exploring mental imagery. Some of the statements rely on author’s earlier research in sketching.1, 2

The first approach is to use sketching pad and pen, if handy. The designer draws the room and thinks of new locations and/or orientations of furniture pieces, then quickly draws the new layout. Looking at the sketch he reviews the new idea for its effectiveness. He will often see and react to what he has sketched and come out with a new alternative to be executed as yet another new sketch. You can capture the process as iterative cycles that involves –

“Think -> Act (sketch) -> Review -> Transform -> Think again”

till he feels that he has met the demands made by the design problem. He may take a holistic approach or take one element or one function at a time and handle complexity incrementally.

If the designer has mastered sketching, this process is cognitively less taxing. There are several reasons. First, it also allows starting with a tentative idea and incrementally moving to a more complete and complex state. In the earlier post “Thinking through the messy sketches” we discussed this approach. Second, designer does not have to budget excessive mental energy to handle the actions of sketching. Third, when he creates a record in a media outside of his body (i.e. as sketches), he does not have to budget energy to memorize and recall his past action and decision. So, if he is quick and good at sketching, it is still a cognitively economical option.

Willful control on the act of sketching, particularly in representing objects in perspective, is a critical issue when implementing a sketch. Sketching is a learnt action and is often taught in a structured way in most design schools. There is enough literature on learning to sketch. My own work on control on sketching is available on D’Source. 2

If designer is bad at sketching, this approach can be counter-productive. He needs to budget more mental energy for the act of sketching, but this is at the cost of mental energy required to come up with new ideas. (Refer to earlier post ‘Why do designer’s sketch?’) Designer’s additional efforts to judge the correctness of the sketch can demand energy budget that could interfere with the ongoing thinking of the design problem. So, designer needs ability to sketch quickly and effortlessly, as if it is a routine and natural act.

That is why in the earlier post, “Out-of-the-box ideas for teaching sketching”, focus was on learning to draw with a ‘feel’ of the line drawn via kinesthetic feedback and by diverting attention away from the pencil tip. Reducing the need to depend too much on visual feedback while sketching, allows you to budget the mental energy saved to attend to the demands of design problem solving. Several ideas were presented to explain how the entire body could participate in the act of sketching. The kinesthetic feedback generated by sketching actions can be used to ‘feel’ the correctness of the sketch being executed. It is critical to learn to sketch effortlessly and with natural actions.

The external display that sketching creates reduces the effortful activity of holding the ideas in your memory. However, the mental imagery is not completely eliminated. You have to mentally decide the new location of a furniture piece first, try to review it and quickly sketch what you have imagined. Thus you avoid spending mental energy on holding ideas in his mind and comparing them later. In short, competent sketching ability helps him conserve mental energy so that he can concentrate on developing the next idea.

There is another way to solve the problem of room layout. Imagine now that designer does not have the paper and pen or chooses not to use it. After all, the

Eureka movements may not occur when you have paper and pencil in your hand. Indeed with Archimedes, it occurred in the bathtub!

Designer will have to then depend on his visualization abilities.

Mental imagery as a medium for conceptualizing of objects

For most people it is so natural to see images in their mind’s eye, that many times they are not even aware of it. So, it is difficult to talk of mental imagery. We see images in our dreams too. Like a dreamer, the blindfolded designer also believes that he is in a ‘real’ environment and is awake.

What is the difference in the images that designers’ see when they are designing blindfolded and what everyone sees in dreams?

Unlike in dreams, the blindfolded designer is indeed awake. More important to note that the designer mostly experiences ‘constructed as well as work in progress’ images that he conceives, modifies and builds in time and space. The site that the architects visualized was ‘real’ space that experienced by walking in and out. It is a highly immersive experience that designers choose to willfully enter in.

It is important to view intentionally constructed images as unique experiences and as ‘work in progress’. This continues till the designer decides to artificially stop. Ideas through images continue to evolve, sometime even when the designer is not consciously working on it. Remember Archimedes?

What is critical is the ability to willfully control image generation and transformation operations in response to the thoughts that drive these operations.

This is why the experiences of imagery during design sessions are different from what happens in dreams. The willful control is needed to handle the unique operations that designer performs. Interestingly, he learns to do this without any formal learning programme.

Designing and the mind’s eye

Let us return to the furniture layout problem, but this time designer does not have a sketchbook with him. Thoughts of the designer continue to drive new solutions and prompt new ideas. He can try out alternatives by physically shifting the furniture pieces around, but that will be not only time consuming but also strenuous. So, the designer would have to solve the room layout problem by relying extensively on his internal resource and working out solutions in their mind’s eye.

Designer could move the furniture pieces around, reorient them, think of creating alternatives mentally. He could view them and compare the alternative solutions. All along, he has to view the current state in the mind’s eye, if necessary bring back the earlier ideas from memory and react to this internal display creatively to generate even newer ideas. He has to hold all of them in memory and compare and contrast them. No doubt a mentally stressful task.

These operations are far more taxing than recalling of the images seen earlier. You can clearly see that the visualization actions obviously go far beyond just recalling the images and inspecting them. Obviously, designer needs more complex capabilities for such a creative use of mind’s eye.

We knew so little about how designers willfully control mental imagery event. By blindfolding the designers in our experiments we forced a situation, so that these capabilities could be studied. The series is biased towards a detailed treatment of mental imagery because, unlike sketching, we were ignorant of the issues involved. If this ability to willfully perform and control operations on the images in the mind’s eye is critical for new ideas,

Should we not identify the operations that designers perform and/or learn to perform?

Fortunately, some of the operations that can be performed on mental images have been identified in cognitive studies. Most well studied operation is mental rotation of objects along three axes.3 Further, Kosslyn identified three broad operations. They included, holding the image and inspecting it (Image inspection technically referred as scanning), generating the image (image generation) and lastly transforming the images (image transform). Last one is a powerful operation and includes several types of transformations.4

Most of these operations are clearly visible in our experiments. The videos clearly show that designers and architects iteratively switch between image inspection, image generation and image transformation. They cyclically inspect the results in their mind’s eye and reflect on these images. We will take these operations one by one.

Image Inspection (Scanning)

This operation usually deals with inspecting the recalled image from the memory (LTM and sometimes STM). Ideal examples of image recalls were when SP as well as the architects were asked to describe their final design after they declared that they have finished designing. The object seem to be virtually there in front of them (SP) or in case of architects, built space enveloping them. When they described, they appeared to be ‘seeing’ in their mind’s eye and ‘reading off’ the images. Similarly, architects occasionally recalled the site plan that they were given to memorize, but subsequently these drawings were often transformed into a 3D view with all its features to see in their mind’s eye.

These designers also recalled images that are part of their memories and use them as precedents. In our case, such images dealt with previously seen, and perhaps liked, examples of objects or built spaces or their features. For instance, in student pair experiments, there is a clear evidence of pergola roof, deck, Japanese garden and water fountain as precedents brought in as images from the memory. Needless to say that these precedents and features get transformed creatively to suit the new context, when they are used in design.

Obviously, the designers are able to hold their images in their mind’s eye, inspect (or scan) them and react to them as they work with the image. They would often inspect images to make judgments, spot inconsistencies, search for opportunities to alter and manipulate elements within the image. It serves the same function that sketches as display serve. It allows designer to review his ideas, creatively react to them and iterate. What exists as current, gradually move towards effective solution incrementally. This is similar to what occurs when sketching is used. Image inspection leads to spatial judgments and all design tasks are based on these judgments.

Yet, there are logical differences because of the nature of the media, in this case the sketchbook and the mind’s eye in which the display occurs. Mental images are fragile and fadeout if you are not attending to them actively. Cognitively, to hold and inspect an image is an active process that demands budgeting of mental energy, more so, when it is done purposefully.

Because designers and architects are involved in creating new objects and built forms, they tend to often inspect what they created and then manipulate it. So, most examples of inspection are part of the process of image generation and transformations.

Image generation

If you watch the video of SP in action in casserole experiment, it shows sufficient evidence of SP’s logic, ideas and thoughts directing his design actions and gestures. These often result in generating images of the object that he was developing. He develops the shape from scratch, builds it up step-by-step as an image. SP choses to ‘treat’ the overall image of the object as ‘real’ and even physically shapes it, making changes interactively as he goes ahead!

Most of SP’s actions in image generation were incremental and became more detailed as the time passed. Observe SP working with a virtual object in the front. See video 1 below.

Video 1: The designer created a virtual object in front of him and shaped it as if it was real for the entire period when he was designing.

As he advances with his design, he fondly sculpts and alters the shape. Similarly, when the architects were blindfolded and let into a big hall, they first create a gross layouts, detail them and as they go ahead, build the 3D spaces around them.

It is common to see the cyclic process that starts with generation of the image, then inspecting it, reacting and implementing changes in their mind’s eye. That explains the incremental nature of the development of the ideas. Note that there is little change in the design process as in both cases. There is only a change in the media used for displaying the current state of design.

Most of these videos have short pauses of few seconds when they are silent. Pauses seem to be useful to judge the current state of design idea. Pauses longer than three seconds were purposeful. They are used for inspections of the images in the mind’s eye and reflecting on it. You can clearly see this in video 1. They appear to be part of spatial judgment efforts and end in actions and decisions immediately after. Typically, they end up with flurry of activity, often resulting in a new idea or a modification that had existed as an earlier idea. SP pauses several times to ‘see’ the object in his mind’s eye and then reflects on it. He ‘looks’ around to inspect what he had created and pass judgment after a pause. After one such pause midway through the casserole session, SP sculpts the shape, pauses and comments like ‘That will look interesting’. See this in video 1 above.

Image transforms

Image transformation is a critical operation on the images. It is also the most often used operation to quickly alter the contents of the image seen. Most surprising finding was that in implementing the operation, designers physically interacted with their creations in the mind’s eye. In fact such interactions were at their best in image transformation. For example, while thinking of support to the casserole body, SP’s palms represented the folding legs of the casserole and movements of the palms simulate the leg movements.5

There are several complex transformations of objects or spaces that architects and designers seem to comfortably deploy while designing. For instance, SP’s casserole video has several examples of the shape or some element of it being reshaped, chopped, moved, flipped, rotated and so on. See this in video 2 and 3 below.

Video 2: Observe designer exploring the shaping (curving) of the bottom of the casserole.

Video 3: Observe designer shifting from shaping of the lid to folding legs.

These videos include actions like selectively moving the objects or elements, manipulating proportions and compositions, changing sizes, altering colours, exploring different backgrounds and even creating exploded structure. They appear to use more than one operation, often in continuous sequences, and that too with amazing ease. The gestures come handy in the process of working with it and on it. More about gestures in part II in the next post.

While modifying their ideas and exploring alternatives, architects too performed different transformations of the images. Architects however use gestures more often to indicate and locate than to shape the spaces, perhaps because of the scale of these spaces. Earlier posts do to indicate few examples of use of gestures in shaping some parts of the building. For instance, see video 4 below. However, these instances are fewer in numbers.

Video 4: See designer shaping the entrance with hands. 

The purpose of most of the transformations is to improve the effectiveness of the solution at hand. It is typically followed by image inspection, where the designer would try to assess the implications of the changes they had just made. The use is similar to the way the sketch is reviewed, except when working with mental imagery, it is lot more difficult.

These changing mental events are available in form of fragile images in their mind. They have to hold these static as well as sometimes-dynamic images in their mind’s eye, which demand budgeting mental energy. Additional energy is required for reacting to them, altering the image with new intentions and hold on to the new image.

What does the series reveal?

The discussion so far has also helped us separate two forms of representations that almost serve similar function. Most discourses on designing focus on the first form i.e. sketching, and do not acknowledge mental imagery as serious contender as a form of representation. It is understandable. Indeed, there are no capturable external representations to write about or comment on, when working with mental imagery. One of the contributions of this series is to discover methods that give access to the actions that occur in the mind’s eye.

The series argues that sketching as well as mental imagery serve similar role but deserve to be treated as separate acts.

We hope that this separation will remove the bias towards sketching and studies in understanding the role of mental imagery in design(erly) thinking will get the attention it deserves. It is hoped that the shift of spotlight will balance the studies of role of representations.

Why is attention to the mental imagery critical?

It is known that this ability is directly correlated with creative efforts. To willfully change, transform or manipulate the image or some of the elements of the image selectively to intentions, distinguish creative people from others. Artists, dancers, architects, designers and some scientists can willfully control what happens in the mind’s eye. Einstein was known to create events in his mind’s eye that he often referred as thought experiments. Mozart was known to compose his orchestra entirely in his mind’s ears, an audio equivalent of mind’s eye. He was also able to hear and follow each instrument in isolation. There is enough scientific evidence to suggest that the ability to handle events in the mind’s eye and creative work are closely related. Finke has spent several years researching the role of mental imagery in inventions using cognitive science framework.6

Understanding nuances of visualization

We now have a clearer understanding of the term visualization and what it means to design community. It is not just externalizing ideas in a sketchbook or on a computer screen. It should include perception and manipulation of representations in the mind’s eye.

Visualization is much used and abused word. At the simplest level, it can be defined as ‘seeing the object when it is withdrawn and is no more available to be perceived.’ We are then asked to recall what it looked like. Can we restrict visualization to just recalling and inspecting the image in the mind’s eye?

Ability to recall or generate an image is not unusual. Most people are able to recall and view images. Dreams also involve recall of images. Dreamers are convinced that they are in a real environment, are awake and are experiencing a real event in time and space. Situation is somewhat similar when the designers were blindfolded, except that they were actually awake. Besides recalling, lots of people generate and see new elements in their images in dreams (also under hallucination) that they have never seen before.

Learning to willfully control mental imagery

The major difference is in the ability to willfully control the events in the mind’s eye. Architects, designers and artists are able to perform several cognitive operations on the images in the pursuit of the creative work. However, unlike sketching which is systematically taught to them, they learn to handle mental imagery with no formal training.

Can we learn to willfully control events during visualization without design training?

To handle mental imagery, at least till this time, there is no structured training available. It is perhaps partly a natural talent and partly perfected on job through practice. We have little knowledge of how willful control on imagery can be taught through training.

The focus leaning towards sketching as a preferred representation could change in future as designers learn the advantages of mental imagery as a creative thinking tool and find strategies to develop educational material for everyone. I saw the absence of training as an opportunity to convert my understanding of controlling mental events by developing visual puzzles that can be solved by use of some of the cognitive operations that we discussed. I hope to present these games and visual puzzles in some later post.

Who else can benefit from such training? There are lots of people who are involved in design but are not competent in drawing. Their creative energies could be harnessed if they could be trained to be comfortable with mental imagery.

Learning to handle sketching

Sketching does come handy when the complexity of the design project is very high. However, it demands rigourous and systematic learning to model 3D ideas as 2D sketches and draw them skillfully. I personally believe that initial training in orthographic drawing trains your mind to switch between 2D and 3D effortlessly. It has long learning period. Obviously, only few professions need it and go through such training.

To think, model ideas in mind and simultaneously sketch them is a task that needs systematic learning and practice. The focus on drawing by feel and by kinesthetic feedback and not by sight has its roots in this idea.

Note that there are other forms of representations including verbal descriptions that are used in communication. Most people use language descriptions accompanied by gestures to explain design ideas. In fact, they are effectively used in brainstorming. I tried to study gestures with minimal use of language with some success.7 But discussion on these topics are outside the scope of this series.

Sum up

This post is an overview of what was presented in the several earlier posts on sketching and mental imagery. It argues that 1] sketching and mental imagery should be recognized as two separate forms of representations. 2] The representations are a critical constituent of design(erly) thinking, 3] Both act like mediums for conceptualizing design ideas, 4] Both of these forms serve similar functions. So, technically, they can substitute one another. The post also compares sketching and mental imagery from the point of expending of mental energy. It suggests necessity of removing the current bias towards studying sketching and treat studies in imagery on par.

Focusing on mental imagery, it identifies the three broad cognitive operations, Image inspection, Image generation and Image transformation, that can be performed on the imagery. These are explained through examples taken from the experiments cited before. It argues that the understanding of visualization in the context of design should necessarily include learning to willfully control cognitive operations in response to the designer’s intention.

At a broader level, the design process has commonalities irrespective of the forms of representations used. What differ are the cognitive operations when working with sketching or exclusively in mind’s eye.

To sketch or not to sketch?

We started our discussion by exploring the role that sketching and mental imagery play as conceptualization tool. Both the forms have their merits and demerits. The choice should depend on designer’s comfort level with these two forms of representations. To be able to competently and skillfully handle both forms of representation would be an ideal situation. The complexity would not trouble the designer. The design response would be quicker. As one of the participant said in the context of use of imagery, and I quote

“I carry my problem with me all the time now”


Preview of the next post

We have restricted the above discussion to the implications of production and modification of display through sketching or through use of mental imagery. However, our discussion does not explain why the architects moved, walked and used the spaces they created, nor does it explain the deeper role that gestures play. In part II, will address questions like,

Why do designers use gestures and movements of the body when they solve problems? Does it support spatial decisions, visualization and design(erly) thinking? And if so, how?

Next post will discuss how these two forms of representation affect design(erly) thinking and why in spite apparent similarities, they are conceptually different.

Notes and references

  1. Athavankar U., (1992) Rediscovery the Act of Sketching: Implication of its Support to the Creative Thought Process, Design Recherche, No. 2, pp 45-60
  2. Discussion on freehand drawing Dec 21,2017
  3. Shepard, R.N., Metzler, J., (1971) Mental rotation of three-dimensional objects. Science 171, 701–703
  4. Kosslyn S., (1983) Ghosts in the Mind’s Machine: Creating and Using Images in the Brain. Norton, New York
  5. Athavankar, U. A. (1999). Gestures, Imagery and Spatial Reasoning. In: Garo, J. S. & Tversky, B. (Eds) Visual and Spatial Reasoning. Preprints of the International Conference on visual and spatial Reasoning, (VR 99) MIT, Cambridge, pp 103-128.
  6. Finke R., (1990) Creative imagery: Discoveries and Inventions in Visualisation. Lawrence Erlbaum, New Jersey
  7. Varshney, S., 1998, Castles in air: A Strategy to model shapes in a computer, in proceedings of the conference ‘Third Asia Pacific Computer Human Interaction (APCHI ’98), Japan, July 1998, 350-355

Can we prevent designer from sketching?

Blindfolded designer in action

So far, we discussed how sketching contributes to the design problem solving and partners with the evolving thought. It externalizes thought, but surprisingly also contributes to clarification of the emerging thoughts. Sketching works as partner in the creative process, support unobtrusively and ideally should demand little expending of mental energy. With their unique role, we characterized such sketches as a thinking tool, a class by themselves. We also divided the act into smaller actions and modeled them as a cognitive act. We looked at how budgeting of mental energy needs to be balanced between problem solving and creating instructions for sketching the idea.

The thinking sketches look like inevitable partner in design thinking and design problem solving. There are no reasons to doubt these statements. However, it automatically implies that you need quick and effortless sketching abilities, if you want to choose design or architecture as a career. There is more than adequate support to such a statement. There are lots of examples of great architects and designers being extremely good at sketching. (and even drafting) FLW is a good example of this. This justifies out effort listed in the last post that documents innovative ways of how sketching can be taught to students who may not necessarily start with good sketching. But

“Can such a statement be generalized. Should careers in design be restricted to those who are already good in sketching?”

Some books on great designers do list their sketching abilities. There is not doubt that sketching is a good skill to have. This has also been a subject research and documentation. (i.e. Leonardo’s sketches) However, why all books on great designers do not discuss their sketching abilities nor show their sketching samples? Should we interpret this silence, as ‘They do not have great sketching skills to talk about’ and yet they have succeeded?

To investigate this, it was possible to study sketches of all great names in design, but this was beyond my reach and means. Besides, there are several creative people connected with art and design, who do not use sketching. I intuitively felt that it was not fair to expect everyone to be competent in sketching. So, I decided to look at

“Is there an alternative route to sketching? What will happen if you stop an architect or a design from sketching?”

I decided to pursue this alternative route to discover answers through properly designed experiments in which designers participated. This led to series of studies and experiments that I conducted between 1995 and 2008 AD. The results inform us of the untapped capabilities of human mind and special abilities used by designers and architects in solving design problems. Reflecting back, I thing it was fortunate that I asked this question to myself. Else we would never have known the hidden capabilities would have remained unreported.

The primary research questions that we address were,

“Would the designer solve a design problem when blindfolded and thus prevented from sketching? If yes, how?”

Capturing designer in action

This demanded developing a strange scenario and a new experimental protocol, where the design process was captured live when the designer is solving a typical design problem.

Design problem

The design project was taken from local company manufacturing range of Thermowares as consumer durable and gift sets. Their range included insulated containers, food boxes and vacuum flasks used in households. The problem selected was to design a casserole that can retain food temperature. It should be used to store as well as serve hot/cold food item and can be offered as a gift item. One of the leading local industrial designer voluntarily participated in it. He is referred to as SP.


We needed to develop experiment protocols to ensure beyond doubt that the actual visualization of the solutions and design actions are completed when the designer is not allowed to sketch.

SP was given a written brief to read and it was cross-checked that he remembered it thoroughly. He was then told that he will wear an eye mask and will develop design in blindfolded conditions. He was asked to concurrently speak-aloud whatever is passing through his mind.1 All the steps in the session were video recorded. When he was satisfied with his creation, he was told to verbally sum up the final design solution. Then, the eye mask was removed and he was asked to sketch his ideas as quickly as he can and not add new ideas during sketching.

Watching design action

Few general observations before we move on to surprises. SP was comfortable with the think-aloud process and finished designing in 56.5 minutes, after reading and recalling the project brief. Further, he took 7 to finish sketching his idea at the end. The session was fully recorded on video with a front and a top camera.

What happened was far beyond what we expected. SP developed alternatives for every design feature, evaluated them by simulating its use mentally and selected the most effective design approach to proceed. (See video 1) He played with different features, manipulated their locations in space to explore the most effective configuration. Halfway through, while simulating the use of design in his mind, he discovered a major functional mistake and reconfigured the new solution quickly, all this in his mind! A quick glimpse at the entire video record showed that throughout the session SP meticulously and mentally sculpted the shape and made sure that he responded to all functional and even production issues. (See video 2 & 3)

Video 1: SP developed features keeping function in mind, evaluated them by simulating its use mentally and selected the most effective design approach to proceed

Video 2: Watch SP as if he is sculpting the shape with his hands, as if he interacting with something real in front of him

SP 3, SP 5


Video 3, 5: Watch SP anticipating even production related issues. Later he assembled the product components with his hands.

He was comfortable taking decisions based on aesthetic judgment, decided on colour and product graphics before he declared that he has completed the design assignment. (See video 4) The detailed account of the session has been published in a paper earlier.2 The only visible difference was that he appeared to be developing the shape, features as well as manipulating and assembling the parts in his mind using hand gestures! The videos bear this out.

SP 4


Video 4: Watch SP take form decisions and refer to product graphics.

Let us return to some of the questions that we started with.

How do we make sure that SP completed the entire design in his mind when he was blindfolded?

How do we make sure that he did not add new ideas during the 7 minutes, when he sketched the final solution? After all, sketching does prompt ideas!

That he was not given enough time to add new ideas during sketching is not sufficient to prove the hypothesis scientifically. To ensure this, the video recording of the session was completely transcribed and later systematically coded. We then listed all the features in his final sketch and matched them with descriptions and references to each feature in the transcripts during the blindfolded conditions. To doubly make sure, we also checked references to features and descriptions in the final summing-up part of the transcripts. Results were surprising. Ninety-five percent of the features in the final sketch had unambiguous matching descriptions in the transcriptions. Obviously, SP had visualized all details in the mind’s eye.

Going beyond doubt

To make it triply sure (validity through triangulation), we asked two new designers to look at SP’s video after the final solutions were edited out. Based on his recorded think-aloud and the gestures, they were asked to reconstruct and sketch the final solution that SP had in his mind. Sketches that both the designers produced were very close to what SP had sketched.3

In the earlier posts, we had emphasized the role that sketching plays as an act, process and as a display in design problem solving. The obvious question that needed to be answered was,

What compensated for the sketching and the lack of visible display?

Is it likely that SP used his mental imagery capabilities to visualize and detail the ideas? For most designers this is not a question worth brooding on. When visualizing, they routinely develop and see their ideas in their mind. It is referred as seeing in the mind’s eye.

Mind’s eye in action

So common is the use of mind’s eye and so real are their experiences, that nobody in the design community ever discusses it, unless someone shows an exceptionally high standard. So, it is not surprising that design literature does list anecdotal evidence on use of mental imagery by the gifted designers. Frank Lloyd Write is known to have visualized the entire idea and details of his famous building ‘Falling water’ in his mind and was able to quickly draft it when Kaufman decided to visit his studio at a short notice. McKim mentions how inventors like Tesla and James watt developed their complete ideas in their mind.4 Mozart had the ability to hear his orchestra and every instrument in his mind’s ear and wrote his final score directly.

Anecdotes and experiences don’t make good science. Besides, there are no accounts of not so gifted designers and creators and their visualization abilities. For this, we need to take a short detour to understand how we use mental imagery and the way mind’s eye works.

On mental imagery and the mind’s eye

Experience of mental image is defined as ‘seeing in the absence of actual visual input in front of you’. To the person experiencing this, the image looks real. (Most convincing and yet difficult to prove example would be experience of dreams) Mental images were not studied because they could not be measured till Shepard and Metzler showed how this could be done.5 Subsequently, there are many studies of mental imagery capabilities. Kosslyn studied mental imagery extensively and listed its characteristics (Fragility, density …) as well as the operations that you can perform on it, like image scanning, image generation and transformations.6, 7 There is also literature that shows how creativity and mental imagery work in synergy.8 With these theoretical back ups the idea of mind’s eye has acquired greater acceptance.

Let us return to the experiment that we started with. Most of the videos above show how SP was continuously using hand gestures to shape an invisible object in front of him. He was obviously working in his mind’s eye. Its virtual-ness turned out to useful, because such a model was quick to manipulate and the change could be ‘observed’ instantly. He interacted with the model with his hand gestures, shaped it, felt the shape and the curves and used the shapes to test if they would work. He used his hand gestures as if he was sculpting a virtual product shape in front of him. (See earlier videos) All these gestural interactions with the virtual model were as real as it would have been with a physical model that he would have created under normal conditions. The gestures were used as much to think and manipulate the virtual object as for communicating the idea.

There is sufficient evidence in research literature to show that there is interrelationship between motor experiences and high-level spatial reasoning. For example, when presented with spatial problems such as mental rotation tasks those who use motor actions (like moving and tilting hands) perform better than those who exclusively depend on visual processes like handling the task in the minds eye. (Ref) That explains surprising accuracy of his gestures and hand movement was surprising. So, when this recording (Audio+video) was shown to two new industrial designers, they could reconstruct the final idea with a fair accuracy. We will focus on the role of gestures and body movements in the future posts.

The structure of the experiment also raised other related questions,

Was the designer’s thinking hampered when he was blindfolded? Was he forced to deviate from the normal design process?

It is difficult to come to a conclusive answer, without comparing this process with the normal process accompanied by sketching. But the transcriptions show that all the typical traits associated with creative problem solving were visible. For example, He systematically identified and tackled all the functional problems one by one. He continued to use ‘moves and reflections’. His moves displayed non-linear shifts, in that he shifted from feature to feature and returned to them again. He iterated extensively, revisiting his earlier decisions several times. His creative explorations remained non-linear.

For most of the ideas that he generated, he simulated its use in his mind’s eye and identified potential problems, and even modified his solutions.

What compensated lack of sketching?

Mind’s eye offered a display that could quickly generate and regenerate image display. It served as a pliable model that he could quickly manipulate in response to his evolving thoughts. It is fast to change, but is fragile and would normally demand budgeting of mental energy to retain and regenerate it. If this is so,

Why the energy budget was not an issue here?

There is no clear answer to this question. I can only venture an answer. Holding images in the mind is indeed difficult. It is true that they need to be regenerated frequently to remain visible in the mind’s eye and that requires budgeting of mental energy. However, most of such findings on energy budget and limitations of short-term memory are based on showing the participants completely new and unfamiliar pictures or words and ask them to recall. As against this, SP used a clear logic and reasoning to evolve the form, which clearly reflected in his speech. So, in case the image is lost due to its fragility, he could regenerate it easily using the logic.

The experiment clearly shows that mental imagery could be one of the viable substitutes to sketching. Perhaps we should correct our earlier statement. What design thinking needs is an ability to represent an object in some form that act as a stable display, that allows you to manipulate it quickly and effortlessly. Such a definition ideally fits sketching, but is inclusive enough to legitimately accommodate other forms of representations like mental imagery.

Could these results have been a freak case? Is the ability restricted to a gifted few? Or is based on years of in design that SP had?

This is a unique ability that designers seem to acquire during their education and practice. In fact, most professional designers who participated in the later experiments told me that, it gives them flexibility to work whenever and wherever they choose. SP himself commented, by using mental imagery “I carry the problem with me in my mind.”

Sum up

The article seeks answer to the question, ‘Is sketching as a representation tool an indispensible part of design problem solving?’ If yes, then this should be treated as an essential skill in design and architecture careers. The answer is explored through a carefully designed experiment, in which the designer is given a design problem to solve and he is blindfolded and thus prevented from sketching.

The fact that designer solved complete design problem when he was blindfolded was ensured by the way experiment was designed. The results show a clear and unambiguous answer that confirms that designers can do without sketching and they compensate this loss with their abilities to create images in their mind’s eye, manipulate them and work with them to develop solutions. In fact, in this case, he created a virtual model in front of him, interacted with it with his hands and altered it willfully. It also showed that he could effortlessly respond to this strange situation and that his design process was not altered.

Mental images are known to be fragile and not easy to work with. They also demand budgeting of more portion of mental energy to retain and process them. So, designer’s visible and effortless switchover to handling of imagery is not easy to explain. Perhaps because the images were generated and regenerated based on his reasoning, he does not seem to face the problem of diversion of excess mental energy. That also explains why designer’s design process does not visibly change.

The designer extensively used hand gestures while generating ideas and for interacting with the virtual model that he created in the front. He perhaps also used them to communicate his ideas. What is worth noting was that his interactions were amazingly accurate.

The results force us to correct our earlier statement. What design thinking needs is an ability to represent an object in some form that act as a relatively stable display, but allows you to manipulate it quickly and effortlessly. Such a definition no doubt fits sketching, but is inclusive enough to legitimately accommodate other forms of representations like mental imagery. Even if designer develops competence is handling one of them, he should be able to make a reasonable headway in design career.

Preview of the next post

When I conceived this experiment reported in this post, I had no confidence that I will discover new findings. Reflecting back, it could have been because of my love for sketching. I was more than surprised by these results and the findings. But it left a nagging feeling,

“Can this result be a freak case? Or is it because of years of experience of designing that SP had?” Or “Is this ability restricted to a gifted few?”

This subsequently led to series of experiments with designers and architects. More about it in the following post.

Notes and references

1 There is sufficient evidence to show that such think-aloud exercises reveals part of the contents of the short-term memory in action. Note that what is captured is what he naturally chose to speak aloud and may not represent everything that passed through his mind. These are referred as think-aloud sessions. Evidence shows that it approximates what he is thinking about. (In fact, most designers and architects are comfortable talking while designing)

2 Athavankar U., (1997) Mental imagery as a design tool. Cybernetics and Systems, 28 (1), 25-42.

3 Athavankar U., (1999) Gestures, imagery and spatial reasoning, In J. Gero & B. Tversky (Eds.), Visual and Spatial Reasoning (pp. 103-128). Preprints of the International Conference on Visual Reasoning (VR99), MIT

4 McKim R. H., (1972) Experiences in visual thinking. Brooks/Cole, California

5 Shepard R. and Metzler J., (1971) “Mental rotation of three dimensional objects.” Sci
ence. 171(972):701-3

6 Kosslyn S., (1983 ) Ghost’s in the mind’s machine, creating and using images in the brain. Norton, New york

7 There are also groups in cognitive psychology who dispute this, leading to what is now termed as mental imagery debate

8 Finke R., (1990) Creative imagery, discoveries and inventions in visualization. New Jersy, Lawrence Erlbaum


Thinking sketches: A messy process and messy results

Almost all of us have learnt sketching and drawing in school. At elementary level, schools focus on the correctness and quality of presentation (colouring) of objects given. The initial objectives are to represent real world objects as correctly as possible. Some have special talent and achieve this quickly. (Picasso was known to have achieved very high level of representation standards in his early childhood). It is only later that you learn to express your views of the world through sketches and drawings. Is not this adequate for career art and design? The answer is yes as well as no!

Visual representations are of course an inseparable part of disciplines like architecture, design and visual arts. These disciplines deal with decisions about spatial creations. All of them demand extensive sketching. That is a reason people believe that if you are good at sketching and drawing in school, design is a good career choice for you.

Sketching externalizes what is slowly evolving in the creator’s mind. However, to lump all the sketching activity in a single category may not be fair. Let is explore how we can split this into classes based on its applications and its location in the art and design processes. Looking at it from this viewpoint, we can see two major classes of sketching as separate. They are, sketching as a display and sketching as a thinking tool.

Sketching as a display

This class includes sketches that are also the end products of artistic activity, and are appreciated for this quality. Most of the works in visual arts fall into this class. (like in caricature, character and scene design or even composition of objects being drawn in schools). These sketches express creator’s thinking and emotions. Semiotics of the representation matters and so does signature style of the creator. These are like signed statement and represents authors interpretations. So, they are valued as an end product. Architects and designers too resort to consciously drawn and well-rendered sketches when they want to display their work to clients and to public. Sketching remains only a means to explain the ideas of buildings and products, so that people see and appreciate. Display remains a keyword in such output.

Sketching as a thinking tool

This class includes sketching meant to complement generation of large number of new and yet unseen ideas of objects and buildings.1 This kind of sketching supports the thinking that is focused on design problem solving, particularly during the initial part of the design process. In such exploratory sketching during early creative phase, designer’s hand is driven by the thoughts and visualized images that are evolving in his mind.

The studies of designers in action shows that the early sketching process includes not just representing spatial ideas or concepts, but reacting to them and thus reconsidering and rebuilding them several times. Architects and designers too sketch extensively in this phase, but draw what they partly or fully visualize in their mind, something that is yet to come in the world. Such a representation must necessarily offer a quick feedback to react to, and allow instant and reversible changes. Architects and designers continue to think and mentally focus on the solutions to the design problem at hand and almost simultaneously sketch their ideas. We have called this class of sketching as ‘thinking sketches’.

On the face of it, freehand sketching should meet demands of both kinds of sketching, but in visual arts and in design, this can happen at the cost of neglecting the differences in the objectives and expectations. This article questions the exclusive dependence on art school tradition of sketching in design programmes. Looking at these differences with a magnifying lens, we hope to argue that during the early creative phase in architecture and design, the emphasis should be also on sketching that can be used as a thinking tool. In this article, we hope to convince the readers that,

“Learning to draw thinking sketches goes beyond the art school teaching of sketching and needs an altogether different approach”

We hope to prove that the sketching taught in schools only partially meets the demands in architecture and design careers. The first section of this article deals with how and why thinking sketches are different and the second discusses the goals and objectives for a course in learning to draw these sketches.

Section 1: Thinking sketches as an end product, as a process and as an act

Let us pick up the magnifying lens and look at thinking sketches in three different ways. As an end product, they are messy and ambiguous. As an act, it is iterative and interestingly it helps designer in not just recording, but also clarify his thoughts. It encourages designer to react, explore what he is looking for, find bugs and induce him to modify the directions of the current thoughts. As a process, it is incremental in nature, building the ideas slowly. It involves reacting to the design problem as well as the solutions being thought of and sketched to search for creative opportunities. So, the ‘way’ the thinking sketches are implemented (drawn), they must prompt continued flow of ideas in their formative stages. This explains why they are being referred as ‘thinking sketches’. How and why do differ?

1 Thinking sketches as an end product and a display

Too cryptic that defies rules

Need to quickly record thoughts and ideas on paper prompt the designer to break rules and procedures of representation. The marks that they make on the paper are like shorthand notes. So, as 2D or 3D representations, such sketches are also profusely annotated when it is faster to record decisions scribbled on a note. (See figure 1) The things that designer feels he will remember are never recorded. The rule followed is, the information is just enough for the creator to reconstruct the ideas and access the logic that had driven these ideas. So, it is no wonder that only the creator can make sense of these sketches. (See figure 2)


Figure 1: Architect’s sketches break all rules of representations. They can be full of annotations, calculations, things to do lists, overdrawn figures, plans, sections and 3D views all mixed together on the same paperIMG_0973_new

Figure 2: Car designers exploring what he is looking for through sketching. Only some of the lines represent some aspects of the object, others are extras! However, in spite of its sketchiness, the creator is able to reconstruct the idea in his mind. (Sketches courtesy Vishnu)

Sketch on top of a sketch

Speed in recording the idea is critical here. So, breaking the rules of sketching, the designers tend to draw next alternative imposed directly on the previous. Thus a single ‘thinking sketch’ may contain many ideas superimposed on each other. They have several alternative overlapping lines representing the objects as well as all its variations. The fuzziness of the sketch at this stage reflects the state of mind not satisfied with a line/curve and becomes a reason to explore more.

The designer knows how to mentally separate these lines to read other alternatives and variations. In this sketches, the creator often reads more than what he had thought of before.1 In revisiting these He also knows which lines are the ‘currently’ final and how to reconstruct the original idea and thinking. For an observer, sketches obviously look like a mess.

Sketches and ideas are distributed across overlays

As the idea develops, there is a tendency to add modifications, either superimposed on the same sketch or on an overlay tracing paper. The efforts end up with stack of tracings with incomplete ideas distributed across them. (In new media, the ideas will be distributed across layers. I have no personal experience of this.)

In a nutshell, the cryptic style of representing ideas, sketching on top of the current sketches and ideas distributed across overlays together create a messy and ambiguous appearance. These are inherent characteristics of such sketches and as we will see in a while that they are important part of the idea generation process.

“The thinking sketches excel by breaking most rules of representations”.

2 Thinking sketches as a process

To imagine that the ideas are available as complete one-shot visualizations will be far from correct.1 Sketching often starts with half formed ideas and these are developed along with the production of the sketches. In our earlier posts on design process, we had touched the iterative nature of design actions without explaining the role of sketching. The question that we must address is ”What role does sketching play (or potentially plays) in iterative processes?”

Ideas to sketching marks and vice versa

In this case, the sketching is expected to contribute to the development of ideas that are just emerging and thus are at best only partially formed. Emerging sketch can be looked at as a display of its current level of completion of conception in the mind. Designer reflects on the incomplete display and reacts, adding new marks that either completes the idea or more often leads to another one, leaving the last one incomplete. As you will see later, the incompleteness of the display is important. The ideas are concretized as the sketch develops, and vice versa.1 The process of sketching is intimately meshed with the iterative thinking process. In fact, in the context of this discussion, they are inseparable.

Reviews to change track

When ideas do not lead to anything exciting, there is a need to change the thinking track. It is a common practice in design to keep intermittently revisiting sketches of earlier explored solutions and reflect on them to discover a solution variation and possibly an altogether new solution direction. The designers review or flip through the stacks of previous sketches to explore if these fragments of solutions in the existing sketches have possibilities of combining in other ways. Thus previous sketches are important sources of ideas during iterations and serve the function of a display to react to.

3 Thinking sketches: A retrospective look at the act

Thinking sketches as an end product tend to be messy, ambiguous and incomplete. As a process too, the actions involved are iterative and with explorations based on pursuing vague directions. As an act, sketching appears to be messy. On the other hand, it appears logical that complete, correct and neat sketches would give a useful feedback to the designer to think of modifications and new ideas. Surprisingly, the messiness seems to have a clear advantage. It may sound counterintuitive.

“Thinking sketches, as an end product are ambiguous, incomplete and messy. As an exploratory execution process it has a messy iterative-ness”.

How and why thinking sketches work?

Why do such crowded and messy sketches contribute to new ideas? To find an answer, we could reverse the question. Why do neatly drawn, well-rendered and pretty sketches don’t contribute to creative efforts? There is always a reluctance to intervene or improve a finished and clean creation, whether it is a sketch, a caricature, a poster or even a cooked new recipe! There completeness discourages intervention. Finished sketches do not leave things ambiguous and are unlikely to act as springboards for new ideas.

MARSJEEPrender r1

Figure 3: Incompleteness of the car sketches suggests that it you can intervene and alter. The rendering is too complete and finished to discourage new design interventions. Pictures courtesy Prof. Sugandh Malhotra

On the other hand, the sketches that are messy and ambiguous encourage interventions.1 Design researchers suggest that this vagueness and ambiguity are critical for generating new ideas. Designer giving a fresh look to a group of lines together often prompts new ideas and new interpretations.

To sum up this section, we can conclude that in the early phases in the design process, the act of sketching is so personal that as an end product sketches communicate very little to others on the team. They are executed to support the creator to develop new ideas. So, it is difficult to conclude that the viewer centered art school tradition of representation is helpful in the early ideation phase. Nor is the excellence in sketching and drawing in school adequate to select design career.

Now that we have some understanding of what thinking sketches are and how they work, we can shift our magnifying lens to the next section dealing with the goals and objectives for a course in learning to draw these sketches. Do we then need to change the teaching of sketching in architecture and design schools? If so,

“What would be the goals of a new learning programme in sketching directed to designers and architects?”

Section 2: Learning to draw thinking sketches

The kind of sketching we are look at should help designer generate and develop his ideas, help in iteration and contribute to clarifying designer’s ideas to them and ‘effortlessly’ record the details. Let us look at this bundle of actions as a cognitive act that makes demands from limited mental energy budget. The budgeting of cognitive energy to drive the act of sketching during the early ideation phase can be divided in two broad activities involved in the act of problem solving.

First segment includes supporting the mental processes engaged in solving design problem and visualizing solutions. We will revisit what we discussed on short-term memory in the earlier post ‘Why do designers sketch?’ and take it forward.2 We know that the ideas are driven by the processes in the brain. Iterations in thinking and solving the design problem and visualizing solutions require budgeting of mental energy. We have seen in earlier posts that designers visualize solutions in their short-term memory. We also know that it is effortful to hold material there, and to work on it and transform it is even more effortful. So, it makes sense to avoid the mental overload by committing intermediate results on to a paper, as most designers do by sketching them out. Sketches serve as a means to record ideas as well as develop new ones, but this requires budgeting of mental energy.

The second segment includes the efforts required to create instructions to sketch, review the sketch during the process of making it, and generate instructions to correct it if necessary. This too demands budgeting energy to draw correct sketches quickly. If one encounters problems of corrections, it is sure to demand additional budgeting of more mental energy and it will be at the cost of energy budgeted for design problem solving.

Why should designer be interested in all this? The problem of budgeting energy for these too activities is a tricky one. We discussed the limitations of memory in earlier post. Kahneman explains this act of balancing the different demands when multiple activities are executed and shows that energy budget allotted to these actions is not consciously controllable.3 (Kahneman D., pp 23-40) He uses the analogy of budgeting of mental energy in directing attention. We can decide (and thus control) “What to do? What to attend?” But we have limited control over how much effort each task will actually use. (This is critical as the nature of the tasks and your abilities to handle the task decides how much mental energy it needs.) Imagine if the designer has difficulty sketching his ideas, it will unknowingly extract greater portion of the mental energy budget and that too, at the cost of slowing down the flow of problem solving efforts.

“Is this a lost case for designers who cannot sketch quickly?”

Not really. They can switch between thinking and sketching serially. But there is one danger. When your mind is occupied with problems of sketching, the idea may slip out of your mind. Haven’t we all experienced that sometimes while writing, a good sentence strikes you, but by the time you complete recording the previous sentence, it slips out of the mind!

It is not a lost case however. We can learn a lot from how people who handle multiple tasks (and now multiple windows on the screen) simultaneously. During driving a car, drivers converse effortlessly with a co-passenger on different topics. Humans effortlessly compose and speak prose. They eat, walk, and also think and hold intelligent conversations simultaneously. After all, did not the Eureka movement occur during bathing? Let is ask a different question,

‘How do humans learn to handle multiple tasks simultaneously? And that too effortlessly?’

All these activities look natural and effortless because of extensive practice and learning that has gone into routinizing these actions. When human actions are routinized they appear effortless and natural. These are often referred as actions in ‘autopilot’ mode. If one of the tasks is routinized, the mental energy can now be spared for the other task/s.

It may appear as a contradiction that you have to work hard and expend efforts to make the act natural. (in our case the actions of sketching). This is not unusual. Cricket shots or football kicks may look effortless and natural. Most learners think that they will be able to repeat the act easily. It conceals the fact that extensive practice (and coaching) has gone into making it look so simple and natural. Watch any Asian eat with chopsticks and you think you can repeat this too! It looks so easy! But is it?

It is known that you become skilled in a task, its demand for energy diminishes. Studies of the brain have shown that pattern of activity associated with an action changes as skill increases, with fewer regions of the brain involved. Talent has similar effects.3 So, the way to solve this problem is to develop skills, to make the act of sketching as natural as possible. Can this root be followed to make it a natural act? This will ensure that the brain focuses its limited energy on solving the design problem and not worry about the problems of sketching. To achieve this, what should be the broad learning goals?

1 Out-of-the-box learning goals

To develop competence in sketching to ensure that sketching is executed with minimal mental energy and ‘comes’ to the designer as an effortless natural act. Let us expand the goal. The idea is to convert it into an effortless mental activity, a routine that demands exercising very little attention. It suggests that the process of representation should become quick, effortless, continuous, and routine and demand only a casual attention from the brain. For convenience, we have referred to it as a natural act. How can you make conscious human acts natural and effortless? Will extensive practice help? The learning programme should not only influence the nature of the sketches, but also the process of sketching.

The discussions so far deal with the act of visualization and problem solving and the act of representation of ideas. Both appear to be intimately meshed together. If the act of sketching is routinized to the extent of becoming near natural, we can spend more efforts on design problem solving. It is known that such an expertise can develop with extensive practice that ensures chunking of initial independent instructions for action into a chunk. The questions that we plan to address are,

“How can we expedite learning to visualize ideas in the mind’s eye?”

“How does one learn to sketch as if it is a routine and natural act that demands very little attention and budgeting of mental energy?”

“Can you reduce time and effort to produce this level of competence by planned and structured efforts?”

We will address all these questions over the next few posts. We plan to touch the first question cursorily here and in a limited context of sketching as it will be addressed in depth in the subsequent post. So, the focus in this article is more on the second and the third questions. Routinizing the act through extensive practice is an obvious solution, but to reduce learning time for sketching is becoming critical, considering that teaching institutions prefer to spend more time on development of design concepts and design thinking.

To cater to these strange demands of effortless sketching one needs to look beyond the design domain. The connection may look strange, but out-of-the-box ideas do need such connections. These ideas are based on sports coaching and what coaches do to players. First, the coaches in sports develop specific short exercises to be extensively repeated by each player so that the response to similar situation ‘comes’ naturally to him. Second, this training is also supported by specially defined exercises that develop the right muscles for a given shot or a stroke in swimming. Third, sport coaches know that you have to ready the body for action by warm ups before launching into action. They have warm-ups and workouts followed by the main task.

“How do these concepts from sports help us in learning of sketching?”

The out-of-the-box source like sport coaching gives uniquely different contents, learning techniques and methods that have proven to be effective in my two decades of teaching sketching to design students. This new approach will be expanded in the next post.

Sum up

Sketching is a word used often, even by children in schools. Do they all mean the same thing? The article points out the mistake of treating sketching as a single inclusive act. It does not contribute to understanding the nuances of the word, nor does it throw light on differences in what it refers to. The article argues that sketching as a tool to think and generate ideas is different from sketching that ends in a display or an end product to be viewed and appreciated. The influence of art school teaching of sketching is justified for the latter, but the former, referred as ‘thinking sketches’, needs a separate treatment.

The article develops the idea of thinking sketches to show how the end product of sketching used as a tool can be messy, ambiguous, incomplete and yet useful for generating new ideas. Next, it treats design as a cognitive act consuming mental energy. So, when sketching is used to solve a design problem and visualize a solution, it will demand most of the limited available mental energy. Little is then available to generate instructions for creating a sketch that records that idea. This suggests that sketching can match the pace of flow of thoughts and produced effortlessly to become a natural partner in thinking, only if it is routinized and demands very little from the mental energy budget. The article then lays down new goals and objectives and suggests out-of-the-box techniques to improve effectiveness and reduce the time that is required for learning to sketch.

The latter part of the article addresses the question, ‘Can you create a structured programme to learn how to draw without having to think about it?’ Such an approach to sketching needs a radically different programme to teach sketching. The article sets up the goals of such a programme and recommends using sports coaching as an analogy to learn how to create thinking sketches. This new approach will be expanded in the next post with videos of how different sports techniques were adopted in classroom situation.

Notes and references

This post is an abridged as well as updated version of the article published on D’Source. You can directly access the earlier version at

1 These ideas expressed here that are partly or fully based on by Gabriela Goldschmidt’s work, particularly

Goldschmidt G., (1994) On Visual thinking: The vis kids of architecture, Design Studies, vol 15, no 2, April 1994, pp 158- 172.

2 Miller G., (1966) The Magical Number Seven, Plus or Minus Two. In Readings in Perception, Eds. Wertheimer M., (pp 90-114). Van Nostrand, New York.

3 Kahneman D., (2011) Thinking, Fast and Slow. Allen Lane, London. pp 23-40


Why do designers sketch?

Art and design, both bring something new in the world through objects and images. Nobody disputes that thinking influences the creations. The creations start as a mental activity, but results deal with real world sensory elements, that you can see, touch, feel and operate. Design thinking often involves composing meaningful physical and visual elements in 2D and 3D space. The thinking leads to intentional actions, so that the resultant ‘composition’ works as a meaningful and effective solution, as close to the original thoughts as possible.1

Such composing is most often accompanied by visualization, but it rarely is a one shot affair. During early creative phase, visualization is base on incremental steps till the clarity emerges. It involves two abilities. First, it deals with working with images in the mind; Second, making them visible in some form. However, these are fragile and are lost if not attended to.

Though visualization can technically remain as a standalone mental activity, it is often supported by some form of dynamic representation/s that keep co-evolving as the thinking explores newer directions.2 In design, initially it is represented in some media and physical form like, notes, diagrams, sketches as well as orthographic drawings, cardboard mockups and so on. See figure 1. Some of these forms of representations are unique to design profession. These actions, and the way they are executed, look like that are natural partners of the thinking process.


Figure 1: Examples of how ideas are represented in early creative thinking. 1) Top: Early sketches of car ideas. Courtesy Vishnu. 2) Early thinking in diagrammes and notes,. Courtesy Ameya 3) Diagrammes to architectural design ideas. Courtesy Kamu Iyer, 4) Mockups with different levels of granularity.

Let us ask a question,

Why is there a need to represent the ideas externally? Why and how do representations contribute to the development of ideas?

To find answers to these questions, we must take a short detour and dig into the ideas on how the brain processes information and the role short-term and long-term memories play. We will then return to design thinking.

Overcoming the limitations of Short-Term Memory (STM)

STM plays a critical role in human actions and specifically in problem solving. Miller3 proved that STM suffers from severe capacity limitations. We can only hold seven plus or minus two items in the short-term memory. Miller also showed that the limited cognitive resource has to be rationed between processing efforts and need to temporarily hold information.

Common examples would convince us of these limitations. For example, you experience these limitations when you learn to drive. You don’t enjoy the drive when you are learning, nor can you give a coherent account of what you encountered on the road! It is a same story when you eat Chinese food with chopsticks for the first time. When you are learning, there are too many micro tasks and sequences of actions to be attended to. With practice, you chunk them together as units. Then you don’t have to process them as micro-steps but attend to them as a chunk. The primary activity (like driving or eating with chopsticks for the first time) takes so much of your cognitive resource that you find it difficult to converse with others around. According to Miller, STM has an upper limit on number of things that it can attend to and process simultaneously.

Here is a short exercise to understand the idea. (See figure 2) There is an unfolded flat cube with 6 sides. You have to mentally fold it and decide if the line closes and forms a continuous loop. Now imagine, while busy with the folding task, if a random list of capital cities were to be read out to you by your friend? Will you remember the cities and get the right answer to the cube problem?

Puzzle_withtextFigure 2: Does the line close into a loop when the opened up shape of a paper is mentally folded into a cube?

The contents of the fragile STM decay and vanish quickly unless you make efforts to retain (rehears) them actively. In solving the cube problem, watch yourself doing it. You complete the folding of first two sides and check if the lines meet on the edge. When you move on to other sides of the cube, the folding operations that you performed little while ago are lost! You divide cognitive resource between folding of planes and holding it in the STM and memorizing the list. The limited cognitive resource available in STM is divided between 1) our efforts to hold the content and 2) process them. If processing demands large part of the resource at the cost of holding it actively in STM and vice versa.

Some of the master chess players, when visualizing and simulating the game in the mind prefer to keep an empty chess board in the front to reduce the contents they would have to hold in their STM. We also encounter the limitations of STM all the time in our routine tasks. That’s why we look for quick and handy forms of representations, like we stick ‘Things to do list’ on refrigerators and Post-its on computer screens. (Traditionally, in India women used to tying knots to their pallu as reminders of things to do.) Obviously, the number of things that we can attend to and process are limited.

Representation and the act of design

We now know that external representations are critical because they primarily serve the purpose of extending the limited capacity of the STM. External representations are part of human strategy developed to overcome limitations of the brain, particularly of STM. To conserve resources, one of the most important strategy that humans use is to quickly represent ideas externally. In design, it includes diagrammes, sketches and quick and dirty mockups. We will concentrate more on sketching. We will start with questions,

How do we overcome the limitations on the capabilities of STM in the act of design? How do we ration the cognitive resource while designing?

Why do designers sketch?

The answer to this lies in the way designers use diagramming and sketching as an extension of STM. But the ideas can’t be routinely extended to understand the role of sketching and the act of design. We will attempt to understand this in two stages. Though there is a clear advantage in external representation, you do need cognitive resource to generate sketches. In the first stage, we will focus on the relative use of cognitive resource in different representation problems. In the second stage, will also add problem solving component and see how the resources are reallocated.

Visualizing, sketching and cognitive resources

We will focus first not on designing, but just on the act of representation. Quick sketching is the most common form of representation. The need for cognitive resource is determined by the nature of sketching tasks and the backup logistics that it expects from the STM. Some tasks may demand more than others. The discussion that follows deals with different representation tasks and indicates potential resource demanded by that representations.4

The idea is modeled in figure below. (See figure 3.1) Let us take a simple task like observing a composition of objects in front and drawing it as seen. (task 1) You may need to hold it briefly in STM and keep refreshing it frequently. But you do have to generate instructions for moving your hand and pencil point. This is least demanding. That’s why children are taught this in schools. Now if you remove the objects after sometime, you must rely on memory. (task 2) You now have to visualize the composition in the STM, hold it actively and refresh it continuously to draw it. That is little more taxing.

Suppose you give a task where the composition remains in the front, but you are expected to draw it from opposite side and that too from a specific viewing angle? (task 3) This makes extra demands on processing, as you have to ‘work out’ and then visualize what it will look like. To overcome processing limitations of STM, most observers do this in stages by segmenting the composition. The processing is similarly demanding if you are given orthographic drawings and asked to draw a cross section! (task 4) Most likely, it will be done in segments and each segment is quickly drawn before moving to the next. Remember the folding of the sides of the cube. The cube problem would have been easy if you were allowed sketching in segments as you solve.

Figure 1-01

Figure 3.1 Modeling the act of problem solving and representation.4

Task 1: [->1 -> 3 -> 4]; Task 2: [->1 ->2 ->3 -> 4]; Task 3 & 4: [-> 1 <-> 2 <-> 3 -> 4 -> 1]

Notice that tasks 3 and 4 demand that you ‘work out’ (visualize) the representation by expending resource and then generate instructions to draw. The task is so heavily processing biased that most people resort to visualizing as well as sketching it in parts. Note that all the tasks so far deal with different levels of processing to understand and/or create representations. None of them demanded any design problem to be solved.

These tasks were recorded in ascending order to reflect increased need of cognitive resource demanded by different representational tasks. The first two are simple sketching tasks typically given to children in schools. The third and the forth require resource to visualize representations. They demanded ‘working things out’ in the head, which requires substantial cognitive resource for processing and subsequently generating instructions on how to represent (draw).

Where is the capacity for design thinking?

Now imagine that you have difficulties in visualizing and in making neat and correct sketches quickly. The problems of making a proper sketch and correcting it will take way most of your processing capacity. Remember the earlier examples of learning to drive and eating with chopsticks?

If implementations take all the cognitive resource, how will additional resource be available for the intensive mental activity of solving the problem in design?

That is why the idea of performing representation tasks (in our case sketching) in autopilot mode, where you expend minimum cognitive resource, makes sense. Can you sketch quickly in autopilot mode?

Added complexities of the act of design5,6

The purpose of the act of design is to find a meaningful solution to the problem that the designers are working on. Designers spend time and efforts in understanding the problem, the context and the functional requirements. The Long Term Memory (LTM) also supports the activity by retrieving relevant stored items into STM. These may include accessing knowledge, design precedents and memories.6 It also takes new inputs from the real world. Is not that the reason to fix those mood boards in the front? See figure 3.2.

It is well known that working out the solution to the design problem as well as accompanying visualization is performed in STM. So, the limited STM capacity has to serve us on several fronts. It is well utilized, if it is used for developing solution directions and visualize how they can work out in space, than on problems of drawing and correcting the sketches.

Imagine now that you are working on a design problem. We need to account for activities like solving a problem that make additional demands on STM. Let us plugin a hypothetical process of problem solving in the earlier model.7 Roughly, we can look at ‘design thinking and problem solving’ as primary activities leading to ‘visualizations’ and subsequently to ‘sketching or mock-ups’.

Figure 2-01

Figure 3.2 : Modeling the act of problem solving and representation.

Task 5: Iterative [5 <-> 2 -> 3 -> 4] -> [ 4 -> 1 -> 2 <->5 <-> 2 -> 3 ->  4]

What are the new plugged in activities? The design thinking is now focused on understanding the problem, generating new ideas and evaluating thoughts and ideas. (Task 5) It demands designing and design thinking and might require that we bring in contents from the direct perception of objects and events in the real world or as memories from LTM and handle it actively in STM. This requires substantial cognitive resource to be deployed.

Next, as the ideas develop, they lead to visualizations of arrangements of elements in the fragile STM. But, this is not a one-way flow. (5 <-> 1). It may start with tentative visualization of ideas that are quickly evaluated and altered in the STM. This cyclic back-and-forth flow is likely to lead to incremental development of design ideas. What are its implications?

Imagine if we had decided not to make those quick and messy sketches and notes? It would mean holding the ideas as visualizations in the STM. Either you have to keep rehearsing it to hold it as an active display in STM or allow it to decay. Quicker you record the ideas, the better it is. So, cognitively less demanding quick representations becomes critical. Take an example. In writing, you suddenly get a wonderful sentence in a flash, but you may loose it, because part of your STM is busy instructing the hand to completing the previous sentence. That is why quick and messy form of recording (noting and sketching) is required.

In design, the representations take form of rapid sketches, diagrammes and quick and dirty mock-ups. Once the current ideas are sketched out and are available to the designer as external entities, capacity starved STM is relatively freed; it can now allow 1) new material to enter and 2) new ideas to evolve in STM. One should treat rapidly recording external devices like sketching as extensions of the STM.

It is advantageous to spend some cognitive resource to generate instructions to the hand muscles, to record them as sketching/diagrammes/notes. Even though they are incomplete and ambiguous, they do serve the purpose. So far as they capture the right contents and reflect the intentions of the thinker, it does not matter that they look messy and confused. Besides, the action releases STM of its burden. In short, these external recordings truly act like an extension of STM, relieves it of the efforts needed to hold the ideas and help to overcome the limitations on its capacities.

So, in spite of its limitations, the STM can handle resource hungry problem solving provided the acts of representation (sketching) do not demand too much resource. Else, the main activity of solving a design problem and visualizing new ideas that steer the thinker to an effective solution are starved of cognitive resource. Most creative efforts are process intensive and make heavy demands on severely limited processing capacities of STM.

Sketching, when the designer are executing it, works like a ‘online’ dynamic depictive display unfolding in front of his eyes. The act of observing the display is important, because it serves as a major input for self-criticism and further incremental development of ideas. Of course, the display needs designer’s continuous attention, which entails expending continuous cognitive resource.

Neat and clean process?

The way the process is described, the operations look too clean and sequential to be realistic. While design thinking and visualization process has its eureka moments, it is often not a one shot process. You iteratively reorder, scrap, rebuild and transform elements till the design intentions are met. The advantage of sketching is easier to understand if you treat act of design as a temporal and cyclic event consisting of series of iterative actions that record ideas. The complexities of handling the temporal iterations obviously demand moving contents in and out of these circles shown in figure 3.2. Such operations are even more cognitive resource intensive. But once the external record is available, even if it is only in parts, STM is relieved of its burden to hold them and is free to take up the next problem solving as well as visualization challenge.

Thinking sketches

Designers use sketches as intermediate representations of the ideas evolving in their mind. They can be called as ‘thinking sketches’. They are messy, incomplete and often ambiguous. The value of these sketches as end products is often not important, at least till you become famous and research worthy! In most sketches, they only represent objects that will be produced on different scale, in a different context and in different materials. For them early sketching is a means of supporting the evolving thoughts. This type of sketching should be treated as a different class. Though they are not important as end products, they have a key role in design thinking.8

We saw how quick and messy representations are key to design thinking. But there is much more to designer’s ‘thinking sketches’. We hope to discuss this exclusively in the next post.

Should design career be reserved for those with good drawing skills?

The discussions so far suggest that, sketching is a critical activity in idea generation. This also leads to interesting questions.

Should design be taught to those who already know how to sketch and draw well? Or should we teach ‘thinking sketches’ and drawing to those who are passionate in pursuing design carrier? And lastly,

Considering that design thinking is widely used in different types of problems, which may not have visual component, are there other forms of representations that could compensate?

In my research on representations, I tried to conceptually explore all contradicting possibilities. We hope to discuss these questions over the next few posts.

Programme to learn ‘thinking sketches’

Our roots in visual arts have always influenced the teaching of sketching in design schools. Would the nature of designer’s sketching be different, particularly during the early creative phase in design, if the design profession had emerged independently? This is a hypothetical question, but answer seems to be yes,

How does one learn to draw thinking sketches? How could such programme for sketching be developed?

Design career should be opened to students who show a creative promise and are thinkers, who could then be taught sketching to support their ideation. I tried to develop a course for sketching based on the conviction that early ideation stage need different kind of sketching. Logically it follows that such efforts should shun the art school influences (at least partly). So, I build the course ground up from scratch.

The focus of the course was to minimally depend on use of cognitive resource while sketching. Observe that we walk, swim, eat and even drive in autopilot mode (i.e. with only a cursory use of cognitive resource). So, why can’t we sketch in autopilot mode? It made sense, because during early creative phase, designers sketch to support problem-solving process. So, they must allot their cognitive resources for their problem solving efforts. Wasting this resource on making correct sketches entails depriving the creator from using it for the main task of solving a problem. So, the focus was on ‘How can we make sketching a near natural act, like walking and eating?’

The course ensured that the students learn to draw quick and correct sketches in a short time. The focus was on correctness of the sketches than on style of sketching. It deployed out-of-box ideas to shorten ‘learning to sketch’ duration. It insisted on using the entire body to participate in the process of sketching. So, it focused not only on moving the hand, but also moving the body and assuming correct postures. I plan to discuss this in the next post.

As mentioned earlier, design thinking is being practiced by many, and they are unlikely to be skilled enough in sketching. Are there forms of representations that can effectively substitute sketching? Can other forms of representations support creative thinking and problem solving? We will have quick glimpses at the other options available, but deal with them extensively in later posts.

What if we prevent designers from sketching?

Out of curiosity, I challenged, the notion that sketching is an inevitable form of representation in designerly thinking. In scientifically conducted experimental series, the designers and architects were blindfolded and they were asked to design. When prevented from sketching, designers used their ability to internally represent their creations in their mind’s eye (mental images). These served as an effective depictive display to react to, almost as the way sketches on paper functioned. Designers seem to have amazing abilities to effortlessly generate new images and manipulate them willfully to solve problems. All of them were able to react to their mental images, transform and manipulate them and at times reject what they did not like, almost as if they were working on paper.

Can this be used by students of design thinking? Is it possible to train oneself in use of mental imagery?


Generating gestures and word strings

Directions attempted so far, is the use of combination of gestures and word strings to represent objects to their teams as well as to self. Together they contained description of evolving shapes and compositions. We are trying to simulate the early thinking during ideation.

I am now trying to concentrate on exploring forms of representations that can be naturally acquired, easily learnt and do not demand much of cognitive resource to be utilized. It is work in progress, but it is likely to be a mix media communication. If potentials of design thinking have to be explored, this will be a critical step.

There are several indicative examples. Ideas in brainstorming sessions are not necessarily sketched. Many times, in face-to-face conversations, ideas are explained using gestures, with only a marginal support from sketching. It is important to explore other forms of representations that could support creative explorations in non-spatial problems. We will pursue this in later posts.

Sum up

In this post, we were exploring design behavior in the early creative phase and have addressed the questions, ‘While solving a problem, why is there a need to quickly represent the ideas? And subsequently, why a mental activity like design thinking quickly ends up in sketches?

Design thinking often leads to visualization and later to its quick representations. The limited cognitive resource that STM works with demands that we ration it carefully to different mental activities involved in design problem solving. We tried to develop a model of these activities in two stages. We prefer to quickly represent ideas externally to conserve limited cognitive resource, which otherwise would have been expended in holding the ideas in STM and work with them.

External representations, particularly visual representation like notes, diagrammes, sketches and mock-ups work as extensions of STM. They have key role in idea generation. As a prelude to the future posts, we concluded the discussion by touching other possible forms of representation.

It is clear that sketching as a form of representation helps designers to overcome limitations of STM. We referred to them as thinking sketches, which serves many more functions than just conserving cognitive resource. In the next post, we will discuss the nature of thinking sketches and how they work. We will also discuss wonderful things that designers do when sketching, along with my efforts to develop a special course to how to draw thinking sketches.

Notes and references

1 This is true of many creative art efforts. We intentionally compose sequence of words to create prose and poetry. We compose musical notes into a composition with intent.

 2 In most cases, representations appear to be critical, however, the way thoughts are represented is largely discipline specific. Designers sketch, musicians play or write score, authors write or word process their thoughts ….

3 Miller G., (1966) The magical number seven, plus or minus two, in Readings in Perception, ed. Wertheimer M., Van Nestrand, New York

4 In spirit, the figure is somewhat close to the interactive visual imagery diagramme that Mckim proposed, but is developed further to include cognitive perspective. Also see

McKim R. H., (1972) Experiences in visual thinking. Brooks/Cole

5 It is not based on specific research on act of design, but extrapolated from what is known of STM. What is explained here can be considered indicative.

6 Discussion on art thinking is not included here. I did not have the fortune of studying how artists think and the kind of problems they solve. So, the discussion is limited to designer’s work. The model proposed will have to be altered to include art. My limited understanding is that the purpose why artists sketch is to explore and present a personal point of view of seeing the real world. This involves intense initial contemplation and self-searching to get clarity on how he should act. Artists are known to keep gazing at the empty canvas in front till some clarity emerges. Initial intermittent representation actions become more frequent as they advance and the depictive display in the front guides their thoughts.

7 There is no way to figure out the actual information flow in the act of design experimentally and show how it should it be represented.

8 Designers also make well rendered sketches as finished products. They are more presentation heavy and are a different ball game. This post does not deal with such cases.