Flat Earth Tech Comm

By | 2013/01/21

Working with my current client has really reinforced for me how much traditional documentation methods involve flattening reality. The client is dealing with a large body of troubleshooting information, in which there are complex relationships between issues the user experiences, the symptoms that help narrow down the issue, the configurations under which symptoms can occur, and the underlying faults, misunderstandings, or frustrated expectations that cause the issues.

The relationships between these factors are complex, many to many, and multi-dimensional. Any presentation of these relationships on paper, or on a piece of glass standing in for paper, involves flattening them. This flattening is necessitated by the media, which is, literally, flat. It is also a result of the limits of our own ability to visualize and express complex multi-dimensional relationships. Whether it is a result of thought patterns developed by a five-thousand year experience with paper as the media for externalizing and preserving our thoughts, or whether the limit is innate, we flatten as we attempt to understand.

The problem is, when we flatten, we distort. Things that should be close are forced apart, things that should be distant are forced together, angles are twisted, three and four dimensions are compressed into two, every valley is exalted and every mountain and hill laid low. If the flattening is necessary to comprehension, it also distorts the thing we seek to comprehend. The distortion increases the cost of understanding, and of acting on information, sometimes with significant economic consequences.

Mercator projection

The Mercator projection flattens a 3D world onto a 2D page, distorting angles, sizes, and distances in the process. Image: Wikimedia Commons.

There is nothing new about this flattening. Examples abound. The problem of how to map a round planet on flat paper has produced a variety of “projections”, each of which distorts the reality of the world in different ways. The classic example is the Mercator projection, which famously exaggerates the size of Europe compared to Africa. It is also particularly generous to Canada, and makes Alaska appear almost as large as the entire continental US.  It represents the poles, which are points, as lines equal to the length of the equator, makes the lines of longitude never meet, and distorts the distances between the lines of latitude. Despite all these distortions, however, the Mercator projection is valuable for marine navigation because it makes rhumb lines straight. In a flattened view of the world, we distort everything else to make a single property easy to work with.

And while a globe represents the world accurately, we find it a difficult image to understand. The flattening of Mercator and other projections may distort, but it also presents the world in a view that is more easy to take in. Our brains find flat easier to deal with.

Perspective drawing deliberately distorts the size of distant objects compared to foreground objects, to create the illusion of depth, but in doing so it destroys the ability to judge relative sizes. The eye accepts the illusion, but for technical purposes, we need to flatten differently, in the form of projections such as the isometric, which preserves dimension but distorts angles and perspective. Again, we flatten in different ways, distorting everything around a single axis that we find useful for a single purpose.

In text, we flatten even more, reducing not to two dimensions but to one: a linear progression of text. No matter how complex or multifaceted a subject may be, when we write about it, we pick a starting point and unravel the picture into a single string of words. Just as the flattening of a drawing distorts distances and relationships, so the linearization of  text distorts the distance between related concepts by moving them apart in the course of the narrative.

When we come to organize content we venture a little into the second dimension by laying out a hierarchy of topics. The hierarchy generally represents what is actually a linear sequence, but it folds the line to present it as a kind of map of the content. But this  hierarchy/map is still a flattening of the real world that the text seeks to describe. Like any flattening, it can only give a true representation along one axis at most, and all else must be distorted. This is why there is no one generally useful hierarchy, and no one generally useful flattened presentation or organization of content.

In text and in line, we have built a civilization and a science on the flattening of information. But this does not change the limits that flattening imposes on our powers of expression, and our ability to convey and acquire real understanding of the world we live and work in. But these restrictions belong to an age in which the predominant tool for the extension and sharing of our intellect was the sheet of paper.

We have a new tool now, and one that is not restricted in its capacity to capture and express multiple dimensions of data. The computer, and, perhaps more importantly, the computer network, are not dimensionaly restricted in the way paper is. They allow us to represent and explore worlds and problems in multiple dimensions. We may not be able to form an unflattened, undistorted picture in the mind’s eye, but we can represent it in the computer’s memory, and create algorithms that can process it and produce usable results from a representation free from distortion. We no longer have to flatten to represent. We no longer have to flatten to understand.

And, we no longer have to flatten the organization of content to make it navigable. We can make it navigable in as many dimensions as we require to represent the multiple dimensions in which concepts and actions may be related in the real world.

Of course, the reader will still receive text as a linear sequence and drawings projected on a flat screen. But we now have the ability to allow the reader to rotate their view, or, through linking, to traverse the information space along multiple axes. We don’t have to give them flattened content. We can give them a projection of multidimensional content and the tools to manipulate the projection as they see fit.

In The Wrath of Khan, Khan proves to be a superior tactician to Kirk, until Spock points out that Khan is limited to thinking in two dimensions.

Like Khan, our training and experience has limited us to thinking in two dimensions. To no small extent, the classic art of writing has involved the artful flattening of information. But our modern tools don’t require such flattening, and flattening, no matter how artful, is now an obsolete skill. We need to start thinking, and designing information, in three (or more dimensions).

Books, or collections of topics with linear dependencies between them (Frankenbooks) are one dimensional. A collections of Every Page is Page One topics create the opportunity to organize, and allow the reader to traverse, information in multiple dimensions.

5 thoughts on “Flat Earth Tech Comm

  1. Jeff Coatsworth

    Psst, Mark – don’t get the fanboys riled up – it’s “Khan” not “Kahn”

    1. Mark Baker Post author

      Fixed, thanks.

      Not so much the fan boys I should worry about getting riled up, I suppose, but Khan himself.

      Wondering why I never noticed till now that the WordPress editor does not have a search and replace feature. It took me a year to discover it didn’t support tables. Two years to discover it doesn’t have search and replace. What a difference working on small pieces makes!

  2. Marcia Riefer Johnston


    You lay out the map-vs-globe example beautifully. Rhumb lines. Who knew? I especially like “every valley is exalted and every mountain and hill laid low.”

    I’m intrigued by your point that “while a globe represents the world accurately, we find it a difficult image to understand.” Usually, “accurate” and “usable” go together. You remind us that accuracy alone doesn’t guarantee understanding, at least when it comes to multidimensional information. Oooh… lots to explore there.

    1. Matt

      Usually, “accurate” and “usable” do not go together. This is why we do not measure times with timers that show ten or so digits, unless we need to. Usable goes together with accurate enough, where enough depends on the context.

      When you need to estimate (in your head) the circumference of a circle, you’ll think of pi as 3, or maybe 3,1. When you write a schedule, every written time is with accuracy in minutes and ends in 0 or 5. This is because a truly accurate schedule would be a bad schedule. The more accurate any information gets, the harder it is to digest.

      1. Mark Baker Post author

        Thanks for the comment, Matt. I think we need to make a distinction here between accurate and precise. A globe is an accurate representation of the world, but not very precise. A Mercator projection is neither accurate not precise.

        We need our instruments be be accurate, and to be sufficiently precise for a particular purpose.


Leave a Reply