You are here: Articles --> 2006 -->
Abstraction: making the complex easier to understand
Vous êtes ici : Essais --> 2006 --> Abstraction: making the complex easier to understand
by Geoff Hart
Previously published as: Hart, G. 2006. Abstraction: making the complex easier to understand. Intercom November 2006:36–37, 40.
Some things are difficult to understand because of their inherent complexity. Fortunately, we have many tools available to simplify that complexity and make difficult concepts easier to grasp. One of the most powerful involves progressively eliminating non-essential details until only the key details remain and can be grasped in their newly simplified form. In addition, we can also approximate the truth rather than striving for complete accuracy. Last but not least, we can use analogies with what our audience already understands to smooth the path to comprehension.
These approaches work, particularly when combined, because once readers understand a simpler version of the truth, they can build on that understanding to tackle a concept's full complexity. The process of simplifying by removing details, thereby increasing the "distance from reality", is called abstraction. Abstract art earned its name because increasing the amount of abstraction distances images from what we can see in the real (non-abstract) world. Abstraction progressively removes details so we can focus our limited supply of concentration on those that remain. Indeed, I recently read of one study (see my bibliography) that reported slower and less certain recognition of realistic photos of people than of caricatures (abstractions of faces). Apparently, our brains may analyze faces by simplifying them into the key features that caricaturists preserve—that is, into abstractions.
Consider a common tool we use for communication: an image. The most abstract depiction of that image might rely solely on words: words are not themselves real physical objects, and must instead be "mapped" to (matched with) real-world things that we can perceive. If I write that I'm going to tell you about a pipe, you might think plumbing until I clarify the context: a device used to burn tobacco. But you won't know whether I mean a cheap clay pipe or a costly meerschaum, and even if I specify clay, you must still imagine what it looks like.
If all I want to communicate is the concept of smoking, the words alone suffice. If I want you to see the same type of pipe I see, I must choose a less abstract way to communicate. The obvious solution would be a photograph or a realistic drawing. But even such images, which are about as close as I can come to the intended concept without handing you the pipe, are abstractions. French artist René Magritte famously captured this concept in his drawing ceci n'est pas un pipe ("this is not a pipe", <http://en.wikipedia.org/wiki/Magritte>). As my hommage in Figure 1 shows, graphics are only representations, not the things themselves.
Figure 1. Ceci n'est pas une plume. ("This
is not a pen.")
Note: This image was purchased from <iStockphoto.com> and should not be downloaded for reuse from my Web site. To obtain a copy of the image for your own use, visit their Web site and search for "feather quill".
If neither words nor images are the actual things they portray, how do we make them useful? By recognizing that both words and images must accomplish a specific purpose. To abtract reality using either words or images, we must start with two questions near and dear to the technical communicator's heart:
The answer to the "need to know" question defines the key elements we must present; the answer to the "tools" question tells us what tools we can use to communicate those elements. The information designer's challenge is to match the tool to the concept so as to communicate effectively.
The example of Magritte's pipe (or my pen) is clear, but the concept being communicated is subtle: it takes a moment of thought to understand the paradox. The power of abstraction becomes more apparent when applied to more complex situations. Consider, for example, the concept of genes and mutations. I'll start with an explanation that's only marginally comprehensible to anyone who hasn't studied modern molecular genetics, then show how progressive abstraction simplifies the concept until even a non-geneticist can understand. Hang in there for a few more paragraphs: it's not as bad as it may at first seem.
Consider an overview of the topic in the most abstract and intimidating form possible: words alone. In a living cell, genes are blueprints for building proteins. Proteins are assembled from components called amino acids, and the order of these components is defined by the genetic blueprint. This blueprint defines amino acids using groups of three bases (triplets). There are five bases, represented by the letters A, T, C, U, and G, and triplets such as CAU represent a single amino acid (in this case, histamine). Different triplets code for different amino acids, and different combinations of amino acids produce different proteins. If we consider genetics as a language whose vocabulary is composed entirely of three-letter words, then combining the correct words, in the correct order, creates sentences that communicate clear instructions to the cell containing the gene. In one common type of mutation, one or more bases change; this changes one or more amino acids, resulting in a different protein and thus, a different sentence.
You can visually represent the bases in a gene using the standard "ball and stick" notation that shows the actual atoms (Figure 2), but only a biochemist would recognize the gene from this image. We can do better. Let's represent the bases using the letters A, T, C, U, and G (Figure 3). Eliminating the three-dimensional complexity of the atoms lets us concentrate on the sequence of bases, which is the important thing. To further focus attention on the groupings, I've inserted vertical bars (|) to separate the triplets that represent individual amino acids.
Figure 2. Three-dimensional structures of the main amino acids.
Figure 3. Simplifying a series of amino acids using letters rather than a chemical formula.
Figure 3 is clearly simpler than the 3D presentation, but again, only a biochemist would recognize the amino acids and decipher the sentence they form based purely on the triplets. Moreover, if we want to explain mutations by showing how changing one or more bases changes the protein, Figure 3 won't help: the groups of three letters mean nothing to a non-expert. But if we extend the metaphor that genes represent chemical sentences, we can further simplify the explanation by replacing these triplets with short, familiar words. Figure 4 provides the final abstraction necessary to explain how one type of mutation results from changes in a single base: replacing a single letter changes the sentence in an instantly recognizable fashion.
THE|DOG|BAT|THE|CATFigure 4. Changing a single letter dramatically changes the meaning.
A comparable genetic change represents a mutation, and just as the English sentence became incomprehensible, mutations may make a genetic sentence incomprehensible. The reality is considerably more complex, but by progressively making the explanation more abstract, I've simplified it to the point at which it's relatively easy to understand. This specific example is based on the brilliant explanation provided in Essential Genetics, 2nd edition, by D.L. Hartl and E.W. Jones. Robert Sapolsky, writing in the March 2006 Discover, demonstrates a different form of mutation—insertion of a base—by wondering rhetorically whether one would rather have a mouse for dessert, or a mousse, and whether it should be served in a bowl or a bowel.
As an information designer, I've used these examples to solve the problem of communicating a difficult concept by answering the two questions I presented in the previous section:
Combining the two answers explains the complicated genetic description of mutation by eliminating everything but the key details, thereby accepting a description that is not completely accurate, and by taking advantage of cognitive tools the reader already has (the ability to recognize incorrectly spelled words).
Though I haven't made you into a geneticist with this explanation, I hope I've made one fundamental concept of modern genetics clearer than it might otherwise have been. Simplification (abstraction) of the form demonstrated in this article provides an incomplete understanding of the full complexity of a truly complex concept, but even that incomplete understanding is better than no understanding at all. Indeed, it can provide the basis for a more complete understanding. In coming articles in this column, I'll explore some additional forms of abstraction that are more directly useful to technical communicators: metaphors and graphics.
Mauro, R.; Kubovy, M. 1992. Caricature and face recognition. Memory & Cognition 20:433–440.
[A look back from late 2006: Did you spot the "Easter egg"? Okay... I confess it's just a tad obscure. Here's a hint: it's in Figure 3. If you substitute the initials that biochemists use to denote the amino acids encoded by these triplets, you'll have a secret message. The only trick you need to know is that the amino acid "glutamine" is abbreviated as an "E" by biochemists, perhaps because all the easier-to-remember letters were already taken.—GH]
©2004–2018 Geoffrey Hart. All rights reserved