“There isn’t any more truth in the world than there was before the Internet or the printing press. Most of the data is just noise, as most of the universe is filled with empty space.”

Nate Silver, The Signal and the Noise

1 A difference that makes a difference

Information consists of data composed according to the chosen syntax. Well-formulated, meaningful data provide semantic content. The elementary unit of information is ‘a distinction that makes a difference’ (D. McKay), often referred to as ‘a difference that makes a difference’ (G. Bateson).

The definition used in information theory describes the distinction as a discrete state, a single data item. When we move beyond the level of noise and into a human context, it’s meaningful bits of information that we analyse and process into gems of insight, reflection and synthesis into knowledge and hopefully love, peace and understanding. For this reason, the value is in the synthesis of the information, not in the data itself.

2 “One needs to see clearly in order to understand.” — Le Corbusier

Information is useful only if it can be understood. Since our ability to collect exponential amounts of data doesn’t match our evolutionarily stable ability to process and reflect on it, we resort to data visualisations or infographics.

The choice of tools and strategies for presenting data is never logical, rational, objective, or neutral at the same time – nor is the technology itself. They’re chosen in an ambitious attempt to meet the needs of users while respecting the conventions, constraints, and possibilities of data design. Even with the best intentions, each data representation highlights some things and hides others, often promoting or denigrating certain moral values and norms.

On the visual side, data design strikes a balance between high-resolution, data-rich visualisations that are often complex and multidimensional, and abstracted, simplified notational systems that condense the dimensionality of data into flat representations of bland generality.

3 “The simpler the form of a letter, the simpler its reading” was an obsession of beginning constructivism.” — J. Albers, Interaction of Color

Roughly speaking, if the presentation of data is to improve understanding or increase the credibility of evidence, dense visualisations should be used. If the goal is to provide actionable insights, a simpler data design makes more sense.

(Of course, there is a point of simplicity at which a data design is no longer necessary and the use of a signal is sufficient. If the data design provides overtly trivial results, perhaps only a verbal synthesis should be used.)

Unfortunately, misconceived visual efficiency blurs this distinction. The notion that all visual queries should be processed quickly is simply wrong. The resulting belief that all design should lead to rapid understanding can only be explained by years of passive-aggressive relationship between graphic design and marketing.

The obsession with expressing any kind of data in layers of standardised charts, graphs, and diagrams became part of the design toolbox and has been used enthusiastically ever since. Since then, data design has become synonymous with processing numbers into pictorial statistics rather than developing elegant visual statements.

“What we seek instead (the simplicity of data design) is a rich texture of data, a comparative context, an understanding of complexity with an economy of means”

E. Tufte, Envisioning Information

4 Clarified Ignorance

Simplicity of data presentation does indeed contribute to clarity of reading, but it has no bearing on clarity or quality of understanding. Clarifying complexity is different from simplifying complexity.

If you selectively engage with the data by cherry-picking the parts that seem familiar and ignoring the context, you put the user in a state of clarified ignorance where they wrongly assume they’re well informed about things that add little to their knowledge or reasoning.

5 Information and Ideas

In The Cult of Information, Theodore Roszak suggested that the mind thinks with ideas, not information. Recognising connections leads from raw data to information, recognising patterns turns information into knowledge, but only understanding principles turns them into wisdom.

The flood of data will continue to increase, but “there is no more truth in the world than before the Internet or the printing press.” The solution may lie in presenting information worth presenting visually, providing narratives that incorporate the subtleties of human context rather than highly abstracted diagrams, and, above all, developing tools that explain complexity by encouraging attentive thinking. Data is only part of a story.

“In this context of information design, with its characteristic emphasis on users’ requirements, an awareness of the semantic dimension becomes all the more apposite: as a continual reminder that understanding is more than just a reception of messages, but entails a construction of meaning and that this ‘meaning’ is subject to influence from a very large set of factors.”

Robin Kinross, Unjustified texts *

Shneiderman’s OZFD model for visualising large data sets:

Visual representations of data should provide:

  1. Overviews to define the scope and categories of a database and provide context for attributes
  2. Zooming to access information at different levels of granularity and to show the user the relationship between levels
  3. Filter to limit the type of information they zoom in and out of based on attributes
  4. Details on demand to allow users to access the finest levels of granularity as needed without cluttering the user interface
    (via: Janet H. Murray, Inventing the Medium)