First Things First: Defining Information

Definitions are neither true nor false; they're useful or useless. (Stuart Kauffman, "The Adjacent Possible" https://www.edge.org/conversation/stuart_a_kauffman-the-adjacent-possible)

Before getting into a working definition of information, try this experiment: Give concise definitions of the terms, Data and Information. (This is a writing test I gave to job candidates. I was looking for a conversation, not a correct answer.)

It turns out information has many different definitions depending on your discipline, whether some branch of philosophy, engineering, economics, biology, etc. Most definitions seem to fall into the following traps:

• Irrelevance
• Ubiquity
• Circularity
• Complexity

Do your definitions avoid these problems?

Examples are in order. Claude Shannon says (roughly) that information is the amount of entropy in a signal. You can get a quantity of stuff, but you don't know what that stuff is - Irrelevant. Cesar Hidalgo (an economist) says "Information... [is] physical order." (in "Why Information Grows") - Ubiquitous. If everything is information, the word becomes meaningless. "Semantic information is well-formed, meaningful, and truthful data." (Luciano Floridi, "The Philosophy of Information") - Circular and complex. Information is data, and data is information. Explaining how data can become "truthful" is hard, if at all possible.

Most of these definitions try to describe information as physical stuff. You can appreciate the urge... You can measure stuff. You can put it into a machine to process, package, and sell. You can calculate ROI. If information is only in our heads, that's harder to do. (Well, folks are talking about the thermodynamics of civilization and culture, which might offer insights. But bleak turns of their outcomes related to planetary physics could moot the effort.)

Information in your head is exactly what I'm talking about:

• INFORMATION - A theory about a data set
• DATA - A record of one or more events

This makes it easy to lay out the production of information. Start with the big bang, which produces events. Then each event produces data. An energy shift in an atom emits a photon, and a year adds a ring to a tree. When a mind looks at the stuff of data, it can build theories. A red shift in a galaxy means the universe is expanding. Blackened rings in petrified trees mean there was a fire long ago. By definition, information is not a form of data. Data and information are completely different; one is physical and the other is psychological. Period.

Do you see the sleight of hand? We just kicked the circularity can down the road. How do we recognize an event? We have data. What is data? The record of an event. Hey, we're technical writers with deadlines. We're too busy to solve the philosophical problems of the universe. The point is that we draw a clear line between data and information, and that line is useful. As a bonus, it's so clear that even an executive can follow it.

Another complaint to deal with... You could restate the definition as "Information is a description of how things are." This is face-plant simple and flirts with tautology. "It is what it is." At best it's painfully obvious, so don't we already know all there is to know about it? If there's any there there, why no scholarly articles to beat it to death? Two points.

First, it's hard to get funding to state the obvious. Academics need grants, institutions need prestige. Who's going to cite your paper if you just say the sky is blue? But more importantly, the interesting stuff here is hiding in plain site. If you look closely, something that appears to be simple can be, and usually is, rich with detail. (Mark Rothco comes to mind.)

And that's what this newsletter is about. Can we say what knowledge and wisdom are? What do people do with information? How can writing inform? Can information be wrong? And simultaneously useful? What things do we wrongly call information? Is AI informative, and how? (Is AI even I?) Are all organisms informed? If not, where's the line? Poking at these questions opens neat problems. Demented as it seems, it's great fun.