How to blog insightful

This is a guide on how to blog insightful. Not insightfully. Because insightful isn't a writing style, it's a type of idea.

There are hundreds of guides to writing like a writer. This is a guide to thinking like a writer.

I suspect that most of us view insightfulness as having this magical and unpredictable character. This is not the case. Insights aren't something you have, they're something you create.

Let me explain.

When mathematicians are proving conjectures, they rely on a standard collection of tools and methods. Like mathematicians, writers also have a standard bag of tricks to draw from. What follows are a few tricks that writers use to create insight.

Specifically, this essay will explain both what makes something insightful, and also where insights come from.

I’m writing not from the perspective of a great writer, but rather from the perspective of someone who knows what creates value for him as a reader.

===What Makes Something Insightful?===

The Oxford American Dictionary defines insight as “the capacity to gain an accurate and deep intuitive understanding of a person or thing.” If we accept this definition, then insights seem to be tools of sorts. Very special tools that help us make sense of the world by illuminating the patterns and relationships between things.

And as tools, insights have the same properties as a hammer or a hacksaw. Let’s brainstorm some ways of writing about tools that would create value for the reader:

We can give readers a new tool.
We can teach them how to use an existing tool more effectively.
We can teach them when and when not to use a tool.
We can replace an old tool with a better tool.
We can replace a complicated tool with a simpler tool that still does the same job.

And so on.

Each insight is like a new tool. And just like a table saw, we must first make the reader aware of this tool, and then teach them how to use it, when to use it, why to use it, what to use it for, etc.

So how do we do this? Through the creation and contextualization of schemas.

Like insights, schemas are models for helping us comprehend phenomena. However, not all writing that deals with schemas is insightful. For example, when we see a pen we know that it is a tool for writing on paper. We know this because we have a schema that connects pen with paper for the act of writing. However, there is nothing insightful here because we are simply using an old schema. Insightful writing is only that which creates value by offering new schemas or helping us better understand existing schemas.

The good news is that there are some common patterns in the ways that schemas are created and connected.

===Patterns in Schema Creation===

Define a phenomenon

A person defines a phenomenon by finding a pattern, describing it, and giving it a name. This creates value by giving others the ability to collectively think about the phenomenon.

A good example of a recently defined phenomenon is Yak Shaving. Yak Shaving is the technical term for when you find yourself at least eight levels deep in a stack of jobs. For example:

I needed to file my taxes, but first I needed turn on my computer, but then I noticed the pillow on my chair was missing, so I had to find it, and when I found it I realized that the filling was coming out the sides, so I had to fly to Bolivia to shave a yak to stuff the pillow.

All this just so I could finish my taxes.

Describing this phenomenon and giving it a label allows others to think about it and discuss it. We can recognize when we’re engaging in the behavior. We can ask questions. For example, how much of our day do we spend yak shaving? Why? And is yak shaving a form of procrastination or a prerequisite to productivity? Giving this pattern a name allows societies to share common schemas—that is, mental models of the way the world works.

Naming the pattern makes it actionable at the level of society, accessible to teachers and policy makers and therapists and clergymen. Once this pattern has been named we give others the ability to recognize it as a negative habit and take action accordingly. By articulating something that many people thought previously but were unable to express we have created value.

In general, the more encompassing and more actionable the new model, the more value created. In situations like this where we already intuitively accept the model before someone articulates it for us, the meme is prone to spreading quickly and becoming part of common parlance.

There are countless examples of people who have created value by finding and naming patterns, but here are a couple of my favorites:

Seth Godin on eBooks

AJAX: A New Approach to Web Applications

Create a Hypothesis

While naming something merely identifies and labels a pattern, a hypothesis attempts to explain causality.

Anyone who watches Fox News knows that trailer parks are frequently ravaged by tornadoes. I have a friend who’s convinced that trailer parks are actually responsible for creating them.

This is a perfect example of a scientific hypothesis. It is eminently testable, because as soon as we observe a non trailer-park-generated tornadoes then we are able to falsify the hypothesis.

Compare this to a non-scientific hypothesis, such intelligent design. While intelligent design could be proven true under the right circumstances, it can never be falsified. Falsifiability is what separates a scientific hypothesis from hypotheses in general. All hypotheses can be evaluated using tools from philosophy, but only scientific hypotheses can be evaluated using the methods particular to science. Thus, while all types hypotheses are potentially useful, new scientific hypotheses tend to be more valuable since we have more tools we can apply to evaluate their truthfulness and create understanding.

Split One Schema into Two

Sometimes we find ourselves working with schemas that clearly have elements of truth, but that for some reason don’t seem to hold up well empirically. Often times this is because the schema we have in our heads is more broadly defined than the underlying phenomenon.

A good example of this is the phenomenon of prodigies. We know there are some people in society who are exceptionally talented in certain areas, and we call these people prodigies. We then have certain schemas that we apply to these prodigies in our quest for sensemaking.

But even though prodigies clearly exist, our schemas often seem to not hold up so well. For example, studies have shown that child prodigies are often not significantly more successful than the rest of us when they grow up. And similarly, many prodigious adults were completely unremarkable as children. Why is this? How is it possible for such exceptional children not to make anything of themselves, and for such exceptional adults to have been completely average as children?

Malcolm Gladwell observes that the reason for this is because when we describe child prodigies, we are describing people who are gifted at learning. Whereas when we describe adult prodigies, we are actually describing people are gifted at doing.

Because we are applying one set of sensemaking tools to both groups, our schemas tend to not hold up so well even though they are based on an underlying truth. The solution to this is to create one set of schemas for understanding and dealing with child prodigies, and another set of schemas for understanding and dealing with adult prodigies.

There are often areas where we engage in fuzzy thinking and apply one toolset to multiple distinct phenomena. As writers we can create enormous value by identifying distinct phenomena, and giving suggestions for how to think about each one.

Examples:

The Myth of Prodigy and Why it Matters

How to Make Wealth

Combine Two Schemas into One

In the same way that the clarity of our thinking suffers when we treat two distinct phenomena as one, the opposite danger also exists. That is, we can mistakenly hold two completely different sets of schemas about apparently distinct phenomena that are actually the same—the intellectual equivalent of a double standard.

For example, support for slavery was ubiquitous in the antebellum south with many reasoning that the practice was morally acceptable since blacks were not actually people. The slave narratives of the mid-19th century changed this. The authors created value by humanizing blacks, their memoirs personifying the emotions and indignities behind the statistics. Even as late as 1861, the year Incidents in the Life of a Slave Girl was published, black women were not yet considered to be women. Harriet Jacobs argued otherwise, persuading America that black women were just as capable as white women of possessing the feminine virtues – chastity, humility, loyalty, submissiveness, etc.

The insight that made the slave narratives so compelling was that blacks and whites both share the same basic set of characteristics that make us human—thoughts, feelings, hopes, dreams… And so by extension blacks and whites should be treated as equals under the law.

This same false differentiation can also be found at the root of most economic bubbles. Francis Wheen, in his book Top 10 Modern Delusions, writes, “Financial sophisticates in the 21st century smile at the madness of the South Sea Bubble or the absurdity of the Dutch tulip craze. Yet only a few years ago they scrambled and jostled to buy shares in dotcom companies which had no earnings at all nor any prospect of ever turning a profit. To justify this apparent insanity, they maintained that such a revolutionary business as the Internet required a new business model in which balance sheets were irrelevant. In short, they thought they had repealed the laws of financial gravity - until they came crashing down to earth.”

Writers can predictably create insight by showing that a phenomenon that appears to be qualitatively new, like dotcom economics, can actually be accounted for by existing models, albeit perhaps with a few twists.

Think On A Higher Order

One of the principle ways for a writer to create value is by making complex ideas easier to understand. There are several techniques for doing this; one of the most important, described here, is called black boxing. Black boxing is the term we use to describe the process of encapsulating a low-level phenomenon and expressing it at a higher level to hide complexity.

For a good example of thinking on a higher order, think back to your introduction to statistics class from college. Unless you were a math major, you probably didn’t learn the advanced calculus that underlies the common statistical tests. This is fortunate, because the math needed to derive these models is very difficult. But because your only goal in stats 101 is to learn how to use statistical models to make sense of data, the task is much easier. In this example the underlying calculus has been black boxed, and so we have the luxury of thinking on a higher order.

Are there even more abstracted ways of looking at statistics? I would argue yes. For example, what if you don’t need to know how to do statistics? What if you just want to learn how to interpret them? Taking a full-year college course seems overkill. Thus, we can create further value by producing a guide that only covers the higher-level skill of reading and interpreting statistics, not the lower-level skills associated with actually performing statistical tests.

We can see many examples of this in everyday life. Of all the people you see driving cars today, how many do you think could design one? Just about none. Why? Driving has been black boxed.

Similarly, you don’t need to understand cell biology to care for a pet. Nor do you need to understand chemistry to know about cell biology. At least up until a certain point…

Think On A Lower Order

Just as we can create value as writers by cutting out what’s not important and focusing on only the higher order details, we can also create value by looking at things on a lower order. For example, ecologists can look at the environment from the perspective of a microbiologist. Biologists can look at organisms from the perspective of a chemist. Chemists can look at molecules from the perspective of a physicist. And physicists can view the natural world from the perspective of a mathematician.

Why should we want to do this? One reason is the law of leaky abstractions. To quote Joel Spolsky:

Abstractions fail. Sometimes a little, sometimes a lot. There's leakage. Things go wrong. It happens all over the place when you have abstractions […] and the only way to deal with the leaks competently is to learn about how the abstractions work and what they are abstracting.

Going to driving school is a great way to learn how to operate a car, but that only works until the timing belt snaps or a spark plug goes bad. When that happens, no matter how good of a driver you are you can’t get the car to work until you open the hood and fix the problem or take it to someone who can.

Thus, in the same way we can create value by black boxing a problem and teaching something on a higher order, we can also create value by teaching the underlying mechanics (no pun intended).

If we can find no explanations for a phenomenon at one level of the stack, we can always look for meaning at a lower level. For example, a psychiatrist finding no explanations for a given behavior at the level of conscious might try searching for an answer at the level of DNA.

Individuals frequently fall into the trap of looking for solutions to problems only at the level of the stack they feel most comfortable working with. As writers we can create novel insights for our readers by learning to recognize when a problem can be more appropriately addressed by looking for answers on a higher or lower order of abstraction.

X is a subset of Y

Often there are times when we find ourselves working with a smaller model that is really part of a bigger model, even though this isn’t apparent at first. When situations like these occur there is an opportunity to create insight.

For example, the concept of ‘weight’ is a subset of the concept of ‘mass’. That is, we can use weight to accurately compare how much matter two objects have as long as we’re on earth. But as soon as we’re in space we need to use the concept of mass instead. By explaining to students how weight is a very specific instance of mass that works to compare two objects as long as we’re on earth we have created insight.

As another example, consider environmentalism. The goal of an environmentalist is to figure out how to equitably divide finite natural resources among multiple stakeholders who want to use these resources for competing purposes. For example, for any given national forest some might want to use the resource for renewable lumber, others might want to use it to open a ski mountain, and still others might want to use it to graze cattle. The goal of an environmentalist is to ethically divide this resource among the competing stakeholders who all want the resource for their own usage. By realizing that environmentalism (unlike conservationism) is really a subset of ethics, we can show that, among other things, environmental problems are best by using the toolsets that normative ethicists have developed over the millennia.

X and Y are both instances of Z

The reason insights are powerful is because they give readers new tools for creating ideas, tools that can later be used whenever the reader wants. Insights are like trees laden with fruit. Our job as writers is to build the reader a tree and teach them how to bang on it to knock some coconuts down whenever they’re hungry.

In his essay Hackers and Painters, Paul Graham’s central insight is that hacking and painting are similar because they are both acts of creating. While you’d have to ask him to be sure, I’d suspect that once he figured this out the rest of the ideas just sort of fell into place. That is, once we know that hacking falls into the same category as painting, it stands to reason that hacking and painting have many elements in common. For example

• The reputations of both have a large random component introduced by fashion.
• In both hacking and painting, it’s best to start by sketching.
• Both software and paintings are intended for a human audience, so both hackers and painters must have empathy to do good work.

Once Graham sells us on the idea that both hacking and painting have commonalities as acts of making, we can draw upon the rich history of painting to gain a better understanding hacking. When we ask ourselves questions such as what makes a good hacker, or how hackers create value, it then becomes logical to look to painting for possible analogs.

Simplify a more complicated model

Writers can create value by simplifying a more complicated model. As explained in the introduction, a model is the set of relationships that explain any given system. (And a schema is a model that is simple enough for us to reason with intuitively.) However, the same system can often be explained through one or more alternative sets of relationships that are logically equivalent, but vary in their complexity.

For example, the Ptolemy’s geocentric model of the solar system technically did work, but it was very complicated. Simulating the solar system required a mechanical device with hundreds of widgets. This gave Copernicus an opportunity to create enormous value by changing the point of reference from the earth to the sun. The insight that the relationships between the heavenly bodies could be conveyed in a simpler way was the basis of the simplified heliocentric model, one of the most important advances in the history of astronomy.

How does one simplify a model? One way is by choosing better assumptions. As I’ve written previously,

The problem with assumptions is that they're usually correct. For certain people, at a certain times, in certain places.

The danger isn't that you'll sometimes be wrong. The danger is that you'll always be right.

That is, your assumptions about human nature will be true, but less useful than those of your competitors.

I’d posit that most models have one or more core assumptions that the rest of the properties of the system are logically derived from. Once one changes the core assumption, the properties of the entire system change. And if the model becomes simpler as a result while maintaining its accuracy, it’s safe to say that we’ve created value.

X is not mutually exclusive with Y

Writers can always create value by correcting false schemas. One cognitive mistake we see often is the belief that two things are mutually exclusive when in fact they are not. Because false beliefs in this category are so common, writers can consistently write insightful articles along the lines of:

• It is possible to support the troops but be against the war.
• Corporations can be environmentally friendly and still make a profit.
• Atheism is not mutually exclusive with morality

And so on. It turns out that outside of mathematics two things are rarely ever mutually exclusive, at least when we’re talking about abstract concepts. Thus, when we hear people making statements that suggest otherwise this is a good area to look for opportunity to create insight.

===Conclusion===

Once you understand the formula above, it should become trivially easy to create an indefinite number of insights. Especially if you're using my KWL method for generating ideas, and thinking about those ideas hierarchically by using software like FreeMind.

What might be less obvious though is that this idea can make you a much better reader as well. I've always said that good writing changes the way you see something, and great writing changes the way you see everything. Insightful writing is great writing, especially if those insights are counterintuitive. (As a general rule of thumb, I figure if it's intuitive I already understand it, so I only bother to seek out books and articles that are counterintuitive.) If you keep this idea of insightfulness in mind when reading through the newspaper, you can skip over 98% of the stories and still be automagically smarter than just about everyone else.

Comments

[Pete Michaud]

Thank you for this, it's, well, insightful about insight. Listing these as you have gives people a jumping off point to generate insight, whereas before they might have just fumbled intuitively. Nice work.

[Sunil01]

Thank you, very insightful.

[steve]

Putting your advice into practice, one can answer the blog's title and summarize the post w/ "Create novel insights by looking for answers on a higher or lower order of abstraction"

:)