How do you maintain original thinking and avoid traditional patterns of thought when learning about a new field?

These are initial thoughts and a request for comments on paradigm blindness, to be compiled into a more thorough article on the concept in the future.

One thing that I've been concerned about lately is how to maintain original thinking when diving into a new field. I think that we are subject to the conscious or unconscious effects of paradigm blindness when we learn about a new field. That is, once we learn how it's done traditionally, it's hard for us to come up with new original ideas. This could mean that we’re less able to come up with great ideas that really change a field, instead of incremental ideas that contribute a relatively smaller amount to it.

The reasoning is that, when we learn something new, we are also learning the frameworks and patterns of thought that accompany that new information. An example is the dual process theory of cognition popularized by Daniel Kahneman and Amos Tversky of System 1 (fast, instinctive thinking) vs. System 2 (slow, deliberative thinking)—when you learn this concept, you’re also learning to think within this framework.

But I think that the more we buy in to a framework of thinking as the ‘right way to think,’ the harder it is to think in novel, original ways. In other words, the more you think in traditional ways, the more you come up with more derivative, less original ideas. Not only that, but I think this happens subconsciously—since these are patterns of thought, we might not even be aware that we are using them, or ignoring that there might be better patterns out there (via confirmation bias).

This phenomenon is known by a number of names, each of which sheds light on a new dimension of it. There’s paradigm blindness (and the related concept einstellung), where thinking in a particular paradigm (framework) makes you unaware and potentially reject other ways to think about something. There’s the curse of knowledge, where people are unable to think outside of what they know to see how others see things—a common problem with teaching. Finally, there’s the beginner’s mind, which is sort of the reverse of the past concepts: the idea that “In the beginner's mind there are many possibilities, in the expert's mind there are few”¹—in essence, experts have learned that some things are possible, and other things not; beginners, who are oblivious to this due to a lack of experience and who lack preconceptions, see more opportunities and (I think) sometimes challenge existing thinking.

In general, the idea is this: people who have learned in a traditional way, by learning everything there is to know about a field, may think in traditional paradigms as well. They may have tunnel vision, where they can’t think in new ways. The ideas they come up with tend to be evolutionary, not revolutionary, since they’re based in the patterns of thinking and concepts of the traditional paradigms. The questions I have are: Is this real? What do we do about that? Here, I’ll be talking about some of the evidence behind paradigm blindness, and potential defenses against it, and I’d like to hear from you if you’ve experienced this or have any ideas with regard to defending against paradigm blindness.

Some evidence for paradigm blindness

(Not interested in this? Jump to the next section, defending against paradigm blindness.)

There is little or no empirical research on paradigm blindness, but there is some theoretical support for the idea. The most apparent manifestation of paradigm blindness is its related concept of the curse of knowledge in teaching. As anyone who has tried to teach something to someone has found, it’s pretty difficult to express complex concepts in a way that’s understandable for beginners. The way that you think of a complex concept is probably more high-level than a beginner might start out thinking about it, and it’s hard to try to think outside of your already-solidified frameworks to express some concept in a way that relates easily to a beginner.

More relevant for paradigm blindness is Thomas Kuhn’s application of the concept in the philosophy of science. In The Structure of Scientific Revolutions, Kuhn talks about how scientific revolutions are caused by a shift in paradigms of thinking. But in the meantime:

In 1996, Kuhn observed that as paradigms shape the ways in which scientists (or managers) are trained, they will then find it difficult to challenge those paradigms, because they “are committed to the same rules and standards for scientific practice. That commitment and the apparent consensus it produces are prerequisites for normal science, i.e., for the genesis and continuation of a particular research tradition.”²

Not only is it difficult to challenge those paradigms because it’s unpopular to, but I think it’s also hard to challenge them since you’ve solidified them as “what you should do,” and subconsciously use those patterns of thought.

A concrete case

Let’s say that you want to enter the field of artificial intelligence. This is an interesting case, since it isn’t as codified as fields like physics or biology (which have foundations that are almost certainly true), but where there is not yet any consensus on the right foundations and where new ideas can still be developed. Would it be a good idea to avoid the traditional frameworks for a while and try to come up with some original ideas, before then diving into the traditional canon of knowledge? When diving into the traditional knowledge, should you approach it in some way—with skepticism, perhaps—to maintain originality? Will this increase the chance you might come up with great ideas?

One related question would be: can people who are not experts in a field, who can think outside the box because they’ve never really been indoctrinated by the traditional set of knowledge, come up with great ideas? Can they come up with what Google[x] director Astro Teller calls 10x ideas, not 10% ideas? Or, are those who really know a field deeply more likely to come up with great ideas, like this:

Defending against paradigm blindness

Writing initial thoughts before diving into a field. I use this approach most of the time, where I write about my own thoughts about a field, which I think are somewhat original, before diving into a field. Most of these ideas suck, but some of them are kind of decent. But this is hard to do, since your initial thoughts on something without knowing anything about it almost always 1) suck or 2) are common ideas that have been thought of before. But sometimes, there’s something decent in here. Maybe as I continue to learn about a field, I can continue to update those ideas and see them in new contexts, which leads to another strategy:

Incorporating reflection while learning traditional knowledge. I've also thought that you might have your most original thinking when you know just enough to know how to think about a field, but not enough to really have solidified your positions on things. If that's true, the originality might look like the shape of a curve, maybe a bell curve. You start out not being original, learn a bit of the foundation and then you're pretty original, and then you learn too much and solidify too much and then you're not original at all. Maybe in the middle of learning about a field, when you know enough information that it points you into interesting directions with your thinking, you can come up with ideas that are reasonably original and also reasonably interesting.

Seeking out different viewpoints. For fields that are reasonably developed, different viewpoints might exist. Learning about different approaches to artificial intelligence, for example, allows you to see the differences between them. You might not know where the boundary conditions of, say, the model of System 1 vs. System 2 thinking until you learn about another model which seems to have some explanatory power that System 1 vs. System 2 doesn’t cover. From there, you might be able to find space in between those models that can lead to an interesting idea – though one that is likely in reaction to (and thus derivative of) those other ideas.

A habit of skepticism. Being able to continually and habitually ask questions about what you’re learning. What are the shortcomings of a concept? What other concepts might similarly explain this? If this concept turned out to be wrong, what would be the weakest link in the concept that could have caused its invalidity? Cultivating a habit of skepticism might help during the process of learning.

Some other ideas: teaching the field to someone else, and learning a field in a nonstandard way.

These seem good, but I’m not fully satisfied with this yet. What are your ideas on how to maintain originality and ‘outside-the-box thinking’ when entering a new field? How do you avoid thinking in traditional patterns of thought and be subject to paradigm blindness?

¹: Suzuki, Shunryu. Zen Mind, Beginner's Mind. Also Shoshin.

²: Fischbacher-Smith, Dennis. Paradigm Blindness, in Encyclopedia of Crisis Management (Paywall).