In our ideal imagination someone would always be able to give us an exact game plan to achieve our dreams, full of steps we know exactly how to do.
That kind of recipe would be comforting and make us more confident.
I can’t give you that.
But what I can give you is a mental picture of the five key phases that make a scientific discovery happen. It’s just one of six core components of my scientific discovery framework (you can read about that here).
Equipped with a mental picture, it will be easier to see where you’re losing momentum and look for ways to fire up your progress.
Let’s dive into this “discovery cycle”.
The five evolution phases of scientific discovery, in order, are:
Question. It all starts with having an unanswered question about the world that needs to be answered. Discovery always begins by actively asking an unanswered serious question. Serious questioning is about generating compelling questions and then choosing one to go out and answer.
Ideation. Next you must form an idea about what might be the answer to your question. Productive ideas are ones that we can chip away at through real-world tests and investigations. Ideation is the process of generating productive ideas and narrowing it down to one idea you move forward on.
Articulation. Productive ideas don’t investigate themselves. You’ve got to put it in a format that lets you determine your idea’s ability to correctly answer your serious question. Transforming something from an idea to a real-world process, procedure, gadget, or systematic concept is articulation.
Evaluation. Now that you’ve articulated the idea you think might answer your question you need to put it to the test. Compare your concept against real examples. Observe and probe your data. Run your model and see if it breaks. That’s the heart of evaluation.
Verification. If your idea survives your evaluation (and most of them won’t) then it’s time to open your idea to deep challenges from others. It’s not a scientific discovery until other people have independently confirmed that your idea answers your starting question and that the way you articulated the answer holds up. Personally, I think two separate independent verifications plus your initial investigation are ideal because good things come in threes.
And that’s the discovery cycle in a nutshell.
The scientific discovery cycle is a human learning algorithm for scientific discovery.
You may move back and forth between scientific discovery phases as you make mistakes and learn new things. That’s normal. But in the end, if you discover something new, you will have evolved through all the phases at some point in the process.
Talking with other scientists, I’ve learned that how long you’ve worked with science (not a project) affects which phase is more likely to trip you up.
People new to science tend to get stuck on the question phase.
They don’t know what a good science question looks like. If this fits you, learning more about creativity, filling your knowledge gaps, and becoming more skilled at asking deep questions and mining published papers can help.
People who have some experience working with science, but haven’t spent a whole career on it, often struggle with the articulation phase.
They’ve got ideas, but they don’t know how to put them in productive testable forms. If this sounds like you, reading up on rapid prototyping, building mental models, and techniques like work sprints can help.
People who have made a career out of working in or around science frequently run out of ideas and struggle with ideation.
They may feel like everything’s been done. Or that every idea is bound to fail (or get ignored) anyway. Sometimes they can’t imagine better solutions than the good solutions they already know. If this describes you, then looking into techniques to get around the Einstellung effect or how to think of more “subtractive solutions” might help.
Those are the three main phases where most individuals get stuck and lose momentum in the scientific discovery cycle: the question, ideation, and articulation phases.
Just to be thorough, if the evaluation phase is where you struggle try things like practicing Fermi questions, toy model techniques, or “why not?” counter-thinking. Verification problems are usually about convincing others to engage with your proposed discoveries enough to test your ideas in a public forum. Learning better communication skills can make the difference.
Most scientific discovery projects must pass through all five evolution stages—question, ideation, articulation, evaluation, and verification—to succeed.
Knowing which stage you’re stuck in can point you toward techniques to help you get past an obstacle.
And being clear on where you are in the discovery cycle can tell you what not to do, like getting lost in brainstorming hacks (ideation) when what you need are strategies to create a new metric to measure something (articulation).
Use this discovery cycle framework like a teacher who points out where you are in your project and what needs more work.
Simply put, a mental picture of the scientific discovery cycle is your ultimate personal coach.
Reflection Question
What phase are you in on a discovery project you are working on, or planning, and what’s keeping you from moving to the next stage?
Bernadette K. Cogswell, “A mental picture of the scientific discovery cycle is your ultimate personal coach”, The Insightful Scientist Blog, September 24, 2021.
[Page feature photo: Photo by DeepMind on Unsplash.]
To make a scientific discovery you need a plan not a map
If I told you there was a study that found what actions you take for the next 10 minutes determines whether or not you will make a scientific discovery in your life…how would you spend that time?
How does thinking about the impact of what your doing right now on your discovery potential make you feel? Guilty, curious, confused, even overwhelmed?
Unfortunately, no such study exists. Instead, there are plenty of biographies analyzing how the Einstein’s of the world spent their time.
Don’t get me wrong.
Getting inspired by previous scientific discoveries and the stories behind them is wonderful motivation.
But it doesn’t tell you how to spend the next 10 minutes of your life to make your own discoveries. For that you need an action plan.
So let me share the scientific discovery framework that I’ve developed, which will give you a plan. It’s helped me see how discovery gets done and it will help you too.
There are lots of parts to scientific discovery, but they all fit together in a logical whole.
In a series of posts, I’ll explain my framework for connecting those parts and how you can prioritize your efforts to get moving on making a discovery.
This first post lays out the big picture of scientific discovery. Get ready for an information download! Stick with it. Don’t worry if it feels like a lot. Shorter follow-up posts will guide you. Jump in and out of the series anywhere – the posts are all standalone. You can take it all in as you have time.
Let’s get to it.
I’ve identified six core areas that power scientific discovery:
1. Discovery repertoire. The personal portfolio of techniques that you use to get science done is your scientist’s repertoire. There are four sections to your internal portfolio: how you think about your science (mindset), what tasks you know how to complete to get science done (activities), the recipes you have for combining outcomes with actions (skills), and what you know (knowledge). When you have a solid plan, but still don’t make progress on your science it means you need to strengthen a weak part of your repertoire.
2. Discovery capacities. Learning new things in science and technology is driven by four human capacities: innovation, invention, insight, and scientific discovery. Capacities get different results because they are driven by different motivations. Innovation motivates us to improve the way something works. Invention motivates us to build devices that will do something useful. Insight motivates us to change how we see the world. Scientific discovery motivates us to explain how the world works. Insight and scientific discovery are core capacities that build on each other.
3. Discovery vital qualities. The difference between a scientific discovery and regular scientific research is that a new discovery-level scientific finding will have at least one of three vital qualities: It will shift our perspective on the world (be radical), it will link knowledge to make a broader range of predictions about the world (be universal), and/or it will be new knowledge (be novel). Your work should have one of these qualities as an objective to aim for discovery-level science.
4. Discovery impact classes. Scientific discovery intuitively feels more high impact than regular science. That impact lies on a continuum from low to high, determined by how many vital qualities a discovery captures. Minor class discoveries possess only one of the three vital qualities. Major class discoveries possess at least two and legacy class discoveries must have all three. Science spans from regular research to legacy class discoveries on an incremental spectrum defined by these qualities. So, start small and build up to the big discoveries.
5. Discovery learning categories. Scientific discovery learns something new about the world. What you learn falls into three categories: something about an unknown object (object-type), something about the properties of an object (attribute-type), or something about how and why the world works the way it does (mechanism-type). Some categories are easier to make discoveries in because the learning curve is smaller.
6. Discovery evolution phases. Most scientific discoveries evolve through five phases, which I call the discovery cycle. First, you ask an unanswered question (question). Then you form ideas for an answer (ideation). You make those ideas into tests in the real world (articulation). You run the tests and evaluate the results (evaluation). And if the results repeatedly prove true then they become a scientific discovery (verification). Troubleshooting your scientific discovery progress is easier if you know what phase you are in because unique problems trip up scientists at each phase.
The framework I’ve developed lets you craft a scientific discovery action plan, troubleshoot your progress, and connect specific activities and techniques with the results you want to achieve.
The simplest starting point? Aim for a minor class, attribute-type discovery that is universal. That represents a baby step from current science to something new. And if you hit an obstacle check your insight in a systematic way and seek out techniques to boost you from one phase of scientific discovery to the next.
No matter where you start, be inspired by the scientific discovery stories of others, but don’t stay stuck in them. Discovery isn’t a sightseeing tour through known territory. It’s a push toward unknown territory.
Simply put, to make a scientific discovery you need a plan for how to tackle the unknown, not a map of the known.
Take Action
Once you’ve got a framework and a plan, spend the next 10 minutes taking action. You’ll be 10 minutes closer to making your next discovery.
Have you have ever been on the receiving end of a colleague, boss, or even stranger sitting next to you on a plane (ah, the good old days before coronavirus), asking you questions like,
“But what impact will your work have?”
“Can you study something more interesting/important?”, or even,
“Who cares?”
If so, then you have come up against a problem all researchers, scientists, and citizen scientists face: How to try and do the best possible, most high impact, most important science you are capable of as often as possible.
What you are aiming for is a scientific discovery. And what well-meaning acquaintances and strangers are asking for is the same thing.
But how do you do that? Trying to aim that high can seem overwhelming.
The Problem: Many existing definitions of scientific discovery are good for textbooks, but not for project planning, follow-through, or troubleshooting.
My Solution: Define “scientific discovery” so that you can achieve it with training and algorithms, and perform quantitative studies to probe it.
The purpose of this post is to lay out all the key components that will help us train ourselves to become better discoverers. You can see this framework drawn in my concept sketchnote below.
My hand drawn sketchnote of the “architecture of scientific discovery” showing all the important parts of the process. (Artist: Bernadette K. Cogswell. For reprint permission, contact me.)
There is a lot of ground to cover, so in this post I am going to give short descriptions of everything in the big picture.
In the following 22 (!) posts in the series, I will drill down into each of the 18 parts in more detail, with examples taken from science history.
Now let’s jump in…
Group 1:
Discovery Capacities
Scientific discovery is just one of four human capacities for discovery in science and technology.
Let’s zoom in on the discovery architecture picture I’ve drawn above.
A close-up of my sketchnote showing the four discovery capacities at the core of the discovery architecture. (Artist: Bernadette K. Cogswell. For reprint permission, contact me.)
As human beings, we have the capacity to discover new things in science and technology. These discovery capacities form the main part of the structure and they house all our abilities and knowledge about science and technology.
These discovery capacities fall into four types—insight, invention, innovation, and scientific discovery. The main differences between types are their results and our reasons and motivations for pursuing a discovery.
So let me give you a definition of insight, a definition of invention, a definition of innovation, and a definition of scientific discovery.
Insight
Definition of insight:
“Insight” is refining the accuracy of your perspective of the real world.
Getting a more accurate perspective, or “achieving insight,” is accomplished in one of three ways. You can add something new to your perspective that you were not aware of before. You can correct something that you misperceived. Or you can clarify something that you only vaguely understood.
Motivation: “I want to…change how I see the world.”
Invention
Definition of invention:
”Invention” is building a machine or process that creates a previously unobtainable result.
Creating a previously impossible result, or “inventing something new”, is brought about by focusing on three aspects of what you build. You want to build something that has not been built that way before. You want to build something that does what it was built to do. And you want what you built to create something that a machine or process like it has not created before.
Motivation: “I want to…build a device that will do something useful.”
Innovation
Definition of innovation:
”Innovation” is improving the functionality of a process or device.
Refining how things work, or “innovating”, is really about making things work better more easily. You can make something function more efficiently. You can make something run faster or with fewer resources. You can make something more likely to produce what it was designed to produce. And you can make something produce a higher quality version of what it was designed to produce.
Motivation: “I want to…improve the way things work.”
Scientific Discovery
Definition of scientific discovery:
”Scientific Discovery” is finding the evidence, interaction, and causes of things that exist in the natural world.
Learning something new about nature, or “making a scientific discovery,” relies on three things. You must acquire knowledge. You must demonstrate that the phenomena exist using evidence and statistical or logical analysis. And the knowledge you acquire must include one or more element of the radical, the universal, or the novel (those are defined in the next section).
Motivation: “I want to…explain how the world works.”
In this framework “applied science” can be defined as a combination of mastering the capacities for invention and innovation, while “fundamental science” or “basic science” can be defined as a combination of mastering the capacities for insight and scientific discovery. By “science” I mean the physical sciences (e.g., astronomy, biology, chemistry, computer science, data science, engineering, geology, medicine, paleontology, physics, etc.), the social sciences (e.g., anthropology, economics, political science, psychology, sociology, etc.), and mathematics.
(Also, you might wonder why insight is listed as its own discovery capacity since it is integral to the other three discovery capacities, invention, innovation, and scientific discovery. This is true. However, it’s more useful to put it on an equal footing with the other three capacities. It’s easier to develop training protocols, algorithms, and quantitative metrics to explore discovery methods, all goals for developing this framework.)
Group 2:
Scientific Discovery Vital Qualities
There are three qualities any scientific study or research must have, or it can’t be called a “scientific discovery”.
The three scientific discovery qualities are the foundation on which we can build any kind of discovery.
A close-up of my sketchnote showing the three qualities that form the foundation of defining something as a scientific discovery. (Artist: Bernadette K. Cogswell. For reprint permission, contact me.)
They are integral to recognizing discovery and generating discoveries.
These three essential qualities form the basis of what makes scientific discovery different from everyday scientific investigation and scientific research.
How is scientific discovery different from scientific investigation? Scientific discovery has a higher and more long-lasting impact on the evolution of science. So let’s define the vital qualities that embody that impact and enduring nature.
Radical
Definition of the radical quality of scientific discovery:
The “radical” quality of scientific discovery means that the new knowledge gained as a result of the discovery represents a meaningful shift in perspective from the previous state of knowledge.
Role in scientific impact and longevity: Scientific discovery is radical—it changes the perspective of science in one of three ways. Something can be added to what we know. Something can be rejected from what we thought we knew. Or something that we know can be changed. By impacting our scientific perspective, the radicality of scientific discovery opens up new avenues of research and creates or ends long-standing practices and beliefs.
Universal
Definition of the universal quality of scientific discovery:
The “universal” quality of scientific discovery means that the knowledge acquired as a result of discovery is valid and reliable and that the knowledge gained has predictive or descriptive power in a range of physical situations.
Role in scientific impact and longevity: Scientific discoveries have a broad impact because the new knowledge they bring has a range of application. The universal nature of a scientific discovery lies on a spectrum from “proximal” to “distal”. “Proximal” means that the new knowledge can be applied to a broad range of areas with few changes to its verified form. “Distal” means that the new knowledge discovered can only be applied to areas and phenomena closely or directly related to the area in which the discovery was made, or that to apply it to other areas requires a lot of translation. The scientific discoveries with the most direct universal appeal have the longest legacies.
Novel
Definition of the novel quality of scientific discovery:
The “novel” quality of scientific discovery means that the knowledge obtained through the discovery has not been previously shown to exist, in a reproducible way, by observation or experimentation.
Role in scientific impact and longevity: Scientific discoveries electrify areas of science because they bring something new to the table. And when those new elements are verified, they shape future research activities and ways of thinking.
The impact effects of scientific discoveries and their longevity, as forces that shape research practice, effort, and interest, are embodied in the three vital qualities at the foundation of the scientific discovery architecture. They give scientific discovery it’s je ne sais quoi factor that inspires the layperson and the scientist alike.
Group 3:
Scientific Discovery Impact Classes
The impact and significance of all scientific discoveries can be grouped into three classes.
The purpose of discovery is progress in some area (understanding, outcomes, effectiveness, and knowledge as we saw from the section on the four discovery capacities above).
The discovery classes represent broad categories that help identify the level of impact (or progress) that our discoveries are capable of achieving or fostering.
These classes, therefore, overarch the specific categories of scientific discoveries.
A close-up of my sketchnote showing the three discovery impact classes that overach all scientific discovery and describe a discovery’s effect and longevity. (Artist: Bernadette K. Cogswell. For reprint permission, contact me.)
In particular, the discovery classes encompass three different levels of impact, from wide-ranging to narrow, as described in their definitions below, and as represented by the fact that the three domes of discovery impact classes are nested in my sketchnote diagram.
Minor
Definition of a minor impact class scientific discovery:
A “minor” class discovery meets the criteria for any one of the three vital qualities of scientific discovery—radical, universal, or novel.
Minor scientific discoveries are either radical, universal, or novel, but not all three at once. Therefore, they have an impact beyond ordinary scientific investigation, but their impact is limited.
Major
Definition of a major impact class scientific discovery:
A “major” class discovery meets any two of the three criteria for the vital qualities of scientific discovery—radical, universal, or novel.
Major scientific discoveries are either radical and universal, or radical and novel, or universal and novel, etc. They have two of the three vital qualities, but are missing the third one. As a result, their impact tends to be more wide-spread than minor class discoveries, but not as high impact as they could be if they embodied all three qualities.
Legacy
Definition of a legacy impact class scientific discovery:
A “legacy” class discovery meets all three criteria for the vital qualities of scientific discovery—radical, universal, and novel.
Legacy class discoveries are the full package—radical, universal, and novel. The impact of legacy class scientific discoveries is wide-ranging and long-lasting. These are the hardest scientific discoveries to achieve, but the ones with the greatest value.
Group 4:
Scientific Discovery Learning Categories
The types of scientific discoveries you could make can be grouped into three categories.
The learning categories are specific to only one of the discovery capacities, scientific discovery (they are not intended to be applied, by analogy, to insight, invention, or innovation).
These categories of scientific discovery divide the field of knowledge obtained through scientific discovery into three areas. These areas are determined by the kind of information you hope to gain, or your learning objective.
Object
Definition of the object type scientific discovery:
An “object” scientific discovery is acquiring knowledge about the existence of a new object in nature.
Learning Goal: Answers the question, “Does something exist?”
Attribute
Definition of the attribute type scientific discovery:
An “attribute” scientific discovery is acquiring knowledge about the characteristics, properties, and/or traits of an object or process in nature.
Learning Goal: Answers the question, “What is something like?”
Mechanism
Definition of the mechanism type scientific discovery:
A “mechanism” scientific discovery is acquiring knowledge about the causes, connections, interactions, and/or sequences of objects and attributes in nature.
Learning Goal: Answers the question, “How does something work?” and/or “Why does something happen?”
A close-up of my sketchnote showing the different types of discoveries you can make as central to the architecture of discovery. (Artist: Bernadette K. Cogswell. For reprint permission, contact me.)
Let’s look at this part of the discovery architecture more closely, as shown above in my sketchnote drawing.
The scientific discovery learning goals are shown under the dome of the scientific discovery classes because they can fall under (i.e., be impacted by or represented in) all the classes of scientific discoveries.
Another way to think of it is that the scientific discovery classes are umbrella terms that cover all the categories or types of scientific discoveries you could make.
(Again, this is just the overview, in future posts I will talk about each of these in more detail and it will begin to make more sense as you see examples and further discussion.)
Onward to the last group I want to cover in this post…
Group 5:
Scientific Discovery Cycle
(Evolution Phases)
Most scientific discoveries must pass through five phases.
All of the above groups—the discovery capacities, the scientific discovery classes, the scientific discovery categories, and the scientific discovery qualities—form the main architecture of scientific discovery.
You can think of these like a very old and sturdy building, where every brick and design element of the structure is built up out of our application of the discovery capacities, classes, categories, and qualities and the knowledge, abilities, devices, and processes that we have created as a result.
There is one more important element in the overall architecture, and that is represented by the sun shown in the upper right hand corner of my sketchnote illustration.
A close-up of my sketchnote showing how the scientific discovery cycle illuminates all other aspects of discovery. (Artist: Bernadette K. Cogswell. For reprint permission, contact me.)
The sun represents the process that drives scientific discovery, or the “scientific discovery cycle”, which shines a light on all the other elements of the architecture so that we can become aware of them and manipulate them in the course of running our projects as scientists.
Below I summarize each of the five evolutionary stages of the scientific discovery cycle.
Question
Definition of the question phase of the scientific discovery process:
The “question phase of scientific discovery” is the stage in the process when the question to be answered, or problem to be solved, is explicitly defined.
Purpose of Stage: Define what you want to find, create, or explain.
Ideation
Definition of the ideation phase of the scientific discovery process:
The “ideation phase of scientific discovery” is the stage in the process when a possible solution or solutions is conceived of to answer the discovery question or solve the discovery problem.
Purpose of Stage: Come up with a guess for how you will find it, create it, or explain it.
Articulation
Definition of the articulation phase of the scientific discovery process:
The “articulation phase of scientific discovery” is the stage in the process when at least one proposed solution is put into a form that can be tested in the real world.
Purpose of Stage: Write an equation or description, or build or code a prototype, embodying your answer.
Evaluation
Definition of the evaluation phase of the scientific discovery process:
The “evaluation phase of scientific discovery” is the stage in the process when the testable solution is probed and its success in answering the discovery question, or solving the discovery problem, is assessed.
Purpose of Stage: Test your equation, description, code, or prototype to see if it answers your question.
Verification
Definition of the verification phase of the scientific discovery process:
The “verification phase of scientific discovery” is the stage in the process when the best available solution to answer the discovery question, or solve the discovery problem, is confirmed to be accurate and reliable by multiple independent analyses.
Purpose of Stage: Subject your “discovery” to public scrutiny and see if it holds up to testing.
Note that, the way I have defined it, the scientific discovery cycle is different from the scientific method.
The scientific method focuses on how to obtain valid and reliable information about the world. But it is not concerned with the impact of that knowledge.
The scientific discovery cycle (or scientific discovery process) is concerned with the impact of the knowledge obtained and its purpose is to obtain knowledge of a certain minimum impact level (namely, knowledge that is either radical, novel, or universal and, therefore, at least meets the standard for a minor class discovery).
The scientific method would be used to obtain relevant insight at various phases within the scientific discovery cycle (such as during articulation, evaluation, and verification).
Therefore, the scientific method is one set of activities in the scientist’s repertoire, which they can use to help them complete the evolution stages in the scientific discovery cycle. They are connected, but distinct.
Summary
Phew!
That brings us to the end of my overview of the architecture of discovery that I have built as a way to develop better discovery training protocols and quantitative methods to identify patterns and correlations in discovery processes.
If you are someone who loves algorithms or self-improvement, then this architecture and way of conceptualizing discovery and how to achieve it is for you.
So much good stuff to talk about!
I’m really looking forward to writing the rest of the posts in this series. Having these new words and concepts in my discovery arsenal has already helped me organize and conduct my research projects in a new way. And it makes extracting nuggets of insight from examples of scientific discovery much more productive.
To wrap up this post, let me give you the very, very short, bullet-list version of “the architecture of scientific discovery” that I covered in this post:
Scientific discovery is just one of four human discovery capacities—including insight, invention, and innovation—in science and technology.
There are three vital qualities—radical, universal, and novel—any scientific study must have, or it should not be called a “scientific discovery.”
The significance of scientific discoveries can be grouped into three impact classes—minor, major, and legacy—which range from low to high impact.
The types of scientific discoveries you could make can be grouped into three learning categories—object, attribute, or mechanism—based on the information gained.
Most scientific discoveries must pass through five evolution stages—question, ideation, articulation, evaluation, and verification—to succeed.
With all these categories and types, it’s easy to fall into the trap of seeing things as black and white, or just buckets to assign things too.
But the definitions and concepts I’ve come up with work well because they allow us to see scientific discovery as a continuum of insight, from narrow to broad, from low impact to high impact, from fundamental to applied.
Every scientific investigation, every research study, every time you use the scientific method, you are placing yourself somewhere on that continuum of insight. Discovery is always just a little way further along that scale. The question you should be asking yourself is not “Am I making, or trying to make, a discovery?” it’s “Where on the discovery spectrum am I working right now?”
So next time someone challenges you on your impact, dream brave and think (or tell them), “I’m working on some (minor/major) research questions right now, but my dream is to leave a legacy discovery for future generations to build on.”
And hopefully with this discovery architecture in your mental repertoire, you will someday be able to do just that.
Bernadette K. Cogswell, “The Architecture of Scientific Discovery: Overview of the Process – On how to define and categorize all aspects of scientific discovery”, The Insightful Scientist Blog, June 13, 2020, https://insightfulscientist.com/blog/2020/architecture-of-scientific-discovery-overview.
[Page feature photo: A quiet and quirky cabin sits among the mountains. Photo by Torbjorn Sandbakk on Unsplash.]
Have you ever watched a movie or TV show, or read a book, where at the end of the story the main character saves the day by doing something unbelievable? By unbelievable I mean that they do something completely out of character. This kind of ending can leave a bad taste in your mouth, as if the writers didn’t do their job in making us believe the character had changed enough to become a person who could behave that way by the end.
When I was working on my degree in creative writing, there was a phrase that summed up the problem:
“Once is an accident, twice is a coincidence, three times is a pattern.”
The idea is that people open up to the possibility that something is plausible by seeing relevant elements happen enough times that we decide a pattern is believable. It’s kind of a “conception through perception” game. If a character behaves in ways that build up to the ending then we consider the ending reasonable. But if we don’t see enough evidence then we find it hard to believe and the ending will seem like a cheap magic trick and a waste of our time (and money).
In my own experience, I’ve found it pays to be aware that this little rule of three affects not only how writers convince us of story endings, but also how we convince ourselves that some of our ideas merit pursuit.
That’s because deciding if a research idea is worth investigating is really about deciding if there’s enough of a pattern there to plausibly lead to an interesting ending…and hopefully that ending will be a scientific discovery.
So let’s talk about how to translate this magic of the number three from creative writing into research in a way that will help us decide if a research idea should move to the top of our to-do list or get shuffled to the back burner.
THREE…Essential Elements of an Idea
Most of our time as scientists is spent in the “articulation” and “evaluation” phases of scientific discovery. Meaning, we worry a lot about defining our ideas and assessing if they are useful, correct, and/or meaningful.
In starting on a research topic, it can be hard to formulate a clear awareness of what we mean by new ideas. And once we’ve jotted something down on paper, or typed it up, it can be difficult to decide if the idea seems worth focusing on. The tendency is to have conversations in your head about it and then put it on the mental back burner because of the feelings of “riskiness” that working on discovery-level science can bring up.
If you’re stuck with a sense that you “have an idea”, but that you couldn’t yet share that idea with someone in a three-minute sound bite then here’s something to try. You can write this down, type it up, do a voice memo, or some combo of all three. Whatever works for you. I’ll use pen-and-paper writing as my example since that’s how I prefer to work:
Write down the idea you are trying to get clear in your head as a one word prompt. Stick to one word, no phrases or sentences.
Spend a few minutes (no more than 15) just thinking about the idea behind your one-word prompt. Now, write down three more essential words that capture the heart of the idea. These new words should sum up the essential elements, features, behaviors, or requirements of your prompt word. Again stick to just three words, no phrases or sentences here either. But you must write down at least three words, no less.
Now create a list numbered one to three. For each number write down what you mean by each of the essential words. You can write in phrases or sentences here. But keep it to no more than 1-2 sentences per numbered item. Start each numbered item with the prompt “By <essential word> I mean…” You can spend up to one whole day to complete this list. But finish this whole exercise (steps 1-3) in 24 hours or less.
This little exercise can help you generate a clearer picture of your idea by forcing you to pick and choose what matters most to you and define it.
That’s where you as a scientist bring your best asset, your personal diversity, to the playing field. Don’t use other people’s words or definitions for this exercise. Set your phone aside. Don’t use Google. Don’t use textbooks or published papers. Just use what you’ve already got inside your head.
I cap the time you spend on it at 24 hours to keep you from overthinking it. The goal here is to make a rapid decision—“research this” or “shelve this.” You want to build momentum, not stall out in the graveyard of analysis paralysis.
The reason I say identify three essential words goes back to the accident-coincidence-pattern idea. Three words is a good sweet spot to help make abstract ideas more concrete. Think of it like triangulating a signal: getting three points of reference lets you narrow down and enclose your idea in a more well-defined area.
THREE…Sources of Information
At this point it’s helpful to get out of your own head and take a look at what other people are saying about your idea. In theory, you probably started out by reading the work of others or listening to someone speak, which helped spark the idea you are working through now. So you may already have some good sources to look over again.
The goal is to get three sources (by “source” I mean a written or spoken piece of work) you can compare against the idea you formulated in the previous exercise. You want to read them (or re-read them) and compare how you formulated your idea to how the author(s) or speaker(s) formulated it.
The most important thing is to find good quality sources to help evaluate your idea.
If you don’t know how to find or consider sources for their quality, here are some tips:
Look for good quality information, not good quality authors. That means you want sources that are complete, accurate and have minimal bias (or consciously acknowledged bias). Authors, writers, scientists, journalists, etc. are only human. No one produces good quality work all the time. Evaluate each information source individually; don’t just assume that famous names, or even people you know who usually do good work, put in that effort this time. We all have off days.
Value sources that speak most directly to the idea you are working through with real data and more references to explore. Be open to traditional (peer-reviewed published articles, monographs, academic books, etc.) and nontraditional (blogs, popular science outlets, podcasts, etc.) sources. Evaluate each source individually. I usually rank items with real data (even if it’s just a thoroughly explained personal example) and that reference other good quality sources I can freely access (no paywalls) more highly than ones that are tangential to my topic or only talk in general terms.
Try to get a good variety in your three sources. Make sure they are all by different authors or speakers. Try to get different perspectives in each one, i.e., the authors are from different fields, different career stages, different job sectors, are different genders, ethnicities, ages, nationalities, etc. The sources don’t need to tick all these boxes, but do the best you can. Try to ensure that you don’t rely too heavily on just one voice in the debate, which could cause you to repeat what’s already been done instead of trying something new.
Again, don’t over think this. I’d limit the time you spend on this to one week. Do the best you can with the information you have access to.
Once you’ve got these sources, spend some time reading them and noting the differences between how you articulated the idea and how they articulated the idea. You’re looking for similarities, differences, things they mention that you left out completely, and things you mention that they ignore (this last one is where scientific discovery lives).
THREE…Mental Examples
Now it’s time to move out of the “rainbows and butterflies” world and into the “bricks and mortar” world.
What I mean by this is that in the beginning we tend to be pretty excited, enthusiastic, and confident about our own ideas when they’ve only existed in our head. This is the “rainbows and butterflies” world. These feelings are a good way to generate momentum to get started on a project and they encourage “thinking.” But they’re not very helpful to encourage “doing.” Doing requires having a clear idea of what the next action is. That’s the “bricks and mortar” part. Rainbows and butterflies are inspiring, they captivate and focus our mental attention, but they are hard to hold in your two hands. With bricks and mortar it’s much easier to grasp how to start building something.
Applying your idea to examples is a way to get started on the bricks and mortar “doing” and to see if you’ve missed out on any major facets of defining your idea so that it’s open to scientific investigation. I like my three examples to cover three types (three is still the magic number!):
An example that fits your idea really well (an “exemplar”).
An example that doesn’t fit your idea at all (a “counter-example”).
An example where it’s hard to tell if it fits your idea or not (a “neutral example”).
Covering these three bases will encourage you to be deliberate and thoughtful and to assess your idea for its strengths (illustrated by the exemplar) its weaknesses (illustrated by the counter-example) and its limits and areas for improvement (illustrated by the neutral example).
You want to develop a more realistic understanding of what your idea is (you could tell someone about the exemplar in conversation as a way to help describe your idea) and to acknowledge its limits and shortcomings.
If the limits make the idea not useful, or the shortcomings show up for examples that are what you were trying to explain, then I find it’s best to go back and trying redefining my idea. Try changing up the essential words or changing their definitions until you have an idea that holds up better to this simple evaluation method.
THREE…Drafts
Now you’re ready to put your idea into a working definition that you can make a decision on.
I know, I know: all of that work just to get to what most people consider the starting point for research!
That’s why the tagline for The Insightful Scientist is “Discovery awaits the mind that pursues it.” Mental preparation and technique are a huge part of being a scientist and trying to make scientific discoveries. Learning processes and strategies to wield our mindset more effectively is one of the best ways to run a winning race in pursuit of discovery.
The point of all this mental preparation is to give yourself a clear picture of where your idea stands and the challenges and advantages to trying to investigate it. That is what gives you the ability to decide if it should move to the top of your to-do list or move to your mental back burner.
This last step ensures that you have something concrete to either (1) return to later if the idea doesn’t make the to-do list for now, or (2) act on right away if it does make your to-do list.
So set aside a day or two for this and type or write (no voice memos here) a formulation of your idea that is in complete sentences and includes both your prompt words, the essential words you identified, and their definitions. Keep the entire working definition to a minimum of one sentence and a maximum of 5 sentences (i.e., a paragraph). If you prefer word count goals, try for something in the 100 to 250 word range.
Write three drafts of your working definition:
First write a “rough draft” that just gets all the basic elements of your working definition (one word prompt, three essential words, definitions of those essential words) in there in grammatically correct language with proper spelling.
Then write a “second draft” that most likely changes some core features of the definition, like the essential words or their meanings, or adds on to clarify exactly what you mean.
Then write a “third draft” that tries to cut down on unnecessary words, overly complicated phrases, or overly technical words. Just include the essential in your definition, not the useful or the interesting.
Once you’ve got your third draft of your working definition it’s up to you to chart your own course and make a decision: are you going to research this idea or not? With all that mental preparation you’re in a much better spot to make a more thoughtful decision and you could explain that decision to someone else. Game. Set. Match.
Good Things Come in Threes
So that’s how I translated the idea of “Once is an accident, twice a coincidence, and three times a pattern” into a way of gathering information to decide what scientific ideas to pursue right now. In fact, I just used it last week to finally decide that one of the many working definitions of “scientific discovery” I have come up with over the last 8 months is worth putting into a paper to submit to the open access philosophy journal Ergo by later this year.
It’s important to point out that this general rule of three is not (necessarily) sufficient for a scientific investigation to be rigorous. That depends on the method being used. This rule of three is more about how to decide if fledgling ideas or flashes of insight from brainstorms are worthy of becoming methodical scientific studies. But as a general mental rule, especially if you’re feeling trepidatious, giving yourself a set of three (sources, examples, key words, ideas, sounding boards, etc.) can be an effective way to help you decide what makes the cut.
There’s another saying that also relies on the number three: “Good things come in threes.” In science accidents spark awareness, coincidences spark curiosity, and patterns spark discoveries.
So maybe there is power and magic to the number three.
Of course there’s only one way to find out if my anecdotal use of the number three will lead you to your own epic story of discovery: take a chance, roll the dice, and jump in with an open mind to try it out.
How to cite this post in a reference list:
Bernadette K. Cogswell, “Good Things Come in Threes”, The Insightful Scientist Blog, July 26,2019, https://insightfulscientist.com/blog/2019/good-things-come-in-threes.
