1 of 59

A Course on the Creative Scientific Process

Adapted from: Itai Yanai & Martin Lercher

In science we like results - but in this course we would like to talk about the process. Everyone talks about the scientific method. But they never talk about the creative process. We are interested in that.
If you think about it: we have a scientific method for scientific ideas but we don’t have a method for coming about with ideas. Today we will ask where do ideas come from - can we train ourselves to
Anyone can be creative - and anyone can become more creative
In this course you will learn thinking tools to become more creative
We hope that you will then take these tricks home with you and incorporate then into your science and make big discoveries.
If we told each other about the process of how our hypotheses actually came about, we might all dive more confidently into our next night science explorations. We may even be able to distill strategies for how to make this creative process more fun and productive.
To a first approximation, scientific research is composed of two distinct domains. The executive domain is concerned with the testing of hypotheses in a robust and consistent manner. This is the domain governed by the scientific method that is taught from elementary schools to universities. Its complement is the creative side of science, which generates the initial insights and ideas that lead to the hypotheses in the first place. The two domains are intimately linked in what we have called the data-hypothesis conversation. Following Francois Jacob [REF], we have called these two domains day science and night science. While scientific progress relies equally on advances in both domains, the methods and practices of night science are not well understood, and perhaps for this reason are not formally taught.

2 of 59

Itai Yanai

Professor of Biochemistry at New York University, School of Medicine

Undergraduate degrees in Engineering and Philosophy of Science

PhD in Bioinformatics, Boston University

Postdoctoral fellow at the Weizmann (Israel) and Harvard

Expertise:

Single-cell RNA-Seq and spatial transcriptomics.
Cancer biology and host-pathogen interactions
Evolutionary and developmental biology
Computational biology

Professor of Computational Cell Biology at Heinrich Heine University Düsseldorf �(Institute for Computer Science & Department of Biology)

Undergraduate degree in Physics

PhD in Theoretical Physics, Cambridge University

Postdoctoral fellow at the University of Bath and EMBL Heidelberg

Expertise:

Computational systems biology
Evolutionary genomics
Researching the functional evolution of bacteria and plants

Martin Lercher

3 of 59

Podcast & Editorials

For more, check out our main website:

www.night-science.org

4 of 59

How does science work?

5 of 59

Karl Popper and falsification

You cannot prove anything to be true; but, you can prove that something is false.

What you should be doing then is subjecting any hypothesis to rigorous testing to try to prove that it is wrong, i.e., to falsify it.

If you cannot falsify it, then you can say that it’s the best that we have for the moment – but we must always be ready to abandon the idea if it is later falsified.

6 of 59

“Science in the works has two aspects: what could be called day science and night science”

François Jacob (1920-2013), discoverer of

principles of gene regulation

But where did you get the hypothesis?

But the bright day is just one half of the cycle. What is on the night side? Reflect for a second on the hypothesis that you are testing. Did you pull it from the ether?

How? There is no single answer to this question. In many cases, we may not even have a coherent answer, which may be why we prefer not to include it in most accounts of the scientific process. As Jacob says: “Night science wanders blind. It hesitates, stumbles, recoils, sweats, wakes with a start. Doubting everything, it is forever trying to find itself, question itself, pull itself back together. Night science is a sort of workshop of the possible where what will become the building material of science is worked out” [3].

Night science is not as clear-cut and easy describe as day science - it’s messy and abstract at times

But it’s crucial because where the ideas come from

7 of 59

“Science in the works has two aspects: what could be called day science and night science”

“Night science wanders blind. It hesitates, stumbles, recoils, sweats, wakes with a start. Doubting everything, it is forever trying to find itself, question itself, pull itself back together. Night science is a sort of workshop of the possible where what will become the building material of science is worked out.”

François Jacob (1920-2013), discoverer of

principles of gene regulation

But where did you get the hypothesis?

But the bright day is just one half of the cycle. What is on the night side? Reflect for a second on the hypothesis that you are testing. Did you pull it from the ether?

How? There is no single answer to this question. In many cases, we may not even have a coherent answer, which may be why we prefer not to include it in most accounts of the scientific process. As Jacob says: “Night science wanders blind. It hesitates, stumbles, recoils, sweats, wakes with a start. Doubting everything, it is forever trying to find itself, question itself, pull itself back together. Night science is a sort of workshop of the possible where what will become the building material of science is worked out” [3].

Night science is not as clear-cut and easy describe as day science - it’s messy and abstract at times

But it’s crucial because where the ideas come from

8 of 59

The actual scientific method: �periods of day science & night science

Night science allows you to float between ideas.

This often happens in associative leaps rather than in logical steps.

Night science is the creative part of science!

In day science, we may test a hypothesis using established protocols, and we may move to neighboring ideas in small, logical steps. But ideas that are unconnected or only loosely connected are out of reach when all we rely on are established protocols and logic. This is why we often have to pop out into the world of night science, where we float between ideas that may be only loosely connected, often moving in associative leaps rather than in logical steps (Figure 1). Intermittently, we may pop back into the world of day science to examine the apparent merit of a night science idea in the light of day, and maybe to even submit it to the rigorous hypothesis testing at the heart of day science - before popping back out into the dream world above to continue our exploration.

Night science is of course not restricted to a particular time of day, just as we can test hypotheses after 10pm. But these two aspects are distinct frames of mind - so different that they seem like day and night.

Night science is a crucial part of science, and the creativity that we find in this realm is not only needed for the generation of novel hypotheses but also, for example, for the development of new methodologies.

And yet when we talk about science, we make it sound as though it is a march of pure rationality, where scientists go from one logical step to another. But science as we know it is not like that; its night side is far from clean and rigorous. To cover up this untidy side, and encouraged by journal editors and reviewers, we take great care to tell the stories of our projects as if they had consisted of rigorous day science from day 1, hiding the night from the eyes of our peers. While we do not advocate that all scientific papers should be written as a diary-style account of the actual process, we do believe that the stories of night science are not only beautiful, but that their explicit study would add an extremely important perspective to the nurturing of young scientists. If we told each other about the process of how our hypotheses actually came about, we might all dive more confidently into our next night science explorations. We may even be able to distill strategies for how to make this creative process more fun and productive.

9 of 59

The day/night distinction is shared with the arts

Music:

Art:

Night Music Day Music

Night Art Day Art

Lady Gaga

Frida Kahlo

The distinction between executive, systematic day phases and exploratory, creative night phases is not limited to science, and may be general to all enterprises that require at least a minimal level of creativity. In the visual arts, for example, one might distinguish between day art and night art. Day art executes an idea in the studio. Night art is the phase that comes before—or sometimes in between—the execution, where the artist develops the idea of what to create—the composition of a painting or a sculpture, for example. By the time the artist knows what she wants to paint, a majority of the creative process may already have happened. In the same way, there may be day music, the act of producing sound or of working out the details of an arrangement, and night music, where musical ideas take shape. In all these fields—science, art, music—the contribution of the creative, night time activity to the success of the whole project is obvious. But there is hardly any human activity that does not require some level of creativity, and so at least a minimal amount of “nightly” exploration may be an integral part of almost everything humans do.

10 of 59

Improvisational science

“The fun is always on the other side of a ‘yes’ ”

Martin De Matt

(improvisational theatre instructor)

11 of 59

Taking a walk with a friend has always been a crucial part of the scientific method

12 of 59

Rule #1: You need someone you can talk to

The benefits of discussing with a colleague:

The other person can make a useful suggestion,
Together you can pool information or ideas that you lack individually,
A way to uncover errors in reasoning,
Receive encouragement about the project,
To escape from an established habit of thought.

From Beveridge’s ‘The Art of Scientific Investigation’

13 of 59

Traditional brainstorming is not ideal

Brainstorming by asking a group of people to suggest ideas on a given problem:

Brainstorming this way does not seem to work - at least not for us.
The topic gets decided in an ad hoc way.
Diffusion of responsibility: In a large group, people tend to stay silent.
There appears to be judgement on the ideas.
Ideas do not get developed.

14 of 59

Group Pressure and Conformity

Conformity - adjusting one's behavior of thinking to coincide with a group standard.

Solomon Asch Experiment (1951): Actors pick the wrong line and then people are coerced by conformity to pick something that they know is wrong:

Alone: <1% made the mistake

In Groups: wrong 33% of the time

‘Group-think’ is a potential problem with brainstorming.

15 of 59

What is the ideal group size?

Nature, 2019

16 of 59

Rule #2: Discussions are better when they are one-on-one

Pairs: Working in pairs seems to be ideal.

No complicated group dynamics & group think.

A person gets attached to their idea.

A second person can help to make progress by making changes.

It’s much more fun to work with another person.

A pair can travel further without getting distracted.

17 of 59

Rule #3: Think of the discussion as an improvisation

Adopt a mindset where you are open to new ideas.

Remember that new ideas tend to sound strange and weird when you first hear them.

The ‘Yes, and’ rule of scientific improvisation

Same as for �improvisational theatre

18 of 59

Rule #4: Make the other person look good

There is improved creativity for those who practice encouraging statements (related to self-confidence & releasing anxieties) and omit self-incapacitating statements.

We all feel stupid �(We all have the imposter syndrome).
We can get better into the creative state and reduce our inhibitions if we feel comfortable with our discussion partner.

19 of 59

Rule #5: It’s not just all talk

We need periods of improvisational science with a colleague.

But the notion of a lone genius is not entirely a myth. There really is a need for periods of intense work that happens in isolation.

20 of 59

Improvisational conversations are all around us and yet they may be the hidden part of the scientific method.

small group size, ideally just a pair of people;�
exercising the “yes, and” principle common to other forms of improvisation;
emphasizing an atmosphere of trust – making the other person look good.

21 of 59

The two languages of science

“If we allow ourselves the license of talking about genes as if they had conscious aims, always reassuring ourselves that we could translate our sloppy language back into respectable terms if we wanted to, we can ask the question, what is a single selfish gene trying to do?”

Richard Dawkins, The Selfish Gene

22 of 59

“The Selfish Gene” by Richard Dawkins (1976)

“We are all survival machines – robot vehicles blindly programmed to preserve the selfish molecules known as genes”

23 of 59

Anthropomorphisms are a special type of metaphor

To ‘anthropomorphize’ is to give human attributes to nonhuman things.
Anthropomorphisms are a special type of metaphors.
Some believe that anthropomorphizing can lead to miscommunication and misunderstanding.

Clearly genes, proteins, and cells do not “feel” or “want” anything, they are not driven by intentions, but instead move and change based on physical and chemical forces.

Human speech is chock-full of metaphors: it has been estimated that in trying to capture the imagination of our brothers in arms, we deploy several metaphors a minute [4]. (There were four in the previous sentence.) Anthropomorphisms are a special type of metaphor that is particularly prevalent in informal speech among scientists. At the same time, many scientists consider them to be inappropriate in scientific discourse, arguing that metaphors and anthropomorphisms belong to the realm of poetry, while science is the domain of rigor and precision. Calling a gene “selfish,” we are taught, can lead to miscommunication and misunderstandings. Indeed, it has been argued that anthropomorphic images of inter-microbial warfare in microbiology have misdirected research and prevented scientists from understanding the diverse functions of “anti-bacterial” small molecule secretion [5]. As a consequence, students are taught not to anthropomorphize when thinking or talking about biology [6], and many a student has had their assignment edited in red ink by their professor with the epitaph “Do not anthropomorphize!”

The case against anthropomorphism in science is simple and straightforward: genes, proteins, and cells do not “feel” or “want” anything, they are not driven by intentions, but instead move and change based on physical and chemical forces. Anthropomorphic thinking is typical of reasoning in early childhood, and understanding the true underlying causes of inanimate “behavior” is an important step in maturation [7]. So why are anthropomorphisms still ubiquitous in informal scientific thoughts and discussions [8]? They are not restricted to biology but are found across all sciences; anthropomorphisms are particularly widespread in artificial intelligence research, where they have been characterized as “at best misleading and at worst downright dangerous” [9]. Is anthropomorphic language a bad habit that we should all strive to eradicate, or could a case be made for its usefulness?

24 of 59

Our brains are wired to adopt the ‘intentional stance’

Social cognition was crucial in our evolution.
To interact with other people we try to predict their actions.
We do this by adopting the intentional stance: we infer beliefs, desires, and intentions
Neurological research supports this notion.
This is analogous to GPUs, which are optimized for processing image-based information

Spunt et al. J Cogn Neurosci. (2015)

A defining feature of the human animal is its reliance on social structures and interactions. Throughout our evolution, a high level of social cognition was crucial: on the one hand, our livelihoods frequently depended on collaborations within a population; on the other hand, much of prehistoric human mortality is believed to have been caused by other humans [10]. Even today, to cooperate and occasionally compete with other people, we must understand and predict their actions with reasonable speed and accuracy. Our brains achieve this by taking an “intentional stance,” utilizing a “theory of mind”: we infer beliefs, desires, and intentions, and we then use them to predict corresponding actions [11, 12]. It is thus reasonable to suggest that the human brain has evolved under selective pressures that favored fast, intuitive predictions for the behavior of other humans based on their presumed intentions.

This idea is backed up by neurological research. Much of our body’s energy budget is consumed by the brain, and most of our brain’s energy is spent on spontaneous activity that continues when the brain is apparently at rest [13]. The brain regions showing most of this costly activity coincide with areas associated with adopting an intentional stance [14,15,16,17,18]. Functional MRI studies have even suggested that spontaneous activity of these regions primes the brain to adopt an intentional stance when presented with a problem [17].

Thus, our brains may have been optimized by evolution for processing information that is cast in an intention-based framework. This specialization is analogous to that of GPUs, which are optimized for processing image-based information—but which can be repurposed for any problem that can be cast in a mathematically equivalent form.

25 of 59

The language of day science

Day science language is used in:

Scientific papers
Conference presentations
Grant proposals
Journal clubs or peer reviews of manuscripts, where we ask if the claims are supported by evidence

Day science starts with a hypothesis that must be expressed in precise, metaphor-free language.

Day science then sets up conclusive experiments with the appropriate controls, and uses statistical methods to assess our confidence in the results

And to make sure that a test is logically valid, we must express the tested hypothesis in precise, metaphor-free language.

The day science language required to describe the testing of hypotheses is the one used almost exclusively in scientific manuscripts, in which we start with a hypothesis, rationalize it (often in retrospect), and then proceed to test it. The language of day science is spoken at conferences, where we attempt to convince our colleagues of our findings. In grant proposals, day science language takes center stage as we describe an approach for rigorously testing an important hypothesis. It is also the exclusive language of journal clubs and manuscript reviews, where the testing of a scientific claim forces us to rigorously challenge the underlying hypothesis, to see if it can withstand the challenge.

26 of 59

The language of night science

As humans, we require a language permissible to intuition—one that gives us a “feeling” for the phenomenon.
In night science language, we are allowed to anthropomorphize freely, helping us to grasp why and how something may be happening.
Night science language is not precise, but what we lose in rigor, we gain in intuition.
We ask “what does the organism want?” We put ourselves into the shoes of a gene. We ponder the best strategy for a genome in a given situation.
We speak in night science language when we are improvising with a colleague.

We need a language inducive to improvisation, intuition, creativity

While the language of day science facilitates the precise formulation and testing of scientific ideas, it is not particularly conducive to the creation of new ideas. As humans, we require a language permissible to intuition—one that gives us a “feeling” for the phenomenon. This is the language of night science. In night science language, we are allowed to anthropomorphize freely, helping us to grasp why and how something may be happening. Night science language is not precise, but what we lose in rigor, we gain in intuition. We ask “what does the organism want?” We put ourselves into the shoes of a gene. We ponder the best strategy for a genome in a given situation.

The typical questions you may be asking yourself or your colleagues differ substantially between the two languages of day and night science (Table 2). To apply the intentional stance in biology, harnessing the intuition founded in our brains’ ability to predict intention-based behavior, you should ask questions that would sound silly in day science. If you were the protein, what would be your goal? If you were the cell, what would you try to achieve with a certain regulatory pattern? This language not only lacks rigor, but may even be misleading; of course, genes, proteins, and cells are not rational agents with beliefs and desires. But this is the nature of night science: anything goes, as long as it may potentially give you valuable ideas about a system. The ideas may turn out to be completely wrong when subjected to the tests of day science, but along the way, we may figure out something important about the underlying system. For the lay public in particular, night science language allows for an intuition about a scientific topic; this is why anthropomorphisms and other metaphors abound in the popular science literature. Without a science apprenticeship, such intuition by itself does not easily lead to a translation to precise day science language, but the understanding it conveys is powerful.

27 of 59

Translating between �night science language and day science language

“If we allow ourselves the license of talking about genes as if they had conscious aims, always reassuring ourselves that we could translate our sloppy language back into respectable terms if we wanted to, we can ask the question, what is a single selfish gene trying to do?”

Richard Dawkins, The Selfish Gene

Night science language appears to commit the error of attributing desires and intentions to non-human entities such as genes, proteins, and cells. But while this kind of language seems inaccurate and misleading, in many ways it is simply shorthand for respectable day science language. Scientific statements expressed in the anthropomorphizing terms of night science can be rephrased in the language of day science. Examples are given in Table 1. One of these is taken from our book The Society of Genes, where we wrote that “a cancer gene aims to secure an unfair advantage” [25]. Of course a gene does not try to do anything—a gene just is. What we meant can be rephrased more rigorously as “a mutation to a proto-oncogene that causes an increased growth rate of the cells that carry the mutation will over time lead to an increase in the total fraction of body cells that carry the mutation.” When we express this concept in night science language, we know that we can translate it back to rigorous day science language. Yet we gain an intuition for the physiological and evolutionary processes the cancer gene is involved in by thinking of it as though it had intentions, and this helps us to think faster and deeper about tumorigenesis. Night science language exploits a particular bias of our brain circuitry, drawing upon its well-honed techniques for revealing intentions and basing projections on them.

28 of 59

Night science language can be rephrased in the language of day science.

Night science language:

“a cancer gene aims to secure an unfair advantage”

Translated into respectable Day science language:

“a mutation to a proto-oncogene that causes an increased growth rate of the cells that carry the mutation will over time lead to an increase in the total fraction of body cells that carry the mutation.”

So here’s an example

Night science language appears to commit the error of attributing desires and intentions to non-human entities such as genes, proteins, and cells. But while this kind of language seems inaccurate and misleading, in many ways it is simply shorthand for respectable day science language. Scientific statements expressed in the anthropomorphizing terms of night science can be rephrased in the language of day science. Examples are given in Table 1. One of these is taken from our book The Society of Genes, where we wrote that “a cancer gene aims to secure an unfair advantage” [25]. Of course a gene does not try to do anything—a gene just is. What we meant can be rephrased more rigorously as “a mutation to a proto-oncogene that causes an increased growth rate of the cells that carry the mutation will over time lead to an increase in the total fraction of body cells that carry the mutation.” When we express this concept in night science language, we know that we can translate it back to rigorous day science language. Yet we gain an intuition for the physiological and evolutionary processes the cancer gene is involved in by thinking of it as though it had intentions, and this helps us to think faster and deeper about tumorigenesis. Night science language exploits a particular bias of our brain circuitry, drawing upon its well-honed techniques for revealing intentions and basing projections on them.

29 of 59

Translating night science language into respectable day science language

Night science language appears to commit the error of attributing desires and intentions to non-human entities such as genes, proteins, and cells. But while this kind of language seems inaccurate and misleading, in many ways it is simply shorthand for respectable day science language. Scientific statements expressed in the anthropomorphizing terms of night science can be rephrased in the language of day science. Examples are given in Table 1. One of these is taken from our book The Society of Genes, where we wrote that “a cancer gene aims to secure an unfair advantage” [25]. Of course a gene does not try to do anything—a gene just is. What we meant can be rephrased more rigorously as “a mutation to a proto-oncogene that causes an increased growth rate of the cells that carry the mutation will over time lead to an increase in the total fraction of body cells that carry the mutation.” When we express this concept in night science language, we know that we can translate it back to rigorous day science language. Yet we gain an intuition for the physiological and evolutionary processes the cancer gene is involved in by thinking of it as though it had intentions, and this helps us to think faster and deeper about tumorigenesis. Night science language exploits a particular bias of our brain circuitry, drawing upon its well-honed techniques for revealing intentions and basing projections on them.

Of course, the notion of there being two languages in science—a night science one and a day science one—is itself an analogy: both use the same words and grammar. But they are as distinct in terms of modes of thinking as perhaps the liberal arts and sciences were when C. P. Snow complained that they were “two cultures” [29]. There is good reason to keep the two languages separate, emphasizing the distinction between the generation of ideas and their subsequent testing. This distinction is the hallmark of modern science. In contrast, alchemists typically did not distinguish between day and night science. They believed in a deeper, fundamental truth behind metaphors, blurring the distinction between metaphorical and real relationships [30]. Without a clear distinction between ideas and tests, their science was severely hampered.

30 of 59

The language of discovery

A truly new idea is born only roughly formed—there may not even be words for it yet.

To make that idea falsifiable and thus testable by day science, it needs to be reshaped and refined, and an inexact yet intuitive language is all we initially have for that task. Day science starts with a hypothesis that must be expressed in precise terms.

Night science language not only helps to provide an intuition about complex ideas. For the generation of some important new ideas, it is absolutely necessary, as not every metaphorical night science idea is translatable one-to-one from the outset into precise day science language in the way shown in Table 1. A truly new idea is born only roughly formed—there may not even be words for it yet. To make that idea falsifiable and thus testable by day science, it needs to be reshaped and refined, and an inexact yet intuitive language is all we initially have for that task (Fig. 2) [31].

As an example, consider selfish genes and how the use of this anthropomorphic language enabled the discovery of ultra-selfish elements in the genome. Since the organism is only a “survival machine,” a gene may “choose” to be so selfish that it does not provide any useful contribution to organismal fitness at all. This notion has led to the discovery of selfish DNA: genes that can secure their survival in the genome with a copying mechanism that provides a survival strategy without an additional function for the organism [32, 33]. Translated into day science, this idea has now been well established with the recognition and characterization of transposable elements. Of course, these may also acquire functions beneficial to the fitness of the entire genome, but this should not distract from their origin as purely selfish elements.

31 of 59

Lost in translation

Night science language can lead to misunderstandings if we do not clearly distinguish it from the realm of day science.
When venturing into night science language in a scientific talk or paper, it should be explicitly noted.
An insistence on day science language during a night science discussion will bring the flow of thoughts to a halt.

The Selfish Gene may be one of the most misunderstood books of all times, in no small part because of its anthropomorphic title. In public events, Dawkins has repeatedly answered questions about how stretches of DNA could be called selfish by pointing out that the questioner apparently did not read the footnote to the book’s title—the book itself. In the foreword to the book’s 30th anniversary edition, Dawkins went so far as to claim that he regrets the anthropomorphism in its title—in retrospect, he should have taken a friend’s advice and called the book “The Immortal Gene.” But Dawkins does not really mean it. The Selfish Gene provided its readers with an intuition for how to think about genes as the central target of natural selection, and hence as the agents of the story of life on this planet.

So why do anthropomorphisms cause so much aggravation? A lot of it can be resolved if we clearly distinguish between the realms of night science and day science and their respective languages. As has been pointed out over and over again, confusion and misunderstandings can arise when we bring anthropomorphisms into the world of day science. If we decide to venture into night science language during a scientific talk or in a journal article, we are leaving the default day science mode. To not confuse our audience, we should explicitly note that we are venturing into night science territory, but that everything we are about to say can be translated into rigorous day science language—if that is indeed the case.

Irritation can also result inversely, when an insistence on day science language infiltrates a night science discussion. When our brain operates in its creative mode, we should not be expected to speak precisely about an idea only being born; this is not the time for day science aspects such as statistics and controls. For example, when speaking in night science language about how a cancer cell might want to evade the immune system with a specific strategy, hearing your interlocutor reply “You know, cells do not actually want anything” may bring your flow of thoughts to a halt.

Our scientific training is conducted in the language of day science: in its terms, we learn how to design an experiment with negative and positive controls, how to find hidden assumptions, and how to use appropriate statistics. This pattern may lead to the assumption that it is the only language appropriate for science. Yet it is equally important to learn the language of night science, in which we try to see the world from the vantage point of your object of study—a gene, a protein, a cell—and ask “What does it want?” Night science language has its place wherever we want to develop an intuition, by taking advantage of the way our brains are wired to adopt an intentional stance, predicting what an object will do given its inferred beliefs and intentions [12]. In many ways, thinking and talking in night science mode may be more related to what Daniel Kahneman has called “fast thinking,” which draws on instinctive heuristics and is intuitive and emotional; conscious, rigorous day science may be closer to the deliberate and logical “slow thinking” mode [36]. We need both of those modes, and scientists advising their students to not anthropomorphize (instead of telling them when to anthropomorphize) may be doing science a disfavor. Night science language unlocks the potential for fast, intuitive exploration; in it we all become poets, and for the better of science.

32 of 59

What is�the question?

33 of 59

We have a false sense of the structure of knowledge:

We see knowledge as fitting together in a ‘wall of knowledge’.
We tend to look for ‘knowledge gaps’

Is our knowledge like a brick wall with some knowledge gaps to be filled in?

34 of 59

We may be forced to build a new wall orthogonal to the first, or even to tear down parts of the existing structure.

A truly new question is an unknown unknown - we are neither aware of it nor do we understand it.

A discovery is unexpected and may not fit neatly into our wall

But this picture gives a false sense of the structure and rigidity of knowledge and its accumulation. The nature of discoveries is that they are unexpected: they may not fit neatly into our existing edifice of knowledge. Although the research may be originally motivated by a perceived gap, the knowledge resulting from the discovery may in fact not complete any part of the wall but instead may lead to the construction of a completely new and unexpected area: we may be forced to build a new wall orthogonal to the first, or even to tear down parts of the existing structure. This is an uncomfortable concept for many of us, who would prefer a tidy and beautiful universe, where a rational process helps us to illuminate the world. And yet, the most interesting unknowns of science are unknown unknowns—gaps that we were not even aware of before chancing upon them.

A truly new question, as an unknown unknown, is not predictable, and generating it requires night science in addition to our day science work. This aspect of the research process is often hidden by the work that follows the invention of the question. In some cases, scientists may spend many years on answering that question, as in the quote on Stephen Hawking at the outset. And while scientists are systematically taught the process of day science—experimental design and controls—they are typically immersed only slowly in the depths of night science: A student joining a lab is often presented with a hypothesis to work on and may see science as a hypothesis-testing endeavor. Many young postdocs have been told that as a PhD student their job was to answer questions—now they have to discover their own unknown unknowns.

35 of 59

Some questions are so general that they do not provide a new direction towards an answer.

Answering them requires a rephrasing (or refocusing) of the original to expose a new aspect - made possible after an insight.

Rephrased questions that led to scientific breakthroughs

Community-generated questions such as those in the left columns of Tables 1 and 2 are typically so general that they do not provide a new direction towards an answer. Answering one of them almost always requires a rephrasing, a refocusing of the original question, which exposes a new aspect of the problem and only becomes possible after an insight into the phenomenon at hand. As an example, “Does the microbiome affect a tumor’s growth?” is a valid question, but it can only serve as a starting point for our explorations. After some initial analyses and much subsequent night science, we might go back and ask, for example, “Does a tumor manipulate the microbiome as a kind of co-conspirator?”, or “Can bacteria become intra-cellular components of a cancer cell?”. These may lead to hypotheses that are testable and novel.

Sometimes, such new questions will not even be posed in response to a specific public question, yet may lead to answering it in an unexpected way. Our ignorance about a certain topic often provides a fertile ground for novel questions [12]. Discovering the question follows from immersion in a particular topic. Francisco Mojica, for example, provided a radically new hypothesis for why bacterial genomes have a structure that previous researchers had termed “clustered regularly interspaced palindromic repeats” (or CRISPR), separated by evenly sized “spacers” of apparently random DNA. Before Mojica’s work, not many scientists were interested in these peculiar structures, and the problem of CRISPR elements could have been stated as “Why do bacteria have CRISPR elements?”. However, this question is too general to be solved, lacking any hints at where to look for the answer. The inconspicuous spacers were largely ignored. Mojica, however, asked [11]: what does the similarity of the spacers to known DNA sequences tell us about their function? Again, it was the question that led the way, generated in night science but requiring rigorous day science for its answer: the spacers are copies of viral sequences, guiding an adaptive bacterial immune system towards their destruction. Table 2 lists more examples of questions whose refocusing led to breakthroughs.

36 of 59

The scientific method:

The perceived (day science) & hidden (night science) view:

“I’m very confused” a student would tell Uri, to which he would reply,

“Oh good - So you’re in the cloud!”

Uri Alon. How to choose a good scientific problem. Mol Cell. 2009

Finding the question can be fun, but it can also be extremely difficult psychologically.
Scientists are often expected by the public to know it all, and yet, “feeling stupid” is a common mode of operation for us [13].
Science is the art of dealing with things we do not know enough about. As Wernher von Braun, the father of German and US rocket programs, phrased it: “Research is what I’m doing when I don’t know what I’m doing.” Science is humbling in this way. For young scientists, it is often very difficult to understand that it is perfectly normal to not know the answer—or even the question. Learning to embrace this uncertainty is part of our maturation as scientists.
Uri Alon has an intuitive image to describe the process of re-finding our questions [14]. Given what we know about a given topic “A,” a researcher predicts that it should be possible to arrive at point “B,” a scientific destination that seems interesting—a hypothesis. However, the plot inevitably thickens over the course of the research project, and new hurdles force the scientist into a meandering path. Soon, the researcher is lost, having lost sight of the start point (which suddenly seems shaky) and end point (which appears unreachable). Uri calls this “being in the cloud”—you have lost your original question, but the reason why this has occurred is strange and thus potentially exciting and itself worthy of study. From inside the cloud, the situation may seem desperate, but Uri sees the cloud as the hallmark of science: if you are in the cloud, then you might have stumbled upon something non-obvious and interesting. “I’m very confused” a student would tell Uri, to which he would reply, “Oh good - So you’re in the cloud!”
Eventually, a new question that arose inside the cloud may lead the way to an unexpected destination “C.”

37 of 59

The perceived (day science) and hidden (night science) view of the scientific method:

The scientific method is often perceived as a simple sequence that leads from a problem to an answer, possibly through long iterations of modified hypotheses. But our reality is much less structured: it often starts with a topic and some observations, leading to the finding of patterns and questions about those patterns, possibly long before we have any explicit hypothesis or any direct tests (Fig. 2). And even if a project starts out with a very specific hypothesis, in our experiences, it still generally arrives at a very different point than expected.
In some way, then, night science may be most productive when it has no agenda, when there are no particular questions it is trying to reshape or resolve. When the scientist does not have a hypothesis, she is free to explore, to make connections. In some sense, any kind of expectation on how things are to behave—a hypothesis—is a liability that could obstruct a new idea that awaits our discovery. Once night science elucidates and reframes this question, the researcher can use the full power of day science to solve it. In this sense, a major discovery is typically both the solution and the problem.
If an idea is truly unexpected, then we could not have arrived at it solely through existing questions; instead, we had to navigate through night science, moving from disparate observations to previously unknown questions. It is freeing and exhilarating to embrace this uncertainty, to fly right into the heart of the cloud, even if we may feel stupid and lost there. Night science, that realm where questions and ideas are born, appears so mysterious that it is often not described at all. But it is our premise that there are patterns to it, and this is what will occupy us in the following installments of this mini-series.

38 of 59

“Is the scientific paper a fraud?”

“I mean that the scientific paper may be a fraud because it misrepresents the processes of thought that accompanied or gave rise to the work that is described in the paper.

That is the question and I will say right away that my answer to it is 'yes'. The scientific paper in its orthodox form does embody a totally mistaken conception, even a travesty, of the nature of scientific thought.”

Peter Medawar

“Father of transplantation"

39 of 59

Scientific advances…

are predictable
fill knowledge gaps
answer specific questions
proceed in a straightforward manner

40 of 59

Scientific advances…

are predictable
fill knowledge gaps
answer specific questions
proceed in a straightforward manner

41 of 59

The Data-Hypothesis Conversation

“ ‘When someone seeks,’ said Siddhartha, ‘then it easily happens that his eyes see only the thing that he seeks, and he is able to find nothing, to take in nothing. [...] Seeking means: having a goal. But finding means: being free, being open, having no goal.’ ”

Hermann Hesse

42 of 59

Night science is an attitude!

It encourages us to explore and speculate. We ask: What could be hiding here? How would we lure it out?
We try look at the data from as many angles as possible.
In this mode, we take on a sort of playfulness with the data, comparing everything to everything else.
We become explorers, building a map of the data as we start out in one direction, switching directions at crossroads and stumbling into unanticipated regions.

“Not all who wander �are lost”

We typically acquire data with the expressed goal of testing a specific hypothesis. But as we have seen with the gorilla experiment, we are likely to miss other interesting phenomena as soon as we are in a mental mode of hypothesis testing. To account for this, we must consciously adopt a different mindset—one of exploration, where we look at the data from as many angles as possible. In this mode, we take on a sort of playfulness with the data, comparing everything to everything else. We become explorers, building a map of the data as we start out in one direction, switching directions at crossroads and stumbling into unanticipated regions.

Essentially, night science is an attitude that encourages us to explore and speculate. We ask: What could be hiding here? How would we lure it out? Night science may occur when we are most relaxed, such as when Friedrich Kekulé dreamingly looked into the fireplace in his study on an evening in 1862, until his mind formed the image of a molecular serpent biting its own tail—an image that he immediately converted into the hypothesis for the ring structure of benzene [7]. However, more often than not, night science may require the most acute state of mental activity: we not only need to make connections where previously there were none, we must do this while contrasting any observed pattern on an elaborate mental background that represents the expected. To see the discovery in our gorilla experiment, all that was needed was some notion of primate appearances. But when you roam the limits of the scientific knowns, you need a deep understanding of a field to even recognize a pattern or to recognize it as surprising. Different scientists looking at a given dataset will do this against a backdrop of subtly different knowledge and expectations, potentially highlighting different patterns. Looking is not the same as seeing, after all, and this may be why some of us may stumble upon discoveries in data that others have already analyzed.

https://www.wired.com/2013/07/why-do-we-love-making-maps/

43 of 59

Exploration requires a deep understanding of a field

Different backgrounds explain why different researchers discover different things in a dataset.

The same data can lead to different conclusions.

In one study, no two teams chose identical work flows to analyze the same dataset, and sets of teams reported contradictory, statistically significant effects.

When you roam the limits of the scientific knowns, you need a deep understanding of a field to even recognize a pattern or to recognize it as surprising. Different scientists looking at a given dataset will do this against a backdrop of subtly different knowledge and expectations, potentially highlighting different patterns. Looking is not the same as seeing, after all, and this may be why some of us may stumble upon discoveries in data that others have already analyzed.

The extent to which the same data can lead to different conclusions has been the subject of recent studies that provide different scientific experts with a dataset together with a set of hypotheses to test. In one study, 70 independent teams were asked to analyze functional magnetic resonance images and to test 9 specific hypotheses [6]. Strikingly, no two teams chose identical work flows to analyze the dataset, and sets of teams reported contradictory, statistically significant effects based on the same dataset. In another study, 73 teams used the same data to test a single hypothesis; the “tremendous variation” in conclusions led the researchers to conclude a “vast universe of research design variability normally hidden from view” [7]. These and other studies [8,9,10,11,12,13] demonstrate that data is not transparent and that converting data into information through statistical analysis has a substantial subjective component. Our own experiment shows that even when studying the same plot, preconceived biases can result in different interpretations.

44 of 59

Ellen Rothenberg on ‘inhabiting the data’

How do you come up with a new idea?

“I think most of it is inhabiting the data. [...] I know that everybody knows the story about Albert Einstein, imagining himself being a photon. But in a sense, when we look at data sets, that's what we should be doing. Even if we're looking at molecular biology data. If you are this transcription factor in the cell's nucleus: How the hell do you figure out which of the million possible genomic sites you're going to go to in this particular cellular context? [..]

�This really helps you identify what I think are really major outstanding problems in one's field.”

Ellen Rothenberg

CalTech

45 of 59

Patternicity, or “just a correlation”?

‘Patternicity’ is the human ability to find patterns and connections.

Many patterns may be spurious – �“Correlation is not causation”
BUT: finding a new correlation may be �the first glimpse of something new.

Data exploration is a generator of correlations �and patterns that can later be tested for causality.
Correlations are the domain of night science – �causation is solidified by day science.
Science relies on this back and forth between day and night, each overcoming the other’s shortcomings.

Identifying patterns

Testing �for causality

“Correlation is not causation”—an aphorism that perhaps all scientists have heard at least once in their careers—warns of putting too much weight on mere covariation of two variables. Undoubtedly, a correlation between two features is not sufficient to infer a causal relationship. But some form of covariation is implied by a causal relationship, and hence, finding a previously hidden correlation may be the first glimpse of something new. We may then think of data exploration as the generator of correlations and patterns that can later be tested for causality.

One of the major facilitators of human intelligence is our minds’ ability to easily find patterns and connections—a tendency called patternicity by Michael Shermer [8]. Patternicity helps us in generating new night science ideas; it is the seed of many discoveries. On the flipside, patternicity makes us vulnerable to being fooled by randomness [9], when we mistakenly infer relationships between genuinely independent things (called apophenia). Clearly, spurious results will be generated during unguided explorations, and this generation of false starts is night science’s own liability.

Day science tempers this liability. In a sense then, correlations are the domain of night science, while causation is solidified by day science. Day science is the adult in the room, rigorously testing hypotheses. But despite its power, the day science mode is not amenable to generating the ideas in the first place. Only the night science realm, with its lack of specific hypotheses that blind us in day science, allows us to think freely in an exploratory fashion. Science relies on this back and forth between day and night, each overcoming the other’s shortcomings; we can let ourselves explore so freely in night science because we trust ourselves to check the generated hypotheses later, in day science.

46 of 59

Beware: Scientists are biased!

Psychologists refer to “confirmation bias” as the propensity of viewing new evidence as supporting one's beliefs.

Also in science, confirmation bias leads scientists to dismiss or misinterpret publications that contradict their own preconceptions.

THUS: We resist confronting contradictions.

Image: The Upturned Microscope

Scientists are biased—especially you

Psychologists refer to “confirmation bias” as the propensity of viewing new evidence as supporting one's beliefs. Writing in the fifth century BC, the historian Thucydides put it this way: “It is a habit of mankind to entrust to careless hope what they long for, and to use sovereign reason to thrust aside what they do not fancy” [14]. In more recent times, experiments have shown that people demand a much higher standard of evidence for ideas they find disagreeable compared to those they hold themselves [15, 16]. As different individuals have different beliefs and experiences, each member of a community may end up perceiving reality differently, each through the prism of their own biased views. Confirmation bias may explain why harmful medieval medical practices were perpetuated over centuries, as only those patients that recovered (possibly despite rather than because of the treatment) were remembered [17]. The same phenomenon may underlie today’s widespread acceptance of “alternative” medicine [18,19,20].

While modern science appears to demand an objectivity that stands above such biases, confirmation bias is also well documented here. It shows itself, for example, in the peer review process of scientific papers. Studies whose findings are incompatible with a reviewer’s own assumptions are reviewed much more harshly than those that support the reviewer’s beliefs [21,22,23]. Confirmation bias leads scientists to dismiss or misinterpret publications that contradict their own preconceptions, up to the point where contradicting papers are cited as if they in fact supported a favored notion [24].

Beyond publications, does the execution of the scientific method itself also fall prey to confirmation bias? In theory, it should not, as emphasized by Karl Popper’s description of the scientific process [25]. For any given new idea, scientists should attempt with all of their experimental powers to falsify it, to prove it to be wrong. But, as philosopher Michael Strevens has argued, humans are not disciplined enough to strictly follow this method in their daily work [5]. Indeed, as much as we hold the philosophy of falsification in high regard, it simply is not very practical. First of all, hardly anyone actually excludes a beautiful and well evidenced hypothesis at the first sight of a falsifying result. If, for example, an experiment contradicts the conservation of energy, we do not simply throw out the first law of thermodynamics. Instead, we search for some flaw in the experiment or its interpretation [26]. But more than that, as we describe next, scientists—just like any other human beings—are hoping for and hence looking for evidence that confirms, not refutes, their favorite notions. If we make a prediction and the prediction seems to be borne out by the data on the surface, we tend to not dig deeper.

Do not stop just because you like the result

Many scientists and the public at large are concerned about a “reproducibility crisis” in science: the observation that many published results cannot be replicated [27]. While it may be tempting to lay the fault at the door of a select group of misbehaving individuals, the wide scope of the observed non-reproducibility suggests that it may be a pernicious aspect of the scientific process itself [27].

As a point of reference, it is useful to recognize one scientific framework that escaped the reproducibility crisis: clinical trials. The acknowledgement and correction of potential biases is built into the rigid process of clinical trials, where study designs are pre-registered and detailed protocols have to be followed to the letter. Confounding variables are clearly identified a priori, the data is blinded to exclude potential biases of the practitioners, and significance is tested only when the pre-specified dataset has been collected. Clinical trials are carefully designed to be exclusively in the mode of hypothesis-testing; if executed with a large enough sample size, they ought to be immune not only to confirmation bias, but also to other sources of systematic irreproducibility.

In sharp contrast to the purely hypothesis-testing mode of a clinical trial, basic research projects typically do not fully know what to expect from the data before analyzing it—after all, “If we knew what we were doing, it wouldn’t be research, would it?” High-throughput datasets are especially likely to contain information unanticipated in their generation, observations one could not have predicted a priori; for this reason, we may often be better poised for a discovery if we conversed with a dataset without having formulated a concrete hypothesis [28]. The natural place for almost every dataset is right in the middle of the data-hypothesis conversation [29, 30], where it is used both in a hypothesis-testing way and in explorations that look for unexpected patterns [31], the seeds of future hypotheses.

Arguably, the central part of the scientific method is to challenge a hypothesis by contrasting its predictions with data. But when we do that, we are typically anxious for the falsification to fail: unless we are testing someone else’s competing hypothesis, we hope—and frequently expect—that our predictions will be borne out. If the results do not initially comply, we will think about problems with the experiment or with auxiliary assumptions (such as how we expect a genetic manipulation to perturb a cellular system). Sometimes, we will identify multiple such problems, and the results may converge to what we predicted them to be. There is, in principle, nothing wrong with this general approach: the scientific process is very much trial and error, and we cannot expect that an initial experiment and our first analysis were without fault. Obviously, we have to be careful not to selectively exclude contradictory data by making sure that similar errors have not equally affected other data points. There is a more subtle point here, though, a more discrete way in which confirmation bias may creep into our science. Our human instincts will lead to a sense of fulfillment once the expected pattern finally emerges. While this may mark the perfect time for a well-earned coffee break, it is not the time when we should stop analyzing the data. Instead, we should continue to think about possible biases and errors in our experiment, its analysis, and its interpretation. If we do not, we may be abandoning our efforts toward falsification too soon.

One extreme example, where confirmation bias is elevated to a guiding principle, is p-hacking [32]: one modifies the specifics of the analysis until the expected result emerges, subsequently reporting only the final configuration. It is important to realize, though, that in this case, it is the biased reporting that contributes to the reproducibility crisis, not the exploratory analysis itself [33]. Just the opposite: an exploration of how our results vary with changes in the specifics of the analysis, if communicated openly, provides important information on the robustness of our interpretation.

An elegant way to counter the drag toward self-fulfilling hypotheses is to test not one, but multiple alternative hypotheses, a core element of a method John Platt called “strong inference” [34]. Platt argued that the fastest scientific progress results from formulating a set of opposing hypotheses and then devising a test that can distinguish between them. While this is indeed a powerful approach, we often do not know initially what may be the best set of competing hypotheses. Forcing ourselves to look beyond one favored hypothesis in order to come up with such competing hypotheses is a serious—and non-fun—night science task, requiring hard and deliberate work.

47 of 59

Interdisciplinarity & Renaissance minds

48 of 59

Modern science is heavily structured

Our discipline decides:

Conferences we attend
Journals we publish in
Study sections at funding agencies
Courses we teach
Departments to which we get hired

Modern science is probably even more categorically structured than music, with disciplines instead of genres.
Similarly to how many folk music fans felt that Dylan should have remained in folk, scientists are generally expected to focus their career on a subfield of their discipline.
If you work on genome biology, your group is likely a subdivision of the Department of Biology within the School of Natural Sciences. Such restricted playing fields will govern almost everything in our professional lives: the specialized conferences we attend, the topical journals where we publish, the study sections we select within our funding agency, the courses we teach, and the departments to which we get hired [6].
There is a good case to be made for why closely knit disciplines promote scientific progress. First of all, its structure makes day science more efficient. A community sharing a common established knowledge can move ahead to new knowledge. When your paper is reviewed for a journal or presented at a conference, it helps if you can rely on a core set of established ideas, so you can focus on the new aspects.
But if the benefits of structured disciplines make you think that the next big idea in your field will come only from inside the field, you might be wrong.

49 of 59

The quintessential Renaissance mind:

Leonardo da Vinci

(scientist, artist, philosopher, writer, inventor)

50 of 59

Optics (white light composition)
Mathematics (calculus)
The three laws of motion (the basic principles of classical physics)
Law of universal gravitation

Isaac Newton’s work transcends any one field

51 of 59

Fields and subfields are just one way of clustering knowledge.

There are no real boundaries between fields.

http://www.rationalskepticism.org/general-science/is-there-any-classification-of-academic-disciplines-t44278.html

Similarly, the borders between scientific fields and disciplines are not natural boundaries; really, there are no boundaries.
Disciplines, fields, and subfields are just one way of clustering knowledge and methodology on increasingly fine-grained levels, but this clustering is not unique, and there is not even an obvious optimality criterion for the clusters.
Many boundaries may simply reflect the way in which a field developed historically. Working within the confines of a field may help us to structure insights and ideas, but—similar to a musician’s fixation on a certain genre—the boundaries can impede our creativity and restrain our advances into certain directions.
During our most creative night science moments, when we come up with potential solutions for problems and dream up hypotheses, when we need to make new and unexpected connections, we are better off if our mind is free to transcend the fields and disciplines.
After all, if there were no boxes, we would not have to think outside of them. This kind of thinking may also be called horizontal [7] or lateral thinking [8].

http://www.rationalskepticism.org/general-science/is-there-any-classification-of-academic-disciplines-t44278.html

52 of 59

Many scientists were originally trained in a different field.

(In an interdisciplinary project, someone will make the connection.)

Francis Crick

Marie Curie

Training:

Contributed to:

Physics

Physics & Chemistry

Physics

Biology

To transgress the boundaries of a field, it is highly useful to have an understanding of multiple disciplines, either as a person or as a team, as this provides more opportunities to make connections. In the modern practice of science, the interdisciplinary aspect is often interpreted as a collaboration between scientists that work side by side in different disciplines. But true interdisciplinarity—even in a collaborative framework—requires us to think across fields. At some point, someone on the team will need to have that idea, and that someone will likely be the one with access to multiple fields.
Thus, while the framework of science is disciplinary, a scientist’s creativity benefits from interdisciplinarity. This may explain why so many eminent biologists were originally educated in a different field: just think of Max Delbrück, Mary-Claire King, or Francis Crick. But there is also an important role for large and diverse teams: if more varied ways of thinking, more diverse ideas come together at the water fountain, they provide a fertile ground for making connections across borders—the modern workplace replacement of the traditional café, where creative people have traditionally met to exchange ideas [9].

53 of 59

Does increased interdisciplinarity really lead to more insights?

There are contradictory results from bibliometric approaches:

Not clear that increased interdisciplinarity is necessarily better.�

Different forms of interdisciplinarity:

Positive correlation between impact and the variance of the references’ fields;
BUT: Lower impact if references too balanced across different fields, or references from very distant fields.�

Those who engage in interdisciplinary research tend to be less productive than the experts.

Does increased interdisciplinarity really lead to more insights?
Meta-research—research on research—has investigated this using bibliometric approaches: a paper’s impact can be approximated by the number of citations it receives, while its interdisciplinarity is reflected in the diversity of the works it cites. The results are contradictory. Some studies found that adding research fields to a work was associated with an increase in impact, while other studies found that more interdisciplinarity is not necessarily better [11,12,13,14,15,16]. Digging deeper into the association between impact and interdisciplinarity has revealed its benefits, but also its costs. Of particular utility is the distinction between different forms of interdisciplinarity. A positive correlation was detected between impact and the variance of the references’ fields; however, if the different fields in the references were too balanced or if the fields were too distant from each other, this was associated with a lower number of citations instead [15]. Night science explorations may thus be most fruitful if they explore the adjacent possible [9]—the undiscovered knowledge that is still within reach from a given discipline, even if we have to stretch out beyond the field’s artificial borders. Viewing these relationships from the angle of the individual scientists, though, unearths an uneasy truth: those who engage in interdisciplinary research tend to be less productive than the experts [17]. At least in part, this may be a reflection of the difficulties involved in publishing interdisciplinary work.

http://www.ed-archmat-ejd.eu/index.php/vision-aims.html

54 of 59

Expert’s dilemma: Finding balance between disciplinary day science expertise and interdisciplinary night science creativity.

Such meta-research has quantified what many of us have experienced first-hand: a tension between the disciplinary and interdisciplinary aspects of science. As we know more about other fields, our creativity can venture further and in other directions. This empowers our night science. But on the other hand, an interdisciplinary scientist will likely lack some of the intimate knowledge expected of experts in his or her main field, and many peers will see this as an indication that this person is not to be trusted. There are a few polymaths, true experts in multiple fields, but for most of us, substantial knowledge about a second field can only be acquired at the cost of reduced expertise in our primary field.

This is what we call “expert’s dilemma”: the more interdisciplinary you become, the less credibility you may have with your peers (Fig. 1).

Thus, while the interdisciplinary person coming in may have a good idea, it might be dismissed because he or she does not know the ins and outs of the field, and when it is time for a grant panel or a journal editor to evaluate the work, why take it seriously? So while scientists sing the praises of interdisciplinarity, actually working in an interdisciplinary manner as a person can be a professional liability. Interdisciplinarity certainly enhances our night science creativity, but it may at the same time stifle our day time careers.

In response to the expert’s dilemma, different scientists have settled on different positions along the spectrum. Some scientists feel comfortable working at the left extreme of the plot, as a highly specialized expert in a particular field. They will spend many years, often decades, to arrive at an intimate understanding of a particular system. The discovery of the molecular pathway of programmed cell death [18] and protein degradation [19] is an example of this strategy. And of course, there are always the Erdös’s of the world who are continuously changing fields—true scientific nomads. But as always, a middle ground may provide the best tradeoff and may constitute the answer to expert’s dilemma.

55 of 59

Importing an idea from another field

Interdisciplinary creativity is a two-way street. You may realize that a concept or a methodology, or maybe even just an analogy, from another field can aid in developing an answer to a question in your home field. Conversely, a concept or methodology from your home field may help in answering an open question in another field; the connection you discover might even lead to a new question in that field. Table 1 lists several examples in both directions. In all of these examples, the link discovered between the fields was not obvious and could not have been made by thinking purely within a single discipline. Rather, the creative act of finding the link was made possible by the scientist’s interdisciplinary thinking.

A classic example of importing an idea from another field is the application of the theory of natural selection to the field of cancer. Wilhelm Roux, a German zoologist born in 1850, is best known for his pioneering work in experimental embryology and for the establishment of the first tissue culture. But aside from his focal work on embryology, Roux was fascinated by Charles Darwin’s books on the role of natural selection in the evolution of species. In what can only be considered a great night science moment, it dawned on Roux that natural selection was such a general principle that it should also apply to competition between cells inside the body. Roux published his ideas in his 1881 book, The struggle of parts in the organism; much of his later day science was devoted to testing the general ideas first laid out in the book [33]. It took the mainstream of cancer research over a hundred years to absorb this idea, but as we move further into the twenty-first century, few cancer researchers doubt that the spread of cancer cells is governed by an interplay between mutation and selection. The principles of natural selection have also been applied outside of biology. As early as 1873, Harper’s New Monthly Magazine wrote: “By the principle which Darwin describes as natural selection, short words are gaining the advantage over long words, [...] and local idioms are everywhere in disadvantage”—the origin of the idea of memes [32], ideas that spread by manipulating human brains.

56 of 59

Exporting an idea to another field

Once an idea is generated in one field, it may prove so widely applicable as to lead people from that discipline to insights in disparate fields—see the right side of Table 1. A good example is the application of a suite of network analyses across disciplines, led by Albert-László Barabási. In 1999, Barabási and his then graduate student Réka Albert reported that many networks—the Internet, the citation patterns in science, or the collaboration graph of movie actors—have a peculiar property in common. These networks are “scale free”: at any level of magnification, they contain a few exceedingly popular nodes with many connections, while most other nodes only have a single connection [34]. Searching to fund this work, Barabási explored different calls far beyond the fields he had ever worked in. In his book Linked [21], he describes coming across calls from the DARPA agency on “technologies that will allow the computer networks of the future to be resistant to attacks and continue to provide network services.” The connection between this call and his work was tenuous at best, but applying their approaches to the topic of robustness, Barabási’s team of outsiders to the field of Internet security had an important insight: such networks are very robust to error, yet remarkably vulnerable to attack [35].

57 of 59

How can we increase our thinking across disciplines?

Expert by day, Renaissance mind by night

Also, go to an art museum and ask yourself which painting is like your project.

Be aware of the distinctions

between day and night science
between disciplines and scientists.�

Read about many things, even if superficial

popular science books!�

Attend talks that have only a marginal overlap with your work.

How can we increase our night science creativity by thinking across disciplines?

The first step is simply to be aware of the distinctions between day and night science, and between disciplines and scientists.

Disciplines are structured frameworks that guide our day science, but that we may have to traverse in our creative periods. Night science is about discovering the unknown and the unexpected, and thus, there can be no established maps to follow.

To expand our interdisciplinary minds, it may be a good idea to read about many things—and doing this in a superficial way is nothing to be ashamed of; on the contrary, it may often be a necessity. While we cannot become an expert on everything, cultivating wide interests—for example, through popular science books—provides a rough map of the questions that people in other fields are wondering about, the methodologies they use, and the concepts and analogies that guide their thinking. Attending talks that have only a marginal overlap with your work can also be inspiring. By connecting their data and their questions to your own expertise, you may come up with new ideas: a way to use their data to shed light on a question in your field, or a way to address their questions with methodologies familiar to you. This way, you will still be able to skillfully navigate your particular field, while at the same time you gain a familiarity with other fields for inspiration: think like an expert by day, but with a Renaissance mind by night.

Experts will remain crucial to scientific progress, and universities, which developed an intricate disciplinary structure for good reasons. Specialists rule by day. But scientific creativity—night science—is enhanced by our ability to move between disciplines. The pendulum swings back and forth over the years on the emphasis of interdisciplinarity in science. Even when it is called for, interdisciplinarity is often seen from a purely day science perspective, emphasizing the role of large teams of diverse but highly focused, disciplinary experts [36]. In contrast, interdisciplinary creativity seems to thrive with “teams” of only one or two scientists, and often occurs when a scientist without formal training in a given field ventures into it in night science excursions (and, in many cases, subsequently even during day science). So if you always wanted to explore other fields, but felt it might be a waste of your time: "don't think twice, it's alright".

58 of 59

SUMMARY

Day Science Night Science

Mode:	Falsification	Improvisational
Goal:	Hypothesis-testing	Hypothesis-generating
Compartmentalization:	Disciplinary	Interdisciplinary
Language:	Precise	Metaphorical
Logic:	Consistent Answers	Questions from Contradictions
Puzzle-state:	Solution-testing	Puzzle-switching

59 of 59

Podcast & Editorials

For more, check out our main website:

www.night-science.org