Science, a collective game

I was at SFI’s Complexity Interactive when I heard Eric Kaplan tell a story about a person searching for the most important question. I couldn’t help but wonder… What would a story about a collective search for the most important questions look like? And Is that what science does? I’ll try to answer these questions in this blog post and discuss it’s implications, but first, let me give a short overview of the story Eric told

The Buddha, who was incredibly wise, was going on a world tour, giving people a chance to ask him one question, any question. One person gets really worried and thinks all day and night about which question to ask. They don’t want to blow up this incredible opportunity to get the answer to one question! It’s the end of the night, and finally, the person asks:

“Oh Buddha! What is the single best question for me to ask you and what is the answer to it?”

And the Buddha says

“That. That is the single best question for you to ask me and that is the answer.”

Eric Kaplan then discussed that we cannot acquire instruction from others but only provocation. Provocation to find our most important question. In his words,

“Who you are — what bothers you, what your family was like, what happened to you that you hated, what happened to you that you love, the people you allow to matter to you — is what made you ask the question.

And since who you are is what made the question – when you answer them they teach you who you are.”

After hearing this talk, I wondered, how this story would look like for a collective? What are the questions whose answers many people would like to know? How does an individual search compare to a collective search for important questions? Let’s try and map this out with thought experiment.

Imagine you are about to play a game where you can get the answer to any question. Yes, any question. Questions ranging from What will I have for breakfast tomorrow? To what’s the origin of the universe? And how did life emerge? Before playing, though, you have to make a choice. You have to decide between getting the answer to one and only one question of your choosing or getting the answer to five questions that were selected democratically by you and the other four people.

Which version of the game would you like to play? (see box if you’d like a more detailed description of the versions)

If you chose the collective version, I think you are interested in science.

Why? You preferred having the answer to five questions rather than one. These five questions are likely to be general interest questions; after all, they were chosen collectively by five people rather than one. And what is science if not looking for answers to general interest questions? The fact that you took the risk of a bargaining game to obtain more answers rather than settling with one question suggests that you have a natural curiosity.

Of course, other factors unrelated to science may have gone into your decision process. Perhaps you are risk-averse and would rather have the answer to one question you are interested in rather than engaging in a social bargaining process that may result in receiving answers to questions that are out of your interest. Perhaps, you generally dislike social bargaining processes and would rather avoid a potentially long discussion with people who may not share your interests. On the contrary, maybe you enjoy social dynamics so much that you chose the collective version of the game without thinking about the end result. Considering different collective game variants might help us better understand your decision process.

A more detailed description of the two versions of the game

You play the game only once, regardless of which version you choose. In the individual version, you ask one question, get the answer and finish the game. In the collective version, you are randomly grouped with four people. These people are randomly drawn from the adult world population and also chose to play the collective version of the game. Once you are in your group, the following steps happen sequentially.

You discuss with the other group members which questions to ask.
Every member postulates one question (any question).
After all questions have been postulated, the group members vote yes/no for each question.
After all votes for all questions have been written, the votes are publicly revealed. That is, votes are cast at the same time, and you can know who voted yes/no for each question.
The set of questions postulated passes if all the questions received a majority of approving votes (3 or more). If the set of questions passes, all group members receive the answers to all questions. If the set of questions does not pass, the group goes back to step 1.

Would your decision change under different variants of the collective game? Define the collective game for any N > 1 players that democratically choose to ask N/a questions, with a greater or equal to 1. The payoff (questions answered) grows with N and decreases with a. As N increases you get more answers, but your voting power (1/N) also decreases. This means that as N grows, you may be less likely to get answers to questions you are interested in, particularly if a > 1. With how many people would you like to play the game? For which values of a would you be willing to play the collective game?

Would your decision change if the people you played within the collective game were sampled differently? For example, what if you only played the game with people with a bachelor’s degree?

What if you were guaranteed to play the game with a diverse group of people (i.e., from different countries, education levels, races, gender, age, sexual orientations, etc.). What if you played the game with the best experts in the scientific field you are most interested in? Perhaps then the process of deciding the set of questions would be similar to a collaborative grant writing proposal where you aim to answer the top questions in your field.

I’d like to think that research is done through a somewhat similar process to this latter version of the game. Experts in a research field discuss what questions are important to answer through direct discussion, conferences, article citations, or peer review processes. Of course, this idealizes how science works. There are many drawbacks and criticisms to the grant writing and peer review processes, not to mention the big problem of lack of diversity in academia.

Nonetheless, knowledge sharing is a key point in this comparison between the collective game and science research. In Paul Romer’s framework, ideas and knowledge are non-rival goods that can be shared infinitely without depletion. In the collective game answers are shared without cost. Although science communication has a cost, it still holds that sharing scientific discoveries does not deplete the knowledge. The trade-off between the individual and collective game is that for the questions to be approved by the group, they will likely have to be of everyone’s interest and not just one’s interest. Similarly, science funding calls usually require researchers to justify the benefit for the society the proposed research will have.

But, what makes us decide what the most important questions are? Recall Eric’s words on what drives one’s most important question

Perhaps the most important questions are the ones that help us answer, who we are, what bothers us, what happened to us that we hated, and how can we improve the quality of life for the people that matter to us?

This underscores the importance of diversity in science. We usually raise questions that relate to our life experiences and knowledge. Researchers from minority groups can bring to light important questions that were previously ignored. Perhaps this is one of the reasons why research students from diverse backgrounds and minority groups innovate at higher rate than majority students. As long as we don’t have a diverse group of scientists, we are likely to keep ignoring questions relevant to many people around the world.

If you are doing research, are you working on solving what you consider the most important questions? Are you working on an important question that you are most likely to answer given your expertise? The wonderful thing about science is that we share answers. In this sense, it is not crucial to work on what you consider the most important question, but just an important question where you can contribute.

When an impactful event happens, such as the Covid-19 pandemic or the emergence of armed conflict, we might be tempted to drop everything and try and research these new events. Before doing so, it’s good ask ourselves if we can contribute to the research or if we will just add noise. If the latter, it does well to remind ourselves that knowledge is shared, and that our research expertise can contribute to society in other ways. As the number of people in across different science disciplines grows we all gain answers to many more questions.

It does, however, not hurt to ask oneself once in a while

What is my most important question? And what do I think is the most important question for our society?