The Art of Scientific Discovery¶

This work is licensed under a Creative Commons Attribution 4.0 International License.

A course in problem-solving strategy and creative thinking

All materials originally developed by Professor Arthur T. Winfree
Department of Ecology and Evolutionary Biology, The University of Arizona, Arizona Board of Regents

About This Course¶

The Art of Scientific Discovery was a 3-credit course taught by Arthur T. Winfree at the University of Arizona (EEB 479/479H/579). Undergraduates (freshmen included), graduate students and postdocs took it.

Original course materials

Every page here derives from Professor Winfree's handout. Read it in his own words in the transcribed syllabus, or see the original PDF.

What the course is for

"The aim of this course is to develop your ability of solve problems like those encountered in scientific investigations. We examine how specific problems were solved in the past, examine 'inspirational readings', and tackle selected puzzles for pencil and paper and for simple lab manipulation."

It is not a lecture course. In Winfree's words, "The objective is not to add to your store of useful facts." He compared it to weight-lifting: "you will benefit from this course in proportion to your investment of time and effort."

What Makes This Course Different¶

The syllabus calls the exercises "practice scrimmages" in the strategy and tactics of:

recognizing ignorance;
posing questions;
cultivating multiple alternative solutions;
eliminating rejectable candidate solutions;
spotting and taking advantage of your own mistakes;
especially, learning a positive attitude toward mistakes, "because they are often the most available doors to discovery."

The problems depend as little as possible on any one subject, so that "everyone is on an equal footing of unfamiliarity."

The Three Main Tools¶

The syllabus names three tools:

Readings. Two required books, Adams's Conceptual Blockbusting and Ehrlich's Nine Crazy Ideas in Science, plus many short handouts. Winfree expected underlining, marginal notes, and "a page per class session to celebrate the best things you find or think of while involved with these readings."
Problems. "Problems, the equivalent of bar-bells for lifting." Write down your approaches, your lucky insights, and how you got into and out of blind alleys: "This is the main thing, not the 'answers'."
Each other. "The genetic diversity and diversity of life-experiences and habits of thought that we collectively bring to the table." Students swapped GamesWorth notebooks every class, and in class pooled data that no one person could collect alone.

The GamesWorth Method ¶

The problems are there to start a habit: a daily "Game's-Worth" of focused thought, named for the thought it takes to play one serious game of chess. Winfree wrote that this "might be the most important (potentially enduring) effect of the course." The introduction 🔓 to his 2001-02 column Adventures in Discovery describes the same idea as a regular daily workout.

Course Structure¶

The thirty class sessions fall into five sections:

Section 1: Detecting Nonsense, Error Checking, False Assumptions, Cherishing Mistakes (sessions 1 to 6): spotting nonsense and hidden assumptions, and putting facts before explanations of facts.
Section 2: Creative Blocks (sessions 7 to 12): the perceptual, emotional, cultural and intellectual blocks of Adams's book, and ways past them.
Section 3: Observations and Questions (sessions 13 to 18): careful observation, outdoors and in a chemical pattern-formation lab, including noticing what isn't there.
Section 4: Patterns, Empirical Generalizations (sessions 19 to 24): finding rules in data, often gathered in class labs such as Cell Shapes and Egg Pouches.
Section 5: Inferences, Hypotheses, Explanations (sessions 25 to 30): multiple working hypotheses and strong inference, then working out the theory of stacked cantilevers, LoShu, and the laws of a toy universe.

Also on this site:

The GamesWorth Method: the daily notebook, how to log it, and the "morning-after" left-hand pages.
Problems: 55 problem pages, one for each problem and lab linked from the schedule, each with its statement, history, hints and sources.
Readings: the reading list, each item marked open, borrowable or paywalled.
The Original Syllabus: Winfree's handout transcribed in full, with the thirty-session schedule. It is the primary source for everything else on this site.

Key Philosophy¶

Can problem-solving be taught?

"Some people suppose problem-solving cannot be taught. They suppose that you are born with innate ability or not, and that's all there is to it. In contrast, I think we are all born with it and mostly lose it during and because of schooling and the general intimidation that comes with any competitive society. We can refine and enhance it about as much as we please."

Getting Started¶

Start the daily GamesWorth habit now. Pick a problem from Section 1, set aside an uninterrupted hour, and log what you try. Winfree's warning: "form the GamesWorth habit immediately and don't let it slip for any 'reason'."

Remember

"In this course actually solving the practice problems is way less important than learning how to try (and, for grading, demonstrating real effort.)"

"This requires joyful playfulness, not grim determination."

Machine-readable: this page as Markdown · raw source on GitHub · llms.txt · llms-full.txt (whole site). See For AI agents.