The Barr Chart: A Lesson in Demonstrating Causality

On a sleepy, sunny Sunday, I was trolling a local bookstore, waiting for something significant to leap into my arms. Most often, this is an exercise in futility. However, this attempt, the art magazine Esopus attracted my gaze with the manic, color pencil war scenes that only a dedicated eight-year-old boy can render.

Wedged into the heavy stock, full-breed pages of esoterica, was an article on the creative process that went into designing a seminal work in information graphics and art history: Cubism and Abstract Art: 1890 to 1935 by Alfred H. Barr, Jr., the man who drove the creation and direction of New York Museum of Modern Art.

This diagram that Barr created to grace the dustjacket of the exhibition catalog connected the various manifestations of art movements through the turn of the century and showed how their commingling led to later art forms. Barr traced these developments with the tools and approach of a genealogist, almost as if he were creating a family tree. This scientific approach and the resulting analysis was the result of many years of reflection on his part.

Esopus includes a reproduction not only of the final work, but five realistic reprints of the pencil drafts, complete with eraser marks and a vellum, typed letter of suggested edits. I felt as if I'd been given a glimpse into the intellectual and artistic process of a designer moving through iterations in search of the truth.

Edward Tufte includes the diagram in his book Beautiful Evidence. On his blog, he writes about the image's significance [see article]:

It is easy to draw a linking line or an arrow of implied causality, but very hard to make credible causal inferences. Linking lines, arrows, and influence trees bring with them many implicit but powerful assumptions. Suppose we take the arrows seriously--how are we to evaluate the evidential quality of influence diagrams? The answer is clear, at least to the open and skeptical mind: by the usual standards for evidence of causality.

In another post on casaulity, Tufte continues: "One of the fundamental cognitive tasks in analytical thinking is to reason about causality. Thus one of the fundamental principles of analytical design is to show causality."

Austin Bradford Hill's classic essay, The Environment and Disease: Association or Causation?, which linked smoking to lung disease and cancer reproduced on Tufte's site. Hill details the core quality that are essential in demonstrating A caused Z:

1. Strength,
2. Consistency,
3. Specificity,
4. Temporality,
5. Biological Gradient (or "dose-response curve"),
6. Plausibility,
7. Coherence,
8. Experiment,
9. Analogy.