D T Campbell

Don Campbell

We often assume that problem solving was undertaken with one fixed goal in mind. This overlooks the opportunistic, serendipitous course of organic evolution and of much of creative thinking.

~ D. T. Campbell

Donald Campbell was one of the most creative and influential social scientists of the 20th Century. Among his many achievements was the development of rigorous guidelines for how to perform quasi experiments outside of the laboratory. In one celebrated article, Campbell (1969) proposed that social reforms could be tested and improved in experimental fashion. More broadly, Campbell was concerned with questions of induction or how knowledge can accumulate and spread over time. His approach was inspired by Darwin’s theory of evolution. Campbell (1974) argued that the growth of knowledge requires many starts, most of which will turn out to be false, but those ideas that work, will be retained and allowed to propagate. There is nothing domain-specific about this theory. If it works, it should work in every corner of human activity where thinking occurs: from the selfish gene to the selfish thought. This includes creative thinking, i.e., the type of thinking that made Campbell famous.

In the closing section of Brutto Tempo, I briefly introduced Karl Duncker—he of candlestick-matchbox fame—and his Gestalt psychological take on creativity. The critical quote is worth repeating. “The decisive points in thought-processes, the moments of sudden comprehension, of the “Aha!,” of the new, are always at the same time moments in which [...] a sudden restructuring of the thought-material takes place, in which something ‘tips over.’” Duncker proposed that in order to earn an “Aha Erlebnis,” we have to work hard and systematically. The problems he used for study (e.g., the candlestick problem or the tumor-destruction problem, had a limited number of features, and a problem-solver could proceed by taking one feature at a time and manipulate it in their mind until a solution pops into place.


Campbell (1969) the experimentalist would agree. At the level of deliberate and overt experimentation in a finite dimensional space, a systematic process of ruling out non-solutions makes sense. Campbell (1960) dug deeper and asked how non-solutions are discarded at earlier stages of perception and cognition, when the number of variables and possibilities is still large. At these stages, he proposed, we find a Darwinian world of blind variation and selective retention. At the psychological levels beneath consciousness, possibilities can spring up in great numbers and at great speed. Templates or targets may also reside beneath consciousness, as well as the processes checking whether proposed solutions fit.

Imagine the target literally as concentric circles and the proposed solutions as randomly generated throws of darts. When a dart hits the bull’s eye, “something tips over,” and the tipping is the breach of the threshold into consciousness. As Campbell (p. 384) put it (and Poincaré (1913) had before him), “once the process has blindly stumbled into a thought trial that “fits” the selection criterion, accompanied by the ‘something clicked,’ ‘Eureka,’ or ‘aha-erlebnis.’” The subliminal struggle for consciousness is efficient; it shortcuts the kind of time and energy consuming exploration envisioned by Duncker, it leads “to intelligent behavior which lacks overt blind floundering, and is thus a knowledge process” (p. 384).

Once a target is hit, it can be replaced by another target of greater resolution. Imagine zooming in on the concentric circles until the bull’s eye becomes the entire target with another bull’s eye at its center, and the process repeating itself. In this way, inductive knowledge can accumulate with increasing refinement. If we left it at that, however, we could never have a paradigm shift or a radical Dunckeresque restructuring of the perceptual field and our understanding of the world. We could only build knowledge of greater precision; this would be fine in a stable environment, but it would be maladaptive in a changeable one. Therefore, the blind production of possibilities must include ideas or hypotheses that stray beyond the target set up as a filter and retention device. Writes Campbell (p. 395) “The greater the heterogeneity and volume of trials the greater the chance of a productive innovation.” In other words, greater variation in blind production guards against being forever committed to a path that could turn out to be mistaken after all.

The Darwin-Poincaré-Campbell theory of thought is radically anti-teleological. It explains how creative ideas can arise without having to assume that a future end reaches into the past to recruit the means it needs. At the same time, this theory explains why creative insights and other advances of knowledge often seem purposeful. If internal thought trials lack foresight, with only the fitting ones breaking the surface of consciousness, it must look to a perceiver residing in consciousness that she was actively looking for the thought which is now conscious. This view is starkly different from the resurgent Neo-Aristotelian view championed by proponents of prospection (Seligman et al., 2013). Campbell rejects the idea that insight is an explanatory concept. It is, instead, the result of processes that have no purpose individually. Indeed, how could insight be an explanation of insight anyway? To say that it can be begs the question of the nature of creative thought.

Campbell (p. 396) concludes that “like the theory of natural selection in organic evolution, [his theory of blind variation and selective retention] provides an understanding of marvelously purposive processes without the introduction of teleological metaphysics or of pseudo-causal processes working backward in time.” Yet, he remains mindful of the limitations of his theory. It says little (nothing really) about the nature of the engine that blindly churns out guesses or how the target is set up, moved, or replaced. Like Darwin’s theory of natural selection, Campbell’s theory of creative selection is haunted by the specter of tautology. How, for example, would you go about trying to falsify it?

Nonetheless, Campbell remains optimistic (p. 397) when he observes that “a creative solution is more likely the longer a problem is worked upon, the more variable the thought trials, the more people working on the problem independently, the more heterogeneous these people, the less time pressure, etc.” So let us purposefully apply ourselves to the task of allowing our subliminal minds to create creativity.

Campbell, D. T. (1960). Blind variation and selective retention in creative thought as in other knowledge processes. Psychological Review, 67, 380-400. doi: 10.1037/h0040373

Campbell, D. T. (1969). Reforms as experiments. American Psychologist, 24, 409-429. doi: 10.1037/h0027982

Campbell, D. T. (1974). Evolutionary Epistemology. In The philosophy of Karl R. Popper edited by P. A. Schilpp, 412-463. LaSalle, IL: Open Court.

Most Recent Posts from One Among Many

We need a folksy free will

Andrew Monroe responds to Free Will Depletion.

Grade Flation

Is grade inflation the heat death of the university?

Free Will Depletion

Should we settle for a folksy notion of free will?