02 / 07
Why Does Stroop Interference Happen? — Processing Speed, Automaticity and Selective Attention
Once you see a word, ignoring its meaning is hard. The theories that explain Stroop interference — relative speed of processing, automaticity, parallel distributed processing and conflict monitoring — and their limitations.
Ignoring the meaning of a word you are looking at is harder than it sounds. Theories of Stroop interference fall into four broad families. None has won outright; each explains a different side of the phenomenon well.
1. Relative speed of processing
This account starts from an old observation. Cattell (1886) reported that people read words faster than they name colors. The speed-of-processing account says the two kinds of information “race” to the response stage. Because the word information arrives first, naming the color means suppressing a word response that is already there, and that extra time is the interference. Color arrives later, so it barely interferes with reading words. This fits the asymmetry in the original paper well.
The limitations are clear, however. Dunbar & MacLeod (1984) showed that even when letters were rotated or inverted so that reading the word became slower than naming the color, interference remained. Speed alone is not enough to explain it.
2. Automaticity
For skilled readers, reading is an automatic process that happens even without intention (Posner & Snyder, 1975; Shiffrin & Schneider, 1977). Naming a color is a controlled process that requires attention, so the word's meaning, which pops up on its own, intrudes and causes interference. This fits the fact that interference grows as children learn to read, and that it is small for words in an unfamiliar foreign language.
Reading is not “fully” automatic, though. Besner, Stolz & Boutilier (1997) reported that coloring just one letter instead of the whole word greatly reduces interference. This is evidence that automatic processing can weaken depending on where you focus your attention.
3. Parallel distributed processing model
Cohen, Dunbar & McClelland (1990) modeled the word pathway and the color pathway as a neural network. The heavily practiced word pathway has strong connections, while the color pathway is weak. The task goal “name the color” boosts the color pathway, but the activation of the strong word pathway never fully switches off, so interference remains. Within one framework, this model explains the asymmetry, practice effects, and the idea that automaticity is a matter of degree rather than all-or-nothing.
4. Selective attention and conflict monitoring
More recent research treats interference as a problem of cognitive control. A leading example is conflict monitoring theory: the brain detects a “conflict” between two competing responses and responds by maintaining the task goal and suppressing distracting information (Botvinick et al., 2001). Brain imaging studies have reported that the anterior cingulate cortex is involved in detecting conflict and the dorsolateral prefrontal cortex in maintaining goals and exerting control (MacDonald et al., 2000).
This view explains trial-order effects well. Interference tends to be smaller right after an incongruent trial (Gratton, Coles & Donchin, 1992), interpreted as the conflict you just experienced ramping up control. The proportion congruency effect is also well known: the more congruent trials there are overall, the larger the interference (Logan & Zbrodoff, 1979, among others).
At which stage does interference arise?
Whether interference arises at the semantic stage, where the word's meaning is processed, or at the response stage, where you choose which button to press, is another long-standing question. Today many researchers think it arises at both. Klein (1964) reported that words related to color, even if they are not color words (words like “fire” or “sky”), cause weak interference, while words unrelated to color cause even less. The closer the meaning is to a color, the stronger the interference.
This is also why the neutral condition on this site uses meaningless strings (XXXX). If a neutral stimulus carried even a slight color-related meaning, the comparison baseline would be skewed.
References
Cattell, J. M. (1886). The time it takes to see and name objects. Mind, 11(41), 63–65.
Dunbar, K., & MacLeod, C. M. (1984). A horse race of a different color: Stroop interference patterns with transformed words. Journal of Experimental Psychology: Human Perception and Performance, 10(5), 622–639.
Besner, D., Stolz, J. A., & Boutilier, C. (1997). The Stroop effect and the myth of automaticity. Psychonomic Bulletin & Review, 4(2), 221–225.
Cohen, J. D., Dunbar, K., & McClelland, J. L. (1990). On the control of automatic processes: A parallel distributed processing account of the Stroop effect. Psychological Review, 97(3), 332–361.
Botvinick, M. M., Braver, T. S., Barch, D. M., Carter, C. S., & Cohen, J. D. (2001). Conflict monitoring and cognitive control. Psychological Review, 108(3), 624–652.
MacDonald, A. W., Cohen, J. D., Stenger, V. A., & Carter, C. S. (2000). Dissociating the role of the dorsolateral prefrontal and anterior cingulate cortex in cognitive control. Science, 288(5472), 1835–1838.
Gratton, G., Coles, M. G. H., & Donchin, E. (1992). Optimizing the use of information: Strategic control of activation of responses. Journal of Experimental Psychology: General, 121(4), 480–506.
Klein, G. S. (1964). Semantic power measured through the interference of words with color-naming. The American Journal of Psychology, 77(4), 576–588.