Bottom-Up & Top-Down in Perception, Behavior & History

Which perspective to choose?
  • Studies in the psychology of perception had established by the 50’s something interesting about perception: perception is influenced by non-perceptual factors, such as needs, motives, expectations, and beliefs.
  • This was eventually attributed to two nearly opposite kinds of processing of perceptual data, bottom-up and top-down (Hunt, The Story of Psychology, 2007, pp. 520-522).
  • Bottom-up processing, more nearly physiological, combines the lowest-level sense-data (e.g., pixels of visual data) into higher-level data (e.g., perceived objects that we recognize, react to, and may manipulate).
  • Top-down processing, more cognitive in nature, works from higher-level schematics, actively looking for matches within the lower-level data.
  • Example #1. Bruner and Postman (1949) showed 28 subjects brief glimpses of playing cards, mostly standard, but with a few not (e.g., a black eight of hearts). 27 of the 28 subjects saw only normal cards, but once the subjects were told about the abnormal cards, mistaken identifications dropped dramatically.
  • Example #2. Else Frenkel-Brunswik (in Bruner and Krech, 1968 [1949]:128-129) first ranked through testing a sample of children on a scale measuring ethnic prejudice, which she used as a proxy for “authoritarian personality pattern”. She then showed two sets of of gradually changing visual stimuli, pictures of a dog that gradually became a pictures of a cat, and cards in which one hue gradually became another. She found that higher “authoritarian personality” rankings correlated to greater persistence of the perception of of the original object or hue.
  • In short, expectations and prior commitment to schema influenced actual perception, yet not to the absolute exclusion of changes in the lower-level stimuli.
  • So, while, repeated perceptions of a new object-stimulus can lead to the recognition of a generalized category, once that category is either constructed from experience or introduced by culture, it starts to shape how the lower-level data are organized and categorized.
  • Now, I want to point out a parallel in the world of human agency.
  • When we approach a new task involving new skills, we initially fumble about until we begin to develop simple habits that seem to bring competence in performing the task within reach.
  • Habits partially automate simpler skills, while allowing us to focus on the next level of mastery. But precisely because we no longer focus on them, taking them for granted, they become top-down organizing principles that we find it difficult to break.
  • In other words, practical life offers the same division between bottom-up and top-down organization.
  • Ideally, these two processes exist in a feedback loop; we blunder about from the bottom-up, until we discover a schematic habit that works. We then rely on that, ceding some control to the habit, so that we can focus on the next level of mastery. But we must remain open to revising the habits that got us to where we are, when we notice that they are sub-optimal or even counter-productive.
  • Those of you acquainted with theory of knowledge will recognize this as a micro, practical parallel to the Kuhnian paradigm, a guiding mindset that is necessary to process information of any kind, yet which can also become a limit and a trap — a box that we can’t think outside of.
  • But there is a second parallel in human agency. For habit, as Aristotle knew, is the basis of character and ethics, the study of good character and good practice. (Our word ethics comes from the Greek ethos, whose root meaning was habit.)
  • For habits not only guide information processing, they also guide habitual choices. They form and develop personality and character, the social characteristics by which we operate within the society of agents.
  • Coming back full circle, consider how character is formed by a feedback loop between bottom-up and top-down choosing.
  • We “feel our way through” novel experiences and situations, developing habits and practices if they are significant or recurring experiences.
  • Practical habits, of both the instrumental and ethical kinds, are considered so important by human societies (including parents) that they are left to chance only by the most careless and ineffective societies.
  • Yet societies develop their cultural habits much as do individuals, initially bottom-up and accidental, later top-down relying on (and enforcing) cultural habits that must be, on balance, useful, if that culture is to survive in the competition for resources among cultures.
  • Yet all cultures will retain some practices that are not optimal (not so different from maladaptive genes in the gene pool). So, the feedback loop between traditional top-down processing and bottom-up fresh processing is crucial.
  • Those societies that use this ongoing feedback loop more effectively will prosper better over the long term. Both processes are essential. It’s the balance that counts.