The question is simple: is this dot purple or blue? That is what a team of researchers asked in a new study published in the journal Science. We all know what colors are – what could go wrong?
Well, at first, nothing. Participants – who had been vetted in advance for color blindness – were shown 1,000 dots ranging in color from "very blue" to "very purple". In the initial test, the researchers ensured that half the dots came from the "blue" part of the spectrum, and half were from the "purple" part, classified according to their RGB (Red, green, blue) values.
But scientists wouldn’t be scientists if they didn’t try to mess with our heads. So for some participants, the number of blue dots was reduced over time – from 50 percent of the first 200 dots down to a measly 6 percent of the last 650. But instead of reflecting this, the participants did something strange: they subconsciously changed their definition of "blue" so it would include more of the dots they were being shown.
It gets stranger. In a second study, the scientists specifically told the subjects that the number of blue dots would decrease over time. It didn’t help. In a third, subjects were explicitly instructed to be consistent in their color classification, and even offered a bonus of $10 for managing to do so. Nothing seemed to work: by the end of the study, they had still changed their definition of "blue" to include dots they had initially called purple.
Of course, these results have wider implications than just improving humanity's dot-color-analysis. The researchers went on to expand the study into more real-world situations, asking participants to view 800 computer-generated faces and decide whether they were a "threat" or "no threat". Again, they reduced the number of threatening faces over time – and, again, the participants reacted by changing their internal definitions to consider previously non-threatening faces as dangerous. In another trial, participants were asked whether hypothetical scientific studies were ethical or not – when the number of unethical proposals was decreased, they started rejecting studies they previously classified as ethical.
These results may explain one of the most enduring aspects of society: human pessimism. "Although modern societies have made extraordinary progress in solving a wide range of social problems, from poverty and illiteracy to violence and infant mortality, the majority of people believe that the world is getting worse," wrote the study authors. "The fact that concepts grow larger when their instances grow smaller may be one source of that pessimism."