In my office, Paul saw black and blue. I saw blue and gold. Our articled student saw gold and black and blue. Our support staff saw white and black. Everyone had a different answer.
I began to think about our subjective views of the colours of the dress, and how similar that was to impaired driving files. It is not uncommon that I have a client who is charged with drunk driving and there are multiple witnesses. Some officers say the client is stumbling. Others note no balance problems. Some police officer say the speech was slurred or the face was flushed. Others do not see these problems. Some witnesses say that the client was drunk after a collision, while others simply say the client seemed disoriented or shaken up.
Of course, there has been extensive discussion in the psychological and scientific communities on the fallibility and unreliability of eyewitness evidence.
Just as we all see something different when we look at The Dress, we also see things differently when we look at a person who is supposedly impaired. And our subjective perception can be influenced by what information we have been given in advance.
For example, if you were told that the dress is actually black and blue, you might look at the dress to see if you can see black and blue. Similarly, police officers responding to a report of a possible impaired driver are often looking at the driver to see if what they see matches their pre-determined notion of what the driver is doing. All of a sudden a lisp or an accent is slurred speech.
It’s human nature that when we are told something is a particular way, especially from a reliable source, that we expect to see it that way. And we will subconsciously craft our perceptions to match our expectations.
It’s a form of bias known as experimenter’s bias.
As defence lawyers, we know about this type of bias. We are on the lookout for it, and we try to structure our cross-examinations and our theory of the case around the possibility of bias where it exists.
The Dress is a good reminder that what we think we see may not always be reliable or accurate. And that this is also true for drinking and driving cases.