When I think about the verbal overshadowing phenomenon I wonder why we, as sketch artists, are able to create sketches of suspects with remarkable likenesses. If the act of verbalizing the description of the face causes issues with recognition later in the photo lineup, why hasn't the practice of the sketch interview been scrutinized?

The psychologists working on these theories bring up some interesting ideas and make some rather valid points about the processing of information. As you all know, I do not use reference images to stimulate memory recall for the face of the suspect. Rather I use calming techniques and ask unbiased questions while sketching the face of the suspect. Once I get the face of the suspect on the page I inquire about the incident to get more details about their focus and duration of the event. I use this information to refine the details of the sketch. Finally, I show the sketch to the eyewitness and we collaborate on the features.

In the first part of the interview the eyewitness is in a visualization mode as they go about accessing the information about the face in response to my prompts. In the second part of the interview the eyewitness will compare their recollection of the suspect to their recognition of the face I have drawn. Looking at the sketch in its entirety, and not feature-based, will keep them in the mode of comparing the face of the suspect to the sketch. The sketch, at this point, should be very similar to their recollection of the suspect and the changes should be minimal. When I ask the eyewitness about the hair, I am asking if the hair reminds them of the suspect's. I'm looking for a yes or a no, with some comments on why. I will make the changes, if needed, and move on to the next feature.

This refinement process can be very complex. Going from the visual (looking at the sketch) to the verbal processing (describing what needs to be changed) is an interesting issue I never found to be difficult. In the paper: Why do words hurt? (pdf) Content, process, and criterion shift accounts of verbal overshadowing, Jason M. Chin and Jonathon Schooler, University of British Columbia, Vancouver, BC, Canada, they talk about how some individuals would be able to perceive and process various nuances of a stimulus, but would not be able to put these perceptions into words. Therefore, subsequent verbalization would be especially unrepresentative of the perceptual experience. To me, this means that when we interview the eyewitness, many of them lack the words to describe exactly what they remember about the face of the suspect. Although they can see the face in their minds, many of them are unable to articulate every nuance. Because of this phenomenon, when they do describe the suspect's face their comments may be less than veridical. Psychologists suggest that because of this disconnect, the reliability of the eyewitness when selecting the photo of the alleged suspect may be suspect.

When you read the article (it is very interesting) you'll be fascinated by the information about processing of visual and verbal information. Interestingly enough, the study focuses on how we see faces and are able to remember them after a few minutes. The experiments usually were comprised of viewing a picture and then five minutes later the witness was asked to either describe the person or pick the person from an array pf photos. Sadly, none of the reports I have read incorporate the collaboration with the sketch artist to create the sketch or to test the verbal overshadowing effects on the various sketch interview techniques. When you read this paper, ask yourself: if we can't ask the eyewitness what the suspect looked like because it will jeopardize their ability to pick them out from a lineup---how can we identify the suspect?

Read the article that started it all: Verbal Overshadowing of Visual Memories: Some Things are Better Left Unsaid. (Schooler & Schooler, 1990)

What do you think about this? Email me!