One popular conception of memory is sometimes dubbed the “video-recorder view.

One popular conception of memory is sometimes dubbed the “video-recorder view.” According to this commonsense view, everything in front of your eyes gets recorded into memory, much as a video camera records everything in front of the lens. This view seems to be widely held, but it is simply wrong. As Chapter 6 discusses, information gets established in memory only if you pay attention to it and think about it in some fashion. Mere exposure is not enough.

Wrong or not, the video-recorder view influences how many people think about memory—including eyewitness memory. For example, many people believe that it is possible to hypnotize an eyewitness and then “return” the (hypnotized) witness to the scene of the crime. The idea is that the witness will then be able to recall minute details—the exact words spoken, precisely how things appeared, and more. All of this would make sense if memory were like a video recorder. In that case, hypnosis would be akin to rewinding the tape and playing it again, with the prospect of noting things on the “playback” that had been overlooked during the initial event. However, none of this is correct. There is no evidence that hypnosis improves memory. Details that were overlooked the first time are simply not recorded in memory, and neither hypnosis nor any other technique can bring those overlooked details back.

Similarly, consider a technique often used in trials to raise questions about an eyewitness’s recall. The attorney will ask the eyewitness question after question about an event and will quickly discover that the witness recalls some facts but not others. Thus (for example) the witness might remember the suspect’s words but not his clothing, or might recall the clothing but be uncertain where people in the crime scene were standing. Later, the attorney is likely to ask the jury, “How can we trust this witness’s recall? With so many gaps in the witness’s recollection, it’s clear that the witness was not paying attention or has a poor memory. We therefore cannot rely on what the witness says!”

However, the attorney’s argument assumes the video-recorder view. If memory were like a video recorder, then we should worry if the playback has gaps or errors in it. (If your actual video recorder, or your DVD player, has gaps in its playback, skipping every other second or missing half the image, that surely does sound like a malfunction.) But this is not how memory works. Instead, memory is selective. People recall what they paid attention to, and if a witness cannot recall some aspects of a crime, this merely tells us that the witness wasn’t paying attention to everything. And, given what we know about the limits of attention, the witness couldn’t pay attention to everything. Hence, it is inevitable that the witness’s memory will be incomplete (i.e., not recording every single detail within a scene), and we certainly should not use that incompleteness as a basis for distrusting the witness.

Related examples are easy to find, with many popular ideas about memory resting on the (incorrect) video-recorder view. Once we understand how memory works, therefore, we gain more realistic expectations about what eyewitnesses will or will not be able to remember. With these realistic expectations, we are in a much better position to evaluate and understand eyewitness evidence.

1. In J. K. Rowling’s Harry Potter books, some of the characters relive their memories, or the memories of others, using a device called the pensieve. How is the “videorecorder” view of memory represented by the pensieve problematic?
2. How can we determine whether to believe the testimony of a witness in court? What kinds of information is a witness likely to remember correctly (and incorrectly)?
3. Many people have experienced significant events in their lives for which they seem to have excellent memories, recalling them in minute detail. Why might some events promote more accurate memories? Alternatively, why might such memories be believed to be highly accurate even when they are not?
4. What do you think is the reason the spacing effect works? How about the reason the testing effect works?