Apr 03 2014

Change Blindness and the Continuity Field

Change blindness is a fascinating phenomenon in which people do not notice even significant changes in an image they are viewing, as long as the change itself occurs out of view. Our visual processing is sensitive to changes that occur in view, but major changes to a scene can occur from one glance to the next without our noticing in many cases.

(See the color changing card trick for an example.)

One group of researchers believe they have a working hypothesis as to why our brains might have evolved in this way. Their idea is that the visual system will essentially merge images over a short period of time in order to preserve continuity – a process they call the continuity field. In essence our brains are sacrificing strict accuracy for perceived continuity.

This is in line with other evidence about how our brains work. Continuity seems to be a high priority, and our brains will happily fill in missing details, delete inconsistent details, and even completely fabricate information in order to preserve the illusion of a continuous and consistent narrative of reality.

Visual continuity is important because otherwise the world would appear jittery to us, constantly morphing as shadows play across an object, or our angle of view changes. This could be highly disruptive and distracting.

The researchers also point out that in the real world objects are fairly stable. They don’t pop in and out of existence, or morph into other objects. So not being perceptive to such changes would not be a big sacrifice and would not be likely to affect fitness. If something is actually moving or changing in our visual field we are very sensitive to that, and our attention will be drawn to it.

Neuroscientists, however, can contrive all sorts of impossible scenarios in order to probe our processing of sensory information. We did not evolve with video or photography, but researchers can use this technology to test how our brains process information.

They also give real world examples, such as the movies. There are often continuity errors in movies, missed by the vast majority of movie-goers.

To test their hypothesis of a continuity field, the researchers had subjects orient a white bar on a video screen to match the just-viewed black bars. They found that the subjects would orient the white bar in a position that was the average of the last several black bar images seen. It is as if their visual systems were averaging out about a 15 second window of similar images in a similar location. The effect was not present when the white bar was distant from the black bars on the screen, only when they overlapped.

Of course, it is difficult to extrapolate from such a specific study result to exact brain functions. This is very tangential evidence. But it was a test of a specific hypothesis, and the results do support the hypothesis, which I also think is plausible and in line with previous research.

While we can’t make too much of this one study, it is a consistent part of a larger picture that is emerging from neuroscience research. Our brains construct our perception of reality. They have algorithms that determine what sensory information to pay attention to, and essentially ignore the rest.

Further, the constructive process highly favors continuity over accuracy. In most everyday situations the sacrifice to accuracy will not be practically significant. However, at times the sacrifice will result in the misinterpretation of  our sensory experience.

Researchers can force this using optical illusions and other contrived experimental setups. But significant misinterpretations can also occur “in the wild” outside of a laboratory setting. In such cases people may have strange experiences they find difficult to explain. They are likely to then reach for culturally available explanations, such as ghosts or UFOs.

People who have such strange experiences may later proclaim that they know what they saw. Chances are, however, they don’t. They have a reconstructed memory of a constructed sensory experience that we know is highly flawed.

6 responses so far

6 Responses to “Change Blindness and the Continuity Field”

  1. SteveAon 03 Apr 2014 at 11:48 am

    I watched the video in the link, but I’m not sure what was going on.

    We were shown two images: ‘white/black bars’ then ‘squiggles’, and asked to align the white bar to them. It looked pretty easy to align the white bar with the ‘white/black bars’ but the ‘squiggles’ looked too random to hold any useful information. I ended up ignoring them. As, apparently, did the person doing the aligning.

    What was the purpose of the ‘squiggle’ pictures? How could anyone use them to align the white bar? Were they there to act as neutral break between ‘white/black bars’?

  2. BillyJoe7on 03 Apr 2014 at 4:15 pm


    -> BLACK LINES -> sqiggles -> BLACK LINES -> sqiggles -> BLACK LINES ->

    The sqiggles act as the discontinuity between the black lines.
    The white line is the subject’s attempts at maintaining continuity.

  3. etatroon 03 Apr 2014 at 5:01 pm

    I wonder if this is one of the reasons it’s so difficult to proofread one’s own writing. Even the smartest person can re-read the same text over and over again and miss a a spelling error or duplicated word.

  4. YtterbiJumon 03 Apr 2014 at 5:59 pm

    The Youtube video linked from that press release seemed to show the exact opposite from what the study conclusions were. The subject was quite accurately matching the angle of the white bar to the angle of the previous black bar, at least to the accuracy of human eyeball angle measurements. There was no “averaging” over several previous sets of black bars.

    Am I missing something here? Admittedly, there is no link to the actual paper and very little explanation in the press release, rather a lot of ridiculousness about movie continuity errors.

  5. SteveAon 04 Apr 2014 at 6:00 am

    BillyJoe7: Thanks. So the squiggles act as a kind of ‘reset’. I understood the purpose of the white line.

    YtterbiJum. Same here. It looked like the test subject was doing a pretty good job of aligning the white bar. Or was that some kind of optical illlusion revealed by subsequent analysis?

  6. BillyJoe7on 04 Apr 2014 at 6:04 am


    When the last black line is oriented in a direction close to the orientation of the previous two black lines, the subject orients the white line closer to that of the previous two. For example, if the previous two black lines are both orientated clockwise of the last black line, the subject will tend to orient the white line slightly clockwise of the last black line. But, if the last black line is oriented in a direction very different from the orientation of the previous two black lines, the subject orients the white line to be more accurately in line with that of the last black line.

Trackback URI | Comments RSS

Leave a Reply

You must be logged in to post a comment.