10 SMIDSY – looked but not perceived; semantic richness

SUMMARY – our visual environment is cluttered with objects… our visual attention tends to be drawn towards the objects with which we have most connection… motorcyclists tend to spot other motorcycles… but other road users with no interest in motorcycles are less likely to see them…


The previous three sections have discussed aspects of cognitive conspicuity; how the ‘prevalence effect’ predisposes us to detect objects that are more common than other objects which don’t appear so regularly, how the complexity of the task spreads our attention, and how focusing on a particularly difficult job can blind us to other, necessary tasks. But there is a yet more subtle problem. Humans, like other animals, need to be able to detect objects of interest in that visually cluttered environment. We could be looking for potential prey, possible predators, or even mates. What is that helps us make the right decision on how to focus our attention?

Semantic meaning

Back in the introduction, I introduced the concept of ‘visual salience’, whereby some objects have a distinct perceptual quality which draws our eyes and grabs our attention. Only at that point we become aware of whatever the visual system has detected. I pointed out that this is the fundamental concept underpinning the conspicuity aids proposed for use by motorcyclists. In a paper published in the journal ‘Nature’ in 2017, Professor Henderson and fellow postdoctoral researcher Taylor Hayes call this visual salience explanation the ‘magpie theory’, because it often assumes that our attention is drawn, just like a magpie, to bright objects. However, it may not be correct.

Rather than being drawn to those parts of a scene that are salient and ‘stick out’, it may be that our visual attention is drawn to parts of a scene that have strong meaning or ‘semantic richness’, because humans are primed towards identifying objects which connect with us in some way. Henderson and Hayes set up a study which:

“…contrasted the degree to which the spatial distributions of meaning and salience predict viewers’ overt attention within scenes. The results showed that both meaning and salience predicted the distribution of attention, but that when the relationship between meaning and salience was controlled, only meaning accounted for unique variance in attention. This pattern of results was apparent from the very earliest time-point in scene viewing.”

They accounted for the results in terms of the observer’s field of interest:

“When viewing a photograph or scene observers most commonly fixate on the most physically salient regions first, but if the observers have relevant content expertise (for example, a history student viewing a photograph of artefacts from the US Civil War) then they focus less on physically salient areas and more on semantically meaningful areas.”

Semantic richness would explain something I have previously written about as the ‘Classic Car Effect’. Start to drive something a bit unusual – an original 2CV perhaps – and suddenly old 2CVs will jump out of the scenery in a way we have never experienced before, despite being surrounded by other, considerably larger, vehicles. That’s not because the 2CVs have suddenly become more visible (visual salience) or that there are a lot more of them around (the prevalence effect), it’s because our new purchase has subconsciously primed us to be more aware of the 2CV – the classic Citroen suddenly holds semantic meaning.

red-89219_960_720

Humphrey and Underwood (2009) found that when inspecting a scene that was highly relevant to the subject – for example, engineers viewing an engineering scene – specialists looked at semantically-interesting features that were in line with their speciality, rather than at more-highly salient but non-specialist features. They found their eye movements responded more closely to the areas of high meaning which related to interest and less to those with high saliency. They went on to construct an experiment which compared car drivers to road users with a specialist interest in motorcycles, by tracking the eye movements of volunteers as they looked at scenes which had been constructed to have both meaning and salience. Not surprisingly, the same holds true for motorcyclists – we tend to look at other motorcyclists.  Humphrey and Underwood stated:

“This conclusion has important implications for current theories of attention across diverse disciplines that have been influenced by image salience theory”.

You will also recall the work done by Lenné et al (2013) which I mentioned earlier, where they investigated how subjects responded to streams of vehicles seeded with buses or motorcycles. They also suggested that the way we detect an object depends on context. By increasing the numbers, they gave those buses or bikes added semantic meaning. The researchers suggested that our prior experience changes the way we search a scene.

These studies reflect an alternative theory of ‘cognitive guidance’, where our attention is directed towards parts of a scene that our prior knowledge tells us are likely to hold meaning. And if we are drawn to areas of meaning within a scene, this makes sense of a body of research which suggests that motorcycle license holders who also drive (so-called ‘dual drivers’) are better at detecting motorcycles when behind the wheel than drivers without experience of riding bikes. Dual drivers eyes are drawn not just to the motorcycles themselves, but to the areas they expect to see motorcycles because they are the ‘semantically meaningful areas’. And so semantic meaning is likely to affect our response. Underwood et al (2011) found:

“It is possible that the expertise and familiarity riders had with motorcycles affected the way they processed the scene. When riders fixated a motorcycle (especially a high saliency one), it was semantically meaningful to them, and thus they processed it as a hazard, decreasing the likelihood of a ‘‘safe-to-pull-out’’ response. However, when drivers fixated a motorcycle, they failed to process it as a hazard, due to their lack of expertise, and thus were more likely to declare a ‘‘safe’’ response, when it was not necessarily safe to pull out (an LBFTS error).”

They further found that:

“When a vehicle was clearly visible (high saliency), riders were quicker to respond (giving a more cautious ‘‘unsafe’’ decision), whereas with low saliency vehicles, riders took longer to make a decision.”

Two quick observations. Firstly, being visible is not the same as being conspicuous. Something can be conspicuous and at the same time not visible, or vice versa. The two terms are not interchangable. I assume that ALL the motorcycles in the study were visible. If not, then it’s hardly a test of the effect of salience.

In any case, the study appears to show that riders saw both the high AND low salience vehicles. However, I think the researchers have totally misinterpreted the meaning of the response time. Because the dual drivers took LONGER to make a decision having seen a low salience vehicle, I would suggest that provided they were not actually committing a ROWV, the longer pause whilst considering their decision means that they were MORE more cautious, not less as implied in the study. I would argue that a QUICKER decision is less cautious.

road-construction-192893_960_720I would further suggest that semantic meaning is also at least a partial explanation for anecdotal claims that motorcyclists (and road safety proponents) who promote the use of conspicuity aids by saying that they find it easier to see other motorcyclists wearing hi-vis clothing and using day-riding lights.

It would be an interesting study to see if motorcyclists who claim to see hi-vis and DRLs more readily also responded more readily to conspicuity aids used by, say, roadside workers. If they didn’t, it would cast doubt on the visual salience theory and support the role of semantic meaning.

In any case, it should be clear there is a danger in using our own perceptions of other motorcyclists to make the intellectual leap which concludes that “other road users will also see motorcyclists using conspicuity aids more more easily”. As other road users do not find semantic meaning in motorcycles or conspicuity aids, any visual salience benefits are likely to be more limited than commonly believed.

The ‘cognitive guidance’ theory suggests that finding meaning within our visual world is, to quote Henderson, “the driving force guiding attention through real-world scenes”. Motorcyclists are much better at spotting other riders at least partly because we are actually interested in motorcycles. And that in turn would explain why ‘Think Bike’ campaigns don’t seem to have lasting effects. However hard we try, to a driver with zero interest, motorcycles will never leap out of the background in the same way that they do to another motorcyclist.

Since you’re here, I’ve a small favour to ask. If you feel able to make a small donation to the upkeep and continued development of SOBS, why not buy me a coffee? Each contribution is much appreciated. Each cuppa keeps me awake and writing! Thank you.

Ko-fi_Red

Next >

 

References:

Henderson, J., M. and Hayes, T., R. (2017) “Meaning-based guidance of attention in scenes as revealed by meaning maps” Nature Human Behavior Vol 1 Oct 2017 743-747

Lenné, Michael G.; Salmon, Paul M.; Beanland, Vanessa; Walker, Guy H.; Underwood, Geoff; and Filtness, Ashleigh. (2013) “Interactions between Cars and Motorcycles: Testing Underlying Concepts through Integration of On-Road and Simulator Studies.” In: Proceedings of the Seventh International Driving Symposium on Human Factors in Driver Assessment, Training and Vehicle Design

Underwood, G., Humphrey, K., van Loon, E. (2011) “Decisions about objects in real-world scenes are influenced by visual saliency before and during their inspection” School of Psychology, University of Nottingham, Nottingham, UK Vision Research 51 (2011) 2031-2038

https://phys.org/news/2018-01-inattentional-drivers-motorcycles-plain-sight.html retrieved April 2018

 

Last updated: 

Friday 23 November 2018 – minor edit for clarity

 

IMPORTANT:

The material is free to all to access and use under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License. That means you can share it with your family and friends, and re-use it for club magazines and websites, so long as you acknowledge the source and author and include the same Creative Commons license in the derived works.

Please note, this Creative Commons license excludes commercial use. If you wish to use any of my work for commercial purposes, including (but not limited to) articles in pay-for magazines or commercial websites, please contact me.

Creative Commons statement

Kevin Williams has asserted his right under the Copyright, Designs and Patents Act, 1988, to be identified as the author of this work.

Photo credit Paul Townsend https://www.flickr.com/photos/brizzlebornandbred/20001313491

One thought on “10 SMIDSY – looked but not perceived; semantic richness

Comments are closed.