I'm sure most Cognitive Daily readers are aware of the massive debate permeating the scientific world these days. No, not evolution versus creationism; I'm talking about object- versus space-based attention.
Haven't heard of this raging debate? Well, then, let me refer you to a fascinating pair of experiments conducted by Massimo Turatto, Veronica Mazza, and Carlo Umiltà. The issue comes down to this: a critical problem for psychologists is the issue of attention. How does the perceptual system decide where to focus our attention when literally millions of bits of information are available to us at a given time? The object-based school believed that attention was focused on objects: when we notice a change to one part of an object, we're more attentive to other changes in that object (e.g. when the car in front of us puts its turn signal on, we're more likely to notice it slowing down for a turn). The space-based school believed that attention was focused on particular regions: we're more likely to notice a change near where we've seen another change, whether or not the changes occur on the same object (e.g. if we see the neighbor's dog running across their backyard, we're more likely to notice the squirrel she's chasing).
But the attention problem isn't limited to things we see. Our sense of hearing, touch, and even smell can inform what we pay attention to -- for example, the turn of the key in the door alerting us to a loved one's arrival. Turatto's team realized that by focusing on hearing, they could address the object versus space debate.
They set up an apparatus to combine visual and auditory inputs at the same time: a large black cardboard box concealing four loudspeakers, all in a dark room. The speakers were arranged in a square 90 centimeters on a side. In front of each speaker were three LEDs: red, green, and orange. Participants sat in looking toward the loudspeakers and were told that their task was to respond as quickly as possible when they saw a red or green LED flash, pressing the B key on a computer keyboard when they saw a red light, or the N key when they saw a green light.
During the experiment, the two speakers on the right emitted an intermittent high-pitched sound, and the speakers on the left produced a lower pitch of the same sound (I've approximated how this sounds here). At a predetermined time (the 7th repetition of the low-pitched sound), an orange LED flashed in front of one of the two speakers used to play that sound. Then, synchronized with the ninth repetition of one of the two sounds, either a red or a green LED flashed, signaling participants to press the corresponding button on the keyboard. Here are the results:
As you might expect, the fastest reaction times occured when the red or green LED flashed at the same spot, with the same sound where the orange LED flashed. The second fastest reaction was when the LED flashed with the same sound, but in a different spot. The slowest reactions occcured when neither the sound nor the spot was the same.
So how does this result impact the raging debate about object- versus space-based attention? If attention were only space-based, then only the when the red or green LED flashed in the same spot where the orange LED had appeared would reaction times be faster. Turatto et al. argue that the different tones constitute "sound objects," analogous to physical objects you can see. Thus, their experiment supports the idea of object-based attention. But the team also believed space-based attention had some merit. To support this argument, they conducted a second experiment, where identical tones were played from each speaker -- but still alternating from left to right, as before. Here are the results of that experiment:
Now only when the orange and red or green LEDs flashed in exactly the same spot did faster reaction times occur. Since the sounds coming from the speakers on the left and right were the same, they did not constitute separate sound objects for listeners, so instead participants attended only to the location of the LEDs. Thus, support for both object- and space-based attention was found in an elegant pair of experiments.
So the next time some object-based attention zealot claims that only object-based or space-based attention should be taught in the schools, you now have the knowledge to retort that both object and space-based attention have merit. Can't we all just get along?
Turatto, M., Mazza, V., & Umiltà, C. (2005). Crossmodal object-based attention: Auditory objects affect visual processing. Cognition, 96, B55-B64.
Unless I'm misunderstanding something, it seems like if you're going to claim this isn't simply a priming effect, you have to explain a lot of data which shows that object-based attention _doesn't_ happen. For example, Posner showed in 1980 that cueing letter identity, letter location, or both resulted in benefits only for location cueing, not object cueing. Theeuwes found something similar in 1989 in which form cues were of no help. Of course there's other work showing that attention is facilitated when directed towards objects but the spacial direction of attention is clearly important (and is to my mind moreso).
One nice demonstration of how object features might facilitate attentional processing, without actually requiring that attention can be directed on the basis of non-spatial information alone is a recent study from Richard Ivry. This study showed advantages for detecting a target object differing only in shade of color from distractor objects if the color had a different name, but only when the object was presented to the right hemifield. This leads to processing by the right hemisphere, which is language dominant; our ability to perceive things of different colors is supported by linguistic representations, which are stronger in the left hemisphere. You can see an example of the stimuli on my blog, here: http://develintel.blogspot.com/2005/12/synaesthesia-part-ii-language-co…
But I guess these kinds of disagreements are exactly why it's a raging debate :)
an interesting take on that article, but in my opinion you were a bit glib about a few things.
first of all, to characterize the object vs. spatial based attention debate in cognition/perception as "raging" may be a bit much. I understand that you were trying to use that idea as a frame for explaining the core point of the findings, but as a grad student in visual cognition who works with some of the people involved in this particular field, there's certainly no recent set of disparate findings that one side uses against each other - at least nothing close to an evolution vs. creationism kind of debate. Fact is, in psychology we're used to both sides being mostly right (this happens in all areas of psych), the really only interesting question is which perspective has more of an influence than the other on the final result.
Secondly, I'd point out that Turatto et al's interpretation of the second experiment could have been motivated by reviewer comments/suggestions rather than their own rethinking of the problem. Certainly, I have no idea which explanation is true, but frequently the experiment that muddles the theoretical clarity of a paper is the result of someone from the "other side" pointing out a flaw in their methodology.
As for characterizing this paper as "elegant," I'm afraid one could argue just as easily that this paper was just one in scores of others that show that the object vs. spatial attention debate isn't very easy to resolve. While they had some object based attention effects, they were only about 20 milliseconds difference across 90 cm. When you consider that their spatial attention effects were roughly comparable: 60 milliseconds across almost a meter of space, it is also hardly compelling to argue that attention travels spatially. Also, the authors of the paper don't specifically point out that they were using a specialized keyboard to collect response data - it is worth noting that regular keyboards only sample at about 60hz (every 16 ms), which affects the temporal resolution of your data & could inflate differences in RT. They also don't explain why their exp 2 same side, same spot data is 20 ms slower than their exp 1 data. (which makes the keyboard refresh rate argument a bit more plausible). Their main point, however, that you can observe within objects effects for auditory objects, is quite interesting.
I don't want to come off as too obnoxious or snotty - I'm glad there are other people besides the dorky guys & gals I work with who find attention research interesting! I just figured there were some esoteric facts that I knew (e.g. the temporal resolution of keyboards) that could help someone form a more accurate opinion of these findings. Hell, I was mildly surprised that I knew some facts about this research that could be considered helpful to anyone in the first place. ;)
Ben and Chris,
Thanks for the comments and information. I agree with you that "raging debate" may be a bit of an overstatement, but I did want to show readers who may not be as well-versed in science as you what a real scientific conflict is (as compared to the evolution-vs.-creationism nondebate).
Didn't Brian Scholl settle this in 2001? :-)
... or maybe that's just at Yale ...
Completely off-topic, but Ben, are you at UIUC (I googled you quickly out of professional interest)? I'm at Marvin Chun's lab at Yale and am running some rapid resumption experiments right now...
hey tim, yeah. I'm in dave irwin's lab.
Perhaps a better argument for space-based attention is found in cuing experiments (e.g. Egly et al, 1994). The typical finding is that when a location is cued within an object, responses are faster to different locations within that same object than to equidistant locations outside the object. My lab actually did some work on the distinction between space-based and object-based attention when I was in grad school at UGA. There's evidence for both space-based and object-based attentional mechanisms; I really don't see why it would be considered a "raging debate."
Please tell Greta her flunkie says hello.
Hi Jenny! My #1 flunkie should email me an update about her life ;)
Who knew space vs object based attention would attract so many personal notes?