Readers of the Brains blog might know about a symposium there concerning a paper by Philipp Koralus. In his commentary on the paper, Felipe de Brigard mentions the problem of captured attention: 

"I have a hard time understanding how ETA may account for involuntary attention. Suppose you are focused on your task—reading a book at the library, say—and you hear a ‘bang’ behind you. A natural way of describing the event is to say that one’s attention has been involuntarily captured by the sound. Now, how does ETA explain this phenomenon?"

Koralus' response is binary: 

"So, you might have been asking, as part of your task of reading the blog, 'What does the blog say?' Now, you are getting the incongruent and irrelevant answer 'There’s a loud noise behind you.' There are now two possibilities, similar to what happens in the equivalent case in a conversation. One possibility is that you accommodate the answer, adopting a new question (and thereby a new task) to which 'There’s a loud noise behind you' would be a congruent answer, maybe, 'what sort of thing going on behind me?…You could also refuse to be distracted and then exercise some top-down control on your focus assignment to bring it back to something that’s relevant to your task.'

When I coined "the problem of captured attention" in my 2012 Synthese paper, "The Subject of Attention" (not cited by Koralus/de Brigard), I took a similar line, but focused on the activity of the subject, rather than on questions and answers:


"Husserl would likely say of the above examples of the large text, the colorful advertisement, and the annoying conversation that the subject is active in allowing him or herself to be distracted; the large text, the colorful advertisement, and the annoying conversation could be actively suppressed by the subject with effort, but are instead allowed to disrupt the subject’s thoughts because the subject 'turns toward' the distracting stimuli in willfully allowing them to intrude. A sign that this turning toward is active in the case of captured attention is that there are instances where the same stimulus does not serve as a distraction—where the large text, the colorful advertisement, and the annoying conversation do not manage to disrupt one’s stream of thought because of greater attention to that stream of thought. It is a comparison with these instances that warrants the claim that the subject actively allows the capture of attention to occur when the subject finds him or herself distracted."

This Husserl-inspired position of mine on the problem of captured attention is not well-loved. Many seem to have the intuition that there is nothing one could do about captured attention, not just that the force of the capturing stimulus is relatively strong. For example, at the 2012 Eastern APA meeting Steven James commented on my paper as follows: 

"The weaker claim that 'attention is a process of mental selection that is within control of the subject' identifies a purported (and very plausible) property of attention–whatever it is, attention is something that is at least sometimes subject to agential control. Of course, this is consistent with some instances of attention being mere occurrences, as for example, when one finds her attention grabbed by the sound of her own name spoken in an unattended conversation. In such cases, the subject does not initially direct the selection of a target stimulus at all but rather simply finds her focus drawn to it."

I wrote on the handout for that talk the following response:

"The fact that attention is sometimes felt as relatively passive, as being drawn rather than directed, is not evidence, I claim, that the subject failed to direct attention to the target. There is a difference in kind between reflexive and non-reflexive response that does not meet that between active and passive attention. That is, all attention involves some goal on the side of the subject and some fulfillment on the side of the object, and these can come in different degrees."

In Q & A at that talk I went on to discuss some empirical work on this topic. Numerous papers discuss the ways in which seemingly captured attention relies on intention, task, and working memory, which I take to support my claim that attention can be relatively passive or active, but that even so-called "captured attention" is not captured against the intentions of the subject (e.g. here and here).

I would like to revisit this topic and wonder whether readers of NewAPPS have examples that they think count for and/or against my thesis on the problem of captured attention. (Feel free to weigh in without reading my paper or the papers linked to above.)

 

Posted in , ,

14 responses to “The Problem of Captured Attention”

  1. Sara L. Uckelman Avatar

    I don’t know if he discusses captured attention specifically, but Tikitu de Jaeger’s Ph.D. thesis “Now that you mention it, I wonder…”: Awareness, Attention, Assumption (http://www.illc.uva.nl/Research/Publications/Dissertations/DS-2009-10.text.pdf) might be of interest to you.

    Like

  2. CJ Avatar
    CJ

    Exploiting my freedom to weigh in before reading anything, I have a question. What do you think the status of a stimulus is between its first appearing in the agent and the agent’s deciding to pay attention to it or not? Doesn’t the agent already have to be paying some attention to it to make a decision about it? I can imagine how an agent can choose to focus more or less on something (perhaps turning towards or away from it, for instance) if it is already an object of attention. I can also imagine that an agent chooses to pay full attention to their current activity so that they will not be disturbed by any stimuli that happen to arrive. But I have trouble understanding how an agent can assess a specific stimulus and choose how much attention to pay it without thereby already having paid some attention to it.

    Like

  3. Gary Williams Avatar

    Hi Carolyn,
    Thanks for the post.
    What do you think of the distinction between “attention” and “orientation” in the context of the problem of captured attention? I’ve seen it used by psychologists to distinguish the loud bang type of attention from other, more voluntary types of selective attention. For example, take a cat that is lightly napping on a bed and next door there is a construction crew. A jack-hammer starts and the cats’ ears immediately orient towards the sound. As the jack-hammer continues the ear-rotation becomes less and less until it habituates totally.There is a sense in which this might be called “attention” but the physiological directness and bottom-up nature of the phenomenon gives credence to the notion that orientation is a distinct process from “top-down” selective attention.
    As for a counter-example to your thesis, how about this. You are a special agent on a train with a ticking time bomb on it. If you don’t defuse it, you and everyone on board will die. In order to defuse it you have to do really complicated math problems in your head. But installed in the train are strategically placed loudspeakers that randomly squawk loud, distracting sounds. Due to cognitive fatigue, these sounds keep distracting you from the mental calculation tasks and you keep having to start over.
    This seems like a counter-example to me because it doesn’t seem plausible that the subject would actively distract themselves – there is zero motivation or reason to do that given the stakes. Personally, I would also distinguish between the “conscious subject” and the “unconscious subject”.I want to allow for unconscious attention to account for things like the cocktail party effect. My view is that the mind is too fractured and multitudinous to be described in terms of a single monolithic subject attending to the world. In the words of Walt Whitman, “Do I contradict myself? Very well, then I contradict myself, I am large, I contain multitudes.”

    Like

  4. Carolyn Dicey Jennings Avatar

    Good question. I think this sort of evaluation would have to occur outside the domain of the subject, but through a range of criteria that are (at least historically) within the domain of the subject. The subject could decide that only a very salient stimulus (the roof caving in) should be allowed to interrupt (as when one is coding or playing a video game) or that any old stimulus (a passing bug) should be allowed to interrupt (as when one is bored). But you are right to point out the oddness of the fact that on this reading once a stimulus passes these criteria it is likely to capture attention, and not because of any evaluation by the subject. So the sense in which it is voluntary/active is very weak. Do you think that works?

    Like

  5. Carolyn Dicey Jennings Avatar

    Lots of great stuff here! I agree that the subject need not be conscious. And I love the Walt Whitman quote. I also agree that orientation need not count as attention. This is a point of disagreement between me and some others: I want to say that focus alone does not warrant the term “attention,” since it could be achieved through mere filtering, and filtering could be done by an inanimate object (whereas the term “attention” is only used for living things with minds). This sort of orienting could account for task-irrelevant perceptual learning, for example, which occurs outside of attention. On to the counterexample: perhaps we could say of cases like this that the subject is active insofar as the subject makes a difference, rather than the subject is active only if he or she has complete control. If the subject is able to make the loudspeakers even a little less distracting, that is enough to show minimal activity. Is that a satisfactory response?

    Like

  6. Charles R Avatar
    Charles R

    How does this work in the context of an actor or performer who seems to completely submerge into character—and then something unscripted and noticeably obtrusive occurs? It seems sometimes people snap out of character and laugh or cuss. Sometimes they keep doing the scene and act as though nothing has happened. Sometimes they respond in character as though the event happens within the world of the character, improvising. Is this kind of layering of attention and behaviors discussed in the literature?
    If attention concerns how the subject makes a difference in the capture, in the third case, is it the character’s subjectivity or the actor’s, or is there no clear way to distinguish the one from the other?

    Like

  7. Carolyn Dicey Jennings Avatar

    This is such a cool example. I have thought some about athletes and musicians, but not actors. I am not sure whether anyone discusses attention as it corresponds with acting, but I would love to look into this at some point. In my view, the other types of “immersions” require little to no attention, which is one of the reasons I think of them as correlating with a different form of conscious experience than standard perceptual experience. But there is a sense in which this immersed state is just a temporally extended form of normal experience–we set an intention and then let the intention guide our actions until we are interrupted, and this period of time can last for hours, rather than minutes. On the last point: I think the subjectivity in question is still the actor’s, but that the subjectivity is constrained by an adopted context. But it would be interesting to explore this more.

    Like

  8. Michael Barkasi Avatar
    Michael Barkasi

    Carolyn,
    I suspect that we might differ in some background assumptions which lead you to discount these two examples, but here are two. The first involves less-than-focal attentional capture''. What I mean is this. A loud bang mightcapture” attention, in the sense that (so it’s natural to say) it displaces for a moment whatever else was the focus of attention. But not all cases of attentional capture are like this. Using reaction time/accuracy as a proxy for attention, it seems that when attention is directed to one thing it it also spreads to other things at nearby locations and to other objects (even far away) that share basic features of the focal object (eg, color). (Here I have in mind Kravitz, Dwight J. and Behrmann, Marlene. 2011. Space-, object-, and feature-based attention interact to organize visual Scenes''. Attention, Perception, & Psychophysics.) So, if I focus my attention on a green ball in the center of my visual field, other green objects in view willgrab” attention (as evidenced by increased reaction time/accuracy to events near them), even if I barely notice them, have no intentions regarding them, etc. So, isn’t my attention being grabbed by peripheral like-colored objects an example of passive or involuntary attention?
    The second example is pretty controversial, but if the work on blindsighters and attention is right (and there really can be attention to things in the blind part of the field) then wouldn’t this have to be involuntary or passive attention? Certainly blindsighters, lacking awareness to what’s in their blind field, don’t form intentions or actively hold attention to things in it.

    Like

  9. Carolyn Dicey Jennings Avatar

    Hi Michael! These are important cases. A short answer is that I do not think that every form of (neural/cognitive/whatever) resource distribution is attention. I think that there are non-attentional resource distributions that can make a difference to reaction time, for example (e.g. the result of neural tuning). What matters here is how to tell when something is the result of mere resource distribution and when something is the result of attention. In my Synthese paper I offered an account of how to separate these, but that account is complex. I will think about this more and see if I have a better answer for you.

    Like

  10. Michael Barkasi Avatar
    Michael Barkasi

    Carolyn, it sounds like to need to look at the Synthese paper! I am especially interested in ways of separating attention from other similar phenomena. As a matter of needing some way to get a handle on it, I mostly just approach it ostensively (phenomena similar to paradigmatic cases of attention are attention) and with good enough stand-ins and proxies (decreased reaction time). Of course, these only provide defeasible criteria.

    Like

  11. Carolyn Dicey Jennings Avatar

    I think that is a good start. The problem is figuring out what type(s) of similarity matter(s). You might also like this: https://www.salk.edu/pdf/faculty/Reynolds_Normalization_Model_of_Attention.pdf .

    Like

  12. Felipe De Brigard Avatar
    Felipe De Brigard

    Hi Carolyn, Great stuff going on here! After reading your post above I couldn’t help but side with your commentator at the APA. But then I read it again, and I thought of something I’m not sure of: almost everyone agrees that dichotic listening tasks produce cocktail-like effects, even when one is focused on a distraction task. But after thinking about it, it occurred to me that I am not really willing to say that the cocktail-effect is mandatory in the strictest sense, that is, that under no circumstance of attentional focus can one fail to have his/her attention captured by the utterance of, say, one’s name. Albert Einstein was notorious for being undistractable (not sure that’s a word), which suggest that, at least under certain circumstances, and maybe for certain people, your Husserlian approach appears to apply just fine. Here’s a suggestion: attentional capture, of the so-called “early-selection” (a la Treisman), may be best modeled by a parabolic regressor, so that there is a certain point in which, with enough attention being paid to the main task, one can achieve successful suppression of unwanted stimuli that under lower demands would have captured one’s attention extrinsically. Maybe there is some research out there that speaks to this.

    Like

  13. Carolyn Dicey Jennings Avatar

    It seems clear to me that hearing our names need not distract us from our current task. (It certainly does not reliably distract my dog, Scout!) In visual attention, abrupt onset is the only feature that reliably captures attention against the intentions of the participant, according to work by Yantis and Jonides. But it makes good sense to me that even if participants are distracted by abrupt onset in the lab, they may not be distracted by abrupt onset in more ecological settings, such as when they take themselves to be in danger (e.g. when presented with a weapon). Tasks performed in the lab are not likely at the peak of participant interest and attention, even when they are reasonably difficult. As to how to best model this relationship, I am not sure.

    Like

Leave a reply to CJ Cancel reply